Fermanagh occurrence

In Fermanagh, this small, evergreen perennial is widespread but not abundant, mainly on strongly acidic, nutrient-poor, peaty, high ground on the Western Plateau plus in a few outlying sites in the Carnmore area. Local habitats include mountain summits and slopes, ledges, screes, open moorland and bogs. It has been recorded in a total of 44 tetrads (8.3% of those in the VC), but it has not been refound in six previous sites recorded in the 1940s and early 1970s. The lost sites are: Gadalough, N of Keenaghan Lough; Lough Scolban; Mullaghmore (the famous Erica vagans (Cornish Heath) site); Lough Doo near Little Dog; Brennan's Rocks, N of Lough Mulderg; and Pollnagollum. While these areas need to be searched to confirm the local extinction of this species, the evidence suggests that Fir Clubmoss has suffered habitat loss mirroring that observed in other parts of its British and Irish range.

The only area where H. selago occurs nowadays in any considerable quantity in Fermanagh is on the exposed summit ridge of Cuilcagh, the highest mountain where the rocks peek through the shallow, acid, nutrient-poor blanket peat. Throughout the year, on the blasted heath, H. selago pokes up its short, stiff, yellowish-green, bushy tufts between the stones on bare peaty ground or where surrounded by a carpet of grey-green Racomitrium moss and wind-pruned, flat-growing dwarf woody shrubs, for example, Erica cinerea (Bell Heather), Calluna vulgaris (Common Heather or Ling), Vaccinium myrtillus (Bilberry), V. vitis-idaea (Cowberry) and Empetrum nigrum (Crowberry).

Ecology

As the Fermanagh tetrad map shows, elsewhere in the VC, Fir Clubmoss is fairly frequent in open areas on scarps and peat bogs on the Western Plateau, and less so in open ground on lowland blanket bogs and heaths whenever competition is reduced. In the latter situation, H. selago usually occurs as individual plants occupying well-drained sites, often on acid, peaty soil close to large rocks, where competition from heather subshrubs and other species is reduced, probably by the shallowness of the substrate. Similar competitive conditions occur when Fir Clubmoss grows on upland cliff and rock ledges. It has also been suggested that the lowland sites of H. selago may lie in damp frost pockets, the cold soils of which would again restrict the growth of competitors (Jermy et al. 1978; Jermy & Camus 1991).

On the other hand, Page (1997) emphasises that all H. selago sites, whether on sandy or peaty soils, seem to be particularly free-draining ones, and that the species is tolerant of very exposed conditions, being able to survive both considerable winter cold and summer sun. However, the species does appear to be susceptible to heavy grazing pressure and populations are even more vulnerable to heathland fires, whether accidental or used to manage heathers, as the plants are readily eliminated by burning (Page 1997).

British occurrence

In Britain, H. selago has become very rare or extinct in many of its former lowland heathland sites, indeed in most of these it has not been seen for nigh on a hundred years (Page 1988). While the reasons for this major decline of the species in Britain are not fully understood, undoubtedly changes in the pattern of land use involving habitat loss, changed methods of vegetation management – perhaps involving heavier agricultural stocking levels, or using herbicides and fire to manage vegetation – plus increased levels of air and soil pollution have all been suggested as possibly significant contributory factors (Page 1988). While air pollution can be absolutely discounted in our area, the most obvious and notable examples of this sort of change in the Fermanagh context are the development of extensive hectares of forestry plantation on the Western Plateau and, until recently, the drainage, cutting, ploughing and fertilizer spraying of our lowland bogs.

Irish occurrence

Despite a possible or real decline, H. selago is almost certainly still the most widespread of the five clubmoss species that occur in Ireland, having been found at least once in all 40 vice counties (Scannell & Synnott 1987). The 1996 edition of Webb's An Irish Flora describes Fir Clubmoss as, "frequent but local, mostly above 300 m, but sometimes on lowland bogs" (Webb et al. 1996).

World distribution

The species is widespread in the northern circumpolar region and numerous varieties also extend worldwide and penetrate far into the southern hemisphere, eg to the Falkland Islands, Tierra del Fuego and Tasmania (Hultén 1962). Elsewhere in W Europe, H. selago has a widespread occurrence throughout western and northern parts of Britain and Ireland, Iceland, Scandinavia and western central European countries south to the Alps and the Pyrenees (Jalas & Suominen 1972).

Fossil record

Fossil spore evidence proves the presence of H. selago in Britain and Ireland, especially in northern parts, from the late phase of the last glaciation onwards (ie from the Late Weichselian onwards, throughout the entire Flandrian period) (Godwin 1975).

Current trends

Historical records show that Fir Clubmoss formerly occurred widespread throughout Britain and Ireland except in some counties in S, E and C England. Nowadays, while it is still common in Scotland, H. selago has lost much of its former ground in England and Wales. A count of the pre- and post-1930 symbols for England and Wales on the BSBI Atlas map indicates that of a total of 140 hectads plotted, 73 had pre-1930 records only for H. selago (Perring & Walters 1976; Godwin 1975). The 1978 Fern Atlas published a revised hectad map which incorporated considerably more records for England and Wales and of the 227 symbols, 127 were for post-1930 records of H. selago, although the overall pattern of losses from the S and E of England was obvious (Jermy et al. 1978). Changes recorded during the BSBI Monitoring Scheme in 1987-88 appeared to indicate a decline in England but an increase in Wales. The authors of the relevant report (Rich & Woodruff 1990), however, reckoned these changes were not significant.

In comparison, The New Atlas of the British and Irish Flora which reports survey data up until the end of 1999, maps H. selago from a total of 1223 hectads. A visual count of Irish hectads in the New Atlas found a total of 213 squares displayed, 61 of which had pre-1970 records only. The data behind the New Atlas map suggests that most Fir Clubmoss losses took place early on (pre-1930), due to habitat changes associated with agricultural intensification. While this process has continued, the overall distribution of H. selago appears stable in these islands (A.D. Headley, in: Preston et al. 2002).

Reproduction

H. selago plants often reproduce prolifically by vegetative means producing sizeable 'bulbils' or 'gemmae' which are budded off from near the tops of stems. In form these are small, leafy, 'trident-like' flattened buds formed in rings around the erect stems, about one cm below the large terminal bud (Page 1997). These abundant propagules are efficiently dispersed by wind in the autumn and they root and rapidly establish new daughter plantlets which grow between one and two cms tall in their first year (Page 1988).

However, Fir Clubmoss plants also produce vast quantities of pale yellow asexual spores, which again are efficiently wind-dispersed. As far as RSF can discover, nothing is known about the efficiency of Huperzia spore germination in the field, nor about recruitment of plants from this biological source (Page 1988, 1997). I think this matter could be accurately described as a field of near total scientific ignorance, which applies not only to clubmosses, but for most pteridophytes and bryophytes.

Even on the question of the longevity of individual plants and their population turnover, little or nothing is definite and everything is qualified; for example, "H. selago appears to be a relatively short-lived plant, which slowly builds up its tufts over several seasons and then probably reaches an abrupt and fairly rapid demise." (Page 1997). Clearly, if in the future Fir Clubmoss is to be actively conserved, whether under a Biodiversity Action Plan designation or not, closer study will be urgently required to clarify the reproductive capacity and population dynamics as well as identifying the significant environmental pressures affecting the species in typical habitats.

Names

The current author has not been able to discover as yet the derivation of the genus name 'Huperzia', which first came to my notice with the first volume of Flora Europaea (first edition), in 1964. The specific epithet 'selago' is a reuse of a previous generic name: it was first used by the Classical Roman Pliny to refer to a plant resembling Sabina herba, a synonym of Juniperus sabina (Gilbert-Carter 1964).

Threats

Afforestation of upland areas, overgrazing, the use of fire to manage hillside vegetation and the possibilities of consequent soil erosion are the main threats.

Growth form and preferred habitats

Unlike H. selago, individuals of Stag's-horn Clubmoss are regarded as being fairly long-lived plants which slowly develop sparse but wide-spreading colonies with long, horizontally running stems, sprawling through or over a mass of accumulated surface leaf and moss litter. The undulating stem, often vivid green in colour, was aptly described by Step and Jackson (1945) as having 'the combined stiffness and pliancy of copper wire'. These spreading stems occasionally branch and root themselves at intervals, anchoring the clone to the substrate. They also send up occasional erect leafy shoots terminating in characteristic, perfectly vertical, bare stalks, on which two or three long, slender cones are borne (Page 1997).

The typical habitat of L. clavatum is on N-facing, acidic mountain grasslands or heaths, subject to high rainfall, but where there is sufficient slope to allow drainage water to flush through the ground (Jermy et al. 1978; Page 1997).

Reproduction

L. clavatum does not produce any 'bulbils' or 'gemmae', but relies entirely on its spores for reproduction, a fact that has probably contributed to its decline to extreme rarity, at least in Ireland. Like H. selago, Stag's-horn Clubmoss is very liable to destruction by fire. Thus the widespread practice of maintaining heather vigour on mountain slopes by the establishment of a cyclical burning regime, has undoubted killed off many local populations, particularly in Scotland and in Ireland where this is a regular form of heath management.

Irish occurrence

While Page (1997) described Stag's-horn Clubmoss as the best known clubmoss in the British Isles, it is much more local in Ireland, indeed rare and declining for at least 60 years. Although it did not fit the criteria for inclusion in the vascular plant Irish Red Data book (Curtis & McGough 1988), L. clavatum features in the list of scarce and threatened vascular plants in the 'Biodiversity in Northern Ireland' discussion document (Brown et al. 1997). The scale of the decline of Stag's-horn Clubmoss in Ireland is demonstrated by the fact that Praeger (1934) listed it as occurring or having occurred in 27 Irish VCs, while Scannell and Synnott (1987) suggest it is still present in just 11 VCs (7 of which were northern and did not include Fermanagh).

The statistics available in the Fern Atlas are telling: of the 62 hectads mapped with records for L. clavatum in Ireland, only 19 have post-1930 symbols (Jermy et al. 1978). The New Atlas of Ferns (2005) confirms the decline, although the map indicates a wider distribution: there now are a total of 81 Irish hectads plotted, 58 of them with pre-1970 records only (Wardlaw & Leonard 2005). The new fern map plots 17 post-1986 hectads in Ireland, all but three in the northern half of the island.

Fermanagh occurrence

L. clavatum was not seen in Fermanagh during the 83 years between Colgan's original find on the slopes of Cuilcagh mountain in 1905 and its discovery by H.J. Northridge in 1988 on a peaty roadside bank on Doon Hill at an altitude of just 250 m. Regrettably, the small population at this new and most unusual lowland site was subsequently destroyed by the unwitting dumping of earth on top of it sometime prior to January 1995. Happily, in April 1995 a third station was discovered by Matthew Tickner on a knoll by a burn flowing through blanket bog on the Pettigo Plateau. Previous to this in January 1990 the species was rediscovered by RHN surviving on Cuilcagh on the slope below the summit cairn, which was very possibly Colgan's original site, bringing the number of extant Fermanagh stations back to two!

The typical habitat of L. clavatum is on N-facing, acidic mountain grasslands or heaths, subject to high rainfall, but where there is sufficient slope to allow drainage water to flush through the ground (Jermy et al. 1978; Page 1997). The Cuilcagh site in Fermanagh exactly fits the typical habitat of L. clavatum, ie N-facing, acidic mountain grassland and heath, subject to high rainfall, but on a slope where drainage water flushes the ground. Plants in the site sprawl in a loosely undulating manner over thin peaty soil, covering hidden boulders amid low-growing heather moorland composed mainly of Erica cinerea (Bell Heather), Calluna vulgaris (Common Heather or Ling), Vaccinium myrtillus (Bilberry) and Empetrum nigrum (Crowberry). Both Huperzia selago (Fir Clubmoss) and Diphasiastrum alpinum (Alpine Clubmoss) also grow near this particular site.

At both its Fermanagh sites L. clavatum is struggling to compete with Calluna. A visit to the Pettigo Plateau station in October 2010 located only two small patches of the clubmoss.

British occurrence

In Britain, L. clavatum is a mainly northern species, but it has more sites in S England than any of the three other clubmoss species with which it most nearly overlaps, Huperzia selago, Diphasiastrum alpinum (Alpine Clubmoss) and Selaginella selaginoides (Lesser Clubmoss)(Jermy et al. 1978).

In Britain and Ireland, and especially in more lowland sites, the species has decreased or disappeared due to the intensification of agriculture and utilisation of previously ignored rough marginal land (Page 1997).

European and world occurrence

Like Huperzia selago (Fir Clubmoss) L. clavatum is a widespread circumpolar species which also extends into the southern hemisphere and worldwide has several named varieties (Hultén 1962). In Europe it displays a mainly western distribution with a relatively continuous range from N Fennoscandia to the Alps and Pyrenees (Jalas & Suominen 1972; Page 1997). In Britain and Ireland it is a mainly northern species, but it has more sites in S England than any of the three other clubmoss species with which it most nearly overlaps, Huperzia selago, Diphasiastrum alpinum (Alpine Clubmoss) and Selaginella selaginoides (Lesser Clubmoss)(Jermy et al. 1978).

Uses

In former years L. clavatum was collected as a source of 'Lycopodium powder', the dry, light, bright yellow spores being used in school physics experiments to display sound waves. The spores were collected commercially chiefly in Russia, Germany and Switzerland in July and August, the cones being cut off and sieved to remove the Lycopodium powder or 'vegetable sulphur', or even 'vegetable brimstone', as it was sometimes referred to (Grieve 1931). The spores, like the rest of the plant, are very flammable, and in years gone by they were used in the manufacture of fireworks and for pyrotechnic stage lighting effects in theatres (Mabberley 1997).

In herbal medicine the spores were used alone (ie apart from the rest of the plant), from the seventeenth century onwards, being employed as 'a diuretic for dropsy, a drastic in diarrhoea, dysentery and suppression of urine, a nervine in spasms and hydrophobia, an aperient in gout and scurvy, a corroborant in rheumatism, and also as an application to wounds' (Grieve 1931). The use of Lycopodium powder was never admitted to the British Pharmacopoeia, but herbalists in the British Isles did use it as a dusting powder for treating skin diseases. The main pharmaceutical use was as a pill powder, to envelope pills and prevent them sticking together when boxed (Grieve 1931). In some rural parts of the British Isles there once was a folk-tradition of using garlands of Stag's-horn Clubmoss for 'personal adornment' in some form of ceremony (Page 1988 & 1997).

Names

The genus name 'Lycopodium' is derived from a combination of two Greek words 'lycos', meaning 'wolf', and 'podion', 'little foot', a translation of the German, 'Wolfsklauen', first used by the German physician and botanist, James Theodore Tabernaemontanus. He fancied that the clubmoss shoot resembled the paw of a wolf in miniature (Gilbert-Carter 1964; Hyam & Pankhurst 1995; Step & Jackson 1945). The specific epithet 'clavatum', is Latin meaning 'club-shaped', which like the general name of the group 'Club-moss', refers to the club-like shape of the fruiting cones (Stearn 1992; Step & Jackson 1945). The English common name 'Stag-horn Clubmoss' is a book name, and alternative local folk names do not appear to exist.

Threats

Overgrazing by sheep, shading overgrowth by Calluna, and fire.

Growth form and preferred habitats

This moss-like, low-growing, blue-green, arctic-alpine (or alpine-montane) clubmoss grows in short grass over shallow, well-drained, acid peat, often over rocks on or near mountain summits. The slender, wiry, spreading stems produce distinctive, short, erect, evergreen branches clothed with four tightly overlapping ranks of leaves, giving the foliage a rather cypress-like appearance. The shoots fork frequently and evenly to produce clusters of branches all of identical length which often develop in a fan-like, decumbent manner. Fertile branches terminate in a solitary stalkless cone 1-2 cm long, which turns pale yellow as the asexual spores mature and are released in late July and August (Step & Jackson 1945; Page 1997).

Fermanagh occurrence

In Fermanagh, there is just one quite large, sprawling patch growing amongst Sphagnum and Racomitrium moss and spilling down a rock outcrop on the NE face of Cuilcagh mountain, near the summit at an altitude of about 580 m. It was found during a survey of the Cuilcagh Plateau for the Royal Society for the Protection of Birds. A herbarium voucher was deposited in DBN. The site has been revisited several times, most recently in September 2010, when the plant was found in good condition.

Irish occurrence

Elsewhere in Ireland, Alpine Clubmoss is known only from a few scattered localities on the acidic high mountains of Ulster, W Mayo, W Galway, Offaly and Wicklow (Jermy et al. 1978). In Northern Ireland, the two main areas for the species remain the granite Mourne Mountains, Co Down (H38), where in the past it occurred on six different peaks, and the basaltic Garron Plateau, Co Antrim (H39). There are also isolated stations in the Sperrin mountains and on Slieve Gallion in Co Londonderry (H40), plus this recent Fermanagh discovery (Hackney et al. 1992; NI Vascular Plant Database 2002). The lower altitude limit for the NI sites is around 300 m, considerably less than the 457 m given for British mainland sites (Page 1997).

British occurrence

In Great Britain, Alpine Clubmoss has a pronounced northern and western distribution on the higher hills of Scotland, N England and Wales, with outliers in the Derbyshire Pennines, the Worcestershire Malvern hills and possibly also in Devon – although the latter records need confirmation (Jermy & Camus 1991). D. alpinum is characteristic of (and in Wales, at least, can become locally dominant on) well-drained, shallow, peaty slopes, on exposed and/or heavily grazed sites, where plant competition is reduced by these environmental pressures (Page 1997).

Recent changes

Like other clubmoss species, D. alpinum populations and its distribution have undoubtedly suffered a decline in the recent past, particularly at lower altitudes (Jermy et al. 1978; N Ireland Vascular Plant Database 2002). The scale of past losses in Ireland is clearly seen from hectad statistics in the New Fern Atlas (2005), which show the species mapped in a total of 49 squares, only 25 of them with post-1970 records (Wardlaw & Leonard 2005).

Land-use changes, including the extension of coniferous plantation on ever higher ground, and increased livestock stocking densities on upland moorland and heaths has resulted in losses of D. alpinum in both N England and N Ireland, and very probably also in other regions of Britain & Ireland. Like all clubmosses, D. alpinum is very vulnerable to fire. In addition, sulphur dioxide air pollution is undoubtedly responsible for the disappearance of the plant from the hills overlooking Belfast in the late 19th century. Despite the advent of clean air legislation and its enforcement in recent years, this factor continues to operate and it certainly must be responsible for some of the losses recorded in sites downwind of major cities in Britain (Jermy et al. 1978; Page 1997).

European and world occurrence

Beyond Britain & Ireland, D. alpinum has a classic disjunct arctic-alpine distribution in Europe (Jalas & Suominen 1972, Map 12), and it would be circumpolar were it not for a rather unexpected absence from much of the arctic region of central Canada, creating a large gap in the distribution (Hultén & Fries 1986, Map 6).

Uses

In past times, D. alpinum was collected and used as a source of dye mordant in place of the usual chemical fixative, alum. Experiment has shown that plant dyes fixed with mordants from clubmoss species produce softer, more permanent colours than those achievable with alum (Page 1988).

Names

I cannot find a derivation for the genus name 'Diphasiastrum', except that it must in some way connect with a synonym of Lycopodium, 'Diphasium'. The word element 'diphasia' suggests something that exists in, or exhibits 'two stages'.

Threats

Increased sheep density on upland moorland and heaths has resulted in losses of D. alpinum in both N England and N Ireland. Air pollution is an increasing threat in some areas. The solitary nature of the Fermanagh plant renders it vulnerable, but the site is rather difficult of access, reducing the likelihood of it being grazed.

Growth form and preferred habitats

A tiny, delicate, moss-like perennial, Lesser Clubmoss is both montane - found above the notional tree-line in areas dominated by blanket-bog, and in heathland, and also lowland, growing in wet flushed, base-rich ground beside lakes and streams. Being very small and rather inconspicuous it needs to be positively searched for in seepage areas, stream sides and near lakeshores, or in open, exposed habitats. It often occurs in shallow soils where the growth of competing species is limited.

The perennial stems of Lesser Clubmoss are generally extremely inconspicuous, weak-looking, prostrate, branching only occasionally, and they trail over and through mossy vegetation and the stem bases of other, more vigorous vascular plants. Slightly easier to spot are the annual cone shoots, of which one or two are produced per stem during July and August.

Pinguicula vulgaris (Common Butterwort) is almost invariably a good indicator of the likely presence of S. selaginoides and it is interesting to compare the distribution maps of the two species. In late August and early September the pinkish-yellow colour of the senescing erect annual cone shoot of Lesser Clubmoss makes it much easier to observe (provided, that is, if one bends right over and looks closely and diligently in the right sort of habitat!). The small size of the plant and the fact that it has a soft, moss-like texture immediately distinguish S. selaginoides from all other species of clubmoss in Britain and Ireland.

Reproduction

The terminal spore-bearing shoots are shaped like an Indian club, or a fox's tail and are held erect, typically between 2 and 10 cm tall. The actual cone is leafy and ill-defined, bearing both sterile leaves and fertile spore-bearing ones (ie sporophylls), carrying separate spore sacs (ie sporangia). The latter structures contain asexual spores of one of two kinds on separate parts of the fertile branch: either four white relatively large female megaspores per megasporangium, or vast numbers of microscopic yellow male microspores per microsporangium (Jermy & Camus 1991; Page 1997). Sexual reproduction follows on a microscopic prothallus produced by each (or some) of the megaspore(s).

Fermanagh occurrence

This tiny, delicate, moss-like perennial has been recorded in 56 Fermanagh tetrads (10.6% of those in the VC), 46 of them with post-1975 records. It is occasionally found on the shores of Lower Lough Erne, but is quite frequent and widespread on the upland limestones of the Western Plateau. The details of the most outlying stations to the S of the county are: Knockninny, 1900, W.N. Tetley; Kilroosky Lough ASSI, 1980, R.S. Weyl; and moorland at Skeaghoge Td, 1989, RHN.

The nature of base-rich habitats

The one characteristic that links all S. selaginoides habitats is the requirement for soil enrichment and aeration associated with the movement of base-rich water. The base (ie the positively charged cation) most typical of such waters is calcium, but in the Lough Navar area of Fermanagh the parent rock has become partially dolmitized and thus base-rich water here contains both calcium and magnesium (Whitten & Brooks 1972).

In other sites, base-yielding rocks may include various mica-schists, volcanic tuffs, lavas, or basalts, some of which may release minerals in quantities that are toxic to certain plants. Base-rich water is not necessarily nutrient-rich as far as plant growth is concerned, and if it is derived from dissolution of limestone or dolomite it is always nutritionally unbalanced, being oversupplied with Ca⁺⁺ ions. As ground-water percolates through soil, however, it can accumulate and transport dissolved nitrogen and available phosphate, plus traces of other elements essential for plant development, which might otherwise be scarce or absent in a particular site and thus become limiting for plant growth.

The best concise account of the concept of base-rich soil and plant nutrient status known to the current author appears in Page's book, Ferns. Their habitats in the British and Irish landscape (Page 1988, pp. 70-3 and 311-2), and this is highly recommended reading for anyone who is puzzled by the usage of this technical term. Page clarifies the essential qualifications associated with understanding and applying this rather difficult and potentially (and actually), very woolly ecological concept.

Fossil record

Micro-fossil megaspores of S. selaginoides have been found at Derryvree, near Maguiresbridge in Fermanagh in a full-glacial freshwater deposit of Middle Midlandian age radio-carbon dated to 30,500 BP (Colhoun et al. 1972). The flora and fauna of this fossil deposit indicated open tundra vegetation and a periglacial climate prevailed at the time it was laid down.

Elsewhere in Britain & Ireland, the fossil record for S. selaginoides is well studied, both microspores and megaspores being readily recognised. The sediment studies prove the species has been present in these islands during the last four glacial stages. Lesser Clubmoss has been less frequently recorded during some of the intervening warm interglacial periods. This is not terribly surprising, since it cannot cope with tall, shading vegetation typical of the forest maximum. However, the species has been found as fossils throughout the entire current heavily studied interglacial, which is called the Flandrian in Britain and the Littletonian in Ireland.

British and Irish occurrence

The fossil record also shows that the plant has contracted in range northwards compared to its late glacial distribution (Godwin 1975). Although we do not understand what factor(s) caused this range contraction, it is particularly well demonstrated in the British Isles, where the species until recent historical times occurred north of a line in Britain from Barmouth in Wales to Skegness in Lincolnshire, and in Ireland from Foynes near Limerick, eastwards to Arklow. There is just one exceptional site below this demarcation, which is on the coast near Wexford town (Jermy et al. 1978).

The distribution of S. selaginoides has been seriously affected by drainage and the intensification of agriculture in the last 50 to 70 years, with significant losses in the SE of England before 1930. Approximately 70 additional sites were lost between 1950 and 1990. Over the same period, similar environmental pressures in Ireland have resulted in the loss of around 35 sites of the species throughout the island, but again losses have been particularly concentrated in the south of its range (Jermy et al. 1978; Jermy & Camus 1991).

European and world occurrence

Beyond Britain and Ireland, S. selaginoides has an amphi-atlantic or circumpolar, boreal-montane distribution (Hultén 1958; Hultén & Fries 1986, Map 8; Preston & Hill 1997). In the view of the current author, the absence of records in Siberia and major gaps in Asia in general, does not warrant describing the distribution as circumpolar. The disjunctions are simply too wide. Furthermore, the published European distribution is very definitely disjunct and appears to fit the arctic-alpine (or arctic-montane) pattern better than the more continuous boreal-montane picture (Jalas & Suominen 1972, Map 13). The distribution of S. selaginoides extends quite far south in Finland, the Baltic region and Denmark. Otherwise its distribution closely resembles that of Diphasiastrum alpinum (Alpine Clubmoss) (Jalas & Suominen 1972, compare Maps 12 & 13). Preston & Hill (1997) regard the latter as circumpolar arctic-montane. Most unexpectedly, Page (1971) has discovered a single extremely disjunct outlying station of S. selaginoides lying 2100 km south of its nearest European mainland station, on the Canary Island of Hierro.

Names

The genus name 'Selaginella' is the diminutive of 'Selago', an ancient name applied by the classical Roman, Pliny to a plant resembling Sabina herba, an old name of Juniperus sabina (Gilbert-Carter 1964). The plant Pliny was referring to was a clubmoss, the whole group being then named 'Selago', including the genus we know as 'Lycopodium' together with subsequent splits from it (Johnson & Smith 1946). The specific epithet 'selaginoides' is Latin meaning, 'Selago-like' or 'clubmoss-like', probably meaning, 'like Lycopodium selago', an earlier name for the current species (Gledhill 1985; Stearn 1992). The English common name 'Lesser Clubmoss' is a typical book name, probably of Victorian origin.

Threats

Afforestation of the species' upland habitat, or improvement of rough or damp ground for agricultural purposes.

ISOETACEAE – Quillwort family

Growth form and preferred habitats

In Fermanagh, this small submerged perennial can sometimes be clearly seen growing in sheltered shallow water on the gravelly bottoms of acidic, nutrient-starved, unproductive lakes, as at Lough Nabrickboy in Big Dog Forest. More often it occurs in deeper, medium-sized, brown-water lakes on the Western Plateau, or in more mesotrophic, ie moderately productive, lowland water bodies. In both these situations, it can form a lawn-like turf and become the dominant bottom-growing plant. The presence of the species is often only betrayed when its stiff, evergreen 'quills' are washed up in the plant debris along the shoreline after stormy weather, sometimes in considerable quantity.

Like the vegetatively similar Lobelia dortmanna (Water Lobelia), I. lacustris is not a competitive species and in nutrient-rich waters it is easily overgrown by algae or by faster-growing aquatic macrophytes or both. Probably for this reason, Quillwort tends to occupy rather deeper water, typically from 0.5 to 2 m, but it can survive down to 6 m deep in order to avoid competition from more light-demanding vascular plants (Page 1997; Jonsell et al. 2000). I. lacustris appears to avoid small lakes and silt- or peat-bottomed mountain tarns, habitats more ecologically suited to I. echinospora (Spring Quillwort).

The four air chambers which traverse the length of each dark green tubular leaf are easily seen when it is sectioned at right angles, and they serve to distinguish the plant from either Littorella uniflora (Shoreweed) or Lobelia dortmanna, two other species with similar stiff 'isoetid' style of leaves, with which it frequently occurs in stony or sandy lake shallows. When Lobelia dortmanna flowers, it becomes a much more conspicuous plant than either of the other two species mentioned, and since it is so similar to I. lacustris in its ecological requirements and tolerances, it is a very good indicator of the likely presence of Quillwort at a site.

Another way in which I. lacustris is very readily distinguished from Littorella uniflora is by its brown (not white) roots (Jermy & Camus 1991). The four elongate air chambers in Isoetes leaves are supported and divided by numerous crosswalls (ie septae). The function of the air chambers within the photosynthetic green leaves is associated with the unusual method of carbon metabolism of these aquatic species, which is called 'Crassulacean Acid Metabolism (CAM)'. This physiology is most frequently found in desert succulent plants such as Yucca brevifolia (Joshua tree) (Crawford 1989, pp. 140-2). The extensive roots of I. lacustris absorb carbon dioxide gas from the mud sediment and store it along with carbon dioxide from night-time respiration as malic acid. The plant can thus recycle and store inorganic carbon, a scarce essential element in this habitat (and see also the species account of Lobelia dortmanna) (Farmer & Spence 1985; Boston 1986; Preston & Croft 1997).

Fermanagh occurrence

Considering the large number of lakes in Fermanagh, many of them upland, it is not surprising that this VC is the N Ireland headquarters of this aquatic species in terms of frequency, having been recorded 35 of tetrads, 6.6% of those in the VC. The two lakes in Fermanagh where it has not been recorded since the 1940s are Lough Scolban in the west of the VC and Lough Skale further east.

Irish occurrence

Compared with Co Fermanagh, the low frequency of Quillwort in Cos Down and Antrim (H38 and H39), and its very slight representation in both Cos Armagh and Londonderry (H37 and H40) is rather unexpected. In Ireland overall, I. lacustris has been recorded at least once in the past in 19 of the 40 VCs (Scannell & Synnott 1987).

The New Atlas hectad map shows that Quillwort is predominantly distributed in the N and W of both Britain and Ireland, although in the latter it is also present in eastern sites, for instance in both the Mourne mountains, Co Down (H38) and the Wicklow mountains south of Dublin (H20). There are also a few isolated sites close to the SE coast of Ireland in Co Waterford and South Tipperary (H6 and H7).

British occurrence

The distribution in Britain is very decidedly Scottish, Cumbrian and Cambrian (ie Welsh), but it does have a very few southern outliers in S Devon (VC 3). The solitary early 19th century E England occurrence at Prestwick Carr in S Northumberland (VC 67) is long extinct (Swan 1993; Preston & Croft 1997).

European and world occurrence

Quillwort has essentially a sub-Atlantic distribution in NW Europe, centred on Scandinavia and the British Isles. However, there are isolated, widely scattered outliers as far afield as Iceland, the Pyrenees and the Urals (Jalas & Suominen 1972, Map 18). Beyond Europe, I. lacustris is also found in S Greenland and in NE North America (Hultén 1958, Map 247; Hultén & Fries 1986, Map 9; Jonsell et al. 2000).

Variation

Hultén (1958) and Hultén & Fries (1986) both map the North American form of this plant as 'I. lacustris var. macrospora' Dur., and it has also at times been elevated to species rank and gone through several synonyms which these authors list. However, many botanists reckon this variety or species is best submerged back into I. lacustris (Preston & Croft 1997). A number of other varieties with one or more unusual characters have been described, including an Irish form, var. morei Syme, from a lake near Bray, Co Wicklow, which has very long leaves (Brunker 1950). An interesting viviparous form, with vegetative buds instead of sporangia, has been recorded from Lake Windermere (Page 1997). These variants, however, are now simply regarded as the evolving products of long geographic isolation and the consequent inbreeding of an ancient species.

Herbivory and dispersal

Upland game birds such as grouse are reported to feed on Isoetes species in North America (Fassett 1957) and their relatives may also do so in Britain and Ireland, although the current author (RSF) cannot locate any mention of this herbivory in the literature. It is considered feasible that the distribution of the species may reflect the North Atlantic migration pattern of wildfowl, most probably that of geese, which might carrying the spores, or much less likely, transport vegetative parts of the plant (Page 1997). The possibility of avian transfer is supported by the existence of the endemic form I. azorica, since it is hardly possible to imagine any mechanism of transport apart from water birds visiting the remote island group of the Azores on their regular migration route (Ridley 1930).

Names

The genus name 'Isoetes' appears to have been coined by the ancient Roman scientist Pliny, combining two Greek words 'isos', meaning 'equal' and 'etos', meaning 'a year', ie 'equalling one year'. This refers to the idea that the plant did not change with the seasons, meaning that it was evergreen. However, it is also thought that Pliny originally applied the name not to this species, but rather to a member of the Crassulaceae (Gilbert-Carter 1964; Johnson & Smith 1946). The specific epithet 'lacustris' is Latin meaning 'associated with lakes' (Gilbert-Carter 1964). The English common name, 'Quill-wort' or 'Quillwort' is an 18th century name, given to this inconspicuous and not well recognised plant from its supposed resemblance to a bunch of quills (Prior 1879; Grigson 1974).

An alternative name, 'Merlin's Grass', is a translation of a Welsh name 'Gwair Merllyns', where 'gwair' means 'hay' and Merllyn was the name of a Welsh prophet. The Welsh name of the plant appears in a manuscript account of Samuel Brewer's botanical journey through Wales in 1726, which is preserved in the British Museum, and it is quoted as follows in Britten & Holland (1886), "At Llyn Ogwen (Carnarvonshire) I saw the horses very greedily eating of that which was cast upon the shore and that on the water; and the people tel [original spelling] me that they wait there every day for it, and leave good grass growing near it; and that it improves cattle better than any grass; and that the fish like it as well. The fish are larger there than any of the other lakes, which they attribute to the eating of [this plant], which they call Gwair Merllyns.'

Threats

Eutrophication (ie cultural nutrient enrichment) of lowland sites and silting of upland ones, the latter at least generally attributed to forestry operations.

Fermanagh occurrence

Just two Fermanagh records exist from peaty or muddy lake bottoms, at Castle Caldwell, listed above and Bunnahone Lough in 1947, made by Meikle and his co-workers. The Revised Typescript Flora noted that, "These identifications need checking, as do all Irish Isoetes records." (Meikle et al. 1975). In Meikle's Fermanagh Flora card index, the Bunnahone plant was originally recorded as I. lacustris but was later reassigned. As far as we are aware, no vouchers exist for I. echinospora from Fermanagh (Osborne & Doyle 1992).

Nowadays, both sections of Lough Erne are eutrophic to hypertrophic and thus have become too nutrient-rich to support either species of Isoetes. However, in 2006 and 2007, palaeoecological studies associated with water quality assessment took sediment cores from several N Ireland lakes including the Trannish region of Upper Lough Erne and Meenatully Lough on the Pettigo Plateau blanket bog. These very different waterbodies contained fossil megaspores of both I. echinospora and I. lacustris. In the case of Meenatully Lough, macrofossils of both species were present in the lower portion of the upper 0-7 cm zone of the core, dated post-1970 (Davidson et al. 2008). The N Ireland Lakes Survey (1988-90) recorded only I. lacustris in this lake.

Irish occurrence and status

Although it did not feature or deserve a mention in the Irish Red Data Book. 1. Vascular plants (Curtis & McGough 1988), I. echinospora is now classified as a scarce species by conservationists in the Republic of Ireland. In our view, it is better described as a rare and possibly declining species. The known Irish stations are extremely thinly scattered down the W coast from Co Donegal to Co Kerry (Preston & Croft 1997). There are voucher specimens in DBN for six of the 40 Irish VCs, but only those of Co Clare (H9) are modern, all the rest being pre-1911.

Only ten of the total of 41 Irish records held at the Biological Records Centre, Wallingford, Oxfordshire are post-1950. They represent occurrences in five Irish VCs, S Kerry, N Kerry, Co Clare, W Galway and W Mayo (H1, H2, H9, H16 and H27). [All but the two records from S Kerry (H1) are from sites below 100 m in altitude (Osborne & Doyle 1992). I. echinospora is probably slightly more frequent in Co Clare (H9) and W Galway (H16), since these two VCs share six of the ten post-1950 records between them (Webb & Scannell 1983; Osborne & Doyle 1992).]

Identification difficulties

It is very difficult to distinguish I. echinospora from I. lacustris (Quillwort) in the field and they can co-exist and hybridise. The fact that specimens require microscopic confirmation deters the more casual recorder, so that I. echinospora may be overlooked or mistaken by field botanists for the very much more common species (Jermy & Camus 1991; Page 1997). Very sensibly in their Flora of Connemara and the Burren, Webb & Scannell (1983) were chary of accepting any reports of Spring Quillwort that had not been verified by microscopic examination of the microspores, a procedure which Osborne & Doyle (1992) also regard as absolutely essential.

In addition to the absence of vouchers for the 1940s Fermanagh records, the systematic survey of our lakes made in recent years has failed to find living specimens of I. echinospora. We therefore believe that either Meikle and his co-workers misidentified their specimens, or subsequent field workers (including ourselves) have not looked carefully enough at Isoetes material, especially that occurring in more oligotrophic waters.

Names

The genus name 'Isoetes' appears to have been coined by the ancient Roman scientist Pliny, combining two Greek words 'isos', meaning 'equal' and 'etos', meaning 'a year', ie 'equalling one year'. This refers to the idea that the plant did not change with the seasons, meaning that it was evergreen. However, it is also thought that Pliny originally applied the name not to this species, but rather to a member of the Crassulaceae (Gilbert-Carter 1964; Johnson & Smith 1946). The Latin specific epithet 'echinospora' means 'spiny spored', which is for once, the major defining character of the species.

Threats

None.

Growth form and preferred habitats

This rare, slow-growing, rhizomatous, evergreen horsetail species, with its distinctive rough texture, is regularly found growing on shady, sloping river banks and streamsides, which represent its predominant habitat throughout Britain & Ireland. It typically grows in heavy, permanently moist, sandy or clayey soils that are rich in silica and other minerals. It can also be found in base-rich moorland flushes, and elsewhere in similar flushes on sand dunes (C. Dixon & T.D. Dines In: Preston et al. 2002).

Identification

The erect, unbranched, dark blue-green stems are pencil-thick and the ash-white, toothless sheaths with black bands around the top and bottom make E. hyemale reasonably easy to distinguish from its hybrid with E. variegatum, E. × trachyodon. When the leaf sheaths are young they do appear to bear short teeth. In reality, these are minute scallops where the true sheath teeth would normally be attached (Page 1997, p. 450), but they are not observable on mature sheaths (Rose 1989). In the current author's experience, E. hyemale and E. × trachyodon only rarely associate with one another.

Fermanagh occurrence

This distinctive horsetail has been recorded in Fermanagh from 14 thinly scattered tetrads, 2.7% of those in the VC. Twelve tetrads have post-1975 records in habitats ranging from moist woods and shaded river banks, to peat covered limestone in the uplands. In addition to Smith's first record listed above, there is another early site at Cloncarn near Magheraveely, where it was recorded by Meikle and co-workers in 1948.

Although it typically grows in permanently moist, sandy or clayey soils, surprisingly it has never been found on any of the many lakeshores in the VC, although further north in Scandinavia it does occur in such situations, plus in a wide range of other very different, often much drier habitats which it never occupies in Britain & Ireland (Jonsell et al. 2000, p. 24).

In Fermanagh, like several other horsetail species, eg E. palustre (Marsh Horsetail) and E. telmateia (Great Horsetail), E. hyemale appears to require some lateral water movement at its roots, either a slow seepage or a flushing of moderately base- or mineral-enriched spring water (Rose 1989; Brewis et al. 1996).

An unusual variant

In terms of habitat, E. hyemale is not a very variable species but one exceptional site occurs on the limestone plateau at Legacurragh above Florencecourt. Here a solitary plant of a completely prostrate form of E. hyemale grows on thin blanket bog peat developed directly over limestone pavement. According to Clive Jermy (pers. comm. 1995), this unusual prostrate form is known from Scottish dune systems and, as here, grows in flushed, shallow peat over limestone.

Reproduction

Fertile stems emerge along with sterile ones in May-June and are similar in appearance, except that they produce a small, black cone which bears a short sharp tip (ie an apiculus). Spores are not produced until early spring of the second year and, in common with all other Equisetum species in Britain & Ireland, reproduction and spread of E. hyemale is mainly (but not exclusively), vegetative, involving lateral growth of the rhizome and water dispersal of stem fragments (Praeger 1934).

Growth rate, silica content and rough stem texture

E. hyemale is said to grow and spread very slowly, even when well established (Page 1997, p. 451), yet at the site on Manyburns River in Fermanagh, and in similar places, the plant grows in abundance in thick clumps. In these situations it locally dominates the riverbank vegetation, presumably due to its tenacious rhizome and the longevity of the species.

The observed slow growth of this species is probably due to its very high silica requirement, which in turn is associated with the colourless siliceous tubercles and other physical structural features which give the plant its characteristic tough, evergreen stems their very rough, abrasive texture.

Uses and English common names

The English common names, 'Rough Horsetail' and 'Dutch Rush', both allude to the fact that from early days, at least from the 17th century onwards, the plant was greatly valued as a scourer, the equivalent of our present day wire-wool (Grieve 1931). E. hyemale stems were collected locally and sold in markets for scouring cooking pots and were also used by artists for fine polishing metal, wood and bone articles (Step & Jackson 1945) and thus the range of English common names 'Pewterwort', 'Shave-grass', 'Scouring-rush', 'Scrubby-grass' and 'Dishwashings' (Prior 1879; Britten & Holland 1886).

Rough Horsetail is mentioned by Gerard (1633) as being used by fletchers (arrow makers) and comb-makers to polish their finished articles. Other more frequent and abundant Equisetum species were also used for these purposes, eg E. arvense (Field Horsetail) and E. palustre and, undoubtedly, some of these common names (apart from 'Dutch Rush') would also have been locally applied to them as well. Grigson (1974) and Mabey (1996) both report that bundles of E. hyemale are still sold as scourers in continental European markets.

Being a slow growing, rather scarce species, commercial collecting in B & I must have very quickly reduced local populations of this horsetail, so that imports from or through the Netherlands became necessary to meet the commercial demand for scourers and hence the name 'Dutch Rush'.

Irish occurrence

Long after the commercial use of the species as a scourer ceased, E. hyemale remains a rare, local and apparently declining species in the whole of the British Isles. In Fermanagh, it occurs in just 12 scattered post-1975 tetrads. Elsewhere in N Ireland, E. hyemale is rather rare in Cos Antrim, Tyrone and Londonderry (H39, H36 and H40), and very rare in Cos Armagh and Down (H37 and H38) (NI Vascular Plant Database 2001). In the Republic of Ireland, Rough Horsetail (often referred to as 'Dutch Rush') is very rare and scattered, and is apparently declining here also.

British occurrence

In Britain, the species overall has a decidedly northern distribution and, while scarce and local even in the northern half of the island, it is very much more rare and obviously declining south of the Mersey-Humber line (Jermy et al. 1978; Page 1997).

European and world occurrence

The European distribution of E. hyemale is quite similar to that of E. sylvaticum (Wood Horsetail), being essentially northern and boreal and stretching from SE Greenland (where it was first found as recently as 1981), through to Iceland, the Faroes (but not the Arctic Isles), and throughout all of Scandinavia. It also extends south to Gibraltar (although only very thinly represented across the Iberian peninsula) and thence eastwards along the northern shores of the Mediterranean to Greece and N Turkey (Jalas & Suominen 1972, Map 30; Daniels & Van Herk 1984). The distribution then continues east through the Himalaya and much of N Asia to Japan and Central America (ie Mexico and Guatemala) (Hultén 1962, Map 174; Hultén & Fries 1986, Map 11; Jonsell et al. 2000).

Is it circumpolar?: Although E. hyemale is classified by Preston & Hill (1997) without qualification as Circumpolar Boreo-temperate, the species only qualifies as circumpolar if we consider the taxon in its very broadest sense. In Europe and in W & C Asia, the E. hyemale we know in Britain and Ireland is a moderately variable species, but in E Asia and especially in N America, it becomes a complex of several forms which, while their taxonomy is incompletely worked out and is a subject of disagreement, have been grouped by American taxonomists into two species, E. hyemale and E. laevigatum and their hybrid (E. × ferrissii). N American E. hyemale is then further subdivided into three varieties and three forma, none of which is identical with our Eurasian plant (Hultén 1962; Scoggan 1978, p. 130). Thus there exists an enormous void in the circumpolar occurrence of the Eurasian form of E. hyemale (ie our E. hyemale), throughout N America. In terms of the plant's plant geography, its presence further south in C America does absolutely nothing to fill this northern, Boreo-temperate gap.

Names

The genus name 'Equisetum' was coined by the ancient Roman writer, Pliny and is thought to have been first applied by him to E. arvense. It is a combination of two Latin words, 'equus', a horse and 'saetum', a bristle or hair, and it is thought to refer to the bristly appearance of the jointed stems with their whorled branches (Gilbert-Carter 1964; Grieve 1931). The same notion also gave origin to the English common name 'Horsetail', which is a direct translation of the medieval Latin name, 'cauda equina', under which it was sold in apothecary shops (Prior 1879; Grigson 1974).

The Latin specific epithet 'hyemale' or 'hiemale', means 'of winter', that is, 'reproducing in winter' (Gilbert-Carter 1964).

Threats

Clearance of wooded stream banks, or excessive trampling or grazing of the sites by cattle pose the two major threats. Drainage might also be significant in other areas of the British Isles.

Growth form and preferred habitats

Plants of this evergreen, rhizomatous hybrid are more robust and more branched than those of E. variegetum, sometimes stretching up to 75 cm tall. This vigour and the fact that the long black teeth on the nodal sheaths are usually very persistent together help to distinguish the hybrid from both its parent species.

The hybrid grows on rocky lakeshores and wooded riverbanks and in Fermanagh is twice as frequent as one of its parents, Equisetum hyemale (Rough Horsetail). It usually occurs some distance from the main concentration of sites of the other parent, E. variegatum (Variegated Horsetail), which seems to require a more calcareous or more base-rich habitat than does the hybrid. E. × trachyodon does occur near E. hyemale at one spot on the upper reaches of the Colebrooke River, and with E. variegatum at Shean Jetty and Magho Jetty along the S shore of Lower Lough Erne, but these are the only exceptions. In the remaining twelve stations, E. × trachyodon avoids both its parents – or perhaps on account of its vegetative vigour, they avoid it!

The Fermanagh plants of this rhizomatous hybrid can form quite large, dense stands, as happens for instance on the Bannagh River and at one spot on Upper Lough Macnean, but it can also occur as just a few straggling, branched stems as on the shore of Lough Lattone. Alternatively, it may be scattered in clumps, as it is along a km or so of the bank of the Colebrooke River below Littlemount Bridge.

In general, the known occurrence of all forms of hybrid horsetails are far from evenly distributed throughout the overlapping portions of their parent species ranges in Europe (Jalas & Suominen 1972). Rather they tend to be thinly scattered, but with distinct local concentrations in certain areas. This is particular the case in western and northern regions of Britain & Ireland (Page & Barker 1985; Page 1997).

Equisetum hybrids in Britain & Ireland

All eight native Equisetum species in the British Isles are involved in producing hybrids, but they are formed strictly between pairs of species within the same subgenus. Six of the British & Irish horsetail species belong to subgenus Equisetum, and two (E. hyemale and E. variegatum) to subgenus Hippochaete. Perhaps surprisingly, only two hybrids from subgenus Equisetum have been found so far in Fermanagh.

Fermanagh occurrence of E. × trachyodon

In the case of the solitary subgenus Hippochaete hybrid, E. × trachyodon, N Ireland undoubtedly has the greatest concentration of known stations for this hybrid possibly anywhere. Within the six-county province, Fermanagh with its 13 main sites and their sometimes many sub-sites has the greatest representation of this hybrid. The Fermanagh tetrad map plots records from 22 post-1975 tetrads, plus two tetrads with older records. The details of the Fermanagh sites where E. x trachyodon has not been refound are: Bunnahone Lough, 1904, Praeger; and Lough Vearty, 1949, MCM & D.

Other Irish occurrences

The Fermanagh coverage is followed by Cos Down (H38), Antrim (H39) and Londonderry (H40), with Tyrone (H36) and Co Armagh (H37) trailing with just one or two sites each (NI Vascular Plant Database 2001). In the Republic of Ireland, Co Cavan (H30) also has one old 1950s record from Gowland, yet so far neither parent species has ever been recorded there (Reilly 2001). Elsewhere there are a further 12 tetrads scattered, mainly in the west, across 8 Irish VCs from Monaghan (H32) and Sligo (H28) to Mid Cork (H4), plus two more inland VCs, N Tipperary (H10) and Kildare (H19) (New Atlas).

Vegetative reproduction and spread

Some of the Fermanagh sites are obviously linked, for instance those along riverbanks, or along lake shores and they may, or may not, represent fragmented clones generated by secondary vegetative spread. Page (1997) reported that small fragments of hybrid shoots root very readily, even after they have floated around for up to ten days, so this vigorous hybrid definitely has a mechanism for increase and local dispersal.

In Fermanagh, we see evidence of local vegetative spread along the Colebrooke River, along the N shore of Upper Lough Macnean and the S shore of Lower Lough Erne. However, other stations are sufficiently remote to certainly represent independent parental hybridisations, and these cases form the majority in Fermanagh.

British occurrence

In comparison with NW Ireland, E. × trachyodon is very poorly represented in Britain, there being only three or four sites in England and six or seven in Scotland (New Atlas).

A strong westerly trend in hybrid horsetail occurrence

All but one of the British stations are in the extreme west and when one examines the distribution of all horsetail hybrids in Britain & Ireland, a westerly trend in their occurrence becomes very obvious. Indeed, if we 'zoom out' to view the whole European picture of hybrid horsetails, the westerly trend appears to be mirrored even at this much larger scale (Page & Barker 1985; Page 1997).

The explanation for this strongly marked trend in distribution presumably lies in the prevailing oceanic climate of the most westerly parts of Britain & Ireland. This is readily summarised as cloudy skies, high rainfall levels that are evenly dispersed throughout the year (ie over 200 wet days) and generally low temperatures with no extremes (ie mild winters and cool summers) (Page & Barker 1985; Page 1997, Maps 6-16; Porley 2001). The described climatic conditions appear to allow horsetail sporophytes of differing species to grow in close proximity to one another. The damp, mild environmental conditions also favour the survival of normally very short-lived horsetail spores, permit prolonged growth of the gametophyte generation and provide a near-constant film of free water, facilitating male gamete transfer conducive to cross-fertilisation (Page & Barker 1985).

The Massenerhebung effect in mountain areas

The ecological overlap of related species in Equisetum and in several other plant groups is probably facilitated by the wonderfully named 'Massenerhebung effect', a German term literally meaning, 'mountain mass elevation effect'. This is a meterological concept that was introduced by A. de Quervain in 1904 to account for the observed tendency for temperature-related parameters such as treeline and snowline to occur at higher elevations in the Central Alps than on their outer limits (Barry 1981). The concept stipulates that climatic and vegetation zones occur at lower altitudes on isolated mountains than they do in mountain blocks of increasing scale (Wardle 1974; Johns 1985). The energy physics relating to this mass-elevation effect are complex, however, and it can only be applied after very careful consideration of the specific local topographical and meteorological factors involved (Barry 1981, pp. 49-50).

Continentality versus Oceanity

The Massenerhebung effect relates to the much more generally applicable climatic concept of Continentality versus Oceanity. Applying either of these two concepts, we find that the low elevation, the small, isolated mass, and the maritime position of British and Irish mountains results in a greatly steepened temperature lapse rate and a marked compression and lowering of our vegetation zones in comparison with continental European uplands, including the Alps (Barry 1981, p. 265). The overall effect of this pattern of climatic variation allows southern species to migrate northwards along western Atlantic coasts, avoiding cold winters. This has produced, for instance, the highly unusual and very famous mix of phytogeographic elements found in the flora of the Burren region, Co Clare (H9). At the same time, montane or alpine plants and animals, with their requirement for cool summer conditions, can descend to lower levels and, in W Ireland, some of these come right down slope, very close to sea level (Praeger 1934, section 67; Webb 1983; Page & Barker 1985; Nelson & Walsh 1991).

Equisetum species mingling

Thus plant species, including sporophytes of our eight native Equisetum species, are often found growing much more closely together in the western parts of Britain & Ireland than they would under more continental growing conditions, where they would commonly be altitudinally, geographically and ecologically separated. Indeed, in the case of some other Fermanagh pteridophytes, colonies of four or more different but closely related species are often found growing intermingled, and ecological and genetic isolating mechanisms no longer apply. This enables the observed increase in frequency of hybridisation.

Reduced competition?

Perhaps we should not ignore the fact that in the case of Ireland, being a more ancient island than Britain, post-glacial species immigration was cut off earlier by sea level rise and thus the species-poor Irish flora presumably presents a less competitive environment than otherwise to newly arrived genetic combinations in the form of the prothalli of Equisetum hybrids. A similar argument would of course apply when comparison is made between the depauperate flora of Britain and that of mainland Europe.

E. × trachyodon is usually remote from its parent species

A comparative analysis of the geographical stations of species in the genus Equisetum and their then known hybrids was made around 1985 by Page and Barker based on updated hectad maps from The Fern Atlas (Jermy et al. 1978). This showed quite clearly that hybrids and their parent species behave quite differently in the two subgenera. In subgenus Equisetum, for every hybrid, examined over the British Isles as a whole, there is an almost 100% coincidence or association between the distribution of hybrid stations and the presence of both parents in the same or immediately adjacent hectad grid-square. The picture for subgenus Hippochaete is quite the opposite and, for E. × trachyodon, both parents were present in the same or adjacent hectads in less than 20% of its sites.

Our finer scale Fermanagh station analysis mentioned above shows the separation of E. × trachyodon and its parent species is even more pronounced than these workers showed. Page & Barker (1985) considered it likely that whilst a small amount of hybridisation may continue to take place in subgenus Hippochaete, the geographical evidence suggested to them that the majority of clones found today are long-established in their particular sites. Indeed, some clones may be very ancient and their parent species appear to have locally died out, or have been ousted by competition with their hybrid progeny.

European and world occurrence

Beyond the shores of Britain & Ireland, E. × trachyodon is locally fairly frequent in Norway, is widely scattered throughout Iceland and occurs in a few provinces in Sweden and Finland. It has also been reported from at least one station in S Greenland, although RSF has heard that other botanists have cast doubt on the identification: one of the parent species, E. hyemale, is said to be unknown there (Böcher et al. 1968; Jonsell et al. 2000). This might not be as significant as it first appears since exactly the same situation applies in Ireland with respect to E. × moorei Newman (E. hyemale × E. ramosissimum) (Moore's Horsetail), since E. ramosissimum Desf. (Branched Horsetail) is absent. In addition to the foregoing, E. × trachyodon is also reported from scattered localities in Czechoslovakia, France, Germany, Switzerland, Hungary and in parts of Russia and temperate N America (Duckett & Page 1985).

Names

The name 'trachyodon' is a combination of two Greek words meaning 'rough teeth' (Gilbert-Carter 1964).]

Threats

Clearing of riverbanks for fishing or agricultural purposes. Locally this is especially problematic along the Colebrooke River.

Growth form and preferred habitats

Fertile and sterile stems of Variegated Horsetail are identical except for the presence or absence of the small, pointed terminal cone which sheds spores in July and August. When taken together, the very slender, generally unbranched, often prostrate, evergreen stems of E. variegatum and the white teeth on the nodal sheaths, which Page so aptly describes as looking like "a broad gothic arch" (Page 1997), clearly distinguish the species from the hybrid it forms with E. hyemale (Rough Horsetail), E. × trachyodon (Mackay's Horsetail).

Variegated Horsetail is a scarce and local, rather variable calcicole species, which in Britain & Ireland occurs in a remarkably wide variety of more or less open, damp to wet, base-rich or calcareous sites, often by running water, or where there is movement of groundwater, even if only at subsoil level. Once established, its rhizomatous growth enables it to form large, compact stands as it still does at Praeger's original fen site at Spectacle Lough, the first Fermanagh site on record. However, it can also be found as scattered individual shoots at some of its more obviously flushed sites, both lowland and at higher elevations. Variegated Horsetail colonises seasonally flooded depressions and damp hollows in several disused quarries in Fermanagh, and it also occurs in similar ground under a hedge in the townland of Clontelaghan near Kinawley.

In other parts of Britain & Ireland, E. variegatum is reported in periodically flooded coastal sand dune slacks and from mountain ledges, neither of which it occupies in Fermanagh (Stewart et al. 1994). As is the case with E. hyemale (Rough Horsetail), in the more northern part of its range, E. variegatum occupies a much wider range of habitats, becoming more or less indifferent to base-status and lime, and being found in much drier situations. In Iceland, for instance, it grows on dry scree slopes and in dry heath (Jonsell et al. 2000).

Fermanagh occurrence

In Fermanagh, E. variegatum has been recorded in a total of 21 tetrads, three of which have pre-1975 records only. It is the eighth most frequent horsetail in the VC. As the tetrad distribution map indicates, the majority of sites are around Lower Lough Erne or on the higher ground to the SW of it.

Irish occurrence

Compared with the five other VCs in N Ireland, Fermanagh undoubtedly has the greater representation of this species. Tyrone (H36) also has three inland sites for Variegated Horsetail (McNeill 2010), but the species is completely absent from Co Armagh (H37) and most of the few remaining N Ireland sites are coastal (NI Vascular Plant Database 2010).

Elsewhere in Ireland, E. variegatum is thinly scattered, occasional to rare down the E coast and across the Midlands (Jermy et al. 1978; Webb et al. 1996).

Variation

Reflecting the wide variety of quite different habitats E. variegatum occupies in these islands, a number of ecotypes have evolved within it. These differ from one another in morphology (ie form, appearance and size), ecology (ie habitat and related features of growth) and in their geographical distribution. Having said this, in some cases ecotype ranges overlap to a degree not yet properly understood. Since ecotypes maintain their distinctive forms when grown together under identical garden conditions, the differences they display must be genetically determined, a feature unique amongst horsetail species in Britain & Ireland (Page 1997).

The most widespread Irish ecotype is var. majus, which as the name suggests is larger than the most typical form found throughout much of Britain, var. variegatum. Var. majus grows erect, rather than decumbent (ie leaning over or lying down, at least in part), and it is generally somewhere between 20 and 80 cm tall, with stems about 3 or 4 mm in diameter, making it half as thick again, or up to twice the width of var. variegatum (Webb et al. 1996; Page 1997).

The existence of these ecotypes undoubtedly provides a useful categorisation and description of the variation within the species (Page 1997). They include a coastal sand-dune form, var. arenarium, common enough elsewhere in Britain & Ireland, but only of incidental interest to us since Fermanagh has no coastline. Another Irish form, but remote from Fermanagh, is var. wilsoni, which is confined to Co Kerry (H1 & H2).

At the same time, it is important to recognise that E. variegatum is a phenotypically very plastic species, each ecotype being capable of displaying considerable modification of form and scale with respect to levels of a wide range of common local environmental variables (Stark 1991). Environmental variables include, for instance, shade, exposure, moisture, competition from other plants, trampling and grazing pressure.

Similar observations of a more local nature can be drawn from Paul Hackney's comparative study of E. variegatum at four sites in N Ireland. Three of the selected sites were coastal sand dunes, but the fourth was the fen shore of Carrick Lake in Fermanagh. Here, Hackney found that the Variegated Horsetail occupied ground with a 90% cover of mosses. It was typically robust, growing up to 50 cm tall, but where the vegetation had been subjected to grazing pressure, the plants only reached 20 cm in height (Hackney 1981). Hackney did not attempt to recognise E. varietatum ecotypes and, since in most of our survey we used the 1977 edition of Webb's An Irish Flora as our field guide, we have not done so either.

British occurrence

E. variegatum is a rather scarce plant in Great Britain, mainly concentrated in the N and W, becoming very rare in the C and S of England (Stewart et al. 1994; Stace 1997). The number of pre-1930 or pre-1970 hectads from which the plant has no recent records suggests the species is in decline and maps indicate that this is happening throughout both Britain and Ireland (Preston et al. 2002). Clearly this is a matter for concern.

European and world occurrence

In terms of geographical distribution, Variegated Horsetail is described as a northern-montane species (Stewart et al. 1994) or a circumpolar boreo-arctic montane species (Preston & Hill 1997). The latter may appear somewhat long-winded but it does summarise a range which stretches in Europe from Iceland and the Arctic Isles, to the northern tip of Scandinavia, thinning markedly towards the S of both Norway and Sweden, and present only at the northern tip of Denmark. In southern areas of these three Scandinavian countries it is also present as an introduction (Jonsell et al. 2000, Map on p. 25). E. variegatum then extends southwards in a somewhat scattered manner until it reaches a further centre of distribution in the Pyrenees, the Alps and other C European mountains (Jalas & Suominen 1972, Map 32).

Further east, E. variegatum reaches the Altai Mountains of C Asia, Manchuria and Japan. In N America, it stretches from the S Rocky Mountains to Labrador and around much of Greenland's coast (Hultén 1962, Map 45; Jonsell et al. 2000). Along the NW Pacific States of N America, the species is represented by a different form, E. variegatum subsp. alaskanum (A.A. Eaton) Hultén.

Names

The Latin specific epithet 'variegatum', means 'irregularily coloured' (Gledhill 1985), and the reference to the black and white banded sheath on the slender green stem is obvious. The English common name is a simple book name translation requiring no comment. As the plant is rare or scarce and it is easily overlooked, it has not accumulated any English folk names.

Threats

A number of sites are vulnerable to changes in agricultural practices. Locally, one quarry site was destroyed when the ground was covered in concrete.

Growth form and preferred habitats

This distinctive, erect, emergent aquatic or semi-aquatic rhizomatous, deciduous horsetail is very variable in size and in degree of branching. It is most commonly and abundantly found in still or slow-moving, shallow water by lakes, ponds and ditches, a habitat where it quite often represents the dominant colony-forming species. Even when dense pure stand communities of Water Horsetail are found, be advised by Wolfe-Murphy et al. (1992), who wrote with bitter experience, that the rhizomes are less robust than those of rather larger emergent species such as Schoenoplectus lacustris (Common Club-rush), and they do NOT form platforms that can bear the weight of the average botanist!

In lakes and ponds, E. fluviatile frequently forms large, dense stands, either pure or accompanied by a very long list of other common emergent wetland species such as Schoenoplectus lacustris, Cladium mariscus (Sword Sedge), Carex rostrata (Bottle Sedge), C. elata (Tufted-sedge), Sparganium erectum (Branched Bur-reed), Glyceria fluitans (Floating Sweet-grass) and Phragmites australis (Common Reed).

E. fluviatile can also persist, but to a lesser extent, in shallows in bays of larger water bodies which by their nature are subject to more water turbulence and wave-induced physical scour. In these circumstances, where Phragmites australis (Common Reed), Schoenoplectus lacustris or Typha latifolia (Bulrush) often represent the deeper water dominants, E. fluviatile regularly replaces Eleocharis palustris (Common Spike-rush) when water is deeper than about 50 cm (Spence 1964). Water Horsetail can also be the dominant species and regularly forms dense, almost pure stands in water 1.5 m deep or deeper (Spence 1964; Wolfe-Murphy et al. 1992; Page 1997).

Like its common relatives, E. arvense (Field Horsetail) and E. palustre (Marsh Horsetail), the rhizome of E. fluviatile runs much deeper in the soil than the underground organs of associated species, thus avoiding most root competition with them, if not altogether.

Identification

Individual stems are readily identified by the large central hollow that occupies between 80 and 90% of the stem diameter. As with several other Equisetum species, when E. fluviatile is in more open, unshaded situations, its stems are typically unbranched, or they bear only short, sparse lateral branches. The latter are usually irregularly whorled near the middle of the aerial length of the stem, but in more shaded situations, for instance in marsh or reed-swamp among fairly dense tall grasses, sedges and rushes, or under trees or scrub in fen carr, Water Horsetail regularly produces regularly whorled lateral branches on the emergent portion of its stems (Preston & Croft 1997; Page 1997, p. 446; Rose 1989, plate 52).

Fermanagh occurrence

Fermanagh has a huge number and variety of lakes, both very large and small, and they exhibit an enormous diversity of water chemistry (Gibson 1988). It is therefore not surprising that the shores of these supply E. fluviatile with plenty of scope for colonisation, and it is by far the most frequently recorded horsetail in the VC. While it has been recorded in 329 Fermanagh tetrads, 62.3% of those in the VC, E. fluviatile is not the most widespread horsetail, a distinction held by E. arvense (Field Horsetail).

Wide ecological tolerances

Water Horsetail is very remarkable for the extreme width of its ecological tolerances with respect to a spectrum of inter-related environmental factors, restricted levels of which typically curtail the growth and govern the occurrence of most other wetland plant species. The tolerances of E. fluviatile include levels of nutrients (from oligotrophic to eutrophic), lime content (from starved to rich), acidity-alkalinity (from pH 4-7.5), light (from full sun to half-shade) and exposure (from still backwater to open, moderately wave-beaten shore). The substrate textures it tolerates vary from clean, firm, mineral sand to silty, smelly, organic mud, deep enough to suck the boots off you!

Water Horsetail can also survive in oxygen-depleted, highly anaerobic, hydrogen sulphide-releasing conditions that exclude many other aquatic species (Grime et al. 1988; Wolfe-Murphy et al. 1992; Preston & Croft 1997). The very large central cavity in the hollow stem of E. fluviatile is considered to be an adaptation allowing air to diffuse downwards to the rhizome, in which latter organ, unusually among Equisetum species, the wide lumen persists. Very possibly it is this unusual morphological property which enables the rhizome of E. fluviatile to penetrate, grow and persist in anaerobic layers of mud (Page 1997).

In N Ireland, Wolfe-Murphy et al. (1992) carried out a detailed lake survey for government from 1988-1991 covering all six counties in the Province. This generated a macrophyte vegetation classification of lake vegetation using the computer program 'Twinspan' (unpublished report to DOE, NI, p. 294). The study found that E. fluviatile was abundant or dominant in eight of the 30 shoreline plant communities these workers defined in N Ireland, indicating the exceptionally wide ecological range of the species. Water Horsetail is common on exposed mud or shallow water in sheltered backwaters of larger lakes, and occasionally so in sluggish rivers and streams, particularly if there is only a minor fluctuation in water levels. It also frequents, in a more scattered manner, the closed turf vegetation of marshes, swamps and Salix-Alder fen-carr habitats, where its status in these plant communities is that of a minor companion species, except when shade seriously restricts the more competitive species around it: reduced light allows rhizomatous E. fluviatile to reassert its vigour and develop a more significant presence (Grime et al. 1988).

Reproduction

Again as is the case in both E. palustre and E. arvense, although E. fluviatile produces vast numbers of cones and spores, conditions for the completion of the full sexual life-cycle are stringent, and prothalii are very seldom observed (Page 1967; Duckett & Duckett 1980). It is thus very probable that increase and dispersal of Water Horsetail is heavily dependent on vegetative reproduction, achieved mainly by fragmentation. Propagation involves either free-floating segments of broken stem, or rhizome fragments. E. fluviatile does not possess tubers (Page 1997).

Herbivory and incidental vegetative propagation

Ducks, geese and other waterfowl feed on the stems of the plant (Fassett 1957), and Coot have also been observed biting off stems and using them for nest building (Praeger 1934). In both these circumstances, the birds were observed deliberately breaking excessive numbers of Equisetum shoots, more of them than they actually used (and in the case of feeding birds, they broke them into many small segments), so that some of the stem pieces could disperse without being used. These stem fragments can develop roots and may thus propagate the plant on suitable wet terrain (Praeger 1934; Page 1997).

British and Irish occurrence

E. fluviatile is a very common, widespread and locally abundant species throughout most of Britain & Ireland occupying a wide variety of habitats. It has declined quite substantially in the last 50 years, due mainly to drainage of smaller wetlands and poor, unsympathetic management of other sites (C. Dixon & T.D. Dines, In Preston et al. 2002).

Environmental pressures and species change

The distribution of Water Horsetail in both Britain and Ireland has undoubtedly been somewhat curtailed in the past century by agricultural drainage and other forms of development. These habitat pressures are most obvious in the most heavily populated and intensively farmed areas in the south of England (Grime et al. 1988; Preston & Croft 1997). At the same time, in other areas of these islands it is possible that E. fluviatile may have benefited from the suppression or demise of other aquatic and marsh species more sensitive to increased eutrophication than it. This is not to suggest that E. fluviatile tolerates extreme levels of organic pollution or sewage-induced accelerated cultural eutrophication. Study of a polluted lake in S Wales showed the species declined noticeably under such circumstances (Wade 1999).

In comparison to crop plants and terrestrial herbs, aquatic macrophytes have not been much studied by scientists working on climatic change. However, in a three year study, Ojala et al. (2002) found growth and reproduction of E. fluviatile was sensitive to a rise of around 2.5 to 3°C.

European and world occurrence

Water Horsetail has a widespread and more or less continuous distribution in boreal and temperate parts of Europe that is very similar to the occurrence of E. palustre. However, the distribution of E. fluviatile thins much more noticeably than E. palustre in the Mediterranean basin (Jalas & Suominen 1972, Map 34). E. fluviatile spreads eastwards from Turkey and the Caucasus, through temperate Asia to N Japan and N America. Only its absence from Greenland prevents it from being circumpolar (Hultén 1962, Map 96; Hultén & Fries 1986, Map 15; Jonsell et al. 2000).

Names

Since it so frequently occurs in muddy ground, E. fluviatile used to bear the name 'E. limosum', 'Mud Horsetail', the Latin specific epithet being derived from 'limosus', meaning, 'of marshy or muddy places'. An earlier name of the plant was 'E. heleocharis', derived from two Greek words, 'helos', meaning 'marsh', and 'karis', meaning, 'charm, grace, or beauty' (Gilbert-Carter 1964; Johnson & Smith 1931). By comparison, the modern specific epithet, 'fluviatile' is derived from the Latin, 'fluviatilis', meaning, 'growing in a river or in running water' (Stearn 1992). In the current writer's opinion, this is an inaccurate and completely misleading indication of the normal habitat of this horsetail, which more often is characterised by still or slow-moving water.

Additional English common names for E. fluviatile include 'Smooth Horsetail', a feature of the stems, said, in the past, to make them acceptable to cattle as food. However, reference to Linnaeus in this connection suggests he was really referring to E. telmateia (Great Horsetail), since the latter was his 'fluviatile' (Grieve 1931; Step & Jackson 1945).

Other interesting local names include 'Paddock Pipes' and variants thereof (eg paddow, paddie and puddock). 'Paddock' is a Scottish name for frogs, making it rather appropriate for this horsetail, but it is devalued by being used for other horsetails (Britten & Holland 1886). Page (1988) mentions the interesting name, 'Trowie Spindles', but offers no details of its origin or derivation.]

Threats

Drainage probably poses the only threat likely to affect this species locally in Fermanagh.

Growth form and preferred habitats

Typical plants of the deciduous perennial hybrid E. × litorale are fairly well branched in the middle portion of the stem, with the upper half to one third of the stem being long, curved and unbranched. The internodes, when squeezed gently give slightly, and then 'bounce back' to their original diameter. In comparison, the internodes of E. arvense when squeezed in this manner do not 'give' at all.

Although like many other horsetails, E. × litorale most frequently colonises open, disturbed ground, Shore Horsetail possesses hybrid vigour and is strongly competitive, often forming large colonies. In fact, like both its parents, E. × litorale can occur in a very wide variety of damp to apparently quite dry habitats (Page 1997).

In sharp contrast to our other most frequent Fermanagh hybrid horsetail, E. × trachyodon (Mackay’s Horsetail), more often than not Shore Horsetail occurs in close proximity to one or both of its parent species (Page & Barker 1985). However, we do have at least one site on the Tempo River where E. × litorale occurs with both its parents, together with E. telmateia (Great Horsetail) and E. × trachyodon!

The phenomenon of many horsetail species growing together in close proximity and the relatively high frequency of hybrids in western parts of British & Ireland has already been discussed at length in the current authors E. × trachyodon account, so interested readers are requested to look there for more information and opinion.

Variation

Plants of Shore Horsetail are usually intermediate between the parent species in most characters, but they are extremely variable, the variation undoubtedly being induced by very local environmental growing conditions. As a result, when occupying relatively dry conditions E. × litorale closely resembles E. arvense and, when the habitat is wetter, its features are most like E. fluviatile.

Fermanagh occurrence

This vigorous, competitive hybrid has now been recorded in a total of 60 Fermanagh tetrads, 11.4% of those in the VC, 52 of them containing post-1975 records. Despite figures like the above, Preston & Croft (1997) still regard Shore Horsetail as frequently overlooked and under-recorded in much of Britain & Ireland.

In Fermanagh, this hybrid between a terrestrial and an aquatic horsetail species most often grows on damp, bare gravelly ground near our larger lakes, in ground that is either permanently wet, or is subject to at least occasional flooding. It is especially frequent around Lower Lough Erne. Very rarely it is found in open peaty mud on bogs, for example at Rossgweer Bog, or on damp roadside verges. Our Fermanagh stations closely fit the types of habitat it occupies elsewhere in Britain & Ireland, described by Preston & Croft (1997) as marshy ground, near bare or disturbed soil.

Of the three hybrid horsetails in Fermanagh, E. × litorale is the more frequent, being known from 52 post-1975 tetrads compared with 22 for E. × trachyodon, and just one for the extremely rare, weak and very probably ephemeral E. × dycei.

Reproduction

E. × litorale produces cones in June and July, but the spores and reported to all abort (Duckett & Page 1985).

British and Irish occurrence

It is interesting to note that Praeger, who first found this hybrid horsetail in Fermanagh in 1939, was writing in 1917 of it having just, "one station in each of the three kingdoms, England, Scotland and Ireland". The New Atlas records the plant in Ireland from a total of 175 hectads with post-1970 dates scattered across the island, and 315 post-1970 hectads throughout Britain, but concentrated mainly in the N & W. The patchy occurrence in isolated VCs however indicates a degree of recorder bias, eg Surrey and Worcestershire (VCs H17 and H37) (New Atlas).

European and world occurrence

Apart from Britain & Ireland, E. × litorale is widespread in Fennoscandia and very likely also occurs throughout the whole northern boreal range of the parent species. It probably occurs everywhere in the world the distributions of the parent horsetail species, E. fluviatile and E. arvense, overlap (Jalas & Suominen 1972, Maps 34 & 38; Hultén 1962, Maps 96 & 98; Duckett & Page 1985; Hultén & Fries 1986, Maps 15 & 19; Jonsell et al. 2000).

Names

The Latin epithet, 'litorale', is derived from 'litus', meaning 'shore' and obviously refers to the most frequent habitat of the hybrid – lake shores (Gilbert-Carter 1964).

Threats

None.

Growth form and preferred habitats

There are no less than five hybrid horsetails in Ireland, the most common of which is E. × litorale (Shore Horsetail). However, despite the high frequency of its parent species, E. × dycei is extremely rare in both Britain & Ireland. This horsetail hybrid is small, slender, with unbranched or sparsely branched shoots and a very long branchless terminal portion. It is very spindly or sickly-looking, lacks hybrid vigour (heterosis) and resembles a smaller, weaker form of E. palustre (Marsh Horsetail) or a debilitated form of the usually very much more vigorous hybrid E. × litorale (E. arvense × E. fluviatile). E. × dycei is mainly confined to man-made, open, muddy roadside ditch habitats (Page 1988 & 1997).

Irish occurrence

On a joint Botanical Society and British Pteridological Society outing to NW Ireland in July 1984, Maura Scannell of Glasnevin Botanic Garden in Dublin and Page together discovered the first Irish station for E. × dycei along the southern shore of Doon Lough, in Co Leitrim (H29), adjacent to Fermanagh. In this case, the hybrid was growing on the marshy shore of the lake, by the roadside, along with both parents and it was recognised by Page on site. A voucher of this find is now in DBN. In July 1986, Scannell found a second station in N Kerry (H2) on the shore of Inisfallen Island, Lough Leane, Killarney. A voucher for this latter discovery was sent to the National Museum of Wales herbarium in Cardiff (NMW), where Page again confirmed the identification (Scannell 1995).

The 1997 second edition of Page's book, Ferns of Britain and Ireland deals with E. × dycei in great detail. The account includes a very small and quite inadequate map, which nevertheless indicates a total of five stations in the Republic of Ireland, the two mentioned, plus one in Kerry, and two additional stations in Co Clare (H9)(Page 1997, p. 492). While one may deplore the poor quality of Page's map, there is no map at all in the BSBI's New Atlas, nor on the accompanying compact disk (Preston et al. 2002).]

Fermanagh occurrence

Sean Wolfe-Murphy made the solitary Fermanagh record listed above when taking part in the NI Lake Survey in 1991. The plant was determined by Paul Hackney from a voucher specimen in BEL. It was growing with E. fluviatile (Water Horsetail) but the other parent was not recorded. This is a new Fermanagh County Record and was the first record of the hybrid in NI. The BEL voucher has the accession number H37907. A second NI station was subsequently discovered in June 1999 by P. Hackney at Grange More Td, just SE of Castlerock on the N coast of Co Londonderry (H40).

Ephemeral nature of the plant

In field notes relating to this hybrid, Page (1997) wrote that from his experience, E. × dycei seems an elusive and possibly short-lived plant and, indeed, by 2008 the Grange More plant had completely disappeared (P. Hackney, pers. comm. 2010). The Edergole Island station has never been refound either. A third find of this hybrid in NI was made in July 2008 by Dave Riley, at Colm Harkin, near Swatragh, Co Antrim (H39). Vouchers for all three NI records are in BEL.

Earlier, Page (1973) suggested that the weak hybrids in this genus are those formed between ecologically similar, but not very closely related pairs of species, while the vigorous ones are those between species pairs which are ecologically more different, yet closely interrelated.

In view of the abundance of both parents and their frequent overlap in many Fermanagh sites, we fully expect that there must be further stations for this perhaps casual hybrid to be found, but only a very experienced eye could detect them.

Names

Chris Page gave E. × dycei its names, both Botanical and English, calling it 'Hebridean Horsetail', after its first discovery in 1962 on these Scottish islands.

Threats

None.

Growth form, identification and preferred habitats

As any gardener or farmer will testify, E. arvense is a too frequently found rhizomatous and tuber producing perennial weed, which grows so deeply and spreads so relentlessly and rapidly that it is widely regarded as nearly impossible to eradicate (Holm et al. 1977; Grime et al. 1988). The acute, green, spreading teeth on the side branches of E. arvense help distinguish its sterile stems from the rather similar ones of E. palustre (Marsh Horsetail), in which the corresponding branch teeth are black-tipped and tightly clasp the stem for their whole length. In terms of identification it is fortunate that at least some of the characteristics of the side branches are unvaryingly reliable: irrespective of the size of the sterile stem of the plant, the lowest internode on each side branch of the stem of E. arvense is always equal to or longer than the adjacent stem sheath on the node from which it arises (Jermy & Camus 1991).

Typical habitats include rough grassland and disturbed areas in damp open woods, hedgerows, field-, river- and roadside banks, lakeshores, cliffs, screes, quarries, gardens, gravel paths and waste ground. Although Field Horsetail is able to grow in a very wide variety of soil types, it has a definite preference for neutral or slightly base-rich conditions. It is really only common and vigorous where there is a high water table and the soil is poorly drained (Williams 1979; Cody & Wagner 1981). Having said this, E. arvense is recognised to be the most plastic and adaptable horsetail in Britain & Ireland and it can survive and function over a very wide spectrum of soil moisture and nutrient regimes (Page 1997).

Field Horsetail occurs in a huge range of more or less open habitats and like Ulex europaeus (Gorse), Pteridium aquilinum (Bracken), Tussilago farfara (Coltsfoot) and other vigorous perennial weeds, it particularly avails itself of soil disturbance. Frequently such disturbance is due to mans' activities, but it can also be naturally occurring, for example in waterside habitats. Indeed, it has been suggested that river banks may have been the original natural habitat of E. arvense (Hauke 1966), and possibly of all Equisetum species.

Whatever the cause of soil disturbance, it creates an opening for the horsetail, providing bare ground or a gap in the vegetation into which the species then readily invades (Salisbury 1964; Grime et al. 1988). The last mentioned authors found that in their main study area around Sheffield, E. arvense was absent in 'skeletal habitats' (ie bare rocks, cliffs, scree and walls). However, in Fermanagh, we do have records of it colonising cliffs and screes, which might be another reflection of our very wet, more or less hyper-oceanic climate.

Fertile sporing and green vegetative shoots

Stems of E. arvense are of two types; the short, stout, pale- or buff-brown, cone-like, fertile sporing branches appearing for a fortnight or so in early April. As the non-photosynthetic sporing shoots appear before the green vegetative ones, they are sometimes referred to as 'precocious'. The fertile shoots are soon followed by the much more slender, green sterile or barren stems, branched to varying degrees. The green shoots are deciduous, growing and persisting until the frosts in late autumn, and then dying and disappearing in November, or by early December at the latest.

Rhizome extent and ecology

The extremely persistent, perennial, underground rhizome is long and branched and can reach a depth of 1.5 to 2 m., or more (Page 1997). In warmer climates it can exceptionally extend down to around 6 m deep (Cody & Williams 1981). Having said this, Williams (1979) found that in a sandy loam in England, 50% of the rhizome dry weight occurred in the uppermost 25 cm of soil, a further 25% in the next 25 cm and only 10% of the total rhizome material was found between 75 and 100 cm deep. He also found that if arable ground containing E. arvense was left fallow for one or two years, rhizomes and starch-filled tubers occurred much more shallowly, more than 80% of them being resident in the uppermost 25 cm of soil. He concluded that this was most likely due to the horsetail being released from competition by the absence of any crop and other work he quotes from different countries appears to support this suggestion (Williams 1979). Climatic factors such as the depth of soil warming are known to affect the depth to which rhizomes and tubers penetrate, as will soil aeration and moisture status (Williams 1979).

The rapidity of rhizome spread reported is both amazing and alarming; for example, a 10 cm length of rhizome planted in a growth room (presumably, although not specified, in the absence of competition) produced in one year a total of 64 m of growth in a vertical direction through branching (Cody & Wagner 1981). As far as the current author is aware, horizontal rhizome growth rates have not been accurately measured under field conditions, but an individual rhizome has been reported achieving a spread of 100 m (Weber 1903, quoted in Cody & Wagner 1981).

Nutrient requirements

The nutrient demands of E. arvense on soil fertility were concisely expressed in classic ‘crop versus weed’ terms by Salisbury (1964) when he wrote, "its filching of soil nutrients is very appreciable". Hill et al. (1999) placed Field Horsetail among other rather nutrient-demanding weeds such as Cirsium arvense (Creeping Thistle) and Rumex crispus (Curled Dock), which typically require moderately fertile soil. Sinker et al. (1985) characterised E. arvense as requiring a medium to rich supply of phosphate, nitrogen and other mineral nutrients. Field Horsetail, however, appears as variable in this respect as in other environmentally controlled characters, since in his Rothamstead pot experiments, Williams (1979) found that the horsetail seemed well adapted to growth in soil with low nitrogen and it produced little growth response to additions of it.

Silica content

The horsetails, particularly E. hyemale (Rough Horsetail), but to a lesser extent E. arvense, are well-known for their ability to take up and deposit silica in the walls of the epidermal cells of their shoots. The silica content of dry E. arvense can vary between 1.2 to 6.9% and the ash can contain anywhere between 6.2 and 76% silica. This is why these horsetails have in the past been widely used for scouring pots and as polishing and buffing tools (Cody & Wagner 1981). The low levels of silica in both limestone and deep peat soils probably limits the growth of Equisetum species where these soils occur (Page 1997).

In mining areas E. arvense and E. palustre (Marsh Horsetail) are relatively frequent on metalliferous spoil heaps, the soils of which most other plants find too toxic to colonise (Grime et al. 1988). E. arvense has an unusual ability to take up and accumulate heavy metals such as copper, zinc, lead and cadmium, so that it has been proposed as a biological tool to monitor levels of pollution (Ray & White 1979). Reports quoted in Holm et al. (1977), that horsetail accumulates gold in quantities of up to 4.5 ounces per ton of fresh plant material, are exaggerations based on mistaken chemical analytic methods (Grime et al. 1988).

Ecological flexibility

It is the degree of ecological flexibility E. arvense displays in terms of both requirements and tolerances, which enables it to become the only horsetail that is a common weed of cultivated ground. Following a detailed systematic study, Hauke (1966) concluded that despite appearances, E. arvense is a hydrophyte! This decision calls into question how we define an aquatic plant, but we can hardly believe that many biologists or ecologists would entirely agree with Hauke's use of the term for this species. He supports his contention by pointing out that although Field Horsetail is sometimes found growing in very dry habitats (for instance along roadsides and railways), on the basis of his observations and measurements, he believes that it is quite sensitive to moisture stress and that it is able to survive in places where other plants cannot, simply because its extensive, deeply-penetrating rhizome system always manages to tap groundwater supplies.

Variation

Being versatile in terms of its ecology, E. arvense also demonstrates a very wide range of environmentally induced modification (phenotypic variation) of both plant size and form, some of which are easily demonstrated even within a single clone (Hauke 1966). Examples of this variation are very clearly illustrated in Page (1997, pp. 439 & 442), but Hauke (1966) concluded that all this variation is only superficial and none of it merits taxonomic recognition.

Asexual reproduction and dispersal

E. arvense dispersal can readily be achieved by its lightweight spores, but without question it also spreads very effectively vegetatively, both by means of rhizome fragments and through its small, oval, root tubers being carried unwittingly in mud attached to machines, animals and boots. Establishment from spores involves the successful negotiation of several high-risk stages of development, involving the ready release and transport of asexual spores and the production of gamete-bearing prothalli, structures which are known to have an extremely narrow habitat tolerance. The prothalli require bare muddy ground of high nutrient status, neither too wet nor too dry, and the entire absence of shading and competition, even that from mosses and liverworts. In view of this, the rarity of prothallus observations in nature is not all that surprising. Nevertheless, the existence of Equisetum hybrids and the few studies of wild gametophytes that have been made, do prove that sexual reproduction does happen. Even if it is a rare event, sexual reproduction will be significant since it maintains genetic variability and heterogeneity within horsetail populations that could not otherwise occur (Duckett & Duckett 1980).

Sexual and vegetative reproduction

As with other homosporous (one type of asexual spores only), sporophyte (spore-bearing plants) (ie the ferns and horsetails), vast numbers of spores are produced and released in the early spring (generally from early April into May). A small proportion of these asexual spores germinate to produce a prothallus, on which separate male and female sex organs develop. Huge numbers of male gametes (sex cells) swim in a film of water to fertilise the much fewer female ova, and young sporophyte plants successfully produced by this sexual cycle then grow and develop. Some of these juvenile sporophyte plants have been known to produce up to two dozen shoots and several tuber-bearing rhizomes by the end of their first season of growth (Page 1967; Duckett & Duckett 1980).

By comparison, vegetative dispersal and successful subsequent establishment appears very much more probable and this is generally regarded as the more usual method of reproduction of the species (Marshall 1986; Grime et al. 1988).

Sporophyte light requirements and colony development

Once the plant has colonised (by whatever means) and a rhizome has been established, growth of the horsetail can be incredibly rapid, allowing the plant to quickly form a clonal patch. E. arvense has such a terrible reputation among gardeners, fruit-growers and farmers (both arable and grass-managing), that along with its relative E. palustre, it appears in Holm et al. (1977) book entitled, The World's Worst Weeds. This might be pitching its status a little higher than it deserves. The reason for saying this is that, having no leaves, E. arvense entirely relies for its photosynthesis on its annual, green, wiry, branching stems, which grow up to 80 cm in height. Thus most Equisetum species cannot tolerate much deep shade, and the horsetail plant casts very little shade itself (Salisbury 1964).

E. arvense has been given an Ellenberg's indicator value by Hill et al. (1999) for its light preference in the British Isles of '7' on a scale from '1' to '9', meaning that it is regarded by these authors as a plant generally found in well lit places, although it may also sometimes occur in partial shade. This light requirement limits the species competitive ability, and being relatively low growing and possessing a comparatively insignificant canopy of its own, it never (or rarely ever), becomes a dominant plant after the manner of taller, aggressive and persistent invading weeds, such as gorse and bracken. Grime et al. (1988) draw a very interesting parallel between many of the features of E. arvense and Tussilago farfara (Coltsfoot), both in terms of their pioneering, colonising ability and weedy persistence.

Toxicity

Pretty well all species of Horsetails are dreaded worldwide, partly for their deep seated growth and their enduring survival ability, but also because the plants are so poisonous to cattle, sheep and horses. Here again, comparison may usefully be made with Pteridium aquilinum (Bracken), since both fern and horsetail contain a toxic enzyme called thiaminase, which destroys thiamine and creates a vitamin B1 deficiency in any monogastric animals which graze them (Cody & Wagner 1981).

Equisetum species also contain high levels of a number of toxic alkaloids, of which the best known is palustrine (Cooper & Johnston 1998). In this case, the poisonous principle is not destroyed by drying and storage and horses can show clinical signs of poisoning when their hay contains as little as 5% horsetail (Cooper & Johnson 1998). Fortunately the plants contain varying quantities of silicates, making them harsh to touch and unpalatable, at least in the fresh state, so that animals tend to avoid grazing horsetails and therefore poisoning is rare in Britain & Ireland.

The levels of toxins in Equisetum species suggest the possibility of allelopathy, ie toxic suppression, directed towards surrounding competing plants. Work on this topic in Russia is regularly quoted, which showed that when tested along with twelve other species, water extracts of E. arvense displayed the strongest inhibitory effect on seed germination and seedling vigour when it was applied to 30 species of meadow grasses (Zelenchuk & Gelemei 1967 (in Russian), quoted by Cody & Wagner 1981).

Weed control

Control of E. arvense is extremely difficult. Forking out, cutting and burning all prove a useless waste of time and effort. Eradication may be achieved in the long term by shading the weed out with taller plants, which is what normally happens to the species in undisturbed natural vegetation (Page 1997). In the garden, even as simple a matter as sowing a patch of Nasturtium has been recommended for this purpose (Allan 1978). Mulching with leaf compost is reputed to stop lateral movement of Field Horsetail, and it has been reported that black plastic sheeting placed over infested soil killed rhizomes in the upper 60 cm of soil within three to four years (Cody & Wagner 1981).

Herbicide experiments

The choice of herbicides used against Equisetum infestation is governed by the crop or vegetation type affected and the scale of the weed problem. For instance, it was found that MCPA applied after the horsetail had completed emergence gave 100% control of aerial growth for the rest of the season and also reduced the number of horsetail stems emerging in the second year (Hoyt & Carder 1962, quoted by Cody & Wagner 1981).

Glyphosate and other translocated herbicides, such as Asulam (which in particular is also effective in similar manner against Bracken), can be used in uncultivated areas. Glyphosate herbicides are most effective if the horsetail is allowed to fully emerge and its stems are then crushed before applying the chemical in late summer (late July or August is probably the best time) (Marshall 1986). Crushing is recommended because the slender stems and branches are covered with microscopic silica spicules and therefore they are not easily wetted.

Fermanagh occurrence

E. arvense is both the most common and widespread horsetail on a world basis and the most widespread horsetail in Fermanagh, present in 433 post-1975 tetrads, 82% of those in the VC. However, since Fermanagh is a county extraordinarily well supplied with lakes and wet marshy or boggy ground, in terms of record frequency, Field Horsetail remains second to E. fluviatile (Water Horsetail) by a margin of over a thousand records. E. arvense is found throughout the county, although it is uncommon on high ground and absent from aquatic situations, except the gametophyte, which is a pioneer coloniser of muddy water margins of lakes and reservoirs, very occasionally producing the sporophyte generation (Duckett & Duckett 1980).

British and Irish occurrence

Field Horsetail is the most common and widespread horsetail in Britain & Ireland, being present in every VC. The distribution thins slightly in areas of predominant deep peat soils in N Scotland and in the SW of both islands (Jermy et al. 1978; Page 1997; Preston et al. 2002).

European and world occurrence

E. arvense is very widespread throughout most of Europe. It extends into the far south of Italy, but is more thinly present in the Iberian Peninsula. It occurs in most of the Mediterranean islands, but not in Cyprus (Jalas & Suominen 1972, Map 38). It stretches eastwards through the Caucasus, the Himalaya, C China and Japan to the S USA. It was introduced into Mexico and New Zealand and may or may not have persisted (Hultén 1962).

E. arvense is the most common and widespread species of the genus, being circumpolar chiefly in boreal latitudes and present though less common in the Arctic, including Greenland, Iceland (Hultén & Fries 1986, Map 19). It also occurs in a prostrate form, subsp. boreale (Bong.) Á. Löve, in the Arctic circumpolar region including the mountains of Norway, Sweden, Iceland and the arctic islands (Jonsell et al. 2000).

Uses

The starch-filled rhizome tubers of E. arvense are eaten by ducks in Alaska (Hauke 1966) and in the past N American Indians both peeled and ate raw stems of the fertile cone-bearing stems and used the dried ashes of the sterile stems to treat sore mouths (Cody & Wagner 1981). Grieve (1931) list many medicinal uses for the fresh or dried sterile stems and their ashes, the main uses being as a diuretic and astringent, to staunch bleeding (including nose bleeds and ulcers) and to counter acidity of the stomach.

Names

The genus name 'Equisetum' was coined by the ancient Roman writer, Pliny and is thought to have been first applied by him to E. arvense. It is a combination of two Latin words, 'equus', a horse and 'saetum', a bristle or hair, and it is thought to refer to the bristly appearance of the jointed stems with their whorled branches (Gilbert-Carter 1964; Grieve 1931). The same notion also gave origin to the English common name 'Horsetail', which is a direct translation of the medieval Latin name, 'cauda equina', under which it was sold in apothecary shops (Prior 1879; Grigson 1974). The Latin specific epithet 'arvense' is a common one, being derived from 'arvum solum', meaning 'arable land', on which the plant is often found growing (Gilbert-Carter 1964).

There are a total of 21 additional English Common names for Equisetum species in general in Britten & Holland’s reference work (1886), many of which carry watery connotations involving pipes, frogs (and tadpoles), toads and rushes, eg Tadpipes, Tad-broom, Toadpipes, Snake Pipes, Water Grass, Cat-rushes. Several other names refer to the bushy appearance of the branched stem, for example, Cat's-tail, Colt's-tail and Bottle-brush (Britten & Holland 1886). The weedy, unwanted nature of the plant, or its malign, pernicious presence is featured in a Welsh name which is translated as 'Evil man's garters', the evil man being a standard euphemism for the devil (Awbery 1984).

Threats

None.

Growth form and preferred habitats

E. pratense is a rhizomatous perennial with two forms of aerial shoot (vegetative and sporing), neither of which overwinter. The species is a northern one with an overall distribution markedly circumpolar boreal-montane or arctic-alpine (Stewart et al. 1994; Page 1982 & 1997). In Ireland it is a rare Irish Red Data Book species, confined to the northern province of Ulster, where it is most frequent in the glens of Antrim (H39), although even here it is really rare and local (Praeger & Megaw 1938; Curtis & McGough 1988, pp. 93-4; Hackney et al. 1992). E. pratense is rarely recorded, perhaps in part because it is not very distinctive in appearance and it often occurs as scattered, diminutive individuals composing small, diffuse colonies. In order to locate it, field recorders first need to encounter a good specimen colony and get to know its particular habitat requirements. Thus it could easily enough be overlooked and under-recorded.

In texture and colour, the shoots of E. pratense most closely resemble the delicate ones of E. sylvaticum (Wood Horsetail), although unlike the latter, the slender primary lateral branches are themselves unbranched. The flat tops of mature shoots are quite distinctive in appearance.

Like the majority of other horsetail species, it is a plant of well-drained slopes (including upland glens and talus slopes below cliffs), where silty or sandy soil is flushed with, or subject to seepage of, base- and mineral-rich groundwater, and where shade, shelter and proximity to running water is sufficient to keep humidity high and prevent any possibility of desiccation (Stewart et al. 1994). A certain amount of winter flooding or mild scouring of base-rich E. pratense sites near streams and rivers may also be significant, producing surface instability and shallow erosion of the wet soil, thus minimising the burden of competing vascular plant species while simultaneously providing plentiful silica (Page 1988). In Fermanagh, this could particularly be the case at the Cladagh River Glen site. To some extent, therefore, Shade Horsetail may be considered a pioneer species, growing best in situations where relatively open conditions are maintained by these types of naturally operating factors. Both of the Fermanagh sites are relatively lowland, but the species is reputed to reach 850 m in Glen Coe in Scotland.

Fermanagh occurrence

In Fermanagh, E. pratense is found in only five tetrads and is the rarest horsetail species in the VC. It is twice as rare as E. hyemale (Rough Horsetail), which is known in twelve post-1975 tetrads), and much less frequent than two of our three hybrid horsetails, E. × litorale (Shore Horsetail) and E. × trachyodon (Mackay's Horsetail). In fact, there are just two main sites (with subsites), in the county for Shady Horsetail in shaded, moist woods and scarps, but they are rather different from one another.

In the Cladagh River Glen (also referred to as the 'Marble Arch Glen'), E. pratense grows in one large patch on a moist, sloping bank in the light shade of an ash dominated woodland. It also has a number of smaller colonies nearby, along the bank of the river either side of the adjacent path. At the second more extensive site on the scarp top of the Magho cliffs and the talus slope below that overlook Lower Lough Erne, Shady Horsetail occurs thinly scattered within the canopy of a six km long stretch of wet, mixed deciduous woodland, growing here amongst Luzula sylvatica (Great Wood-rush) and Stellaria holostea (Greater Stitchwort).

Even when failing to cone and restricted to an entirely vegetative condition, the persistence of E. pratense is quite remarkable. Records of Shady Horsetail at Co Antrim sites (H39) extend back up to a century or more. The plant has not, however, been refound in Fermanagh at Praeger's original 1904 site, which he (in the days before grid references) rather loosely described as being amongst rank heather on an open moorland scarp, behind (ie south of) Poulaphouca cliffs – a name rather widely printed across several cliff ranges on the one inch OS map (Praeger 1904). Nevertheless, the species persists on adjacent scarp tops and scarp woodland along the Cliffs of Magho and, indeed, if one can call it such, the species has its Fermanagh headquarters here.

Sterility and changing winter temperatures

In Fermanagh, in common with elsewhere in Britain & Ireland in recent years, observation suggests that E. pratense is sterile and cones only very rarely and spasmodically, if at all. E. pratense appears to rely almost exclusively on vegetative reproduction for its increase and survival, and on rhizomatous creep for any local spread that it manages to achieve. The species is, however, long persistent in the vegetative state in Britain and Ireland, most of the sites known a century or more remaining extant (Stewart et al. 1994).

As Page (1982 & 1997) points out, herbarium specimens from the 19th century indicate that E. pratense used to cone abundantly, and possibly did so regularly. However, it is possible to imagine that field botanists may have favoured or restricted their collecting to 'complete', ie sporing specimens, thus unintentionally distorting the frequency of cones in the herbarium record.

Since the great majority of Shady Horsetail sites in Britain & Ireland are fairly remote, they are left relatively undisturbed by man's activities. In the absence of human disturbance, the near sterility of modern clones suggests some other environmental factor(s) controlling cone induction has changed sufficiently during the last 100 - 200 years to prevent spore production. The most likely rapidly changing environmental factor that might affect horsetail fertility is winter temperature. Since about 1850 AD there has been a gradual climatic warming in Britain & Ireland producing milder winters. This trend began just prior to current CO₂ and other greenhouse gas-induced global warming associated with the industrial revolution. Even slight climate warming may have provided the significant biological threshold beyond which E. pratense fails to produce sporing cones (Page 1997).

Other evidence of species decline

In Fermanagh, E. pratense is at the very southern edge of its W European geographical range and, because of its poor to negligible reproductive capacity in modern times, Page (1982 & 1997) fears that the species is in the process of dying out in Britain & Ireland. Species living at the extreme limits of their distribution are often represented by small, scattered, isolated populations, which frequently are either completely sterile or are reproductively weak. Compared to populations in the central parts of the species distribution, they tend to occur, as E. pratense now does here, in smaller numbers and in a much more restricted range of habitat conditions than occurs elsewhere within their overall distribution. These features indicate such species are suffering from both poor competitive ability and a lack of recruitment into fresh suitable sites.

In genetic terms, small, isolated populations inevitably lose vigour over extended periods of time, gradually losing genetically variability (sometimes referred to as 'genetic erosion'), as a consequence of restricted gene flow and 'genetic drift' (ie the tendency for gene alleles to fix within small populations at random, or even somewhat against selective forces) (Richards 1997, p. 46-9).

These processes associated with inbreeding lead to the accumulation of homozygous gene alleles and deleterious recessive mutants but, in polyploid species, such as most pteridophytes, these processes may develop very slowly indeed. Thus we cannot involve them in the much more rapidly occurring loss of fertility that appears to be the case in E. pratense populations.

As a northern-montane or circumpolar boreal-montane species, E. pratense is also likely to find the current rather dramatic rapid climatic warming in NW Europe unfavourable to its continued survival in these islands, a situation leading to slow decline and eventual extinction as Chris Page (1982 & 1997) has predicted. Thus E. pratense in Britain & Ireland already demonstrates several of the features associated with genetically weakened species populations (reproductively inadequate, small, scattered, more or less isolated populations lacking recruitment, displaying poor competitive ability). When this is compounded with the unsettling effects of an unfavourable warming climate, plus the thinning effects of random mortality inevitable in all small populations, with regret the current writer is forced to agree with Page's depressing verdict.

Along with other arctic-alpine, circumboreal and northern-montane plants, E. pratense must be recognised as a relict species in our latitudes of past cooler environments, and as such it is doomed to local extinction, probably in the not very distant future (Briggs & Walters 1997, pp. 411-419).

British occurrence

In England, E. pratense is already completely confined to a few scattered stations in the northern Pennines, but it is much more widespread in Scotland, typically occurring as small patches in the lower valleys in the Highlands and islands, but occasionally found on more open upland moorland where drainage and spring water flushes provide nutrient enrichment of the turf (Jermy & Camus 1991; Stewart et al. 1994; Page 1997).

European and world occurrence

In Europe, E. pratense has a very pronounced circumpolar boreal-montane range that is distinctive from any other European horsetail. It ranges from Iceland, north to within the Arctic Circle in Scandinavia, and south to the C European mountains and the Caucasus. However, it is completely absent from both the French Alps and the Pyrenees (Jalas & Suominen 1972, Map 37; Jonsell et al. 2000, p. 21). The distribution then stretches eastwards across most of N Asia to Manchuria, N Japan and across N America, south to about 40^o N. It is absent, however, from Greenland and from much of W Europe (Hultén 1962, Map 83; Hultén & Fries 1986, Map 18; Jonsell et al. 2000).

Names

E. pratense is neither distinctive enough nor sufficiently common to have been given local English common names, 'Shady Horsetail' or 'Shade Horsetail' being mere invented book names. The Latin specific epithet, 'pratense', means 'growing in meadows' (Gilbert-Carter 1964), which fits the behaviour of the species in Scandinavia and undoubtedly in other northern parts of its range.

Threats

Both Fermanagh sites are protected; the Cladagh River Glen is a National Nature Reserve and part of the Magho cliffs is a Forest Nature Reserve. They could still be vulnerable to grazing or trampling.

Growth form and preferred habitats

The only horsetail with branched lateral branches, this feature, together with the graceful, fine textured, drooping, pale yellow-green stems and branches makes Wood Horsetail quite unmistakable and the most beautiful of our horsetails. Sterile and fertile stems are similarly green and branched, the latter differing only in that in May each year they bear a terminal, narrowly oval cone, about 10-20 mm long.

E. sylvaticum is a small to medium-sized, deciduous, colony-forming, rhizomatous horsetail of moderately wet, usually deep, acid to neutral, clay or peaty soils, which, since the species is calcicole, must be more or less constantly flushed with moderately calcium- or base-rich groundwater. In its preference for half-shade and the root-flushing requirement, Wood Horsetail, at its lowland sites at least, is rather similar in its ecological demands to E. telmateia (Great Horsetail) and, to a lesser extent, E. palustre (Marsh Horsetail), with both of which it frequently associates. The roots of both E. sylvaticum and E. telmateia have an absolute requirement for some degree of water movement, while E. palustre can tolerate almost no flow (Sinker et al. 1985; Page 1997).

While Wood Horsetail tends to be most prevalent in damp, sheltered, at least partially shaded sites with constant high humidity (for instance, in fen-carr, woodland margins, glades, hedgerow-, river-, stream- or ditch-banks), it can frequently be found in very much more open and exposed upland positions. In Fermanagh, examples of the latter habitat occur on the Drumbad Scarps, Poulaphouca cliffs and Topped Mountain. In these more exposed sites, as might be expected, it is both rather dwarfed and much more sparsely branched, which makes it quite difficult to recognise (eg see illustration in Page 1997, p. 470).

In Fermanagh, near the NW coast of Ireland, E. sylvaticum can also be found in open, more sunny conditions, as well as in more typical semi-shade. Open habitats here include stabilised scree, quarries, flushed heath and moorland, cut-over bog and even in churchyards. The important essential proviso is that sufficient moisture and the nutrient requirements of the species must be met. The upland tendency of E. sylvaticum on damp, peaty heath and moorland, as well as on cliffs and more sheltered gullies and stream-sides, together with its known woodland shade preferences, suggests the possibility that the upland plants of this horsetail might be relicts of former forest in these sites (Jermy et al. 1978). As with several pteridophytes of moist western heathland, Wood Horsetail appears to have swapped the shade, shelter and constant high humidity under woodland canopy, for an overall heathland region 'Atlantic' climate of overcast, grey, cloudy skies and regular, almost daily, precipitation, evenly spread throughout the year (Gimingham 1972, pp. 11-13; Page 1988, p. 288).

Fermanagh occurrence

E. sylvaticum is the fifth most frequent horsetail in the Fermanagh Flora Database, being represented in 147 tetrads, 27.8% of those in the VC. Eight tetrads contain pre-1976 records only, suggesting some loss of habitat. As the tetrad distribution map indicates, Wood Horsetail is very widespread, but especially frequent on the western limestones. Typical local habitats are wet woods, shaded meadows, flushed heath and moorland slopes, cliffs, streamside banks, damp, shaded ditches and roadsides.

British and Irish occurrence

In Britain, E. sylvaticum is common and widespread in the N & W, but decidedly rare in most areas of C, E and SW England (New Atlas). There has also been a marked decline in the species in lowland England and Wales, which has been going on since before the first BSBI Plant Atlas (Perring & Walters 1962) and continues to the present (C.Dixon & T.D. Dimes In: Preston et al. 2002).

In Ireland, Wood Horsetail is frequent and widespread in the north, but it is much more widely scattered and only occasional in the southern half of the island (Jermy et al. 1978; An Irish Flora 1996; New Atlas).

European and world occurrence

The north-western trend in the British & Irish distribution is mirrored in the horsetail’s European occurrence, covering almost the whole of N and C Europe south to N Spain, the Alps, N Greece, the Balkans and Turkey, but almost entirely absent from the Mediterranean coasts (Jalas & Suominen 1972, Map 36). The species also stretches, sometimes commonly and with very little taxonomic variation, across N Asia to Japan, and it completes its circumpolar range in the higher latitudes of N America, from Alaska to Labrador and on to W Greenland, Iceland and the Faeroes (Hultén 1962, Map 86; Hultén & Fries 1986, Map 17).

Names

Despite its very distinctive and beautiful appearance, E. sylvaticum has not in the past been recognised by most people as anything other than simply another horsetail and for English Common names it has really only got what we refer to as 'book names', such as 'Wood Horsetail' and 'Bottle-brush'. The latter name it shares with E. arvense and also with Hippuris vulgaris (Mare’s-tail) (Britten & Holland 1886). The Latin specific epithet 'sylvaticum' simply gives us the same information as the English common name, meaning 'of woodland' (Gilbert-Carter 1964).

Threats

None.

Growth form and preferred habitats

Marsh Horsetail is a characteristic rhizomatous perennial of wet to moist, moderately to slightly base-rich, often calcareous or dolomitic (ie magnesium rich limestone) habitats. It tends to occur as scattered individuals or clonal patches on the margins of small streams and ditches in marshy ground, including lakeshore fen-scrub woods, damp to wet meadows that seasonally flood and hedgerow ditches. In upland areas, Marsh Horsetail frequents calcareous flushes in acid, peaty, heather or grass dominated moorland. It grows beside lakeshores and pools at every altitude, including those on cut-over bogs, in disused quarries and gravel pits – always provided there is some lateral water movement and a degree of base-enrichment.

As its English common name suggests, Marsh Horsetail is more or less confined to these wetland habitats and in this respect it differs somewhat from its close relative, the much more weedy E. arvense (Field Horsetail), which is frequently found growing in relatively dry ground, including arable fields – if there were any of those around Fermanagh!

Plants of E. palustre vary greatly in the amount of branching they produce (see below), but they can readily be distinguished from the more frequently found E. arvense by the first internode of the side branches always being shorter than the adjacent stem sheath. Also, when the plant bears its black spore-producing cones, they are found on the tips of the slender green branches.

Like E. arvense, E. palustre readily colonises disturbed ground, and on marshy ground and in peatland situations it often becomes abundant after the original surface vegetation has been broken or trampled, for instance along tracks and beside streams, or along freshly dug boggy ditches. Borg (1971) has reported similar behaviour of E. palustre on Finnish bogs and fens.

In a detailed Dutch study of 1,000 agricultural fields, Marsh Horsetail was found to be more common on 29% of the fields surveyed that were regularly mown, than on those which had been under continuous grazing (8% of the fields surveyed). E. palustre was also much more frequent on soils low in potassium or phosphorus than when these nutrients were in good supply. In general, the study found Marsh Horsetail was a rare weed of well-drained, grazed pastures that were supplied with a manure rich in phosphorus and potassium (Sonneveld 1953 (in Dutch), quoted by Holm et al. 1977). Typically, Marsh Horsetail is a minor component of moderately fertile, grazed, herb-rich wet grassland conditions.

Variation

E. palustre is extremely sensitive to environmental change, varying markedly in size and form with degree of exposure. In the past, this variability led to numerous growth forms being named which are nowadays regarded as taxonomically insignificant. Plants in sheltered, somewhat shaded positions are much taller and are well furnished with whorls of upswept branches compared with the stunted and almost or entirely unbranched forms that occur in more open or windswept upland sites (Page 1988, 1997). These often very small, unbranched horsetails require careful identification, as sparsely scattered shoots of several horsetail species occasionally co-exist, eg E. arvense, E. palustre and E. variegatum (Variegated Horsetail).

Shade, trampling, spring frosts, or even air turbulence associated with road traffic can cause damage to the delicate growing point of E. palustre, causing the normally erect plant to become prostrate and bushy (Holm et al. 1977).

Spore production, dispersal and sexual reproduction

Dispersal in all horsetail species, including E. palustre, is surrounded by a degree of mystery, but we have to presume that in the case of transport to occupy a freshly created vegetation gap, it either involves wind dispersal of spores, their germination in a suitable soil and then subsequent delicate prothallial sexual process or, alternatively, vegetative spread. The sexual production of a new generation of the sporophyte plant is probably the least likely, or least frequent mechanism of increase and dispersal, since spore production and release is seasonal, the spores are short-lived and the many stages of the sex process require very specific environmental conditions. All stages of the sexual reproductive process involve high levels of risk and a failure of any of them negates the whole venture (Duckett & Duckett 1980).

Hybrids

E. palustre forms hybrids with E. arvense ( E. x rothmaleri C.N. Page) and E. fluviatile (E. x dycei C.N. Page). The former has not yet been recorded anywhere in Ireland although there are 30 records widely scattered in Britain (Stace et al. 2015). E. x dycei has been found in five sites across Ireland and in 33 well scattered sites, mainly in the N & W of Britain (Stace et al. 2015).

Vegetative reproduction

Vegetative reproduction or spread, on the other hand, may prove less risky and could be achieved in several ways, either by the fortuitous arrival of a tuber or rhizome fragment, perhaps transported by water, or in mud attached to animal or human traffic. Alternatively, if an already established clone exists nearby, E. palustre may spread laterally by means of normal rhizome growth, both horizontal and vertical, eventually to occupy the vegetation gap at the soil surface, a form of dispersal known as 'diffusion'. In addition, soil disturbance could fragment the hardy and resilient rhizome and attached tubers, allowing the possibility of transport to fresh suitable sites.

Underground stem or rhizome

The rhizome of Marsh Horsetail is remarkably robust and wide in diameter when compared with its slender aerial stems and while the individual rhizome lives only a few years, it produces buds at its nodes which allow it to branch and spread underground and then develop fresh aerial shoots. Occasionally the rhizome bears tuberous outgrowths. Individual plants in a favourable habitat are known to have spread over 100 m in a few years (Holm et al. 1997). The plant does not however tolerate firm, packed soils, and rhizomes have been known to travel considerable distances in soft subsoil to avoid compacted upper horizons before the photosynthetic shoots re-emerge in more porous substrates (Borg 1971).

A pioneer colonist, competitive ability poor

It appears that Marsh Horsetail only competes effectively if its optimal growing conditions are met. As a result, it frequently behaves as a pioneer colonist, being abundant in suitable open, disturbed, muddy habitats in the early more open stages, but it declines thereafter as other species arrive, compete and become dominant. However, E. palustre is usually able to survive as a scarce, often diminutive companion species, provided the potential dominants are restricted in their vigour by grazing, cutting or some other destabilising environmental factor(s). Marsh Horsetail has a rather sparse canopy of slender shoots and this, together with its comparatively poor competitive ability, means it is not normally a dominant species in any plant community in which it occurs (Borg 1971; Grime et al. 1988).

While E. palustre occurs in a very wide range of wetlands, yet in all the vegetation communities it frequents, it remains a minor component. It is difficult to typify its role and status beyond that of a pioneer colonist of gaps and bare ground associated with disturbance. It may be reasoned that since the main part of the horsetail species grows well below ground level, it does not really belong to the same vegetation stratum as most of the more shallow-rooted species with which it associates. This revelation helps explain the wide range of vegetation types in which Marsh Horsetail occurs, albeit in many instances having a very low presence in them, and its relationship with associated species and vegetation communities are not firmly set. This matter is further complicated, since the rhizome of E. palustre confers exceptional persistence on the species, even in a changing environment. Possessing a vigorous rhizome allows the horsetail to survive and effectively ignore dynamic vegetation modifications happening around and above it, so that the species may become a relict of past plant communities in a particular spot of ground (Borg 1971).

Plant associates and communities

Having said this, we can identify some of its more regular associates in particular habitats. For instance, in moderately fertile marsh grasslands, the common associates of E. palustre include Filipendula ulmaria (Meadowsweet), Caltha palustris (Marsh Marigold), Geum rivale (Water Avens), Lycopus europaeus (Gypsywort), Angelica sylvestris (Wild Angelica) and Cirsium palustre (Marsh Thistle), together with numerous species of rush and sedge. In some of the marshy grasslands, swampy fens, woods and heathland, E. palustre overlaps and associates with E. arvense and E. fluviatile (Water Horsetail), and especially in hedgerow ditches, it can associate with E. sylvaticum (Wood Horsetail) and E. telmateia (Great Horsetail).

In more upland acid grassy blanket peat moorland, E. palustre frequently occurs in sloping flushes and beside small streams where the flow of base-rich spring water enhances both the nutrient supply and the aeration of thinner layers of peaty or silty soils. Here, and around moorland lakes where there is similar spring water enrichment, for instance in Fermanagh, around Spectacle Lough and Drumcose Lough, Marsh Horsetail occurs with species such as Potentilla palustris (Marsh Cinquefoil), Galium uliginosum (Fen Bedstraw), Eriophorum latifolium (Broad-leaved Cottongrass), Parnassia palustris (Grass-of-Parnassus), Schoenus nigricans (Black Bog-rush or Black Beak-sedge), Pinguicula vulgaris (Common Butterwort), Carex dioica (Dioecious Sedge), Selaginella selaginoides (Lesser Clubmoss), Cirsium dissectum (Meadow Thistle) and, occasionally, E. variegatum (Variegated Horsetail). In a similar flush, but at lower altitude below the cliffs of Poulaphouca, close to the shore of Lower Lough Erne, E. palustre has again been found growing along with E. variegatum.

Fermanagh occurrence

In Fermanagh, it is the third most frequent horsetail being found in 248 post-1975 tetrads, 47.0% of those in the VC. It is especially frequent around Upper and Lower Lough Erne and on the upland limestone, bog and heath in the west of the county.

British & Irish occurrence

A common and widespread deciduous species throughout Britain & Ireland, E. palustre is perhaps somewhat less common in the S and W of Ireland due to the prevalence in this part of the country of heavily leached, base-poor, excessively acid soils (Jermy et al. 1978).

European and world occurrence

In Europe and beyond, its widespread distribution closely matches that of both E. arvense and E. fluviatile, extending from Iceland to almost all of Europe southwards to Spain, Italy, Greece and Turkey (Jalas & Suominen 1972, Map 35). The world distribution stretches eastwards through much of N Asia to almost encircle the arctic, although there is a distinct gap in N America in the central provinces of Canada (Hultén 1962, Map 89; Hultén & Fries 1986, Map 16).

Toxicity

In common with other species of Equisetum, E. palustre contains high levels of silica, plus the thiamine-destroying enzyme thiaminase, and the toxic alkaloid(s), palustrine and/or equisetine. When animals are allowed to graze selectively they avoid horsetails. However, in cut and stored hay, there is little way they can do so, and the toxins survive drying and storage. Tests in Finland showed that as little as 2 gm of dried horsetail in the daily fodder of cattle caused their milk yield to decline. Larger amounts of horsetail caused lack of appetite, diarrhoea and general illness of the animals (Borg 1971). Horses are not as sensitive as cattle, but if fed small amounts over long periods, they can also suffer serious poisoning (Holm et al. 1977).

Weed control

In spite of numerous and prolonged trials, a really effective method of controlling the growth of E. palustre has not yet been developed. The aerial shoots are easily destroyed by any of several herbicides, but the deep, wide-ranging rhizome survives and persists, and its starch reserves simply cannot be exhausted by attacking the aerial parts of the colony. The systemic herbicide MCPA gives the best penetration into the rhizome, and if low rates of application are used, the aerial shoots are not destroyed as quickly as at higher rates. This allows more time for the herbicide to be translocated downwards into the rhizome. Using this MCPA herbicide procedure in conjunction with both sub-surface drainage and intensive arable cultivation, rather than allowing prolonged grass leys to develop, appears to be the best available agricultural treatment of the Marsh Horsetail problem in pastures (Borg 1971).

Names

The Latin specific epithet 'palustre' is derived from 'palus', meaning a swamp or bog, and translates as 'growing in swampy places' (Gilbert-Carter 1964). E. palustre shares some of the English common names applied to horsetail species in general, and they are? similar to those applied to the more common species E. arvense. Example include Cat-whistles, Marshweed, Paddock Pipes and Snake Pipes (Britten & Holland 1886).

Recommended further reading

The detailed review of the biology and ecology of E. palustre published in English by the Finn, Pekka J.V. Borg (1971) is highly recommended reading for anyone requiring further information and speculation on the properties and behaviour of this species.

Threats:

Drainage and other agricultural improvements and acidification associated with forestry plantation.

Growth form and preferred habitats

This imposing rhizomatous species always grows in colonies and the large size of the tall, sterile, annual, deciduous shoots, their abundant lateral branches and ivory-white stems make Great Horsetail both conspicuous and completely unmistakable. E. telmateia is closely tied to permanently wet soil, typically either poorly drained clay or shallow peat, which is more or less constantly flushed with base-rich, generally calcareous ground water springs or seepage emerging from overlying, more permeable sedimentary rocks (Brewis et al. 1996; Page 1997). It often grows along the edge of roadside hedges and ditches, on eroding river banks, or at the base of wooded cliffs where it is shaded or semi-shaded by overhanging tree and shrub branches. It is associated with man and some habitats he creates, and therefore is apophytic on clayey slopes of roadside banks, railway cuttings and embankments, provided there is the necessary degree of base-rich water seepage through the soil (Jonsell et al. 2000).

The ecological requirements and tolerances of Great Horsetail overlap quite closely with those of several other horsetails, since it commonly associates with E. palustre (Marsh Horsetail), less frequently with E. fluviatile (Water Horsetail) and rarely with E. hyemale (Rough Horsetail).

E. telmateia is essentially a lowland plant, unable to tolerate upland exposure and rarely found over 250 m, but the requirement for moving groundwater rich in calcium or other bases, together with the laws of gravity, confine it both geographically and geologically in any case.

Plant structure and spore production

The thick, fleshy-looking, white sterile stems of Great Horsetail can reach 2 m or more in height and are heavily furnished with whorls of long lateral branches. Despite their robust appearance, the tall, stout sterile stems are in fact very brittle, making the species intolerant of any trampling or grazing whatsoever (Sinker et al. 1985).

As in the case of E. arvense (Field Horsetail), separate, whitish-brown, unbranched fertile stems up to about 20 cm in height appear before the photosynthetic sterile stems in April and May. The cone-like fertile spikes are produced most abundantly on the drier, warmer, marginal areas of the typical habitat (Jonsell et al. 2000). The cone-bearing shoots are very ephemeral, typically sporing, withering and dying off within two weeks of their initial production (Jermy & Camus 1991; Page 1997).

The deeply running branched rhizomes of Great Horsetail are about 1.0 cm in diameter, have large air canals embedded in their conducting tissues, similar to those in E. palustre, and they bear numerous starch storage tubers at their nodes (Page 1997).

Vegetative and sexual reproduction

As with all other horsetails, the main form of reproduction is probably vegetative by means of the rhizome tubers and fragmentation of other parts of the plant. Since E. telmateia forms a number of rare or very rare hybrids with other Equisetum species, some successful spore germination and prothalial sexual reproduction must also occasionally take place (Page 1997).

Hybrids

E. telmateia is capable of forming hybrids, and does so very rarely with E. fluviatile (one record from Ireland), E. palustre (2 hexads in Co Sligo), E. sylvaticum (Wood Horsetail) (no Irish records) and E. arvense (again, no Irish records) (Stace et al. 2015).

Fermanagh occurrence

E. telmateia is the fourth most frequent and widespread horsetail in Fermanagh, occurring in 171 tetrads, 32.4% of those in the VC. Six tetrads contain pre-1976 records only. As the tetrad distribution map shows, it is widely distributed, especially in the S & W of the county. The most extensive stands in Fermanagh are along the shore road of Lower Lough Erne, below the Cliffs of Magho (also known as Poulaphouca). However, it occurs in quite a wide variety of habitats chiefly in limestone areas, including lakeshores, river and stream banks (especially in gullies), cliff bases, below waterfalls, in quarries and even along railway lines disused for over 50 years.

Toxins and uses

Like other horsetails, stands of E. telmateia contain large quantities of silicates and a variety of toxic principles, including alkaloids and an enzyme called 'thiaminase' that destroys thiamine (vitamin B) (Cooper & Johnson 1998). The plants are usually avoided or ignored by grazing horses, but owing to the smoothness and fleshy softness of its stems, it is said to be acceptable to cattle as food. The 18th century Swedish botanist Linnaeus commented that in N Sweden E. telmateia was cut and given to cows for fodder and that in Lapland reindeer would eat it, although horses always refused to do so (Grieve 1931).

There are reports of young stems of Great Horsetail being eaten like asparagus from Roman times onward, particularly by the starving poor. Unfortunately, as a vegetable they are neither palatable nor very nutritious and like other horsetails they undoubtedly contain toxins (Grieve 1931).

British and Irish occurrence

E. telmateia is frequent to locally abundant throughout most of the British Isles, although it is more occasional in the southern half of Ireland and is rare or absent in many parts of C, E and N Scotland (Jermy et al. 1978; Stace 1997; New Atlas).

European and world occurrence

Great Horsetail reaches the most northernly point of its European range at Caithness in Scotland (VC 109), and the European distribution extends southwards through W, C and S Europe to N Africa, Turkey, the Caucasus and eastwards to Iran. In the Pacific States of N America, it is represented by a separate subspecies, subsp. braunii (Milde) Hauke (Jalas & Suominen 1972, Map 39; Hultén 1958, Map 258; Hultén & Fries 1986, Map 20; Jonsell et al. 2000).

Names

The Latin specific epithet, 'telmateia' means 'of marshes', or 'of muddy waters' (Gledhill 1985). The few English common names that are specific to this species include 'Fox-tailed Asparagus' and 'Horse Pipes'. The former obviously refers to the similar appearance of the spikes of the young plants to a foxtail and the previously mentioned fact that both have been eaten by humans. This name first appeared in notes made by Lyte when he was preparing his Newe Herball (Lyte 1578), based on his translation of the French edition of Dodoens' Dutch language Cruydeboeck, that was originally published in Antwerp in 1554 (Britten & Holland 1886; Anderson 1977).

Threats

None.

Growth form and preferred habitats

As this fern is small and generally occurs as thinly scattered individuals, it is very easily overlooked. It is therefore hard to be certain despite our efforts that O. vulgatum is fully recorded in the county. It needs to be deliberately searched for from late April to September when the annually produced aerial parts of this perennial species are visible. The texture of the sterile frond and the fact that it has no mid-rib, together make it fairly easy to spot once you know the sort of moist, usually calcareous or lime-enriched loamy, clay soil it favours, and 'you've got your eye in for it'!

It does, in fact, occur in a wide variety of calcareous or mildly acidic, moist, unimproved grassland and rocky habitats, including calcareous lakeshores, limestone pavement, scree and sometimes in scrub woodland when it is invading such sites. In our experience, however, it is never really abundant, even on or around grassy turloughs' floors (ie vanishing limestone lakes which drain through their base), which ought to suit its preferences perfectly (Webb & Scannell 1983; Page 1997).

Fermanagh occurrence

In Fermanagh, this little, deciduous, rhizomatous, taxonomically polymorphic fern has been recorded in a total of 64 tetrads, 12.1% of those in the VC. As the distribution map indicates, Adder's-tongue is widely distributed in Fermanagh, mainly to the SW of Lough Erne, but with a few scattered stations further east. The most isolated easterly recent station is on a roadside at Knocknalosset Td, where RHN found it in June 1990.

Vegetative Reproduction

Adder's-tongue is capable of vegetative reproduction, producing shoot buds on its spreading roots, so that patches of the plant may develop over time. Having said this, like Botrychium lunaria (Moonwort), another fern species closely dependant on a mycorrhizal fungal partner in the soil, O. vulgatum population densities fluctuate from year to year, apparently being very sensitive to prevailing local climate (Page 1997).

Irish occurrence

The second edition of the Census Catalogue of the Flora of Ireland indicates that O. vulgatum has been recorded at least once in all 40 Irish VCs (Scannell & Synnott 1987). Careful manual inspection of the Botanical Society Atlas (Perring & Walters 1962, 1987) shows that in many Irish VCs this is literally the case, there being a solitary plotted symbol. On the hectad map of Ireland, 71 of the symbols represent pre-1930 records. While the most recent Irish Flora continues to describe this easily overlooked fern as 'occasional' (Parnell & Curtis 2012), in reality Adder's-tongue is much more scarce and scattered, and in some areas it is in long term decline. On the other hand, Irish field recording has greatly improved over the last 50 years, and the New Atlas hectad map plots 110 Irish hectads with post-1987 records, a high proportion of these being in the northern province of Ulster (Preston et al. 2002).

Population decline

There is evidence of a long term decline of O. vulgatum populations in lowland areas of Britain & Ireland and the trend appears particularly obvious in Ireland. The species losses are widely attributed to drainage, agricultural intensification, a changing grazing pattern and the move from hay to silage production (C. Jermy, in: Preston et al. 2002). At the same time, O. vulgatum seems to be holding its ground in the areas of Fermanagh where agriculture is less intensive, while undoubtedly it is now less frequent in the N & E of the VC, where drainage and disturbance associated with pasture improvement will probably eventually eliminate it entirely. Even in areas 'improved' in this manner, however, the fern may manage to hold on in adjacent, more rocky marginal ground. The decrease in the number of unimproved limestone or lime-enriched pasture sites in Fermanagh is very real, however, a rough measure being given by the fact that there are 15 tetrads where the species was recorded prior to 1976, but has not been re-recorded. This compares with 49 tetrads which have post-1975 records of the fern.

The decline of this species in Ireland appears obvious from the map in The Fern Atlas (Jermy et al. 1978, p. 28), and this has also been commented upon in other modern Irish VC Floras, notably that of Co Dublin (H21), and of the three VCs in NE Ireland (Cos Down, Antrim and Londonderry (H38, H39 & H40) (Doogue et al. 1998; Hackney et al. 1992). However, two recently published Floras from the far south of the island indicate that in the areas covered, Adder's-tongue has either been under-recorded in the past, or has recently increased its presence: in Co Waterford (H6), Green (2008) lists the species from eight sites post-1996 and in Co Limerick (H8), Reynolds (2013) lists six sites with modern records.

British occurrence

O. vulgatum appears much more frequent and widespread in lowland England and S Wales than is the case in Ireland. The New Atlas map, displaying plant data recorded up to the end of 1999, indicates that there has been a considerable improvement in the recording of this fern in Britain over the last 40 years in comparison with the original Botanical Society Atlas (Perring & Walters 1962). However, despite the enhanced recording effort, the presence of O. vulgatum thins and becomes sparse and declining to the N of a line between Morecambe and Middlesbrough (Preston et al 2002; Wardlaw & Leonard 2005). Drainage and the widespread agricultural use of fertiliser and slurry-spreading in particular, have undoubtedly accelerated the previously gradual process of Adder's-tongue population decline, a process which in Britain has been operating for more than three centuries (Page 1997).

European and world occurrence

O. vulgatum is widespread in W & C Europe, becoming more coastal in Scandinavia and very scattered in the Iberian Peninsula and along the N Mediterranean coast (Jalas & Suominen 1972, Map 42). Elsewhere, the species (considered broadly) extends in an extremely widely-spaced, disjointed manner across boreal Asia to Japan, Burma and N America as a very disjunct circumpolar temperate species. It also occurs further south in equatorial W Africa and in Mexico (Hultén 1962, Map 91; Hultén & Fries 1986, Map 21).

Folklore and uses

In earlier centuries Adder's-tongue, like Moonwort, was very much more plentiful in the British Isles, and it was better known than it is today so that it developed a folklore reputation. It was regarded both as a herbal antidote for snakebite from Adders (another case of 'the Doctrine of Signatures' based on a fanciful resemblance), and as having a malevolent influence on the health of pasture grasses (Step & Jackson 1945; Page 1997). An ointment called 'Green Oil of Charity' was prepared using the fern and was still in demand as a salve for wounds in the 1940s in parts of Britain (Grieve 1931; Step & Jackson 1945). The expressed juice of the leaves, drunk either alone or with distilled water of Horsetail, was once widely used by country people for internal wounds and bruises, vomiting or bleeding at the mouth or nose (Grieve 1931). On the other hand, doubts as to the authenticity of the reports of folk use of this small, rather scarce native are expressed by Allen & Hatfield (2004), who suspect the possibility of imports of plant material and possible 'borrowings' of reported use from older herbals.

Names

The genus name 'Ophioglossum' is a combination of two Greek words, 'ophis' = a serpent and 'glossa' = a tongue, said to be because of the appearance of the fertile branch, although in reality no snake's tongue has any likeness to it. The Latin specific epithet 'vulgatum' means common or ordinary (Step & Jackson 1945).

Additional English common names locally applied to O. vulgatum (and sometimes also to other unrelated species), include 'Edder's-tongue', 'Serpent's-tongue', 'Dragon's-tongue', 'Adder's Spear', 'Adder's-grass' and 'Cock's-comb' (Britten & Holland 1886).

Threats

Disturbance of older pastures, including drainage, agrichemical application and grazing pressure from sheep and rabbits would all tend to depress the frequency of this small fern.

Growth form and preferred habitats

Moonwort is a small, rather fleshy, hairless, perennial but deciduous fern that typically grows on near-neutral to strongly basic soils. On ericaceous heaths, B. lunaria is limited to ground which contains at least a little lime and where the dominant subshrubs, Calluna and Erica are absent. B. lunaria usually occurs as scattered individual shoots in short, grassy turf. It is a small plant, the sterile frond rarely being more than 15 cm in height and often only half or one third of this figure, so it is very easily overlooked. The sterile fronds appear in early May, each followed by the adjacent but separate fertile frond. When the latter is eventually fully developed in mid-June, it is somewhat taller than the sterile branch and bears a pinnately divided, triangular spike of globular orange-brown spore-sacs, aptly likened to a miniature bunch of grapes (Page 1997). This is alluded to in the genus name, see below.

Fermanagh occurrence

The local Victorian field botanists West, Abraham and McCullagh all saw this little fern in Fermanagh around the 1900s, as did Meikle and his co-workers in the 1940s. Yet despite all the surveys of the last 30 years, Moonwort has only been seen on four occasions recently, a fact that is difficult to understand as the vegetation at the plant's original sites at Belmore Mountain (recorded in 1946) and Drumkeenagh (recorded in 1902), have probably not changed much during the whole of the 20th century (Revised Typescript Flora, Meikle et al. 1975). The map shows records occurring in seven scattered tetrads, four of them with post-1975 dates. None of the records give information on the species degree of presence, but we recollect that the Legacurragh discovery was a single very small plant, and the latest late-June 2013 record at Doagh Lough was a solitary plant just 3 cm in height.

Details of the other seven Fermanagh records are: Drumkeenagh, near Black River, 1902, J.T. Abraham & F.R. McCullagh; Belmore Mountain, July 1946, MCM & D; scree slope above Doagh Lough, 1947, MCM & D; Callow, Monawilkin, 11 July 1985, ASSI Team, DOENI; Legacurragh, above Florencecourt, 1991, M. Tickner; Isle Namanfin, Lower Lough Erne, 1990-5, D. Hughes; limestone hill south of Dough Lough, 27 June 2013, H. Northridge.

Irish occurrence

In Ireland, while B. lunaria has been found at least once in all but two of the 40 VCs (Scannell & Synnott 1987), it appears to occur only rarely and fleetingly in small numbers as a casual species on dry heathy grassland, mountain ledges and in old lawns or sand-dunes (Webb & Scannell 1983; Webb et al. 1996). The New Atlas hectad map indicates how very scarce B. lunaria has become in Ireland: the map displays scattered modern records in just 14 VCs (28 hectads), mainly in the north of the island.

It could be that unlike earlier generations, modern field botanists do not have an eye for the plant, but more likely it really has declined or become even more transient than previously was the case. RSF is very familiar with Moonwort from continental field trips, but he has never found an original site for the fern in Fermanagh, which does suggest genuine rarity.

In Co Dublin (H21), B. lunaria was already a rare species at the turn of the 19th century, whereas formerly it was more common in upland, usually base-poor pastures. There are only three modern records in Co Dublin: one site with a solitary plant and another with only four individuals. Both the mentioned sites are in unimproved pasture, but interestingly, the third Co Dublin site is from an unusual habitat, in woodland beside a reservoir (Doogue et al. 1998).

British occurrence

In Britain, Moonwort is widely scattered, but while it is predominantly a northern and upland species, it also occurs in a wide range of open, exposed, short-grassland habitats, including at low altitudes sand-dunes, golf-links, old unimproved meadows (now an exceedingly rare habitat), grassy banks and downs. Upland habitats in Britain include grassy moors and heaths, old stabilised grassy screes, alpine meadows and pastures, as well as on cliff ledges (Page 1997).

B. lunaria has suffered a gradual decline in Britain as well as in Ireland as evidenced, for instance, by the map in the Atlas of Ferns (Jermy et al. 1978), where the number of pre-1930 stations recorded almost matches those more recently recorded. The New Atlas hectad map confirms the continuing decline of Moonwort sites in Britain, although often being very small and either solitary or in small populations, it could well be overlooked and under-recorded (A.C. Jermy, in: Preston et al. 2002).

One reason why B. lunaria might fare less well in Irish conditions may be the tendency for it to occur in conditions of high rainfall on very shallow peaty soils formed over limestone. English experience, on the other hand, suggests growth of the fern really benefits from deep, well-aerated soils which allow adequate scope for its deeply running root system and which presumably facilitate the success of the associated mycorrhizal fungi (Page 1997).

European and world occurrence

Beyond our shores B. lunaria is widespread in N, C & S Europe, extending well into the arctic region and as far south as Sicily and the Greek Peloponnese. The distribution peters out in much of W France and in the southern parts of the Iberian Peninsula (Jalas & Suominen 1972, Map 44).

B. lunaria, taken in a broad taxonomic sense, extends in a circumpolar temperate manner around northern latitudes through Asia to Japan, and across northern N America from Alaska to Labrador and on to S Greenland and Iceland. A typical form of B. lunaria occurs also in southern S America along with closely related varieties, and it is also found in SE Australia, Tasmania and New Zealand (Hultén 1962, Map 103; Hultén & Fries 1986, Map 23).

Folklore and medicinal use

Moonwort has long had a reputation for being sporadic in making its appearances and disappearances and, perhaps partly for this reason, it has gathered a considerable body of folklore around it. Magical properties include the ability to unlock locks and unshoe horses' feet. The former notion perhaps derived from the key-like outline of the sterile frond, the latter maybe somehow too associated with its power over iron locks (Step & Jackson 1945; Page 1997). Other well known myths describe Moonwort's powers of alchemy and witches collecting it by moonlight for their magic spells.

In herbal medicine, it was said to heal wounds like Ophioglossum vulgatum (Adder's-tongue) (Grieve 1931). The significant weight of folklore that exists suggests that this small fern was once much more prevalent than now. It was widely known by country people, compared to the few who might recognise it today (Page 1988, 1997).

Names

'Botrychium', which is a diminutive based on the Greek 'botrys', meaning 'a bunch of grapes' (Gilbert-Carter 1964). The specific epithet 'lunaria' is derived from the Latin 'luna' meaning the moon, the curved individual lobes of the sterile frond being likened to the crescent moon (Gilbert-Carter 1964; see also figure in Arber 1970, p. 257).

'Moonwort' is the general widely accepted English common name of the fern, a translation of its specific epithet, but also linked to the magical folklore attached to the plant. Other common names include 'Lunary' or 'Lunarie' (Turner 1568), 'Unshoe the Horse' (Culpeper 1653) and 'Shoeless Horse', another reminder of the folklore (Britten & Holland 1886).

Threats

Improvement of upland pastures for agriculture.

Growth form and preferred habitats

This large, robust, rhizomatous, clump-forming fern with its tall, broad, erect, bipinnately divided, distinctive light-green sterile fronds up to 3 or even 4 m tall also produces separate, smaller, fertile fronds with rusty spike-like clusters of sporangia. O. regalis is a very conspicuous and unmistakable calcifuge species of fens, bogs, lakeshores, streamsides and ditches. While Royal Fern is typically a species of wet, acid, peat conditions, when necessary it can tolerate the considerable base-enrichment associated with salt-laden coastal winds on sea cliffs and the like (Jermy et al. 1978; Page 1997).

Ecology and longevity

In spite of the large robust nature of the sterile fronds and their leathery appearance, the aerial parts shrivel and die-back at the first serious touch of frost (Step & Jackson 1945). While in nature it is relatively slow growing, Royal Fern is probably one of the longest lived native ferns in Britain & Ireland. Large individuals in garden cultivation are known to be over 100 years old, and their thick, erect, rhizome bases display no signs of decline. Comparisons of massive wild ferns, based on the size of such garden specimens, suggest some individuals must be many centuries old (Page 1997).

Fermanagh occurrence

The tetrad map of its Fermanagh occurrence gives an over-optimistic picture of Osmunda's frequency in the VC. While it appears to be widely distributed in the S & W of the county and scattered elsewhere, at many of the 70 post-1975 tetrads there are only one or two plants recorded. Furthermore, the high frequency of pre-1975 sites where the plant has not recently been seen (20 tetrads on the map), suggests that either there has been a decline of the species in the area, or it was merely casual in its occurrence at these sites.

British and Irish occurrence

It has a mainly, but not exclusively, western and southern distribution in Britain & Ireland, although it is more coastal in Britain and more definitely western in Ireland (Jermy et al. 1978; New Atlas).

From at least the time of the Victorian fern craze in the 1850's, plants of this fern have been collected for garden use, both as a decorative subject in itself, and because the fibres of its large leaf bases provide an ideal horticultural medium for growing orchids (Allen 1969). We have no evidence of this wholesale collecting ever happening in Fermanagh, or if it has, it is at an insignificant frequency compared with some places in England, for instance in Westmorland where the species has been depleted by collecting, almost to extinction (Halliday 1997).

As a result of the widespread garden cultivation of the fern, particularly in Britain, it is not a simple matter to separate native occurrences from naturalised garden escapes or discards, a point which must be borne in mind when studying the detail in distribution maps of the species (Jermy et al. 1978; New Atlas).

O. regalis has undoubtedly declined considerably in N Ireland due to habitat loss associated with drainage and, in the E of the province, from development, for instance in bogs around Donaghadee in Co Down (Hackney et al. 1992). Apart from the Lough Neagh basin, the Fermanagh records probably constitute the most easterly large concentration of sites of the fern in Ireland. However, the species is very much associated with areas of high rainfall, and thus it is decidedly western in its distribution anyway.

European and world occurrence

Royal Fern being very variable and polymorphic, on the world scale it is best examined in the broad sense as an aggregate species. In Europe, O. regalis has its main area of distribution in the W and S of the continent, from the S Swedish coast to the Azores and it ranges discontinuously across the Mediterranean to Crete, Turkey and the Caucasus (Jales & Suominen 1972, Map 50; Page 1997).

On a world basis, Hultén (1958), and again in Hultén & Fries (1986 Map 29), map this polymorphic species along with three varieties they recognise, so that in total it stretches in a decidedly disjunct manner from Europe and C & S Africa to the Cape, a small pocket in N India and the Himalaya, to China, Burma and Japan, to eastern N America and parts of S America.

Names

The genus name 'Osmunda' is derived either from 'Osmund the Waterman' (an English Common name given by Lyte (1578), since it is a fern of bogs and streamsides), or from the Anglo-Saxon equivalent of the name 'Thor', the Scandinavian god of thunder (Grigson 1974; Gledhill 1985). The specific epithet 'regalis', Latin meaning 'kingly' or 'royal', apparently refers to the dignified and impressive appearance of the plant, and its great longevity (Gilbert-Carter 1964).

The plant has a long list of local English Common names, some of which date back to Anglo-Saxon, others of the 16th century. One of the most interesting is 'Herb Christopher' or 'St Christopher's Herb', which appears in Lyte (1578), Gerard (1597) and later authors. These names allude to the waterside habitat of the fern which the saint frequented and where, before he was converted, he exercised his self-imposed task of carrying people across fords. A very odd name is 'Bog Onion', from Cumberland, the explanation of which entirely defeats the current writer (Britten & Holland 1886; Vickery 1985).

Uses

The root or rhizome of O. regalis has been used not only as a potting compost as mentioned above, but also medicinally. The supposed curative powers are attributed to the salts of lime and potash, amongst others, which it obtains from the bog soil and water in which the fern grows. It was prescribed by herbalists for treating jaundice in its early stages, or for removing alimentary obstructions. An ointment made from the root was also recommended for healing wounds, including bruises, dislocations and lumbago (Grieve 1931).

Threats

Mechanical peat cutting on lowland raised bogs is the main threat, in many of the sites.

Growth form and preferred habitats

This small, clump-forming, deciduous, polymorphic, montane fern with its distinctive, finely cut sterile fronds and separate, narrower fertile branches is very much rarer in Ireland than in Britain. In both islands, it is strictly confined to calcium-free, silica-rich, acid soils. The species is so sensitive to bases that it is absent from otherwise suitable geological areas, if such habitats are subject to any enrichment with base-rich cations, for instance from salt-laden onshore winds (Jermy et al. 1978).

C. crispa prefers well-drained sites on steep, but relatively stable, screes, but it also occupies crevices on cliff ledges. In some areas it can be found on artificial habitats, namely mortar-free dry stone walls (Page 1997; T.D. Dines, in: Preston et al. 2002).

Fermanagh occurrence

The solitary mid-19th century Fermanagh record of John McDonald (of whom we know absolutely nothing at present), from "the East side of Caulteach (Cuilcagh) Mountain, near Florencecourt; sparingly", is from a very suitable site well furnished with acidic rock cliffs and steep screes (Cybele Hibernica 1866). Regrettably, this appears to have been one of the fleeting, casual occurrences that appear to typify the behaviour of this very strict calcifuge species in Ireland. No one has seen the fern in Fermanagh since, despite diligent searching for it on the appropriate parts of this very remote mountain area.

Interestingly, another fleeting occurrence of the species was recorded in 1937 from adjacent Co Cavan (H30): "a solitary tuft of the plant found growing out of a crack in the overhanging face of a small boulder on the W side of Bruse Hill at about 183 m altitude" (Cole 1938). It was reported as being killed the following year, probably by drought (Praeger 1946; Reilly 2001). On account of wider geographical distribution and biodiversity aims, this fern is included on the NIEA list of Priority Species of special concern requiring conservation action.

Irish occurrence

The presence of C. crispa in Ireland has declined during the 20th century from ten VCs to just three at present. Recent sightings are mentioned in the Irish Red Data Book, from one site in W Galway (H16), three in Down (H38) and two in Co Antrim (H39) (Curtis & McGough 1988). Plants in N Irish sites tend to be small and inconspicuous, growing in rock crevices on cliffs (Hackney et al. 1992). Flora of Connemara and the Burren noted that the fern is, "very rare in Ireland, and in some of its stations little more than a casual" (Webb & Scannell 1983).

British occurrence

In Britain, C. crispa has a strongly marked northern and western distribution, and its headquarters very obviously occurs in the Scottish Highlands. Losses in England occurred in the Southern Pennines prior to 1930, perhaps as a result of fern collecting, but the distribution now appears stable (New Atlas).

European and world occurrence

This variable, polymorphic fern species has several named lower taxa with separate distribution areas. In Europe, it has an Arctic-alpine distribution, the main areas of occurrence being the mountains of W Scandinavia, the British Isles, the Alps and the Pyrenees, with scattered stations south within the Iberian Peninsula, Corsica, N Italy and the Macedonian mountains (Jalas & Suominen 1972, Map 61; Page 1997).

In more continental areas of Europe, C. crispa tends to be found in the upper subalpine to mid-alpine belt in sites with reliable blanketing winter snow-cover (Jonsell et al. 2000). Indeed, while Parsley Fern evades severe frost in this manner, the species also chooses sites that allow it to avoid high summer maximum temperatures. The present-day distribution of C. crispa in Ireland and the Scottish Highlands correlates closely with the 24°C maximum summer temperature summit isotherm (ie for the highest places in the landscape). In the Scottish Lowlands, England and Wales the equivalent temperature limit is the 26°C isotherm (Conolly & Dahl 1970).

Beyond Europe, C. crispa, in the broad sense, extends to Asia Minor and there are at least four geographically differentiated races which carry the plant to the Himalaya, China, Alaska and N America (Hultén 1958; Hultén & Fries 1986, Map 32).

Names

The genus name, 'Cryptogramma', is derived from two Greek words, 'kruptos' = 'hidden', and 'gramme' = 'a line', an allusion to the fact that the lines of the sori are not as evident as on most other fern species, being covered by the rolled frond margin (Step & Jackson 1945; Jonsell et al. 2000). The specific epithet 'crispa', meaning 'curly', or 'with wavy margins', refers to the Parsley-like appearance of the deeply cut sterile fronds (Gilbert-Carter 1964).

The English Common name most commonly used is 'Parsley Fern', which Grigson (1974) reckons dates from the 18th century, the plant resembling curled forms of parsley. Other Common names less often heard include 'Rock Brakes', and 'Curled Brakes'; 'Brakes', 'Brake' or 'Brake-fern' are general names for the larger ferns dating back to Turner, Lyte and Gerard in the mid-16th century (Britten & Holland 1886). 'Stone Fir' and 'Mountain Parsley' are two 19th century 'book names' (ie, a derogatory term for plant names invented by other authors), also listed by Britten & Holland (1886).

Threats

None.

Growth form and preferred habitats

This is a small to moderate sized, delicate-looking, rhizomatous fern with fronds divided into numerous small, fan-shaped segments borne on thin, blackish, wire-like branches. It is frost-tender and a rare or very rare species mainly of damp, mild, limestone coastal areas of W Ireland and SW Britain. In very sheltered sites fronds are sometimes semi-evergreen, surviving overwinter. Native populations are threatened by unthinking collectors.

Apart from its scattered natural sites, Maidenhair Fern is frequently cultivated (mainly indoors), and can sometimes be found 'escaped' on sheltered, damp, lime-rich mortar on garden walls near old greenhouses.

Variation

Plants are rather variable in size, degree of frond dissection and margin serration, so that extreme forms have been brought into cultivation and have become named as horticultural cultivars (Page 1997).

Fermanagh occurrence

Praeger's original Fermanagh station was given as, "On outer side of wall opposite main door of Crom Castle" (Praeger 1939). It continues to thrive on the boat house at Crom, approximately 100 m from the old castle, and it is also abundant on the garden wall of Florencecourt against which a greenhouse once stood. These naturalised plants are the only recognised records of this fern in N Ireland, several 19th century finds in coastal stations in Down (H38) and Co Antrim (H39) having been discarded as either errors or unconfirmed and, in any event, now extinct (Hackney et al. 1992).

European and world occurrence

Maidenhair Fern is essentially a Mediterranean Basin and S Atlantic species, although Hultén (1962, Map 139) maps it as circumpolar in warm-temperate latitudes of the northern hemisphere. It is also widespread around the southern hemisphere, including appearances in some very remote island groups.

Irish occurrence

A. capillus-veneris occurs as a native species in Ireland only in scattered stations along the W coast, most abundantly in the Burren, Co Clare (where it is especially luxuriant and impressive on Inishmore, Aran Islands) (H9) and in SW Donegal (H35) (Jermy et al. 1978; Scannell & Synnott 1987). The latter, since the Antrim stations are discarded, must now be regarded as the most northerly site of the species anywhere in the world (Hultén 1962; Jalas & Suominen 1972). The New Atlas map displays nine Irish hectads scattered across the island where this fern has been recorded as an introduction.

British occurrence

In Britain, A. capillus-veneris is native in scattered sites from the Channel Isles and the SW coast, up the W coast from Cornwall, through S Wales to Cumbria and the Isle of Man (Jermy et al. 1978; Page 1997; New Atlas). This rather delicate-looking fern also crops up in Britain from time to time as established escapes from cultivation at inland sites in lime-rich mortar or on limestone walls, exactly as it does in Fermanagh (Jermy & Camus 1991).

Names

The genus name 'Adiantum' is a Greek plant name used by ancient botanical writers, derived from 'adiantos' meaning 'dry' or 'unwetted', an allusion to the non-wettable character of the foliage, a feature known to Pliny (Step & Jackson 1945; Gilbert-Carter 1964). The Latin specific epithet 'capillus-veneris', means 'Hair of Venus', and hence the English Common name 'Maidenhair Fern'. 'Capillus veneris' was the medieval Latin name used by apothecaries for this medicinal fern, a name first found in the Herbarius of the fourth century AD ascribed to Apuleius, and it refers to wiry pubic hair which are likened to the blackish stalks of the fern fronds (Grigson 1974).

Additional local common names include 'Capillaire', 'Lady's Hair' and 'Dudder-grass', the latter a strange Norfolk usage, apparently making comparison with Briza media (Quaking-grass or Dodder-grass), because of the trembling motion of the frond segments resembling the movement of the grass spikelets (Britten & Holland 1886).

Uses

Medicinal use dates back to Dioscorides, as a remedy in pectoral complaints and pulmonary catarrhs (Grieve 1931).

Threats

The species has died out at a number of previous stations along the coasts of Ireland, England and Wales. Either of the two Fermanagh sites could easily be destroyed by re-plastering of stonework or other excessive or unknowing 'tidying' operations.

Growth form and preferred habitats

Tunbridge Filmy-fern tends to be the only small, moss-like, mat-forming fern species growing on vertical or near-vertical, permanently and deeply shaded rock faces on mountain slopes and in woodlands, on and under rocks in deep, damp shade.

Identification

The long, flat, overlapping bluish-green fronds easily distinguish H. tunbrigense from the somewhat more common, blackish fronds of H. wilsonii (Wilson's Filmy-fern). Confirmation is often provided by the irregularly toothed margin of the pocket-like indusium covering the spore-sacs, a feature visible with a good hand-lens. The indusia are not always present however when these species are found growing in conditions of very moist heavy shade.

Fermanagh occurrence

In Fermanagh, H. tunbrigense has been recorded in a total of 24 tetrads, 23 of which have post-1975 records. The main areas where it occurs are on Cuilcagh mountain and the sandstone scarps and wooded glens of SW Fermanagh. The isolated 1950 station where it has not been refound was at Annaghmore Glebe Lough.

At some Fermanagh sites, H. tunbrigense colonies can cover several square metres of rock, and sometimes the fern is sufficiently profuse to completely fill rock crevices. Locally, Tunbridge Filmy-fern is almost always found growing on acidic rocks, most frequently on sandstone, but occasionally it may also occur on fairly hard, basic igneous and metamorphic rock types, particularly those providing crevices and with a texture, location and position that enables them to retain moisture for prolonged (or relatively long) periods (Richards & Evans 1972).

H. tunbrigense occurs in close physical proximity with its near relative H. wilsonii right up on the summit ridge of the highest mountain, Cuilcagh (c 600 m). At this site and others on the N-facing slopes of Cuilcagh, on sandstone scarps on the Western Plateau (ie in and around the Lough Navar Forest Park in particular), and in oak and mixed deciduous woodlands nearby (eg the Correl Glen NR), H. tunbrigense is always confined to very sheltered conditions. Typically it grows in deep, shaded hollows under very large overhanging boulders, cliffs or trees, growing on bare rock surfaces or in crevices. It is also found much less frequently on the peaty or uncompacted humus soils of woodland floors. Only rarely does it occur as an epiphyte on the bases of oak or old ash trees in very damp woodland, eg in the Correl Glen, and at the base of the heavily wooded Cliffs of Magho.

Comparison with H. wilsonii

By comparison, H. wilsonii occupies less shaded, somewhat more open and exposed conditions and it is often intermingled and embedded in cushions of moss and leafy liverwort. H. wilsonii does not form large single-species mats to quite the same extent as H. tunbrigense does.

On Cuilcagh, both these species also occur under and around huge, house-sized, rocks on block screes on the northern slopes just below the long, whale-back summit ridge, and again together on further massive rock falls at Cuilcagh Gap, and in similar situations around Lough Atona lower down these same slopes (at c 500 m). There are no trees on any of these heathy moorland slopes, which at this altitude and exposure are dominated by a canopy of ericaceous subshrubs and upland grassland.

Comparative tolerance of desiccation

The normal belief is that of these two species of perfectly frost-hardy filmy-fern, H. wilsonii is better able to tolerate high altitude exposure and the associated risk of desiccation than can H. tunbrigense (Jermy & Camus 1991; Page 1997). In mountain environments in the W of Ireland, rainfall is so very plentiful, frequent and regularly distributed, that droughting of the delicate fronds is not anything like as great a risk as might at first appear. At various sites in Fermanagh drought affected Hymenophyllum plants are occasionally found, particularly affecting the usually more exposed individuals of H. wilsonii, but occasionally also those of H. tunbrigense. The plants may be discovered looking very shrivelled, brown and desiccated, sometimes apparently dead. However, the rhizome and even the fronds have greater powers of recovery than their appearance and structure might suggest, and they can recover surprisingly well from temporary desiccation, or even from the effects of a light heathland fire (Richards & Evans 1972).

Comparative experimental studies by the latter authors have shown that H. tunbrigense suffers the effects of desiccation more immediately than H. wilsonii, but it also recovers more quickly than the latter. However, of the two, H. wilsonii has greater drought resistance, its protoplasm coping better with desiccation and its cells avoiding severe diurnal mechanical stress during drought periods.

A most interesting and rather unexpected finding is that the fronds of both Hymenophyllum species possess the capacity for indeterminate apical growth. This allows individual fronds to survive and continue growing for several years (perhaps four, five, or even more seasons), and thus both species manage to produce sporangia and new indusia in waves, maybe twice a year under favourable growing conditions (Richards & Evans 1972).

Growing as it does in more sheltered and more deeply shaded conditions, it is not really surprising that H. tunbrigense consistently has a lower photosynthetic compensation point than H. wilsonii. On Cuilcagh, the two Hymenophyllum species have found adjacent but distinct habitats, and although their microclimates definitely overlap, obviously they are both able tolerate the prevailing environmental conditions and have found ways of avoiding direct competition.

In the past, doubt has been cast on the finding of H. tunbrigense at altitudes above 460 m in Britain and Ireland (Richards & Evans 1972; Page 1997). However, the local botanists who identified this fern on the ridge of our highest mountain, Cuilcagh, are very familiar with both species of filmy-fern, and I am confident that the identifications are correct. Having emphasised this, H. tunbrigense is recorded only half as frequently in Fermanagh as H. wilsonii, so that of the two, Tunbridge Filmy-fern still clearly has the narrower ecological range (Richards & Evans 1972). While both filmy-fern species grow extremely slowly, c 2.5 cm/yr, H. tunbrigense appears to possess less biological vigour and suffers more from desiccation than H. wilsonii does.

Intolerance of liquid water wetting

One of the interesting facts to emerge from the study of Evans (1964) is that although Hymenophyllum species demand humid conditions, they absolutely do not tolerate being directly wetted with liquid water, for instance, by splashes from streams or waterfalls, which often are the source of the atmospheric humidity they do require. In other words, a rapid change of water content is much more harmful to a filmy-fern than a gradual one (Richards & Evans 1972). Evan's finding appears to directly contradict a statement regarding the habitat of H. tunbrigense made by Page (1997, p. 248), where the latter suggests the fern thrives in the splash zone of cool, permanently tumbling streams. While this is often quoted as one of the preferred habitats of another very much rarer species of filmy-fern, Trichomanes speciosum (Killarney Fern), RSF does not believe he has ever observed Hymenophyllum growing right in the spray zone anywhere in Ireland.

British and Irish occurrence

H. tunbrigense is restricted to a very discontinuous occurrence in the N, W and S of both Ireland and Britain, extending from Cornwall to Skye. In Britain, apart from an outlying group of sites in East Sussex (where it has markedly declined in recent years) and a couple of very scattered sites in NE Yorkshire, S Northumberland and Cheviot (VCs 62, 67 & 68), it is otherwise completely absent from the east of the island (Preston et al. 2002).

European and world occurrence

The distribution of H. tunbrigense in continental European is very sparse and disjunct, even in comparison with its representation in the British Isles. It is known only from a few stations each in France, Luxembourg, N Spain, Italy and the eastern coast of the Black Sea. Beyond this, it does also occur on the Azores, Madeira and the Canaries (Jalas & Suominen 1972, Map 69; Richards & Evans 1972).

Fossil history

Fossil spores of H. tunbrigense have been recorded from the Hoxnian interglacial in Ireland by Watts (1959). This fact, taken together with the species' present-day widely disjunct and sporadic European distribution, undoubtedly confirms it is a relict species. Previously the species had a larger and much more continuous range (probably most recently around the current post-glacial climatic optimum), but it has declined and continues to do so (Richards & Evans 1972). Gradual climatic deterioration, compounded in recent centuries with habitat destruction by man, has resulted in the fragmented distribution of H. tunbrigense we observe in Europe today.

Name

The specific epithet 'tunbrigense' is a Latinised reference to the first known British site, found by Daire a few years previous to Ray's (1686) published report, at High Rocks, Tunbridge Wells, Sussex (for a full history see Evans & Jermy (1962)).

Threats

Some Fermanagh sites were for a time threatened by being overgrown by coniferous forest plantations, but the threat has eased since the trees have been felled and they are not being re-planted.

Growth form and preferred habitat

This moss-like, rhizomatous, cushion-forming perennial fern grows in similar habitats to H. tunbrigense (Tunbridge Filmy-fern) on ± vertically erect, damp, shaded, mountain rock crevices and ledges, and on mossy rocks and tree boles in damp woodlands, but while usually in shade it always occupies slightly more open situations than the latter.

Fermanagh occurrence

In Fermanagh, H. wilsonii has been recorded in 34 tetrads (6.4%), 31 of which have post-1975 records. As the tetrad map indicates, it is quite widely distributed on the Western Plateau, Cuilcagh Mountain and Florencecourt. The most isolated recent station is on a scarp at Drumskinny Td in the far north of the county, where RHN found it in May 1990.

Comparison with H. tunbrigense

The fronds of H. wilsonii are generally longer, narrower and more upright than those of H. tunbrigense and in overall appearance they are less flattened and of a deeper shade of olive-green, sometimes rather blackish, or indeed brown if it has recently suffered drought. This little fern is often found growing amongst mosses where its distinctive colour makes it quite easy to spot.

Wilson's Filmy-fern generally forms rather smaller patches than Tunbridge Filmy-fern and sometimes when it is growing through cushions of moss, the fronds are few and quite distant from one another, so it could easily be overlooked or mistaken for one of the larger species of leafy liverwort. The frond is stiffer and of thicker texture than that of H. tunbrigense and, because it is more drought resistant and can thus thrive better in illuminated sites than the latter, it sometimes competes with lichens as well as with mosses and liverworts.

Although the two filmy-ferns occur in very similar habitats and their micro-environments overlap, in any type of habitat where they occur together they typically occupy different vegetation zones. For instance, H. wilsonii occurs higher on the bark of trees and on block screes on mountainsides it occurs in more exposed, better lit sites. Like H. tunbrigense, H. wilsonii is usually absent from the spray zone of waterfalls, but while it normally avoids running or dripping water, it is slightly more tolerant of being wetted in this way and occasionally it is found on rocks which are flooded from time to time (Richards & Evans 1972).

Of the two species, H. wilsonii is better able to withstand relatively prolonged drought. Sometimes, however, drought can result in colonies peeling away from rock or soil surfaces, which may lead either to their destruction (Richards & Evans 1972) or, very occasionally in the W of Ireland, this may serve as a method of vegetative dispersal.

Very little is known about the life-history of Hymenophyllum species, save that they are very slow-growing and that the individual fronds may survive up to about five years. The frequent occurrence of large colonies of Hymenophyllum suggests that they may be long-lived, perhaps surviving in stable environments for centuries. This is clearly an area which would repay further field study.

Irish occurrence

The distribution of Wilson's Filmy-fern overlaps that of Tunbridge Filmy-fern in the Atlantic-influenced N and W of both Britain and Ireland, but it is quite a lot more widespread – recorded in more than twice as many hectads as H. tunbrigense. In Ireland, H. wilsonii is locally abundant in all the main mountain ranges (which with the exception of the Galtees in Tipperary, all happen to be quite coastal), and in addition it is thinly scattered in damp woods throughout the island (Richards & Evans 1972; Jermy et al. 1978).

Survey work for Atlas 2000 in the north of Ireland has proven that H. wilsonii is much more widespread than had been previously considered (NI Flora Web site 2001; New Atlas).

British occurrence

In Britain, the species is again locally abundant in the N & W of the island, but unlike H. tunbrigense it has no south-eastern outlier. Instead, it is confined west of a line from Start Point in Devon to the mouth of the River Tees. Within this area of Great Britain, it does extend further east and much further north than H. tunbrigense does, extending to Shetland (Richards & Evans 1972).

World occurrrence

H. wilsonii has a much more confined world distribution than H. tunbrigense, being largely restricted to W Europe and N Atlantic islands (Hultén & Fries 1986). It is known only from the Faeroes, one station in Iceland plus the more southern Atlantic islands (Azores, Madeira and the Canaries), which are not really part of Europe but where species populations are generally conspecific with European flora. On the European mainland, the distribution of H. wilsonii runs from a broad coastal band of SW Norway plus just four coastal sites in N France (Jalas & Suominen 1972, Map 70; Page 1997; Jonsell et al. 2000).

Names

The genus name 'Hymenopyllum' is a compound of two Greek words, 'umen' meaning 'a membrane or thin skin' and 'phyllon' meaning 'a leaf', a reference of course to the membranous fronds which are a single cell thick (Step & Jackson 1945). The Latinised specific epithet 'wilsonii' commemorates the bryologist, W. Wilson, who in 1830 drew the differences between the two filmy-ferns to the attention of the elder Hooker (Professor, and later Sir William). Hooker named the species but subsequently reunited the two species under H. tunbrigense once more (Richards & Evans 1972). For a full history, see Evans & Jermy (1962).

Threats

Some sites could be threatened by forestry operations.

Fermanagh occurrence

The discovery of this very distinctive fern in Fermanagh was announced in the Irish Naturalist of December 1900 when R.Ll. Praeger wrote, "The investigations of Messrs. West and Tetley have added the Killarney Fern to the Flora of Co Fermanagh. They sent me a specimen last August, describing the exact locality – a deep crevice in limestone rocks; but I think it better not to publish the station, so terribly has this lovely plant suffered from the depredations of unscrupulous collectors." The following year Praeger included mention of the station in his monumental Irish Topographical Botany simply as, "Near Carrick". The site was more clearly identified as the Correl Glen in the unpublished Typescript Flora of Co Fermanagh of 1951. The rocks in the area around the Correl Glen are limestone and calcified or dolomitised sandstone. The glen supports old, moist, mixed deciduous woodland, growing on peaty soils at an altitude of around 100 m.

Growth form and preferred habitat

Killarney Fern is closely related to the genus Hymenophyllum and is another type of filmy-fern having rather larger, dark-green, membranous, translucent fronds, 7-60 cm in length and much divided (Ratcliffe et al. 1993). Unlike both of the Hymenophyllum species, however, T. speciosum typically resides near waterfalls or mountain streams in sheltered, deep, shady rocky gorges, growing in rock crevices between boulders, in dripping caves or under overhangs where spray or seepage water keeps the plant permanently moist and dripping (Page 1997). Almost all of its British & Irish stations past and present have been in mild oceanic districts in the far west and often at low altitudes. Evidently it is a warmth-demanding species requiring a great deal of shelter, and probably it is rather frost sensitive, a feature which contrasts strongly with the much more hardy Hymenophyllum species (Wigginton 1999).

Gametophyte generation

The prostrate, sexual, gamete-bearing gametophyte generation of Trichomanes was first recognised by Mettenius in cultivation as long ago as 1864. Uniquely amongst European ferns, the gametophyte reproduces vegetatively in the absence of the asexual spore-bearing generation, the ± upright, or else dangling, frond-producing sporophyte. The independent gametophyte populations were completely overlooked in Britain until 1989 (Rumsey et al. 1992).

In February 1993, Nick Stewart visited Fermanagh with an international group of bryologists. When examining mosses and liverworts in the Correl Glen and on the overhanging scarps of Lough Navar Forest Park, he twice found the gametophyte of the Killarney Fern. RHN was on the outing and learnt how to identify this green, thread-like sexual stage of the fern life-cycle. Since 1993 he and his wife Hannah have found it in twelve tetrads spread across three hexads. Many of these gametophyte finds have been verified by staff at the Natural History Museum, London. The southerly, most isolated site shown on the map is in a deep, dark hole in a scarp at Aghnahoo, on the slopes of Cuilcagh mountain.

The gametophyte grows in Fermanagh under deep, dark rock overhangs, usually about an arm's length removed from the light. It is the last form of plant life to grow before light levels become totally inadequate. The gametophyte forms patches which vary in size from several square centimetres to almost the size of one's hand in exceptional cases. It is light green, looks like a filamentous alga, and has a spongy texture when pressed down. When viewed under a hand lens it looks like wire wool with small spikes sticking out from it like the thorns on certain rose bushes. The inability of all these gametophyte plants around the British Isles to carry out successful sexual reproduction and develop new mature sporophyte frond-bearing plants, means that the high conservation status of T. speciosum, ranking it as vulnerable, must remain in force. Along with other oceanic species, Killarney Fern is considered by some conservation biologists to be at further risk from the predicted effects of global warming (Plantlife 1991).

Further detailed study of the exact growth requirements of T. speciosum is urgently needed, and an investigation as to why the gametophyte fails to produce new sporophyte plants would also assist conservation and rehabilitation efforts associated with increasing biodiversity awareness.

Robert Northridge rediscovers the sporophyte

On 1 May 2005, while searching for the gametophyte under a large rock at a location within the H05 hectad, RHN discovered a solitary plant of the sporophyte. It consisted of two fronds each 8 cm long, one frond 6 cm long, another frond 6 cm long but with the top half dead, plus the broken stipe only of another frond. There was also a small patch of the gametophyte about 15 cm × 10 cm close by. When the sporophyte site was visited a year later on 26 May 2006, as well as the previously seen fronds, there was a new frond unfurling which was just over 5 cm long.

This site was in a very sheltered, cave-like area under large fallen rocks through which there is no perceptible flow of air that might dry and damage the wet filmy fronds. The site is under trees and the entrance to the 'cave' is overhung by dense straggling leaves of Luzula sylvatica (Great Wood-rush). Exact details of the location of this site have been lodged with the Northern Ireland Environment Agency (NIEA) in Belfast.

The rediscovery of the Fermanagh Trichomanes sporophyte has been the recording highlight of this whole VC Flora project.

Gametophyte Irish occurrence

The Trichomanes gametophyte has been recognised in numerous Irish sites and by 1998 was known from 13 VCs. In Britain, the same published report listed the gametophyte from a total of 38 VCs and provided a map for both islands (Rumsey et al. 1998, Figure 4). Since then isolated finds of the gametophyte have been made in Cos Tyrone and Antrim (H36 and H39) and we are confident that this generation of the species remains under-recorded.

Previous Irish occurrence of the sporophyte

In the early 19th century, the sporophyte of T. speciosum was found to be frequent and widespread in and around Killarney and SW Ireland in general (Jermy et al. 1978; Wigginton 1999). On account of the beauty and comparative rarity of the fern, T. speciosum became the supreme target of collectors and gardeners during the long-running Victorian Fern Craze (1830-1920) (Allen 1969; Page 1988).

The Victorian Fern Craze in SW Ireland

Kerry and other parts of the west of Ireland attracted many visitors on account of its wealth of fern species, and since T. speciosum appeared in such abundance in the region, it probably seemed harmless to remove a little of it in order to grow the lovely fern in one's front room in a glazed Wardian Case or in a Conservatory, or to press and dry it for the family herbarium collection. Visitors found the fern easily, and there was no difficulty in pulling it down from its wet rock under-hang or cave roof habitat. In fact, the long strands of bristly, black, rhizome came away far too readily, and it is very likely that visitors accidentally removed far more of the plant than they ever wanted to transport. At one stage, tinkers collected Killarney Fern and hawked boxes of it around the Killarney hotels, selling it to visitors, until eventually the extremely slow-growing plant was reduced to extreme rarity and local extinction (Marren 1999).

A lingering collective cultural guilt is the reason why naturalists are now so extremely circumspect about the remaining stations of the Killarney Fern in the British Isles, and it is now recognised as a Red Data Book species, protected by conservation laws in both Britain and Ireland.

Irish occurrence

In Ireland, over the years, the sporophyte T. speciosum has been recorded from a total of 43 scattered sites. The Irish Red Data Book (1988) reported that it had been recently found in just ten of these sites. A subsequent study of T. speciosum habitats in the British Isles made by Ratcliffe and his associates found the sporophyte was still quite widespread in the hill country of Cos Kerry and Cork, where they located a total of 26 'colonies' (ie individual patches, each representing a single rhizome system). These workers reckoned they had searched only a small part of the total possible ground in SW Ireland that might support the sporophyte generation. Nevertheless, they concluded that T. speciosum is a much reduced species, and that some of the more outlying Irish colonies were also collected out of existence besides the ones previously known about in Cos Kerry and Cork (Ratcliffe et al. 1993).

Similar depredations befell the much more thinly scattered populations in England, Scotland and Wales, where sporophyte T. speciosum is now confined to very few localities. Ratcliffe and his associates located a total of just 13 of their individual 'colonies' in Great Britain, details of which are kept so secret that their map references do not even appear on the confidential computer records at the Biological Records Centre at Monkswood (Ratcliffe et al. 1993; Marren 1999).

Field observations suggest that under existing environmental conditions Killarney Fern has been incapable of producing any new sporophyte populations for over a hundred years. Again, like other filmy-ferns, growth of the fern is very slow. Each frond can continue growing for five or more seasons, and individual colonies that have been observed and measured over many years, do not appear to change in appearance or show any significant growth (Ratcliffe et al. 1993).

World occurrence

Reflecting the climatic limitations discussed above, T. speciosum is confined to a very limited mild, damp oceanic area of Europe and the Atlantic isles (Azores, Madeira and the Canaries). Besides the British Isles, it occurs only in Brittany, the Pyrenees, near Gibraltar and near the west coast of N Italy in the Alpi Apuane at between 180-250 m (Jalas & Suominen 1972, Map 71; Ratcliffe et al. 1993; Pignatti 1997, vol. 1, p. 53; Rumsey et al. 1998).

Uses

There does not appear to be any folklore about the fern, nor is it credited with any medicinal properties. It has, of course, been valued for its showy beauty and rarity and has been cultivated by keen gardeners since its discovery. Trichomanes is a cosmopolitan genus and contains a total of 25 species, most of them tropical, and the majority in horticulture originate from the southern hemisphere (Griffiths 1994).

Names

The genus name 'Trichomanes' was given to an unknown fern species by Theophrastus (possibly Asplenium trichomanes, Maidenhair Spleenwort). Gilbert-Carter (1964) regards its origin as obscure, but it is definitely a Greek compound involving 'thrix, thichos', meaning 'a hair', 'a bristle', or 'hairy' (Stearn 1992). One idea for the derivation of the other half of the genus name is the Greek 'manos', meaning 'flexible', but there is some doubt as to the relevance of this suggestion, and the correct derivation therefore remains a mystery (Step & Jackson 1945).

The specific epithet 'speciosus' is Latin meaning 'showy' or 'handsome'.

Threats

All the currently known sites are within ASSIs. Further study of the ecology and biology of the species is urgently required in order to properly manage its conservation.

Taxonomy

What was for many years treated as one Polypodium species in B & I has been recognised, since 1960, as a polyploid complex of two rhizomatous perennial species (tetraploid Polypodium vulgare and diploid P. cambricum (Southern Polypody)) and the fertile allopolyploid hybrid between them (hexaploid P. interjectum (Intermediate Polypody)) (Shivas 1962; Jermy & Camus 1991). Some forms of Polypody require microscopic investigation to identify with certainty and, therefore, field records are still very frequently made at species aggregate level (ie as P. vulgare agg. or s.l.), and the constituent species and hybrids seem destined to remain under-recorded.

Growth form and preferred habitats

All forms of Polypodium, like Bracken, do not develop a crown, but instead possess a scaly, creeping rhizome, which in this case runs along the surface of the ground rather than being buried in the soil. From the rhizome, leathery, evergreen, hairless, aerial fronds resistant to both frost and drought arise at intervals, generally only a few centimetres apart. Fronds vary enormously in length from just a few centimetres in stunted forms growing in dry or exposed situations as in rock crevices or on walls, but they can develop up to 45 cm in length in the case of hybrids which display 'hybrid vigour' or heterosis.

Representatives of the Polypodium vulgare agg. typically grow amongst mosses and leaf-mould on semi-shaded rock outcrops, under hedges, on walls, on roots and stumps of trees, or on the trunks and thicker branches of mature trees in damp woods.

Reproduction

The sporing sori are formed in pairs on the underside of the herringbone-like frond either side of the stalk (technically called a 'rhachis') in the upper one to two thirds of its length. The sori are large, rounded, naked of any covering indusium tissue and are a conspicuous, golden-orange colour when fully developed, anytime between June to September. The sporangia are so securely attached that indentations appear on the upper side of the frond opposite the sori positions beneath.

Fermanagh occurrence

In Fermanagh, P. vulgare s.l. is very common and widespread, having been recorded in 401 tetrads, 76% of those in the VC. The species aggregate typically grows amongst mosses and leaf-mould on semi-shaded rock outcrops, under hedges, on walls, on roots and stumps of trees, or on the trunks and thicker branches of mature trees in damp woods.

British and Irish occurrence

The New Atlas hectad map shows P. vulgare s.l. occurring throughout the vast majority of B & I, but absent from the Channel Isles and less prevalent in the English Midlands and up the E coast from the Wash to Newcastle-upon-Tyne. The New Atlas map also indicates areas of less common occurrence or absence over parts of the Irish Midland plain, and also in the more exposed, wet, nutrient-starved, boggy ground of Cos Clare (H9), Roscommon (H25), W Mayo (H27) and W Donegal (H35).

World occurrence

On a European basis, the aggregate species occurs widely and commonly throughout continental Europe extending from Gibraltar and Crete in the south, to the northern tip of Norway, Iceland and the southern tip of Greenland, but absent from Jan Mayen and the Arctic Islands (Svalbard, etc.) (Jalas & Suominen 1972). Eastwards, P. vulgare agg. occurs in the Caucasus, Urals, the Himalaya, and, in an even broader definition (P. vulgare sens. lat.), it becomes circumpolar and stretches around middle latitudes of E Asia and across northern and central regions of N America (Hultén 1962; Hultén & Fries (1986, Map 74).

Medicinal uses

The rhizome of Polypodium had several uses in herbal medicine and was often referred to as Polypody of the Oak, or Oak fern (not our modern, much more delicate Gymnocarpium dryopteris). There was a belief that ferns and flowering plants that grew on oak roots or branches were especially efficacious as remedies. Its principal use was as a mild laxative, but it was also considered useful for coughs and catarrh. An infusion made from the crushed rhizome was drunk like tea as a treatment for the early stages of consumption or for rheumatism. Polypody was used to treat jaundice, dropsy, scurvy and other skin complaints, and dried powdered rhizome used alone or mixed with honey was also said to remove nasal polyps (Grieve 1931; Vickery 1985).

Names

The genus name 'Polypodium' is Latin but is derived from two Greek words 'polus' meaning 'many' and 'pous' meaning 'a foot', the notion being either that the plant having a branched rhizome has many feet (Grigson 1974), or more likely, that the comb-like, pectinate frond resembles a centipede (Prior 1879).

In addition to the English common names mentioned above, this easily recognised and well-known fern aggregate has been given at least nine other names including: 'Brake of the Wall', 'Adder's-fern', 'Everferne', 'Wall Fern', 'Wood Fern', 'Golden Locks', 'Golden Maiden-hair', 'Golden Polypody' and 'Moss Fern'. The latter, a name proposed by Gerard (1597), is rather apt (Britten & Holland 1886).

Threats

None.

Growth form and identification

Some of the earlier Fermanagh records of this rhizomatous species were determined from herbarium vouchers in BEL by Paul Hackney. Typically P. vulgare s.s. has narrow, rather leathery, evergreen, parallel-sided fronds with the lowest pair of pinnae not bent forward (ie not inflexed) like those of P. interjectum (Intermediate Polypody) (the allohexaploid derivative of its hybrid with P. cambricum (Southern Polypody)).

New fronds of P. vulgare s.s. are produced in early summer and the species, which is a tetraploid, has numerous pairs of naked sori in the upper portion of the frond which ripen their spores in mid-summer (Jermy & Camus 1991).

Preferred habitats

In contrast to the two other species of the genus in our survey area, P. vulgare s.s. is a definite calcifuge. It typically grows on cliff ledges and in rock crevices on steep, peaty banks, in between the rocks in old, dry-stone walls or along the tops of such walls, and it also occurs as an epiphyte on the bark of mature deciduous trees in damp woods.

A study of the preferences of a number of ferns with respect to soil reaction carried out in W Europe by Koedam et al. (1992) found that soil taken from the root mass of P. vulgare s.s. had a median pH of 4.13, and consequently the plant was regarded by these workers as 'acidiphilous' (ie acid loving, or acid tolerant). Such species are adapted to soils with high levels of exchangeable aluminium and hydrogen, and relatively low levels of exchangeable calcium.

British and Irish occurrence

P. vulgare s.s. is the most common form of Polypody recorded in both Britain & Ireland, and in N Ireland it is recorded in almost every hectad.

Fermanagh occurrence

In Fermanagh, this is the most common and widespread species of polypody, having been recorded so far in 94 tetrads, 17.8% of those in the VC. Nevertheless, since many field recorders work at the species aggregate level and do not distinguish the separate Polypodium species, we regard it as definitely under-recorded. As the tetrad distribution map indicates, this form is widely scattered throughout Fermanagh, but more prevalent in the wetter, more acid, rocky upland environments of the Western Plateau.

Threats

None.

Growth form and preferred habitats

All ten records for this sterile pentaploid hybrid in the Fermanagh Flora Database have been determined by Paul Hackney working on specimens deposited in the herbarium at BEL. The most recent eight were collected by RHN either growing in moss on rocks or as epiphytes on tree trunks in damp old woods. In view of the frequency of the parent species and their degree of ecological overlap, this intermediate hybrid is probably quite widespread and is likely to be the most common of the three sterile Polypodium hybrids.

In a W European study assessing the pH preferences of ferns and their root cation-exchange properties and preferences, Koedam et al. (1992) found that P. × mantoniae follows its tetraploid parent P. vulgare s.s. in being 'acidiphilous' (ie strongly calcifuge), since it occurred on soils with a median pH value as low as 3.8. Furthermore, no calcium carbonate was detected in any of the soil samples taken from the roots of this fern hybrid. On the other hand, Page (1997) reported it occurring in humid conditions on mossy boulders over a wide range of rock types including limestones.

British and Irish occurrence

Page (1997) reported that this hybrid had been found in at least 29 VCs in Britain & Ireland and that the distribution was slanted towards western areas. The New Atlas hectad map shows that P. × mantoniae is now known quite widely across Britain and N Ireland, and it occurs on soils derived from both siliceous acid and limestone rocks.

Reproduction

Although this is usually regarded as a highly sterile hybrid, some of the plants appear to have a high proportion of apparently normal spores. The backcross with P. vulgare s.s. has been suspected in at least one of the Fermanagh records. This hybrid is vegetatively vigorous and can form large clones.

Fermanagh occurrence

The details of the other nine Fermanagh records follow with their collector; all were determined by P. Hackney and vouchers exist: Finlane Td, Florencecourt Forest, 1976, P. Hackney; scarp SW of Lough Achork, December 1987, RHN; Castle Archdale, Lower Lough Erne, December 1987, RHN; Ely Lodge Forest, Lower Lough Erne, February 1988, RHN; Sillees Wood, March 1989, RHN; Clonelly, NW of Kesh, April 1989, RHN; Arney River, April 1989, RHN; Crossmurrin NR, December 1989, RHN; Ballindarragh Bridge, Colebrooke River, 1988-90, RHN. The last listed record is considered a possible backcross with P. vulgare s.s.

Growth form and preferred habitats

Typical specimens of this species (a fertile hexaploid hybrid formed by allopolyploidy between the other two British Isles Polypodium species) have ovate to narrowly oval fronds, generally longer than those of P. vulgare s.s. and with at least the lowest pair of pinnae bent forwards (ie inflexed), to form a 'V' shape. The fronds are leathery, evergreen and frost and drought resistant, fresh ones being produced in late summer and autumn, ie later in the year than P. vulgare s.s. and before those of P. cambricum (Southern Polypody) (Jermy & Camus 1991).

The plant is weakly calcicole or may prefer near-neutral conditions. A study of fern species with respect to root cation-exchange properties carried out in W Europe discovered the pH of soil samples at the roots of P. interjectum had a median value of 6.67, so that Koedam et al. (1992) classified it as 'neutrocline'.

P. interjectum grows in very much the same types of habitat as P. vulgare s.s., ie on rocks, cliffs, stony banks, mortared walls and also as an epiphyte on trees in damp woods.

Fermanagh occurrence

As with P. vulgare s.s. (Polypody), this perennial polypody is definitely still under-recorded in Fermanagh. Many of the existing records have been determined or verified by Paul Hackney at BEL. Intermediate Polypody is the second most common species of polypody in Fermanagh having been recorded so far in 90 tetrads, 17.1% of those in the VC. As the distribution map indicates, it is widely scattered throughout, but more frequent in the west of the county.

British and Irish occurrence

The Fern Atlas hectad map gave an early picture of the known distribution of P. interjectum in 1978, where the sub-Atlantic influence appeared quite strong in Britain, while the Irish distribution was then very much more evenly, or randomly scattered, although inland sites in NI appeared almost entirely absent (Jermy et al. 1978). The New Atlas hectad map now shows Intermediate Polypody is common and widespread throughout most of NI, but with apparent gaps in Cos Tyrone and Armagh (H36 and H37). The very obvious patchiness of the plotted distribution on both islands strongly suggests that recording of P. interjectum, although greatly improved in comparison with the Fern Atlas, remains incomplete.

Name

The Latin specific epithet 'interjectum' means 'intermediate in form', and of course refers to the fact that this species arose as a fertile hybrid (Gledhill 1985).

Threats

None.

Growth form and preferred habitats

P. × shivasiae is the rarest of the three Polypodium hybrids in Britain and Ireland. However, since the parent species are becoming quite frequently recorded in Fermanagh (particularly P. interjectum (Intermediate Polypody)), and as they are both weakly calcicole and undoubtedly overlap geographically and ecologically, this hybrid might be rather less rare in the VC than the current few records indicate. Having said that, as Roberts (1970) pointed out, survival of the relatively small, often sporadic populations of P. cambricum might be adversely affected by the formation of its hybrids with other species, since the offspring will compete with the diploid parent in a habitat in which it is already rather restricted. Under these circumstances any genetic barrier to hybridisation would tend to be strengthened.

P. × shivasiae is the most spectacular in appearance of the three sterile hybrid polypodies found in the British Isles, combining as it does large size with vigorous growth. As to be expected, it is intermediate in its morphological and physiological characteristics with respect to its parents. It grows on sheltered sections of limestone walls and on equally sheltered, shaded, parts of steep, wet, basic rock cliff faces, usually with both parent species nearby (Page 1997).

Fermanagh occurrence

All four Fermanagh records of this rare hybrid have been determined by Paul Hackney, the first record verified by R.H. Roberts. The details of the other three Fermanagh records, all collected by RHN, are as follows: Keenaghan Lough, Tievealough Td, September 1988; Levally House, 1 km SE of Roosky, October 1990; and W end of Lough Acrussel, 1988-90. This last record is marked "requires checking".

Name

The specific name 'shivasiae' commemorates the work of the British fern taxonomist M.G. Shivas, who contributed greatly to understanding of the genus in the 1960s and 1970s.

Growth form and preferred habitats

The fronds of P. cambricum are broadly triangular and can be up to 50 cm in length, with the edges of the pinnae often quite serrate. New fronds are produced later than in the other two Polypodium species, in the autumn and winter, and they are reputedly the least frost hardy of the three species. It has been suggested that in Britain and Ireland Southern Polypody is mainly confined to lowland sites within a few hundred metres of sea-level, often coastal and lying within the 4°C winter minimum isotherm (Page 1997). Spore-producing sori are confined to the upper third of the frond; young sori are more oval than in other forms of polypody, and when the plant is vigorously vegetative, no sporangia are formed, rendering the plants sterile.

Species identification

Variation in both vegetative and reproductive characters is so great in the genus Polypodium that P. cambricum (the diploid) and P. interjectum (the hexaploid allopolyploid fertile hybrid derived from a cross between P. cambricum and P. vulgare s.s., followed by spontaneous chromosome doubling), tend to overlap in many respects. The presence of relatively large, branched colourless threads (ie paraphyses), within the sorus among the sporangia, are unique to P. cambricum, and taken in conjunction with 'good spores' (ie fully formed, fertile, regular-shaped, non-aborted ones), they provide the best distinguishing character for the species (Roberts 1970; Page 1997).

Having said this, these are very much laboratory microscopic characters, and furthermore care must be taken not to confuse the much larger, more branched paraphyses with the minute glandular hairs which occur scattered on the lower surface of the frond in all European species of Polypodium. These hairs are much smaller than true paraphyses, being usually only 3 or 4 cells long – with their glandular, inflated terminal cells mostly coloured a dark reddish-brown (Roberts 1970 – see p. 128, Fig. 6 to clarify these differences).

Fermanagh occurrence

While this perennial is the most distinctive and readily identified of the Polypodium species or hybrids in Britain & Ireland, it is also very definitely the least frequent of the three species in Fermanagh and elsewhere in these islands (Page 1997). So far, P. cambricum has been recorded in just 37 of Fermanagh tetrads, 7.0% of those in the VC.

Southern Polypody is a strongly calcicole species and, as the Fermanagh tetrad map indicates, it typically grows in the W & S of the county on limestone natural rock outcrops. Often it occupies the more sheltered parts of cliffs, crevices in old lime-mortared walls, or ground on the steep-sloping rocky floors of hazel woods, where it is capable of forming large stands. Locally, it has never been found in artificial habitats such as quarries, nor growing as an epiphytic on the bark of trees, behaviour that may be confined to more maritime districts than occurs in Fermanagh (Jermy & Camus 1991).

In drier, more exposed upland sites, which are usually limestone gorges and cave mouths, P. cambricum associates with Asplenium ruta-muraria (Wall-rue) and Ceterach officinarum (Rustyback). On the other hand, when occurring in shade in Ash and Hazel woodland in constantly damp and humid conditions, it consorts instead with other shade-tolerant species such as Hedera helix (Ivy), Polystichum setiferum (Soft Shield-fern) and Phyllitis scolopendrium (Hart's-tongue).

British and Irish occurrence

The current knowledge of the occurrence of P. cambricum in Britain & Ireland, as displayed at the hectad level in the New Atlas and the 2005 New Atlas of Ferns, shows it is both southern and western in Britain, although stretching northwards as far as mainland Argyll in W Scotland (VC 98) (Preston et al. 2002; Wardlaw & Leonard 2005). The same two maps show it very much more widely scattered across the whole of Ireland. The distribution looks patchy, however, suggesting that the recording effort is distinctly uneven. More work is required to arrive at an accurate picture of the real distribution of this fern.

World occurrence

P. cambricum is entirely confined to W and S Europe, being sparse and rather disjunctly spread from the British Isles to Portugal, N Africa and eastwards through the Mediterranean islands to Greece and Turkey (Jalas & Suominen 1972, Map 140; Page 1997).

Name

The species Latin epithet 'cambricum' means 'of Wales' (Cambria) (Gilbert-Carter 1964).

Hybrids

While diploid P. cambricum is by far the least common of the three Polypodium species in Britain & Ireland, its hybrids with the other two species are even rarer. The hybrid between P. cambricum and P. interjectum (P. × shivasiae) should be actively looked for as both species are calcicole and P. interjectum (being very much the more common parent) probably occurs at or near every P. cambricum site. So far only four records of this hybrid have been discovered in Fermanagh.

Threats

Clearance of woodland.

Introduction

Bracken is a serious opportunistic, invasive, colony-forming weed almost everywhere in Britain and Ireland, and indeed world-wide on under-used, under-managed or abandoned farmland (Cody & Crompton 1975; Taylor 1990). Current land-use changes in Britain and Northern Ireland induced by socio-economic and political forces, including the policies of 'set-aside' and 'Environmentally Sensitive Area' farmland designation, may very well end up promoting Bracken encroachment. Bracken is such a vigorous invader, persistent and so extremely difficult to eradicate, that its many researchers have formed an International Bracken Group, which regularly holds conferences and issues publications (Smith & Taylor 1986, 1995; Thomson & Smith 1990).

Taxonomy

P. aquilinum is world-wide in its distribution, but it is both genetically and ecologically variable and it is also phytochemically polymorphic (eg with respect to the levels of animal and plant toxins it contains), so that several different schemes of subdivision have been proposed involving subspecies (or species) and up to a dozen varieties (Tryon 1941; Cody & Crompton 1975; Page 1997).

Taken at its taxonomic widest, P. aquilinum has had a couple of hundred varieties and forms described (Hultén 1962, Map 131); but that way madness lies! In Britain and Ireland, what previously was regarded as the solitary species P. aquilinum has been subdivided by Page (1997) into two species, P. aquilinum and P. pinetorum, the former being further subdivided into three subspecies, and the latter (which is confined to scattered sites in Perthshire and Inverness-shire), into two subspecies. The New Flora of the British Isles (1997), on the other hand, takes a very much more conservative view of the variation, downgrading Page's species to subspecies, and regarding his subspecies merely as local ecotypes. We have not differentiated the subspecies of either of these authors in Fermanagh, but almost undoubtedly the Bracken in our area is P. aquilinum subsp. aquilinum, by far the most vigorous and abundant weedy form of the plant throughout the whole of the British Isles.

Growth form, size and timing

The thick black rhizome runs deeply buried, between 10 to 50 cm below the surface in well-drained acid soils (usually less than pH 6.5). The spreading rhizome branches and sends up to the surface the familiar annual branched aerial green fronds. The fronds vary greatly in height, but can be up to 2.5 m tall (a maximum of 2.75 m tall has been recorded (Step & Jackson 1945)), when developed over fertile deep soil. Often fronds are only half that height when the fern is growing on more nutrient-poor substrates.

The fronds begin poking above ground in mid-April, and when the hook-like croziers first appear they are very sensitive to frost and to damage by trampling (Grime et al. 1988). The fronds are slow to develop, requiring time for the great length of stipe and rachis to harden before the physical strain is put on them of the fully spread-out pinnae. In fact the fronds do not stop growing and using up the rhizome energy reserves until the latter half of July. This timing determines the optimum timing for cutting or the use of systemic herbicides for the eradication of Bracken, since the plant's reserves are then almost non-existent. The energy-containing products of photosynthesis are then being translocated down into the underground parts of the plant for storage, and the herbicide will be carried down into the rhizome with this flow of material.

Preferred habitats

Bracken is very frequent on hillsides, sloping, rocky waste ground, woods and roadside embankments. It is especially common on reasonably deep, well drained, acidic upland pastures, where it can produce fronds up to 2.5 m tall and become locally dominant. The plant does not have a high requirement for soil nutrients and the ideal habitat in Scotland was described as ground providing free-draining slopes of brown earth soils on sheltered, lower hill faces of glen sides. It may also occur on similar soils in glen bottoms, but only if the drainage there is unimpeded (MacLeod 1982).

P. aquilinum is much less common on base-rich or shallow substrates where its growth is often stunted. In general, Bracken avoids aquatic sites, shallow soils, poorly drained, deep, acid peatlands and densely shaded conifer plantations. It is also absent from well managed farmland, from frost hollows and from very exposed situations (Biggin 1982; MacLeod 1982).

While primarily a plant of woodland shade and open moorland pastures, Bracken is also common on rough grassland on waste ground and roadsides. Although P. aquilinum abhors waterlogging, it frequently occurs near waterways and on the drier margins of lakes and bogs at all altitudes, excepting the most exposed sites. In general, for reasons that are not at all clear, Bracken seems to frequent shade more often in lowland sites than it does in upland areas (Grime et al. 1988). It is also frequently found on cliffs, screes, limestone pavement and other forms of rocky ground. This includes artificial habitats such as old stone quarries and neglected areas of sand and gravel pits.

Soil preferences

As is the case with other primarily or mainly calcifuge plants, Bracken frequently occurs in limestone areas of Co Fermanagh. The explanation of this unexpected behaviour is that soils in the west of Ireland have been so thoroughly leached by prevalent heavy rainfall, acid peat often develops directly on top of calcareous rock. The suggestion has been made that P. aquilinum on limestone soils in places such as Andalucia in southern Spain, may differ to the extent of having only half the normal chromosome number (ie 2n= 52) (Molesworth Allen 1968). Studies on the soil preferences of ferns with respect to acidity carried out in W Europe by Koedam et al. (1992), found that P. aquilinum was 'acidiphilous' (acid-loving or acid-tolerant). In this study, soil samples taken from the fern root mass had a median pH of 4.0. On the other hand, Jonsell et al. (2000) suggest that in Scandinavian countries, Bracken is probably quite indifferent to lime.

Bracken litter ecology

Being a very large colonial, deciduous plant, Bracken has the potential to produce an enormous amount of leaf litter when the aerial shoots die off in late autumn. In communities where P. aquilinum forms virtually pure stands, litter production can be between 8,000 and 14,000 kg/ha/yr (Watt 1976). The rates of both litter accumulation and its decay depend very much on the particular habitat conditions, but in open heathland it can be the actual depth of litter (and the toxic and allelopathic substances it contains), that allows the species to dominate other plants. In deciduous woodland, on the other hand, litter decay is rapid and very little or none remains after just one year (Watt 1976).

An excessive accumulation of frond litter can prove detrimental to Bracken itself, so that it can fall victim to its own success. When this happens, the rhizome degenerates, gaps in Bracken cover occur, and this can allow colonisation by woody species and other plants (Watt 1976). It was his prolonged studies of Bracken and its competition with heathers which led Watt (1947, 1955), to the influential idea of 'cyclical succession' in vegetation, and to his describing the phases of Bracken 'sociology' in terms of a 'pioneer, building, mature and degenerate' phases that form this cycle (Watt 1964).

Between the extremes of litter persistence mentioned above, there are all degrees of dead frond litter accumulation. This phenomenon is often affected directly and indirectly by mans' interference through grazing, trampling, burning and cutting of the live or dead fronds (Watt 1976). From an ecological point of view it is probably wisest to consider Bracken as a pioneer species, that will most likely give way to colonising tree seedlings of species such as birch or oak in succession towards an eventual climatic climax forest vegetation. Within the process of succession, P. aquilinum occupies a niche as a stable, controlled, often very long-lived component species (Page 1986).

Fermanagh occurrence

In Fermanagh, Bracken shows a definite preference for acidic conditions, but it certainly is also frequent in limestone areas. This large, deciduous fern has been recorded in 303 tetrads, 57.4% of those in the VC.

Factors involved in the historic spread of Bracken

Studies in Scotland in particular suggest that colonisation by Bracken has greatly increased from about 1750 onwards, and that until at least the first three or four decades of the 20th century, farmers in areas of upland grazing noticed rapid invasion of their pastures. In this manner, Bracken has become a real menace, and one that continues to spread (Rymer 1976; Page 1997). P. aquilinum was probably originally a plant of open woodland and forest margin, occurring mainly on calcifuge terrain where a degree of shade and competition from other species kept its growth in check. Destruction of the woodland by man's activities, extending from the period of the Neolithic farmers onwards, has removed much of these constraints, allowing this vigorous rhizome-possessing species to form the extensive stands it occupies today on virtually all sorts of sloping rough ground (Rymer 1976; Page 1982(b); 1997, p. 361).

Apart from woodland destruction, numerous other factors are involved in Bracken spread. Probably most important has been the decline in rural (especially upland) human populations, and the consequent abandonment of areas of farmland previously cultivated or more or less intensively grazed. This would also account for ground where Bracken was previously cut, either for its eradication, or more purposefully for economically significant folk use of the plant (see below).

Subsidiary factors that have encouraged Bracken spread include the general move from cattle to sheep grazing, with a consequent reduction in trampling pressure; a rising rabbit population with its selective grazing pattern avoiding Bracken; poor understanding of heather burning regimes leading to a loss of heather cover; plus a general climatic rise in temperatures occurring over the last 250 years (Rymer 1976).

Colonisation of burnt ground

Bracken is an extremely invasive plant, and it is particularly so on burnt ground. The sporophyte rhizome and the sexual prothallus stage of the species are both very well adapted to rapidly colonise the almost-virgin ground of recently burnt areas in both woods and heaths, while surviving, competing plants have yet to recover from the effects of the fire. While the sporophyte plant shows a definite preference for calcifuge conditions, the prothallus stage rather unexpectedly is base-demanding. The potash and other minerals released by fire clearly provide particularly suitable conditions for rapid spore germination, and the released soil minerals promote the growth of both the Bracken gametophyte prothallus and the juvenile sporophyte arising from it after fertilisation (Conway & Stephens 1957; Page 1997, p. 362).

Grazing at either the pioneering or re-colonisation stage of vegetation development, especially close cropping by sheep, is capable of tipping the competitive balance between available species towards Bracken (Page 1982(b)).

The extent of Bracken in Britain

Estimates of the land area covered by Bracken in the United Kingdom of Britain and Northern Ireland range from 3,000-6,000 km², with the most serious infestations in upland regions of the west and north (Fowler 1993). Undoubtedly the scale of the Bracken weed problem in Britain is large and the landscape and biological conservation implications are quite frightening. The fern causes problems for agriculture, forestry, conservation, shooting interests, recreation, health and water collection (Pakeman et al. 1994). Several surveys made between 1978 and 1990 estimate that Bracken dominates approximately 3,600 km² of Britain's land area, representing around 1.5% of the total land cover. It is present, but not necessarily dominant, in around 17,000 km², or 7.3% of British land cover, and there is a considerable risk of these figures increasing due to changes in both land management and climate (Pakeman et al. 1994).

Bracken continues to spread, and Page (1997) reported this occurring in Britain at a rate of 1-3% per year, which seems an alarmingly high figure. Very probably the move during the last 100 years away from cattle to sheep grazing in upland areas has contributed to the extension of Bracken on such pastures, since apart from the differing grazing pressure, the emerging fronds can more readily survive sheep trampling than that of heavier beasts (Step & Jackson 1945).

Control measures

Efforts to control Bracken largely consist of ploughing it in, regular cutting, crushing, or the use of herbicides, especially Asulam, which is Bracken specific. However, all of these methods of attack are expensive, labour-intensive and require safe access to the land by agricultural machinery. Since Bracken often infests steep or rocky slopes, aerial application of herbicide is often the only current option for control, making it both expensive and problematic, since such widely broadcast spray may well endanger other desirable or protected species. In addition, the problem remains that unless all of the rhizome buds are destroyed, the plant will survive and reappear at a later date when control measures are eased (Taylor 1990).

Rehabilitation of sites is a very important part of conventional Bracken control programmes and this greatly adds to the cost, particularly if fencing is required to prevent access by grazing animals (Fowler 1993). A programme of experiments on biological control using two moth species specific to Bracken imported from S Africa showed definite promise during testing under semi-natural conditions (Fowler 1993). However, at the eleventh hour funding was refused by government for field trials in Great Britain (Taylor 1995). Reviewed in a global context, Bracken is encroaching and not retreating where it occurs, and within the scope of current technology and economics, it is well nigh impossible to reverse, control or eradicate the plant (Taylor 1990).

Reproduction

Under dense woodland canopy Bracken fronds are quite often sterile. However, spore production is enormous in unshaded habitats, where a single frond is capable of producing up to 30 million spores (Conway 1957). Having said this, Bracken spore production is sporadic, development being affected by plant and frond age, degree of shading, exposure, weather conditions, perhaps soil characteristics, and by the genetic make-up of the individual. In most years spore output is generally poor, at least in Britain and Ireland, and some populations appear to be consistently sterile, even when others nearby spore copiously (Dyer 1990).

Despite the potentially enormous spore production, established stands of Bracken often reproduce exclusively by vegetative means due to the amount of frond litter they produce smothering the surface of the ground and preventing spore germination. Thus spores are probably only important in the colonisation of new sites on burnt or otherwise disturbed ground (eg, animal burrows, damp hollows and lime-rich cavities in rocks, old walls and rubble (Grime et al. 1988; Dyer 1990). More work is urgently required on the colonising potential of Bracken spores and the significance of Bracken spore banks in the soil.

Bracken nectaries

As long ago as 1877 Francis Darwin first observed that Bracken plants secrete sugars through numerous foliar nectaries found all over the plant, but especially on the under-surface at the junctions between pinnae and the rachis, and also in smaller amounts at the junctions between pinnules and the pinna midrib (Page 1982(c)). The size and prominence of the nectaries varies with the particular habitat occupied, being larger on plants growing in more open sites (Page 1982(c)).

The function of sugar glands like these in Bracken and other plants (including, worldwide, a few other unrelated ferns), has been a topic of debate since the early years of the 20th century. Some biologists suppose them to be purely excretory, while others believe them to attract pugnacious ants into a mutualistic relationship with the plant, whereby in return for a food reward they protect it from herbivore attack (Tempel 1983; Page 1997). The real question here is whether or not possession of such nectaries provides the species with an ecologically significant advantage over plants without such structures? Numerous studies carried out have found that it is not easy to answer this apparently simple question.

Bracken and ants

Experimental work in New Jersey by Tempel (1983) confirmed that Bracken nectaries are most active in the young expanding frond, and that they did attract ants. However, she also showed that mature plants continue to secrete small quantities of sugar, even though levels of ant activity sharply fell away on fully expanded fern fronds. Despite indications that Bracken is adapted to some form of mutualistic relationship with ants, Tempel concluded that no such interaction actually existed in her particular geographical region, since the ants in her study were non-aggressive and they did not protect the fern from herbivore damage. This still leaves a number of open questions relating to the significance of foliar nectaries on Bracken in Britain and Ireland, amongst which must be, is it a topic worthy of further study?

Toxicity

All parts of the plant, including its airborne spores, contain carcinogens as well as various other poisons. Some of these remain toxic after the plant has been cut and dried, so that it can be a danger to both man and livestock. The carcinogenic and immuno-suppressive effects of Bracken are an active area of medical research (Cooper & Johnson 1998). As with other plant toxins, the role of poisons in Bracken is to deter herbivores and inhibit the growth of neighbouring plants (ie it is allelopathic to potential competitors). The chemical armoury of Bracken is generally extremely effective in these respects. Allelopathic toxins are contained primarily in Bracken roots, rhizome and litter, and they are released into the soil environment to suppress the growth of associated plants (Gliessman 1976).

However, not all Bracken populations possess the full complement of animal deterrent toxins, and these plants may sometimes be detected by livestock and become heavily grazed. Sheep usually avoid Bracken, but if starving they will graze it and they can then become addicted to it. The same situation applies with horses. Another Bracken constituent is known to cause thiamine deficiency in non-ruminant animals such as horses and pigs (Cooper & Johnson 1998).

Other ill effects of Bracken in pastures include the shelter it provides for the sheep ticks that transmit Louping ill virus to both grouse chicks and sheep. Sheep ticks are also implicated in the transmission of Lyme disease to a range of animals including man (Fowler 1993).

Uses

Before the health risks inherent in handling Bracken were known, Bracken was collected and used in farms for animal fodder, bedding, kindling, thatch, compost, fertiliser (on account of its potash content) and as packing material for fruit. It was even used as human food: the rhizome contains a lot of starch, although in reality it tastes very astringent (Grieve 1931; Step & Jackson 1945; Rymer 1976). Young frond croziers were previously eaten in Japan like asparagus, while in Siberia and Norway expanded fronds were used in the past along with malt to brew some dreadful form of beer (Grieve 1931). Rhizomes were dried and powdered to make flour from which bread was baked either directly, or after mixing with wheat flour, a practice which was found in native cultures as far apart as New Zealand and Normandy (Rymer 1976).

The astringent properties of the rhizome also saw it being used to dress and prepare kid and chamois leather, but although this has been reported many times from Lightfoot (1777) onwards, we do not know where or when this was ever the case (Rymer 1976). The ash of Bracken contained enough potash to recommend its use in glass making, and it was also boiled with tallow and used as soap in parts of the East.

Bracken is a light and quick-burning fuel (making it a severe fire risk when it is abundant in or near recreational areas), and it produces a very violent heat (Rymer 1976). In many parts of the British Isles, it has been used in the past for burning limestone, for heating ovens used in baking and brewing, and for firing bricks. The cutting of more or less dead Bracken fronds for fuel, thatch, bedding, packing material or other purposes, had the effect of removing some, but not all, of the fern's frost-protective litter layer. Since it was gathered late in the season, however, this will have had relatively little ecological consequence on the performance or survival of the species.

A most informative, thoroughly researched review of the ethnobotany of Bracken has been published by Rymer (1976) and it is highly recommended reading.

Names

The origins of the plant's botanical names are quite fairly described as 'obscure'. The genus name 'Pteridium' is derived from the Greek diminutive of 'pteris', fern, from 'pteron', meaning 'a wing', 'winged', (ie 'little wing'), or 'a feather', an allusion suggesting that some fern fronds resemble a bird's wing (Stearn 1992). This notion, except with respect to size, fits the expanded Bracken frond really rather well. The English term 'fern' is similarly derived from the Anglo-Saxon 'fepern', meaning 'a feather' (Grieve 1931).

The Latin specific epithet 'aquilinum' means 'eagle-like', a notion reputedly suggested to the Swedish botanist, Carl Linnaeus, by the pattern of vascular bundles that is observed when the lower stipe is cut across obliquely, which supposedly resembles a spread eagle (Grieve 1931; Gilbert-Carter 1964). For those interested, this and other similar notions of the name origins are given by Step & Jackson (1945).

Folklore and folk medicine traditions are given full accounts by Grieve (1931, p. 305) and Vickery (1985, pp. 44-45). The most interesting and widespread folklore tradition is that Bracken, or its spores, confers invisiblity (Rymer 1976).

The English name 'Bracken' (sometimes 'Brecken') is the plural of 'Brake', and apparently is derived from the Old English 'bracu', possibly referring to something broken. This would be appropriate to the dead fern in winter, forming as it does a dense tangle of broken stems (Grigson 1974). It should be noted that the name 'Brake' or 'Bracken' was used in pre-scientific days for large ferns generally, and also more particularly applied to Pteridium aquilinum (Britten & Holland 1886). Another possible origin of 'Brake', 'Brakes' or 'Bracken' is suggested by Prior (1879), who derives it from the German 'Brache' or 'Brach-feld', meaning uncultivated land or land that is breakable, or open to tillage after a term of years, ie land that is not preserved as forest. This particular etymology suggests that Bracken has long been known as an active coloniser of abandoned arable land (Rymer 1976).

Threats

Reviewed in a global context, where Bracken already occurs it is encroaching and not retreating. Within the scope of current technology and economics, it is well nigh impossible to reverse the spread in these areas, or to control or eradicate the plant from them (Taylor 1990).

Growth form and preferred habitats

An erect, medium-sized perennial fern with a creeping rhizome from which arise solitary or clumped annual fronds, T. palustris is a decidedly rare species in Fermanagh, confined to the permanently wet, peaty, but not too acidic, muddy ground dominated by sedges, alder and willow around Upper Lough Erne and several of the smaller lakes in the county. Marshes, sedge fens and wet fen-carr woods by small lakes are the typical habitats of the species.

Fermanagh occurrence

As the distribution map indicates, T. palustris is represented in ten tetrads, just eight of which have post-1975 records. It is mainly concentrated in the south of the county, in wet ground around Upper Lough Erne and beside lakes along the SE border of the county. The first Fermanagh record for this species is a recently noted early herbarium specimen in DBN which was collected by Prof Scott from near his home in Scottsborough. Marsh Fern has not been re-found at Scott's site, nor at Hart's pre-1887 station on limestone shingle by the River Erne close to Belleek (a record so vague it probably refers only to the Co Donegal side of the international border (H34)) (Hart 1898), nor at the W end of Inver Lough where it was recorded by Meikle and his co-workers in the period 1946-57.

The sites in the River Finn catchment represent the main concentration of the species in Northern Ireland. The sites in S Fermanagh from which the species has been recorded are: Derrymacrow Lough, Abacon Lough, Farmhill Lough, Clonshannagh Lough, Lough Garrow, Killynubber Lough, Inver Lough, the lakelet by the avenue at Crom and the River Finn near Gortnacarrow Bridge. T. palustris is even rarer in adjacent Co Cavan (H30), where only one of four stations has a recent (1996) record (Reilly 2001). Similarly in Co Tyrone (H36), which had two late-19th century sites for the fern, it only persists at Enagh Lough, near Caledon (McNeill 2010).

Page (1997) regards Marsh Fern as a plant of essentially Continental climatic conditions. Since Fermanagh is decidedly oceanic (or Atlantic) in its climate, and the local T. palustris sites lie within 12-25 km of the west coast of Ireland, the species must be close to the extreme margin of its range, where it is often and extensively replaced by Osmunda regalis (Royal Fern).

Irish and British occurrence

Throughout both Ireland and Britain, T. palustris is a scarce and widely scattered species. The Census Catalogue of the Flora of Ireland lists past records from a total of 21 VCs, but the map in the 1978 Fern Atlas records only 14 Irish hectads with post-1930 records (Jermy et al. 1978), while the New Atlas map plots 18 hectads with post-1987 records (Preston et al. 2002).

In Britain, while slight concentrations occur on the Isle of Wight (VC 10) and S Hampshire (VC 11) (Brewis et al. 1996), and again in Norfolk (VCs 27 & 28) (Beckett et al. 1999), Marsh Fern becomes much rarer north of a line between Hull and Liverpool (Jermy et al. 1978; Stewart et al. 1994; Preston et al. 2002).

European and world occurrence

Marsh Fern ranges widely across warm-temperate latitudes of mainland Europe stretching east to Siberia. In the north it reaches N Finland, and it stretches southward to the southern tip of the Peloponnese in Greece, the distribution thinning markedly in both directions (Jalas & Suominen 1972, Map 73). Taking the species in the broadest taxonomic sense, it is disjunctly circumpolar, with gaps in N Asia and in the eastern half of N America. The NE American and E Asia forms of the plant are now sometimes recognised as varieties of a separate species, T. thelypteroides (Michaux) Holub. A form of T. palustris s.l. is also found in S Africa, S India and New Zealand, now referred to T. confluens (Thunb.) Morton (Hultén 1962, Map 170), or to T. palustris subsp. squamigera (Schlecht) Hult. (Hultén & Fries 1986, Map 38; Page 1997). Without the insight of a trained taxonomist, I am amazed and a little disconcerted by the fact that Thelypteris palustris has been referred in the past to as many as six other genera, some nowadays totally unfamiliar, so that index searching in older texts often works much better using the more stable English common name (Hultén 1962).

Names

The genus name 'Thelypteris' is a Greek compound meaning 'Lady-fern', a name first used by the ancient Greek botanist, Theophrastus, for an unspecified fern. The specific epithet 'palustris' is Latin (given a masculine ending), and means 'of swampy places' (Gilbert-Carter 1964). The standard English common name of T. palustris is nowadays 'Marsh Fern', a folk name suggested by its habitat. Previously it was also called 'Marsh Buckler-fern' and, in the Isle of Wight, 'Ground Fern' (Step & Jackson 1945).

Threats

In Fermanagh, as elsewhere in Britain and Ireland, agricultural drainage, cultural eutrophication and scrub encroachment is making inroads on suitable habitats, and Marsh Fern is becoming increasingly rare (Hackney et al. 1992).

Growth form and preferred habitats

The triangular annual fronds of Beech Fern with their deflexed pair of lower pinnae in a different plane from the rest of the blade are quite unmistakable, but they are rarely enough seen in Fermanagh. This very misleadingly named fern never occurs under beech trees, the English common name simply being a mistranslation of the Greek 'phegos', which means 'oak'. Nevertheless, oak or beech being equally inappropriate names, this perennial fern is a plant of moist, shady cliffs and damp banks, often near streams, in upland ravine woodlands. The creeping rhizome sometimes allows the plant to develop extensive colonies in undisturbed sites (Jermy et al. 1978). Stunted fronds of Beech Fern are also found where water drips through the roots of other plants on cliff ledges, in crevices and among boulders on rocky slopes.

The species frequents a wide range of rock types and soil pH, but while it appears to prefer soil with a reasonable base content (Page 1997), it can also occur under very acidic conditions, eg on the Mourne granites in Co Down (H38). In the latter situation, it must be presumed that there is some slight base enrichment, however undetected it remains. A degree of inaccessibility tends to assist the survival of this fern since P. connectilis is known to be intolerant of grazing (Sinker et al. 1985).

Fermanagh occurrence

This creeping, rhizomatous fern has been recorded in a total of eight Fermanagh tetrads, but only five of them have post-1975 records.

P. connectilis was first reported in Fermanagh in the grounds of Brookeborough Deerpark by Smith in 1860. Meikle and his co-workers refound it there in the 1950s, but it has not been seen since then at this station (Meikle et al. 1975).

The most interesting site in Fermanagh for the Beech Fern was found by the Rev W.B. Steele in 1929 on the S shore of Lower Lough Erne at Carrickreagh. On its first discovery, it was remarkably abundant on a relatively dry, flattish, limestone woodland floor, under mixed oak, birch and hazel. It remained abundant until 1945 when the site was largely destroyed by the clear felling of the woods and an extension of the nearby quarrying operation (Carrothers et al. 1946). Praeger visited the Carrickreagh site in 1933 and described it in enthusiastic terms, "The fern grows here on limestone rubble thinly covered with humus and mosses, among Primula, Endymion, Hedera, Lysimachia nemorum, Thymus, and Sesleria, forming dense patches up to 20 feet [6 m] across with fronds up to 2 feet [60 cm] high." (Praeger 1934a). In his book The botanist in Ireland, he commented on the Beech Fern at the Carrickreagh site as, "the only habitat of the kind which I know in Ireland, its characteristic stations being wet chinks or ledges in the mountains" (Praeger 1934i). He visited it again around 1938 and reported, "More abundant in the woods of Carrickreagh than anywhere else I have seen it in Ireland; one dense patch measured 250 ft [76.2 m] by 50 ft. [15.24 m]." (Praeger 1939). At present, just one tiny patch survives in this area, consisting of just a couple of fronds on the bank of a stream.

Other current local Fermanagh sites include a strong lowland colony on the Bannagh River near some waterfalls, and the fern also maintains a precarious existence on high ground as tiny fronds in rock crevices on the north face of Cuilcagh mountain.

Irish occurrence

Elsewhere in N Ireland, the species is rare and widely scattered, although locally plentiful in the Mourne Mountains and the wooded Antrim Glens (Hackney et al. 1992). It has many fewer modern stations in N Ireland than was previously the case and a rather similar situation pertains in the Republic of Ireland (An Irish Flora 1996; New Atlas).

British occurrence

P. connectilis is widespread and locally frequent in upland parts of the N and W of Britain, the distribution thinning markedly further south (New Atlas). In these regions, it is most frequent in ancient woodlands dominated by Quercus petraea (Sessile Oak) developed over neutral to acidic soils. In steeper, less accessible gullies in these woods, it frequently occurs on deeper soils percolated with base-rich water (R.J. Cooke, in: Preston et al. 2002).

Throughout the British Isles, P. connectilis tends to follow the distribution of a mountain type of climate, ie cool and with frequent precipitation and high humidity in summer when the fronds are present and growing, and fairly cold in winter when they are not (Page 1997). The summer regime clearly applies throughout the oceanic area of W Ireland including Fermanagh, while the winters in this area are very much milder than in mountainous areas of Great Britain.

The creeping rhizome appears to grow slowly (measurement of just how slowly would make an interesting project), yet the plant is capable of forming very large patches, fully occupying moist, sheltered, generally sloping sites. It therefore appears likely that the plant is long-lived, and thus is an excellent indicator of long-undisturbed sites, perhaps in some instances, with a timescale measured in thousands of years (Page 1997). The evidence gathered for the BSBI New Atlas survey immediately prior to 2000, indicates that the distribution appears stable (Preston et al. 2002).

Reproduction

The nucleus of P. connectilis plant cells contain three sets of chromosomes (ie they are triploid). While the species does manage to produce the gametophyte generation, cell division is unbalanced when meiosis occurs and thus it cannot form normal gametes (ie male and female sex cells). Nevertheless, the fern shortcuts the sexual process and produces new sporophyte plants without fertilisation taking place (ie it reproduces apogamously) (Page 1997).

World occurrence

Beech Fern is widely distributed throughout northern and central temperate parts of Europe and western Asia (Jalas & Suominen 1972, Map 74). Forms of it, including the diploid and tetraploid parents of the British and Irish triploid form, extend it widely around the northern hemisphere making it circumpolar (Hultén 1962, Map 107; Hultén & Fries 1986, Map 39). The taxonomic uncertainties of this group of ferns can be appreciated when one sees that Phegopteris connectilis has belonged in the past to no less than six other genera, including Dryopteris, Thelypteris, Lastrea and Polypodium (Hultén 1962; Hultén & Fries 1986).

Names

The genus name 'Phegopteris' was invented by the Swedish taxonomist, Linnaeus, the Greek 'phelos' referring to a species of oak, not the beech, though the word is cognate with the Latin 'fagus', the name of the Beech tree. The second part of the genus name is also Greek, 'pteris', meaning 'fern' (Gilbert-Carter 1964).

The Latin specific epithet 'connectilis' means 'well connected' (Hyam & Pankhurst 1995). An alternative English common name for the plant is 'Long Beech Fern' (Hyam & Pankhurst 1995) which, like the much more frequent 'Beech Fern', is a 'book name' rather than derived from folk usage (Britten & Holland 1886). The fern does not appear to have had any uses and there is no folklore associated with it.

Threats

The remnant of the Carrickreagh site is threatened by cattle trampling, but the populations at the other Fermanagh sites are kept safe by their remote nature.

Growth form and recognition

O. limbosperma is a rhizomatous, deciduous species that can be very readily overlooked, being easily mistaken for young Dryopteris filix-mas (Male-fern) or D. affinis (Scaly Male-fern), although the shorter, more slender, less scaly stipe, the pinnae tapering nearly to the extreme frond base and the naked, marginally set sori are all useful distinguishing features. The young fronds of O. limbosperma also give off a distinctive lemon or orange scent when lightly brushed, although as always, appreciation of this depends very much on the individual's sense of smell (Webb et al. 1996).

Preferred habitats

O. limbosperma is a strongly calcifuge plant of open sunlight, but at the same time it often occupies more or less sheltered situations. In terms of soil, it prefers damp to moist, peaty slopes which are sufficiently steep to create either surface run-off or sub-surface seepage after typically frequent rainfall. Moving groundwater is characteristic of the specific habitat and Lemon-scented Fern is reckoned to be indifferent to lime, and sensitive to both frost and summer heat (Jonsell et al. 2000). This set of growing conditions creates the cool, constantly moist, well-aerated root and rhizome environment which O. limbosperma demands, and as a result the species is most frequently found on rather steep peaty banks beside mountain or moorland streams, or in open, acidic woods at somewhat lower altitudes (Jermy & Camus 1991; Webb et al. 1996; Page 1997).

In Flora Nordica (Volume 1), the authors describe the habitat of this fern as oligotrophic forest and heath, ie nutrient poor, unproductive vegetation in terms of growth rate and biomass (Jonsell et al. 2000, p. 50).

Fermanagh occurrence

This is an exceedingly rare and vulnerable fern in Fermanagh, there currently being very few plants in the county spread over just three widely spaced sites. Previously, there were two other old stations in the county, but the fern is probably extinct in these now. O. limbosperma has been known in Fermanagh from Brookeborough Deerpark since the 1860s, and it is still there, having been refound in 1981 by a waterfall on the edge of the forest. Very sparse colonies, each of a few fronds, have also been discovered in two new sites in the last 30 years: at Tullynanny Lough in the SW of the county (Scannell, M.P.H.; June 1974), and in a ravine beside a stream at Stranahone in the NE (RHN & RSF; August 1994). It has not been seen, however, at West & Tetley's 1899 site, simply listed as "Florencecourt", nor at Praeger's similarly vague site, described as, "two stations on the lower hills W of Church Hill" (Praeger 1904).

Irish occurrence

In N Ireland, Lemon-scented Fern is rare and extremely local, and as with Phegopteris connectilis (Beech Fern), it is very much centred on the Mourne Mountains, Co Down (H38), in a few glens in NE Antrim (H39), and in the Sperrin Mountains and surrounding moorlands of Cos Londonderry and Tyrone (H40 & H36) (Hackney et al. 1992; Northern Ireland Flora Website 2014). In the Republic of Ireland likewise, Lemon-scented Fern is rare and very widely scattered, although it is locally abundant in a few sites in mountainous regions such as Connemara and Wicklow (Jermy et al. 1978; Page 1997; New Atlas).

British occurrence

O. limbosperma is much more frequent in suitable habitats in Britain than in Ireland, the species increasing in abundance as one travels both westwards and northwards. This trend culminates in W Scotland having the greatest abundance of Lemon-scented fern in Britain and Ireland, and probably also on a world scale (Jermy et al. 1978; Page 1997; New Atlas). The New Atlas survey prior to 2000 found that the distribution in upland areas of Britain remains stable, and concluded that many of the losses in lowland areas that occurred before 1930 resulted from the destruction of heathland habitat (T.D. Dines, In: Preston et al. 2002).

European and world occurrence

On mainland Europe, the distribution of O. limbosperma most closely resembles that of Dryopteris dilatata (Broad Buckler-fern), occurring predominantly in NW Europe, more or less continuously from the Pyrenees northwards to western Norway and up to just within the Arctic Circle. It also occurs further south on the Atlantic isles of the Azores and in Madeira. On a world basis, it is regarded as circumpolar, but it only manages this in an extremely disjunct manner (Hultén 1962, Map 144; Jalas & Suominen 1972, Map 72). Despite its very wide but sparse geographic distribution, in terms of abundance W Scotland still very probably represents the world headquarters of what really looks like a very fragmented relict species (Page 1997).

Comparison of Irish and Scottish presence

The very extensive and abundant presence of O. limbosperma in western and central Scotland contrasts so sharply with its rare and sparsely spread occurrence in Ireland, and especially so when compared with N Ireland in particular, that one cannot but wonder at the scale of the difference and ponder on the possible reason or reasons. The climate, the rock structure and geological history of the two regions are strikingly similar and are intimately linked, the geology possibly most obviously so with respect to their Tertiary igneous activity (Whittow 1974, 1977). This being so obviously the case, the speculation by Page (1997, p. 275) on the different representation of O. limbosperma in the two regions seems unlikely to offer an explanation, since in Ireland he suggests poorer drainage in lowland stations might be responsible, and at higher altitudes he believes that there are more base-rich sedimentary rocks than in reality is the case.

As generally happens, however, it is easier to criticise others' explanations than it is to suggest a better alternative! A careful examination of the habitat requirements and tolerances of the species might generate a fresh hypothesis, but we should not ignore the historical factors either, including for instance, differing human population pressures and land management regimes stretching across many centuries. The earlier and almost total deforestation that took place in Ireland, for example, could well be significant, as might differing grazing, tillage and burning patterns. It is difficult to imagine anything that would affect O. limbosperma on quite the scale necessary to account for the enormous difference in its presence evidenced on the New Atlas British and Irish hectad map (Preston et al. 2002).

Names

The genus name 'Oreopteris' is a combination of the Greek words, 'oreos' meaning 'mountain' and 'pteris' meaning 'fern' (Gilbert-Carter 1964). The specific epithet 'limbosperma' is derived from the Latin 'limbatus' meaning 'a border' or 'margin', and 'sperma' meaning 'seed' (or in this case, 'spores'), referring to the characteristic marginal position of the sori on the underside of the frond (Gledhill 1985).

As with Phegopteris connectilis (Beech Fern), Oreopteris limbosperma has moved through six genera over the years and, partly as a result of this has also been given a long sequence of English common names including 'Mountain Fern', 'Mountain Buckler Fern', 'Heath Fern', and in reference to the scented glands, 'Sweet Mountain-fern', 'Scented Fern', 'Lemon-scented Fern', 'Hay-scent Fern' and 'Tea-scent Fern', the latter two names being local to Cumberland (Step & Jackson 1945).

There does not appear to be any tradition of use of this fern or folklore associated with it as the species was not commonly differentiated from the Buckler-ferns, Dryopteris species.

Threats

The minimal populations in Fermanagh could very easily be eliminated by any disturbance of their streamside habitats. This is very clearly a species in urgent need of local habitat management to encourage and support its survival.

Growth form and preferred habitats

A characteristic leathery, entire-leaved, strap-shaped, wintergreen perennial fern shuttlecock appearance, Hart's-tongue is a plant of more or less permanently damp, base-rich habitats. The species is very widely distributed and common throughout most of Britain and Ireland. P. scolopendrium is essentially a plant of lowland areas, rarely found above an altitude of 185 m (Jermy & Camus 1991; Page 1997). It is most frequent on limestone and scarce or rare in upland areas where peat cover is extensive. Typical habitats include woods, roadside banks, shady rocky slopes and damp walls.

In parts of Britain and Ireland, where limestone pavements occur, P. scolopendrium is particularly abundant and luxuriant in the cool, moist, shade, shelter and protection of narrow fissures (called grykes or grikes) between the limestone blocks (clints). It is also well developed on damp, shaded limestone cliffs, where it is often overhung by hazel or ash, the frond blades sometimes in such habitats reaching 60 cm or more in length.

The plant is sufficiently shade-tolerant to be a typical species of ash and other mixed, damp, deciduous woods, especially where these occur on the broken, rock strewn ground of talus slopes below cliffs, or in sheltered, narrow valley woodlands. It is also quite commonly found rooted in artificial habitats, such as the mortar of old, damp walls, especially those of bridges, both those over rivers and of the disused or neglected railway variety, and around old wells.

Growth rate and reproduction

The entire, long, tongue-like, sterile fronds are produced each spring from the basal rhizome and they usually survive about a year before browning off and dying. The base of the stipe is thickened and serves as a storage organ (Jonsell et al. 2000). Individual plant rosettes are very slow-growing, and they generally take between 2-5 years to produce their first fertile frond, which in appearance is like the sterile fronds. The fern rosette is long-lived, individuals perhaps surviving for several decades, if not longer. The blades of large fertile fronds may bear up to 60 lines of paired sori either side of the midrib, and thus have the ability to produce vast numbers of spores (Page 1997).

Fermanagh occurrence

Hart's-tongue is very widely distributed and common throughout Fermanagh being present in 415 tetrads, 78.6% of those in the VC. It is most frequently found on the limestone, and is scarcest on uplands with extensive peat.

Irish and British occurrence

Although P. scolopendrium is very widespread in Ireland, England, Wales and lowland Scotland, it is most abundant in W and SW parts of these islands, where conditions of general high humidity and a long, mild growing season are conducive to its growth and individual frond survival (Jermy et al. 1978; Wardlaw & Leonard 2005).

European occurrence

In Europe, P. scolopendrium has a widespread and more or less continuous southern sub-Atlantic distribution. It extends in a scattered manner from extreme SW Norway to the Iberian peninsula, NW Africa, Madeira, Canaries and the Azores, then eastwards to Turkey and the Caspian Sea, but is only abundant in the British Isles and on the seaboard of W Europe south to the Pyrenees (Jalas & Suominen 1972, Map 102; Page 1997).

World occurrence

Elsewhere in the world, Hart's-tongue has a very disjunct distribution, with a different chromosome count in eastern N America (diploid whereas the form in Britain and Ireland is tetraploid). What may or may not be the same diploid taxon (or varieties allied to it) occurs in N Japan and Mexico. It is probably best to consider these three widely separated taxa as at best subspecies of P. scolopendrium (Hultén 1962, Map 147; Page 1997). In Germany Denmark and Switzerland, and probably in other sites close to its European distributional limits, the distribution of P. scolopendrium has contracted in the last 50 years and it has become locally extinct in areas where previously it was considered native (Hultén 1962; Welten & Sutter 1982; Jonsell et al. 2000).

The predominance of P. scolopendrium in mild, western, oceanic areas of Europe, where the climate is ameliorated throughout the year by the warming influence of the Gulf Stream, is undoubtedly associated with this slow-growing fern's requirement for a long growing season and constant high humidity to prevent desiccation. In its more heavily shaded habitats, much of the plant's photosynthetic gain probably takes place during the autumn and winter when higher levels of illumination operate after deciduous leaves fall (Page 1997).

Cultivated forms

P. scolopendrium is genetically very variable and easy to cultivate, so it is not surprising that it has given rise to many cultivated variants. During the Victorian 'Fern Craze', Lowe's Ferns British and exotic listed 65 cultivars of this species (Lowe 1865). The modern Royal Horticultural Society Dictionary Index of Garden Plants lists a mere 16 cultivars, most of which feature fronds with frilled margins such as var. 'crispum' (Griffiths 1994). Jones (1987) does mention that hundreds of cultivars are known, a rough estimate also given by Step & Jackson (1945).

Locally, an unusual form of the plant with fronds that fork at their tips has been recorded on limestone pavement on Heron Island, just off Tully Castle on the shore of Lower Lough Erne.

Names

The name 'Phyllitis' is the ancient name given to this fern by Dioscorides, which it has retained ever since, although occasionally it wanders temporarily into genera such as Asplenium and Scolopendrium (Gilbert-Carter 1964). The large number of paired sori either side of the midrib on the fertile frond gave rise to the Latin species epithet 'scolopendrium', a name given by the classical herbalist Dioscorides, that compares the plentiful sori to a millipede's leg count (Gilbert-Carter 1964).

It is a rather sad fact of life that fern Latinised botanical names have in the last 40 or 50 years been much less stable than their English common names. The most frequent English common name of this distinctive, entire-leaved fern is 'Hart's-tongue', which first appeared in print as the Middle English 'hertes tongue', in the Grete Herball (Anonymous 1526, see mention of its origin under Ceterach officinarum) (Ryden 1984). The name obviously refers to the tongue-like shape of the frond blade, and is a translation of the apothecaries' Medieval Latin, 'Lingua cervina' (Gerard 1633; Prior 1879). The tongue name also appears in several variants in parts of the British Isles as, 'Adder's Tongue', 'Fox-tongue', 'Horse-tongue' and 'Lamb's-tongue'. Additional names suggesting blade length are 'Long-leaf', 'Finger-fern', 'Seaweed-fern' and 'Snake-leaves'.

Two unusual local names refer to other specific physical features; 'Button-hole' from E. Sussex, rather aptly refers to the appearance of the young sorus on the back of the frond, while 'Christ's Hair' or 'God's Hair', names that originate from Guernsey, Devon and elsewhere. The ‘hair’ in the name is said to refer to the black fibrous bundles (one or two) present in the vascular tissues of the stipe. Both the name and the explanation of its supposed origin are provided by Britten & Holland naming Mr W. G. Piper as the source of the botanical information. The associated story, as quoted in Vickery (1995), is charming and goes as follows: "The fern was once the pillow for the Son of Man, when He had nowhere to lay His head. In return for this service, He left two hairs of His most blessed and dear head, which the plant treasures in her ripe stem, as His legacy – two auburn hairs which children find and show."

Herbal medicinal uses

The name 'Burntweed' originates from Westmeath, Wales and the Scottish Highlands, and refers to the local use of the frond to manufacture an ointment for the treatment of burns, scalds and piles (Britten & Holland 1886; Grieve 1931). Other medicinal uses in herbalism were as an astringent treatment for diarrhoea and dysentery, and as a remedy for removing obstructions from the liver and spleen (Grieve 1931).

Threats

None.

Growth form and preferred habitats

Black Spleenwort is a very variable evergreen perennial of well-drained, somewhat base-enriched rocky places, including limestone screes and dolomitised sandstone scarps, especially in more western, coastal and lowland parts of Britain and Ireland. The plant typically produces a loose tuft of triangular fronds from a short, creeping rhizome. It favours sheltered, lightly shaded situations where competition is reduced for a variety of reasons. In very sheltered, more deeply shaded woodlands and hedgebanks, frond length can occasionally reach 50 cm. Reduced competition often involves fern habitats with little soil such as, for instance, tree-shaded rock crevices on cliffs and on old quarry faces. Here, fronds sometimes reach up to 30 cm in length, while in more exposed sites with similar strictly limited soil resources they are always very much smaller, often only 10 cm in length or less.

Black Spleenwort also occurs, though not quite so abundantly, on limestone cliffs, and together with the modified sandstone rock mentioned, this behaviour reflects this species known soil nutrient requirement for at least a trace of base-rich elements, most likely calcium or magnesium (Jermy et al. 1978; Jermy & Camus 1991). This minimal but necessary base requirement excludes A. adiantum-nigrum from acidic siliceous rocks, such as quartzite, granite or normal sandstone (Webb & Scannell 1983), but it is not the case, as is sometimes claimed, that this species avoids limestone (Hultén 1962; Sinker et al. 1985), at least not in Fermanagh, nor in the Burren, Co Clare (H9).

Apart from hedgebanks, the fern occasionally occurs in other artificial habitats, such as on old lime-mortared walls and bridges in lowland areas.

Variation and taxonomy

In recent years, A. adiantum-nigrum has been recognised as having two subspecies: the common and widespread subsp. adiantum-nigrum, and a second one, to some extent associated with, but perhaps not totally confined to ultra-basic (especially serpentine) rock, subsp. corrunense Christ (Page 1997). The latter was previously confused with the continental European species, A. cuneifolium, and its relationship with this species remains the subject of continuing research. Serpentine and other ultrabasic (or ultramafic) rocks often supply toxic or near-toxic levels of heavy metals such as nickel, cobalt and chromium. The derived soils are extremely infertile, which greatly reduces plant vigour and competition (Brooks 1987). Within the Fermanagh western plateau, there is a crescent-shaped outcrop of intrusive dolerite and basalt which just might support subsp. corrunense, and while it has not yet been discovered, it should certainly be looked out for (Woodland et al. 1977).

Fermanagh occurrence

Black Spleenwort is a characteristic species of the dolomitised, somewhat base-enriched, sandstone scarps in the more upland SW of Co Fermanagh that is referred to as the Western Plateau. It has been recorded in 73 tetrads in the VC, representing 13.8% of the total area. The occurrence is most unevenly spread however, the fern being predominantly confined to the western half of the county. The most elevated site in Fermanagh for Black Spleenwort is at Cuilcagh Gap, around 550 m. Otherwise, the generally lowland pattern that this species displays elsewhere in the British Isles is reflected in Fermanagh (Page 1997). It also occurs, but only very occasionally, on old lime-mortared walls and bridges in the lowlands. Local examples of this occur at Tubbrid churchyard and on the old bridge at Roogagh River.

British and Irish occurrence

A. adiantum-nigrum s.l. rarely occurs in large populations, but is widely distributed throughout these isles, being most frequent in the S and W, and especially so in mild, coastal districts where high levels of humidity and illumination are the norm (Page 1997; Wardlaw & Leonard 2005).

European and world occurrence

A. adiantum-nigrum s.l. is widespread in Europe north to 63^o on the W coast of Norway, and south to the Peloponnese, NW Africa, the Caucasus, N Iran, C Asian mountains, SW North America and is also found on some isolated tropical mountains and islands (Jalas & Suominen 1972; Jonsell et al. 2000). The northern hemisphere map of the species (Hultén (1962, Map 142), shows that it absent from vast tracts of temperate and boreal Asia and N America, to such an extent that to this author’s mind it does not warrant inclusion in an atlas of circumpolar plants. The later publication of A. adiantum-nigrum s.l. world distribution (Hultén & Fries 1986, Map 44), again highlights how very far removed the species is from being circumpolar. Preston and Hill (1997) classified the fern as European temperate, at the same time noting that it has an additional very restricted presence in N America, occurs in C Asia and also in widely disjunct parts of the tropics and the Southern Hemisphere, including Australia and Hawaii (Hultén & Fries 1986).

Names

'Asplenium' is derived from the Greek 'a' meaning 'not' and 'splen', 'splene' or 'splenon' referring to the spleen, alluding to the supposed medicinal properties of the fern genus. The herbal medicinal use is also invoked by the English common name applied to the genus, ‘Spleenwort’ (Hyam & Pankhurst 1995). The specific epithet is ‘adiantum-nigrum’. ‘Adiantum’ is Greek, ‘a’ meaning ‘not’ and ‘diantos’, moistened, and thus the combination ‘adiantos’ means ‘dry’, ‘not wetting’ or ‘unwetted’, referring to the fact that fronds of the fern genus ‘Adiantum’ remain unwetted under water. ‘Nigrum’ means ‘black’, so ‘adiantum-nigrum’ translates literally as ‘dry black’ or ‘unwetted black’. However, as the fern Adiantum capillus-veneris has the English common name ‘Maidenhair Fern’, Asplenium adiantum-nigrum in the past was given the book name ‘Black Maidenhair Spleenwort’. Since there already is a fern with the given (book) name ‘Maidenhair Spleenwort’ (Asplenium trichomanes), to avoid confusion A. adiantum-nigrum is most usually referred to as ‘Black Spleenwort’ (Step & Jackson 1945).

Although A. adiantum-nigrum is very widely distributed in both Britain and Ireland, it does not appear ever to have had a genuine folk-name or English common name, only the invented, given, book names listed above. Lyte (1578), in his ‘Niewe Herball’ mentions this fern under the names ‘Black Oak-fern’ and ‘Petty-fern’, but in reporting this information Step and Jackson (1945, p. 51), express their doubt that these names would have been in use among the people at that early date.

Herbal medicinal uses

Black Spleenwort is said by herbalists to have similar medicinal virtues to other Maidenhairs, “a decoction of it relieving a troublesome cough and proving also a good hair wash. Dosage of infusion: 3 tablespoonfuls” (Grieve 1931, p. 303). Allen and Hatfield (2004) reported that in Ireland a cough cure known as ‘maidenhair’ was once popular among country people in Londonderry (David Moore unpublished report 1834-5). These latter authors assumed that this referred to Maidenhair Spleenwort (A. trichomanes), but it might equally well have been A. adiantum-nigrum that was used. Adiantum capillus-veneris does not come into the question on this matter since it was always a rare plant of very restricted distribution.

Threats

None.

Growth form and preferred habitats

Except at the western extremities of its markedly Atlantic distribution, A. marinum is restricted to a very narrow zone, seldom more than 20-30 m above sea level, where the winter air temperature is ameliorated by warm sea spray from the Gulf Stream. The usual habitat requirements of this glossy, evergreen, singly-pinnate, perennial fern are a cool, moist crevice, sheltered from full sun in summer and entirely frost-free in winter (Page 1997).

British and Irish occurrence

Sea Spleenwort is entirely coastal throughout its British and Irish range. In Britain it stretches almost continuously from the Isle of Wight on the middle of the south coast of England, westwards to Land's End and up most of the west coast through Wales to Shetland. On the east coast of Britain it is again represented from the far north of Scotland as far south as Scarborough, although here it is less continuous and definitely more thinly scattered.

In Ireland, the New Atlas hectad map shows A. marinum is again very well represented on almost all of the S & W coasts, but it is absent from a few stretches of the E coast, mainly from Dublin Bay to Wexford (New Atlas).

Fermanagh occurrence

In view of this information, the solitary Fermanagh site in Lough Navar Forest Park is extremely abnormal, being 17 km inland from the tidal estuary at Ballyshannon and situated at an altitude of 210 m. Originally the plant occurred at two places on the same set of N-facing sandstone cliffs, there being about 20 plants at one site and about 50 at the other (Northridge et al. 1988). On a visit in July 2002, there was a very noticeable decline in the smaller more accessible population, the result of aggressive competition from young Hedera helix (Ivy) stems also present in the crevices. Subsequent visits up to September 2010 showed three plants surviving at the smaller patch; the larger patch was thriving. In addition, two inaccessible plants growing higher up the cliffs were identified through binoculars!

Ecology

The Fermanagh sites being so remote from the sea, it cannot be that A. marinum has an absolute requirement for sodium chloride. The red sandstone of the Lough Navar cliffs has been dolomitized by the vertical percolation of waters rich in calcium and magnesium sulphates, so that mineral replacement has occurred and the rock has become base-enriched. At both Fermanagh sites, A. marinum grows where the top of the cliff overhangs the base to a considerable degree, providing a measure of frost protection, although one would think that this would hardly be sufficient to entirely avoid freezing temperatures at this altitude. Many of the plants grow in a shaded pocket of the cliff while other small individuals grow along fault line crevices. A. ruta-muraria (Wall-rue) grows close to some of the A. marinum plants, but many have no nearby competitors.

The winter-green, coarsely divided, leathery, glossy fronds and stiff, dark purple-brown stipes are readily recognised. It is the opinion of the current author that the thick cuticle of the fleshy frond enables A. marinum to withstand the extreme dryness of this site. A visit to the site in July 2002 found that Ivy had invaded some of the cliff crevices and rendered A. marinum locally extinct.

Fortunately another cliff outcrop further west maintains a vigorous population of the species at its base, and the Ivy stems appear to have grown up and beyond the fern, so that they are unlikely to compete directly for space and light.

Reproduction and possible hybridization

The fronds produce spores abundantly from August onwards and overwinter before decaying the following spring as new fronds are produced (Page 1997). The breakdown in the Fermanagh site of the usual ecological barriers between A. marinum and A. ruta-muraria suggests the possibility of a hybrid, but none has yet been found here. Page (1997) suggests the reason for this failure to hybridize with any other native northern species of Asplenium is that A. marinum's phylogenetic affinities appear to lie with an extensive group of species characteristic of wet forests in the tropics and sub-tropics.

European and world occurrence

The wider distribution of A. marinum is Atlantic-Macaronesian (ie along the W and SW coasts of Europe, plus the island groups of Madeira, Canaries and Azores (Jalas & Suominen 1972, Map 78; Page 1997, p. 72). This, together with its essentially frost-free ecological requirement, tends to support the idea that it associates more closely with Asplenium species of warmer climates than those encountered at present in the British Isles. In their phytogeographical survey of British and Irish plants, Preston and Hill (1997) classified A. marinum as Suboceanic southern-temperate, although they noted that it also occurs in the Southern hemisphere. Hultén and Fries (1986, Map 40) indicate that this refers to presence on the Cape Verde Islands and St. Helena.

Trends in Britain and Ireland

Page (1997) points out that in general A. marinum is much less frequent and less luxuriant in many of its current British and Irish habitats than it was in the 19th century. Past ravages of fern collection, together with slowly changing factors, for example, pollution and contamination of shores, and natural changes in climate, that is, global warming, with respect to which this species appears to be in an extremely delicate balance, are almost certainly involved. Thus this species is under threat, and should on no account be collected.

Names

'Asplenium' is derived from the Greek 'a' meaning 'not' and 'splen', 'splene' or 'splenon' referring to the spleen, alluding to the supposed medicinal properties of the fern genus. The herbal medicinal use is also invoked by the English common name applied to the genus, ‘Spleenwort’ (Hyam & Pankhurst 1995). The Latin specific epithet ‘marinum’ means ‘sea’, an obvious reference to the preferred habitat and hence the English common name ‘Sea Spleenwort’.

Like A. adiantum-nigrum, while A. marinum is widely distributed along coastal sites in both Britain and Ireland, it does not appear ever to have had a genuine folk-name or English common name, only the invented book name, ‘Sea Spleenwort’. Presumably it is too uncommon to have merited folk interest, and I cannot locate any mention of folk use of the plant.

Threats

Part of one of the Fermanagh sites is being overgrown by Hedera helix (Ivy).

Growth form and preferred habitats

Maidenhair Spleenwort is a small, distinctive, evergreen perennial species that colonises crevices on upland limestone outcrops and cliffs. In lowland areas it frequently occurs on the mortar in walls. It is most abundant and luxuriant when the habitat offers high humidity and it becomes distinctly stunted in drier, more exposed sites.

Taxonomy

Asplenium trichomanes really consists of a complex polyploid series of forms, within which a simplified account segregates three ecologically and morphologically recognisable subspecies in Britain and Ireland. All of the Fermanagh records almost certainly refer to by far the most common of the three forms, the tetraploid A. trichomanes subsp. quadrivalens D.E. Mey. (but see the separate account below). Essentially this is a calcicole subspecies, but to a surprising degree it can tolerate habitats with soils having very little calcium present (Jermy & Camus 1991; Jonsell et al. 2000).

Fermanagh occurrence

In Fermanagh, this fern species has been recorded in 181 tetrads, 34.3% of those in the VC. Eight scattered tetrads have pre-1976 records only. It is widespread throughout the county, but is especially frequent in crevices on upland limestone outcrops and cliffs. It also occurs less commonly on sandstone scarps in the Lough Navar area which have been dolomitized by seepage of water rich in both calcium and magnesium carbonates derived from overlying base-rich mica-schists.

In lowland areas of Fermanagh, A. trichomanes is restricted to man-made habitats such as the lime-rich mortar in old walls, bridges, basement areas around older houses, walled gardens and the like.

British and Irish occurrence

A. trichomanes subsp. quadrivalens is common throughout Britain and Ireland, but it is perhaps slightly less frequent in Ireland than in Britain (Jermy & Camus 1991). Until the publication of the New Atlas in 2002 there were insufficient data to permit the subspecies to be mapped separately. The New Atlas maps and those in the New Fern Atlas (Wardlaw & Leonard 2005), both suggest that only subsp. quadrivalens occurs in Ireland. The other two subspecies are very much less common in Britain, and subsp. pachyrachis (H. Christ) Lovis & Reichst. is recorded from just eight hectads in England.

European and world occurrence

The collective species, A. trichomanes is common and widespread in W and C Europe, thinning out to both north and south although reaching well within the Arctic Circle. However, it is currently declining in Scandinavian countries (Jalas & Suominen 1972, Map 81; Jonsell et al. 2000). In the widest sense, A. trichomanes is a circumpolar southern-temperate species (Hultén 1962, Map 130; Preston & Hill 1997), but is also very well distributed around the southern hemisphere, being present in S Africa, C and S America, New Guinea, S Australia and New Zealand (Hultén 1962, p. 138; Hultén & Fries 1986, Map 41).

Names

'Asplenium' is derived from the Greek 'a' meaning 'not' and 'splen', 'splene' or 'splenon' referring to the spleen, alluding to the supposed medicinal properties of the fern genus. The herbal medicinal use is also invoked by the English common name applied to the genus, ‘Spleenwort’ (Hyam & Pankhurst 1995).

The Latin specific epithet ‘trichomanes’ is the genus name given to Maidenhair Spleenwort by Theophrastus, the ancient Roman doctor. ‘Trichomanes’ is Greek and it refers to a thin hair or a bristle (Johnson & Smith 1946). After the green pinnae die and drop off the black or deep brown stipe of the frond, the bare remainder adds to the short, dense, wiry tuft of old stipes attached to the plant. Certainly, no female would be flattered to have the hair on her head compared to the wiry tuft of a Maidenhair Spleenwort plant. It follows that ‘Maidenhair’ must refer to hair located elsewhere on the body, and it is not too difficult to imagine where.

Uses

A tea made with the fronds was described as ‘ sweet, mucilaginous and expectorant’ and was used in herbal medicine to treat lung disorders. It was also considered a laxative (Grieve 1931, page 303).

Threats

None.

Growth form and preferred habitats

This little evergreen perennial fern is generally described as a calcicole species, ie one which is lime-tolerant and/or preferring or requiring base-rich conditions. The frond has a delicate texture and the green rachis is extremely distinctive. A. viride requires much higher levels of humidity than A. trichomanes (Maidenhair Spleenwort) and will grow in conditions of considerably deeper shade. Unlike A. trichomanes and its subspecies, the rachis of A. viride does not shed its pinnae, but rather when it withers the whole frond shrivels and only a very short persistent brown base remains (Page 1997; Jonsell et al. 2000).

Normally the species does not tolerate drought or high summer temperatures, and while in natural rock habitats it regularly associates with A. trichomanes s.l., A. ruta-muraria (Wall-rue) and Cystopteris fragilis (Brittle Bladder-fern), unlike these three ferns it is generally absent from walls. All Fermanagh records for this species are from naturally occurring rock habitats (ie cliff faces, crevices, ledges and limestone grykes and swallow holes), but in some areas of Britain and Ireland A. viride has also been rarely found growing on lime-mortared walls (Jermy & Camus 1991).

Fermanagh occurrence

In Fermanagh, Green Spleenwort is rare, never abundant and has been recorded in a total of just eight tetrads. Although a rare species, it is most abundant in several of the deep limestone swallow holes near Legacurragh above Florencecourt in the south of the county, where it grows as tufts on the vertical faces of the rock. It also occurs sparingly on both the dolomitized Upper Visean sandstone and shale, and the dolerite and basalt scarps in the Lough Navar area (Woodland et al. 1977). As the tetrad distribution map indicates, in addition to the rather isolated Legacurragh site, A. viride also has a second outlier on an isolated scarp near Lough Alaban in Tullyloughan Td, Carrigan Forest. The fern population here consists of just six plants. The range of substrates mentioned demonstrates that while Green Spleenwort always occurs on soils derived from base-rich rock, it does not necessarily require a high calcium content in the soil it occupies (Jermy & Camus 1991).

British and Irish occurrence

In Britain, Green Spleenwort has a distinctly N, W and upland distribution, with the greatest concentration of records occurring N of a line from Morecambe to Bridlington. In Ireland, by comparison, it is very much less frequent overall, and it is distinctly western and montane, being confined to moist, shaded habitats in areas with low summer temperatures (Jermy et al. 1978).

European occurrence

In Europe, A. viride is widespread on calcareous and other base-rich rocks, but further south it is found mainly in the mountains, while in Fenno-Scandinavia it has a north-western and Atlantic concentration (Jalas & Suominen 1972, Map 83; Jonsell et al. 2000). We could thus describe its European range as being rather similar to that of an Arctic-alpine or Arctic-montane species, but with an additional more lowland boreal occurrence (Page 1997). In Fenno-Scandinavia, A. viride has been described as being particularly frequent on ultra-basic rock (eg serpentine), and under such unusual and generally toxic soil conditions in those latitudes, it can also be found in much drier, sun-exposed sites than is the case elsewhere (Jonsell et al. 2000).

World occurrence

The world map features the species with an uneven distribution in boreal northern and eastern Europe, and then a rather discontinuous, disjunct circumpolar occurrence through NW Africa, Turkey, the Caucasus, the Urals, mountains of S Siberia and C Asia, W Himalaya, Japan, W & E North America, S Greenland and S Iceland (Hultén 1962, Map 92; Hultén & Fries 1986, Map 43; Jonsell et al. 2000).

Names

'Asplenium' is derived from the Greek 'a' meaning 'not' and 'splen', 'splene' or 'splenon' referring to the spleen, alluding to the supposed medicinal properties of the fern genus. The herbal medicinal use is also invoked by the English common name applied to the genus, ‘Spleenwort’ (Hyam & Pankhurst 1995). The Latin specific epithet ‘viride’ means either ‘youthful’, or rather better, ‘fresh green’. The English common name is a simple translation of the Scientific name, ‘Green Spleenwort’.

Threats

None.

Growth form and preferred habitats

This small tufted evergreen perennial is locally abundant on natural habitats such as narrow crevices in rocks and cliff faces, almost exclusively growing on limestones or other forms of base-rich rock. Otherwise this lime-tolerant, strongly calcicolous species is almost exclusively a plant of the lime-rich mortar of old walls and bridges, and it is particularly abundant on rural examples of these man-made habitats. It develops most luxuriantly in half-shaded, damp sites.

Page (1997) rather tentatively suggests that A. ruta-muraria is, "a fast-growing and perhaps rather short-lived spleenwort. In most situations, frequent re-establishment occurs ...". There appears to be very little evidence of dead plants and turnover taking place on walls in Fermanagh, and the reproductive behaviour of Wall-rue requires and deserves further study.

Fermanagh occurrence

Widespread in Fermanagh, this small, tufted, evergreen fern has been recorded in 153 tetrads, 29.0% of those in the VC. Nine of the tetrads have pre-1976 records only. It is very common throughout the county, mainly on calcareous rocks, including in shaded crevices in the limestone pavement above Florencecourt and around Knockmore. However, it also grows on the dolomitized sandstone scarps in the more upland, rather exposed and somewhat wetter Western Plateau. Some stretches of the red sandstone scarps here have been chemically altered by long-term percolation of calcium- and magnesium-rich water derived from overlying rocks (ie., they have become dolmatized).

British and Irish occurrence

Wall-rue is very common throughout Britain and Ireland. However, being calcicole and sensitive to atmospheric pollution, A. ruta-muraria is much less frequent or scarce in more urban and industrial areas of Britain and Ireland, and in regions where soils are predominantly acidic. This includes substrates derived from siliceous igneous and metamorphic rocks that make up the structure of N Scotland and N & W Ireland. Rarity also applies in oceanic areas of very high rainfall, where strongly acidic peaty soils develop, irrespective of the geochemistry of the underlying rock (Jermy et al. 1978; Jermy & Camus 1991; New Atlas).

European and world occurrence

This fern is widespread across W & C Europe and extends north of the Arctic Circle (Jalas & Suominen 1972). The species has its headquarters in more central areas of Europe, and geographically is considered a continental species, albeit with a particularly extensive range near to the western Atlantic coast (Page 1988, p. 86).

The circumpolar distribution of A. ruta-muraria is decidedly disjunct, the species being entirely absent from W and C regions of N America, and poorly represented with very few, widely spaced sites in E Asia (Hultén 1962, Map 156).

Hultén (1962) described the species as being, "fairly variable, at least 15 varieties having been described from C and S Europe, Morocco and China". Two subspecies of differing chromosome count are recognised in Europe, but only the tetraploid, subsp. ruta-muraria, is recorded in Britain and Ireland (Page 1997). A counterpart, confined to eastern N America, has been described as a distinct species, A. cryptolepis, but Hultén (1962) preferred to consider this plant a further subspecies of A. ruta-muraria.

Extremely rare hybrids

Although Wall-rue commonly occurs immediately adjacent to A. trichomanes subsp. quadrivalens (Maidenhair Spleenwort) in both natural habitats and on walls, the hybrid between them, A. × clermontiae (Lady Clermont's Spleenwort), has only once occurred in the British Isles – in 1863 on a mortared wall in Co Louth, Republic of Ireland (H31). It has not been found in Co Down (H38) as erroneously reported by Stace (1975) and repeated by Page (1997) (Hackney et al. 1992).

A. ruta-muraria does also form another very rare hybrid with A. septentrionale (A. × murbeckii), which has never been recorded anywhere in Ireland, but has been found three times in Britain, and has also been reported in 13 Scandinavian provinces (Stace 1997; Jonsell et al. 2000).

Names and Uses

'Asplenium' is derived from the Greek 'a' meaning 'not' and 'splen', 'splene' or 'splenon' referring to the spleen, alluding to the supposed medicinal properties of the fern genus. The herbal medicinal use is also invoked by the English common name applied to the genus, ‘Spleenwort’ (Hyam & Pankhurst 1995). The Latin specific epithet 'ruta-muraria' translates as 'Rue of the wall' or 'Wall-rue'. Folklore suggested that wearing A. ruta-muraria could protect the individual from witchcraft. The unrelated flowering plant, Rue (Ruta graveolens) with bluish-green leaves, was a symbol of sorrow and repentance, and as such could be used either to bless or curse, help or harm. In the absence of flowering Rue, the fern, Wall-rue, could be used as a substitute (Vickery 1995).

Wall-rue had the alternative and rather mysterious English common name, 'Tentwort' (Page 1988). The original version of this name was 'Taintwort', from its use as a remedy for 'the taint', better known to us as rickets (Step & Jackson 1945). Other names applied to the fern include 'White Maidenhair', from the fronds sometimes taking on a blue-green appearance (Gerard 1597), and 'Stone Rue' (Lyte 1578). Herbalists considered it a good remedy for coughs and ruptures in children. It was also thought to prevent hair falling out, and was used for treating shortness of breath, yellow jaundice, diseases of the spleen, stopping of the urine, and to help break up kidney stones. Grieve (1931, page 303) details these and many other herbal medicinal uses of this fern.

Threats

None.

Growth form and preferred habitats

This distinctive, definitely calcicole, evergreen perennial typically occurs either in crevices in cliffs and screes of limestone or other basic rocks, or in the artificial habitat of old lime-mortared walls.

The prevailing damp and relatively mild climate of Fermanagh allows the evergreen fronds to grow most of the year round, and since our high precipitation is rather evenly distributed, checks on C. officinarum growth due to summer drought are few and far between. In the rare event of prolonged dry weather, Rustyback may wilt so severely and curl up so that it appears to have suffered terminally. However, the fronds and their thick backing of overlapping scales have amazing powers of recovery, and thus individuals are very persistent. The fronds can even develop a second flush of sporing sori after such an event (Jermy & Camus 1991). While the fern often appears to be rooted in very dry wall or rock crevices, in reality its roots are frequently embedded in cushions of water-retaining moss or in pockets of damp, black humus derived from dead moss.

Fermanagh occurrence

Only on the limestone cliffs and screes near Boho, at Carrickreagh and in disused quarries at Goladoo near Ederny has Rustyback been found on natural rock surfaces in our survey area. Lime-mortared walls are not all that frequent in Fermanagh, so the fern's real local stronghold is the mortar on old bridge parapets. C. officinarum has been recorded in 69 Fermanagh tetrads, 13.1% of those in the VC. More than half the records are from old bridges scattered throughout the survey area. As the tetrad map shows, it is widely spread across the county, but eleven tetrads have only pre-1975 records, suggesting there has been some local decline of this species.

British and Irish occurrence

C. officinarum is widely distributed throughout the whole of Ireland, but is probably most common to the S and W of the island. The latter comment could also be applied to Britain, but here it is even more south westerly, being frequent only in SW England and Wales and Cumbria, scattered in the Pennines and SW Scotland, but very rare north of the Grampian Highlands and scarce east of the Pennines in England (Jermy et al. 1978; Jermy & Camus 1991).

Europoean and world occurrence

Beyond Britain & Ireland, C. officinarum is widely distributed in SW Europe and the Mediterranean basin where it has its main centre of occurrence, but it also extends eastwards to the Crimea, the Caucasus and C Asia. It also stretches southwest to the Cape Verde Islands (Jalas & Suominen 1972, Map 99; Hultén & Fries 1986, Map 47; Page 1997). The species contracted during the last century and became locally extinct in several areas, mainly along the easterly margins of its natural range in Britain, Germany, Holland, Switzerland, and no doubt elsewhere in mainland Europe (Jalas & Suominen 1972).

Names

Conflicting derivations are given for the genus name 'Ceterach' in the literature. Several sources (eg Gledhill 1985) suggest it is an Arabic name for an unspecified fern, while Gilbert-Carter (1964) believes it is derived from a German word meaning 'itchy', referring to the covering of scales which are said to resemble a cutaneous skin eruption! Step & Jackson (1945) refer back to Turner (1548) The names of herbes (which is always a good place to start!). Turner refers under the name 'Asplenum' or 'Asplenium' to the apothecary's 'Citterache'. Step & Jackson (1945) also regard 'Ceterach' (or, alternatively, 'Chetherak'), as possibly being of Arabic origin, apothecaries "using it as a medicine for troubles of the spleen and liver".

The species name 'officinarum' as always is a reference to the apothecary's shop, the 'officina', where medicinal plants were kept, and thus infers that the species was used in medicine (Gilbert-Carter 1964).

In his Herbal of 1568, Turner is quoted as saying that he had heard of no English name of the fern, although the ancient name 'Asplenum' and the French(!) 'Ceterache' were familiar to him. Turner then suggested several English names himself: "It may well be called in English 'Ceterache' or 'Miltwaste', or 'Finger ferne', because it is no longer than a manne's finger; or 'Scale ferne', because it is all full of scales on the inner syde." (Turner 1568; quoted in Step & Jackson 1945, p. 57). ‘Ceterach’ or ‘Chetherak’ is said to be of Arabic origin, probably handed down by apothecaries who used it as a medicine for troubles of the spleen and liver (Step & Jackson 1945).

The name 'Miltwaste' refers to the spleen (the 'milt'), the suggestion being that animals eating the fern rootstock (although considering how and where the plant grows, this would in reality seem almost impossible for them to achieve), were said to suffer wastage of their spleen and liver (Step & Jackson 1945). The origin of other names like 'Rustyback', 'Brown-back' and 'Scale-fern' are obvious to anyone examining the plant. 'Stone-fern' reflects the habitat in which it grows.

The name 'Saxifrage' (ie 'Stone-breaker') has also been applied to this fern. This name first appears in 1526 in a work printed by Peter Treveris in Southwark, London entitled The grete herball, an anonymous early English translation of an anonymous French work published in Paris c 1498, known under the title Le grant herbier en françoys (Henrey 1975, Volume 1, page 6 & pp 15-22). In this work, 'Saxifrage' is simply illustrated as a fern, which was subsequently identified by Britten and Holland (1886) as Ceterach officinarum. This name could refer to any rock or wall growing fern, including several members of the genus Asplenium. C. officinarum was previously called Asplenium ceterach L., and occasionally it reverts to this name (eg Jermy & Camus 1991). The idea behind the name 'Saxifrage', whatever plant it is applied to, is that they assist the disintegration of the rocks on which they grow. Another English name is simply 'Common Spleenwort' (Grieve 1931).

Uses

The appearance of the frond was considered spleen-like, and following the ancient ‘Doctrine of Signatures’ it could therefore be used to treat ailments of the spleen and other viscera. For instance, an infusion of the fronds was perscribed by herbalists to treat gravel in the liver and spleen (Grieve 1931, pages 302-3).

Threats

Re-pointing of bridges, or replacement of old bridges by new ones.

Fermanagh occurrence

A patch about five metres square of this large, distinctive deciduous fern with its tuft of fronds giving it an attractive shuttlecock appearance, was noticed growing, well established in damp ground below an untarred track at the edge of alder and willow scrub at Knocknabrass Td on the Crom estate. The spot is approximately 400 m NE of the present occupied castle and quite remote from any garden.

RHN collected a sample portion of the plant, including its stoloniferous base, and grew it on in a garden pot. Originally it was mistaken for Oreopteris limbosperma which it quite closely resembles, but in 2009 when the potted plant was fully developed RHN recognised it was this species.

Origin, ecology and history in cultivation

M. struthiopteris is a native fern of boreal and montane areas of central and eastern Europe, Asia and N America. It was introduced to garden cultivation in Britain and Ireland as long ago as 1760, and was first recorded in the wild in 1834 (T.D. Dines, in: Preston et al. 2002). As it tolerates shade and waterlogged soils of almost any pH, it has become a popular subject for water gardens in recent years. Instances of it occurring as a garden escape are quite frequent but widely scattered in Britain from Cornwall (VC 2) to N Ebudes in W Scotland (VC 104) (New Atlas).

Irish occurrence

In Ireland, it is of much rarer occurrence, the Catalogue of Alien Plants in Ireland listing just three VCs, Cos Leitrim, Antrim and Londonderry (H29, H39 & H40) (Reynolds 2002). Famously, two largish colonies spread from an abandoned garden at Shane's Castle, Antrim, into damp woodland at Massereene on the NE shore of Lough Neagh. These established patches of the fern, first discovered in 1948 by Carrothers, Moon and Davidson of Fermanagh Typescript Flora fame, survive and appear naturalised, competing with natural vegetation (FNEI 3). We expect the Crom plant will do the same.

Names

'Matteuccia', is named in honor of Carlo Matteucci (1800-1868), an Italian physicist. The Latinised epithet 'struthiopteris', from the Greek, 'strouqeios' or 'stroutheios', meaning "of an ostrich", and 'pteris', fern. The English common name 'Ostrich Fern' is from a supposed resemblance of the fronds to the plumes of the large flightless Afican bird. Alternative common names include 'Fiddlehead Fern', and the singularly uninspired 'Garden Fern' and 'Hardy Fern' (http://rook.org/earl/bwca/nature/ferns/matteuccia.html)(accessed Nov 2014).

Growth form and preferred habitats

In Britain, this deciduous, rhizomatous, perennial garden species of North American and Eastern Asian origin is a widely scattered, but quite frequent established garden escape or discard. It tends to colonise wet lake margins and areas where ground water seeps. Other typical habitats include damp woodland marshy meadows and riversides. The plant, which is frequently grown in gardens, has a horizontally spreading rhizome which quickly spreads allowing this attractive fern to form large clonal patches. The fronds are heterophyllous, the plant producing yellow-green pinnatifid fronds up to 90 cm tall and separate, shorter, purplish-brown, fertile fronds that bear bead-like sori tightly clustered like grapes.

Since it may quite rapidly outgrow available garden space, Sensitive Fern gets passed around between gardening acquaintances and also becomes discarded in refuse tips or in areas of waste ground where fly tipping takes place (Clement & Foster 1994). The species may also spread naturally by means of its spores and colonies in Britain have been reported considerable distances from habitation. There is some evidence that this species is currently spreading in Britain at least (T.D. Dines, in: Preston et al. 2002).

Fermanagh occurrence

A recent, solitary record exists in the Fermanagh Flora Database of a large, well-established patch that was found growing beside a stream in the outer Pleasure Grounds of Florencecourt House, in an area that otherwise appeared unplanted.

British occurrence

O. sensibilis is mainly distributed in S & W parts of Britain and according to Jermy & Camus (1991) has also been reported in Ireland. Oddly, in view of this, there is no mention of it in the Catalogue of Alien Plants in Ireland (Reynolds 2002), and while it is mapped in the CD-rom distributed with the New Atlas, the latter plots no hectads for Ireland.

Names

'Onoclea' is from the Greek 'onos', meaning 'vessel', and 'kleio', 'to close', referring to the closely rolled pinnules of the fertile fronds that enclose the sori. The Latin specific epithet 'sensibilis' means 'sensitive'. The English common name 'Sensitive Fern' originates in N America where the fronds are observed to be very sensitive to frost, the aerial parts quickly collapsing and dying off when first touched by it. An alternative common name, 'Bead Fern' refers to the bead-like sori (http://rook.org/earl/bwca/nature/ferns/onoclea.html)(accessed Nov 2014).

Growth form and preferred habitats

This widespread and abundant usually calcifuge fern is a large, deciduous, perennial with fronds regularly up to 120 cm long, finely-dissected, delicate-looking (and hence lady-like). Although extremely variable, it is generally easy to recognise, especially when it is a distinctive light, yellowish-green colour. Some plants of a much darker colour do occur, however, and these need to be checked more carefully to ensure correct identification. Lady-fern is a very common and widespread species of damp but well-drained, usually (but not obligatorily) shady, acid to neutral habitats.

In earlier centuries, botanists looked for and found their Male- and Female-ferns, their Filix-mas and Filix-femina, the former being coarse and aggressive, the latter contrastingly delicate, finely-cut and lady-like. The female fern originally chosen was our present day Bracken (Pteridium aquilinum), whose lace-like tripinnately-cut fronds suggested feminine grace and delicacy (Step & Jackson 1945). It was the Swedish botanist, Carl von Linné (since he wrote his texts in Latin, most often referred to by the Latinised form of his name, Linnaeus), who transferred the name 'filix-foemina' (nowadays spelt, 'filix-femina') to the present species, likewise regarded as characteristically feminine (Grigson 1974).

Fermanagh occurrence

Lady-fern has been recorded in 380 tetrads, 72% of those in the VC. In reality, we would estimate that it probably occurs in just about every tetrad in Fermanagh except those on: a. the very highest ground, where a combination of altitude and exposure are too great for its tolerances; b. in very heavily disturbed or waterlogged sites; and c. in county boundary tetrads where only a very small parcel of land lies within our survey area. Typical local habitats include deciduous woods, cliffs, rocky ravines, damp meadows and by water, including along ditches, paths, shady roadside banks and hedges. Occasionally, Lady-fern is also found growing on damp to wet walls, though plants generally fail to mature in this circumstance due to the lack of adequate moisture and soil.

Although Lady-fern is a lime-avoiding, calcifuge species, in our damp climate it frequently occurs in shallow, acid, peaty soils overlying calcareous or base-rich rocks. This is especially the case on slopes, which while they are regularly wetted by our skies, permit drainage adequate to support this fern (Page 1997).

Variation

A. filix-femina is one of the most variable of ferns, so that at least 30 named subspecies, varieties, forms and hybrids have arisen in the wild and under cultivation. Some of these named entities are based on the form and cutting of the frond and occurring in widely separated localities and within several races, while others are more geographically limited (Hultén 1962; Hultén & Fries 1986). In Fermanagh, as long ago as 1860, Prof Smith described var. convexum Newm., and var. incisum Newm. as, "abundant everywhere" (Smith 1860; Meikle et al. 1975). Some of named forms are difficult to separate on account of the great variability of the species, not to mention the inherent confusion of the taxonomy and nomenclature. An excellent account of this fern in Britain and Ireland, its variation, recognition and ecology, is provided by Page (1997).

British and Irish occurrence

A. filix-femina is a very common and widespread species throughout the British Isles, being especially common in the wetter and more mountainous western counties. It becomes local only in the somewhat drier and colder Irish Midlands and in the more continental climate of East Anglia (Jermy et al. 1978; Page 1997; New Atlas).

European and world occurrence

A. filix-femina in the broadest sense is a very widespread circumpolar species of middle latitudes around the northern hemisphere. The distribution thins southwards towards the southern peninsulas of the European mainland and the Mediterranean isles. It also becomes decidedly more sparsely scattered in eastern parts of continental Europe (Hultén 1962, Maps 168A & 168B; Jalas & Suominen 1972, Map 105; Hultén & Fries 1986, Map 49). In the southern hemisphere, Lady-fern has also been reported from tropical forests and mountains in Zimbabwe, Natal, Java, Peru and Argentina (Hultén 1962; Grime et al. 1988).

Names and Uses

The derivation of the genus name 'Athyrium', is somewhat obscure, but possibly comes from the Greek 'anthoros', meaning 'breeding well', perhaps alluding to the varying form of the sori (Gilbert-Carter 1964; Hyam & Pankhurst 1995). The Latin specific epithet, 'filix-femina', translates as 'Lady Fern' (Step & Jackson 1945). In herbal medicine, the uses of Lady Fern are as in Male Fern (Dryopteris filix-mas), but it is considered less powerful in its action (Grieve 1931).

Threats

None.

Growth form and preferred habitats

Brittle Bladder-fern is a widespread, very variable, small deciduous fern in our survey area. Strongly calcicole, it is frequent in permanently shaded, rocky ground, Ash woodland, or walls and quarries where plant competition is limited. It frequently associates with Asplenium trichomanes (Maidenhair Spleenwort), Phyllitis scolopendrium (Hart's-tongue) and Ceterach officinarum (Rustyback). Its thin, delicately cut, deciduous fronds sometimes reach 25 cm or more in length, although usually they are much shorter, around 10 cm long. Separate but similar fertile fronds bear numerous sori. Each young sorus is protected by a thin, membranous indusium which is slightly inflated and is pear- or bladder-shaped. The sori turn dark brown or black in colour as the sporangia ripen, and by this stage the indusium has shrunk and become inconspicuous, facilitating the release of spores on the slightest breeze.

Fermanagh occurrence

This little fern is a typical plant of the Fermanagh limestones, base-enhanced dolomitized sandstone scarps or other calcium-bearing rock habitats of the west of the county. In Fermanagh, it has been recorded in 64 tetrads, 12.1% of those in the VC. Seven tetrads contain pre-1976 records only. Typical local habitats include ± permanently shaded, damp crevices on N-facing, sometimes wooded cliffs and ravines, rocky slopes, screes, swallow-holes, caves and narrow fissures (ie grykes) in limestone pavement. As the tetrad distribution map indicates, apart from the calcareous or base-rich natural rock outcrops of W Fermanagh, Brittle Bladder-fern is only occasional elsewhere in the county and here it is confined to scattered man-made habitats, such as quarries and the weathered lime-mortar of shaded old walls and bridges.

British and Irish occurrence

In Britain & Ireland, the distribution of C. fragilis is markedly northern and western, being strongly associated with the wetter and higher rocky ground in these regions. In other parts of these isles, C. fragilis ascends well over 915 m in sheltered, moist cliff crevices, but in the very oceanic climate and with the relatively low relief base-rich rocks in Fermanagh, the highest it reaches is the very modest 400-420 m of Trien Mountain. The complete inability of the species to tolerate summer drought means that it is more or less absent or very rare in the S & E of both Britain and Ireland (Page 1997).

In Ireland, the distribution and frequency of Brittle Bladder-fern most closely matches that of Hymenophyllum wilsonii (Wilson's Filmy-fern), another very delicate, even thinner-textured fern species, the distribution of which is also very much governed by constantly high atmospheric humidity levels (Jermy et al. 1978; New Atlas).

Variation

C. fragilis is a very variable polymorphic fern with a number of named varieties, some of which are very probably linked to the different chromosome numbers recorded for the species which form a polyploid series. Tetraploid and hexaploid plants, plus the pentaploid hybrid between them, have been found in the British Isles. The degree of frond dissection appears to be correlated with chromosome number: the more dissected it is, the more likely the plant is hexaploid. An octoploid form has been found in Europe, which might also crop up if searched for (Jermy & Camus 1991; Page 1997).

European and world occurrence

In Europe, the species is very widely distributed throughout moister northern and middle latitudes, becoming somewhat less evident in the Mediterranean basin, yet reaching the Azores (Jalas & Suominen 1972, Map 110; Page 1997). A closely related species, previously regarded as C. fragilis subsp. diaphana (Bory) Litard., occurs throughout Madeira (Press et al. 1994), and another related form also occurs in the Canary Islands (Page 1997).

Taken in the broadest sense, C. fragilis s. lat. is an extremely widespread circumpolar species, its natural range stretching right across N America as well as throughout Europe and Asia. Taxonomic varieties of it also occur in large parts of Africa down to the Cape, in Australia, Tasmania, New Zealand, in Kerguelen and South Georgia, as well as in S America and the Falkland Isles (Hultén 1962, Map 55; Hultén & Fries 1986, Map 52).

Names

The genus name 'Cystopteris' is Greek, combining 'kustis', 'kystis' or 'cystis', meaning "a bladder", and 'ptěris', "fern", referring to the bladder- or pear-shaped outline of the indusium (Gilbert-Carter 1964). The shape of the latter always reminds me of the large flasks of coloured water that in my Londonderry childhood of the 1950s often graced pharmacists' premises, and which still act as icons of their profession. The Latin specific epithet, 'fragilis' usually translates as 'brittle', 'fragile' or 'easily broken', which in this case applies to the brittle stipe or stalk of the frond. Translating the Scientific name thus, we derive the English common name 'Brittle Bladder-fern'. However, an alternative translation of 'fragilis' is 'wilting quickly' (Stearn 1992), which is also appropriate for this fern and can readily be applied to its delicate fronds.

Threats

Upgrading and 'tidying' of old walls and bridges.

Growth form and recognition

This large, common rhizomatous fern develops dense shuttlecocks of many fronds. In the middle part of the frond the base of every pinnule has a larger lobe, ie the ultimate segments are highly asymmetrical at their base (Webb et al. 1996; Rich & Jermy 1998, p. 29). This feature, plus the soft texture (always best appreciated by touching the plant), and the grass-green colour of mature fronds make this an easily recognised species, although some individual plants whose texture is less soft will need to be examined more closely to distinguish them from P. aculeatum (Hard Shield-fern) (Page 1997).

The fronds expand in late April or early May and are generally wintergreen with us in Ireland, the old fronds only dying off as the new fronds begin to unfurl. (In colder parts of Britain the fern is regarded as semi-evergreen, fronds dying off after the first hard frost.) Sporing is copious and begins in July or early August. It may continue until the fronds wilt and collapse in the following spring.

Preferred habitats

Soft Shield-fern occurs on a variety of moderately acidic to neutral, damp but not wet soils of medium base-status. The species is sometimes described as a moderate calcicole (T.D. Dines, In: Preston et al. 2002). It is very common and widespread, and under suitable conditions it can dominate the ground layer in humid woods at low altitude. P. setiferum is generally absent from upland areas above c 200 m, and also from strongly acidic peaty ground. Apart from damp mixed deciduous woods, the most frequent habitats of P. setiferum in Fermanagh as elsewhere in Britain and Ireland are on slopes in scrub, along hedgerows, riverbanks and in similar sheltered, shady, damp, but well-drained places.

Fermanagh occurrence

In Fermanagh, this large, rhizomatous fern is common and very widespread having been recorded in 344 tetrads, 64.6% of those in the VC. In terms of tetrad frequency this makes it the eighth most widespread fern in our survey area.

Irish occurrence

Soft Shield-fern is very common throughout most of Ireland except in the more exposed, strongly acidic, constantly wet peatlands of the W & C where it becomes decidedly rare or absent (Jermy et al. 1978; An Irish Flora 1996; New Atlas).

British occurrence

Soft Shield-fern is abundant in suitable sites in SW England and SW Wales, but as one moves either east or northwards it becomes steadily rarer, until in Scotland it is virtually confined to western coasts. The most northerly point P. setiferum reaches anywhere in the world is near Ullapool on the west coast of Scotland (Jermy et al. 1978; New Atlas).

European and world occurrence

Overall P. setiferum is very definitely an Atlantic or oceanic species, most abundant along the Atlantic coastline, but occurring inland throughout W and S Europe and along the northern shore of the Mediterranean, becoming more discontinuous further east but reaching the Caspian Sea area. It occurs on all the islands in the Azores, and further south reaches Madeira and the Canary Islands (Jalas & Suominen 1972, Map 121; Hultén & Fries 1986, Map 62; Page 1997; Vertag 2002).

Hybrids

P. setiferum hybridizes with P. aculeatum to form P. × bicknelli, which has been recorded once in Fermanagh (see our account of this hybrid).

Names

The genus name 'Polystichum' is derived from two Greek words 'polus', 'many', and 'sticos' or 'stichos', 'row' or 'file'. The genus name thus translates as 'many rows', a reference to the regular rows of sori on the fertile frond (Gilbert-Carter 1964). The specific epithet 'setiferum' means 'bristle-bearing' and is derived from the Latin words 'seta' or 'saeta' meaning 'a bristle' or 'a stiff hair', and 'fero' meaning 'to bear' or 'to carry' (Gilbert-Carter 1964).

In past years, the general population did not distinguish any species of Polystichum other than P. lonchitis (Holly-fern), so there are no folk uses, nor any English common names other than obvious, invented 'book names' (Step & Jackson 1945). The name 'Shield fern' is given because the circular, centrally stalked (peltate) indusium that protects the young sorus is considered reminiscent of the Medieval circular Buckler shield.

Threats

None.

Growth form and appearance

P. × bicknellii is evergreen and calcicole like its P. setiferum (Soft Shield-fern) parent and it can be distinguished by the presence of mostly abortive spores. The plant closely resembles a large, leathery, robust, dark green form of P. setiferum.

Fermanagh occurrence

This hybrid has been found only once in Fermanagh as detailed above, but it should be looked for in shaded, mainly lowland areas where both parent ferns occur (Northridge et al. 1988). Apart from the limestones of the Western Plateau, the Florencecourt and Maguiresbridge areas and the ground lying north of Kesh are the main localities in Fermanagh where both parents commonly meet and this sporadically occurring hybrid is most likely to occur.

British and Irish occurrence

Throughout Britain and Ireland, this hybrid occurs thinly scattered as isolated individuals, mainly in ecologically open, calcareous or base-rich soils. Typical habitats include rocky woodlands, limestone gorges, stream banks and old disused quarries. It seems particularly associated with ground where both parents are involved in recolonisation after disturbance (Jermy et al. 1978). The hybrid is very easily overlooked, and Page (1997) therefore believes it is under-recorded.

Threats

None.

Growth form and preferred habitats

While P. setiferum (Soft Shield-fern) is quite gregarious and forms large 'shuttlecocks', each composed of numerous fronds, in comparison P. aculeatum rarely occurs as anything more than scattered individual plants. Also, individuals of Hard Shield-fern are typically small to medium-sized, often with just a few fronds per shuttlecock crown.

P. aculeatum usually grows in shade, and it is very often associated with exposed calcareous or other forms of base-rich rock. Of the two shield-ferns, the evergreen P. aculeatum tends to be the more upland in character, growing in damp pockets of soil in a wide range of shaded places where outcrops of limestone rock occur, including stabilized scree, around swallowholes, in deep grykes in limestone pavement, on steep wooded slopes, river banks and by streams and waterfalls.

The hard leathery texture of the fronds of P. aculeatum and their glossy, dark-green colour when mature makes most plants easy to distinguish from Soft Shield-fern. If in any doubt, then the plant is far more likely to be P. setiferum. A useful and reliable characteristic of P. aculeatum is that the ultimate segments of pinnae in the middle of the frond are more or less symmetrical at their base and the innermost pinnule on the upper side of each pinna is usually very much larger than all the rest and it is more deeply divided (An Irish Flora 1996; Page 1997).

Fronds of P. aculeatum expand from the beginning of May onwards and they begin to spore in mid-July. Like Soft Shield-fern, the fronds are wintergreen and they persist longer in their second year than those of P. setiferum (Page 1992).

Fermanagh occurrence

Although, as the tetrad distribution map indicates, this is a rather widely scattered species throughout much of Fermanagh, having been recorded in 130 tetrads, 24.6% of those in the VC, there really are only two areas in Fermanagh where Hard Shield-fern occurs with any notable frequency. These are along damp, shaded riverbanks and roadsides in the Maguiresbridge district and to the north of Kesh. In these two localities, it is more frequent than the generally much more common and widespread P. setiferum.

In addition to its shaded and rocky natural or semi-natural habitats, in Fermanagh Hard Shield-fern is also recorded in urban situations around Enniskillen, and in more rural areas it is often associated with man-made structures such as old walls, bridges, weirs and even along old disused railway lines.

Similar behaviour is described in England, where additional man-made base-rich sites colonised include canal-sides, locks and bridges (Jermy et al. 1978). In SW Scandinavia, Hard Shield-fern occurs on mountain cairns, 'stone-fences' (presumably this refers to our 'dry stone walls'), and along ditches (Jonsell et al. 2000). The leathery fronds of P. aculeatum are considerably more hardy than those of P. setiferum, enabling it to survive in more easterly areas of the British Isles and Europe which suffer heavier winter frost than western areas like Fermanagh.

Irish occurrence

In the past, Hard Shield-fern has been recorded at least once in every Irish vice-county except Mid-Cork (H4) (Scannell & Synnott 1987). While this is a fact, in reality over most of the island the fern is only thinly scattered, seldom frequent, and occurs chiefly in the N & W of Ireland (Jermy et al. 1978; Webb et al. 1996; New Atlas).

British occurrence

In Britain, P. aculeatum is much more widespread than P. setiferum, extending from the English Channel coast to Orkney in the far north (but not Shetland, where it is considered a rare introdution). It is absent, however, from considerable areas of the English East Midlands. It is most common in N England and Scotland, but less common than P. setiferum in SW England and S Wales (Jermy et al. 1978; Jermy & Camus 1991; Stace 1997; Preston et al. 2002). In areas of the British Isles where their distributions overlap, P. aculeatum and P. setiferum meet in suitable habitats intermediate in altitude, and their hybrid, P. × bicknellii (recorded once in Fermanagh), is not uncommon and should be looked for (Preston et al. 2002).

European and world occurrence

In Europe, Hard Shield-fern is widespread but uncommon, predominantly occurring in the W and C temperate regions, extending to 64°N up the coast of SW Norway, and stretching south to the Azores, Majorca, Corsica, the S Peloponnese and the coast of N Africa (Jalas & Suominen 1972, Map 120). Eastwards the fern spreads ever more discontinuously, reaching the Caspian Sea and possibly beyond. Probable related taxa occur in the Himalaya and in Japan (Stace 1997). Hultén (1962, Map 141) and Hultén & Fries (1986, Map 61) treat P. aculeatum in an extremely broad sense, merging it at least in part with P. setiferum and related species, so that their maps and accounts are for once singularly unhelpful.

P. aculeatum is sometimes grown for ornament, and Jones (1987) lists four garden varieties, of which 'Pulcherrimum Gracillimum' is described as "the most beautiful British fern" [in cultivation].

Name

The genus name 'Polystichum' is derived from two Greek words 'polus', 'many', and 'sticos' or 'stichos', 'row' or 'file'. The genus name thus translates as 'many rows', a reference to the regular rows of sori on the fertile frond (Gilbert-Carter 1964). The Latin specific epithet 'aculeatum' means 'having prickles, thorny or prickly' (Gledhill 1985), which in the present author’s opinion, rather overstates the case as the frond does not feel prickly when handled.

Threats

Excessive clearance or tidying of riverbanks.

Growth form and preferred habitats

Holly-fern produces its glossy, evergreen, rather leathery fronds in shuttlecock-like rosettes from a short, stout rhizome, often embedded in tight crevices on exposed rock faces. The English common name 'Holly-fern' is not really all that appropriate, since it is not 'holly-like' at all, except in that the simply pinnate fronds are definitely evergreen (each persisting two or three years). The fronds are shiny, and the individual pinnae bear teeth that look rather spiny, but are really quite flexible. Both the growth of fronds and the establishment of new plants are slow, but compensating for this fact, individual plants and fronds are clearly long-lived.

P. lonchitis is a definite calcicole, being confined to calcareous or base-rich rocks, which includes the dolomitised sandstone it frequents in Fermanagh. The species typically grows in cool, moist, well-drained positions at the base of cliff faces, or in crevices on or near ledges. In numerous sites, it frequents stabilised boulder scree and, in England, occasional plants of P. lonchitis are found in moist, deep grykes in limestone pavement and around the entrances of sinkholes.

Fermanagh occurrence

One plant of this species is known to have survived for 30 years at its site in a crevice on a N-facing dolomitized sandstone cliff in Lough Navar Forest (Northridge et al. 1988). In 2000, the original plant was joined by a small plant 20 cm higher on the cliff, which we regard as most probably an offspring of the established plant. By 2010, there were five plants at the site, the original plus two small daughter rosettes slightly higher up on the cliff face, plus two nearly mature rosettes on the ground at the base of the rock face, which previously had been overlooked.

In June 1999 R.D. Porley, an English bryologist, surveying mosses and liverworts on the Lough Navar scarps for the Environment and Heritage Service, found another small plant which he identified as P. lonchitis on Bolusty Beg, almost exactly 2 km due north of the earlier known station. The present author and his Botanical Society joint-Vice County Recorder, Robert Northridge, only discovered this claim when the bryological results were published in September 2002 (the Irish Naturalists’ Journal was a year late in its appearance). In 2003 Robert Northridge visited the site and saw an immature Polystichum plant near a red marker stick. The plant was too immature to determine to species level. On 1 February 2004, RHN revisited the site and the plant had matured enough for it to be determined as P. aculeatum (Hard Shield-fern). It is difficult to distinguish juvenile specimens of these closely related ferns and the mistaken identification is a perfectly understandable one.

P. lonchitis is a circumpolar arctic-alpine species, and at its Fermanagh site it grows beside a plant of Asplenium viride (= A. trichomanes-ramosum) (Green Spleenwort), another northern or arctic-alpine species similarly confined to calcareous or other base-rich rock habitats. Normally both these species occur in upland areas in Britain and Ireland, which are typically cool in summer (summer maximum around 27C), and where base-rich pockets of soil are kept permanently moist by water seepage (Page 1997).

Sori are produced on the top portion of the frond only, and spores are released from mid-summer until the following spring. Clearly conditions at our Fermanagh site are suitable for reproduction, since new young plants have become established near the original specimen in recent years. Apart from the well-colonised stable block scree stations in Glenade, Co Leitrim (H29) (the undoubted Irish headquarters of the plant, where the species grows large, luxuriant clumps), Holly-fern occurs in most of its Irish sites as small, widely scattered, individual cliff crevice plants. It is quite possible that somewhere in the Lough Navar area another plant might occur on nearby cliffs or screes of suitable rock chemistry.

Irish occurrence

The plants at the solitary Fermanagh site are the only representatives of this species known from Northern Ireland, although elsewhere in the Republic of Ireland there are a few scattered localities down the W coast in Co Donegal (H34 & H35), Co Leitrim (H29) (definitely its Irish headquarters), Co Sligo (H28), Mayo (H27), Co Galway (H16) and Kerry (H1 & H2) (Jermy et al. 1978; Census Catalogue of the Flora of Ireland 2).

In limestone pavement in N England, occasional plants of P. lonchitis are found in moist, deep grykes and around sinkhole mouths, making it all the more odd that the species is absent from the identical habitats which are so vastly more abundant in the Burren district of Co Clare. Having said that, A. viride (Green Spleenwort) is another inexplicable Burren fern absentee.

British occurrence

In Britain, P. lonchitis is similarly rare and local from N Wales, N England and S Scotland, but it becomes more widespread in the mountains of C & NW Scotland (Jermy et al. 1978; Page 1997). The records in the 1978 Fern Atlas suggest that there may be a contraction in the number of sites for Holly-fern in parts of Ireland, England and Wales, but this does not appear to apply in the main area of the species in N & W Scotland (Jermy et al. 1978). In limestone pavement in N England, occasional plants of P. lonchitis are found in moist, deep grykes and around sinkhole mouths.

European and world occurrence

In Europe, P. lonchitis is widespread in cooler areas, ie montane and high latitude regions. The distribution extends northwards along W Scandinavia reaching well inside the Arctic Circle. The species is also present in Iceland and the Faroes. In continental Europe, it is mainly associated with the Alps, Pyrenees and other outlying mountain areas, reaching its southern extremities in SE Spain, Italy, Corsica, the Peloponnese and W Crete (Jalas & Suominen 1972, Map 119). The overall European distribution pattern is remarkably similar to that of Asplenium viride (= A. trichomanes-ramosum), Green Spleenwort (Jalas & Suominen 1972, Map 83). As in the case in Dryopteris carthusiana (Narrow Buckler-fern), P. lonchitis in Scandinavia occupies a much wider range of habitats than it does in the British Isles, including forest (deciduous forest as well as mixed and spruce forest), together with crevices in lava-fields, and even occasionally on man-made structures e.g. stone walls ('stone-fences') (Jonsell et al. 2000).

Beyond Europe, Holly-fern ranges around cooler parts of the northern hemisphere in a circumpolar manner from the Caucasus, the N Urals, mountains of C Asia, the W Himalaya, Japan, W and NE North America and Greenland (Hultén 1958, Map 219; Jonsell et al. 2000).

Names

The genus name 'Polystichum' is derived from two Greek words 'polus', 'many', and 'sticos' or 'stichos', meaning 'row' or 'file'. Taken together 'many rows' is a reference to the regularly arranged rows of sori on the fertile frond (Gilbert-Carter 1964). The specific epithet 'lonchitis' is a Latinised form of the Greek 'loncho' meaning 'spear-shaped' or 'lance-shaped', obviously referring to the outline of the frond. It was also a name given to an unknown fern by Dioscorides (Gilbert-Carter 1964; Gledhill 1985).

The English common name, 'Holly-fern', is a rather curious one, the only similarities between the two plants being the evergreen, thick textured leaves. The numerous teeth on the fern pinnae only look prickly or spiny, but are in fact quite soft, unlike the real Holly's leaf prickles. In common with the 'Oak Fern' and 'Beech Fern', there is no ecological linkage between the fern appellation and the tree for which it is named, and one is left to wonder at the philosophy behind the obscure names humans give to the things about them!

Threats

The Fermanagh Holly-fern station is vulnerable having survived one extensive fire on its cliff in recent years; another such event could easily destroy it.

Growth form and preferred habitats

Male-ferns in Britain and Ireland were originally conceived to be a single species and the history of the subsequent splits and their naming is a complicated one. In Fermanagh, we have records for a total of five species of Dryopteris plus two hybrids. Among the male-ferns, we recognise and distinguish D. filix-mas and D. affinis (Scaly Male-fern), the latter an apomictic species, several forms of which have in the past been considered and recorded as separate species. D. filix-mas is a large, common deciduous fern that is distinguished from D. affinis by the lack of a black mark at the point where the pinna midribs meet the rachis, and by its sparser, paler, straw-coloured scales on the stipe.

D. filix-mas is a large, vigorous, shuttlecock-forming perennial, the vertical rhizome of which is long-lived and can become quite massive. It is common in woods, hedgerows, streamside banks, ditches, roadside verges, rocky slopes, screes and cliff ledges. The species has an unusually wide tolerance of soil pH (from pH 3-8), nutrient status and water supply, which is reflected in the huge variety of habitats it occupies. Experimental measurements made in W Europe to determine the role of root cation-exchange properties of fern species found that D. filix-mas was indifferent to both soil pH and calcium carbonate content (Koedam et al. 1992).

D. filix-mas performs best in lowland sites with relatively fertile soil and little disturbance. On well-drained slopes, within mixed deciduous woodland, it can dominate stretches of the floor vegetation. Often in this type of seasonally shaded habitat, its main competitors are other large ferns. Most frequently these are D. dilatata (Broad Buckler-fern) and Athyrium filix-femina (Lady-fern), but sometimes it also overlaps with the closely related D. affinis.

Unusually for a fern, D. filix-mas is very tolerant of atmospheric pollution and it colonises shady, damp, urban areas and industrial sites including rubbish dumps, old brickwork and other less salubrious situations. It is probably true to say that its occurrence is limited only by extremes of exposure, wetness (ie permanent water-logging) and shade, and by heavy disturbance including grazing pressure.

Fermanagh occurrence

This species and D. dilatata (Broad Buckler-fern) are the two most common ferns in Fermanagh, the latter perhaps being slightly more widespread. The Fermanagh Flora Database contains records of D. filix-mas from 462 tetrads, 87.5% of those in the VC, and it is clearly widespread and abundant almost everywhere except in aquatic, heavily disturbed, or very exposed situations. It dominates stretches of the woodland floor at the base of the cliffs of Poulaphouca, on Bilberry Island on Lower Lough Erne, and in parts of the Cladagh River Glen.

Phenology

Fresh annual fronds of D. filix-mas unfurl from early May and are fully expanded by the end of that month. D. filix-mas is one of the hardiest ferns in the British Isles and its fronds are wintergreen to semi-persistent in Fermanagh, depending on degree of exposure, although in truth under our dull, grey skies they probably are semi-senescent for much of the winter (Grime et al. 1988). By late February or earlier, the old fronds are often broken down and they are certainly dying off, soon to be replaced by the growth of fresh new croziers on the upright rhizome.

Reproduction

Male-fern plants produce colossal amounts of spores which are released from August to November (Page 1997). Despite this fecundity, it is not at all clear with what degree of success it achieves establishment of the sporophyte generation. That this does happen is attested by the presence of young, small, sporophyte plants embedded in cushions of moss on little rises on the floor of woodlands, and occasional plants developing in crevices at various heights on walls (Willmot 1985). A preliminary population study in Derbyshire woodland suggested that recruitment of small D. filix-mas plants was a relatively rare event when compared with D. dilatata, but that once plants of the former established, they might live for a long time (Willmot 1985). A study in Russian woods suggested that plants mature and spore only when over six years of age, and they may survive for 30 to 40 years or longer (Pogorelova & Rabotnov 1978).

On walls, observation indicates that small plants frequently persist for many years, but they seldom achieve maturity and sporing ability. Possibly reasons for such failure are excessive dryness, lack of nutrients, or perhaps because of eventual, inevitable disturbance (Grime et al. 1988).

British and Irish occurrence

D. filix-mas is extremely common throughout the whole of the British Isles, and in fact in terms of spread it is second only to Pteridium aquilinum (Bracken) in the number of hectads in which it occurs (Jermy et al. 1978; New Atlas). Page (1982, 1997) suggests that the presence of Male-fern in Britain and Ireland has diminished in the last two thousand years with the gradual removal of much of the forest vegetation, the natural habitat of the species. Considering just how common and widespread a plant it remains, and its almost unrivalled ability to colonise artificial, man-made habitats, it is hard to see that there is any cause for concern about the species yet, since it clearly manages extremely well in the substitute habitats it now so fully occupies!

European and world occurrence

In Europe, Male-fern is widespread at middle and northern latitudes, thinning northwards but reaching the Arctic circle in Scandinavia, and while present in Iceland it is absent further north in Svalbard (Spitsbergen, Bear Island and Jan Mayen) (Tutin et al. 1993). The distribution also thins towards the south in the Iberian Peninsula, and Male-fern is absent from the Azores, Madeira and the Canaries. Of the Mediterranean islands, it is present only on Corsica and Sicily (Jalas & Suominen 1972, Map 123).

On the world scale, taking D. filix-mas in a broad sense (and recognising that the taxonomic and nomenclature confusion that exists within this group of ferns creates problems when it comes to estimating distribution), beyond its main European base it displays a disjunct circumpolar distribution. In the past, several forms were given separate taxonomic recognition, especially in the far east of Asia, S Africa, S America and a number of island groups in the southern hemisphere, including Madagascar, Hawaii, the Falklands and the Galapagos (Hultén 1962, Map 110; Hultén & Fries 1986, Map 64).

Names

The genus name 'Dryopteris' was first given by the Greek physician, Pedanius Dioscorides (c. 40-90 AD), to a fern growing on oak trees, and is a compound of the Greek 'dryas' = 'oak', and 'pteris' = 'fern' (Gilbert-Carter 1964). The specific epithet 'filix-mas' was first given by the German botanist, Leonhart Fuchs (1501-66), and is derived from Latin 'filix' = 'fern', and 'mas' = 'male', from the supposedly big, bold 'masculine' appearance of the species in comparison with the much more delicate, finely divided fronds of the 'Lady-fern' (Gilbert-Carter 1964).

Uses

The above mention of these ancient and medieval herbalists gives an indication of the fact that Male-fern has a very long history of medicinal use. An oil extracted from the rhizome was used from ancient times as a vermifuge (ie used to kill and expel flatworms, tapeworms and liverfluke). It has also been applied for worming in veterinary medicine (Vickery 1995), although the rhizome also contains a dangerous toxin, thiaminase. The latter is known to cause thiamine deficiency in animals such as horses, cattle and pigs which have eaten the fern. When oil of Male-fern was used for worming, a single sufficient dose was reputed to produce a cure at once, but too much of the toxic drug is poisonous and can cause coma and blindness.

D. filix-mas also contains filixic acid (filicin), and the main toxic activity of the species is due to a phloroglucinol derivative of this substance (Cooper & Johnson 1998). As with all herbal lore quoted here, BE WARNED, and do not attempt to administer any such drug without the guidance of a qualified, licensed medical or veterinary practitioner. Another past herbal use for powdered root of Male-fern was for treating rickets in children, and an ointment was also made and used for healing wounds (Grieve 1931).

Male-fern had other uses similar to those of Bracken, the ash of the burnt fern having applications in glass-making and soap. Young croziers were boiled and eaten like Asparagus. In Norway, at times of hardship, it was used to make beer, the dried fronds being said to make an excellent bitter. Fronds infused in hot water have also been regarded in the past as good fodder for sheep and goats (Grieve 1931).

Threats

None.

Fermanagh occurrence

There are just eleven records, therefore, in the Fermanagh Flora Database, occurring in nine tetrads. The majority of them were made by Matthew Tickner of the RSPB when he was surveying the flora of small islands in Lower Lough Erne in summer 1989. The seven additional islands on which he claimed D. × complexa grew were as follows: Bingham's Rock, Coghran's, Gravelly, Inishmakill, Inishturk, Sam's and 'Stone Park West'. During a BSBI field meeting on 24 August 2004, the visiting English botanist Ken Trewen added a further three records with sites listed as follows: woodland in Correl Glen NR, determined by KT; scrub covered scarp on S side of Glencreawan Lough, determined by KT; woods beside Lower Lough Erne below the Cliffs of Magho, determined by KT.

Growth form and variation

This is an extremely variable apomictic species of complex ancestry which is divided into four subspecies by Page (1982) and into five morphotypes by Jermy & Camus (1991). Page (1997) reverts to three subspecies, and in the second edition of the New Flora of the British Isles 1997, Stace takes the same approach. Like other entirely asexually reproducing species complexes, D. affinis does not carry out meiosis or reduction division during spore formation, so that its spores are of the same chromosome number as the parent plant. When the spores germinate and produce prothalli, no sexual fusion takes place, so the new sporophyte arises by apogamy (ie an unreduced female gamete, or a cell associated with it, forms the embryo of the next sporophyte generation). It is therefore genetically identical to the original parent plant and the sexual mechanism has been by-passed. Over time, genetic mutations occur, however, and, if the offspring are viable, the mutations are maintained by this method of non-sexual spore reproduction. Eventually, given sufficient time, this produces a multitude of self-perpetuating varieties or 'micro-species' as in this particular case (Jermy & Camus 1991).

These varieties, morphotypes, subspecies or micro-species – whatever we decide to call them, can probably all act as male parents in crosses with other species of the genus Dryopteris, and we do have a limited number of records in Fermanagh of the hybrid that D. affinis forms with D. filix-mas (D. × complexa).

Identification

Despite the above, D. affinis is usually easily separated from D. filix-mas (Male-fern) by the presence of a dark lead-grey or blackish spot at the point at which each pinna meets the rachis, which itself is densely clothed in masses of orange-brown or golden-brown, chaffy scales. The new annual fronds of D. affinis are also produced two to five weeks later than those of D. filix-mas, and their stipes are more densely clothed with chaffy golden-brown or light orange-brown scales (Page 1997).

The three subspecies recognised in New Flora of the BI (1997) are really very difficult to differentiate in a consistent manner, and Robert Northridge and the current author (Ralph Forbes) as joint Vice-county Recorders of Fermanagh, have not tried to distinguish them here. There is certainly more than one subspecies or 'morphotype' of D. affinis present in Fermanagh, although with limited manpower we have not yet been able to differentiate and properly record them. As Page (1997) points out, "satisfactory identification of the variants of this taxonomically very complex apogamous species usually requires a symphony of characters to be taken into simultaneous account."

Preferred habitats

In general, D. affinis occurs in the same wide range of natural and artificial (man-made) habitats as both D. filix-mas and D. dilatata with which it frequently overlaps (Page 1997). These include deciduous woods, along open rides or fire breaks in conifer plantations, hedgerows, streamsides, well-drained places on open hillsides and in mountain glens, plus on or in crevices in urban brick walls.

Toxicity

As with other Dryopteris species, D. affinis contains toxins and is intolerant of grazing pressure.

Fermanagh occurrence

In Fermanagh, D. affinis has been recorded in 345 tetrads, 65.3% of those in the VC. This makes it common and more or less widespread throughout, although it is still scarce in some parts of the lowlands, particularly those with less acid or more base-rich soils. Typical habitats are woods, shaded banks and mountain screes. At higher altitudes, D. affinis certainly appears more common than D. filix-mas which it appears to replace in these circumstances. In Fermanagh, it is very likely that the triploid subsp. borreri (Newman) Fraser-Jenk. will become recognised as the common form of the species, a pattern of occurrence that is perhaps just beginning to appear elsewhere in Britain and Ireland.

British and Irish occurrence

Like D. filix-mas and D. dilatata (Broad Buckler-fern), D. affinis is a long-lived plant, most frequent in the wetter, more ceanic climate of western parts of the British Isles, frequent but more local in the east (New Atlas). It ascends in Fermanagh to moderately high levels, around 300 m on mountain cliffs and screes. Being difficult to identify, the three subspecies recognised by Stace (1997) were poorly recorded during the BSBI Atlas 2000 survey, and they were not mapped in the 2002 New Atlas.

Elsewhere in Ireland, Hackney et al. (1992) listed three subspecies occurring in the three VCs covered by The Flora of the NE of Ireland; subsp. affinis in open situations at all altitudes; subsp. borreri in similar situations to the previous, but also in woodland and more frequent than subsp. affinis at lower levels; and subsp. cambrensis, which is confined to open, upland situations, for example, on mountain cliff ledges and in crevices. D. affinis was not subdivided in The Flora of County Dublin (Doogue et al. 1998), and there are just two individual records by Clive Jermy of subsp. affinis and subsp. borreri dating from 1984 listed in The Flora of Cavan (Reilly 2001). In the Flora of County Waterford, Green (2008) lists nine records of subsp. affinis and two of subsp. borreri made by four different recorders.

European and world occurrence

On account of the taxonomic and nomenclature changes this taxon has gone through in the last 40 years it is difficult to assess the real distribution of D. affinis on a European, let alone a world basis. Nevertheless, it is mapped by Hultén & Fries (1986, Map 65), and regarded by them and most other botanists as a mainly European temperate species.

Names

The genus name 'Dryopteris' was first given by the Greek physician, Pedanius Dioscorides (c. 40-90 AD), to a fern growing on oak trees, and is a compound of the Greek 'dryas' = 'oak', and 'pteris' = 'fern' (Gilbert-Carter 1964). The specific epithet 'affinis' is Latin meaning 'related' or 'similar to', presumably referring to D. filix-mas, the other scaly Male-fern (Gledhill 1985).

Threats

None.

Growth form and preferred habitats

The somewhat crimped appearance of its distinctive light-green fronds, and the long, purple stipe of the plant make this wintergreen plant a distinctive and easily recognised fern. The fronds when lightly bruised in the field give off a slight, sweet smell, but when collected and dried for the herbarium, they at first give off a much more distinct coumarin odour reminiscent of new-mown hay, and hence the English common name (Page 1997). D. aemula is a fern of permanently moist, but essentially well-drained acidic to neutral, often peaty soils of low base content. It typically occupies wooded slopes, shaded banks and sea-cliffs (Wardlaw & Leonard 2005).

Fermanagh occurrence

In upland mixed deciduous woodlands in Fermanagh, such as the Correl Glen NR, D. aemula can form the dominant ground cover over quite large areas of shaded, rocky ground on damp, acidic soils. The species is also a characteristic plant of the sheltered, N-facing, more acidic scarps of the Western Plateau. It is also found on some of the wooded islands of Lower Lough Erne, and on steep, wooded streamsides elsewhere in the county. Altogether, D. aemula has been often recorded in 80 Fermanagh tetrads, 15.2% of those in the VC. The fern is frequent in the western half of the county, particularly in moist woods and shady banks, but it is only occasional and very scattered elsewhere.

Irish occurrence

In Northern Ireland, D. aemula is noticeably more widespread in the wetter, more oceanic western parts of Cos Fermanagh, Tyrone and Londonderry (H33, H36 & H40), while in the east of the province it is largely but not entirely confined to the wooded coastal glens of Co Antrim (H39), and to more upland woods and stream-sides of south Down (H38) (Hackney et al. 1992). In the Republic of Ireland, D. aemula is quite frequent and widespread in counties along the western and southern Atlantic coasts, but is encountered much more rarely or completely absent along coastal counties in the east adjacent to the Irish Sea, and similarly rare in inland situations (New Atlas).

British occurrence

In Britain, as in Ireland, the distribution of this distinctive fern is predominantly western, extending right from the SW tip of Cornwall to Orkney (but not reaching Shetland). There are a few eastern outlying populations in damp, acid, mainly coastal ground in both Britain and Ireland, but nevertheless the predominant distribution is markedly western (Jermy et al. 1978; Wardlaw & Leonard 2005).

The slow growth rate of this species and the gradual, unhurried deployment of additional fronds after a relatively rapid spring flush of growth, means that the immature fronds of D. aemula are rather susceptible to frost both in late spring and in the autumn (ie an early winter or cold snap). The length and reliability of the frost-free period is the most likely factor restricting the distribution of the fern to markedly oceanic areas, and at the same time helps explains its absence elsewhere (Page 1997).

The inference immediately drawn from the distinctive distribution pattern of D. aemula is that it is sensitive to winter cold and late frosts, and according to Page (1997), its habitats are low-lying, "most being within about 30 m [100 ft] of sea-level, although it occasionally ascends higher, especially in Ireland". The mild influence of the Atlantic Gulf Stream has a more pronounced effect on winter, late spring and early summer temperatures in western Ireland than is the case in Britain. In eastern Ireland, D. aemula is known to ascend mountains to 370 m (1200 ft) in Co Wicklow (H20), and 440 m (1430 ft) in Co Down (H38), while in south Co Kerry in the extreme SW of the island (H1), the fern reaches an altitude of 646 m (2100 ft) on the high Reeks (Hart 1891; Brunker 1950; Hackney et al. 1992). Page (1997) therefore appears to be somewhat overstating the difference in the fern's behaviour between Britain and Ireland, since by comparison D. aemula reaches 220 m (715 ft) on the Furness Peninsula in Cumbria (VC 69), 770 m (2500 ft) in E Perthshire (VC 89), and an incredible 1015 m (3300 ft) on Braeriach in the Cairngorms (VC 96) (Wilson 1956; Halliday 1997).

European and world occurrence

This distinctive species has a pronounced western distribution in Europe as a whole, and Ireland is one of its strongholds (Page 1997; NI Vascular Plant Database 2014). It was listed as vulnerable in the Council of Europe report on the Rare, Threatened and Endemic Plants of Europe (Anon. 1977). In France, D. aemula is strictly confined to western parts of Normandy and Brittany. The remainder of its disjunct mainland continental distribution is thinly scattered along the Cantabrian coast of N Spain and Portugal. The only other known world locations for this fern are on the higher mountains of the Azores, the Canary Islands and Madeira (Jalas & Suominen 1972, Map 133; Jermy et al. 1978). In Madeira it is described as "frequent amongst rocks, in woods and along levadas throughout" (ie beside artificial open water channels) (Press et al. 1994).

The fact that Hay-scented Buckler-fern has its world distribution centred and concentrated in the British Isles, means that although it is not overall a rare or even a scarce species here in these islands, we do have a special duty to conserve and manage its sites and study its requirements on the grounds of our International Biodiversity responsibilities.

An interesting species comparision

Examination of both the British Isles and European species distribution maps indicates that there is a definite similarity between the pattern of D. aemula occurrence and that of Hymenophyllum tunbrigense (Tunbridge Filmy-fern). The match of these two physically very different species is particularly close within Britain and Ireland, but on the continental mainland H. tunbrigense has a number of additional stations in E France, SW Germany and NW Italy that are not shared with D. aemula (Jalas & Suominen 1972, Map 69; Tutin et al. 1993). D. aemula and H. tunbrigense both occur on the Sussex Weald and in a number of other disjunct sites in the cooler, more eastern areas of southern England, a fact apparently associated with local pockets of high humidity in these parts of the country (Jermy et al. 1978; Jermy & Camus 1991).

The habitat requirements of these two very different looking ferns are startlingly similar. Both require free-draining yet permanently moist soils, year-round high atmospheric humidity, plus shelter from full sun and desiccating winds. Both ferns grow in sheltered shade in a similar manner, rooted in peaty, acid soils on mossy boulders or on mossy rock slopes, forming carpets or curtains of cascading pendulous fronds. They grow rather slowly, and both can also be epiphytic on mossy tree trunks, as they are in Fermanagh, eg in the Correl Glen Nature Reserve.

Names

The genus name 'Dryopteris' was first given by the Greek physician, Pedanius Dioscorides (c. 40-90 AD), to a fern growing on oak trees, and is a compound of the Greek 'dryas' = 'oak', and 'pteris' = 'fern' (Gilbert-Carter 1964). The specific epithet 'aemula' is Latin meaning 'striving', 'rivalling' or 'imitating' (hence our familiar word, 'emulate'), and presumably this refers to D. dilatata or D. carthusiana which species D. aemula rivals in beauty and competes with in Nature (Step & Jackson 1945; Gilbert-Carter 1964; Gledhill 1985).

Threats

None.

Growth form and preferred habitats

A medium-sized, finely-divided, stiff, upright-fronded, deciduous Buckler-fern with a distinctive dull, greyish-green, mealy surface, this is a rare plant of base-rich rocks, including deep crevices (grikes) in limestone pavement, coarse limestone screes (block screes) and rock crevices where moist, humus-rich, peaty soils develop. D. submontana prefers a degree of shelter from weather and adequate protection from grazing is essential to its survival, but is intolerant of all but light shade. It therefore tends to occur in relatively inaccessible places, such as rock ledges, deeper, wider grikes in limestone pavement and amid thorny or evergreen scrub that provides shelter and protection.

British Isles occurrence

This is a rare plant confined to a limited area of limestone terrain in the northern English Pennines, although there are also rare outlying stations in N Wales and the NW Midlands.

Fermanagh occurrence

This fern, of which there is just the solitary record listed above, was recorded at the time as Lastrea rigidum (= L. rigida (Sw.) C. Presl). However, it must certainly be wrongly identified, since this very rare, deciduous, calcicole species which demands sheltered, moist, humus-rich soils, has never been found anywhere else in Ireland. Neither Meikle and his co-workers (who in 1957 and 1975 referred to the plant as Dryopteris villarii (Bell.) Woynar), nor we, can identify what fern might have been taken for this in error by an expert pteridologist like Rev Prof Smith.

Growth form, identification and preferred habitats

This deciduous, bipinnate fern is very much a species of wet, peaty, cut-over lowland raised bogs and lakeshore marshes and fens, especially those overlying rich alluvial soils. The lightish-green upright fronds with long basal stipes are usually produced in sparse, irregular groups (never in tight shuttlecocks), which makes it easy to distinguish the species even at some distance from the much more robust and very much more common D. dilatata (Broad Buckler-fern).

The plant has either a short, decumbent rhizome crown (ie reclining but rising at the tip), or a more slender creeping rhizome, the latter type spreading through wet peat and mossy cushions and branching to produce new crowns which send up groups of aerial fronds at intervals.

In our Fermanagh experience, there often may be only one or two individual fronds sprouting in 30 cm high vegetation on a large expanse of bog, so that this is a species that must be actively searched out. Once one has developed an eye for its particular habitat and manner of growth, however, it can be found quite frequently.

The literature suggests that D. carthusiana occurs as a plant of wet woodlands, usually with an alder-willow-birch canopy and with a floor dominated by Sphagnum bog mosses (Page 1982, 1997). However, we do not find it under these conditions anywhere in Fermanagh.

Fermanagh occurrence

In Fermanagh, D. carthusiana has been recorded in 62 tetrads, 11.7% of those in the VC. Seven of these tetrads have pre-1976 records only, a proportion that suggests that the habitats this fern occupies are under threat (see below). As the distribution map indicates, Narrow Buckler-fern is very widely but rather thinly scattered, mainly across wetter areas of the Fermanagh lowlands.

Reproduction

In addition to vegetative spread, mature fronds produce numerous asexual sori, which spore freely from July to September (Hyde et al. 1969). The fronds are summer-green only, dying and disappearing quickly after the first winter frost (Jermy & Camus 1991; Page 1997). Interestingly, in Scandinavia, Jonsell et al. (2000) suggest that it is the fertile fronds which die off, while most sterile ones persist overwinter.

British and Irish occurrence

D. carthusiana is widespread in lowland area of both Britain and N Ireland, especially in Britain south of a line between Stranraer and Berwick-upon-Tweed, and in Ireland, north of the International border with the Republic. However, its frequency has quite rapidly declined throughout these islands since the 1930s. The decline is perhaps most obvious in Ireland where the species was never all that frequently reported to begin with, although there remains an outside and unlikely possibility that the species may not be discriminated by sufficient Irish field workers to give an accurate picture (Jermy et al. 1978; Webb et al. 1996; Page 1997). Irrespective of this possibility, there can be no doubt that Narrow Buckler-fern was much more familiar and frequently found by Victorian field botanists than it is today, since it is now a locally frequent to occasional, or even a rare species in parts of the British Isles (Webb et al. 1996; Page 1997).

Being a plant of wet, peaty habitats which naturally follow a transitional pattern of dynamic succession gradually moving towards drier seral stages as organic matter accumulates, it is not surprising that D. carthusiana populations are eventually eclipsed by these environmental and vegetational changes. They are also vulnerable to the much more drastic and rapidly operating effects of artificial drainage for farming, peat-cutting or other land-development processes which have increasingly affected lowland wetlands in Britain and Ireland during the last 50 or more years (Jermy & Camus 1991).

European and world occurrence

In continental Europe and W Asia, D. carthusiana is widespread in mid-temperate latitudes of N and C Europe thinning somewhat northwards (although reaching within the Arctic Circle in Scandinavia), and southwards to the Mediterranean (Jalas & Suominen 1972, Map 129). Related forms or species occur in eastern N America allowing Hultén (1958, Map 155) to include D. carthusiana (as D. spinulosa) in his amphi-Atlantic group of species. In eastern and central N. America there are closely related taxa that Hultén and Fries (1986, Map 67), plot as var. intermedia and var. fructosa.

In Scandinavia, D. carthusiana appears to occupy a much greater range of habitats than in Britain and Ireland, including much drier sites such as rock crevices, screes, tall-herb meadows and dunes, as well as on stone walls and urban situations (Jonsell et al. 2000).

Name

The genus name 'Dryopteris' was first given by the Greek physician, Pedanius Dioscorides (c. 40-90 AD), to a fern growing on oak trees, and is a compound of the Greek 'dryas' = 'oak', and 'pteris' = 'fern' (Gilbert-Carter 1964). The Latin specific epithet 'carthusiana' refers for some unknown reason to the Grande Chartreuse Monastery of Carthusian Monks, near Grenoble, in France (Gledhill 1985).

Threats

Drainage of fens and bogs, and mechanical peat cutting.

Growth form and preferred habitats

The dark-centred scales on the stipe, the down-turned margins of the pinnules and the dark-green colour of the frond readily distinguish D. dilatata from two much rarer Dryopteris species, D. carthusiana (Narrow Buckler-fern) and D. aemula (Hay-scented Buckler-fern).

The typical habitats of this very common deciduous fern are woods, hedgerows and shaded banks, but it also appears in upland areas on open rocky slopes and in rock crevices and, as a weedy species, in more urban and waste ground situations. Thus, like D. filix-mas (Male-fern), it occurs in a large variety of damp, lowland shade or, in upland, more open habitats throughout almost the whole range of altitude. The biology, ecology and distribution of these two extremely common ferns in Britain and Ireland are very similar and their ecological niches clearly overlap considerably at many shared sites. D. dilatata is more frequently found and is the more abundant of the two species on permanently wet, but not waterlogged soils, at pH levels below 5.0, on bogs, or on acidic, moderately fertile, organic substrates in woods, scrub, hedgerows and on the banks of rivers and streams.

In mixed deciduous acid woodland generally dominated by oak, D. dilatata can carpet the damp, shady floor vegetation, and developing from massive, old, more or less upright rhizomes it forms a dense, mid-green sward of fronds up to 1.5 m tall. In old, less disturbed woods of this type, D. dilatata sometimes also grows as an epiphyte in mosses on the rugged bark of the larger trees in the same way that Polypodium species very often, and Blechnum spicant (Hard-fern), occasionally does (Page 1997). Like Male-fern, D. dilatata also invades conifer plantations, most frequently being found along tracks and fire-breaks, and particularly along the sides of ditches and drains associated with these less shaded, better drained conditions.

D. dilatata is perhaps slightly less frequent than D. filix-mas in limestone areas of Fermanagh, although in our wet western oceanic climate, an insulating layer of peat regularly forms over base-rich rocks, and thus plants that are widely regarded as calcifuge can frequently be found also growing in limestone districts.

Of these two common fern species, D. dilatata is also more often found than D. filix-mas in more open sites on higher ground, such as on more or less steep, rocky slopes, stabilised screes and in rock crevices. Relatively dwarfed plants of D. dilatata are abundant in the clefts between rocks for instance on the summit of Cuilcagh, our highest mountain.

D. dilatata is such a rapidly growing and maturing fern that in less natural, urban and disturbed habitats it can also behave like a weed species, colonising crevices in damp brickwork in the manner D. filix-mas sometimes does, but doing so even more effectively than the latter. It is also quite commonly found in less well-tended gardens, growing out of soil on steps, competing with decorative species in tubs and in greenhouse pots. The pronounced reproductive ability and wide range of variation within the species, suggests that D. dilatata is possibly still capable of further increasing its distribution and range of habitats within Britain and Ireland (Grime et al. 1988).

Fermanagh occurrence

D. dilatata is almost ubiquitous throughout the county. It is both the most frequent and the most widespread fern in Fermanagh being present in 462 tetrads, 87.5% of those in the VC. In the wetter Western Plateau uplands of Fermanagh, D. aemula tends to replace D. dilatata both in wet, acidic, shallow rocky ground in shade and also as an epiphyte in oak woods, eg in the Correl Glen NR.

Reproduction

Broad Buckler-fern produces sporing sori on all but the smallest plants, a feature rather different from Male-fern, which instead takes up to six years growth to achieve sporing fertility (Page 1997). Fronds are less frequently wintergreen than those of D. filix-mas, but D. dilatata produces its fresh annual fronds and sporing sori much earlier in the season than Male-fern. Spores are clearly produced in massive quantities, and from the wide range of habitats and geographical spread in the British Isles, dispersal is very efficient. Probably it is only the essential requirement for free moisture to enable the functioning of the delicate prothallial stage which limits the plant and prevents the even more common occurrence of the species.

A field study by Willmot (1985) found that small sporophyte plants of Dryopteris dilatata and D. filix-mas in woodland, developed in cushions of moss. In the several woods he studied, all populations of D. dilatata produced an excess of small, sterile plants over larger, older fertile ones, while the age structure in D. filix-mas populations did the opposite. Several interpretations of this observation are possible, but a likely one is that Broad Buckler-fern either produces more new sporophytes each year, or that members of this vulnerable stage survive better than those of Male-fern under the site conditions studied and are recruited into the mature population more successfully. There appears to be very few field studies of the population behaviour of any fern in the British Isles, and Willmot's work urgently requires to be followed up and emulated with other species.

Vegetative reproduction

D. dilatata also has a greater tendency to carry out vegetative reproduction than D. filix-mas; the rhizome of some plants, typically when they are growing on shallow soil overlying rock, very occasionally produce long, slender, creeping, offset branches which bear a sequence of small crowns each producing new fronds (Page 1982; Grime et al. 1988). Page (1997) believes that this variant is an environmentally induced form, a suggestion which appears very likely the case.

Toxins

As with Male-fern and other Dryopteris species, D. dilatata is intolerant of grazing pressure and contains the toxic substances thiaminase and filixic acid which can cause blindness and, very rarely, the death of cattle which have eaten the rhizome through lack of other more suitable grazing material (Cooper & Johnson 1998).

European occurrence

In Europe, the distribution of D. dilatata is very much more limited than D. filix-mas, being rather confined to the western region of middle temperate latitudes, at the same time thinning considerably towards the Mediterranean. It does however extend northwards along the Atlantic coast of Norway, and just reaches the Arctic Circle near Bodo (Jonsell et al. 2000).

World occurrence

Beyond Europe (in the Florae Europaea sense), D. dilatata only occurs (presumably rarely) in Asia Minor and the Caucasus (Jalas & Suominen 1972, Map 130). Related forms occur in N America, and possibly also in the Far East (Hultén 1958). Forms in Greenland and Iceland were previously recorded as D. dilatata (Böcher et al. 1968; Löve 1983), but these have been reassigned to D. expansa (Northern Buckler-fern), an amphi-Atlantic species which is now regarded as one of the diploid parent species of tetraploid D. dilatata (Kristinsson 1987; Grime et al. 1988; Jonsell et al. 2000).

Name

The genus name 'Dryopteris' was first given by the Greek physician, Pedanius Dioscorides (c. 40-90 AD), to a fern growing on oak trees, and is a compound of the Greek 'dryas' = 'oak', and 'pteris' = 'fern' (Gilbert-Carter 1964). The specific epithet 'dilatata' is Latin meaning 'broad' or 'spread out', and is derived from the past participle of 'dilato', itself from 'latus' meaning 'broad' (Gilbert-Carter 1964). As the species in the past was not differentiated from ferns in general, it has no local English common names nor any folklore (Step & Jackson 1945).

Threats

None.

Growth form and preferred habitats

A very characteristic, easily recognised wintergreen, rosette-forming, heterophyllous, strongly calcifuge fern of wet, acidic, generally peaty conditions in a wide variety of habitats, B. spicant is especially common in damp woodlands, both deciduous and coniferous. The species is also commonly found on more open upland heathy moorland and montane blanket bog, especially along stream and ditch banks in such sites. Hultén (1962) comments that, "in most parts of its area [ie he is referring here to its whole range], it is a calcifuge, but this is not always so in Scandinavia." Jonsell et al. (2000) instead regard it as a plant of, "mostly oligotrophic ground [ie nutrient-poor, unproductive]; apparently indifferent to lime."

Ecology

Hard-fern does not tolerate drought, but rather it requires permanently damp, humid, yet relatively freely drained, sheltered and usually somewhat shaded conditions for optimum growth and competitive ability (Page 1997). Frequently it becomes locally dominant, forming extensive patches of overlapping rosettes in damp hollows in woodland, or along steep, damp, acid riverbanks, where the depth of shade varies from light to moderate. Apart from woods, cliffs, stream and roadside banks and old, acidic-rock quarries, the fern is much sparser or absent in lowland habitats, particularly in the areas of better agricultural soils.

Hard-fern is well named, the fronds being rather leathery, not to say rigid, so that they appear unattractive to all but extremely hungry grazing animals. Putting this another way, the rosettes are tolerant of moderate but not heavy grazing pressure. Management of upland grazing involving cyclical burning to create a mosaic of young and older vegetation is undoubtedly detrimental to this fern, which can be locally eliminated by such practices. (Sinker et al. 1985; Page 1997).

Reproduction

The plant produces separate, quite dissimilar sterile and fertile fronds: the latter are longer and bear much narrower pinnae that on the underside bear paired linear sori covered by two long indusia either side of the midrib.

Fermanagh occurrence

B. spicant occurs on moors and upland bogs in sheltered spots right up to near the highest levels on mountains in Fermanagh. Very dwarf specimens of Hard-fern grow in sheltered damp hollows, along peat banks and between rocks close to the summit of Cuilcagh, the highest mountain. B. spicant finds very many suitable sites in Fermanagh and has been recorded from 366 tetrads, 69.3% of the tetrads in the VC, making it considerably more frequent and widespread than even Pteridium aquilinum (Bracken) (301 tetrads, 57.0%). It is found almost throughout the VC, but avoids lime and base-rich conditions and therefore is absent particularly from some fertile, intensively cultivated lowland areas.

In the very wet oceanic climate of Fermanagh, peat can develop directly over limestone and calcareous sandstone so that B. spicant can occur, but only very locally, in small pockets or on wider stretches of damp, acid, organic soil, even in what appears from the map or general appearance to be geologically unsuitable limestone terrain.

British and Irish occurrence

Hard-fern is very common and widespread throughout most of the British Isles, especially in the wetter N and W areas. It is much less prevalent, or indeed absent, in parts of the east and midlands on both islands. It is most markedly absent on the clay, chalk and limestones of S England. B. spicant has contracted to an unknown extent in the Irish and English Midlands, probably due to a combination of factors causing habitat loss, including woodland clearances, destruction of lowland heathland to create improved pasture, general intensification of farming practices, building development, and industrial and domestic air pollution; it is sensitive to all of these (Jermy et al. 1978; Jermy & Camus 1991; Page 1997).

European occurrence

B. spicant is widespread in W and C Europe, oceanic conditions allowing it to spread northwards up the coast of Norway to within the Arctic Circle. It is also found on the Atlantic islands (Iceland, the Faeroes, Azores, Madeira and the Canaries). Towards the Mediterranean it becomes more dispersed and local, but it is found (however rarely), on the southern tip of the Iberian Peninsula, Corsica, Sardinia, Sicily and Crete. There appears to be considerable doubt about its presence on the Balearic Isles (Tutin et al. 1993) and Jalas & Suominen (1972) do not map it there. In SE Europe, it reaches, but is scarce and local, in Turkey and in Asia Minor (Jalas & Suominen 1972, Map 139; Page 1997).

World occurrence

Beyond Europe, B. spicant has a very widely disjunct discontinuous range in middle temperate latitudes around the northern hemisphere, occurring locally in N Asia, Japan (where a var. nipponicum (Kuntze) Miyabe & Kudo is recognised), Alaska and the Eastern Pacific states of N America (Hultén 1962, Map 143; Hultén & Fries 1986, Map 73). In the opinion of the current author (Ralph Forbes), the distribution is so extremely disjunct it really is stretching the concept to breaking point to refer to it as circumpolar, but in some heavily qualified context it might fit this description.

Names

The genus name 'Blechnum' is derived from a Classical Greek fern name 'blechnon', which in view of the rarity of the modern species of this name in the Mediterranean area, probably was not applied to the same plant at all (Gilbert-Carter 1964; Stearn 1992). The Latin specific epithet 'spicant' means 'tufted' or 'spiked', probably referring to the often shuttlecock manner of growth of the whole plant, or to the relatively rigid, spike-like appearance of the pinnae (Johnson & Smith 1946; Step & Jackson 1945; Gledhill 1985).

Alternative local English common names include 'Deer Fern', 'Foxes Fern', 'Herrin'-bone Fern' or 'Fishbone Fern' (the latter two both being fitting names, especially when applied to the shape of the distinctive fertile frond), 'Rough Spleenwort' and 'Snake Fern' (Britten & Holland 1886; Step & Jackson 1945).

The species is not very variable, but a small number of varieties of the fern, including a crested form, are quite commonly grown in gardens.

Threats

B. spicant is susceptible to the more intensive agricultural practices and associated habitat changes, eg heather burning, bog drainage, removal of hedges and woodland felling. On the other hand, it invades coniferous plantations, where it thrives along the sides of firebreaks and drainage channels. On balance it is holding its own and is much too common to be under any immediate threat. However, changes in the management of upland areas could affect this species either way.

Fermanagh recorded occurrences

In Fermanagh, we have only four records of A. alba in semi-wild situations on wood margins and thickets. Even with these few records, we have no definite evidence and cannot be absolutely certain that they are self-sown rather than planted. There is evidence of other alien trees being planted at Cladagh River Glen for instance. The details of the other three stations are: two records by D.M. Smith, W. McKenna & Ms E. Kennedy in July 1990, in Glen Wood, Florencecourt and Corry Point Wood FNR, Lower Lough Macnean; and Cladagh River Glen NR, 6 June 1991, Ms B. Hamill & M. Bradley.

Other Silver Firs

The other Abies species most likely to be found self-sown and naturalised in parts of Britain and Ireland is A. procera (Noble Fir), although we do not as yet have any records for it in Fermanagh. It is a very handsome tree with blue-grey foliage and it is frequently planted as a single row around blocks of forestry plots (including some in the Lough Navar Forest Park), to make the plantation more attractive in appearance.

A. alba is now very seldom planted because its foliage is more susceptible to rust fungus and woolly aphids than other Silver Firs, eg A. grandis (Giant Fir) and A. procera. In amenity plantations, A. alba has been largely replaced by A. nordmanniana (Caucasian Fir), of which Forestry Service timber trials are also under way. In the last 20 years or so, A. nordmanniana has also begun to oust Picea abies (Norway Spruce), as the most popular 'Christmas tree', since unlike the latter it does not drop its leaves indoors over the holiday period (Stace 1997). The gaunt 'stag-head' tops of surviving old trees of A. alba are often rather noticeable from a distance in estate parks and woodlands.

Like the other species of Abies and indeed all exotic conifers, A. alba is inconsistently recorded in the flora of Britain and Ireland, making it very probably seriously under-recorded (M.E. Braithwaite, in: Preston et al. 2002).

Names

The genus name 'Abies' is from the Latin 'abire' meaning 'to rise', which is considered by some to refer to the great height some species of the genus can attain under good growing conditions (Hyam and Pankhurst 1995). The specific epithet is Latin meaning 'white', possibly referring to the distinctive whitish stomatal bands on the undersurfaces of the leaves.

Threats

None.

Fermanagh recorded occurrences

In Fermanagh, it has been recorded so far in just 21 tetrads in woods and lakeshores on islands in Upper Lough Erne and thinly scattered in the near vicinity of obvious plantation woods from which seed must have originated. As the distribution map shows, most of these plantations lie close to the county boundary. The other record details are: Inishleague Island (West), 17 June 1987, R.S. Weyl & Mrs J. Whatmough; Inishlught Island, 1987, R.J. Bleakley; Glen Wood, Florencecourt, 1990, D.M. Smith, W. McKenna & Ms J. McConnell; Cornagague Lough, 1991, W. McKenna & Ms J. McConnell; Knocks Td, ENE of Lisnaskea, 1995, RHN & RSF; 1 km E of Eshcarcoge Td, August 1995, RHN & RSF; shore at plantation, 1.5 km NW of Garrison, July 1996, EHS Habitat Survey Team. RHN & HJN added six more records in autumn 2010 as follows: Cashel Crossroads; N of Rotten Mountain bridge; Meenawanick NE of Brickagh; SW of Tonymore; Corraleek SE of Garrane; and Tullykeeran Td, Pettigo Plateau.

Names

'Picea’ is the classical Latin name of the genus and is derived from the Latin 'pix' meaning 'pitch', a reference to the resin obtained from the tree

(Hyam & Pankhurst 1995). The specific epithet 'sitchensis', means 'of Sitka', geographical reference to a borough and city in Alaska, NW America.

Uses

Although popularly denigrated and even vilified by "the undiscerning public" and "green conservationists", as creating, "huge sterile blocks of alien conifers blanketing the beautiful moors in ugly rows of uniform dark green", Mitchell (1996, p. 91), from whom these are quotes are taken, puts up a strong case for the virtues of Sitka Spruce, especially in terms of the many species of birds it supports.

Threats

Sitka Spruce is another species susceptible to the cankerous killer disease caused by the readily transmitted fungal pathogen Phytophthora ramosum which is currently spreading in Britain and Ireland.

Names

'Picea’ is the classical Latin name of the genus and is derived from the Latin 'pix' meaning 'pitch', a reference to the resin obtained from the tree. The specific epithet 'abies' is derived from the Latin 'abire' meaning 'to rise', which is considered a reference to the considerable height some species of coniferous trees can attain under good growing conditions (Hyam and Pankhurst 1995).

Threats

None.

Fermanagh recorded occurrences

The details of the remaining eight records are: W end of Inishleague Island, Upper Lough Erne, 17 June 1987, R.S. Weyl & Mrs J. Whatmough; shore of Lower Lough Erne, N of Strahenny Point to Temple Hill, Rossfad Td, 13 July 1987, RHN; wood on Doocharn Island, Upper Lough Erne, 1 October 1987, B. Nelson; Glen Wood, Florencecourt, July 1990, D.M. Smith, W. McKenna & Ms E. Kennedy; Correl Glen woodland, 7 May 1992, J. Farren & T. Waterman; cutover bog 1 km NE of Dresternan Lough, towards Rosslea, 4 August 1995, RHN & RSF; Bilberry Island, Lough Melvin, 17 July 1996, EHS Habitat Survey Team; forest track, Tullyvocady Td, N of Derrin Mountain, 26 October 2010, RHN & HJN.

Names

'Larix' is the classical Latin name for this tree, Larix decidua, and the specific epithet 'decidua' denotes that it is deciduous, somewhat unusual for a coniferous tree.

Threats

A leaf blight and canker-causing fungal disease organism, Phytophthora ramorum, potentially fatal to a range of woody species including both Japanese and European Larches, is actively spreading northwards in Britain and Ireland at present. Between 2002-2009, the disease was found at 34 sites in N Ireland, mostly on Rhododendron and other ornamental species at sites which included plant production/retail premises, private gardens, private estates and public parks. All outbreaks were successfully controlled. The disease first attacked Japanese Larch in eastern N Ireland in 2010. It has not yet been found on Hybrid Larch. A second fungal species, P. kernoviae, has also appeared, although so far it is attacking mainly Rhododendron. These pathogens can be spread on footwear, vehicle wheels, tools and machinery, by the movement of infected plants and in rain, mists and air currents. Felling and burning of infected trees is the only effective containment measure known.

Fermanagh recorded occurrences

There are only three records of this hybrid in the Fermanagh Flora Database. The details of the other two are: roadside near Scarfield Bridge, Colebrooke River, 12 April 1996, RHN & RSF; and Derrysteaton Td, Galloon Island, Upper Lough Erne, 1 September 2001, RHN & RSF.

Names

The Latin specific epithet 'kaempferi' is given in honour of the late 17^th century German naturalist, physician and traveller Engelbert Kaempfer, who explored regions from Russia to Japan between 1683 and 1693 and wrote both a flora and a history of Japan. The specific epithet 'marschlinsii' is another genitive case meaning 'of Marschlins', possibly or apparently referring to an area in Switzerland that seems most famous for its castle.

Threats

This hybrid faces the same threat from fungal attack as L. decidua.

British and Irish occurrence

A very commonly grown, extremely familiar conifer Scots Pine is almost always deliberately planted in Fermanagh as is generally the case elsewhere in most of Britain and Ireland. The fossil record proves that P. sylvestris was present throughout both islands early in the Post-glacial period, having rapidly spread north as glaciers retreated. Nowadays, although remaining very widespead, it is regarded as native only in the Caledonian forests of the eastern Scottish Highlands (mainly Deeside and Speyside)(Proctor 2013).

Fermanagh occurrence

In Fermanagh, very occasionally we suspect it of seeding itself spontaneously, and when this occurs it is almost always close to habitation or to plantation woods. Scots Pine is still by far the most commonly found gymnosperm in Fermanagh, having been recorded in 155 tetrads, 29.4% of those in the VC. While no longer widely planted by the Northern Ireland Forest Service for commercial timber, it is still actively planted by landowners in plantation woods, on boundaries and for wind-breaks. It is widespread throughout Co Fermanagh, in woods, hedges and rocky glens, but is especially frequent in the east where more intensive farming takes place. Such trees are capable of regeneration and their seed may disperse, self-sow and become naturalised.

Status in Ireland

Although P. sylvestris was present in Ireland after the last Ice Age and was thus originally a native species, it was over-exploited and died out around the Sub-Boreal to Sub-Atlantic periods (ie 3000 BC to around 1500 AD). It was subsequently re-introduced in the early 18th century using Scottish seed (McCracken 1971), so that nowadays it is generally considered native nowhere in Ireland (Carlisle & Brown 1968; Mitchell 1986).

Fossil record and Uses

There has to remain some slight doubt on this interesting matter, however, since the fossil record shows differences in the decline of pine in separate parts of the island. It had disappeared from NE Ireland by around 2000 BC, but seems to have survived in the SW in marginal, exposed sites with poor soils until around 200 AD. The fossil record shows that whenever bog surfaces became relatively dry, pine could invade them, and we know that some (probably stunted), trees were still growing on Midland Irish raised bogs at around 300 AD (Mitchell 1986). It probably also hung on into historic times in other areas of Ireland with exposed conditions and dry limestone soils, such as the Burren, Co Clare (H9), and on heaths. An early Irish Law text from the eighth century AD survives and it lists the penalties for unlawfully interfering with trees and bushes. Evidence form this text indicates that pine, which was greatly valued for its resin and was used for making pitch to caulk boats, was still fairly common in Ireland in the eighth and ninth centuries AD (Kelly 1997, p. 383). The question of pine survival is an academic one, however, and to all intents and purposes, the botanical and associated entomological evidence points to P. sylvestris either becoming extinct or surviving in such minute numbers that it can no longer be represented in the gene pool of present-day pines in Ireland (Webb & Scannell 1983; Speight 1985).

World occurrence

Scots Pine is a polymorphic Eurasian species and it is the first or second most widely distributed conifer in the world: it’s rival for this distinction is Larix sibirica (Siberian or Russian Larch). The natural range of P. sylvestris stretches from beyond the Arctic Circle in Scandinavia to southern Spain, and from western Scotland to the Okhotsk Sea in eastern Siberia (Hultén & Fries 1986, Map 80). Within this range it grows at elevations from sea level to 2,400 metres (8,000 feet), the elevation generally increasing from north to south.

A forester’s verdict: In his last and postumously published book, Alan Mitchell's Trees of Britain, the author explains that at forestry school Scots Pine was known as "the facile snare", because it caught out the uninitiated tree planter. The point being made is that Scots Pine is easy to raise and it establishes well on a wide range of soils and sites, including some of the most difficult, but in almost all circumstances it will be outgrown and out-yielded by other timber species. "It is unable to exploit better sites and even at its best is a smaller and slower growing tree than any other used in forestry." (Mitchell 1996, p. 118).

Threats

None.

Fermanagh occurrence

This evergreen conifer is widely used in forestry plantations in Fermanagh, but it has only been recorded in the county on seven occasions as a naturalised escape from cultivation. The first case was growing on the margin of a cut-over bog and on the adjacent roadside at the site listed above. Two further examples of self-sown plants alongside forest tracks were noted by RHN & HJN at Tullykeeran Td, Pettigo Plateau, 4 October 2010 and Tullyvocady Td N of Derrin Mountain, 26 October 2010. Other sites include near Moysnaght, and in two places around the shores of Lough Vearty where RHN recorded in December 2010.

British and Irish occurrence

Elsewhere in these islands, Stace (1997) mentions P. contorta being self-sown from only two areas, Cardiganshire (VC 46) and the Scilly Isles (VC 1b), but clearly it does so also in Fermanagh. The New Atlas hectad map now shows that P. contorta has been quite frequently and widely recorded across both Britain and Ireland, but especially so in the N & W regions of both islands.

Between 1945 and 1980, the two forms of this species were the second most abundant trees planted in Britain and Ireland, being used in re-afforestation to raise the planting limit above 100 m, and to extend planting to deep peat soils. The Shore Pine (var. contorta) was first introduced to forestry by A.C. Forbes of the Irish Forest Service, who discovered it as a rogue growing among a batch of seed of Pseudotsuga menziesii (Douglas Fir) imported from western N America in about 1920. He noticed that this form made good strong plants, but later it was found that when grown in the very exposed sites to which they appeared to be ideally adapted, Shore Pine soon became bowed at the base, the effect of strong wind on their very rapidly growing young stems.

Subsequently this led to growth trials of the inland form of the species, var. latifolia (Engelmann) Critchfield (Lodgepole Pine), and it is this particular variety which today forms much of the maturing pine forest in plantations, especially in the wetter, cooler N & W parts of Britain and Ireland (Mitchell 1996).

Growth form and preferred habitats

This evergreen, dioecious conifer varies from a prostrate, dwarf in exposed upland and coastal situations, to an upright shrub up to 10 m tall usually growing in damp, lakeshore or riverbank sites but it can also occur over limestone, chalk and slate soils on heaths, moors and in pine and birch woodlands.

The prickly, spreading needle-like leaves borne in whorls of three represent so-called 'juvenile foliage', and the subsequent adult foliage is comprised of tightly appressed overlapping scale leaves in opposite pairs, resembling the genus Cupressus (Cypress). All, or almost all forms of juniper occurring naturally in Britain and Ireland produce only more or less prickly, linear, juvenile foliage. Fertilised female cones ripen in their second or third year, the overlapping scales swelling and becoming fleshy to form the familiar 5 mm, blackish globular edible berry-like fruit.

Fermanagh occurrence

J. communis has been recorded in a total of 16 Fermanagh tetrads, 14 of which contain post-1975 records. They probably constitute juniper's main stronghold in Northern Ireland following its decline in Co Antrim in the 1980s and 1990s (N Ireland Vascular Plant Database 2002).

Response to grazing pressure

The juniper population decline in Co Antrim is very probably a direct result of the drastically increased grazing pressure on the Antrim plateau due to European Community sheep headage payments in the 1970s which encouraged overstocking of upland pastures.

In the Republic of Ireland, a similar outcome has been found in upland grasslands in Connemara, causing peat erosion and a loss of species diversity (Bleasdale & Sheehy Skeffington 1995). Studies on J. communis conservation and regeneration in the English Chiltern downlands, the Lake District and in the Scottish highlands, all indicate that the species is sensitive to grazing pressure, particularly by sheep (Fitter & Jennings 1975; Ward & Lakhani 1977; Miles & Kinnard 1979a & b; Dearnley & Duckett 1999).

British occurrence

Juniper is a locally common and very variable shrub in various areas of Britain over a wide range of soils and habitats, but it is most widespread in the north and west of the country and is especially frequent in Scotland. It occurs in NW Wales and is also rather thinly scattered in southern England, where there have been many lowland extinctions over the last century or so (Ward 1981; Preston et al. 2002).

Reproduction

In common with most other species, juniper requires open, bare ground or very short turf for regeneration from seed to occur, and while the necessary vegetation gaps are usually present on steep slopes, on more level terrain openings that might permit colonisation are generally provided by heavy grazing pressure, or more rarely by fire. After germination takes place it is vital for the establishment and longer-term survival of juniper seedlings that they are allowed to grow on under managed conditions providing a protected, considerably lightened grazing regime.

With respect to seedling survival, the seasonal timing of sheep browsing, as well as the extent or degree of grazing pressure, has been shown to be significant in chalk downlands in Oxfordshire. Autumn and winter grazing increases juniper seedling mortality and stunting, whereas summer grazing of the grassland is apparently beneficial: the sheep having sufficient browsing material, simply leave the young juniper plantlets alone, and their grazing curtails competition from grasses, tall herbs and hawthorn scrub without killing the juniper (Fitter & Jennings 1975).

Variation

In Fermanagh, all juniper shrubs are confined to the limestone areas of the county, on scarps, scree, pavement and rocky grassland. It is thus restricted to the Monawilkin, Hanging Rock, Marlbank and Florencecourt areas of the county. Although the growth form in Fermanagh is invariably prostrate, the plants are almost certainly exposed ecotypes of subsp. communis. In truth, however, this form of the plant is very difficult to distinguish from subsp. nana (Hook.) Syme in many parts of Ireland, and probably the same is true elsewhere. It is therefore possible that the latter may also occur in some of our most exposed Fermanagh sites (Webb & Scannell 1983; Dearnley & Duckett 1999).

In many places in Fermanagh, juniper grows on scarps and rocky ground closely associated with Taxus baccata (Yew). Some individual juniper plants cover several square metres, and since prostrate growth forms of the species are very slow-growing, this suggests they might possibly be ancient clones. Studies elsewhere in Britain, however, have shown that juniper is a notoriously difficult species to accurately age without cutting live samples of the stem – a destructive process and something we could not justify under any circumstances. The difficulty in measuring the age of individual specimens arises because, firstly, the stem diameter of J. communis is not closely related to age, and secondly, the stems are usually eccentric in shape. This makes sampling for girth measurements or taking cores to ring count for age an inherently inaccurate process (Dearnley & Duckett 1999).

Seed viability

The maintenance of high seed viability has been shown in several studies in England to be important for the conservation of J. communis (Ward 1989). In the English Lake District, a study found that sites with large populations of juniper (ie more than 1,000 bushes) had significantly higher seed viability than those in small populations, and a reference site which had for 70 years been protected from sheep grazing, produced the greatest juniper seed viability index of all (Dearnley & Duckett 1999).

As far as we are aware, no study has yet been made of the ability of juniper in N Ireland to regenerate, and in the light of the obvious contraction of the species on the Garron Plateau in Co Antrim, not to mention the question of the effect on this northern-montane species of Global warming (or as we have come to experience it over the last decade - Global wetting and winding!), conservation background research work needs to be carried out to ascertain the threats to the species, and its ability to survive at all its northern Irish sites under current management practices and levels of disturbance.

European occurrence

J. communis is very widespread over most of northern and western Europe, while to the south it becomes increasingly rare and mainly a mountain plant (Jalas & Suominen 1973, Map 181).

World occurrence

A very variable, polymorphic species, Hultén & Fries (1986, Map 82) recognised six subspecific taxa (either varieties or subspecies), and they plotted the total species distribution as almost completely circumpolar.

Toxicity and uses

Juniper 'berries' are the ripe female cones of the plant, and they contain high levels of resin (10%), an unnamed essential oil, plus terpene derivatives and a bitter substance (probably an alkaloid), which has been given the name 'juniperine' (Cooper & Johnson 1998). The berries have a long history of culinary use, as a flavouring of both meat dishes and gin. The plant is poisonous if eaten in quantity, being particularly dangerous to pregnant animals since it can cause severe cramps and even abortion. Fortunately, however, no specific cases of such poisoning have been reported in Britain or Ireland (Cooper & Johnston 1998).

Names

The genus name 'Juniperus' is Latin not Greek, being a name first given by Virgil to the plant (Chicheley Plowden 1972). The Latin specific epithet 'communis' means either 'common' or 'clumped', ie 'growing in company' (Gledhill 1985).

Threats

Apart from excessive grazing pressure throughout its range in Northern Ireland, a couple of Fermanagh sites on the Marlbank could be threatened by improvements for agriculture. Elsewhere in parts of both Britain and Ireland there has been loss of scrub habitat suitable for juniper due to burning of moors and heaths, succession to woodland and afforestation (M.E. Braithwaite, in: Preston et al. 2002).

Growth form and preferred habitats

A widespread European native and widely cultivated dioecious evergreen tree casting a very dense shade, Taxus baccata L. s.s. (Yew) is one of eight species that formerly comprised T. baccata L. s.l. (Hultén & Fries 1986). The linear leaves are arranged in two lateral ranks and the seed is surrounded by a dull red, fleshy, cup-like edible aril. The Yew tree is unusual for a native conifer, or rather, a 'Taxad', in Britain and Ireland, in being both very slow growing, and also very long-lived (see below). Most particularly, it is unusual through being capable of both vegetative reproduction by layering, in addition to sexual increase and dispersal by seed production (Milner 1992).

It is a small tree in stature when compared with many of its relatives, reaching only a maximum of around 20 m in height, and very often achieving only half this measurement under natural or semi-natural growing conditions. While it is of restricted stature, the Yew can spread its canopy extremely wide in relation to its vertical measurement. On limestone pavement, especially if grazed by goats, Yew can also rarely appear as a prostrate, mat-like shrub.

Sexual reproduction

T. baccata is usually dioecious, having separate male and female trees. Very occasionally, however, monoecious plants are recorded. For instance, an otherwise female tree may bear one or a few branches with small, yellow male cones, rendering the tree monoecious (Nelson 1981; Nelson & Walsh 1993). The female reproductive structure is produced on a short side bud and each consists of an insignificant fleshy disc with a single central ovule (Ross-Craig 1967-70, Part 27, Plate 45; Milner 1992). Pollen release takes place between February and April, wind pollination being the rule, although honey bees do frequently visit male cones to collect the early season pollen that they need to feed their developing brood. Although native Yew trees are hardly all that common, thanks to the frequency of planted specimens the amount of pollen the male trees release is so great that the species is in the top ten plants for pollen abundance in Britain (Milner 1992).

From August to October, the ripe red fruit of the Yew is unmistakable (except possibly for an insect-induced gall that attacks some other alien taxads), consisting of a single, smooth, brownish-purple seed surrounded or embedded in a fleshy, sweet, edible red or pink translucent aril that attracts birds and other animals. (The aril is an outgrowth of the seed coat, which actually is an extra integument layer of the ovule) (Holmes 1979; Lang 1987).

Fermanagh occurrence

T. baccata is one of the very few Fermanagh plant species for which we have records dating before 1800. William Henry recorded it from both the Hanging Rock area and from around Upper Lough Macnean prior to 1739 (Henry et al. 1987). Rutty (1772) also recorded it, "from the islands of Lower Lough Erne". Yew trees, which are certainly native rather than planted specimens, still do occur both on the cliffs and rocky slopes of Hanging Rock Nature Reserve and on the shores and islands of Lower Lough Erne, though there are no recent records from Upper Lough Macnean. Having said that, probably there are more planted Yew trees in Fermanagh than naturally arising specimens, a situation that is now common throughout these islands (Nelson & Walsh 1993).

T. baccata has been recorded from 64 Fermanagh tetrads, 12.1% of those in the VC. As a native species it occurs on the old, pre-drainage shores of Lough Erne and on limestone cliffs and steep scarps in the Monawilkin, Knockmore and Florencecourt areas. Elsewhere, the species has either been planted or is bird-sown in hedges. As the tetrad distribution map shows, it is quite widespread. The Fermanagh Flora Database records include some planted trees in demesnes, gardens and graveyards, together with trees in hedgerows which undoubtedly are bird-sown.

The tetrad map indicates that the main area for the species in Fermanagh lies within the region with limestone soils, particularly around Lower Lough Erne and the upland limestone plateau lying to the south of it. However, the species is not confined to well-drained, lime- or base-rich soils, although it may prefer them, rather it also grows on more acidic terrain. It occurs in mixed deciduous woods over limestone rock, mainly but by no means exclusively associated with ash and hazel.

British and Irish occurrence

While the BSBI's 1962 Atlas attempted to record and display the predominantly native occurrence of Yew in Cumbria, S England and Wales, together with a very thin and mainly coastal occurrence in Ireland (Walters & Perring 1962), the editors of the four year (1996-1999) survey for the New Atlas found that it proved impossible to distinguish native from introduced trees, and consequently they published a map which treats all Yew records as native (Preston et al. 2002). Even with this limitation, while the New Atlas hexad map shows the species widespread throughout the British Isles, it remains most prevalent south of a line between Carlisle and Newcastle, while in Ireland it is thinly scattered throughout (M.E. Braithwaite and M.J. Wigginton, in: Preston et al. 2002).

European occurrence

The Yew tree is sensitive to frost, which limits its northern distribution both in Scotland and elsewhere in W Europe, since it is confined in Scandinavia to southern coastal districts of Norway and Sweden (Jonsell et al. 2000). On the continent it is quite widespread in C and S regions, but strangely absent from most of France, N Germany and Denmark (Jalas & Suominen 1973, Map 194).

Historic timber uses

Yew is the tree most closely linked with history and legends throughout the British Isles. In particular, probably on account of large dimensions and supposed great age, the tree has been associated with notions of immortality and with religious and/or revered burial sites dating back well into Druidical or indeed Neolithic pre-history. Some of the notions associated with the tree continue up to the present day (Mitchell 1996). In early Christian Ireland, Yew was rated in an eighth century law text along with just six other trees as a 'Noble of the wood'. Apparently it rated so highly because Yew wood was the preferred material for domestic vessels, such as eating utensils (Kelly 1997, pp. 380-3). In addition, the timber being extremely hard and both water and insect resistant, it had very many other uses. Yew timber was greatly valued before the use of iron became general, both for its durability and its elasticity. The latter property, for instance, recommended its use for the manufacture of longbow weapons for hunting and in war for well over 300 years (Grieve 1931, p. 866; Milner 1992, pp. 40-3).

In a wide ranging essay extolling the Yew tree, Grigson (1952) discussed how valued the trees were for sheltering dwellings from the prevailing winds. When used in this manner in Fermanagh, they are almost always planted on the SW side and close to the house.

Famous Fermanagh specimens

The story of the 'Florencecourt Yew', T. baccata var. fastigiata, is covered separately below. The other most famous Yew tree in Fermanagh is 'The Crom Yew', or rather 'Yews' in the plural, which probably are the oldest trees of any species in Northern Ireland. Having said this, they are estimated to be only 800 years old at most (Browne & Hartwell 1999). The Crom Yew stands on the eastern bank of Upper Lough Erne near the ruins of Old Crom Castle. The late Alan Mitchell (1996, p. 157), a very well-known tree expert of the second half of the 20th century, gives a very amusing account of 'The Crom Yew', of which he read several descriptions in Irish books and magazines, and examined several drawings of the tree. Accounts of the tree mentioned its extraordinary crown, spread in a low canopy supported by 16 oak posts (and previously by 34 brick pillars), and under whose shelter 200 guests of Lord Erne were once served tea. In 1895, the said Lord wrote that it had a 6 foot [1.83 m] bole, girthing 12 feet [3.66 m] at ground-level, and a spread 77 feet [23.47 m] in diameter. Elwes & Henry (1902) quoted Lord Erne and his "60 supports", and described the tree as, "resembling an enormous green mushroom".

Alan Mitchell visited Crom in 1983, and to his great surprise found not one but two very similar trees, brother and sister, planted about 20 ft [6 m] apart! The trees have had their branches interwoven so that they share the same crown and its supports. The second tree is clearly a twin of its partner, and must have been there for between 400 and 800 years (depending upon which account of its planting you believe). The unsolved mystery must remain why these biological details had been ignored or overlooked in the various accounts of The Crom Yew and by the artists who depicted it? (Mitchell 1996, p. 157). The trees have been 'tidied' in more recent times, and as a result they have lost some of their mystery. It is now easy to get under and through the combined canopy of the two siblings. Browne & Hartwell (1999) provide three photographs in their booklet, but Packenham (1996) has captured their appearance even better in his book Meetings with remarkable trees.

Yew toxicity and seed dispersal

Yew is very poisonous. Indeed, it is claimed that every part of the species except the fruit aril is toxic, but even it may be slightly so (Cooper & Johnson 1998). The fleshy, mucilaginous aril surrounding the seed is eaten by a wide variety of birds, the poisonous seed passing undigested through their intestines and voided with their faeces (Lang 1987). Uneaten fruits remain in good condition for many weeks, but eventually they begin to ferment on the branch. Birds that eat these old, decomposing arils may become intoxicated or ill.

Fruit eating normally reaches a peak in November, and by January only a few, more concealed arils remain on the Yew branches. Members of the thrush family (especially Song Thrushes, Blackbirds and Mistle Thrushes) are the main feeders, but Robins and Starlings are also important. Greenfinches are significant seed-predators, and Great Tits are to a much lesser extent. The Greenfinch has a technique of removing the aril and seedcoat first, before consuming the remainder of the seed, which strongly suggests that the toxins are contained in the testa (ie the seedcoat) (Snow & Snow 1988).

A great deal of study has been carried out on the toxins, a number of which have names based on the genus name. One of the many is 'taxine', which is present in all parts of the tree, and is a complex mixture of at least 11 poisonous alkaloids. Taxine is rapidly absorbed from the digestive tract and interferes with the action of the heart. Another poison is a cyanogenic glycoside called 'taxiphyllin', and there is also an irritant volatile oil. In the last 30 years, a chemically altered alkaloid derivative of taxine extracted from T. baccata leaves, named 'taxol', has proven useful for treating ovarian and breast cancers. Due to a vast effort by biochemists worldwide, taxol can now be efficiently semi-synthesised from Yew leaf hedge clippings.

The toxicity of Yew is not diminished by wilting or drying, so that clippings or even fallen leaves are still highly poisonous (Cooper & Johnson 1998). Yew is considered by some to be the most toxic plant in these islands, yet there are conflicting reports of its toxicity to grazing animals. In general, the trees should always be considered highly toxic, but if eaten regularly or often in small quantities, there may be no adverse effects. Cooper & Johnson (1998) report deer regularly grazing on Yew on the North Downs in Surrey without being poisoned, and we have observed the same thing happening with feral goats in the Burren, Co Clare. Despite these two exceptions, a long list of stock and wild animals are known to have been fatally poisoned, often collapsing and dying within a couple of hours of ingesting a lethal dose, which may be as low as 0.5-2.0 g per kg body weight for animals such as horses (Cooper & Johnson 1998).

Medicinal uses

The fact that Yew is potentially lethal to humans has been known since ancient times, and as with many other poisonous plants, this has somehow led to its medicinal use. It was given to 'steady the heart', and as an antidote for adder bites and against rabies (Cooper & Johnson 1998). As usual, it should be emphasised that NOBODY SHOULD EXPERIMENT WITH POTENTIALLY LETHAL POISONOUS PLANTS.

Most human poisoning with Yew involves children eating quantities of the red arils and the seeds they surround. Provided the poisonous seeds are not chewed, they should pass through the gut harmlessly, or with only minor digestive disturbances being noticed (Cooper & Johnson 1998).

The question of Yew longevity

The longevity of Yew trees and shrubs has been appreciated for hundreds of years, and huge trees are venerated or have even been worshipped in ancient times. However, it has only quite recently been realised that some specimens may be thousands of years old (Milner 1992; Thomas 2000). Unfortunately, the yew trees of greatest girth, and all those over about 400 years age, are invariably hollow (Mabey 1996). For this reason, accurate tree-ring counts or radio-carbon dating of specimen trees are impossible, the early growth wood being absent (Mitchell 1996). Indeed, we do not even know at what age the central heartwood begins to rot, or when and why the trees go into slow-growth mode (Mabey 1996).

Consequently, researchers have turned to documentary evidence, comparing girth measurements and calculated growth rates and annual ring counts of presumed very old and younger trees (Milner 1992). This method of estimating age applied to 70 trees over 300 years old scattered throughout England and Wales, allowed Allen Meredith to calculate that during the first 500 years of growth, a churchyard yew increases in girth an average of 1.1 cm per year until it reaches approximately 5.5 m (A. Meredith, quoted in Milner 1992, p. 82).

However, it is clear from documented measurements made of particular trees, that the accuracy of this method is very suspect, or that many old specimen trees grow much more slowly than this calculation suggests, and in some cases their girth may cease increasing for periods of 300 years or more! Furthermore, Yew is known to be capable of growing without the formation of tree-rings (Thomas 2000, p. 45), and thus, even when accurate ring counts exist, they may provide an underestimate of the real age of the specimen.

The growth of Yew is so irregular, variable and anomalous, that unfortunately no generalised growth curve can be fitted to the data (Mitchell 1996). Dated trees however do exist, for instance that at Dryburgh Abbey, Selkirk, which the monks are known from documentary evidence to have planted in 1136. In 1894, John Lowe measured the girth of this yew to be 11ft 4in [3.45 m], and 90 years later it had added ten inches [25.4 cm] (Mitchell 1996). Thus a tree 858 years old had a girth of 12ft 2in [3.71 m]. Really large old yews may have girths of up to 30ft [9.14 m], but with the known possibility of slowing of the growth rate, and the decay of central wood, we can only guess at their age. As far as we can tell, they are probably several thousand years old. The best account of this interesting but problematic topic appears in Alan Mitchell's Trees of Britain (Mitchell 1996, pp. 153-6).

Variation

The extent to which T. baccata varies is demonstrated by the fact that including the fastigate 'Irish Yew', or 'Florencecourt Yew' (dealt with separately), there are as many as 49 cultivars listed in Dallimore & Jackson (1966) Handbook of Coniferae and Ginkgoaceae. Even allowing for inevitable duplications among cultivated varieties, this still represents a striking array of genetic variation within the species.

Folklore

A considerable body of folklore attaches to the Yew tree, which is conveniently summarised in Grigson (1987), Milner (1992) and Vickery (1995). Several of the folk notions associated with the Yew recounted by these authors might not be all that ancient, but may really have originated from the speculations of the famous late 18th century cleric, Gilbert White, who pondered at length on a specimen in Selborne churchyard (Grigson 1952, p. 12). The Selborne tree was a much measured individual, which unfortunately became uprooted by a gale on 25 January 1990 (Mabey 1996).

Names

The genus name 'Taxus' is probably derived from the Greek 'taxon' meaning 'a bow', the strong, flexible wood having been used for making longbow weapons for centuries. The Latin specific epithet 'baccata' means 'berried', although a better spelling would be 'bacatus' as it is derived from 'baca', meaning 'berry' (Melderis & Bangerter 1955; Gilbert-Carter 1964).

The English common name 'Yew', or 'Yeugh' has various spellings in old authors, for example, 'Ewgh', 'Ewe', 'Ife', 'Ugh', 'Uhe' or 'Uhe tre', 'Vew' and even 'View'. In Anglo-Saxon, it was 'iw', Medieval Latin, 'ivus', 'iva', or 'iua', and there are cognate names in other Germanic languages and in the Celtic languages (Grigson 1974). However, it appears that this Anglo-Saxon name was applied to several different plants (Prior 1879). Some authorities derive the name from Gaelic 'iw', meaning 'green' (ie evergreen), but according to Prior (1879) there does not appear to be any such word. As the current author has no skill in Gaelic, he awaits knowledgeable opinion on this matter. Prior links the Medieval Latin version of the name 'iva', to that of Ivy, and also to 'chamaepitys' (Mediterranean Black Cypress), "through a train of blunders" (Prior 1879), which sums up the confused situation rather aptly.

Yew 'berries' are variously referred to as, 'Snat-berries', 'Snottle-berries', 'Snottergall', 'Snotty-gogs' and 'Snoder-gills', an elegant reference to their sliminess (Prior 1879; Britten & Holland 1886).

Threats

Nowadays, most truly native trees grow on inaccessible cliffs, though some other rare possibly indigenous specimens are browsed by sheep and goats, despite being poisonous.

Growth form and origin

This is the famous upright or fastigate 'Florencecourt Yew', planted world wide and known to all, although probably hated by some on account of its very dark, almost blackish-green, often dusty, or even sooty, fastigate foliage and its rather gloomy funereal associations with church and other graveyards. Although it originates here in Fermanagh, we have not bothered to record its local distribution in our vice county botanical survey as it is always planted.

The original one or two specimens of this unusual growth form were collected as juveniles some time before 1767, or perhaps as early as 1740 (Nelson 1981). It was discovered at a place called Carraig-na-madadh or 'The Rock of the Dog', on the NE slopes of Cuilcagh mountain above Florencecourt by a Mr George Willis, a farmer who lived at Aghtenroark (actually, Aghatirourke), in the parish of Killesher (see pseudonymous account by Norval, The Gardeners' Chronicle 1873, p. 1336). Willis planted one specimen in his own garden which eventually died around 1865, but fortunately he presented a second to his landlord, Lord Mountflorence (later created the Earl of Enniskillen), who lived in Florencecourt house. This plant has survived to this day and is the mother tree of all 'Florencecourt Yews' everywhere.

The original station

Charles Nelson and John Phillips investigated the original find area on Cuilcagh thoroughly in 1980 with the help of W. Forde, a former gamekeeper of Lord Enniskillen whose family had handed down knowledge of the spot through the generations. No trees whatsoever now grow in the rock strewn area of blanket bog near the boundary of Aghatirourke and Beihy townlands, below an exposed sandstone outcrop about 2 m high, which has been renamed locally 'Willis's Rock', grid reference: H141297 (Nelson 1981; Morton 1998).

Reproduction

The Florencecourt Yew is a female 'berried' tree, so that all its descendants raised vegetatively must also be female. The fastigate form does not breed true from seed, so the gene producing the upright growth form appears to be recessive (Milner 1992, p. 42). Very occasionally a mutant branch with male cones is produced. This also happens rarely with the common, or 'English', yew (Morton 1998, pp. 196-7).

Garden introduction

The fastigate form of Yew is very easily propagated from cuttings, and when around 1780 the Florencecourt specimen was admired by George Cunningham, a Liverpool nurseryman who obviously could see its commercial potential, Lord Enniskillen was persuaded to give him some slips. Probably it was Cunningham who first introduced the tree to the horticultural trade (Nelson 1981), and possibly not Lee and Kennedy of Hammersmith (Bean 1970-80), although they too, along with other private Irish gardeners, may have been given cuttings sometime around 1780 (Nelson 1981).

Irrespective of exactly who introduced the tree to commerce, by 1838 the 'Irish Yew', or better, the 'Florencecourt Yew', was available to the public at a low price (Bean 1980, 4, p. 566), and was widely disseminated as we can easily see by today’s many large specimens and avenues of the tree around the country in gardens, demesnes and church associated sites.

Thomas Packenham featured and photographed the mother tree with himself standing under it in his excellent book, Meetings with remarkable trees (Packenham 1996), although he comments rather disparagingly of its present day appearance. It is also illustrated in the late Dinah Browne's attractive booklet of Northern Ireland's special trees, Our remarkable trees, photographed by Mike Hartwell (Browne & Hartwell 1999).

Condition of the original tree

The original fastigate tree at Florencecourt grows beside a small stream in a glade surrounded and sheltered by laurel and other taller trees, on land owned and managed by the N Ireland Forest Service. The tree survives rather than thrives, being rather looser in habit than normal, scraggy and ragged due to its uncongenial position, which was heavily shaded in the past by laurels, and in a soil too damp for good growth. It is misshapen by the repeated taking of cuttings, and while some clearing of trees around it and careful pruning of the specimen carried out in 1980 has helped rejuvenate it, the tree is still somewhat overgrown with lichens. Nevertheless, it remains recognisable and while it lives is of botanical interest.

Threats

None.

Growth form and preferred habitats

Asarum europaeum is a patch-forming perennial of shady places, which was widely grown in medieval times by herbalists from at least 1200 AD onwards, according to Harvey (1990).

Fermanagh occurrence

This unique record originates in Dickie's (1864) Flora of Ulster under the heading, "Species which may be considered not strictly indigenous". It is unique since there does not appear to be any other record for this species anywhere in Ireland at any time (Preston et al. 2002). Dickie recorded the plant from, "waste places at the village of Pettigoe". Being a village on the border with the Republic of Ireland, the record might equally apply to Co Donegal (H34) as to Fermanagh.

George Dickie

The finder of the plant in Pettigoe, George Dickie (1812-1882), was a native of Aberdeen and a medically qualified graduate of both Aberdeen and Edinburgh. He came to Belfast in 1849 as Professor of Natural History at the then new University of Queen's College (now The Queen's University of Belfast). He returned to Aberdeen as Professor of Botany in 1860. While in Ireland, Dickie collected the material for his Flora of Ulster (1864), which Praeger (1949) described as, "an excellent little book which embraced not only that province" (ie Ulster), "but included the interesting area of Sligo and Leitrim." Dickie produced two further Floras dealing with parts of eastern Scotland, and he became an FRS in 1881 (Praeger 1949).

Regarding the accuracy of Dickie's Asarabacca record, Robert Northridge has suggested the possibility that the professor could have made an error: the two dark green and rather glossy, evergreen, kidney-shaped leaves of A. europaeum are smaller, but similar in shape to those of Petasites fragrans (Winter Heliotrope), a plant which today is abundant on roadsides around Pettigo. However, I reckon that a man of Dickie's medical and botanical learning and experience would certainly know a Birthwort from a Butterbur! This is not to imply that even the most eminent professor cannot make mistakes. I recall David Webb detailing some of his own errors at a BSBI AGM held in the Botanic Gardens, Glasnevin in Dublin. The odds are in Dickie's favour when it comes to a medicinal plant like Asarabacca.

British status and occurrence

This species is also a declining and rare plant in Britain, with 60 of the 77 hectad squares plotted in the New Atlas having pre-1970 records only. Although previously claimed as being native, it is no longer regarded as anything but an ancient introduction (ie an archaeophyte) (Coombe 1956; G.M. Kay, in: Preston et al. 2002).

Medicinal uses

The root and leaves of A. europaeum are acrid and contain a volatile oil, a bitter matter and a substance with properties like camphor (Grieve 1931, p. 64). It was used amongst other things as a cure for hangovers, as a purgative and to promote sneezing, although Grieve indicates that even in the 1930s it had been replaced, "by safer and more certain remedies". It is clear from a quotation in Grigson (1987, p. 225), taken from a book by John Pechey (1694) The Compleat Herbal of Physical Plants, that Asarabacca was also used as an abortifacient, "Tis diuretick also, and forces the Courses: wherefore Wenches use the Decoction of it too frequently, when they think they are with Child."

European occurrence

A. europaeum is a European temperate species, widespread in middle latitudes and towards the east, but absent as a native from most of W Europe (Jalas & Suominen 1976, Map 368).

Names

Of the curious name, 'Asarabacca', Grigson (1987) says, "Dioscorides described the plant very precisely in his Di Materia Medica, under the name 'Asaron'. He also wrote of a bacchareis, which some herbalists took to be Asarum europaeum as well, though the two descriptions do not tally. Virgil in the Eclogues wrote of a baccar which grew with ivy, in the way of A. europaeum. As if to compromise and resolve the matter, apothecaries squashed the two names into one, to give the strange word 'Asarabacca'".

Threats

None.

Growth form and preferred habitats

This conspicuous rhizomatous perennial is frequently found in open water floating leaf plant communities in still, shallow-water lakes of all sizes in our area, but it also extends into adjacent reedswamp shallows. In other parts of Britain and Ireland, White Water-lily is also found to a much lesser extent in slow flowing ditches and in river backwaters, but while we have over 280 records for N. alba in Fermanagh, only once has it been listed from this type of habitat, from a drain at Cornaleck, Upper Lough Erne in 2007.

In the type of still water bodies that N. alba frequents, the bottom is typically mud, silt or peat, and it is only rarely of shingle or rock. N. alba tolerates a wide range of water chemistry and it is found in calcareous turloughs (ie vanishing lakes) and marl lakes, as well as in decidedly acidic upland lakes, often over peaty bottoms. In the Britain and Ireland, it is normally regarded as a plant of relatively shallow, still, lowland waters, but it can survive in waters up to 5 m deep (Heslop-Harrison 1955b). It reaches an altitude of 405 m at Tarn Fell in Cumberland (C.D. Preston, in: Preston et al. 2002). Colonisation of deeper waters must be entirely by means of vegetative reproduction involving rhizome fragments since seedlings cannot grow at the low light levels that prevail at depth. N. alba also occurs in waters of all levels of productivity ranging from eutrophic to oligotrophic (Preston & Croft 1997).

Although generally found in open, full sun conditions, N. alba is occasionally found in the shade of overhanging trees, or more frequently among tall emergent stems, such as those of Phragmites australis (Common Reed), where light levels may be reduced by up to 50% of full sun (Heslop-Harrison 1955b).

Under favourable growing conditions, N. alba can dominate open water in smaller lakes, the large leaves of the species completely covering the surface. Such large clonal individuals may be of very great, or indeed, indefinite age, and can measure 10 m or more in diameter. Large colonies are generally ring-like in form, the central older parts of the clone having eventually died off (Heslop-Harrison 1955b).

Recognition

In the absence of flowers, the leaves can be distinguished from those of Nuphar lutea (Yellow Water-lily), by being almost circular in outline, having the lateral veins branching at wide angles and forming a rather inconspicuous net-like web near the margin. The leaf stalks are also rounded or oval in cross section, not angular as in Nuphar (Haslam et al. 1975; N.F. Stewart, in: Rich & Jermy 1998).

Alternative growth form

The species is also known to possess a terrestrial growth form with tufts of erect leaves with rolled margins, occasionally reported in marshy habitats in several areas of Europe, eg in W Ireland on Achill Island (Praeger 1934i, paragraphs 24 & 408), by a pool south of Lough Gill near Castle Gregory (H28) and around the shores of Lough Carra (H26), where it grows on wet, calcareous marl (Heslop-Harrison 1955b). Although we have similar marl lakes in Fermanagh, to date we have not observed this rather unusual form of the species anywhere in the VC.

Fermanagh occurrence

N. alba has been recorded in 98 Fermanagh tetrads, 18.6% of those in the VC. As the tetrad distribution map indicates, it is frequent in the W, C and SE of the county, but absent from the open-water areas of our largest lakes, Upper and Lower Lough Erne. The species lacks submerged leaves and is, therefore, much less tolerant of disturbed water than Nuphar lutea (Yellow Water-lily). Thus in Fermanagh's larger lakes, and especially in the navigable channels between their many islands where water can readily become turbid through wind and wave action or as a result of boating activity, White Water-lily is absent. In these lakes, it is confined to sheltered bays and backwater shallows protected from excessive disturbance by reedbeds.

Phenology and sexual reproduction

The floating leaves are produced annually, first making their appearance at the water surface around May. The flowers, which also float, are produced from early June to August. In comparison with Nuphar lutea, the blooms open much wider, and at up to 20 cm in diameter, they are the largest individual wild flower in these islands by a wide margin. The white flowers, which are very primitive (probably dating back to the Jurassic Period), have large numbers of separate, spirally arranged parts (tepals). The flowers close at night, or when it rains, during the 4-7 days each one lasts. The female organs ripen first (ie the flowers are protogynous), with a subsequent prolonged male phase. Pollination is effected either by bees or by other insect visitors, but should this fail, selfing occurs (Heslop-Harrison 1955b; Velde 1986).

After pollination the flower is pulled about 30 cm below the water surface by contraction of the stalk. The flower rots and the fruit ripens and eventually sometime between September and early November it bursts irregularly, releasing a floating mass of mucilage and embedded seeds which resembles frog-spawn and thus attracts both water birds and fish.

Seed production and dispersal

Typical seed production averages around 500 per fruit, although the number of flowers per plant and the seed set per flower, both vary considerably with growing conditions. The seed mass mucilage consists of the transparent arils around each seed aggregated together. It entraps air, floats and enables dispersal across the water body. According to Dutch studies, this flotation may be of short duration, since any rainfall releases the entrapped air and the seeds then immediately sink (Velde 1986; Smits et al. 1989). Duck, Coot and other water birds actively eat the seed, and fish such as Common Carp will also do so, but only if they are starving. Unfortunately, at least in the case of these particular animals, ingestion does not appear to aid the plant's dispersal since, unlike Potamogeton species (Pondweeds), the seed coat is weak, allowing it to be either completely digested or killed by passage through the animal's gut. This fact does not rule out the possibility that other birds and herbivorous fish with less efficient digestion, might internally transport and excrete viable seeds (a mechanism technically known as 'endozoochorous transport'), but as far as I know there is no positive evidence of this occurring at present (Smits et al. 1989).

Transport of species between water bodies

Due to this inability of seeds to survive passage through the animal gut, aquatic plants like N. alba having a weak testa, can transfer between water systems (ie long-distance or jump-dispersal), only through external adherence of seed on the body of an animal vector. This might be achieved either through chemical stickiness, or physical projections involving some kind of burr-like seed or fruit with a spiny or a hooked surface. Our two Water-lily species, Nymphaea alba and Nuphar lutea, have no such fruit or seed properties, and thus they are not specifically adapted for external transport on animal bodies. They might still, however, adhere for a short time in mud on the feet, coat or down of an animal moving between adjacent lakes or streams.

Thus it is not completely impossible to imagine the occasional or rare stochastic longer range dispersal event occurring. Studies of very long distance bird transport of seed to remote oceanic islands has shown that the down on young fledglings may well prove significant, since seed attaches very much more readily to this type of surface than to feathers, particularly pre-flight, preened ones (Falla 1960; Carlquist 1974). In the case of Water-lily species, it is likely that external animal attachment in mud would persist only for a short space of time, perhaps a few minutes, since it has been shown that the seed is susceptible to dehydration. However, I believe that the experimental test used to examine this factor was unrealistically extreme, as it involved air drying seed in a dissector for a period of 28 days before testing their viability! (Smits et al. 1989).

Long-distance dispersal must be a repeatable, recurrent, even if random event; it is not sufficient for it to be fortuitous. Thus animals which seasonally migrate along regular routes, such as birds, are favoured vectors to carry the plant propagules (Cruden 1966).

Seed properties

Seed germination is usually successful and seedlings can sometimes be found in large numbers, but few of them survive overwinter. In water, seed may survive dormant for up to three years (Jonsell et al. 2001), although other studies suggest the seed bank is only ephemeral (Thompson et al. 1997).

Irish occurrence

In Northern Ireland, N. alba, like Nuphar lutea, is very much more frequent in the southern half of the six counties, but overall it is the less common of the two Water-lilies (NI Flora Web Page (accessed 2015) http://www.habitas.org.uk/flora/species.asp?item=2757). The overall Irish distribution shows N. alba concentrated in the N and W of the island, with a thinly scattered occurrence in the centre of the island, and a remarkable almost total absence from the SE, where it is introduced in Wexford (H12), Carlow (H13), Laois (H14) and Dublin (H21) (Praeger 1934i, paragraphs 270 & 495; Scannell & Synnott 1987).

British occurrence

The British distribution in the New Atlas fails to distinguish native populations from the many introductions that are now known to occur as a result of the fashion for garden ponds. 'Escapes' and 'deliberate releases' from such ponds are especially frequent in SE England. The New Atlas map therefore shows N. alba occurring widespread in the lowlands throughout the whole territory, but becoming much less frequent in the E and the NE as one travels northwards (C.D. Preston, in: Preston et al. 2002).

Allowing as best one can for the introduced populations, it clearly remains the case that the somewhat erratic pattern of N. alba distribution in Britain and Ireland must depend both on the current presence of suitable lowland, nearly-still-water habitats, and the long-term effectiveness of dispersal by water, and perhaps by other means, between discrete water bodies and separate catchment areas.

Clinal variation in Europe

In Europe, the genus Nymphaea is generally recognised as consisting of four or five taxa at species or subspecies level. Of these, N. alba and N. candida are the most similar, to the extent that they are regularly considered as subspecies of N. alba. While these latter two forms usually remain geographically separate, N. candida being more NE European and Asian than N. alba, intermediate forms and small areas of territorial overlap do occur between them, so that in northern Europe a 'cline' may exist, ie a gradient of gradually changing variation running between the extremes of the two forms (Heslop-Harrison 1955b).

European occurrence

Beyond Britain and Ireland, the form of the species that occurs with us (N. alba) is common and widespread in W and C Europe, and to the NW it stretches beyond 68°N in coastal Norway, although absent north of 60°N in Sweden and present, but thinning northwards in Finland and the Baltic States (Jonsell et al. 2001). In southern Europe, the distribution thins, both in the Iberian peninsula and towards the Mediterranean, where there are a number of extinctions, eg on Sicily and the southern end of Sardinia. This thinning and local extinction pattern is also repeated towards E Europe and SW Asia (Jalas & Suominen 1989, Map 1509).

World occurrence

The southern limit of N. alba is reached in Algeria, and the SE extremity in Kishmir and the Himalaya (Heslop-Harrison 1955b; Hultén & Fries 1986, Map 812).

Fossil history

The fossil seed and pollen record of N. alba extends back to the Cromer Forest Bed interglacial series and it appears in all subsequent warm periods including the present Flandrian/Littletonian in Britain and Ireland. Indeed, the species is capable of surviving considerable frost, being present within the Arctic Circle at present, and a continuous fossil record runs back to the middle and late stages of the last major Ice Age, called the 'Weichselian' in Britain and the 'Midlandian' in Ireland. It is possible, therefore, that N. alba might have survived in situ the British Isles during much colder phases than the present, perhaps living quite close to the edge of the ice sheets under periglacial conditions (Godwin 1975).

Toxicity and uses

All parts of the plant, except the seeds, contain the alkaloid nupharine, the amount present varying with the season (Heslop-Harrison 1955b). The rhizome and leaves have a history of use in herbal medicine, as in addition they contain tannin, gallic acid, resin and mucilage. Grieve (1931, p. 484) details the plant's use in cases of dysentery, diarrhoea, gonorrhoea and leucorrhoea. The leaves and 'roots' were also used to poultice boils, tumours, ulcers and inflamed skin, and an infusion was gargled for mouth and throat ulcers.

Other folk uses include the starchy rhizome as a food in parts of Finland and Russia, and the same plant organ was a source of purple-black dye for wool and yarn in the Inner Hebrides (Heslop-Harrison 1955b; Vickery 1995).

Names

The genus name 'Nymphaea' is from the Greek name 'Nymphe' given by Theophrastus to an unknown water plant after one of the three half-divine water nymphs who in mythology inhabited seas, streams and woods (Gilbert-Carter 1964; Chicheley Plowden 1972). The Latin specific epithet 'alba', means white.

There are at least 16 local English common names listed in total between Britten & Holland (1886) and Grigson (1987). Some names are poetic and very evocative of the beautiful floating flower, eg 'Lady of the lake' and 'Swan amongst the flowers'. In his famous evocation of Cotswold village life, Cider with Rosie, the author Laurie Lee (1959) spoke of the white flowers, "... they poured from their leaves like candle-fat, ran molten and then cooled on the water."

A high proportion of the English common names refer either to 'water', eg 'Water Bells', 'Water Blob', 'Water Socks', 'Water Rose', or to 'floating', for instance, 'Floating Dock', of which 'Flatter Dock' is probably just a variant. References to the conspicuous broad leaves of the plant are only to be expected, and comparison with 'dock', as already mentioned, eg 'Can-dock', 'Can-leaves'. 'Can' is a reference to the carafe or jug-like shape of the fruit capsule (Grigson 1987). 'Bobbins' is a name applied to both Nymphaea alba and Nuphar lutea, being a further shape comparison of the globular fruit to a lace or weaving bobbin. The name 'Cambie-leaf' is a curious one of northern Scottish origin, applied there again to both of the common water lilies. I would welcome any explanation of the derivation. Yet another name applied to both species in early herbals is 'Nenuphar', which Grigson (1987) reports came down via Mediaeval Latin from the Sanskrit 'nilotpala', the name for the Blue Lotus of India, Nymphaea stellata. By comparison, Caltha palustris (Marsh Marigold) is 'Petie nenufar' (ie 'Petty nenufar'), in Turner's The Names of Herbes, 1548 (Watts 2000).

Threats

None, except perhaps undue water turbidity resulting from disturbance.

Growth form and preferred habitats

A familiar floating-leaved water-lily of lowland lakes, lakelets and slowly flowing streams and rivers, N. lutea is a perennial with a creeping rhizome. It is usually found in both lake and river water between 0.5-2.5 m deep, over mud and silt bottoms, and it typically bears annually renewed leaves of two kinds: the familiar floating, rounded leathery plates, plus large, crumpled, inconspicuous ones which are translucent and are kept permanently submerged. Possession of these submerged leaves allows N. lutea to dampen the physical effects of water movement, and it tolerates much more mechanical disturbance and associated water turbidity than the related Nymphaea alba (White Water-lily) can manage. However, N. lutea is less tolerant of base-poor sites than the latter and it tends to occur in basic to only moderately acidic waters, seemingly restricted to pH 6.0 and above (Heslop-Harrison 1955a; Preston & Croft 1997). In sheltered, still water sites where the two water-lily species cohabit, which they quite often do, N. lutea typically occupies the deeper, more nutrient-rich water.

The floating leaves of N. lutea are clearly oval rather than circular in outline and the lateral veins divide repeatedly and forking regularly in a herring-bone pattern until they reach the margin, while in cross section the leaf stalks are triangular or semi-circular (ie definitely angular) (Haslam et al. 1975; N.F. Stewart, in: Rich & Jermy 1998).

Again, like Nymphaea alba, N. lutea occurs in both the lowland floating leaf open water plant community and occasionally along the margins of reedswamp, particularly where the water depth shelves steeply, or where there is a faster current beyond. Colonies of Yellow Water-lily can vary enormously in size, from isolated plants in less favourable sites, to situations where it dominates hectares of water at a stretch, spreading both by vegetative growth and branching of the horizontal rhizome, and by seed if water depth, flow and turbidity permits germination and establishment (Heslop-Harrison 1955a; Jonsell et al. 2001).

Flowering reproduction

Individual plants of N. lutea in cultivation are known to survive for at least a century and it is likely that under favourable growing conditions in the wild, plants may persist absolutely indefinitely. In a suitable environment, young plants flower annually and they do so quite freely from their third year of growth onwards. It is not unusual to find a mature plant bearing 15 or more flowers at some stage between early June and late August. The cup-shaped solitary yellow flowers, 4-6 cm across, are borne a few cm above the water surface on very long rigid peduncles. Unlike White Water-lilies, once the flower buds open they do not close again during their 4-8 day flowering period.

The flowers are protogynous (ie female first), the stigmas ripening slightly earlier than the stamens (Velde 1986). Pollination is carried out by a variety of flies, bees and beetles, probably in that order of importance, and there may even be Nuphar specialist pollinating flies (Lippok & Renner 1997). The insects are attracted by a strong flower scent, reminiscent to many humans of plum brandy or a combination of fruit and alcohol. The active ingredient of the perfume is actually ethyl acetate, which is an organic chemical combination of acetic acid and ethyl alcohol (Genders 1971, pp. 27 & 42). The alcohol component is produced in the plant's roots, which appear to tolerate and cope with anaerobic conditions in the muddy bottom exceptionally well. Alcohol is a by-product of the partial breakdown of starch and sugars in the absence of oxygen and, of course, in high concentration it is a lethal toxin that can kill cells. N. lutea transports the alcohol aloft from the submerged roots to the floating leaves and to the flowers, from which it evaporates harmlessly away. Effectively, this is a form of plant excretion, getting rid of a toxic waste-product into the atmosphere (Fitter 1987, p. 218). Consequently, one of the better known English common names of the plant is 'Brandy bottle', although in part this might also be due to the globular shape of the ripe fruit.

The flower visitors are rewarded by a copious and freely available supply of both pollen and nectar foods. The stamens are unusual in that while the anthers are introrse (ie their slit-like openings are directed towards the centre of the flower), as they split to release their pollen they arch over backwards and thus they present their pollen outwards towards the petals. Consequently, automatic self-pollination is avoided and out-breeding is strongly favoured, although an insect walking or staggering around on the flat and relatively broad stigmatic disk, can still occasionally manage to self-fertilise the flower (Heslop-Harrison 1955a).

Fruit and seed production

Again, in further contrast to Nymphaea alba, the fruit develops and ripens just at the water surface, rather than submerged (Velde 1986). The number of seed per fruit is very variable and appears to be under both genetic and environmental control. In cultivation in Sweden, for instance, a mean of 361 seeds per fruit was measured for a sample of 70 fruits, the range of seed production being 45-651 per fruit (Heslop-Harrison 1955a). When ripe, the whole fruit may detach and float around for two or three days before the contained mucilage swells and the fruit bursts or irregularly disintegrates, releasing the ten to 20 fruiting carpels, each containing numerous seeds.

Seed dispersal

In contrast with Nymphaea, seeds of Nuphar do not possess an aril to assist their flotation in water, but the slimy and spongy pericarp tissue which contains small air or gas bubbles serves the same function in N. lutea. Smits et al (1989) concluded that Yellow Water-lily carpels had very poor buoyancy, and that after a day or so (or even less if it rains and releases the air from the bubbles), this spongy mucilage disintegrates and the seeds then immediately sink to the bottom. Since water transport appears to be the most obvious means of dispersal for aquatic macrophytes, the plant propagules of N. lutea appear to be strangely ill-equipped and poorly adapted by evolution for flotation over anything other than short distances. It is known that young seedlings of some aquatic macrophytes also possess powers of flotation and thus a secondary, additional phase of dissemination may occur at a later stage, as is the case with some emergents, eg Baldellia ranunculoides (Lesser Water-plantain) and also some submerged plants, eg Hottonia palustris (Water-violet) (Sculthorpe 1967, p. 329), but as far as we know this possibility has never been observed in N. lutea.

The findings of Mason (1975) with regard to the behaviour of long-established alien aquatic plants in New Zealand underline the poor dispersal abilities of many of them, including N. lutea. Mason showed that although these introduced species were locally abundant, they colonised other isolated water bodies only slowly, if at all, when they were spread deliberately, or accidentally, by man. Without this human intervention they were static and completely confined to existing sites.

Animals that live in, or frequent water, are often considered important dispersal agents, capable of carrying seeds or other transferable vegetative plant parts (Ridley 1930; Sculthorpe 1967). In the case of N. lutea, however, Smits et al. (1989) found no evidence of any adaptation to enable any imaginable form of such animal transport, either attached externally, or ingested and voided in a viable condition through the alimentary canal of fish (Carp), or birds (Mallard duck and Coot). However, viable seeds of N. lutea were once discovered in the excreta of a Heron. The seeds were presumed to have been eaten by a fish, which was subsequently eaten by the bird! Heslop-Harrison (1955a) also quotes the work of a Finnish researcher, who found that seeds removed from the gut of Ridd fish (Scardinius erythrophthalmus), germinated more readily than seeds sown directly from the plant.

More recent work in a sheltered muddy site on the River Rhone in France found that while seed production of N. lutea was at least 600 per m², not a single seedling of the species was recorded in numerous sample bare mud quadrats examined over an entire five year study period. The only explanation which could be offered was that the muddy sediment in question was very loose and easily re-suspended in water, so that the seeds could sink into the mud to a depth from which seedling emergence simply became impossible. The dynamics of the Nuphar population at this particular site therefore rested entirely upon vegetative extension of the rhizome system (Barrat-Segretain 1996). Water Hens and Grebes might assist in vegetative dispersal, since aquatic birds such as these have been observed carrying uprooted portions of Nuphar rhizome during their nest building (Heslop-Harrison 1955a).

Seed germination

The findings and conclusion of the Rhone work described above is rather surprising in the light of Dutch studies which indicate that germination in this species, while often erratic, is actually better under anaerobic conditions when compared with aerobic, provided the seed has been cold-treatment stratified. The same Dutch study also found that the presence of ethanol and ethylene helped stimulate germination in both Nuphar lutea and Nymphaea alba (Smits et al. 1995). Both seedling and mature N. lutea plants are also known to produce allelopathic compounds, eg resorcinol, plus a number of

alkaloids such as nupharolutine, which appear capable of inhibiting the growth of competing species (Elakovich & Yang 1996; Sutfeld et al. 1996).

Fossil record

The fossil pollen and seed record of N. lutea extends back through all the interglacials to the Pastonian. Both pollen and seed are also frequently recorded in all zones of the current Flandrian interglacial (also known as the Littletonian in Ireland), and pollen appears in low frequency throughout the Late Weichselian glacial stage (the Midlandian in Ireland). Godwin (1975) concludes, therefore, that there is a strong case for regarding the species as persisting in Britain and Ireland right through the Pleistocene, although as one would expect, it must have been restricted during the glacial stages.

Fermanagh occurrence

Nuphar lutea is four times more commonly recorded and at least twice as widespread in Fermanagh as Nymphaea alba (White Water-lily). There are records of N. lutea in 188 Fermanagh tetrads, 35.6% of the total in the VC. Yellow Water-lily is particularly abundant and ubiquitous throughout the whole of the Upper Lough Erne catchment, but is very much less frequent in Lower Lough Erne, where indeed it is absent from most of the shore.

Irish occurrence

In Ireland, N. lutea is much more common and widespread than Nymphaea alba, especially in the Midlands, although like the latter species, it also becomes rare or absent from most of the SE of the island.

British occurrence

Having smaller flowers that are much less decorative than White Water-lily, N. lutea is not nearly so likely to be introduced in English sites and the New Atlas therefore attempts to map the native British distribution. This shows frequent throughout C and S England, but absent from the extreme SE, the Channel Islands, and also from much of the N of Scotland including the northern isles. There appears to be little change in the distribution when it is compared with that in the 1962 Altas (Walters & Perring 1962; Preston et al. 2002).

European and World occurrence

In Europe, it is widespread throughout, extending to 67^oN, but thinning markedly southwards in the Iberian and Italian peninsulas, and is scarce throughout the Mediterranean basin, although present and rare in E Algeria (Jalas & Suominen 1989, Map 1514). Eastwards, it stretches through the Middle East, the Caucasus to C Asia, Siberia and Manchuria and there are closely related forms in N America (Heslop-Harrison 1955a; Hultén 1971, Map 158). As with Nymphaea alba, the wide range of N. lutea in Europe and Asia rules out broad regional climatic factors as agents creating the observed uneven British Isles distribution. This must instead reflect the availability of suitable water bodies. The fact that it is more or less a lowland plant again probably reflects its minimum pH (ie pH 6.0), rather than lower temperatures, although higher exposure and more frequent wind on heights might create wave and turbidity problems which would also limit the species (Heslop-Harrison 1955a).

Names

The genus name 'Nuphar' is derived from the Iranian 'nufar' or 'naufar' meaning or referring to a water-lily (Gilbert-Carter 1964). The Latin specific epithet 'lutea' meaning 'yellow' is derived from the name 'Lutum', of the Dyer's Wintergreen or Weld, Reseda luteola, which yields a yellow dye (Gilbert-Carter 1964; Stearn 1992).

Being a common and conspicuous water plant it is not surprising that N. lutea has at least 25 local English common names, of which five are shared with Nymphaea alba (Britten & Holland 1886; Grigson 1987). The name 'Blob' or 'Water-blob', is one of six names shared with Caltha palustris (Marsh-marigold). The word 'blob' or 'bleb' is Anglo-Saxon and means a bladder, bubble or blister (sometimes, as in the Ranunculaceae, indicating the capacity to raise a blister on skin), and the adjective 'blub' refers to something swollen, plump or round, all descriptive terms which could be applied equally well to flowers of Nuphar and Caltha (Prior 1879; Britten & Holland 1886; Watts 2000). Several names suggest the yellow flower in referring to butter and other allusions, for instance 'Butter Churn' and 'Butter Pumps' both refer secondarily to the shape of the fruit. The name 'Golland', 'Water Golland', or in numerous variant dialect spellings as, for example, 'Gowan', 'Gowlan' and 'Gowland', is shared not only with Caltha palustris (Marsh-marigold), but also with Trollius europaeus (Globeflower), the yellow buttercups and, indeed, with almost any yellow flower, or even with one which has a yellow centre, like Bellis perennis (Daisy). 'Golland' is thought to be derived from the Anglo-Saxon 'gold', or if we prefer a remote ancestry, the Suio-gothic 'gul', 'gol', meaning 'yellow' (Britten & Holland 1886). It could also be related or compared to the Norwegian 'gal' or 'gaul', again meaning simply 'yellow' (Grigson 1974). 'Clot' and 'Clote-leaf' are names applied to both N. lutea and Verbascum thapsus (Great Mullein). 'Clote' is Old English and can mean 'wedge', which is probably a reference to the large, broad leaf shape (Watts 2000). Other names and their allusions are discussed under Nymphaea alba, as are the herbal medicinal uses that Nuphar lutea shares with that species.

Threats

None.

Growth form and history of discovery in Fermanagh

This submerged, floating, truly aquatic species typical of eutrophic, still or slow-flowing water, was found in quantity on 6 October 2010 by Robert Northridge at a jetty near the National Trust's Crom Castle estate Visitor Centre on the shore of Upper Lough Erne. The Rigid Hornwort was floating in a tangled mat of aquatic species including Elodea nuttallii (Nuttall's Waterweed). A specimen of the hornwort was collected and sent to DBN for confirmation. Dr Matthew Jebb confirmed the identification and we were confident that this was a new County Record.

On checking this supposition with the CEDaR Database for N Ireland, Robert Northridge and the current author were very surprised to discover a previous Fermanagh record had been listed. The record details were given as: 8/9 August 1968 at Mill Lough near Bellanaleck, by staff of the Department of Agriculture & Rural Development NI, Fisheries Division. We presume the record was made during survey work on water quality or fish stocks. In checking this Mill Lough record, we found that no voucher exists, nor any note of the recorder's name (or names). CEDaR supplied three other plant records for the same date and lake: Littorella uniflora (Shoreweed), Phragmites australis (Common Reed) and Schoenoplectus lacustris (Common Club-rush). All three are common widespread species in Fermanagh and, together with five other common wetland species on a total list of 18 records across several lakes, they give us no real indication of the recorder's identification skill.

This particular Mill Lough (there are three others in Fermanagh) has been visited by RHN on three occasions (1979, 1985 and 2000) and also by the NI Lakes Survey 1988-90, without occurrence of C. demersum. RHN revisited Mill Lough on 13 October 2010 and made a thorough search of floating plants at seven jetties around the lake. C. demersum was not found, but masses of Water Crowfoot were present, either Ranunculus circinatus (Fan-leaved Water-crowfoot) or R. trichophyllus (Thread-leaved Water-crowfoot), with very short submerged leaves that might easily be mistaken for Rigid Hornwort by inexperienced botanists or non-botanical field workers. In view of all this, we have discounted the 1968 Fisheries record entirely and have asked for it to be removed from the CEDaR Database.

On mentioning the above error to CEDaR staff, Robert and I were disconcerted to be then given additional lake data for Fermanagh collected at two sites for NIEA by the English consultant company, ENSIS in 2006 and 2007, and from another four sites on Upper Lough Erne discovered by the NIEA's own Lake Ecology Team in August 2010. Since the timing of these October 2010 revelations coincided with evidence given to the current authors regarding the recent arrival of Elodea nuttallii in Upper Lough Erne, RHN carried out a speedy late season survey of Upper Lough Erne shores (3-5 November 2010) to search for these two species. This led to the discovery of C. demersum at a further eight sites in the southern part of Upper Lough Erne. Thus Rigid Hornwort is now known to have first appeared in the VC in 2006, and during the last four years it has spread to a total of 11 Fermanagh tetrads as shown in the map.

Irish occurrence

Although occasionally abundant in lakes, pools and canals, C. demersum remains a relatively rare species in Ireland. The 1987 Cen Cat Ir Fl 2 listed 17 VCs where it had been recorded at least once. Inspection of the New Atlas hectad map using a transparent overlay of the 40 Irish VCs (not as accurate or as straight forward an operation as it sounds) increases the number of VCs where it has occurred up to the year 2000 to 23. The VCs featured in these two sources do not completely agree, however, so while a wider distribution is indicated, we cannot conclude that the species has spread to six additional VCs. The major change in C. demersum distribution in NI has been its arrival in Lough Neagh in 1988 and its subsequent spread around the lake basin. Nevertheless, this was preceded in Co Tyrone (H36) by a find in 1982 at Ballagh Lough, 5.5 km NE of Fivemiletown (McNeill 2010). This site is only 4 km from the boundary of Fermanagh.

Although it flowers regularly and freely in B & I, Rigid Hornwort very rarely sets seed, reproducing almost entirely by vegetative fragmentation. While it is considered native in some parts of Ireland, the species is commonly used to oxygenate garden pools and indoor aquaria. We therefore suspect that the Crom colony at least is an introduction originating from recently discarded cultivated material. Crom jetty is a heavily visited public amenity.

British occurrence

C. demersum is widespread in lowland aquatic habitats across the length and breadth of Great Britain. However, it is very unevenly spread, being much more frequently recorded in the English Midlands and the South East of the country (Preston et al. 2002). This pattern of distribution, taken together with the fact that fruiting in Britain and Ireland is rare at present, suggests that temperature may be limiting both the occurrence of the species and its sexual reproduction.

Rigid Hornwort is not confined to lakes, but occupies a wide variety of still and slow-moving eutrophic habitats including ponds, rivers, canals and ditches. In smaller water bodies such as ponds and ditches, it often forms large, dense mats of growth that can rise above the water surface. Comparison of the New Atlas hectad map with the 1962 BSBI Atlas shows it is now more frequently recorded, but this is probably a reflection of better co-ordinated plant recording, rather than a real increase in presence. In any event, at least the distribution in Britain appears stable, if not definitely increasing (C.D. Preston, in Preston et al. 2002).

World occurrence

Rigid Hornwort is an extremely widespread circumpolar species, not only in Eurasia and N America, but also in C and S America, C and S Africa, S India and S Australia (Hultén & Fries 1986, Map 817).

Fossil record

The long-spined fruits of C. demersum are very easily recognised fossils and the sedimentary record shows that this species has persisted in Britain as far back as the Cromer Forest Bed series and through all the subsequent interglacial periods including the present Flandrian. The record shows that it was also present in the less severe terminal parts of the glacial stages. Godwin (1975) quotes Samuelsson (1934) regarding the post-glacial immigration of C. demersum back into northern Finland in the early Post-glacial period as glaciers retreated, where fossil records show the species established a considerable distribution. However, the record in Finland also demonstrated a subsequent contraction and readjustment of the species distribution, possibly due either to changes in water chemistry or to subsequent climate deterioration.

A study of lake deposits on the Scottish isle of Skye by Birks (1969) suggests that C. demersum suffered a restriction of range southwards since the Late Weichselian glacial period, a notion which supports the case for a present day temperature limitation on the species.

Names

The genus name 'Ceratophyllum' is from the Greek 'keras', meaning a horn, and 'phyllon', a leaf, the many divisions of the leaves suggesting the shape of horns. The Latin specific epithet 'demersum' means 'down under the water' or 'submerged'. The English common name 'Rigid Hornwort', is a so-called book name and the plant does not appear to be associated with any folk traditions. Despite having 'wort' as part of the name, the species does not have any use in herbal medicine that the present author can detect.

Growth form and history of possible Fermanagh discovery

Although this free floating or lightly anchored aquatic has a similar ecology and reproductive capacity to C. demersum (Rigid Hornwort), in Ireland it is extremely rare, having been mapped in the New Atlas in a total of just three hectads in two widely separated regions near the east coast in Cos Wexford and Down (H12 and H38). A water quality survey carried out by the Water Management Unit (WMU) of the Northern Ireland Environment Agency (NIEA) at 15 sites around Upper Lough Erne in August 2010 produced two unexpected records that were communicated to the current authors in late October, 2010. The two sites were the one listed above and another at Knockninny Quay. Plant material was located at depths of 0.7-1.6 m at Cornaleck, and a small quantity only was found at a depth of 1.6 m at Knockninny Quay.

C. submersum was first recorded in Ireland in August 1989 when the NI Lakes Survey discovered it in two lakes on the Lecale Peninsula near Downpatrick (Smith & Wolfe-Murphy 1991). To date, these are the only known sites for Soft Hornwort in Northern Ireland. We therefore made enquiries to ascertain if vouchers had been collected at the Upper Lough Erne sites, and whether or not the identification of the species material had been checked by a known botanical authority. Plant material had indeed been collected and was identified in the WMU laboratory in Lisburn. Unfortunately, specimen samples were not retained. The determination of C. submersum from C. demersum was based entirely on the dividing of the leaves, "into threes, rather than dichotomously" (B. Walker, pers. comm. 18 February 2011).

The significant leaf distinction between these two Ceratophyllum species refers to the number of times the leaf segments divide dichotomously: once or twice forked for C. demersum, three or four times forked for C. submersum. The leaf segments are obviously toothed in C. demersum and much less so in C. submersum, and the leaves are more rigid and darker in colour in C. demersum in comparison with C. submersum (Stace 1997).

Without vouchers these two C. submersum records are very doubtful and cannot stand as First and Second County Records. We believe the plants were incorrectly identified. However, the possibility remains that Soft Hornwort just might be present in eutrophic, base-rich Fermanagh waters growing along with C. demersum and it should certainly be looked for. The point has been well made by the recently retired Curator of the BEL herbarium, that, "it is a regrettable tendency these days for agencies and consultants not to collect or retain vouchers. It undermines all the time and expense spent in doing surveys." (P. Hackney, pers. comm. 23 February 2011).

C. submersum is reputed to flower more freely in Britain and Ireland than C. demersum, and although vegetative reproduction by fragmentation is certainly the norm in both species, the small, hard, nut-like fruits may well be dispersed by waterfowl, thus enabling jump dispersal of the species (Preston & Croft 1997). On account of wider geographical distribution and biodiversity aims, C. submersum is included on the NIEA list of Priority Species of special concern requiring local conservation action.

Names

The genus name 'Ceratophyllum' is from the Greek 'keras', meaning a horn, and 'phyllon', a leaf, the many divisions of the leaves suggesting the shape of horns. The Latin specific epithet 'submersum' means 'submerged'. The English common name 'Soft Hornwort', is a so-called book name, and the plant does not appear to be associated with any folk traditions. Despite having 'wort' as part of the name, the plant is rare and does not have any use in herbal medicine that the present author can detect.

Growth form and preferred habitats

A perennial with a short, thick, rather tuber-like horizontal rhizome, C. palustris occurs in a very wide variety of wet habitats, including the margins of rivers, streams and ditches, in wet woods and fen carr around lakes and ponds where it grows well in partial shade in low-lying ground. Vegetative growth begins early in the year and, like Ranunculus ficaria (Lesser Celandine) and Anemone nemorosa (Wood Anemone), it is pre-vernal in its flowering when growing in deciduous woodland. It also occurs more locally in seasonally flooded meadows and pastures where it can sometimes form large dominant patches, but it does not tolerate permanently flooded ground (Grime et al. 1988).

Typically C. palustris prefers neutral to base-rich mineral or fen-peat soils, rather than very acidic ones and it is, therefore, rare or absent on boglands and on soils with a pH below about 4.5 (Grime et al. 1988). Like several other members of the Ranunculaceae, it tends to frequent areas of disturbed ground with relatively high exposures of bare earth and it often grows on banks of streams and lakes close to the water's edge where fluctuating water levels help minimise competition from taller, tufted, potentially dominant emergent species (Grime et al. 1988). As with most wetland species, it is chiefly found in lowland situations, but it is known to ascend to 1,100 m in Scotland (Wilson 1949). With increasing altitude the species occurs more frequently in open, completely unshaded situations.

Reproduction

In the normal upright growth form of the plant there does not appear to be any form of vegetative propagation, rather reproduction and dispersal relies entirely on seed. However, Grime et al. (1988) reported that shoots detached during disturbance are capable of regeneration, a feature which deserves to be further investigated, since, for example, trampling by grazing livestock and drain cleaning operations must often result in plant fragmentation.

The showy and very familiar flowers of Marsh-marigold are produced early in the year from March onwards and, depending on the season, they continue flowering through to July or even August. The flowers are petal-less but have five to eight sepals, glossy yellow above, often greenish beneath, which take on the protective and insect-attracting, advertising roles of petals. Flowers of C. palustris are self-incompatible and are visited and pollinated by a variety of insects, including flies, bees and beetles, which collect both nectar (secreted by the several carpels at their base and, therefore, partly concealed, though generally copiously produced) and plentifully produced pollen as food sources (Hutchinson 1948, p. 139; Proctor & Yeo 1973; Fitter 1987; Jonsell et al. 2001).

Fruit and seed production

The distinctive star- or crown-like fruit (hence the English common name 'Kingcup'), usually contains five or six compartments (ie follicles), which split open along their upper sides to reveal a double row of seeds (Melderis & Bangerter 1955). Salisbury (1942) found that a considerable proportion of the ovules formed normally abort, so that while the usual number of ovules per follicle is 17 or 18, the number of seeds is generally much fewer, the calculated mean being nine per follicle. The average plant examined by Salisbury (1964) produced 2,700 seeds, though as Butcher (1961) earlier pointed out, the species is very variable in both reproductive and vegetative respects, and therefore such estimates should be regarded as merely a guide.

Seed dispersal

There is no specialised seed dispersal mechanism: seeds simply drop out of the boat-shaped open follicle and since they possess a corky swollen attachment, they may float for between one and four weeks (Ridley 1930, p. 198; Jonsell et al. 2001). After flowering the leaves of the plant have been observed to increase very considerably in size (Step & Blakelock 1963).

Seed germination and longevity

Caltha seeds have a well-defined chilling requirement which ensures that they remain dormant during winter and germinate in the spring (Grime et al. 1988). A survey of soil seed bank data in NW Europe has provided information from 18 sources, 12 of which regarded the seed as transient, two as short-term persistent (ie surviving for 1-5 years), and two as long-term persistent (ie surviving for at least five years). The two remaining studies were undecided as to which category the species fitted (Thompson et al. 1997). Clearly there is no consensus on this important matter and further study is required.

Genetic variation and the 'creeping' form: In the Typescript Flora of Fermanagh, which details finds up until the early 1950s, Meikle remarked on variants of C. palustris with slender, hairless, creeping stems rooting at the nodes and bearing sparse inflorescences of small flowers that are frequent on lake shores and in wet woods in Castlecoole and elsewhere in Co Fermanagh. The extremes of variation in C. palustris are very distinct and the different forms are said to remain distinct (though rather less so) in cultivation. However, it is not easy to distinguish them satisfactorily from typical C. palustris at a taxonomic level, because numerous intermediate forms exist. Stace (1991, 1997) believes that these few-flowered, procumbent plants are best recognised as var. radicans. In the critical Flora Nordica volume 2 (Jonsell et al. 2001), Piirainen elevates this variety to subspecific level as subsp. radicans (T.F. Forst.) Syme. Whatever we decide to name it, this creeping form was first recorded in Fermanagh by Barrington (1884) as C. radicans (Forst.), and although Meikle states that the variant has been recorded by him and his co-workers in all four of their district subdivisions of the county (Meikle et al. 1975), subsequent local recorders have tended to ignore the form, and in fact there are just three records of this form of the plant in the Fermanagh Flora database. The prostrate creeping form of the plant spreads both by seed and by rooting shoots, so it definitely demonstrates vegetative reproduction. The seeds of this form of the plant lack the corky appendage and do not float, which must restrict its dispersal considerably (Jonsell et al. 2001, p. 331).

Genetically, in terms of chromosome numbers, C. palustris contains a polyploid series ranging from diploids to octoploids, plus a number of aneuploid

forms. However, the chromosome variation is not correlated with variation in form, nor with geography (for details see Jonsell et al. 2001, p. 329).

It is interesting to note that argument regarding the taxonomy of C. palustris and the status of its varieties has rumbled on since 1768, when Philip Miller produced the 8th edition of his Gardeners Dictionary. [NB There is no apostrophe in Gardeners in the original work]. In this work he proposed distinguishing Caltha minor for montane (ie upland) plants having round, crenate, heart-shaped leaves and smaller flowers than normal C. palustris. This suggestion was made only 15 years after the Swedish taxonomist, Carl Linnaeus, described and named C. palustris (Panigrahi 1955). Despite its age, the author of this latter paper has conveniently summarised the history of the debate on the several forms of C. palustris, their morphology, cytology and geography. Panigrahi gives full weight to several significant contributions based on Irish material by Robert Lloyd Praeger. Anyone requiring a more modern treatment of the subject should consult Woodell & Kootin-Sanwu (1971) or Jonsell et al. (2001).

Grazing and toxins

The fact that it can become so prevalent in wet grassland is unfortunate for the farmer and his animals, since like other members of the Ranunculaceae, Caltha contains a significant quantity of the dangerous toxin protoanemonin (for details of this poison see the Ranunculus acris synopsis). This makes plants acrid tasting and they are avoided by stock animals to such an extent that Cooper & Johnson (1998) have not reported any recent cases of poisoning in the British Isles. Salisbury (1964) comments that C. palustris acts as a purgative and that it leads to diminished milk production in cattle which graze it. He also reports it being known to cause death in both cattle and horses.

Fermanagh occurrence

Marsh-marigold is the tenth most frequently recorded vascular plant in Fermanagh in terms of records and it is the ninth most widespread species in the VC, being found in 376 tetrads, 71.2% of our total tetrads. It is absent only from the highest altitudes, permanently flooded, or excessively acidic soils below pH 4.5.

British and Irish occurrence

C. palustris is common and very widespread throughout Great Britain, but is absent from areas of intensive farming such as Cambridge. It is also common and widespread in Ireland, but more or less absent from the more acid boglands of the S & W. Rather remarkably, it is a rare and introduced species in the Channel Isles (Garrard & Streeter 1983; Preston et al. 2002). Despite the destruction of many wetland areas in Britain and Ireland during the past 40 years, the index of change calculated and provided in the New Atlas is -0.26. At least at the hectad scale, this suggests there is little evidence of any change in the species distribution in these islands (R.A. Fitzgerald, in: Preston et al. 2002).

Fossil record

Seed and pollen of C. palustris are well represented in the fossil record although Caltha type pollen also includes Aquilegia vulgaris (Columbine). Godwin (1975, p. 119) suggests that the abundance of site records in the Late Weichselian ice age, strengthens the possibility that the species, "survived the glaciations in this country". One has to presume that as an English writer Godwin was referring to his native island (Great Britain) alone, but perhaps he did not mean to be so insular, and really meant to specify periglacial survival of C. palustris in all unglaciated areas within Britain and Ireland. An opinion on this matter from a suitably qualified authority would be appreciated.

World distribution

In Europe, C. palustris is widespread in both temperate and Arctic areas, but in Europe it becomes scarcer towards the south and around the Mediterranean, reaching only N Portugal. It is absent from the Greek Peloponnese and from all the Mediterranean islands (Jalas & Suominen 1989). Due to differing views of the taxonomy of Caltha palustris, the species has been mapped in different ways; for example, Hultén recognises and maps several different subspecies so that it becomes necessary to mentally combine these variants to achieve an overall picture (Hultén 1971, Maps 75 & 76). Once this is done, C. palustris in this broad sense is shown to be circumpolar in Arctic and boreal, northern temperate regions.

Uses

Considering the fact that the plant is not a Near Eastern nor a southern species, and thus it never attracted the attention of the older Classical civilisations who handed on a knowledge of plants and their uses to Mediaeval and Renaissance Europe, the long list of English common names that Grigson provides must derive almost entirely from the colourful show the plant makes in the countryside and in the water garden. Its prominence, even in habitats disturbed by man and his animals, and the religious and folk beliefs associated with it, rather than any widespread use in folk or herbal medicine, must have made Caltha palustris very much a favourite plant (Grigson 1987).

Being a severe irritant and causing blisters, it was widely known to have caused serious effects when attempts were made to use it in herbal medicine. Nevertheless, it did have some medicinal use associated with mythical beliefs on the treatment of fits (Grieve 1931, p. 519). The leaves were also boiled and eaten like spinach and flower buds were occasionally used like capers − although even Grieve regarding this type of usage as "rather inadvised" [my italics].

Names

The genus name 'Caltha' is derived from the Greek 'kalathos' meaning 'a cup', a reference to the bowl-like shape of the flowers. The Latin specific epithet 'palustris' means 'of marshes' (Melderis & Bangerter 1955). C. palustris has a huge number of English common names. Grigson (1987) lists no fewer than 93. The most widespread English name, 'Marsh-marigold', refers to its use in church festivals in the Middle Ages, as a flower devoted to the Virgin Mary. Several other names also refer to Mary and to May-Day festivals (Grieve 1931; Grigson 1987).

Threats

Many of the plant's meadow sites are vulnerable to drainage operations and could easily be destroyed. However the plant is so widespread and abundant in Co Fermanagh that the threat level to the species is nil.

Growth form and preferred habitats

In Britain and Ireland, this rather lovely perennial typically occurs in cool, damp, often shaded habitats, including lake margins, stream and river banks, open woodland or their margins, traditionally managed hay-meadows and rarely on upland rock ledges where the plants may never flower. It needs protection from grazing by sheep and goats to survive. Even when it is not flowering, the large basal palmate leaves of the species are perfectly recognisable. The plant prefers basic soils and is generally associated with limestone districts, and especially in sites where there is some degree of nutrient enrichment from mesotrophic flushing groundwater.

Fermanagh occurrence

T. europaeus occurs, sometimes in considerable quantity, in just one very specialised Fermanagh habitat. It occupies a narrow zone about 2 or 3 m wide, just below the winter high water-mark, on the rocky limestone shores of two of the larger lakes in the county under the light shade and shelter of alder, willow and rose scrub woodland. The species has been recorded from a total of five Fermanagh tetrads on the shores of Rosskit, Gorminish Island and Bilberry Island on Lough Melvin, and from Rushin Point and a couple of nearby sites on Upper Lough Macnean, where it also occurs on the Co Cavan side of the lake (Northridge 1995). RHN has visited the Upper Lough Macnean shore when only the flowers and upper leaves were emergent from the water surface, presumably after a period of heavy May rainfall!

Irish occurrence

Apart from the sites just mentioned, the only other extant Irish stations for T. europaeus are in Co Donegal (H34 or H35 or both?) (three or four sites), although previously it also occurred in Co Leitrim (H29). It might well survive on another, different Bilberry Island, which occurs in Co Leitrim, the one that lies just south of Patrick's Island, but this still needs to be investigated (Northridge 1995).

British occurrence

T. europaeus is a decidedly northern species in the British Isles, being much more common in Scotland and Cumbria than elsewhere. The British distribution is totally confined to the area north of a line between Cardiff and Whitby (R.A. Fitzgerald, in: Preston et al. 2002). This distribution pattern is very similar to that of Cirsium heterophyllum (Melancholy Thistle) (another Irish rarity present in Fermanagh), and in Britain the two species are often associated in damp calcareous grassland (Clapham 1978).

European occurrence

In Europe, T. europaeus is native and widespread throughout N and C regions, but it is absent from most of the westernmost parts of the continent, although widespread and frequent again in the mountains further south from Spain eastwards to Greece (Jalas & Suominen 1989, Map 1557). In Scandinavia, it extends beyond 70°N and it also stretches eastwards to W Siberia (Jonsell et al. 2001), and on the margins of its indigenous distribution, it is occasionally recorded as an established introduction (Jalas & Suominen 1989).

It was been shown by Conolly & Dahl (1970) that the distribution pattern of T. europaeus lies north of the current 27°C maximum summer temperature isotherm, quantifying the basis on which it occurs in Britain and indeed on most of the European Continent. In these areas, it is considered a species of northern or montane to sub-alpine damp, calcareous meadows and pastures, stream banks, the margins of woods and rarely on cliff ledges (Halliday 1997).

Indigenous populations in Scandinavia occupy a rather wider range of habitats than in Britain and Ireland: there it occurs in varying conditions of light, open to moderately shady habitats, on moist to mesic, nutrient-rich soils that are usually refreshed and enriched with moving groundwater. Typical habitats include tall-herb vegetation, scree, fen margins, Willow and Alder thickets, and the banks of water courses. Also in Scandinavia, T. europaeus is closely associated with man-managed 'apophytic' situations, including wooded pasture, forest glades, former hay-meadows, field margins and roadsides (Jonsell et al. 2001). In many of these latter more regularly disturbed stations, T. europaeus, in Scandinavia at least, is recognised as an established introduction, its seed very probably being transported around the countryside by cutting machinery and the hay fodder market.

Decline in range and environmental pressures

In both Britain and Scandinavia, T. europaeus appears to have declined in range over a long period, probably for at least a century. As usual in such situations, decline and retreat is particularly obvious at the margins of the previous distribution. (Halliday 1997; Jonsell et al. 2001; R.A. Fitzgerald, in: Preston et al. 2002). In part, T. europaeus appears to be retreating under pressure from grazing animals, yet like other members of the Ranunculaceae, it contains the blister-inducing bitter toxin protoanemonin, which in theory ought to deter browsers and give the species a measure of protection (Cooper & Johnson 1998). However, further north, Arctic Reindeer are known to graze T. europaeus and other northern representatives of the plant family, and Globeflower seeds are internally transported and occur in reindeer excreta (Ridley 1930, p. 373). The levels of toxin in Trollius may be insufficient to deter deer and other large herbivore species, so it might be possible for cattle, sheep and goats to browse on it without suffering ill effects, again assisting the plant in its seed dispersal. Work is required to investigate these possibilities in Britain and Ireland.

In Britain, the main cause of the obvious decline in the species appears to be the agricultural improvement of land (especially in hilly areas), by drainage, the earlier cutting of herbage for silage than previously for hay, and the widespread application of fertilisers stimulating the growth of more vigorous species with which T. europaeus simply cannot compete (Halliday 1997; R.A. Fitzgerald, in: Preston et al. 2002). In Ireland, populations on or near river and stream banks have also been destroyed by flash floods (eg in Co Donegal), a weather related phenomenon which seems to be increasing in frequency and may well be related to global climate change (Curtis & McGough 1988, p. 100; Sheppard 1991).

Garden escapes and relict status

Globeflower is a longstanding, popular garden border perennial that is undemanding and easy to cultivate. While it must occasionally 'jump the garden fence', as it is claimed it does in Scandinavia (Jonsell et al. 2001), in Britain and Ireland there are only a few stations on the New Atlas map which suggest or indicate such a happening (Preston et al. 2002).

In Ireland, T. europaeus and the other members of the Northern Montane phytogeographical group are regarded as relict plants of past climate stages and much changed environments that are approaching the end of their local occupation (Matthews 1955, p. 117). This understanding fits very well with the restriction in habitat found in Fermanagh, the contraction in range elsewhere, especially in Ireland, and with the developing picture of Global warming (or in our corner of the British Isles, more accurately Global wetting and blowing) (Plantlife Report: Death Knell for Bluebells? Global Warming & British Plants, Anon. 1991).

In view of the large number of apparently suitable sites for T. europaeus on Lower Lough Erne, it is sad to think that the species nowadays appears to show no capacity whatsoever for natural dispersal and colonisation over the relatively short distances involved (8-13 km).

Fossil record

The fossil record is sparse but is sufficient to confirm the native status of Globeflower (Webb 1985). It consists of a tentative Flandrian post-glacial seed record from Ireland of Mesolithic age, and pollen records exclusively from Scotland throughout the entire post-glacial (Godwin 1975).

Sexual reproduction

T. europaeus has a short, stout, erect, fibrous rootstock and it has no powers of vegetative reproduction, instead relying entirely on seed production for increase and dispersal of the species (Clapham et al. 1962).

Insect pollination

Globeflower has a most interesting relationship with three species of small flies belonging to the genus Chiastochaeta. The globular yellow flowers produced from June to August never open their overlapping petal-like sepals, yet despite this (or maybe because of it and the concealment and protection this floral behaviour affords the flower's pollen and nectar food supplies), they are visited by adults of these species of fly which meet, mate and feed on the hidden food sources. Individual Trollius flowers last for only about five or six days and they are self-incompatible. After mating inside one of the closed flowers, the female insects lay their eggs (usually just one per flower).

The Chiastochaeta fly species differ slightly in their sexual behaviour, for example, the stage of flowering at which they lay their eggs, the exact positioning of the eggs, and the paths along which the larvae bore inside the carpels during their development, and in this manner the fly species manage to avoid competition to a remarkable degree, ie strict resource partitioning is present (Pellmyr 1989). The insect species which lays its eggs earliest during the short life of the flower is the one which is most effective in achieving pollination, but this is entirely incidental as far as the insects' movements are concerned.

Flower-insect mutualism

Each flower is reputed to produces a rather large number of ovules (around 400 or so) and only a few young seeds of each flower are eaten by the fly larvae, so the cost to the plant is fairly minimal. This is reckoned to be a fine example of an obligate mutualistic relationship occurring in the temperate zone. Mutualism is a phenomenon that is associated much more frequently with biology in the tropics (Pellmyr 1989). Although there is a cost to the plant, the relationship with these small flies essentially benefits both partners, and the trade-off by the plant of a few of its many seeds enables cross-pollination of its un-opening, self-incompatible flowers.

When fertilised, the numerous carpels of T. europaeus develop into many-seeded follicles, which when ripe split to release the seed.

Seed dormancy and germination

The seed is dormant at dispersal and requires cold-stratification during the following winter to break dormancy (Milberg 1994). Experimental germination was equally effective in both light and dark treatments, which suggests that seeds might germinate even when they are too deep in the soil for seedlings to emerge. In turn this suggests that Trollius does not form a persistent soil seed bank and indeed this appears to be the case. In an experimental study comparing Primula veris (Cowslip) with T. europaeus, after 16 months burial, 85% of the Primula seeds, but only 8% of the Trollius seeds remained viable (Milberg 1994).

Competitive ability

T. europaeus is a poor competitor and in Finnish meadows has been known to suffer a prolonged delay in maturity, successful flowering and fruiting being curtailed for up to eight years. In the absence of competition, the species can fruit in its first season (Linkola 1936, quoted in Salisbury 1942, p. 54). The species is polycarpic and probably long-lived, continuing to fruit for a number of years after attaining maturity, so that the loss of between two and seven years' seed output in Finland may represent only a small percentage of the total production throughout the plant's life. However, the prolongation of the juvenile phase represents a greatly increased risk of mortality prior to any reproduction and this could be a significant factor determining the survival ability of the species in a particular plant community (Salisbury 1942). This delay in plant maturity is sometimes referred to as 'the demographic penalty', affecting the overall 'fitness' of a plant population (Silvertown & Lovett-Doust 1993, p. 158).

Plant and flower size plasticity

The flowers of T. europaeus at the Fermanagh lakeshore sites are much smaller, approximately a third the size of those the author has often observed elsewhere, including those in Teesdale, Co Durham (VC 66) (Clapham et al. 1978) and regularly found in the Swiss and Italian Alps and the Pyrenees at very much higher altitudes. This suggests a particular genetic clone with smaller flower occurs in Fermanagh, perhaps a response to a challenging waterside environment.

All plants allocate their limited photosynthetic energy and mineral nutrient resources under constraints and limits imposed by their local environment which ultimately acts upon the plant genotype, but is expressed in the readily observed and more easily measured variation of the population phenotype (Silvertown & Lovett-Doust (1993). Conflicting demands inevitably lead to trade-offs between different activities. Two classes of trade-off are virtually universal: (i.) a trade-off between reproduction and other activities which is manifest as 'the cost of reproduction', or in proportional terms as the 'reproductive effort' (RE); and, (ii.) a trade-off between the size and the number of offspring produced. For perennial plants, absolute allocation to reproduction usually increases with plant size. However, depending on the species under examination, the proportional reproductive allocation, ie the reproductive effort (RE), can either decrease, increase, or remain independent of plant size.

Comparative reproductive effort

A study of reproductive effort of populations of T. europaeus and Ranunculus acris (Meadow Buttercup), in subarctic Swedish Lapland compared the species at differing altitudes up to and above the tree-line (at 870 m). This showed that for both species plant size, measured in terms of 'mean plant biomass' (ie mean dry weight), was approximately two or three times larger below the tree line compared with above it (Hemborg & Karlsson 1998). For Trollius, plants at the high altitude sites showed no relationship between RE and size, while below the tree line, RE varied inversely with plant size, ie reproductive effort generally decreased as the Trollius plants became larger. This might be due to a limitation imposed by the number of flowering meristems the Swedish population could support, since all the plants in the study, irrespective of size only bore one flower per stem. Moreover, floral structures, eg flower size and ovule number, may be limited by low plasticity, features which are very likely both genetically and environmentally regulated (Schmid & Bazzaz 1992).

At low altitudes in Britain and Ireland, and in the French Alps, most plants of T. europaeus have two or three flowers per stem, the lateral flowers usually being smaller than the terminal. This fact alone would produce different size-effects on reproductive effort in comparison with this Swedish study (Hemborg & Karlsson 1998).

Toxins and herbal use

Globeflower contains protoanemonin, the same toxic principal present in other members of the family. Being a decidedly local plant, T. europaeus has never had the herbal medicinal reputation and uses of its more common and readily available relatives. The exception to this is perhaps found in Russia and parts of Sweden where the species is very prevalent (Grieve 1931).

Names

The genus name 'Trollius' is a Latinised version of the older Swiss-German name 'Trollblume', first coined by Gesner in the 16th century, which translates as 'rounded flower' (Gilbert-Carter 1964; Gledhill 1985). 'Trollblume' may well be a contraction of 'die rolle Blume', referring to the rolled in, or closed in, petals of the flower (Grigson 1987, p. 31). The Latin specific epithet is geographical and obvious.

There are 16 local English common names listed by Grigson (1987), of which various forms and spellings of 'Locken Gowan' are the most frequent. 'Gowan', 'Gowlan' and so on are derived from the Anglo-Saxon 'gold' meaning 'yellow', while 'Locken', 'Lockety', 'Locker', 'Lapper' and so on, refer to the locked or closed petaloid flower parts (Britten & Holland 1886; Grigson 1987). 'Locken Gowan' has also been sometimes used to refer to Caltha palustris (Marsh-marigold) and there is a fair degree of overlap with some of the other English common names between these two species. Other local names are more predictable not to say prosaic, eg 'Goldilocks', 'Golden Balls' and, of course, 'Globeflower'. The current author particularly likes the imagery generated by the name 'Bull-Jumpling', which hails from Kinross-shire: the 'bull' part is most probably a corruption of the Old English, 'boll', meaning, 'any globular body' (Prior 1879). The origin of 'Jumpling' we can only guess at. Another name mentioned by Gerard (1597) in his Herball, he being the first English writer to mention the whereabouts of the species in northern England, was 'Troll floures', or 'Troll flower'. Grigson (1987) is dismissive of this name, emphasising his belief that Trollius is not a flower in any way linked to the evil Nordic trolls. Prior (1879) disagrees though, comparing the name with the Scottish 'Witches Gowan'. According to him both these names were given on account of the plant's acrid poisonous properties.

Threats

In Fermanagh, reclamation of shore lines and, at least on Bilberry Island, grazing by goats. Potential shoreline building development could also prove deleterious.

Names

Some authorities derive the genus name 'Helleborus' from the Greek 'helein' or 'elein', meaning 'to kill' or 'to injure', and 'bora', meaning 'food', indicating the poisonous properties of the plants (Melderis & Bangerter 1955; Cricheley Plowden 1972). Other writers suggest the name 'helleboros' was the ancient classical Greek name for H. orientale (Gilbert-Carter 1964; Stearn 1972).

The Latin specific epithet 'foetidus' translates as 'stinking' or 'bad smelling' (Gledhill 1985).

The English common name 'Stinking Hellebore', listed above, is a straight translation of the botanical name and as such is a mere book name (Britton & Holland 1886). Other, much more interesting common names with genuine, local folklore connections do exist (Prior 1879; Britten & Holland 1886; Grigson 1955, 1987). These include 'Bear's-foot' and 'He-barfoot' (presumably referring to the digitate leaf shape), and 'Setter' and 'Setterwort', the latter a name used in his early English herbal by John Gerard (1597). The derivation of the 'Setter' names, which are associated with healing cattle, is given in this work under Helleborus viridis and will not be repeated here. Other local English names for H. foetidus with cattle connections are 'Gargut root' (originating in parts of Norfolk), 'gargut' being "a disease incident to calves" (Britten & Holland 1886, p. 199), and 'Ox-heal', from the Anglo-Saxon 'oxnalib', again referring to settering cattle with the plant root (Prior 1879, p. 174).

Irish and British occurrence and status

In Ireland, H. viridis has always been accepted as being a persistent alien introduction, but in Britain small, permanent, supposedly or traditionally native populations are widely distributed. The plant usually occurs on chalk or limestone in shady lowland habitats such as wooded glades, rocky stream-sides and in old hedgerow banks. However, H. viridis has been grown in gardens throughout Britain and Ireland since medieval times (Harvey 1990; Landsberg 1996) and it was first recorded 'in the wild' around 1562. It is therefore difficult or virtually impossible to distinguish native from introduced populations and, after so long a period in cultivation, it appears rather foolish to even try to do so.

Fermanagh occurrence

H. viridis has been rarely recorded in ten tetrads, seven of them with post-1975 records. In the late 1940s and early 1950s, when Meikle and his companions were exploring the flora of Fermanagh, they recorded five widely spaced stations of the plant. Nowadays we know of six scattered populations, including two that they originally discovered − at Rossclare Bay and near Monea Castle. The typical habitats include woods, scrub, shaded river banks, and on scree below cliffs, usually on limestone. Meikle recalled seeing the plant in scrub on the limestone shore of Lower Lough Erne at Rossclare on family holidays in the 1930s (Carrothers et al. 1947), and a large patch of about 18 plants survives to this day in the same area. At Coffey's Ford, SW of Kinawley, there are 20 large patches growing beside a stream, clearly indicating that the plant is reproducing successfully.

In the Revised Typescript Flora of Fermanagh, Meikle remarked that the plant was often found far removed from gardens and that it appeared, "perfectly spontaneous". He also suggested that some of the existing populations might have been derived from stocks of the plant deliberately cultivated for cattle medicine (see below) (Meikle et al. 1975).

Growth form and reproduction

The species has a short, stout, ascending, blackish rootstock as its perennating organ, but the attached pair (or more) of leathery radical leaves and the aerial stem do not overwinter. Being a long-lived perennial, the current rootstock of the plant tends to be surrounded by a substantial woody cluster of old decaying stem-bases left from previous years (Ross-Craig 1948, Part 1, plate 38), but the plant has no real means of vegetative reproduction.

Growth begins early in midwinter, and the individual plant produces its two to four yellowish-green flowers around the second half of February or early March. Flowering generally continues on into April. Each flower contains 9-12 pocket-like green nectaries that attract early flying bees as pollinators. The insects collect both nectar and pollen from the flowers, both of which are urgently required to feed the developing brood of the bee colony in the spring.

Seed production and snail dispersal

The flower usually has three carpels and the fruit is a many-seeded follicle. The individual seeds are rather large, 4.5 × 3 mm, and dark brown in colour. Salisbury (1942) examined a small sample of 25 plants and calculated the mean annual seed output as 191, ± 48, per plant. The seeds possess an obvious appendage growing out from the seed coat. This is an edible elaiosome or oil-body that attracts ants which carry off the seeds when they are shaken out of the open fruit onto the soil or other surface, thus assisting the species' dispersal (Beattie 1985).

It has also be reported that the elaisome of the related species H. foetidus attracts snails which devour the oil-body and in so doing get some of the seed adhering to their slimy body. Dymes (1916) observed that the snail resented the presence of seeds on its head or tail and actively sloughed them off. However, the snail was quite unconcerned if the seed stuck to its body near the shell on the head side. A snail was observed to carry a seed in this manner for a distance of 35 cm. The Garden Snail (Cornu asperum = Helix aspersa), on average travels 5 cm per minute and Dymes measured them travelling a distance of 5.4 m for food. While molluscs move slowly, they do get around and they are abundant and widespread in many types of plant community across Britain and Ireland. If snails commonly carry seeds in the manner described, they would certainly play a significant role in seed dispersal (Ridley 1930, p. 150).

This is yet another plea for a return to careful natural history observation, something for which people in the British Isles were once famous, and could be again. No laboratory required; just eyes, imagination, time, patience and a notebook, although a hand-lens and a stereo-zoom low-power microscope would certainly also be useful!

The solitary determination of the soil seed bank in the NW European survey indicates that the Green Hellebore seed is transient (persisting in the soil for less than a year) (Thompson et al. 1997). A correspondent of Salisbury's indicated that the rate of germination success is low (Salisbury 1942, p. 178).

Reasons for population decline

The waning of H. viridis populations in Britain and Ireland that is apparent from the New Atlas hectad map may partly be attributable to records of introductions which failed to persist. Additionally, some long-established populations are known to have been lost as a result of changes in agricultural and land management practices taking place during the last 50 years, including the clearance of hedges and copses, and the gradual cessation of woodland coppicing (R.A. Fitzgerald, in: Preston et al. 2002).

Toxins

The principal toxin hellebore plants contain is protoanemonin, as found in most other members of the family Ranunculaceae. The content of this irritant varies widely with the species. The scale of the variation is demonstrated by the fact that measurements showed H. foetidus (Stinking Hellebore) contained 672 µg/g, while H. viridis had just 28 µg/g of the toxin present (Cooper & Johnson 1998). Numerous other references (including Clapham et al. 1962) mention the presence of two further toxic glycosides named helleborin and helleborein, and the fact that the plant has a burning taste.

Use in herbal medicine and related plant names

Decoctions of hellebores (both species covered here) were used in former years as purgatives, local anaesthetics, abortifacients, or to clear parasitic infestations of the skin or animal coat (Cooper & Johnson 1998). The association of garlic with veterinary application of Green Hellebore root in cattle follows an old belief, possibly dating from the ancient Classical Period. This suggested that since hellebore was such a powerful herb, a certain amount of prayer and ritual should be observed when lifting its rootstocks. "The person digging them up had either to chew on, or shortly before have eaten, several cloves of garlic, simply to ward off the poisonous effluvia of its roots." (Le Strange 1977, p. 136).

In the Revised Typescript Flora of Fermanagh, Meikle records being told by a local hotelier when he was on an outing in the Lough Melvin area during 1949 that, "the plant is (or was) used for a disorder of cattle, being pounded with butter and garlic and rubbed into an incision in the animal's tail, 'until you could smell the garlic on its breath.'" We are not told what the cow was suffering from, but other sources indicate it was used either as a purgative for worms, or to clear the skin and coat of lice (Grieve 1931; Le Strange 1977).

While the hotelier in Fermanagh described the application of the remedy to the tail of the animal, other accounts tell of an incision being made in the cow's dewlap (the loose fold of skin hanging under the animal's throat), and the Green Hellebore rootstock, or a preparation made from it, inserted into the wound. One of the less well known English common names of the species is 'Setterwort', which is derived from the term used to describe the aforementioned process, which is 'settering' or 'pegging' the dewlap. Prior (1879, p. 213) who details this plant name and the medicinal term, mentions that 'setter' is a corruption of 'seton', derived from the Italian 'setone', meaning a large thread of silk. Possibly the thread was used to sew up the wound in the dewlap. An alternative name for Green Hellebore is 'Pegroots', from the operation of 'pegging the dewlap'. Prior (1879) also indicates that settering was used to treat lung problems in cattle, such as coughs or wheezes (Prior 1879; Grigson 1987).

Cases of cattle poisoning have occurred as a result of this process, as one could easily imagine, the symptoms of which included prostration, loss of appetite, swelling of the neck, loss of condition of the coat, muscular tremors and difficult breathing (Cooper & Johnson 1998). Clearly the plant is not one to be handled more than necessary!

European occurrence

H. viridis has a strictly western discontinuous, native distribution on continental Europe, centred on France and stretching south to N Spain and N Italy, northwards to C Germany and east to Poland. The species has established alien status both within this range and to the north of it (Jalas & Suominen 1989, Map 1524). In plant geography, its distribution is summarised as suboceanic temperate (Preston & Hill 1997). The plant was introduced to New England for its somewhat dubious medicinal properties and it has become naturalised in N America (Grigson 1987).

Names

Some authorities derive the genus name 'Helleborus' from the Greek 'helein' or 'elein', meaning 'to kill' or 'to injure', and 'bora', meaning 'food', indicating the poisonous properties of the plants (Melderis & Bangerter 1955; Cricheley Plowden 1972). Other writers suggest the name 'helleboros' was the ancient classical Greek name for H. orientale (Gilbert-Carter 1964; Stearn 1992). The Latin specific epithet 'viridis' simply translates as 'green'.

English common names additional to those already mentioned above include 'Bear's-foot' and 'Boar's-foot' − a bear is called a 'boar' in Scotland, especially in northern Scotland, according to a source quoted by Britten & Holland (1886, p. 55); the allusion is to the digitately lobed leaf of the plant (Prior 1879). Another interesting name is 'Fellon-grass' which was applied to several quite different plants of which Hellebores were just one. In Westmorland, the name was applied to H. viridis. A 'fellon' was a boil or swelling, most commonly encountered in children. Housewives grew the plant to treat these childhood skin problems and partly also to treat against worms. It was a dangerous treatment, however, and it sometimes killed both the worms and the patient (Grigson 1987).

Threats in Fermanagh

The old, well-established Rossclare and Monea Castle sites could be threatened by building development.

Irish status

When it is found in Ireland, which is only rarely the case, A. napellus is always regarded as a naturalised garden escape. The plant generally occurs on roadsides and open areas on field margins.

Fermanagh occurrence

A. napellus s.l. has been recorded in Fermanagh on a total of eight occasions and at only four scattered sites in recent decades. A possible reason for the near disappearance in the 'wild' of this conspicuous, tall, beautiful, blue-flowered garden escape could be its well-known extremely poisonous nature, which may have led to its eradication by landowners when found in order to protect grazing stock.

The details of the other seven Fermanagh records are: Galloon Td, Upper Lough Erne, 1951, MCM & D; Arney village, 1952, MCM & D; Clonelly, NW of Kesh, 25 July 1976, Miss N. Dawson; Colebrooke Church, 1 July 1997 & 24 June 2003, RHN; roadside at Killadeas, near hotel entrance, 24 May 2002, RHN; and roadside at Cornamucklagh Td, NE of Brookeborough, 29 May 2004, RHN.

Growth form and reproduction

Monk's-hood has a blackish tuberous taproot or rootstock as its perennating (ie overwintering) organ, from which arises a usually unbranched flowering stem up to 1.5 m tall, but generally less. The plant has no means of vegetative reproduction and relies entirely on seed for its increase and dispersal. The cowl-hooded or helmet-shaped deep reddish-violet or purplish-blue flowers are produced in a long terminal raceme that may consist of around 30 flowers if the stem is unbranched and up to 100 when branched. The irregular flower conceals two long nectar-secreting spurs inside the hood, which are interpreted either as petals or as staminodes (modified sterile stamens) (Blamey & Grey-Wilson 1989). Each flower contains 3-5 carpels which, after the stigma has been pollinated by long-tongued bumblebees, ripen to form upright, many-seeded follicles.

The seeds are rather large (6 × 3 mm) and they bear three wings, one of which is slightly wider than the others (Butcher 1961; Clapham et al. 1962). The 'wings' are not very large, but they undoubtedly assist the seed to travel slightly further through the air when shaken out of the censer-like fruit. As with Aquilegia vulgaris (Columbine), one would not expect the plant to have great powers of dispersal, yet its occurrence in the wild indicates it is well able to escape from gardens on a regular basis. Apart from this, the reproductive ecology of A. napellus appears a completely closed book. I have not located any information on seed dormancy, longevity or germination.

A supposed native British variant

A form of the plant referred to as A. napellus subsp. napellus (or subsp. anglicum) has traditionally been considered indigenous in S Wales and parts of SW England (Watson 1883; Druce 1932; New Atlas). The semi-native habitats it frequents are characterised by calcareous to slightly acidic soils along stream banks that are often shaded, in damp, open woods or meadows (R.A. Fitzgerald, in: Preston et al. 2002). This supposedly native British population is mapped by Jalas & Suominen (1989, Map 1576), showing in addition to its British distribution a solitary record of a plant of similar form in the Pyrenees.

Since A. napellus has been grown in gardens for centuries and first made a 'wild' appearance in the British flora as late as 1821 (R.A. Fitzgerald, in: Preston et al. 2002), I find it quite amazing that anyone today would simply assume and assert that a species or subspecies is native (and/or endemic) to a region without first making a rigorous objective examination of all the circumstantial evidence that might be assembled to support such a status (Webb 1985; Forbes 2000). The 'endemic native' plant is reputedly found flowering in early summer in shady riverside sites in SW England and S Wales (A.J. Silverside, in: Rich & Jermy 1998).

Aconitum napellus subsp. napellus is distinguished from many of the more widespread garden forms of the plant (which are sometimes of hybrid origin), by having less deeply cut leaves, but with more finely pointed ultimate leaf segments. It also has a slightly earlier flowering period and the helmet of the flower is hemispherical, not elongated (A.J. Silverside, in: Rich & Jermy 1998).

Apart from its taxonomy and alkaloid content, the plant appears little studied and I cannot locate any recent references on its biology or ecology, let alone on its status, apart from the brief treatment by John Akeroyd in Scarce plants in Britain (Stewart et al. 1994).

Irish occurrence

A. napellus s.l. is slightly less rare than elsewhere in Ireland in the six county province of Northern Ireland, where it has records in five counties (the exception is Co Armagh (H37). In the Republic of Ireland, the species has only five widely scattered records (Preston et al. 2002).

British occurrence

In Britain, A. napellus s.l. is widespread throughout as a garden escape, but it has a greater presence in western districts of England and Wales, while north of the Scottish border it becomes somewhat more frequent in eastern areas. It is very possible that some of the mapped plants really are the garden hybrid A. × cammarum L., a cross between A. napellus and A. variegatum (R.A. Fitzgerald, in: Preston et al. 2002).

European occurrence

A. napellus s.l. is a very variable species endemic to W and C Europe, its distribution on the continental mainland extending south to C Spain and stretching eastwards to the Carpathian mountains. It is absent, however, from most of the Mediterranean basin (Jalas & Suominen 1989, Map 1575). The taxonomy of the species (or polymorphic aggregate of forms) is greatly confused by the recognition by some of a range of subspecies (which others elevate to species rank), plus a history of very many name changes. As is the case with Aquilegia vulgaris (Columbine), which sometimes occurs in similar shady and damp habitats to A. napellus, the comparative scarcity of the species and its insect pollination syndrome would hardly lead us to expect the appearance of either its pollen or its seed in the fossil record, and indeed none exists (Godwin 1975).

Toxins

The plant contains a cocktail of at least four alkaloids including aconitine, which even on its own is highly toxic, so that Monk's-hood has the reputation of being the most poisonous plant in the British Isles (Cooper & Johnson 1998, p. 188). Perhaps because its poisonous nature is so well known and hence its subsequent removal from sites where grazing animals might find it, there are very few reports in recent years of animal or human poisoning by the species in these islands.

Names

The genus name 'Aconitum' is Latin and is thought by some to be derived from the Greek name 'Akoniton' (although Gilbert-Carter (1964), for instance, regards the etymology as doubtful). 'Aconitum' is a classical name first given to an unknown poisonous plant by Theophrastus (Gledhill 1985; Stearn 1992) and later reused by the Swedish botanist, Linnaeus, for the current genus. The Latin specific epithet, 'napellus', is a diminutive of 'napus', which means 'little turnip', an obvious allusion to the tuber of the plant (Stearn 1992).

The English common name 'Monk's-hood' was first given by Lyte (1578), who described "The flowers be as little hoodes", translating the name directly from Dutch and German (Grigson 1974). In his excellent The Englishman's Flora, which deserves to be on every Celt's bookshelves too, Grigson (1955, reprinted 1987) remarks that the local names of the garden form of this very poisonous plant are all charmingly innocent. Most of them relate to the odd form of the flowers, "and especially to the fluttering, dove-like nectaries". They include 'Doves in the Ark', 'Lady Lavinia's Dove Carriage'. Many alternative names refer to bonnets, caps, helmets or hoods, for instance, 'Old Woman's Nightcap' and 'Face in Hood' (Britten & Holland 1886).

Fermanagh threats

None, as it is much too rare to be a threat to natural vegetation, or to grazing stock. With its recognised status as a rare, neophyte, garden escape, we are not concerned about threats of any kind to the survival of Aconitum napellus.

Growth form and preferred habitats

One of the prettiest if not the earliest of the welcome harbingers of spring in its typical habitat of deciduous woods and hedges, A. nemorosa is a perennial geophyte with a shallow, brittle, slender, brown rhizome which branches and forms clonal patches up to 5 m in diameter (Shirreffs 1985). As with other vernal species, A. nemorosa is really a shade-avoiding rather than a shade-tolerant species since it exploits the light phase in woods, scrub and hedges before the leaf canopy develops. While it tolerates a wide variety of soils, as a rule A. nemorosa is best developed on soils that are moist to wet in spring. Such soils are often of relatively heavy texture or rich in humus, ie between 7% and 20% organic matter (Grime et al. 1988), or even more than this in Fermanagh.

The very superficial root system of A. nemorosa runs at a depth of just 5-10 cm, a feature which exposes the plant to early season drought but which allows it to survive in much wetter soils than its common competitor, Hyacinthoides non-scripta (Bluebell), the bulbs of which are drawn down year-by-year by contractile roots, often reaching depths of 15-25 cm where they may easily suffer waterlogging (Grabham & Packham 1983). While this is the case, A. nemorosa is not really a wetland species, it can merely tolerate moist soils on the fringes of marsh, fen and bog, although in some of these situations it may have to endure periods of submergence.

While Wood Anemone appears to prefer less fertile limestone woodland soils, it is not in any way confined to them, but rather it may be expected in any shady situation including low growing ericaceous heath. It does not, however, penetrate peat bogs where the soil pH falls below 3.5, which appears to totally exclude the species (Shirreffs 1985; Grime et al. 1988).

In hedges, limestone pavement and long-established grassland, the presence of A. nemorosa can frequently be associated with previous scrub or woodland cover, so that the plants are seen as remnants of previous vegetation. When found on cliffs, however, or in sheltered spots above the tree-line, there can be no question of relict status, and one wonders exactly how the species was transported to these elevated sites. Then the plant typically occurs in rock clefts, or on sheltered, often north-facing ledges, or shaded by overhanging sub-shrubs or trailing vegetation, all conditions providing the shelter, shade and high humidity the species requires.

Fermanagh occurrence

Locally, the species is very common and widespread, especially in lowland Fermanagh, having being found in 290 tetrads, 54.9% of those in the VC. The most typical habitats it occupies are deciduous woods, hedgerows and river banks but, in the prevailing wet conditions of Fermanagh, A. nemorosa is capable of extending into grassland and other open habitats, including more rarely, cliff ledges and scree on the talus slopes beneath cliffs.

British and Irish occurrence

Wood Anemone is common and widespread throughout Britain and Ireland, although absent from Orkney and Shetland and rare in areas like the English Fens and other exposed situations where woodland (and indeed anything approaching dry land) are sometimes scarce. The previous dearth of records from the Republic of Ireland, which was regarded as probably under-recording (Shirreffs 1985), was remedied to a considerable extent in the New Atlas survey. The distribution remains fairly patchy in the Republic, except in parts of the far south and in the Dublin and Wicklow area, where there are more resident plant recorders (New Atlas). However, one must never overlook or underestimate the likely limiting ecological factor(s), and local excesses of soils, exposure and wetness must certainly also restrict distribution.

Phenology and growth

The shoot emerges from below ground in March, pushing up through the leaf litter crozier-like, with three folded leaf-like bracts surrounding and protecting the solitary flower bud. The shoot soon straightens, the palmately cut bracts unfurl and expand, and the flower stalk elongates carrying the flower well above the ring of three involucral bracts. The flower bud then loses its green tinge and, since the perianth in this species consists of just one set of leaf-like segments, the tepals or petal-like sepals expand (plant anatomists tell us they are not true petals), and the first anemone flower opens 'for business'. This usually occurs around the end of March or the beginning of April depending on season, habitat and geographical location (Shirreffs 1985). The true leaves are very similar in appearance to the bracts, but they are produced a short distance further along the rhizome from the flower-stalk and they do not appear until after the flowers have opened (Step & Blakelock 1963).

Growth rates in A. nemorosa are extremely low: seedlings take at least five years to form a viable rhizome, plus perhaps another five to ten years for the plant to become capable of flowering. The average annual extension in adult plant rhizomes is only 2.5 cm, so vegetative spread is also extremely slow (Ernst 1983; Shirreffs 1985). By June, the sexual reproductive cycle has been completed, and in the shaded floor vegetation the plant rapidly dies down and disappears completely below ground by around the middle to end of July.

Flower variation and pollination

All parts of the anemone plant are very variable (ie phenotypically plastic), and this is particularly true of the flower. Flower stems vary in height between 10-30 cm above ground (with a mean of 14 cm), each bearing a solitary blossom 10-40 mm in diameter. The perianth is composed of from 4-11 elliptic sepals (most frequently six or seven). The sepals are usually white inside, purple tinged or streaked outside, but purple, blue and pink forms also occur, the former sometimes quite frequently (Shirreffs 1985). Stamens are numerous, usually about 45 in three ranks of differing filament length, and as Wood Anemone flowers offer no nectar, unspecialised insects visit them to collect the openly presented protein-rich pollen as food for themselves and their brood. The insects attracted by the perianth's appearance, movement and food reward range from honey-, bumble- and solitary-bees, to beetles, flies, thrips and bugs (Proctor & Yeo 1973).

When the flower first opens the maturing stamens are crowded over the stigmas and prevent them from being pollinated, though pollen is already being shed from the ripe, outer anthers at this stage. After about a week, the rest of stamens ripen and diverge, and during the second week the white, translucent stigmas can be pollinated. Self-fertilisation is prevented by an incompatibility mechanism, making cross-pollination obligatory (Proctor & Yeo 1973). The flowers are held erect during the day and move in the slightest breeze, but they droop and fold at night, or in dull or wet weather. The stigmas shrivel and blacken after pollination and a crowded head of single-seeded achene fruits then develops.

Reproductive biology

The number of carpels in the flower varies from 9-42, with a mean of 22 (Salisbury 1942). Shirreffs (1985) found the mean number of carpels ranged from 16 to 31 at different sites, the lower figures being associated with non-woodland sites, such as open grassland. Flowering density is greatest in woodlands since this is where the species forms dominant carpets. The mean number of fruiting flowers in a typical A. nemorosa carpet is around 152/m² (Salisbury 1942), with higher figures in coppice (380 flowers/m²), but much lower figures than this are found in densely shaded areas.

The cluster of achenes breaks up and the individual fruits are shed from May to June, depending on the local climatic and micro-climatic conditions. When shed, the achenes of A. nemorosa contain an immature embryo that requires a moist, cold, after-ripening period lasting from 4 to 6 months before they ripen sufficiently and become capable of germination (Vegis 1961). Germination occurs in the following spring, but the typical rates that occur in the wild are poor. Ernst (1983), however, found somewhat improved figures of between 5% and 35% germination occurred after long, cold winters in Germany, with lesser figures being obtained after mild winters. This indicates that seedling recruitment into the existing mature population is irregular. Ernst concluded that recruitment of seedlings and young plants (ie 2nd to 5th year classes) in the study area was inadequate to maintain a viable long-term population. The mortality of the young plants was very high in the first and second years (88.2% ± 13.5%), which is comparable to that shown by other, bulbous vernal species, namely Allium ursinum (Ramsons) (Ernst 1979) and Narcissus pseudonarcissus (Daffodil) (Barkham 1980).

In his study, Ernst (1983) calculated that A. nemorosa generally does not invest more than 5% of its biomass resources into sexual reproduction, while most of the fixed energy (production) is used to maintain the rhizome at 40 to 50 % of the total plant biomass. The annual increase of the rhizome can be as much as 150 mg. Rhizomes older than 15 to 25 years are brittle and readily separate from the parent plant. This fragmentation constitutes a rather unspecialised form of vegetative reproduction and, in the sites studied, it appears to be the main mechanism maintaining the population.

Population persistence

Wood Anemone showed a remarkable degree of persistence in neglected coppice uncut for 30 to 40 years in E England, being present in 70% of plots in five such woods. This figure was surpassed only by Rubus fruticosa (Bramble) which had a 100% occurrence (Brown & Oosterhuis 1981). It is interesting that during the same study, germination tests carried on for two years with soil samples taken from the upper 15 cm of the profile (after litter removal), found no seedlings of A. nemorosa present, nor indeed any seedlings of Hyacinthoides non-scripta (Bluebell) or Mercurialis perennis (Dog's Mercury), although all three of these often dominant carpet-forming woodland species had survived in at least 50% of the neglected, overgrown woodland coppice investigated. This agrees with the general finding that species of shaded habitats tend to lack mechanisms for widespread and rapid seed dispersal (Webb 1966; Brown & Oosterhuis 1981): the seeds of such species are heavy, seed production is relatively low (Salisbury 1942) and they are in the main dispersed by rainwash, in clinging mud, or by ants (Ridley 1930).

Possible dispersal mechanisms

Anemone nemorosa has no specialised means of achene dispersal, although it has been suggested that ants may be involved (Ridley 1930; Oberdorfer 1970). The achene has no attached food body, so that ants and other animals are unlikely to show any interest in them. However, if ant dispersal (ie myrmechory) really is the sole, or even the principal method of seed dispersal, then the efficiency and efficacy of the process must be rather severely limited since Ernst (1983) found the distance between parent plants and established seedlings was never more than 13 cm. Brown & Oosterhuis (1981) observed that even a relatively narrow strip of non-woodland habitat around 50-100 m wide, would create an ecological hurdle that most plant species of shaded habitats could hardly ever cross. The apparent lack of an effective seed dispersal mechanism in A. nemorosa, together with the exceedingly slow diffusive spread of the rather long-lived, creeping rhizome, results in the observed single species carpet of intermingling clones of the plant which we see mainly in the relative stability of woodland or undisturbed scrub vegetation.

Competing woodland species

The carpet growth of A. nemorosa enables it to shade out many smaller competing species, allowing it to become dominant at sites which are wet in the spring. However, it is usually unable to compete with taller growing vernal species, including for instance Hyacinthoides non-scripta (Bluebell), Allium ursinum (Ramsons) and, in Britain, but not in Ireland, Mercurialis perennis (Dog's Mercury). In these instances, A. nemorosa is either completely ousted by the shade of the taller plants and their competition for other limited environmental resources, or it survives in smaller numbers only as a subsidiary companion species (Shirreffs 1985). However, where there is a considerable degree of woodland disturbance (eg grazing, trampling or coppicing), A. nemorosa is often better able to withstand these external pressures than can Bluebells, Ramsons or Dog's Mercury, especially where such disturbance is combined with seasonally wet soils (Grime et al. 1988).

Toxins

A. nemorosa has some degree of protection from grazing animals since it contains the volatile, oily, irritant substance, protoanemonin, the concentration of which reaches its peak when the plant is flowering and most conspicuous. The toxin has an acrid taste and causes burning in the mouth and throat, effectively deterring animals from eating much of it (Cooper & Johnston 1998). This said, a study in Warwickshire woods made 50 years ago, when rabbit populations were very much more active than now, found a number of widespread woodland herbs, including A. nemorosa, suffered heavy grazing pressure near warrens, sometimes almost to vanishing point (Knight 1964). Several fungi, both Ascomycete and Basidiomycete, attack A. nemorosa leaves and rhizomes and can suppress flowering partially or completely (Ernst 1983; Shirreffs 1985).

Fossil record

Fossil pollen of Anemone-type has been found in Scotland from the late-Glacial period (13,000-10,000 BP), but it is not specifically that of Anemone nemorosa, but could also come from the related species Actaea spicata (Baneberry) or Pulsatilla vulgaris (Pasqueflower) (Shirreffs 1985).

World occurrence

Beyond the British Isles, A. nemorosa occurs throughout the suboceanic northern temperate zone of both Europe and W Asia and reaches 67° N, just within the Arctic Circle in Norway (Shirreffs 1985; Hultén & Fries 1986, Map 827; Jalas & Suominen 1989, Map 1630; Jonsell et al. 2001). Forms of A. nemorosa are widely grown in gardens within and beyond the natural range of the species. Griffiths (1994) lists 19 garden cultivars of the species and Jonsell et al. (2001) mention an additional yellow form cultivated in Sweden.

Herbal use

A. nemorosa has entirely fallen out of use in herbal medicine today, although the older herbalists such as Gerard and Culpepper listed numerous ailments it was supposed to alleviate, eg headache, rheumatic gout, lethargy and for cleansing ulcers (Grieve 1931). As the plant is decidedly poisonous, the modern advice is to BEWARE of any such remedies.

Names

The name 'Anemone' is often said to be derived from Greek 'anemos', wind, plus the feminine patronymic suffix, making it 'daughter of the wind' (Gilbert-Carter 1964). The connection with the wind is somewhat obscure in this particular species, however, although the flower does dangle and flutter in the breeze if it is strong enough, so perhaps the alternative explanation may fit better. This suggests the name is a corrupted Greek loan word of Semitic origin, referring to the lament for slain Adonis, or Naaman, whose shed blood produced the blood-red flowers of Anemone coronaria (Crown Anemone), or Adonis annua (Pheasant's Eye), both common spring species in the Mediterranean region (Gilbert-Carter 1964; Stearn 1992). The Latin specific epithet 'nemorosa' means, 'growing in woods' or 'in shady groves' (Gledhill 1985).

There are dozens of English common names listed in Grigson (1987), the two most frequently used being 'Nemony' or 'Neminies', a simple contraction of Anemones, and 'Wind-flower'. Both of these names are borrowed from Anemone coronaria, famous in Greek legend as mentioned above. Other names include 'Wood Crowfoot', 'Moonflower', 'Cowslip' (the latter rather odd), and two names that refer to an odour, 'Smell Foxes' and 'Smell Smock', both of which might be derived from the sharp, unpleasant taste and the faint smell of A. nemorosa (Grigson 1987).

Threats

Anemone nemerosa regenerates mainly by rhizome growth and while large clones do develop and genets can persist for a long number of years, seed persistence is low, dispersal is very poor and seedling establishment is extremely slow. Thus the plant is a poor colonist of new sites and Grime et al. (1988) believe it is decreasing in England in grassland habitats at least, and perhaps also in some woodlands. Certainly it is not equipped for jump-dispersal and the colonisation of new habitats, so it may well be a relict in many of its existing sites.

Growth form and preferred habitats

This vigorous deciduous perennial with its climbing, scrambling and trailing stems belongs to the only genus in the family Ranunculaceae that contains woody members. Typically it clambers over shrubs and trees, or clings on walls or rocks, holding onto its support by the twining stalks (petioles) of its compound, once-pinnate, opposite leaves that act like tendrils. The grip of the petiole-tendrils is very tight, and as they age they harden and become wire-like, so that C. vitalba sometimes strangles and kills the stems of the plants supporting it (Melderis & Bangerter 1955; Step & Blakelock 1963a).

Being deciduous, the leaves of C. vitalba drop off in the late autumn, although their twisted tendril-like petioles persist for a while after the leaf blades have disappeared. During the winter, the plant relies for its entire support on the entanglement of its woody stems with those of the tree or shrub on which it is climbing and despite winter storms this always seems to suffice (Fitter 1987).

Soil preferences

The species performs best on base-rich or calcareous soils, of which at least in certain parts of Britain and Ireland, it is a useful and reliable indicator species (Lousley 1969, p. 15). In central Europe, however, Ellenberg (1974) found that C. vitalba grows on a wide range of soils from weakly acid to weakly basic. However, to really thrive it requires a soil with moderate to high fertility and medium to good drainage. In another English study, low calcium levels in soil appeared to retard the growth of the species (Buxton 1985).

A detailed experimental investigation of the species soil nutritional requirements in New Zealand found that growth of C. vitalba increased with increasing levels of lime; this was especially so when this was accompanied with increasing rates of applied phosphate. Maximum growth occurred at pH 4.7, while plants were killed by a pH as low as 3.7, presumably due to the toxicity of available aluminium at this acidity (Hume et al. 1995). This study also showed that the plant's response was to high pH and/or low aluminium concentrations, rather than to high concentrations of calcium, indicating that at least in New Zealand, C. vitalba is not a true calcicole species (ie lime-loving) (Hume et al. 1995).

While showing a definite preference for calcareous or base-rich soils, C. vitalba is not completely confined to them, but may also occur on other dry, stony sites and on disturbed, enriched soils, including on waste ground and in rock quarries (Sinker et al. 1985).

Origin of most Clematis plants

The garden source of the C. vitalba plants found in the wild in Britain and Ireland is often not immediately obvious. The genus Clematis is a very popular, indeed at present a very fashionable horticultural subject, with many very beautiful species and legions of novel varieties widely available in the nursery trade. However, C. vitalba itself is much too rampant and weedy a plant to be grown by many gardeners on its own account. The answer to this apparent garden usage, yet lack of decorative worthiness, has to do with the horticultural production of rapidly flowering new Clematis varieties and especially of sterile hybrids such as the familiar C. × jackmanii. Hybrids and other varieties can be multiplied by internode cuttings, but a flowering plant of the desired hybrid or variety is more quickly and more certainly obtained by grafting the variety on to a seedling rootstock of C. vitalba, or alternatively on another cultivated form, C. viticella (Purple Clematis).

If the grafted individual is planted with the region of the tissue union below the soil surface, after a few years the grafted variety will have developed its own root system (Salisbury 1935, p. 151). On the other hand, if an inexperienced gardener plants the graft with the union exposed, there is nothing to prevent the C. vitalba rootstock developing its own competing stems. In addition, if the grafted plant is incorrectly pruned the scion may be damaged, killed or entirely removed, again allowing the C. vitalba rootstock to take over. Such plants invariably prove too dominant or rampant in the garden setting and they eventually end up being discarded on refuse tips, manure heaps, or on waste ground, where they may survive and continue to grow, reproduce and release their wind-borne seeds into surrounding wild habitats.

Fermanagh occurrence

C. vitalba is either deliberately planted, or much more probably, a naturalised wind-dispersed garden escape in Fermanagh, growing in just a few old, rather neglected hedgerows and thickets. The first record of this climber in Fermanagh was made as late as 1951, incidentally providing a very good example of the lack of previous recording of alien and introduced species in the county. It was discovered in what was then a very typical habitat of the species − by a railway crossing. C. vitalba is very much associated with railways throughout the British Isles (Hackney et al. 1992). Previously, when the Fermanagh railway was in operation, C. vitalba grew alongside the permanent way, and having plumed achenes for fruit, it is very easy to imagine it readily spreading along open areas on embankments, cuttings and road crossings. After the closure of the railway in 1957, although it could equally disperse along roadsides, the species had fewer opportunities for colonisation and it has almost certainly dwindled to arrive at the current level of rarity. We cannot know this for certain, the plant having only ever been recorded in seven Fermanagh tetrads, six of them with post-1975 records. Occasionally, however, as at Gubbaroe on the limestone shore of Lower Lough Erne, when it grows in full sun, colonies of C. vitalba become so vigorous the plant can cover and dominate very large sections of hedgerow and scrub thicket. In such situations, it may totally obscure and overwhelm young trees and shrubs and even threaten older, established plants.

The other Fermanagh record details are: Carrickreagh Bay, Lower Lough Erne, 1983, RHN; Scottsborough lakelet, 28 August 1988, RHN & RSF; Cloughmore, 2 km SE of Rosslea, 28 August 1988, RHN & RSF; ride in conifer plantation, Gubbaroe Point, Lower Lough Erne, 1 January 1990, RHN; Cleenishgarve Island, Lower Lough Erne, 17 June 1990, RHN; Killadeas Td, Lower Lough Erne, July 1993, I. McNeill; near old house, E of Gubbaroe Point, Lower Lough Erne, 18 April 1998, RHN; near old house, Mullaghfad Td, E of

Brookeborough, 21 September 1998, I. McNeill; Gublusk Bay, Lower Lough Erne, 8 August 1999, RHN; same site, 12 October 2002, I. McNeill; fence on Irvinestown Road, Enniskillen, 19 September 2010, RHN & HJN; Carrickreagh Bay, Lower Lough Erne, 29 September 2010, RHN.

Flowering reproduction

C. vitalba flowers rather late in the year from July to September and the branched clusters of creamy or greenish-white, star-like 2 cm diameter flowers, like those of Caltha and Anemone, are petal-less. The usually four, but occasionally five or six sepals, which are hairy on their outer surface, again take on the role of the missing flower whorl (Webb et al. 1996). The Clematis flower has a faint but pleasant vanilla fragrance (Genders 1971), but it does not produce any nectar for pollinating visitors. The flowers are visited by flies and by bees that collect the abundant protein-rich pollen as a food reward and carry out cross-pollination. Wind-pollination may also be involved, but the extent of this in unknown.

Fruiting and dispersal

The characteristic fruits are produced in the autumn in large numbers, and usually sufficient silvery achenes are retained on the receptacle to keep the plant conspicuous in the hedgerow throughout the winter months. It has been estimated that around 17,000 seeds (achenes) are produced for every 0.5 m² of C. vitalba canopy and they are dispersed by wind, water, people and other vertebrates (Cronk & Fuller 2001, p. 70). In linear habitats, eg along roadsides and beside railway lines, the dispersal of the many achenes of C. vitalba with their firmly attached white or silvery grey, long-plumed, feathery styles is obviously enormously facilitated by the sucking linear slipstream of swiftly moving traffic.

Germination and seed survival

Study of C. vitalba germination ecology in New Zealand has shown that achenes (ie single seeded dry fruits), retained on the vine over the winter, have a high degree of dormancy and viability, and the sporadic release of the seed from the parent plant effectively acts as a form of aerial seed bank (Bungard et al. 1997). The published survey of soil seed banks in NW Europe contains very little information on C. vitalba, but one report did suggest that a short-term persistent buried soil seed bank may exist (ie seed surviving between one and five years) (Thompson et al. 1997). In addition to a minimum light requirement (equivalent to 5% of full sun), the achenes require a period of chilling to break dormancy, conditions that effectively time germination to the spring following their production.

Behaviour as a serious weed in New Zealand

Although over most of its European range C. vitalba is usually an innocuous climber, the species can become a serious weed in young forestry plantations causing losses due to overgrowth of saplings. In New Zealand, it is a particularly vigorous and harmful invasive alien in forestry plantations and also in conserved remnants of native podocarp forest and it is responsible for losses of both forest structure and indigenous species biodiversity (Ogle et al. 2000; Hill et al. 2001). C. vitalba was first recorded as a weed in New Zealand in 1940 (Webb et al. 1988), although it was known much earlier in gardens and as a local garden plant escapee. It is thought to have arrived as a garden plant from Europe, and the first herbarium specimen of a wild plant of the species was collected in 1936 (West 1992). C. vitalba now occurs as an adventive species almost throughout the lowlands of New Zealand, except for regions north of latitude 37˚S (Webb et al. 1988; West 1992). It is probably the most publicised environmental weed in New Zealand, and community groups, government departments, local authorities, schools and paid contractors have tackled infestations over large and small areas, either mechanically or chemically (Timmins 1995). It has come to public notice mostly because it invades and smothers indigenous forest. In 1998 it was the subject of 37% of the complaints about plant pests made to the Regional Council which oversees the indigenous forest area around Taihape in central North Island, New Zealand, which was more than any other species, including agricultural weeds (Rowatt 1998). The seriousness of its weed status is illustrated by the fact that research is underway to identify an insect suitable for biological control, and the introduction of the European leaf-miner Phytomyza vitalbae Kaltenbach which attacks the species, is being actively considered (Hill et al. 2001).

In the warm temperate and moist to wet conditions prevailing in New Zealand, C. vitalba can regrow from fragments after cutting and this is recognised as important in its invasive spread in this part of the world. Regeneration of fragments is related to age, since older stem sections have better water retention and larger nutrient resources available than softer, young tissues (Kennedy 1984). Under the prevailing growing conditions the species has a high growth rate, with young plants and new shoots extending up to 2 m per year. If given full sun, plants also reach reproductive maturity early in life, producing seed when one to three years old and reproducing vegetatively after just one year's growth (Cronk & Fuller 2001, p. 71).

Irish occurrence

C. vitalba is an established alien in Ireland, where although widely scattered throughout, it is very much more frequent south of a line between Limerick and Dundalk.

Fossil record and species status in Britain

Despite an extremely skimpy fossil record in Britain from the Atlantic Period onwards (7,500-5,200 BP)(Godwin 1975, p. 119; Rackham 1980, p. 108), C. vitalba is at least traditionally regarded as native in most of the area of England and Wales south of a line between Anglesey and The Wash. North of this juncture it is considered an alien introduction and its presence diminishes rapidly towards S Scotland (Preston et al. 2002). It is difficult to envisage exactly what criteria determine the distinction between native and alien in these circumstances (Webb 1985). It is sensible to take a cautious approach when attempting to interpret the published map (Preston et al. 2002) and, indeed, for any species great care is needed when distinguishing native from alien status for regions on the same land mass.

European and world occurrence

Beyond Britain and Ireland, C. vitalba is considered native in S, W and C Europe and has alien status in a narrow zone north of its native range in Holland, Denmark and Germany (Jalas & Suominen 1989, Map 1679). It is also regarded as native in N Africa, W Turkey, the Lebannon, the Caucasus, N Iran and Afghanistan (Griffiths 1994; Jonsell et al. 2001). As already mentioned, it is a very invasive naturalised introduction in New Zealand and is also naturalised in parts of both S Australia and N America (Cronk & Fuller 2001).

Knowing the extent of the major weed problem this alien climber has created in New Zealand over the past 50 years and the threat it poses to the survival of remnants of indigenous vegetation and rare species on those islands, it is rather surprising that at least one horticultural supplier is still offering C. vitalba for sale to gardeners in N America, although it does warn of its vigour and potential spread (http://www.botanical-journeys-plant-guides.com/clematis-vitalba.html, accessed 19 January 2016). As Cronk & Fuller (2001) point out, "It is important that the problems associated with it are made known so that future introductions to potentially 'invasible' [a horrid, invented word] areas are prevented."

Toxins

Like other members of the family, C. vitalba contains an appreciable amount of the poisonous substance, protoanemonin and, although animals rarely browse the plant because of the acrid taste and irritant effect on the mouth, it is known to have killed cattle. Contact with the plant sap can also blister the skin (Cooper & Johnson 1998).

Names and uses

'Clematis' is derived from the Greek 'klema', meaning 'a vine branch', alluding to the vine-like twining climbing habit of the plant. The Latin specific epithet 'vitalba' translates as 'the white vine' and refers to the wild species (Johnson & Smith 1946).

There are a large number of English common names in existence: Grigson (1955, 1987) lists 36. 'Travellers’ Joy' was a name coined by John Gerard for his English herbal of 1597, presumably because he knew that the plant grows on waysides and hedges. The most widespread English common name is 'Old Man's Beard', a reference to the silvery-white twist of long, feathery styles that adorn the fruit achenes. It should be remembered that very often in folklore the 'old man' referred to is the Devil. Similar allusions include 'Bushy Beard', 'Daddy's Whiskers', 'Grandfather's Whiskers' and 'Father Time'.

Other names refer to the rope-like nature of the older climbing stems, eg 'Bullbine', 'Hag-rope' and 'Devil's Guts'. The woody branches with their characteristic flaking stringy bark have been used to make lightweight baskets in the past (Hutchinson 1972). In Devon, Clematis stem was woven to make the bottoms of pots for catching crabs (Vickery 1995, p. 375).

Another group of names indicates the use of stems as a cheap tobacco substitute, eg 'Boy's Bacca', 'Gipsy's Bacca', 'Tom's Bacca', 'Smoking Cane' and 'Poor Man's Friend'. According to Grigson, the young and the poor used to smoke cigar lengths of the dry stem, "as they draw well and do not burst into flame". Personally, I cannot imagine the desperation and the taste! However, there is evidence that this type of cigarettes was also smoked elsewhere, as there are equivalent names for the plant, eg 'Rauchholz' ('smoke wood') in German, 'Smookhout' in Dutch, and 'Fumailles' and 'Bois à fumer' in French (Grigson 1955, 1987). If anyone is tempted to try this, beware; unless the stem is dead and completely dry the irritant toxin is present and can cause ulceration of the lips and mouth (Mabey 1977, p. 159).

There are also a number of other peculiar, unexplained ideas contained in further quite widespread alternative English names, for instance, 'Honesty', and some connection with the virtuous Virgin Mary, for example, 'Lady's Bower' and 'Maiden's Hair' (Grigson 1987). Prior (1879) reckoned the name, 'Virgin's Bower', which was also used by Gerard (1597), alluded not to the Virgin Mary, but rather to the virgin Queen Elizabeth. In names, as in other matters, truly it is often easier to pose a question than to answer it!

In herbal and homeopathic medicine, several European species of Clematis, being diuretic and diaphoretic, have been used in treating ulcerous diseases such as syphilis, gonorrhoea, cancer and other inflammatory conditions including those of the eye (Grieve 1931). According to this herbal source the roots and stems of C. vitalba bruised and boiled for a few minutes in water and then digested for a while in sweet oil, made a cure for itch. There is folklore indicating that a C. vitalba stem twisted into a ring and worn round the neck was used in herbal medicine to cure convulsions in children (Vickery 1995, p. 375).

Considering the poisonous nature of the species, an interesting usage of it as food has been recently published. In a remote and isolated valley in NW Tuscany called 'Garfagnana', traditional gathering and use of a wide range of herbaceous plants has survived, so that for example, the inhabitants make a vegetable broth containing at least 20, but often more than 40 wild plants (Pieroni 1999). The soup recipe does not contain C. vitalba, but it is the main ingredient in a very popular vegetable omelette. Young shoots are boiled before they are incorporated with eggs, and sometimes with cheese, and the mixture fried so that the protoanemonine is inactivated (Pieroni 1999).

Threats

There are no conservation problems locally, but rampant C. vitalba can smother other plants.

Growth form

This variable, wintergreen perennial is a widespread and abundant medium-tall herb (up to 50 cm), of moist to seasonally wet (but not waterlogged), pastures, mown meadows and roadside verge grasslands. It is also present in unmanaged grasslands, eg in woodland clearings and on all forms of rock outcrop. Although it possesses a small wintergreen leaf rosette the species produces very little growth until February or March and for its energy requirements the overwintering plant relies on starch stored in its short, stout rootstock (Harper 1957). Coles (1971) refers to the compact rootstock as 'premorse', as opposed to a longer, more spreading, rhizomatous one. The term is derived from the Latin 'praemorsus', meaning 'bitten off', which is a rather appropriate description of the rootstock (Holmes 1979).

Possible native habitats

As with Ranunculus repens (Creeping Buttercup), while fossil evidence proves R. acris is undoubtedly a native species in Britain and Ireland (see below), it is very definitely a 'follower of man'. Nowadays, more often than not it occupies habitats managed, opened up, or created – intentionally or not – by human activity (Harper 1957). It is not really possible to be certain what the natural habitats of R. acris were in these isles prior to human arrival, although Harper (1958) has suggested that it most likely frequented a variety of damp ground communities, ranging from marshes and Carex elata (Tufted-sedge) dominated fens, to mountain grasslands above the climatic forest limit.

Variation

R. acris is a polymorphic species that in Europe can be subdivided into four subspecies (see Flora Europaea 1, Tutin et al. 1993, p. 274), two of which occur in in the British Isles, the common subsp. acris and the much rarer northern subsp. pumilus (Wahlenb.) A. & D. Löve, that only appears in Scotland. R. acris can also be subdivided into three varieties, a larger, very widespread var. acris, a considerably smaller (up to 20 cm tall) var. pumilus Wahlenb. that is restricted to the Scottish Cairngorm mountains, and a hairier var. villosus (Drabble) S.M. Coles, which is common in undisturbed areas of N Scotland (including the isles) and W and C Ireland (Stace 1997, p. 88). Var. villosus was not recorded in Co Fermanagh and certainly is under-recorded across its range.

Soil moisture preferences

The soil moisture preferences of R. acris (although it is not strictly confined to them) are intermediate between those of two other, very common, closely related buttercups, R. bulbosus (Bulbous Buttercup) on drier forms of rocky ground or on shallow soils and R. repens in wetter hollows, or poorly drained soils. Certainly, R. acris is always absent from areas which suffer serious midsummer drought, but thanks to its geographical west Atlantic situation, prolonged drought very rarely, if ever, occurs in Fermanagh (Harper & Sagar 1953; Harper 1957). Meadow Buttercup is taller than these other two mentioned buttercups and thus is better able to compete and survive in the more closed, tightly knit turf of low to moderately fertile, fairly productive, herb-rich water meadows that are such a significant conservation feature of lakeland Fermanagh. These damp, waterside meadows that flood regularly or from time-to-time, harbour some rather aggressive competitive plant species, including the grasses Lolium perenne (Perennial Rye-grass) and Agrostis stolonifera (Creeping Bent) and two even more locally common and dominant herbs than Meadow Buttercup, Juncus effusus (Soft-rush) and Filipendula ulmaria (Meadowsweet).

Fermanagh occurrence

Throughout Fermanagh, R. acris is common everywhere except at the highest, most exposed ground, or in permanently wet, or very acidic habitats. Despite these substrate limitations it is still present in 487 tetrads, 92.2% of those in the VC, making it one of our most widespread species. In terms of tetrad numbers, it ranks fourth equal with Angelica sylvestris (Wild Angelica) and it is the eighth most frequently recorded species in the Fermanagh Flora Database.

Even at this high level of frequency of records, it is possible that our statistics might understate the presence of this species. One of the acknowledged errors in any botanical field survey is the likelihood that very common and familiar plants tend be overlooked, simply because they have become so unremarkable in the recorder's mind, and regularly it will be assumed that they have already been entered on another field recording card and consequently they are overlooked or ignored!

Land management and plant species competition

The farmland grasslands where Meadow Buttercup populations persist to the greatest extent are regularly grazed or mown. This form of ecological pressure is important in maintaining the herb's population size, since without regular disturbance opening up the vegetation and offering periodic temporary release from competitive stress, the species would survive only at much lower frequency. If the grazing or cutting regime becomes relaxed, R. acris cannot thrive and maintain its abundance in the longer term among more vigorous, gregarious and aggressive plant species.

Writing prior to the latest (and greatest) intensification of farming in the last half-century, Harper's (1957) research indicated that the abundance of R. acris in pastures and meadows provides an index of the age of the grassland and that its frequency increased with overgrazing or with regular cropping for hay. Meadow Buttercup has an acrid taste that makes herbivores avoid it and thus it tends to spread and increase in heavily grazed communities (see below for details of its poisonous properties).

The widespread introduction and development of intensively managed 'improved' rye-grass based sown grasslands, replacing old permanent meadows and pastures, together with regular or frequent application of chemical fertilizer or slurry manure, has led to a considerable decline in the abundance of R. acris and, indeed, of most other lowland grassland herbs throughout the British Isles. These forms of government subsidized agricultural 'progress' are aimed at greatly increasing farm productivity and together with other changes in farmland management, eg changing over from arable to grassland farming, improved land drainage operations, swapping from sheep to cattle husbandry and from hay to silage-making, are all practices which have resulted in depletion of plant species diversity on managed land.

Fermanagh has suffered the accumulated effects of these changes along with almost everywhere else in these islands from the second half of the 20th century onwards. The major transformations have been in the more intensively managed productive farmland, which in Fermanagh is most frequently situated in the eastern lowlands. However, where farming has become intensified, drainage generally leads to increased pollution of both surrounding semi-natural ground and adjacent water bodies. This results from inevitable agricultural chemical runoff, both nutrient and toxic. As a result, permanent 'unimproved' pasture has become restricted to less productive ground, either on less accessible steep or rocky slopes, very shallow soils, or in more upland areas featuring small pockets of grazing that are inaccessible to tractors and spraying machinery. Thus, apart from on lightly or occasionally managed lowland wayside grasslands, R. acris is now most frequent and abundant on pastures at higher altitudes, except where such ground becomes excessively acidic (ie below pH 4.0) and nutrient leached, or where grassland gives way altogether to either ericaceous heath or bracken.

British and Irish occurrence

The New Atlas hectad distribution map shows R. acris having almost total cover across Britain and Ireland, the few absences being on exposed western coasts and in a few scattered squares in Irish and Scottish VCs which probably have no resident botanist. In these latter areas, recording even of such a common species inevitably tends to be incomplete, despite the efforts by non-residents. The Relative Change Index of R. acris, measured in Britain (but unfortunately not surveyed in Ireland), between the recording periods of the two BSBI plant Atlases, is calculated as +0.30. The index is interpreted as indicating that R. acris is a stable species in the area studied (Preston et al. 2002).

Buttercup population studies

Detailed demographic studies of field populations of three common buttercup species, R. acris, R. bulbosus and R. repens, were made in the early 1970s at Bangor in N Wales by Dr J. Sarukhan, supervised by Prof. John Harper. Despite its age this pioneering study remains highly recommended reading since the authors thought deeply about plant population processes and used their data, and that of earlier workers, to deduce important generalisations regarding plant behaviour.

Plant ecology has few general concepts that scientists totally agree about, and absolutely no 'Laws' exist as they do in subjects like Physics. However, it is a useful and illuminating exercise to try and generate such general principles whenever possible. Harper and Sarukhan give us a clear picture of the recruitment rate of each buttercup species from seed and also from vegetatively produced daughter plantlets (ie ramets). This fecundity is then compared with the species population mortality and thus the flux or rate of turnover can be calculated (Sarukhan & Harper 1973; Sarukhan 1974, 1976).

The concept of 'a plant population' is an extremely dynamic one, even more so than the concept of 'vegetation', and in the case of the former it involves the gains and losses of individuals happening at the same time. Thus a census approach is essential to population studies, marking and following the fate of individuals, including some that are actively growing and others that may be going through a dormant phase (ie buried viable seeds have also to be considered as part of the total plant species population picture).

Comparative patterns of population turnover in three buttercups

In his research studies, Sarukhan found that individuals of the three common buttercup species had quite different levels of longevity. This reflected their methods of reproduction and the balance of strategy between seed and vegetative ramet reproduction that each employed. Reproduction in R. bulbosus is exclusively by seed, that of R. repens is predominantly vegetative with some seed, and R. acris is mainly seed, but very occasionally it reproduces vegetatively by branching of its characteristic short rhizome.

During the two and a half year period of Sarukhan's study of the three buttercup species, the population flux was greatest in R. acris. Indeed, he found that in some study plots there was no permanent population, only a series of temporary but overlapping short-term cohorts, establishing rapidly from seed and then soon dying. R. acris seed germinated in April and May, with a swift and high mortality peak in May and June, and mortality continued gradually at very low levels until the following spring. A similar picture of high seedling production and mortality occurred in R. bulbosus, although germination in this species took place in the autumn rather than the spring. In both these buttercup species, the high initial seedling mortality is balanced by a long life span for those minority of individuals that manage to survive and become mature established plants.

The high risk of death at the early seedling stage probably reflects not only the very obvious innate risks involved in establishing a viable root and shoot system, but additionally, and probably even more so, the genetic load of unfit genotypes being carried by the species. The new gene combinations that are produced by sexual reproduction are, by their nature, experimental and elimination of the plants with the most unfit genotypes can be expected to occur early in life, as was the case with all three buttercups (Sarukhan & Harper 1973; Sarukhan 1976).

In comparison with the other two buttercups, R. repens (Creeping Buttercup), the only one of the three with appreciable vegetative multiplication, suffers little loss of vegetative individuals (ie ramets) at the very beginning of their life, but it replaces a large number of them more frequently than the other two species do. The life expectancy of a R. repens ramet was short, ranging only between 1.2 and 2.1 years. Very few plants of R. repens derived from seedlings lived for more than 1.5 years, whereas mature plants of both the other buttercups could survive for much longer periods; in the case of R. acris, individuals were shown to live up to eight or more years (Sarukhan & Harper 1973).

R. repens, with its dependable, high level of vegetative reproduction, comparatively weak flowering and seed production, but great seed longevity, contrasted strikingly with the other two buttercup species which had little or no vegetative reproduction, high seed output, rapid germination, high seedling mortality and a rather short half-life of the buried dormant seed (ie just 5 months in the case of R. acris and 8 months in R. bulbosus) (Sarukhan 1974).

Buried seed longevity

Other estimates of buried seed longevity are given in the detailed European survey of this topic published by Thompson et al. (1997). The survey lists a total of 40 estimates for R. acris divided as follows: transient (less than one year) = 19; short-term persistent (between 1 and 5 years) = 10; long-term persistent (at least 5 years) = 3, and present in soil but unassigned to any of these three categories = 8. The equivalent figures for R. bulbosus seed are: 8, 4, 4 and 5 – a total of 21 estimates. It is not unusual for seed longevity estimates to vary, but for both R. acris and R. bulbosus the predominant impression remains that buried seed is relatively short-lived.

Flowering reproduction

Flowering in R. acris stretches from May through to August, peaking in mid-June. Seeds are shed from July onwards. The female flower parts ripen first (ie the flower is protogynous) and a large variety of short-tongued insects visit them to collect nectar, including Honey Bees (Apis mellifera) (Harper 1957). Spatial fragrance patterns within the bowl-shaped flower guide the insect visitor to the nectary, which is partially concealed by a flap at the base of each petal (Bergstrom et al. 1995). Self-incompatibility in the species is described as, "often very marked, but not [occurring] in all populations" (James & Clapham 1935). Although cross-pollination is frequently achieved, some degree of apomixis (ie agamospermy − seed production without any fertilisation or pollen involvement), does occur, perhaps in as many as 1% of flowers (Marsden-Jones & Turrill 1952).

Flowering in R. acris is so very variable in response to its environment that it has proved difficult to measure its flowering capacity. Even Salisbury (1942, 1964), who was very keen to compute such statistics, did not attempt to do so for this species. In his comparative study, Sarukhan (1974) was more daring than Salisbury and in his sample he found that the number of flowers per plant could range from 1-20. Sarukhan also found a high proportion of plants of both R. acris and R. bulbosus each produced between 40 and 140 seeds in total. The maximum seed number produced by an individual plant in his populations was 281 for R acris and 287 for R. bulbosus.

Due to grazing and other forms of disturbance, Sarukhan's study showed that not all plants that flowered managed to set seed, but in R. acris about 40% of flowers did. Over a two year period, similar proportions of R. acris plants flowered in Sarukhan's study, but the ratio of seeds per flowering plant fluctuated wildly, from 26 in 1969, to a little over 1.0 in 1970 (Sarukhan 1974). Since Harper (1957) reported the number of seeds per flower varying between 0 and 40, and the number of flowers per plant was also very variable in his study, the variability helps explain Salisbury's reluctance to measure the reproductive capacity of these buttercup species.

Seed dispersal

As with other buttercups, R. acris has no specialized seed dispersal mechanism, but viable seed has been found in the droppings of birds including the House sparrow (Collinge 1913). Furthermore, Dore & Raymond (1942) calculated from analysis of the seed content of farmyard manure, that a single cow might disperse around 22,000 R. acris seed per ha during a 165 day grazing period. However, as Sarukhan (1974) pointed out, while the seeds are 'dispersed' in this manner, they are also concentrated in 'local droppings'! Voles and Field mice are very probably the main rodent seed predators of Ranunculus species in grasslands in Britain and Ireland, but birds such as Pigeons and Pheasants must also consume and destroy huge quantities of buttercup seed.

Absence of Ranunculus acris hybrids: Hybrids between R. acris and other buttercup species have never been recorded anywhere in Britiain and Ireland, and while crosses with both R. repens and R. bulbosus have been reported from several European countries, Stace (1975, p. 124) regarded them as doubtful and requiring confirmation, and the new Hybrid Flora of the British Isles (Stace et al. 2015) makes no mention of them whatsoever.

Fossil history

Fossil seed (ie achenes) have been found in sediments of the four most recent interglacial periods (from the Comerian onwards) and also from the two most recent glacial stages. This evidence conclusively proves that although R. acris populations increase when human populations disturb natural or semi-natural forms of vegetation, the species very definitely is native in these islands (Godwin 1975).

Toxins

Meadow Buttercup contains the acrid, irritant, poisonous principle protoanemonin, an unstable compound derived from the glycoside ranunculin. The concentration of protoanemonin increases during growth of the plant and reaches a maximum during the flowering period. Being an unstable chemical, however, the drying involved in hay-making readily converts protoanemonin into an inert, non-toxic substance called anemonin and thus dried fodder containing buttercups is perfectly safe to give to animals. Also, R. acris contains much lower toxin concentrations than is found in Bulbous Buttercup (Cooper & Johnson 1998). Due to the presence of this acrid, bitter-tasting toxin, both of these buttercup species are unpalatable when fresh and they are avoided by grazing animals unless the beasts are actually starving (Harper 1957).

R. acris has been known to poison both grazing cattle and sheep, but few cases have been reported in recent years, probably because pastures nowadays contain much smaller proportions of these species than previously was the case for the reasons discussed above (Cooper & Johnson 1998). In Norway in 1988, however, five cows in late pregnancy were turned out to graze on a field of poor pasture containing abundant R. acris. They developed severe diarrhoea, a rapid pulse and noisy respiration and all of them died or had to be destroyed because of their deteriorating condition (Heggstad 1989). Other experimental studies in Canada found that cattle gradually fed increasing amounts of R. acris in the flowering stage could cope very well with between 7 and 25 kg of the plant per day for two weeks towards the end of the experimental trial (Therrein et al. 1962; Hidiroglou & Knutti 1963, both references quoted in Cooper & Johnston 1998).

European and world occurrence

Beyond Britain and Ireland, R. acris s.l. (ie the polymorphic species we are considering here) is common over the whole of C and N Europe including the Faeroes and Iceland. There is some dispute as to whether or not it is native in Greenland, since material from there was indistinguishable from plants known to have been definitely introduced to Spitzbergen and N America (Coles 1971). R. acris has a more limited distribution in S Europe; it is absent from Portugal and areas S of latitude 40^oN (Jalas & Suominen 1989, Map 1714). The coverage of the species in Italy was under-recorded by the latter map, since Pignatti (1997, 1, p. 306) maps a complete cover of the peninsula. R. acris is also found in Morocco, and while it is very rare on Madeira, it is absent from all of the Canaries (Hultén 1971, Map 288; Press & Short 1994).

R. acris is very widely naturalised across the world and as a result has become circumpolar in the N Hemisphere (Hultén & Fries 1986, Map 844; Preston & Hill 1997). It is also introduced in a few temperate areas of the S Hemisphere, eg in S Africa and New Zealand (Hultén 1971; Hultén & Fries 1986).

Names

'Ranunculus' is derived from the Latin 'rana' meaning 'a frog', an allusion to the fact that so many members of the plant genus live in or near water, the habitat of frogs (Johnson & Smith 1946). The Latin specific epithet 'acris' means 'acrid', ie with a sharp, burning, peppery taste (Gilbert-Carter 1964). This is very descriptive of all parts of the plant. Contact with skin can cause severe blistering (Grieve 1931; Cooper & Johnston 1997). Despite this, the caustic sap has been used in herbal medicine to remove warts and the plant has also been used to treat headache, gout and even cancer (Grieve 1931). As ever, a health warning should here be attached to these comments and no recommendation whatsoever is intended or implied by the inclusion here of this information.

The English common names include a recognition of the danger of handling the plant, eg 'Blister cup' and 'Blister plant' are both listed by Grigson (1987). Britten & Holland (1886) provide a list of 38 varied common names including 'Clovewort', 'Crowflower' and 'Bassinet'. The latter name means 'a small basin', apparently a reference to the bowl shape of the flower and, therefore, quite widely applied to a whole range of flowers, including all species of buttercup, Caltha palustris (Marsh-marigold) and many Geranium species. 'Crowflower', 'Crowfoot' and other versions of this name, refer to the deeply cut leaf shape of many buttercup species and their relatives plus, again, some species of Geranium, eg G. pratense (Meadow Crane's-bill) (Grigson 1974). Other common names refer to the gold or yellow flower colour, but the strangest name of all appears to be 'Crazy' (see below in the species synopsis of Ranunculus repens) (Prior 1879).

Threats

None. R. acris is readily exterminated by modern systemic herbicides and the species has only a negligible, short-lived, soil seed bank. It has survived for thousands of years despite these limitations.

Growth form, ecology and Fermanagh occurrence

This vigorous, rosette-forming, wintergreen, short-lived perennial is very variable in form (especially with respect to leaf shape and degree of hairiness), and it can grow in almost any habitat provided it is damp (except very acidic peat bog), meaning virtually everywhere in Fermanagh! Creeping Buttercup is in fact the second most frequently recorded and widespread plant in the Fermanagh Flora database, being found in 515 tetrads, 97.5% of those in the VC. Creeping Buttercup thrives and, indeed, is most abundant in heavy mineral or clayey soil where drainage is naturally impeded (Harper 1957). In a survey of the Sheffield area, Grime et al. (1988) found R. repens almost entirely absent from infertile acidic soils with a pH below 4.5 and from permanently flooded sites. Creeping Buttercup is also common on wooded, open marsh or fen-fringed lakeshores, riversides and stream banks, on ditches and by roadsides, especially on wet, heavy soils. In these situations, its long creeping stoloniferous stems, rooting at their nodes and their deep, stout, tenacious roots emanating from a short erect rhizome, make the plant extremely difficult to eradicate or control.

R. repens can regenerate from very small root fragments, plus its seeds show dormancy and sustained viability in the soil seed bank (see below for details) and it is also resistant to many herbicides. These factors combine to make it particularly difficult to devise an effective weed control strategy for it in cultivated ground (Lovett-Doust et al. 1990).

In lakeshore grasslands, Creeping Buttercup occupies a zone between drier ground, where it is forced to compete with R. acris (Meadow Buttercup), and wetter soil lower on the shore where Caltha palustris (Marsh-marigold) becomes dominant. R. repens is also a common weed of disturbed soil and gravel and in these more open habitats it can tolerate very much drier conditions, rapidly establishing and spreading vegetatively, its numerous stolons quickly forming large clonal patches (Harper 1957). Salisbury (1964, p. 203) reported that, under favourable soil conditions, an individual plant could spread vegetatively over 4 m² in a single year. However, the species is phenotypically very plastic, and particularly so with respect to stolon production. Stoloniferous growth closely reflects both soil fertility and the intensity of plant competition.

Effect of trampling on populations

Creeping Buttercup can tolerate frequent disturbance and a considerable degree of soil compaction and as a result it is very common around field gates, along paths and on forest and woodland tracks and clearings. Damp, heavy soils frequently become 'poached' or puddled by the hooves of cattle or other stock animals (especially around gates or feeding troughs) and grasses are often killed under these conditions. R. repens is an efficient pioneer species colonising this type of disturbed, bare ground. It invades rapidly through germination of re-exposed buried dormant seed and from nearby plants by the extension of their stolons and the attached rooting plantlets.

The number of daughter plantlets (ie ramets) has been shown to increase with trampling of the vegetation (Diemer & Schmid 2001). The leaves of trampled plants spreading vegetatively in this manner are especially large, which enables R. repens to cover and hold on to previously open patches in the turf or soil and prevent invasion by competing pioneer species (Harper 1957).

Since R. repens possesses an impressive ability to rapidly colonise disturbed ground, to a large extent it has become a follower of man. The plant is a significant weed of gardens, waste disposal areas, building sites, dredgings of river banks, hedgerows, roadsides and, indeed, it occurs on any form of disturbed ground, including in the depressions made by animal hooves which tend to be damper than surrounding ground through holding rain, dew, or water from other sources (Harper 1957).

Turloughs ('vanishing' lakes) − an unusual and specialised habitat

Other open habitats which R. repens occupies in Fermanagh include limestone pavement and scree. In sharp contrast to these latter relatively dry conditions, it also occurs on the shores of a special type of limestone lakes called turloughs (ie so-called 'vanishing lakes' that drain vertically into underground cave systems).

Recent studies in W Ireland have shown that turlough populations of R. repens differ from more typical broad-leaved ruderal plants in both their leaf form (they have more highly-dissected and more glabrous leaves) and in their physiology (the turlough plants have a higher rate of aerial and submerged photosynthesis) (Lynn & Waldren 2001, 2002).

Life expectancy in comparative Buttercup population studies

During detailed population census studies of R. repens and two other buttercup species (see the R. acris species account above), the life expectancy of individuals of R. repens decreased significantly with increasing density of the plant population (Sarukhan 1976). The study also clearly showed that the highest mortality rates per week were obtained, not in the unfavourable phases of the physical environment (ie during the winter), but rather they coincided with active growth phases of the plant (Sarukhan & Harper 1973).

Flowering reproduction

R. repens flowers are quite variable in size and in the number of shiny yellow petals they possess. They contain nectar and are pollinated by honeybees over a 4-9 day lifespan. The curve of the honeybee's body closely mirrors and 'fits' that of the stamen cone of the buttercup to a remarkable degree but while Percival (1955) noted bees actively collecting nectar from R. repens flowers, he remarked that pollen collection has very seldom been seen. Having said this, all buttercup flowers are primitive and unspecialised, meaning that their nectar and pollen flower foods are available to all types of insect visitors. Thus they attract a great many different insect species and are probably pollinated by many of them. The flowers are so unspecialised, they could possibly also be self-pollinated by raindrops (Van den Berg et al. 1985; Proctor et al. 1996; Jonsell et al. 2001). Having said this, cross-pollination is very much the norm, but a low level of selfing is also possible. There is no evidence of apomixis (ie the asexual formation of seed without fertilization taking place) (Coles 1977).

Fruiting performance

Under favourable conditions of slight or negligible competition, the average R. repens plant individual produces five fruit heads although the number ranges from 0-38. However, the frequency distribution is very heavily skewed, the most frequent class having just three fruiting heads per plant (25%). Each fruiting head contains a mean of 30 achenes, giving a mean total output of 150 ± 10 achenes per flowering plant. When subjected to marked competition, however, R. repens flowering becomes suppressed and, if the plant persists, as it does in some wet habitats, the reproductive balance is even more completely directed towards stolon development, with seed production then becoming meagre or completely absent (Salisbury 1942, p. 226).

Individual plants normally die off after they have successfully fruited, being replaced by a daughter plantlet produced vegetatively on a very short stolon, ie the plant is usually (but not always) monocarpic. This is another example of a biological concept (ie monocarpic versus polycarpic reproduction), which is not absolute, but rather, it is somewhat 'leaky' or facultative in its mode of operation (Forbes 2000, p. 187). The connecting stolons generally (but not always) die off in the autumn from September onwards, leaving the daughter rosettes produced at their nodes as independent plantlets (Van den Berg et al. 1985; Jonsell et al. 2001).

Population studies and plant plasticity and variation

R. repens is one of the best studied weedy plants from a population biology or demographic perspective, having been the subject of major studies by Prof. John Harper and several of his co-workers, including amongst others, Lovett-Doust (1981). The latter showed that Creeping Buttercup populations studied in grassland and open woodland habitats follow similar seasonal patterns. The populations examined remained remarkably stable from year-to-year, but the density of plants in the woodland was significantly greater than that in the particular grassland examined. This suggests that some form of density related self-regulation of population size occurs at a figure referred to as 'the carrying capacity'. The latter varies according to a number of environmental factors. The average time for complete turnover of rosette populations was calculated as 2.17 years for woodland and not significantly different at 2.27 years for grassland rosettes (Lovett-Doust 1981). The creeping habit of the species is a response to the pressure of close grazing (or mowing) and if this is removed the plant will grow more upright (Harper 1957).

The population dynamics of R. repens in pastures in N Wales were examined over a four year period and computer modelled by Soane & Wilkinson (1979). These workers found little evidence of selection among families of clonal rosettes, or against new seedling recruits within populations. Their measurements showed that the number of original genotypes present in a population of R. repens declines continuously at an approximately exponential rate. Local dominance by a few clones is therefore to be expected unless new genotypes are recruited into the population, eg by seedling establishment. There was no evidence that selection was maintaining a diversity of genetic individuals (ie genets) within the R. repens population, but although recruitment of new seedlings was low enough to be described by them as 'occasional', nevertheless it clearly plays a very significant role in the longer term, through determining the number of genets represented in the population, and thus maintaining genetic variation within these populations (Soane & Watkinson 1979).

Variation

Creeping Buttercup is continuously variable in so many of its characters that Coles (1977) found no justification for the distinction of any infraspecific units within the species (ie forms, varieties or subspecies).

Possible allelopathic effect on competing plants

There is a suggestion that R. repens may exert an allelopathic effect (ie a chemical inhibition) when competing with the roots of other plants, possibly involving phenolic compounds (Whitehead et al. 1982). Hatfield (1970) regarded R. repens as responsible for serious depletion of potassium and other elements from soil and he proposed that the roots secrete a toxin causing neighbouring plants to suffer from a nitrogen deficiency (Lovett-Doust et al. 1990). More work is required to clarify the real position on this topic, but as yet nobody has proven that any allelopathic effect exists, although it might be more likely occur in soils rich in lime, or after lime has been applied (Whitehead et al. 1982).

In upland leached acidic soils and in other situations of low fertility, for instance in peat bogs and in wetter, marshy areas, R. repens is much less competitive than in drier, lowland situations and here it tends to be replaced by R. flammula (Lesser Spearwort).

Colonising ability

The immediate opportunistic response of R. repens to disturbed environmental conditions permitting colonisation of new territory tends to be increased stolon development, rather than seedling production. This is thought to be due to stolon production and flowering being largely coincident in May and June, plus the fact that vegetatively produced offspring often do not flower in their first year of growth. However, juvenile plants may do so if the habitat is very open and in these circumstances they generally flower rather late in the season, up until about October (Harper 1957).

On the other hand, seed germination is greatest in late spring, with just a few seedlings appearing in autumn, and then only if there is a combination of high soil temperature and abundant moisture. Germination is very rapid (almost immediate) when seeds are exposed by soil disturbance. Seedlings establish readily where ground is open and particularly when the water table is high but the soil is not completely flooded.

Very rapid colonisation of bare ground may be achieved in the year of germination (Harper 1957). Salisbury (1942, p. 225) illustrates a case where a plant in open garden soil occupied more than 0.5 m² in its first growing season, producing 35 rooting nodes, of which 23 bore inflorescences and twelve remained vegetative. Clearly, this is merely an isolated instance, and growth rate will be dependent upon habitat conditions, but it provides a helpful indication of the level of colonising ability the species is capable of achieving.

Further detail of R. repens population behaviour is present in my R. acris species account under the heading, 'Comparative patterns of population turnover in three buttercups'.

Seed dispersal

No specialised dispersal mechanism exists in R. repens, most of the dry, smooth seed simply being dropped beneath the parent plant. However, cattle and horses are known to help disperse R. repens by transporting the seed in their gut, it having been found in their droppings. Various birds do likewise, notably the House sparrow, but while Partridges, Pheasants and Pigeons are found frequently with a high crop content of R. repens seed, it is unlikely to pass through their guts in a viable condition. Plants growing beside water may disperse seed in flowing streams and occasionally whole plants will dislodge and migrate downstream in the adjacent flow. These processes may possibly be assisted by the disturbing activity of various ducks and water hens (Harper 1957; Salisbury 1964).

Other animals may also act as agents of dispersal, including those which transport propagules externally on their surfaces. This form of dispersal agency includes man and his vehicles. Darlington (1969) found that washings of mud from motor tyres in the month of June contained a considerable number of R. repens achenes. In another novel and unusual, not to say quaint study of the trouser turn-up fluff of schoolboys who walked across fields to school, he showed that of the 70 plant species the boys unwittingly transported, 11% of the propagules were of R. repens (Darlington & Brown 1975). These authors pointed out that with the exception of adherent burrs like those of Galium aparine (Cleavers), the majority of the fruits and seeds in the turn-up are carried loose in the contained dust and fluff, "so that the wearer becomes a sort of peripatetic censer mechanism for scattering propagules, notably the smoother kinds (R. repens and others), as he walks about" (Darlington & Brown 1975, p. 34).

Seed dormancy and survival in soil

Dormancy is enforced by burying the seed and large populations of buried viable seed have been reported. Seed survival ability varies enormously, presumably dependent upon soil moisture, nutrient levels, stability and disturbance.

The survey of soil seed bank of NW Europe tabulates results of no less than 98 records of buried seed survival. Of the four seed bank categories listed, the representation of R. repens appears as follows: transient (surviving less than 1 year) 21; short-term persistent (between 1 and 5 years) 26; long-term persistent (at least 5 years) 30; and present in soil (but not assigned to one of the foregoing) 21 studies (Thompson et al. 1997).

British and Irish occurrence

R. repens is extremely common and widespread over almost all of Britain and Ireland, becoming slightly less frequent in the NW and highlands areas of Scotland (Preston et al. 2002).

World occurrence

It has an almost continuous Eurasian boreo-temperate native range (Jalas & Suominen 1989, Map 1712). R. repens is also introduced and partly naturalised in both N & C America and has thus become circumpolar in its Northern hemisphere distribution. It is also an introduction in South America, South Georgia, New Zealand, Tasmania and Great Barrier Island (Hultén 1971, Map 225; Hultén & Fries 1986, Map 842; Preston & Hill 1997).

Toxicity

Unlike other common buttercup species, R. acris, R. bulbosus and the more scarce R. scleratus (Celery-leaved Buttercup), it appears that Creeping Buttercup in the British Isles normally contains only a low concentration of the Ranunculus poisonous principle protoanemonin (a toxic cardiac glycoside). Consequently, R. repens does not harm stock animals − including horses − and they frequently browse upon it (Cooper & Johnson 1998).

In the wider geographical range of the species however, there are instances where the levels of toxins in populations of R. repens are sufficiently high to make the plant distasteful, or even seriously poisonous, making it capable of causing diarrhoea and abdominal pain in cattle and sheep with symptoms that persist for up to 14 days (Lovett-Doust et al. 1990). In Chile, for instance, in recent times R. repens poisoning was held responsible for causing abortions in a herd of dairy cattle (Morales 1989).

Herbicidal control

Although resistant to a wide range of herbicides, R. repens is very sensitive to translocated selective herbicides such as 2,4-D, MCPA-salt, MCPB-salt, paraquat and aminotriazole (Lovett-Doust et al. 1990). Selective herbicides containing aminopyralid, such as Milestone and VM, can be used to kill Creeping Buttercup. Aminopyralid products such as these are available at farm supply stores and should only be used in areas listed on the label, ie pastures, hayfields and other agricultural settings. Fortunately, aminopyralid products do not harm livestock, provided all precautionary advice is followed. The Nature Conservancy Wildland Invasive Species Team publishes an online Weed Control Methods Handboook that is regularly updated (Tu, M. et al. 2001: http://invasive.org/gist/products/handbook/methods-handbook.pdf

Accessed 25 January 2016). This tabulates herbicide advice and makes recommendations on other ecological manipulations which help limit the weed population. To eradicate Creeping Buttercup from grassland it will probably be necessary to apply herbicide up to three times, since mature plants can often recover, and seed in the soil seed bank will germinate and may re-establish the plant. Sprayed ground will need to be monitored and seedlings removed before they develop runners.

Uses

As R. repens tissues normally contain only a small percentage of protoanemonin, the species has not been used in herbal medicine in the same way as its close relatives, R. bulbosus (Bulbous Buttercup) and R. acris (Meadow Buttercup). In fact, Grieve (1931) does not mention it at all in her comprehensive book, A Modern Herbal.

Names

'Ranunculus' is derived from the Latin 'rana' meaning 'a frog', an allusion to the fact that so many members of the plant genus live in or near water, the habitat of frogs (Johnson & Smith 1946). The Latin specific epithet 'repens' means 'creeping'. As one might expect, R. repens shares many English folk or common names with R. acris. Additional ones include 'Devil's Guts', 'Gold-balls', 'Granny-threads', 'Hod-the-Rake', 'Lantern Leaves', 'Meg-many-feet', 'Ram's Claws', 'Sitfast', 'Sitsicker', 'Tether-Toad' and 'Toad-tether' (Britten & Holland 1886). Many of these names refer to the spreading stolons and/or the tenacity with which its roots cling to the ground, making the plant difficult to eradicate.

Threats

None.

Growth form and habitat preferences

Early in the growth season, when plants of this perennial are not yet in flower, the lower leaves of each individual's basal rosette lying flat to the ground are a distinctive identification feature of Bulbous Buttercup and, in addition, the corm-like, swollen stem-base can often be felt underneath the leaves. The lowest leaves being close-pressed to the ground (a feature described by Harper (1957, p. 332) as, 'strongly epinastic' − an interesting term), appears to confer a distinct competitive advantage over other plants in the sward, since it allows the R. bulbous individual to form a 5-7 cm saucer-shaped hollow in the turf, from which it manages to exclude other species. When Bulbous Buttercup is in flower, the down-turned (ie reflexed) sepals are very obvious and they provide another ready means of recognition.

While Bulbous Buttercup is a characteristic perennial of Fermanagh's limestone pastures, it is not a strict calcicole, ie it may prefer and occur most abundantly on well-drained, lime-rich soils, but it is not entirely confined to them. Rather, the species also occurs on dry roadside banks and verges as well as on well-drained unimproved meadows and pastures, including some over more neutral to moderately acidic soils (ie around pH 5.0 and above). This can also involve the shallow peaty soils that form over leached Carboniferous limestone in Fermanagh and other areas in western Ireland.

Essentially a lowland plant of open, full-sun situations, R. bulbosus can withstand moderate amounts of disturbance such as trampling and since it contains protoanemonin and is unpalatable, stock naturally avoid grazing it (Cooper & Johnston 1998). Severe trampling and the associated degree of soil disturbance and compaction will, however, obliterate the plant and its complete absence from frequently used pathways in grassland is often very obvious. Bulbous Buttercup also avoids strongly acid, wet, shaded or indeed overly fertile, productive conditions. In the latter circumstance, taller and more vigorous plants out compete it (Harper 1957; Grime et al. 1988).

Fermanagh occurrence

R. bulbosus is much less common in Fermanagh than either R. acris (Meadow Buttercup) or R. repens (Creeping Buttercup) and has only been recorded in 66 tetrads, 12.5% of those in the VC. Nine of these tetrads contain only pre-1976 records, indicating a loss of suitable habitat which is easily identified with agricultural improvements made to lowland limestone grasslands in recent decades. As the distribution map shows, Bulbous Buttercup is widely scattered across Fermanagh, but it is definitely most frequent in the Monawilkin, Knockmore and Marlbank limestone areas.

Flowering

R. bulbosus flowers earlier than its closest buttercup relatives, R. acris and R. repens, the peak of its flowering normally occurring in mid- to late-May. Fruit has normally ripened by the end of June and the aerial parts may brown and die off soon afterwards, so that the summer-dormant plant survives as the corm just below the soil surface − a drought-avoiding mechanism not very necessary in the wet climate of western Ireland! The species can thus be very inconspicuous for a few weeks in mid-summer, particularly if we have a dry spell of weather. However, in our damp, mild climate, with rainfall typically occurring regularly throughout the summer, Bulbous Buttercup plants produce fresh basal leaves soon after flowering (ie by the end of July or in early August), and they maintain their growth until winter cold eventually stops them in October or November. The leaf rosette is maintained throughout the winter and it recommences vegetative growth early in the year once temperatures begin to rise. Although the corm-like stem base and all the other parts of the plant are renewed annually (Harper 1957, p. 333), this method of perennation means that the plant tissues never become old and senescent and thus established individuals of R. bulbosus can be exceedingly long-lived and, indeed, in stable habitats they may persist indefinitely. There is nothing against saying that, in certain circumstances, some individuals could be thousands of years old!

As with other common buttercup species, R. bulbosus shows considerable variation in form in relation to prevailing environmental conditions (ie phenotypic plasticity), especially with respect to the size of corms and leaves and the number of flowering stems per corm (Harper 1957; Coles 1973). The typical plant has a corm 1.5 cm in diameter which produces just one flowering stem, but larger corms can carry up to eight or more flower stems and Harper (1957) reports a record breaking plant bearing 42!

The flowers open daily for about 4-7 days and are visited for both pollen and nectar by honey bees and many other short-tongued insects. Cross-pollination is the norm since a high degree of self-incompatibility exists, but while some evidence of selfing and a low degree of agamospermy, ie seed formation without fertilization, had been reported in the past (Harper 1957). However, Coles (1973) concluded after he had carried out greenhouse tests that R. bulbosus is totally self-incompatible, cross-pollinated and sexual, ie the flowers are fully outcrossing (xenogamous), and the breeding system is ‘panmictic’ (ie it involves random matings)(Richards 1997, page 6). Each flower produces between 20 and 30 achenes or seed (ie the achenes are single-seeded dry fruits).

Variation in seed production

The scale of seed production, or the reproductive capacity of the species, is another characteristic that varies enormously with the environment and particularly with the competitive situation of the plant. It is hard to measure or summarise reproductive capacity, but Salisbury (1942) provided experimental evidence by comparing the growth and productivity of plants in a meadow under three regimes: no competitors, and slight and severe levels of plant competition. As Harper (1957) pointed out, the competitive measurements Salisbury made were carried out without any experimental control, yet they still give a clear impression of the overall scale of the effect of competition on the reproductive capacity of the species. In the case of R. bulbosus, the sample under severe competition produced 69 fertile carpels per plant, while that with no competitors produced 687, a tenfold difference in sexual productivity.

Seed dispersal

The ripe seed drop off the receptacle and there is no specialised means of dispersal. Internal transfer in the gut of animals through being eaten by birds or stock animals, plus external carriage in mud by animals, including man and his vehicles, are probably significant. Wind and rain-wash may also play some role in dispersal of R. bulbosus achenes (Harper 1957). Other bird species, such as Pigeons, are known to be responsible for a considerable level of seed predation, and Voles and Field mice also make depredations.

Phenology

Bulbous Buttercup germinates mainly in the late summer and early autumn on bare ground in gaps created in turf by disturbance such as trampling and overgrazing (Sarukhan & Harper 1973; Sarukhan 1974, 1976). In Britain, molehills may be significant in this respect, providing fresh, bare soil, but fortunately there are no moles in Ireland. Although tremendous rates of seedling loss are involved, effective establishment from seed is the only reproductive mechanism that Bulbous Buttercup displays, corm division being extremely rare.

Seed survival in the soil seed bank

Buried seed longevity frequently appears to be brief: eight out of 15 studies in a major survey in NW Europe which gave survival estimates in years, indicated that the seed bank of R. bulbosus in soil is transient (surviving less than 1 year), and four indicated that it is short-term only (surviving more than 1 year but less than 5) (Thompson et al. 1997). The remaining three studies gave estimates of R. bulbosus seed longevity that ranged from 5 years to over 30! Clearly there is considerable variation in estimates of seed longevity, but the predominant mode remains transient to short-term only.

Fossil record

The fossil record of R. bulbosus is more slight than that of other species in the genus, but the pattern is similar through numerous glacial and interglacial stages and thus provides conclusive proof that the species is native in Britain and Ireland (Godwin 1975).

British and Irish occurrence

R. bulbosus is widespread in Britain and Ireland, but less so than either R. acris or R. repens. In Ireland, it is widely distributed but is less frequently recorded in western and southern areas of the island. The pattern in Britain shows the species is less frequent in the N & W of Scotland. While this may in part be due to the upland, acidic nature of the terrain in the N & W of both islands, the fact that the species is only conspicuous and readily identified in the early summer, probably means that it is under-recorded in areas where botanical recorders are themselves scarce, or indeed, rare (New Atlas).

The calculated Change Index value measuring change between the two BSBI plant Atlases (published in 1962 and 2002) is -0.48, which is taken to indicate very little loss over the 40 year period (R.A. Fitzgerald, in: Preston et al. 2002). The results of the 2003-04 BSBI Local Change re-survey in Britain of the 1987-88 BSBI Monitoring Scheme survey squares suggest that R. bulbosus is actually on the increase in Britain. Unfortunately, this re-survey was not extended to the island of Ireland. In comparison with other widespread plant species of predominantly calcareous or neutral grasslands, Bulbous Buttercup can cope relatively well with factors which lead to ranker swards, including eutrophication (essentially, nitrogen enrichment) (Braithwaite et al. 2006).

European occurrence

In continental Europe, depending upon which source is consulted, either two or three subspecies of R. bulbosus are now recognised: subsp. bulbosus which occurs in the British Isles, and subsp. adscendens which is confined to the Mediterranean region, presumably including N Africa (Coles 1973). Subsp. adscendens may be subdivided into subsp. castellanus in NW Spain and subsp. aleae in S Europe, extending eastwards to Hungary (Tutin & Akeroyd, in: Tutin et al. 1993). Plants of R. bulbosus in N Africa, N Turkey, Georgia and Azerbaijan also lie within subsp. aleae (Hultén & Fries 1986, Map 845).

Subspecies bulbosus is widespread over W and C Europe, but it is absent from much of the N and E of the mainland. Having said this, it is frequent in Denmark and southern Sweden, rarer on the south coast of Norway, and extends east to Belarus and the Balkans (Hultén & Fries 1986, Map 845; Jalas & Suominen 1989, Map 1743; Jonsell et al. 2001, p. 289). Some outlying occurrences in C Finland and C Russia are considered occasional only (Hultén & Fries 1986).

World occurrence

Bulbous buttercup is a naturalised alien in N America, apparently invading from both E and W coasts, although much more frequent in eastern states. It is also a commonly naturalised weed in New Zealand (Harper 1957; Hultén & Fries 1986, Map 845; Jonsell et al. 2001).

Names

'Ranunculus' is derived from the Latin 'rana' meaning 'a frog', an allusion to the fact that so many members of the plant genus live in or near water, the habitat of frogs (Johnson & Smith 1946). The Latin specific epithet 'bulbosus' means 'having a bulb', but of course technically it is not a leafy storage organ, that is, a bulb, but rather stem tissue (Gilbert-Carter 1964). Thirty-eight English common names are listed by Britten & Holland (1886), many of which are widely applied and some of which are merely mis-spellings, for example, 'Bolt', for 'Bout', which is derived from the French 'Bouton d'or', referring to yellow flower-buds of this and other species (Prior 1879). One of the more interesting names is 'Lodewort', said to be an Anglo-Saxon name for the species, but also applied by some to R. aquatilis (Common Water-crowfoot) (Britten & Holland 1886). Another name of interest is 'St Anthony's Rape' or 'St Anthony's Turnip', from its corm being a favourite food of pigs, and he being the patron saint of pigs (Prior 1879, p. 204).

Uses

Although in Britain and Ireland it is not nearly as common a species as several other buttercups, R. bulbous is very conspicuous early in the growing season and, perhaps for this reason, it has long been used in herbal medicine for its blistering properties. Grieve (1931) summarises the many uses, some of which like the cure for a headache which involves applying the acrid juice to the nostrils, makes the present author's mind boggle at the very thought. THIS DEFINITELY IS NOT ADVICE TO BE FOLLOWED.

Threats

Improvement of grasslands involving ploughing, reseeding and application of fertiliser.

Growth form and habitat preferences

Goldilocks Buttercup is a perennial with a short, stout rootstock and in Fermanagh it mainly occurs (or perhaps is most noticed) beneath hedges, especially on roadsides where its distinctive pale yellow flowers can be spotted even from a car. Less frequently it is found in quite deep shade in hazel woods on limestone, eg in the Screenagh River Glen. These are moist to fairly dry sites and R. auricomus appears to require moderately fertile, base-rich, generally calcareous soils. It really thrives when it is supplied with a good depth of rich leaf-mould in undisturbed corners of old woodland or under scrub in light to half-shade conditions.

R. auricomus avoids both very acid and very dry sites (Garrard & Streeter 1983) and it appears to be both a weak competitor and intolerant of grazing or cutting, tending to shun unshaded meadow grassland for these reasons (Salisbury 1942, p. 54; Sinker et al. 1985). R. auricomus also tends to be a lowland species in Britain and Ireland, although in some parts of Scotland at least, it can rarely be found on open moorland when it is protected from grazing by boulders, and it may also occur rarely on mountain ledges (R.A. Fitzgerald, in: Preston et al. 2002).

Further north, in the Nordic region of continental Europe, R. auricomus shows a very much wider habitat range than is observed in Britain and Ireland, appearing in much more open situations, eg in meadows and grazed pastures, in wetter littoral and riparian habitats with bare soil, in mountain snowbeds and on scree. It is also a weed in cultivated and disturbed ground in the more northern part of its species distribution. Another difference is that it appears indifferent to lime in these northern territories (Jonsell et al. 2001).

Overwintering (perennation) and phenology

The rootstock of R. auricomus overwinters with its bud at or just beneath the soil surface (ie it is a rosette-forming hemicryptophyte, or a buried geophyte). It begins growth in the very early spring, enabling it to flower a couple of weeks before R. bulbosus (Bulbous Buttercup) and several weeks earlier than all the other buttercups in Britain and Ireland. In a comparative study of the flowering behaviour of five buttercup species in central Germany, R. auricomus was first to flower and it had a five week flowering period (Steinbach & Gottsberger 1994).

British and Irish occurrence

R. auricomus is almost certainly an under-recorded species in Britain and Ireland and particularly so in the less-frequented corners of these islands, due to its early flowering season. This runs from March to May and peaks in April, so that it may be missed by 'summer botanists'. Despite their awareness of this, Webb & Scannell (Flora of Connemara and the Burren, 1983) regarded Goldilocks Buttercup as a rare species in W Ireland and it is generally considered that R. auricomus declines in frequency in Britain and Ireland as one goes northwards and westwards, a belief reinforced by the distribution shown in the New Atlas (Preston et al. 2002).

R. auricomus is frequent and widespread in England and to some extent at least it mirrors the distribution of chalk and limestone. In Wales, Scotland and Ireland, however, R. auricomus is very much more scattered and the mapped distribution in these areas does not reflect the calcareous geology. This is especially the case in Ireland where the distribution of the species is much better represented than was the case in the previous 1955-60 BSBI Atlas survey (Perring & Walters 1962, 1976; Preston et al. 2002, page 3).

Fermanagh occurrence

In Fermanagh, there were just 18 pre-1975 records for R. auricomus, but when RHN started looking for it around 1986, he found that it was very widespread, quite frequent and locally abundant. We now have over 250 records and the distribution map shows Goldilocks Buttercup present in 121 tetrads, 22.9% of those in the VC. Harron in his Flora of Lough Neagh (1986) made the very same discovery around Lough Neagh, where he considered R. auricomus present in greater abundance than anywhere else in Ulster (ie in the nine county Irish province).

Apomixis, an unusual breeding system producing an exceptional range of variation

R. auricomus is a facultative apomictic, ie in addition to the normal sexual process it displays pseudogamous agamospermy − which is a shorthand technical way of saying that it can set seed asexually, but only after pollination takes place (either by crossing or selfing). Despite the requirement for pollination, in the case of apomixis, no actual fertilisation takes place, yet seed is produced. For a simple introduction to this complex matter see Proctor et al. (1996), pp. 348-349, and for a more detailed explanation see Richards (1997a), p. 405 and pp. 411-420).

As a further complication, the apomictic microspecies created are all at least tetraploid (Jonsell et al. 2001). As a result of its dual reproductive methods, R. auricomus is extremely variable in form, especially with regard to petal development. A very full treatment of the microspecies has recently appeared in Flora Nordica, 2 (in English), where a total of no less than 605 microspecies in the Nordic countries are described (Jonsell et al. 2001). The agamospecies have not yet been formally described within the British Isles, but Stace (1997) reckons at least 100 R. auricomus microspecies exist in these islands.

Flower variation

In Fermanagh, as elsewhere, plants of R. auricomus are sometimes found with five perfect petals as, eg in the Teemore district, but more frequently the petals are reduced in number, distorted in their development, or even in some cases totally absent. When the petals are much reduced or are completely absent, the sepals, which are usually greenish, may develop yellow colour and become shiny and petaloid, thus taking on the advertising role as insect attractants (Hutchinson 1972).

Pollination compared with related species

R. auricomus has either a cup-like nectary without a covering scale, or it has a small or abortive scale (Butcher 1961; Clapham et al. 1962). The flowers attract insect visitors, but in smaller numbers than most other terrestrial buttercup species. A study of five common buttercup species in central Germany found that while in its natural habitat R. flammula (Lesser Spearwort) attracted up to a mean of 35.7 insects per hour, R. auricomus achieved a mean of just 2.3 per hour. When cultivated in a garden bed R. auricomus was visited by ten species of insects (evenly divided between Diptera, Hymenoptera and Coleoptera), whereas R. acris (Meadow Buttercup), R. flammula, R. bulbosus (Bulbous Buttercup) and R. repens (Creeping Buttercup) had visits from 54, 41, 37 and 28 species, respectively. Thrips were the only insects observed collecting nectar from the Goldilocks Buttercup, while the other nine visiting species took only pollen. In its natural habitat, R. auricomus was visited only by Coleoptera at the very low rate quoted above (Steinbach & Gottsberger 1994).

Seed production

The present author does not know of any study of the reproductive capacity of R. auricomus (certainly nothing published in English).

Seed dispersal

Nordic studies of seed dispersal in their much wider range of habitats emphasised the role of human activities (eg transport of hay), for both local and long-distance microspecies movements (Jonsell et al. 2001). However, this is obviously irrelevant, or much less relevant, when the plant lives in woods and hedges rather than in meadows, as is the situation in most of Britain and Ireland. How does Goldilocks Buttercup get about? Does it have a dispersal mechanism at all? The seed (actually achenes) are described as, "very shortly pubescent" (Stace 1997), perhaps suggesting they might adhere to animal coats. Another project beckons!

Buried seed longevity

The survey of soil seed banks in W Europe found a total of 15 estimates , ten of which regarded R. auricomus seed as transient, two considered it long-term (ie persisting for at least 5 years), and three did not specify any duration (Thompson et al. 1997).

A call for basic ecological information

A careful but necessarily incomplete internet search failed to unearth any other information on the life history or ecology of this interesting species group. Perhaps if less research emphasis were placed on its genetics and reproductive strategy and a preliminary study initiated on the population biology, life-table and natural history of the species group, it might prove worth the effort.

European and World occurrence

In W & N Europe, R. auricomus s.l. is widespread everywhere except Spain and the Mediterranean mainland, where it becomes rare and scattered towards the south. It is absent from all Mediterranean islands except Corsica (Jalas & Suominen 1989, Map 1809). R. auricomus s.l. also occurs in N Asia, Alaska, NE Canada and Greenland (Jonsell et al. 2001).

Uses

Despite its attractive-sounding English common name, the species does not appear to have any folk-lore or use specifically associated with it. This is probably because the plant is too rare, or is seldom recognised.

Names

'Ranunculus' is derived from the Latin 'rana' meaning 'a frog', an allusion to the fact that so many members of the plant genus live in or near water, the habitat of frogs (Johnson & Smith 1946). The Latin specific epithet 'auricomus' is a combination of 'aurum' meaning 'gold' and 'coma', meaning 'hair of the head' or 'locks', and thus translates as 'with golden hair', presumably a poetic likening of the spring carpet of yellow flowers to a blonde head of hair (Gilbert-Carter 1964). The English common name 'Goldilocks' or 'Goldylocks' is a straightforward translation of its Latin specific name, first used by William How in his Phytologia britannica of 1650. An alternative name is 'Wood Crow-foot' (Britten & Holland 1886).

Threats

Removal of both hedges and small patches of woodland.

Growth form and habitat preferences

R. sceleratus is quite a tall (30-70 cm), conspicuous but rather scarce, much branched, yellowish green plant producing a host of rather small yellow flowers. This many-seeded winter or summer annual is a pioneer coloniser of shallow water, or wet, disturbed, nutrient-rich (especially nitrogen-rich), bare mud after it has been thoroughly disturbed, eg heavily trampled and poached by drinking livestock. The animals also provide, of course, the required nitrogen in their excretion (van der Toorn 1980; R.A. Fitzgerald, in: Preston et al. 2002). The wet, open, almost always lowland habitats it frequents are generally flooded and deeply submerged for part of the winter months and often indeed remain so into the late spring or even the summer in our wet Oceanic (or Atlantic) climate. R. sceleratus is totally absent from soils below about pH 4.0 and it never occurs in permanently flooded aquatic sites (Grime et al. 1988).

With regard to its ecological status, R. sceleratus is always a pioneer species colonising bare mud. Among its many associates are Myosotis scorpioides (Water Forget-me-not), Rorippa palustris (Marsh Yellow-cress), Persicaria hydropiper (Water-pepper), Alisma plantago-aquatica (Water-plantain), Veronica beccabunga (Brooklime), Lemna minor (Common Duckweed), L. trisulca (Ivy-leaved Duckweed), Callitriche stagnalis (Common Water-starwort) and Bidens cernua (Nodding Bur-marigold). In terms of plant communities, R. sceleratus belongs chiefly to the NVC OV32 Myosotis scorpioides-Ranunculus sceleratus open, nitrogen-rich, often muddy and disturbed, intermittently wet ground community (Rodwell et al. 2000, page 434), an Association of the Bidention Alliance which goes under various names in different parts of W Europe (White & Doyle 1982). It does also occur however, as a regular associate in seven other aquatic and swamp communities listed in Rodwell et al. (1995).

Like most pioneer colonisers of bare ground habitats, the presence of R. sceleratus tends to be ephemeral. Where disturbance occurs more rarely or irregularly, it is gradually crowded out by the arrival of taller, more permanent colonising vegetation dominated by species such as Phragmites australis (Common Reed), Typha latifolia (Bulrush), Schoenoplectus lacustris (Common Club-rush), Equisetum fluviatile (Water Horsetail), Cicuta virosa (Cowbane) and Iris pseudacorus (Yellow Flag). The latter, together with various sedges and a collection of other invading and carpeting species, can very quickly cover and occupy previously bare mud, out-competing and excluding R. sceleratus. In addition, should the inhabited site dry out during a prolonged drought, R. sceleratus quickly succumbs. It is too fleshy and succulent to survive dry conditions for long (Grime et al. 1988).

Fermanagh occurrence

This is a rather scarce annual which had only been seen twice in Fermanagh before 1980, but since then it has been recorded at 15 new sites covering 21 tetrads, 4% of the total in the VC. As the distribution map shows, it is thinly scattered in seasonally flooded water meadows around the Upper Lough Erne basin, mainly in the south of the county, with one outlying station at Derryclawan near Enniskillen. The latter is the only station where it was found in any real quantity and here it grows on cattle poached, wet, well-dunged, anaerobic or very poorly-aerated mud on the bed of an old lake exposed in summer after a spell of dry weather.

Flowering

R. sceleratus flowers are remarkable for their outsized, elongated, pineapple-shaped receptacle, which sits quite incongruously amongst the encircling small petals and tends to dwarf them. Flowers are produced mainly from May to September, but chiefly from June to August. They attract flies and bees with freely presented nectar and they may be cross-pollinated by wind or by their winged insect visitors (Clapham et al. 1962; van der Toorn 1980). In addition, if this fails to occur, the travels of thrips and aphids crawling around individual plants enables self-pollination (Baker & Cruden 1991).

Seed production

The many seeds of the plant (ie the achenes − single seeded dry fruits), are produced at rates of between 70-100 per receptacle and up to 45,000 per plant (with a mean of 26,000, however) (Salisbury 1942). Another estimate given by van der Toorn (1980) indicated that a large plant in very good growing conditions and with little or no competition can produce up to 50,000 achenes. The individual seeds/achenes are smaller than those of R. flammula (Lesser Spearwort) and many times smaller than those of R. lingua (Greater Spearwort) (Clapham et al. 1962).

Seed dispersal

Seed dispersal involves wind and water. Seeds float for at least an hour, but generally somewhat longer, the distance travelled obviously dependent upon rate of water flow and density of waterside vegetation (van der Toorn 1980).

Germination and life-cycle/growth strategy

Celery-leaved Buttercup seed which has been stratified by winter cold for 4 to 6 months and subsequently exposed will germinate in the spring or early summer (van den Toorn & ten Hove 1982). The species can complete its life-cycle in two months (ie behaving as a summer annual). If it germinates later in the summer, for instance in August or September, it is frost resistant and may persist through the winter as a submerged dormant leaf rosette. It then recommences growth when the mud it occupies is exposed the following year − ie it behaves as a winter annual, thus giving the species a dual life strategy (Bakker 1966; van der Toorn 1980). During the second year, these winter annuals will probably seed early in the summer and, since a proportion of the fresh seed can germinate immediately, there may be sufficient time for a second generation to grow and complete their life-cycle in the same season (Bakker 1966; Grime et al. 1988).

In the first year or so after germination, R. scleratus can build up its population very quickly due to its enormous seed output and their easily achieved dispersal by wind and water.

Buried seed survival

Despite their small size, the seeds can lie dormant and survive for many years on the muddy margins of lakes, ponds and ditches, until low water levels expose the bare mud and disturbance brings them up to the light, triggering germination. Five of the 13 records quoted in the survey of NW European soil seed banks reckoned that R. sceleratus seed is long-persistent: one estimate reckoned survival is possible for over 50 years (Thompson et al. 1997).

British and Irish occurrence

In suitable muddy habitats in other parts of Northern Ireland, Celery-leaved Buttercup is much more common than it is in Fermanagh, eg in Co Down (H38) and in all the vice-counties around Lough Neagh in particular, ie Cos Tyrone, Armagh, Down, Antrim and Londonderry (H36-H40) (Harron 1986). It is also more frequently found in coastal areas both around the province and in the wider British Isles. R. sceleratus is tolerant of brackish alluvial mud conditions and therefore it is also a frequent pioneer coloniser of mud on grazed or otherwise disturbed estuarine saltmarshes (NI Flora Website 2002; R.A. Fitzgerald, in: Preston et al. 2002). Overall, however, R. sceleratus occupies a rather restricted niche habitat and it is not surprising that although it is widespread and frequent in C & SE England, the species is uncommon in many inland parts of Ireland, Wales and Scotland (Gray 1970; Stace 1997).

European and world occurrence

In Europe, R. scleratus is widespread, especially in C & W areas, becoming scattered to rare and increasingly coastal to both the north and south of its range (Jalas & Suominen 1989, Map 1828). The species extends through the Near East, Siberia, C Asia, Japan to eastern N America (where it is in fact also represented by an additional subspecies, subsp. multifidus) (Hultén 1971, Map 291). A further subspecies, subsp. reptabundus (Rupr.) Hult. is later mapped in N Europe and NW Siberia (Hultén & Fries 1986, Map 857). Hultén remarks that as the plant is so often apophytic (ie occupies man-made or strongly man-influenced habitats) it is difficult to decide exactly where R. sceleratus is native. He is particularly suspicious about its native credentials in eastern N America, and I would go further and say that it is always introduced in N America. In the southern hemisphere, Celery-leaved Buttercup is a certain introduction, for instance, in C & E Africa, New Zealand, Queensland and Tasmania (Hultén 1971; Jonsell et al. 2001).

Fossil record

Living as it does in nutrient- and nitrogen-rich muddy habitats and requiring disturbance to initiate germination, R. sceleratus is very often associated with human settlements. This naturally raises the possibility that in some places it might be an ancient introduction, ie an archeophyte. However, these muddy habitats also favour preservation of fossil pollen and achenes and this evidence clearly indicates that R. sceleratus has been continuously present in Britain and Ireland from the Pastonian stage onwards. Thus Celery-leaved Buttercup is very definitely a native species (Godwin 1975).

Toxicity

Like other buttercups R. sceleratus contains the glycoside ranunculin, which on hydrolysis breaks down to yield an irritant oily substance protoanemonin, plus glucose (Saber et al. 1968). Protoanemonin is responsible for the toxicity of all Ranunculus species, and R. sceleratus is reputed to be the most poisonous of all. It is possible, however, that because of the rich, luxuriant, somewhat succulent growth of the plant, it may be eaten in larger quantities than other buttercup species. Protoanemonin poisoning is reported most frequently in cattle, the acrid juice causing blistering of the mouth. In an experimental trial, one goat fed with R. sceleratus died and two others became severely ill. Celery-leaved Buttercup does not normally invade pastures, but a horse that grazed an area where it had access to R. acris (Meadow Buttercup) and R. sceleratus temporarily developed paralysis, convulsions and a loss of sight and hearing (Cooper & Johnson 1998).

Names

The plant's toxicity undoubtedly explains why it was given its Latin specific epithet 'sceleratus', which means 'wicked' or 'vicious' (Gilbert-Carter 1964).

Apart from the aptly descriptive English common names, 'Celery-leaved Buttercup' and 'Celery-leaved Crowfoot', the plant is also known in N America as 'Cursed Crowfoot'. Presumably farmers whose animals attempt to eat it, refer to it in this way. An alternative common name listed by Britten & Holland (1886) is 'Ache', apparently through a connection with the old French name for Parsley, and thus a connection via 'Apium' to the vegetable, Celery. A further name 'Thiretelle' originates in two dictionaries of obsolete English which refer to, "The herb apium risus", which is identified by Britten & Holland (1886) as R. sceleratus. Another name these authors mention is 'Blisterwort', which originated with Lyte (1578), and is a useful reminder that buttercup sap very readily causes burn-like blisters on skin.

Uses

In past times, beggars were said to use buttercup species commonly, and especially R. sceleratus, in order to induce sores on themselves to excite compassion and gain alms from the public (Lightfoot 1777, p. 291; Vickery 1995, p. 63). According to Mrs Grieve who reports this nefarious use of the plant, after 'working' their 'con', the beggars afterwards would cure their blisters by applying fresh Mullein leaves to the wounds (Grieve 1931, pp. 182 & 235). Funnily enough, Grieve does not mention this healing property under her entry for Mullein (Verbascum species). Poor beggars! Grieve warns that R. sceleratus, "is one of the most virulent of native plants: bruised and applied to the skin, it raises a blister and creates a sore by no means easy to heal". She goes on to indicate that if the plant is boiled and the water discarded, it can be eaten as a vegetable, and was peasant food in Wallachia (an old name for Romania). Grieve mentions a tincture, used in small doses, as an herbal cure for "a stitch in the side and neuralgic pains between the ribs".

R. sceleratus is widely used to this day in homeopathy and numerous internet websites deal with this topic. YOU ARE STRONGLY ADVISED NOT TO MAKE ANY ATTEMPT TO MAKE USE OF THIS VERY DANGEROUS CAUSTIC PLANT.

Threats

Due to the general nutrient enrichment of aquatic habitats, we may see this species increasing, provided that climatic change does not reduce the water level fluctuation that provides bare mud for colonisation.

Growth form and habitat preferences

Greater Spearwort is an erect, robust, semi-aquatic, emergent perennial up to 120 cm tall with a creeping rhizome or stolon up to 50 cm long, bearing shallow fibrous roots (Jonsell et al. 2001). As such, it is a bigger plant that shares some of the characteristics of both R. flammula (Lesser Spearwort) and R. repens (Creeping Buttercup), ie like R. flammula it is a helophyte, growing in soil frequently saturated with water or it stands in shallow water with its base submerged. Again, like R. repens, the underground stoloniferous stem branches and produces offsets, daughter plantlets or ramets, which can play an important role in vegetative reproduction, clonal local diffusion and longer-distance dispersal of the species within a lowland, wetland system (Johansson & Nilsson 1993).

R. lingua is a plant of mesotrophic to eutrophic, rather nutrient-rich, sheltered, shallow water, rich-fen and swamp vegetation. It seems to prefer places where there is a fairly gentle inflow of stream water, presumably adding nutrients. It is often rooted in organic, peaty, lime- or base-rich mineral mud, usually with a pH between 5 and 6.5 (Spence 1964; Garrard & Streeter 1983; R.A. Fitzgerald, in: Preston et al. 2002). As is the case with R. auricomus (Goldilocks Buttercup) and some other species, which in the British Isles frequent calcareous or base-rich soils, in Nordic countries R. lingua appears indifferent to lime (Jonsell et al. 2001).

Fermanagh occurrence

Prior to 1975, there were a total of just 27 records for R. lingua in Fermanagh. However, thanks to the extensive recording in the VC from 1977 onwards, which has particularly focused on lowland wetlands, there are an additional 200 post-1975 records for Greater Spearwort in the Fermanagh Flora Database. This stoloniferous perennial has now been recorded in a total of 64 tetrads, 12.1% of those in the VC.

In Fermanagh, R. lingua grows among tall, often dense, fen and marsh reed vegetation, especially around the small inter-drumlin lakes that form the fretted margins of Upper Lough Erne. It is also found around a variety of other lakelets and ponds and on the muddy banks of rivers, canals and ditches, eg in marl ponds along the River Finn, on the banks of the Swanlinbar River, the Old Ulster Canal and in gravel-pits at Gortaree. It typically associates with Phragmites australis (Common Reed), Typha latifolia (Bulrush), Carex elata (Tufted-sedge), Equisetum fluviatile (Water Horsetail), Cicuta virosa (Cowbane) and Sium latifolium (Greater Water-parsnip). It is present, but not a constant or even a very frequent species in one aquatic (A4) and six swamp plant communities in the NVC listing (S1, S4, S17, S22, S24 and S27) Rodwell et al. 1995).

Although R. lingua is stoloniferous and potentially clonal patch-forming, in Fermanagh it is always a rather sparse and local component of the type of tall waterside vegetation it frequents. R. lingua is a decidedly inconspicuous plant until about mid-June when it comes into flower. It could thus be very easily overlooked during early season field recording. Seven Fermanagh tetrads contain only pre-1976 records, possibly indicating a loss of suitable habitat, or the need for more timely recording.

Greater Spearwort is said to be intolerant of trampling and grazing (Sinker et al. 1985), and since both these pressures must occur on the grazed water meadows along most of the Fermanagh lake shores where the species occurs, these two factors might well be limiting its local occurrence.

Flowering

R. lingua begins flowering in late June and continues until September, reaching a peak in July. The rather large, creamy yellow, very glossy- petalled flowers are usually borne in a few-flowered cyme, but sometimes flowers are solitary. The flowers are protogynous (ie their female stigmas ripen before their pollen) and they attract insect pollinators (mainly flies) by producing copious nectar (Clapham et al. 1962; Hutchinson 1972). Broad-leaved forms of R. flammula can easily be mistaken for R. lingua, but they have much smaller flowers, about half the diameter of those of Greater Spearwort.

Fruiting, overwintering and vegetative reproduction

R. lingua is described as 'thermophilous' by the Dutch botanists van der Voo & Westhoff (1961), meaning that it prefers relatively warm temperatures, or it requires such conditions to really thrive (Holmes 1979). Further north in Sweden, fruit-set is often poor and recruitment from seed is regarded as very infrequent. It is possible that from time-to-time in a poor summer fruiting may also be poor in parts of Britain and Ireland. I have not located any measurements or estimates of the reproductive seed capacity for this species in either Britain or Ireland and, clearly, there is need for further study in our latitudes. A Swedish study at around 60^oN reported that individual plants (ie ramets) of R. lingua live for just one year, and that propagation occurs by means of overwintering rhizomes. These overwintering organs are up to 10 mm thick and 25-75 cm long and consist of 5-10 nodes. They are produced in late summer from basal stem nodes lying just below the sediment surface (Johansson & Nilsson 1993). This system of growth and perennation results in clones of physiologically independent ramets or daughter plantlets. It was observed that each established ramet went on to produce one or two daughter ramets each year of the study (Johansson & Nilsson 1993).

Dispersal

The overwintering rhizomes or stolons are very efficient water-borne dispersal units, being perfectly buoyant for long periods due to their having hollow internodes. In comparison, seed only floats for one to two hours, which must severely limit their efficiency as water-borne dispersal units (Romell 1938, cited in Johansson & Nilsson 1993). There is, however, a high mortality of dispersed ramets during the first year after dispersal, so that successful establishment, even from organs as large as rhizomes, must be rare events. In their study the Swedes concluded that (at least in their area) R. lingua is a "pseudo-annual clonal plant", and that annual clonal disintegration (of the individual plant) can be viewed as a form of (ecological and biological) risk-spreading (current author's inserted brackets) (Johansson & Nilsson 1993). The probability of extinction decreases because some ramets are always devoted to dispersal to new sites. "In reality, however, only a few of all rhizomes are dispersed, and this can also be interpreted as a safeguard against local population extinction." (Johansson & Nilsson 1993).

Although seeds in many wetland species are ineffective for long-distance dispersal within water systems, their smaller size and longer life must still allow them to be the most effective means of dispersal between water systems by transporting agents such as birds and other animals (Smits et al. 1989). Only one reference is given in the comprehensive survey of soil seed banks of NW Europe (Thompson et al. 1997), and it suggests that seed of R. lingua is transient, surviving less than one year.

Changes in distribution

As mentioned above, R. lingua has been in decline for perhaps as long as 200 years (Harron 1986; Hackney et al. 1992), a situation demonstrated for the British Isles in the 1962 and 1976 Atlases (Perring & Walters 1962, 1976). This decrease in range is also described for the Nordic countries by Jonsell et al. (2001), where the plant has retreated south of the Arctic Circle (see their map, p. 277). Although changing temperatures may have had some effect on the sexual reproductive capacity of the species further north, it is much more probable that the main factor causing losses in Britain and Ireland was land drainage, since R. lingua is not very sensitive to either water pollution or eutrophication (nutrient enrichment) (Jonsell et al. 2001). However, at least in Britain, if not elsewhere, previous losses have been reversed, as is clearly illustrated by the maps published in the New Atlas (Preston et al. 2002).

Introduction to gardens

During the last 40 years or so, Greater Spearwort has gained appreciation from horticulturalists and is now considered a sufficiently decorative, appropriate and easily enough cultivated subject for planting around the fashionable, indeed almost obligatory garden 'water feature'. Thanks to this trend, there have been so many 'escapes' and deliberate introductions of the plant to the wild in Britain, that Fitzgerald commented, "the distinction between native and alien populations is now hopelessly blurred" (R.A. Fitzgerald, in: Preston et al. 2002).

Irish occurrence

The New Atlas shows that R. lingua is widely scattered and locally frequent in Ireland, but the main areas of concentration are undoubtedly in the C & NE of the island. However, while it may remain scattered in NE Ireland, the area around Upper Lough Erne now appears to constitute the main stronghold of this rather scarce and local emergent aquatic species in the north of Ireland. Harron (1986) described Greater Spearwort as being widespread but very sparingly distributed around Lough Neagh and he considered it was probably decreasing there. Hackney likewise regarded it as rather rare in the three counties in the FNEI 3.

British occurrence

Due to the increased use of the plant in gardens in recent decades, it is almost impossible to distinguish many native populations from introduced populations of R. lingua in S & C England in the New Atlas map (Preston et al. 2002). However, if we mentally subtract the obviously alien concentrations of R. lingua around the major conurbations, the encouraging impression remains discernible, that the species has at least survived in those native areas where it appeared in the earlier BSBI Atlas (Perring & Walters 1976). The better recording coverage in some areas also helps offset some definite losses.

European occurrence

In Europe, the species is widely represented in W & C areas, declining to both N & S, and only occasional and very scattered throughout the Mediterranean region. There have also been significant extinctions in Belgium, N and W France, S Germany and Hungary (Jalas & Suominen 1989, Map 1861).

World occurrence

Beyond Europe, Greater Spearwort ranges from SW Asian Russia (to the Altai), the Caucasus and Kashmir in W Asia. It has been introduced in New Zealand (Hultén & Fries 1986, Map 868). There is very little variation within the species and none of taxonomic importance (Jonsell et al. 2001).

Toxicity

Like other buttercup species R. lingua contains bitter-tasting toxins which undoubtedly help deter cattle and other stock, although it is not specifically reported as being responsible for poisoning any such animals (Cooper & Johnson 1998).

Names

The Latin specific epithet 'lingua' means 'tongue' and refers, quite aptly in this instance, to the shape of the stem-leaves (Gilbert-Carter 1964). The English common names 'Spear-wort' and 'Spear Crowfoot' were applied from the 14th century, again on the basis of the leaf shape, to both the more common R. flammula (Lesser Spearwort) and to R. lingua (Grigson 1974). Thus to distinguish them in modern times, they have been called 'Lesser' for the smaller, narrower-leaved R. flammula and 'Greater' for the larger of the two. The only additional name is 'Sparrow-weed', which Britten & Holland (1886) list only from Co Londonderry (H40).

Threats

Hyper-eutrophication of its habitat, or destruction of the vegetation surrounding the lakes where the plant grows.

Growth form and preferred habitats

This is a very variable, abundant and widespread perennial of soft, wet, marshy ground found mainly around lakes and ponds and in seasonally flooded water-meadows. R. flammula is the type of plant described in Raunkiaer's Life Form Classification as a 'helophyte', a term derived from Greek literally meaning 'a marsh plant'. The term refers to a perennial with an underground storage organ which grows in soil saturated with water and which, therefore, has submerged winter buds and is thus typically shallow rooted like R. flammula (Raunkiaer 1934; Holmes 1979; Grime et al. 1988).

R. flammula occurs in situations where the dense shading effect of tall fen vegetation is excluded or kept minimal by a range of ecological factors, perhaps including nutrient limitation, but more typically involving a combination of seasonal flooding and occasional spates, plus grazing, trampling, cutting or other forms of disturbance, including that generated by popular water-based human leisure activities. It is also very frequent or almost constant in hollows in damp grassland, by flowing water in streams, ditches, springs and in flushes on moors, bogs and upland woods, eg in Fermanagh's wooded Correl Glen NR.

In Britain, R. flammula is mainly associated with moderately oligotrophic to mesotrophic waters (Preston & Croft 1997), but in Fermanagh it also features around or near some of our more decidedly eutrophic waterbodies and it is particularly frequent around Upper Lough Erne. In coastal regions of Britain and Ireland, R. flammula additionally occurs in dune slacks and on damp sea cliffs (Grime et al. 1988).

Under terrestrial habitat conditions R. flammula typically grows erect, but it can also sprawl horizontally to some extent (ie it can be decumbent). However, when the base of the plant becomes submerged in water the habit often becomes creeping. Under these circumstances, it then roots at the nodes, ie it becomes stoloniferous, and given relatively open substrate conditions to colonise, such as bare mud or disturbed soil it may branch and spread to form a clonal patch (Cook & Johnson 1968; Grime et al. 1988). As with other waterside species, this feature undoubtedly plays an important part in the reproduction of R. flammula, since detached portions disperse readily in flowing water.

Preferred soil types

In terms of substrate, Lesser Spearwort generally occurs on wet, moderately acidic, peaty mud, stony gravel or mineral soils, but in Fermanagh it is also very common in limestone and marl situations, eg around Lower Lough Erne and along the pools in the River Finn. This range departs from the British Isles norm to some extent, since Hill et al. (1999) gave it an indicator value for soil reaction of '5', meaning, "of moderately acid soils, only occasionally found on very acid or on neutral to basic soils". The original soil reaction value associated with the species by Ellenberg (1988), based on his experience in continental Europe, was '3', which is significantly (or strongly) inclined towards the acid end of the nine-point scale used in his soil classification. In the Sheffield region of England, Grime et al. (1988) found R. flammula occurred, "mainly on mildly acidic soils between pH 4.5 and 6.5, but extending locally to soils of pH up to 7.5". The latter represents a similar range to that noticed in Fermanagh. In contrast to our local experience, however, they went on to comment, "Rare on calcareous soils." Perhaps the interesting thing is that the species DID also occur on highly calcareous soils, as it very definitely does to a greater extent in Fermanagh. Further north in Scandinavia, R. flammula is regarded as indifferent to lime (Jonsell et al. 2001).

Fermanagh occurrence

Lesser Spearwort is very widespread and abundant in Fermanagh and has been recorded in 438 tetrads, 83% of those in the VC, a situation very much expected in an area of the country with so many lakes and a large expanse of marshy ground, fens, bogs and ditches. Although obviously very widespread throughout the county, it is especially frequent in the area lying south of Lough Erne.

British and Irish occurrence

R. flammula is also frequent and widespread throughout the whole of the British Isles, but has declined to some extent in SE England, presumably due to the usual factors of pressure for land, development, drainage and intensive agriculture (R.A. Fitzgerald, in: Preston et al. 2002).

Is R. flammula really an aquatic plant?

The definition of what constitutes an aquatic plant is not an easy matter to decide, since the boundary between land and water fluctuates over several time scales. Lesser Spearwort is not really a true 'aquatic species' in the ecological sense that fits all the related Water Crowfoots, including the Ranunculus subgenus Batrachium species with the exception of R. hederaceus (Ivy-leaved Crowfoot) and R. omiophyllus (Round-leaved Crowfoot), ie which in Professor Cook's view is limited to "species with a submerged phase during their generative history" (Cook 1966). This very strong delimitation of an aquatic plant makes the submerged phase absolutely obligatory. However, like the two previously mentioned water-crowfoot species, R. flammula is primarily a terrestrial wetland species that very commonly is found standing emergent in water. It may be temporarily submerged, either regularly, or only occasionally whenever local water table levels are higher than normal. It is NOT quite patently a species that, "characteristically grows in water which persists throughout the year", which is the definition of an 'aquatic plant' provided by Preston & Croft (1997). Despite the failure of R. flammula to comply with their definition, the latter authors included a species account of it in their excellent book, Aquatic plants in Britain and Ireland. At the same time Preston & Croft (1997) omitted other native species characteristic of tall-herb fen, which in Fermanagh, and other parts of Ireland, are capable of forming partially emergent floating mats in lakes and rivers. Many of these ignored or neglected species appear equally deserving of treatment as emergent aquatics, eg Cicuta virosa (Cowbane) and Sium latifolium (Great Water-parsnip) (Cook 1998). As with all published selections of plant species, considerations of time, space and cost undoubtedly intrude and they can determine the resulting quorum.

Variation

Living in a fluctuating environment on or near the boundary between land and water, R. flammula shows a high degree of phenotypic variation in form with respect to changes in its environments (ie morphological plasticity). This affects a wide range of characters including size, habit, leaf shape, flower size and even achene shape (Jonsell et al. 2001). Leaf development is particularly variable and plastic in form, to the extent that when the leaf develops under aerial conditions it is lanceolate, whereas when it develops under water, it is linear. This is also a case where heterophylly is of a reversible kind, for although change in shape of an individual leaf is impossible, an extending shoot can produce first one leaf form and then the other, in response to changing environmental growing conditions. It is important to distinguish this kind of heterophylly from that more often met, which is associated with maturation of the plant or entering a flowering phase, where the change in leaf form follows an irreversible one way sequence from youth to maturity (Cook & Johnson 1968).

Three subspecies of R. flammula are recognised by Stace (New Flora of the British Isles, 1997), of which subsp. flammula is the common and widespread form, while the other two, subsp. minimus (A. Benn.) Padmore and subsp. scoticus (E.S. Marshall) A.R. Clapham, are very much more rare or under-recorded. We have not attempted to distinguish the subspecies in Fermanagh, although a few old records of subsp. scoticus exist in the Fermanagh Flora Database (see separate account below).

Possible identification problems

Large specimens of R. flammula, sometimes distinguished as var. ovatus Pers. (= var. major Schult.), can easily enough be mistaken for R. lingua (Greater Spearwort) (Padmore 1957; Preston & Croft 1997). However, in Fermanagh the latter species is almost invariably found growing around Upper Lough Erne and it generally occurs within the taller lakeshore vegetation that R. flammula eschews. Despite the above mentioned phenotypic variation that occurs in R. flammula, after considerable field experience, we find that the small lanceolate leaves of non-flowering specimens are quite distinctive and recognisable.

The more procumbent phenotypes of R. flammula are also sometimes confused with its very rare relative R. reptans (Creeping Buttercup), which is confined to two sites in Scotland and a few, perhaps transient sites in Cumbria (Padmore 1957; Gibbs & Gornall 1976). An equally rare hybrid also occurs between these two species (Gornall 1987; R.A. Fitzgerald & C.D. Preston, in: Preston et al. 2002). The hybrid, R. × levenensis (R. flammula × R. reptans), has been twice recorded in Northern Ireland, from Lough Fea in Co Londonderry (VC H40), near Lough Neagh, which happens to be a major arrival site for waterfowl, which are presumed to transport R. reptans (Gornall 1987).

Flowering

R. flammula flowers abundantly from June to August. It attracts insect visitors and being largely (but not absolutely) self-incompatible, it mainly carries out cross-pollination (Cook & Johnson 1968; Gibbs & Gornall 1976). The flowers are described by Jonsell et al. (2001) as, "weakly protandrous" (ie the pollen matures first) and these Nordic authors also report that selfing is possible when pollination is carried out by raindrops falling into the flower bowl.

Seed production, germination and dormant survival

Published estimates of viable seed production by R. flammula are unknown to the current author, but plants are known to produce between 0-20 flowers, each containing up to around 20 achenes (ie single-seeded dry fruits), thus potentially each plant may produce up to 400 seeds. In a Canadian study, freshly collected seed kept moist germinated sporadically over a period of six months (Cook & Johnson 1968). In a survey of seed bank data in NW Europe, Thompson et al. (1997) reported 31 estimates for R. flammula, of which six regarded it as ephemeral, eleven short-term, and seven reckoned it produced a long-term seed bank (ie capable of surviving longer than 5 years).

Seedlings produce a rosette of leaves and if submerged they will also produce spreading stolons. In Canadian populations, flowering only occurred after the plant had been exposed to terrestrial conditions for an unspecified period of time (Cook & Johnson 1968).

European occurrence

R. flammula subsp. flammula is widespread in Europe except in both the southernmost and northernmost areas. It is absent from Iceland, and occurs at isolated stations only in Greece, Turkey, the Caucasus and W Siberia (Jalas & Suominen 1989, Map 1856; Jonsell et al. 2001). It extends south as far as N Africa and eastward to W Asia (Hultén 1958, Map 147). R. flammula is rare and probably accidentally introduced on the E and W coasts of N America (Jonsell et al. 2001) and it is a definite alien in New Zealand (Preston & Croft 1997).

Medicinal uses

R. flammula was another buttercup used in herbal medicine as a rubifacient for blistering, that is, for raising blisters. An ancient belief was that by irritating the skin and raising a blister, disease would be drawn out of the body. It was extensively used for this purpose during the bubonic plague in the 16th century, and was also used to treat scabs and running sores (Ui Chonchubair & Mhic Daeid 1995).

Threats

None.

Threats

None.

Growth form and habitat preferences

While R. ficaria is very commonly recorded in over half the area of Fermanagh (see below), we still regard this familiar yellow-flowered, patch- or carpet-forming tuberous perennial as under-recorded to some extent. The plant is a geophyte, having its resting or over-wintering buds protected by shallow burial in the soil. The reason we consider R. ficaria may be under-recorded is because the plant is pre-vernal, completing its annual growth and reproductive cycle early in the year and then quickly dying down and becoming inconspicuous, usually by the middle of June or earlier in some sites (Grime et al. 1988). Dead leaves, however, can still be found and the species recognised in suitable habitats later in the season and there is no month in which we do not have at least a few records of Lesser Celandine in the flora database. The short growing season of the shoots characterises R. ficaria as an ephemeroid perennial, the shoots disappearing before the summer (Rogers 1982).

R. ficaria occurs in a wide range of habitats and light environments, from deciduous mixed woodland where it is most frequent (except in those on the most acid soils), to hedges, roadside banks, verges, stream-sides and damp pastures. It is also found in gardens and can become a persistent weed of old lawns and flower beds (Taylor & Markham 1978; Sinker et al. 1985). Rapid growth very early in the season (ie its prevernal habit) enables the species to avoid the shade and competition of other plants, particularly in deciduous woodland but also to some extent in its other habitats.

Lesser Celandine tolerates a wide range of soils in these islands, both chemical and physical and, indeed, it appears to show a wider amplitude in this respect in Britain and Ireland when compared with its behaviour in Continental Europe (Hill et al. 1999). Measurements of soil reaction by Taylor & Markham (1978) showed the pH of the substrates occupied by the species ranged from 4.4-6.9 in the British Isles and that it frequently occurs on seasonally wet or flooded situations. It does not grow on very acidic mull soil (pH 3.9 or lower) and is absent from both permanently waterlogged or regularly droughted dry soils.

The small tubers or bulbils and their roots are buried only shallowly in the top 5-10 cm of soil and litter so that R. ficaria is confined to soils that are moist in spring, but which may possibly dry out later in the summer months. This dehydration does not matter as far as the plant is concerned, since by the time moisture stress occurs in the substrate, it is likely to be in its resting phase and therefore completely resistant to drought.

Competitive ability

Given adequate spring moisture, Lesser Celandine is most frequent where soil and vegetation is average with respect to fertility, productivity, level of disturbance and extent of bare uncolonised soil surface. Although it is a poor competitor in well illuminated sites, R. ficaria plants grow and perform better in ecologically open conditions in terms of producing a larger than average shoot, bigger individual tubers in the soil and more of the plants produce flowers and fruit (Taylor & Markham 1978).

The flora associated with Lesser Celandine typically has a relatively low level of diversity and this is particularly so in damp mixed woodland. Here, under the shade and protection of tree trunks, bare branches and the occasional evergreen or wintergreen leaf and frond, as the population of R. ficaria increases and begins to form extensive patches, species diversity declines further as it begins to oust other herbs. Thus under damp, shaded growing conditions, Lesser Celandine can sometimes form a more or less dominant, single-species carpet of ground vegetation (Taylor & Markham 1978; Grime et al. 1988).

Tolerance of disturbance

Plants can survive grazing and trampling by farm stock including horses, and mowing on grassy paths, roadside verges, and to a lesser extent in lawns. Due to the early season nature of its active period, annual growth and reproduction of R. ficaria plants is often near completion before these sorts of ecological disturbance reach any great level of intensity (Grime et al. 1988).

Phenology

Growth of R. ficaria is initiated by low temperatures in the autumn, and buds on the buried tubers begin to develop in December. The long-stalked, deeply notched, cordate foliage leaves of the basal rosette are very variable in shape. They are also shiny, hairless and often blotched with radiating pale or dark markings. The leaves begin to appear above ground in mid-January and they are well deployed by early February. Seedlings of the diploid, subsp. ficaria, require at least two year’s growth before they are capable of flowering (Marsden-Jones 1935).

Flowering

On established plants the flowering stem develops in February: it varies from 3-20 cm long and is spreading and weakly decumbent, branched and typically it bears one or two opposite leaves, similar to the basal ones but generally more distinctly lobed. Solitary flowers are borne at the end of each branch and flowering reaches a peak in late March and continues into April. The flowers are 20-30 mm in diameter and appear in a range of yellow shades (Grime et al. 1988, 2007).

Compared with other buttercup species, individual flowers of R. ficaria vary much more widely in the number of their floral parts, including the sexual organs. Early in the season many more stamens and carpels are produced per flower than in the ones opening later (Lee 1902). Thus the flower is atypical for the genus Ranunculus, usually having just three sepals and anything between seven and 13 petals instead of the more normal five of each organ characteristic of the genus (Sell 1994). Together with the apparently opposite and unequal stem leaves, these floral part numbers have led some taxonomists to separate the species off into its own genus and name it Ficaria verna Huds. (Hutchinson 1972).

There is not any concrete evidence suggesting that more open habitats favour flower and seed production by individual plants, which is the particular strategy of subsp. ficaria, the diploid (but see below). Higher light levels do allow more diploid plants to flower, however, and as would be expected, very heavy shade prevents any flowers being initiated. The apparent preference of R. ficaria s.l. for damp habitats may well be related to the conditions required to maximise the formation and survival of the perennating root tubers which are common to all ploidy levels (Nicholson 1983).

Pollination

However we decide to name this species, the flower structure is completely open and available to all insect visitors and the presence of nectar attracts short-tongued bees, honey bees, small beetles and flies whose visits achieve pollination. In wet cold weather the flowers remain closed and no insects visit. The amount of sun per day is a significant factor affecting early season insect-pollinated flowers, since whenever the sun becomes obscured there is always a sudden drop in temperature and a consequent great falling off in the number of insect flower visits (Marsden-Jones 1935). In the absence of visitors, self-pollination can occur, but the number of seed set is low in such circumstances, some flowers proving completely self-sterile and others producing very much reduced numbers of viable seed (Marsden-Jones 1935).

Fruit

The fruit of Lesser Celandine is a cluster of single-seeded achenes that are shed from the receptacle of the flower by early June. The number of achenes produced per flower varies with the fertility and suitability of habitat growing conditions (see below). Between 15 and 35 achenes may form on the receptacle of the flower, although often many of these potential fruit fail to develop or simply abort (Taylor & Markham 1978; Jonsell et al. 2001). Shortly after the achene clusters on the stem tips ripen and break up, the aerial parts of the plant wilt and quickly decay.

Genetic variation − two subspecies

It has been realised since the mid-1930s that Ranunculus ficaria occurs in the British Isles as two races with different chromosome numbers, a diploid, subsp. ficaria L. (2n=2x=16) and an almost sexually sterile tetraploid, subsp. bulbilifer Lambinon (2n=4x=32) (Marsden-Jones 1935). However, there is a problem distinguishing the diploid and tetraploid forms in the field, since whenever the species is flowering and is at its most conspicuous, the main character (the only field one), which distinguishes the two subspecies − the aerial bulbils of subsp. bulbilifer − have not yet begun to develop (Gill et al. 1972). In the 2nd edition of Stace's New Flora of the British Isles, additional vegetative distinguishing characters of a critical nature are provided, but this involves microscopic examination to count the chloroplasts in leaf pore guard cells (stomatal guard cells).

Vegetative reproduction − two forms of bulbil

The diploid form of R. ficaria is fully fertile and generally sets seed, but unlike other situations involving polyploidy where complete fertility attaches to even numbered chromosome sets, for some reason in this species the tetraploid form has a very low level of sexual fertility. By way of reproductive compensation, as it were, and undoubtedly encouraging or enabling their survival, the tetraploid plants possess aerial bulbils in addition to the subterranean vegetative bulbils that are typical of the species. Aerial bulbils are not produced at all on the diploid plant (Nicholson 1983). The small white aerial tubers are produced in the angle between a stem leaf and the shoot bearing it and should not be confused with the white or buff, fig- or spindle-shaped tubers which are produced just below ground level around the basal leaf rosette and which sometimes become exposed if the soil around the plant becomes disturbed. Basal tubers occur in all forms of Lesser Celandine plant irrespective of chromosome number and are both the perennating organ and a primary means of population increase and dispersal in many situations. The aerial bulbils are derived from swollen axillary adventitious roots exactly as the basal tubers are, so the structures are distinguishable really only by their position on the plant (Gill et al. 1972). It is clear from where they are produced on the plant that both types of bulbil originate as modified buds.

Reproductive strategy

The reproductive strategy demonstrated by R. ficaria s.l. is very common in, or even characteristic of, Arctic environments where the frequency of polyploid species is much higher than is found in the floras of temperate regions. It is estimated that 51% of the British and Irish flora is of polyploid origin, whereas the comparable figure for Iceland is 72% (Löve & Löve 1974). In Arctic regions, the season of growth in some years is sometimes too short for plants to complete their sexual processes and form viable seed. In such circumstances, natural selection favours the survival of species with a fall-back option of well-developed vegetative reproduction. The growing season of vernal herbs is also brief in woodland and other deciduous seasonally shaded habitats, growth in pre-vernal and vernal herbs being confined to the light phase of the year, from autumn leaf fall to the expansion of the new leaf canopy the following spring. During at least part of the long dormant or overwinter period, low temperatures obviously restricts or prevents active plant processes (Taylor & Markham 1978). When temperatures again increase, the active growth phase of the pre-vernal plant is so early in the year, changeable weather conditions make activities like pollination uncertain and some processes, including seed set, may actually fail. Therefore species such as R. ficaria, Hyacinthoides non-scripta (Bluebell) and, in Great Britain (but not in Ireland), Mercurialis perennis (Dog's Mercury), rely very heavily on vegetative reproduction for their increase and survival.

R. ficaria tetraploids have a greatly reduced seed output, but as discussed above, they have two methods of vegetative propagation − aerial bulbils and basal tubers. In seasonally shaded, relatively undisturbed habitats where opportunity for seed dispersal is poor, this may well prove the better arrangement in terms of both plant survival and spread into new ground.

Fermanagh occurrence

In Fermanagh, R. ficaria s.l. is commonly recorded in 297 tetrads, 56.3% of those in the VC. As previously mentioned, it is probably under-recorded to an unknown extent since its growth is early season, pre-vernal and the plant dies down completely by June. The habit of distinguishing these two subspecies only gained ground slowly in Britain and Ireland and it is regretted that they have not been distinguished in any of the Fermanagh fieldwork to date.

British and Irish occurrence of the two subspecies

The distribution of subsp. ficaria is described by Stace as occurring, "throughout the British Isles", and for subsp. bulbilifer, "almost throughout, but apparently absent from Shetland and the Channel Isles" (Stace 1997). The British and Irish occurrence contrasts strongly with that in Nordic regions, where the tetraploid is the common and widespread subspecies and the diploid is very rare and confined to Denmark and the southern tip of Norway (Jonsell et al. 2001).

In the 6th edition of An Irish Flora (the book relied on for most of the current Fermanagh survey by Irish recorders), Webb (1977) indicated that a variant (ie subsp. bulbilifer) existed and was, "occasionally found, mainly in the E, and usually as a garden weed". This impression is partially confirmed by the Flora of County Dublin, where subsp. ficaria was described as occurring, "very common in woods and hedges, and very rare in gardens". In contrast, the same work described subsp. bulbilifer as being, "very common in gardens, disturbed roadsides, and also in woods and hedges" (Doogue et al. 1998).

Triploid forms and aerial bulbils

Where the diploid and tetraploid overlap,

triploids result from their crossing (2n=3x=24) (Gill et al. 1972). Triploid plants may flower but they are totally sterile. They sometimes produce aerial bulbils, but these are smaller and fewer than those of the subsp. bulbilifer tetraploid plants.

Comparison of subspecies with respect to light and shade

While R. ficaria s.l. is most commonly found in summer shaded habitats such as woods, scrub and hedgerows, as with the Wood Anemone (Anemone nemorosa) and the Bluebell (Hyacinthoides non-scripta), in the west of Britain and Ireland it also grows in open situations in meadows and rough pastures. Clapham et al. (1962) suggested that the diploid, subsp. ficaria, is the more commonly found form in the British Isles, especially in sunny sites and that the tetraploid, subsp. bulbilifer, is more restricted to shady places. In Yorkshire, Nicholson (1983) found that while the diploid occurred under the full range of light levels, the tetraploid was associated with moderate to heavy shade and was absent from high ground, including the Wolds and the cliffs at Flamborough Head, and it was also absent from low sandhills in the coastal region. This is a rather surprising claim since these areas appear to represent more physically demanding environments, and by analogy with the pattern of arctic plants where higher degrees of ploidy and associated vegetative apomixis occur in the more severe growing conditions. One would therefore expect the tetraploid form with its additional means of vegetative reproduction to be favoured, rather than excluded from the more demanding environments.

Field observations and experimental measurements both suggest that subsp. ficaria is the more light tolerant form of the two, rather than suggesting subsp. bulbilifer is more shade tolerant (Nicholson 1983). The observation that R. ficaria is able to persist in more open meadow conditions in the N & W of the British Isles is thus more probably due to the prevalent moist, cool conditions in these regions, rather than suggesting any requirement the subspecies has for shade.

There is not any concrete evidence suggesting that more open habitats favour seed production by individual plants, which is the particular reproductive strategy of the diploid subsp. ficaria (but see below). However, higher light levels do allow more diploid plants to flower and very heavy shade will prevent flowers from being initiated. The apparent habitat preference of R. ficaria s.l. for damp ground may well be related to the conditions required to maximise the formation and survival of the perennating root tubers, which are common to all ploidy levels of the plant (Nicholson 1983).

Comparative seed production

In measurements of sexual reproductive performance, the mean seed output computed for sun and shade diploid plants was very similar (75 and 71 per plant), giving an overall mean of 73 ± 5.8. (Marsden-Jones 1935). Flowers on tetraploid plants in open conditions tended to be produced earlier than those on the diploids, but even if they become pollinated, only a low proportion of their carpels set viable seed. Marsden-Jones (1935) collected a total of just 23 apparently viable achenes from ten tetraploid plants, a figure representing around 2% of the ovules produced by these plants. A large proportion of the pollen of tetraploids is also non-viable, unlike that of the diploid form (Taylor & Markham 1978).

Seed dispersal

After fertilization in late May and early June, the head of achenes ripens and dries somewhat so that it breaks up at the slightest touch. By this time, or soon after, the other aerial parts of the plant have already begun to wilt and decay. There does not appear to be any special dispersal mechanism for the seed and presumably it does not travel more than a few centimetres. The fruiting stems do elongate, however, and become more horizontal as they age which helps dispersal, and some seed might be carried further by rain wash, especially over bare compacted soil or along tracks and paths.

Seed dormancy and germination

Seeds of R. ficaria require a period of after-ripening as the embryo is only partly differentiated when they are shed. A delay of four to six months and a period of low temperature is necessary to break dormancy. Germination begins in the following spring and continues until early summer. Germination rates are poor for both subspecies, ranging from 15-46% for the diploid and from 7-29% for the much rarer seeds of the tetraploid (Taylor & Markham 1978).

Seedlings have only one cotyledon

Seedlings of R. ficaria, like that of at least some populations of Conopodium majus (Pignut) are highly unusual for a dicotyledonous plant group, in having only one embryo leaf or cotyledon and not two (Taylor & Markham 1978). This fact has been known since the 1850s and was the subject of much speculation among botanists for many years. The seedling's typically bilobed blade and the venation in the single cotyledon suggested to some botanists that it might have arisen by the fusion of two cotyledons. However, in a beautiful piece of anatomical and embryological investigation carried out at Kew Gardens, Metcalfe (1936) clearly and carefully argued and conclusively proved that the cotyledon is in fact a single foliar organ brought about by the suppression of one of the cotyledons during embryo development, and not the product of two embryo leaves becoming fused together. Part of the evidence for this is the rare occurrence of R. ficaria seedlings with two bilobed cotyledons, and trilobed single cotyledons have also been observed. Metcalfe's discovery involved recognising the existence of a rudimentary second cotyledon which normally fails to develop and thus seedlings only produce a solitary embryo leaf or cotyledon.

A comparison might be made here with the genus Cyclamen (Primulaceae) members of which also have only one cotyledon, the other aborting during early development. Another case in point is the genus Peperomia (Piperaceae), where some species are heterocotyledous: one cotyledon being a green, aerial, assimilating organ, while the other is retained inside the seed when it germinates and serves primarily for the absorption of food reserves from the seed (Metcalfe 1936). Other examples of this type of odd phenomenon include some species in the genus Claytonia including C. virginica (Stebbins 1974; Mabberley 1997). [N.B. This very interesting paper by Metcalfe has been incorrectly attributed to Marsden-Jones and given the wrong date by two recent authors, one of whom also got the volume number wrong! It is correctly quoted, attributed and cited here.]

Tuber and bulbil dispersal

All clones of R. ficaria irrespective of chromosome number possess root tubers (around twelve per cluster) and the tetraploid subsp. bulbilifer also has its aerial axillary bulbils. Fragmentation of the basal tuber cluster by soil disturbance is a very efficient means of vegetative propagation. In addition, the tetraploid produces up to 24 bulbils per plant that separate off and drop to the ground as the shoot bearing them dies down. A heavy shower of rain will sometimes wash these aerial bulbils away from the parent plant, heaping them together on the sides of run-off channels when the rain ceases. In such places, the quantity of bulbils aggregated is often so noticeable that the idea arose that they had fallen from heaven with the rain, and the myth developed of 'potato rain', or a 'rain of wheat' (Grieve 1931; Taylor & Markham 1978).

Tubers and bulbils provide the most rapid means of increase and spread for the species (and subspecies), as they are readily detached and may be spread by disturbance of any kind, but especially by digging, ploughing or mowing. Bulbils being considerably larger than seed, can regenerate the plant quicker and some may even flower in their first season of growth (Marsden-Jones 1935).

Origin of polyploidy in R. ficaria: In view of the quite different ecological behaviour and reproductive strategies of the diploid and tetraploid forms, it appears very likely that subsp. bulbilifer is an ancient autotetraploid, ie it is the result of mutant chromosome doubling, followed by a long period of divergence of the two ploidy levels (Nicholson 1983). Genetic isolation barriers have not yet been established, however, since in Britain and Ireland (though apparently not in Nordic areas (Jonsell et al. 2001)), triploids are formed by interbreeding where the two subspecies overlap (Gill et al. 1972). Thus Lesser Celandine is a very good example of species evolution in action, and indeed in other parts of Europe several additional subspecies are recognised (Sell 1994; Jonsell et al. 2001).

Toxicity

R. ficaria plants contain low levels of the toxin irritant protoanemonin, an unstable compound derived from the glycoside ranunculin. The concentration of this toxin increases during growth and it is at its highest during the flowering process. Unlike many other buttercups it has not been known to poison grazing stock animals (Cooper & Johnson 1998). Despite the presence of the toxin, young leaves of R. ficaria often have holes eaten in them in early February, and we have seen entire blades removed at this time of year, presumably eaten by slugs before the protoanemonin levels become a functional deterrent.

European occurrence

External to the British Isles, the diploid R. ficaria subsp. ficaria is restricted to W Europe, north to SW Norway and Denmark, where it is extremely rare, and south to The Iberian peninsula and the W Mediterranean region (Jalas & Suominen 1989, Map 1836). The bulbil-bearing tetraploid, R. ficaria subsp. bulbilifer is also confined to Europe, but it is very much more widespread in C and S parts of the continent than the diploid form (Jalas & Suominen 1989, Map 1835; Taylor & Markham 1978).

Fossil record

There is no fossil record of R. ficaria s.l. as the pollen and possible macrofossils are indistinguishable from other Ranunculus species (Taylor & Markham 1978).

Names

The specific epithet 'ficaria' is derived from the Latin name 'Ficus', the fig, and it means 'small fig'. This is a reference to the supposed fig-like shape of the root tubers (Gledhill 1985).

Two of the numerous English common names are 'Figwort' and 'Pilewort', 'fig' and 'pile' being alternative names for haemorrhoids, which the root tubers resembled. The similarity in appearance led herbalists by the 'Doctrine of Signs' to believe that a plant organ that looked like a haemorrhoid could be used to cure the complaint (Grieve 1931; Grigson 1987). Certainly the herb contains an astringent, but while Vickery (1995) includes a folklore recipe for a skin cleanser, it should be left well alone since the sap can cause blisters on or in the body (Grieve 1931).

Other English common names such as 'Brighteye', 'Butter and Cheese', 'Golden Stars' and 'Goldy Knog', refer to the shiny, yellow petal colour. Grigson (1987) also suggests the root tubers were regarded as reminiscent of a cow's udders, and hence the milk and butter references in some common names. An early botanical name for the species was Chelidonium minus, which translates as 'Lesser Celandine'. This erroneously linked the plant to the unrelated Chelidonium majus (Greater Celandine). The name 'Chelidonium' was derived from the Greek 'chelidon', a swallow (the bird), supposedly because of coincidence between the time of the plant flowering and the arrival of the migrant bird.

Threats

None.

Fermanagh occurrence

In Fermanagh, Ivy-leaved Crowfoot has been recorded in 89 tetrads, 16.9% of those in the VC although, as the distribution map indicates, it is widely scattered throughout the county; ten of the tetrads contain pre-1976 records only, which could be argued as indicating some degree of local decline during the past half-century.

Growth form and habitat preferences

R. hederaceus regularly occurs in Fermanagh as something of an invading, ephemeral, companion species of a number of quite different submerged floating or emergent vegetation communities. The habitats where this happens include the muddy banks of slow-flowing rivers, streams (together with springs and upland flushes) and drainage ditches (especially in late autumn after ditch cleaning) and beside relatively still waters in smaller lakes and ponds and on the sheltered shores of backwater bays of the larger loughs in the VC. It does not occur, or is only occasional and sparsely developed, where water stagnates. Typically, R. hederaceus requires sufficient seepage of ground water to create at least a slight fluctuation in levels and a consequence of this will be an additional in-flow of oxygen and mineral nutrients.

R. hederaceus was also once found in muddy ground by temporary pools in an old quarry beside Keenaghan Lough and it also appeared on an urban waste tip in Enniskillen, which mirrors Segal's experience in Holland (Segal 1967). However, the most predictable habitat in Fermanagh of this small, prostrate annual or short-lived perennial appears to be by field gateways and along wet, muddy tracks where water lies in shallow quagmire pools and puddles in mud that has been trampled, poached and dung-enriched by cattle. We find it both where the parent rock of the soil is acidic or of a calcareous nature, and it clearly tolerates a rather wide range of pH and appears indifferent to lime (Cook 1966b).

The authors of the critical Flora Nordica regard R. hederaceus as a calcifuge throughout their northern European region (Jonsell et al. 2001). In Great Britain, Hill et al. (1999), summarising the preferred environmental growing conditions of species, gave R. hederaceus an indicator value of '5' on a nine point scale, indicating that while it mainly occupies moderately acid soils, occasionally it ranges wider and it can occur on both very acid and more neutral to basic (alkaline) substrates. In the latter situation, due to our very wet climate, a leached calcareous soil usually develops. This generally becomes overlain with a thin, acid, peat horizon, which forms the substrate over which Ivy-leaved Crowfoot spreads and shallowly roots at its nodes.

Cook (1966b) made a significant point when he speculated that if R. hederaceus and the closely related species R. omiophyllus Ten. (Round-leaved Crowfoot) are not directly competing (the latter does not occur anywhere in N Ireland), they may exhibit wider ecological amplitudes than when they overlap with one another. The habitat description of R. hederaceus is rather similar to that of R. sceleratus (Celery-leaved Buttercup), but the latter is more frequently found by lakes, or in somewhat less disturbed, more nitrogen-enriched, perhaps slightly better drained mud which is more liable to drying out temporarily than that of R. hederaceus. Also, R. sceleratus never takes on a truly aquatic 'submerged' – or more realistically – 'floating' existence, in the way that Ivy-leaved Crowfoot is capable of doing in water even as shallow as 5 cm!

Apart from its freshwater habitats, R. hederaceus can also be found in sheltered coastal parts of both Britain and Ireland, usually on the upper edges of salt-marshes (Preston & Croft 1997).

Variability and identification

R. hederaceus is classified within the Subgenus Batrachium of the genus Ranunculus. Most species of this subgenus are notoriously variable, frequently hybridize and backcross, and thus are often very difficult to identify (Cook 1966a; Holmes 1979; S.D. Webster In: Rich and Jermy 1998). R. hederaceus and its close relative R. omiophyllus (Round-leaved Crowfoot) (the latter apparently absent from Northern Ireland), differ from all the other members of the subgenus by being essentially semi-terrestrial and by invariably having un-dissected, laminar, floating leaves. Unlike all other water-crowfoots in the subgenus, these two species do not possess thread-like submerged leaves (Cook 1963, 1966b). Thus, in our Fermanagh survey area, Ivy-leaved Crowfoot is very easily recognised, despite the species being remarkably variable in form in response to environmental changes, ie it has a very plastic phenotype with respect to many characters. This high level of variability is quite characteristic of emergent aquatic species (Segal 1967; Cook 1966a, b).

The creeping or surface floating habit of the plant, its small white flowers, and ivy-shaped leaves with dark markings following the veins are all very distinctive identification features, and the typical plant species associates of Ivy-leaved Crowfoot include Montia fontana (Blinks), Callitriche spp. (Water-starworts) and Stellaria uliginosa (Bog Stitchwort).

Summary of environmental requirements

From the above, it is clear that R. hederaceus has the wide ecological tolerances that one would expect of a species inhabiting situations where slight but significant fluctuating water levels are the norm, and its requirements can be satisfied in a diverse range of habitats encompassing a mosaic of vegetation communities, moist, wet, semi-aquatic and shallow aquatic, lowland and upland, coastal and inland. The most constant requirements of the species are for high levels of illumination, wet, moderately acid, waterlogged soil, or very shallow water only a few centimetres deep, of medium (mesotrophic) fertility, or richer more productive eutrophic levels, plus shelter from strong water currents.

Temperature limits and distribution

An equable, fairly low water temperature throughout the year is recognised by ecologists and plant geographers as another very important environmental factor controlling the growth and occurrence of R. hederaceus. Temperature is sufficiently significant to actively govern the wider distribution of the species, confining it to the truly oceanic or Atlantic region of W Europe. A summer maximum of around 16C, and mild winter temperatures with little in the way of severe frost, characterises the required regime (Segal 1967).

In his study in Holland, Segal noticed that R. hederaceus tends to occur in specific landscape situations where small scale water bodies lie on the junction between hilly, acidic, infertile (oligotrophic), non-calcareous soils and much more fertile, mineral-rich (mesotrophic to eutrophic) conditions, on lower ground at the base of slopes. In the habitat gradients that occur when two very different ecological environments of this nature meet, and especially where the zone of contact is kept open by some form of disruption, either physical (eg trampling and grazing), or chemical (eg a moderate level of pollution, including manuring and other forms of farm effluent run-off), or both of these, then R. hederaceus and other interesting and quite scarce species such as Catabrosa aquatica (Whorl-grass) and Veronica catenata (Pink Water-speedwell) appear to find growing conditions to their liking. On the other hand, in excessively enriched sites containing high levels of nitrogen and phosphorus, ground that tends to be rapidly overgrown by algae and by Lemna species (often L. gibba (Fat Duckweed)), R. hederaceus cannot compete for very long under such conditions and becomes ousted.

Flowering and pollination

Ivy-leaved Crowfoot flowers early and for a prolonged period, often stretching from April to August. However, flowering is frequently curtailed by the habitat becoming overgrown later in the summer, or sometimes by it drying out. The small white flowers produce nectar, are sweet scented and are protogynous (ie the female parts develop first, followed by the anthers). This difference in sexual timing in the individual flower is generally considered an adaptation favouring or enabling cross-fertilisation between flowers, but in reality the blooms are highly self-compatible and they appear to habitually inbreed. Indeed, self-fertilisation often takes place at the unopened bud stage, making it obligatory. Despite this sexual behaviour, the timing of flower opening depends on the prevailing weather and some level of opportunity for cross-pollination does exist in other flowers. However, observation also indicates that the flowers attract few insect visitors (Cook 1966b).

Occasionally R. hederaceus flowers are produced underwater. When this occurs, a gas bubble is formed within the bud, allowing pollination to proceed as normal (Cook 1966b). Since R. hederaceus is seldom submerged and, if so, then usually only for short periods, submerged pollination is not likely to occur very often.

Fruiting and dispersal

As the fruit develops the flower stalk bends away from the light, forcing the developing achenes into the mud (Cook 1966b). This is another example of a negatively phototrophic movement by a fruit stalk (ie a growth movement away from the direction of light), similar to that of the fruit stalk of Cymbalaria muralis (Ivy-leaved Toadflax). It appears odd that both species which commonly show this unusual physiological feature, should possess ivy-shaped leaves!

Dispersal of the seeds (achenes – single-seeded dry fruits) is most probably achieved in mud by attachment to animals or vehicles. There might possibly be some degree of water dispersal too, if a fast current were to develop in the locality and dislodge seed shed onto the soil surface around the plant.

Germination and growth form

Seed germination is reported by Cook (1966b) to be very irregular if the achenes are kept wet – which one might imagine would be the normal condition of the habitat. On the other hand, if the mud dries out after the seeds ripen and the seeds themselves become dried, then Cook found that when they are re-wetted, the seeds gave nearly 100% germination. He also showed that, depending on the local regime of water levels and competing species, R. hederaceus (and also R. omiophyllus (Round-leaved Crowfoot)), can behave either as winter or spring annuals, or individual plants may persist and reproduce on multiple occasions for periods up to six years, thus achieving perennial status.

Overwintering and factors affecting survival

During the winter months, the resting plant survives as a small, tight, rosette of leaves. In this state, it is very resistant to freezing, desiccation and shade. However, in summer when individual plants develop their normal spreading habit, they becomes very susceptible both to the three mentioned physical factors and to pressure from taller, more aggressive competing plant species (Cook 1966b).

Ignorance regarding seed output, reproductive strategy and longevity in the soil seed bank

In common with many other species in the flora of Britain and Ireland, after careful searching, we do not appear to have any figures for typical plant seed output. Nor do we know anything regarding the relative significance of seed versus vegetative reproduction for perennial populations of R. hederaceus, nor even if a persistent soil seed bank exists (Thompson et al. 1997).

British and Irish occurrence

In Britain, R. hederaceus has quite a distinct northern and western distribution, partially created by the destruction of suitable habitats in much of the SE due to development, drainage and a shift from livestock to arable farming. The change to arable farming means that existing wetlands are no longer subject to the trampling and disturbance of grazing animals that previously opened the ground to colonisation by Ivy-leaved Crowfoot (Preston & Croft 1997; Preston et al. 2002). In Ireland, R. hederaceus, being essentially a lowland wetland species, has a somewhat scattered, almost disjunct distribution, featuring a heavy concentration of records in northern counties, but it is more coastal in the Republic and appears quite scarce in the Midlands.

European and world occurrence

R. hederaceus occurs throughout the Atlantic region of W Europe from Portugal to the S tip of Sweden (Jalas & Suominen 1989, Map 1884). It was previously present further N up the coast of Norway at Tronheim and considered native there by some, but it became extinct there in 1950 (Jonsell et al. 2001). Cook (1983) regarded the species as endemic to Europe (and indeed one of only a few ancient 'palaeoendemics'), and he felt that it has declined throughout Europe, probably due to changes in agricultural practices over the past 50 or so years which involved the destruction of wetland habitats. Preston reminds us that the species was losing sites in SE England before 1900 for the same reason and also because of urban expansion (Preston et al. 2002).

The species is also present in disjunct locations in eastern N America where it has been known since 1821 – in Newfoundland and the Chesapeake Bay region (Hultén 1958, Map 137). The general shape of the distribution of R. hederaceus in N America and its history discussed in Cook (1983, 1985), suggests that it was introduced from Europe. Against this view there is the existence of two fern species, Schizaea pusilla Pursh and Woodwardia areolata (L.) Moore, which have very similar N America distributions to R. hederaceus and they are certainly not European introductions.

Names

The Latin specific epithet 'hederaceus' simply translates as 'ivy-leaved' (Gilbert-Carter 1964) and the plant, having no herbal or folk-lore usages, does not appear to have any English common names other than the one chosen by the committee of the Botanical Society of Britain and Ireland (Dony et al. 1974).

Threats

None. The species may be benefitting, like R. sceleratus, from the current general eutrophication of water bodies, and considerable physical disturbance helps it colonise by keeping the habitat open.

Growth form and habitat preferences

This aquatic species, which only produces dissected capillary leaves, is a spreading to erect perennial when growing in permanent water, but it becomes greatly dwarfed and behaves as a small annual if its muddy substrate dries out during the summer. It is most frequently found in still or almost still shallow water less than 50 cm deep in small lakes or ponds and water-filled holes in disused quarries.

Thread-leaved Water-crowfoot roots mainly in bare clayey bottoms and it can tolerate a wide range of water chemistry, from oligotrophic to eutrophic, lime-rich or otherwise. It chiefly frequents mesotrophic to eutrophic conditions in lowland water bodies (C.D. Preston, in: Preston et al. 2002). R. trichophyllus also occurs in brackish waters around the coastline of Great Britain and Ireland (particularly in Scotland and Wales), eg in dune-slacks and around sheltered estuarine bays (Preston & Croft 1997).

Like its close relatives, R. aquatilis (Common Water-crowfoot) and R. peltatus (Pond Water-crowfoot), this species tolerates disturbance such as trampling by stock animals, human activities and/or a substrate which seasonally dries out. It can therefore behave as a pioneer colonist of bare mud around temporary water bodies, including recently dug or cleaned pools and drains (Cook 1966a, p. 134). Later in the season, as the vegetation in such muddy habitats matures, R. trichophyllus and its pioneer associates tend to be ousted by increased shade and other forms of competition produced by rhizomatous aquatic plants such as Potamogeton species (Pondweeds), Hippuris vulgaris (Mare's-tail), or by any of the numerous vigorous emergent plants of swampy ground.

Fermanagh occurrence

In Fermanagh, there were only two records for R. trichophyllus prior to 1980 and there have been just eight additional finds of the species since then. There are records from nine Fermanagh tetrads, seven with post-1975 dates. The species is undoubtedly under-recorded here, as is also the case elsewhere in Britain and Ireland. In the New Atlas, Preston suggests that part of the reason for the paucity of records is the early flowering season of the species, and this, together with the well-known discouraging taxonomic difficulties of the Batrachian group of Ranunculus, is very probably responsible for the dearth of observations (Preston & Croft 1997; New Atlas).

Most of the Fermanagh records for this species will have been determined on the basis of the flower size being less than 12 mm in diameter, with the petals being non-contiguous. Non-flowering plants in this group are almost impossible to identify, contributing to their all being under-recorded. Regrettably, none of the ten Fermanagh records of R. trichophyllus is supported by a voucher.

As the tetrad map indicates, most of the records are confined to the Lough Erne basin, with just four outlying stations, one at Green Lough turlough and three scattered further east. Although more frequently found by smaller water bodies, it does also occur in well-sheltered areas around larger lakes, especially where shallow backwater bays are protected by islands close to shore. Local examples of the latter habitat are the West Point islands and Lowery Island, both at the W end of Lower Lough Erne.

Flowering and pollination

Thread-leaved Water-crowfoot is the first Batrachian Ranunculus species to flower in Britain and Ireland, which it does from April through to July. The flowers are habitually self-pollinated, this often (but not always) taking place in the bud (ie the flowers are cleistogamous). As a result of this, seed set is perfect (100%) and completely assured (Cook 1966a, p. 183; Hong 1991, p. 53).

Seed output

Statistics of mean seed production per plant do not appear to be available, or have not been located.

Germination

The seeds can germinate in wet conditions at any time of year, but in winter seedlings require the protection of submerged conditions. If frozen under terrestrial conditions they die, but they can survive being frozen in ice for several months (Cook 1966a).

Survival in the soil seed bank

Seed longevity is unknown, but it is probably similar to that of R. peltatus which is considered to be only short-term persistent, ie the seeds survive in soil for periods ranging between one and five years (Thompson et al. 1997).

Vegetative reproduction

Like most other aquatic macrophytes, R. trichophyllus can reproduce vegetatively from plant fragments after disturbance or uprooting of an original established individual. An experimental study in France examined the fate of five different types of plant fragments of R. trichophyllus and five other wetland species. The plant parts used ranged from roots, to stems with or without nodes, to apical buds. After ten weeks greenhouse incubation floating in pans of shallow water (7-8 cm deep) and sediment, the R. trichophyllus results showed that within two or three weeks, between 30% and 50% of the aerial parts rapidly rooted and established into the sediment, while the remainder of the fragments of the species died off without showing any root development. None of the fragments of R. trichophyllus developed new buds before they rooted (Barrat-Segretain et al. 1998). The French workers recognised that the six aquatic macrophyte species they studied reacted to disturbance by exhibiting one of two survival tactics: 1. the fragments either develop roots and establish rapidly into the sediment (eg R. trichophyllus and Sparganium emersum (Unbranched Bur-reed), or 2. the fragments develop many new propagules that may be dispersed, but they fail to establish within the 10 weeks of the experiment (eg Hippuris vulgaris (Mare's-tail) and Elodea canadensis (Canadian Waterweed)) (Barrat-Segretain et al. 1998).

The species examined in the French study appeared to present a trade-off between the plant's regeneration and colonisation abilities: each species practiced both tactics, but high levels of regeneration implied low colonisation ability and vice versa (Barrat-Segretain et al. 1998). One of the most interesting and unexpected conclusions of this work was that the different survival tactics involving vegetative reproduction were not related to the morphological types of the species and thus they were unpredictable. At the same time, it appears that species possessing narrow ecological amplitude promote dispersion of propagules, thus increasing opportunities of finding suitable habitats, whereas other species (including R. trichophyllus), that are more tolerant of varying ecological conditions, try to anchor themselves as soon as possible at the nearest sites to the original population (Barrat-Segretain et al. 1998).

A subsequent similar experimental study showed that some aquatic species behaved quite differently with respect to these two vegetative strategies in the spring and in the autumn, but this did not apply to either R. trichophyllus or Sparganium emersum (Unbranched Bur-reed) (Barrat-Segretain & Bornette 2000).

British and Irish occurrence

In Ireland, R. trichophyllus is occasional in the C & SE, and rare and widely scattered elsewhere including NI (Webb et al. 1996). In Britain, it is much more frequent in the SE of England, below the line from Hull to Bristol, but it is widely scattered in the lowlands elsewhere, becoming distinctly coastal in the W and throughout Scotland (Preston & Croft 1997; Preston et al. 2002). Although the New Atlas contains a high proportion of pre-1970 records, the authors consider that this does not indicate any real decline in what is a difficult enough species to identify.

European occurrence

Beyond the British Isles, R. trichophyllus is probably the most widespread Batrachian species in the world. The only reason for qualifying this with the word 'probably' is, of course, the great amount of variation and plasticity within the group, plus the infamous taxonomic difficulty of the subgenus which creates problems in terms of delimiting the species and their field recognition. For instance, when examining material of R. aquatilis (Common Water-crowfoot) and R. trichophyllus in S Sweden, Hong (1991) found that all the distinguishing characters for these two species, including nectary shape, were variable and readily modified. Also, sexual crosses between typical forms of the two species were found to be fully fertile, a feature undoubtedly allied to polyploidy within the complex. In S Sweden, both of these species are regarded as hexaploid taxa, whereas in his taxonomic review Cook (1966a) dealt entirely with tetraploid R. trichophyllus and hexaploid R. aquatilis material from C and W Europe. Reflecting this degree of variation, in Flora Nordica 2 (Jonsell et al. 2001), R. trichophyllus is downgraded from a species to a variety of R. aquatilis, referred to as var. diffusus With.

Apart from these problems of polyploidy and status change, R. trichophyllus is regarded as widespread throughout Europe, especially frequent in western regions and thinning eastwards. Unlike many other members of the subgenus, it is even frequent in the Mediterranean basin and is present on many of the smaller islands, for example, Majorca, Malta, Cephalonia, Zante and Crete (Jalas & Suominen 1989, Map 1895).

World occurrence

Beyond Europe, it is a very widespread species being both Arctic and circumboreal stretching across N Asia, the Himalaya, N and W China, Japan, SE Australia, Tasmania and central areas of N America (Cook 1966a; Hultén 1974, Map 80; Hultén & Fries 1986, Map 874). R. aquatilis is the only other water-crowfoot that approaches this extent of world distribution, but it is not as well represented as R. trichophyllus in both northern regions and in the Mediterranean region.

Dispersal of aquatic species

We have to ask ourselves why should R. trichophyllus be so much more widely distributed than all the other aquatic Ranunculus species? It is listed by Cook (1985, Table 1), along with Potamogeton pectinatus (Fennel Pondweed) and 24 other aquatic species widely distributed in the world, where their native ranges are impossible to distinguish from their areas of introduction. Prof. Cook points out that it is surprisingly difficult to get recent, accurate information on the distribution of aquatics and it is even more difficult to discover or decide where a particular species is native rather than introduced. In connection with dispersal ability, Cook also points out that aquatic plants in general lack very light wind-carried diaspores or propagules and they also lack large seawater-resistant ones; they are thus ill-equipped for long-distance dispersal. The widespread plants, listed in his Table 1 of unknown native status, have probably spread more or less naturally, or at least they dispersed before botanists were around keeping records!

The majority of aquatic plant species have become mobile as man has increased his own global mobility (Cook 1985). Reviewing the distribution of 172 aquatics, Cook found that marked imbalances existed between Old and New World regions, in terms of donor and recipient areas with respect to introduced species. He concluded that the problems of successful establishment in new lands are quite obviously very great and that it is extremely naive to regard the aquatic environment as globally uniform, thus allowing or expecting aquatic macrophytes to have a wide geographical range. It is known that some aquatic macrophyte genera (eg Brasenia and Dulichium), became extinct in Europe before or during the last Ice Age, but they persisted in N America. These plants have not re-established in Europe despite being in garden cultivation there, an observation, which lends strength to the argument that ecological niches in the aquatic environment are more complex than we commonly assume. Cook argues that we should not combine the terms 'introduced aquatic' and 'weed', pointing out that weeds (ie undesirable plants) essentially tend to flourish in sites that are disturbed by man. "Usually it is the nature of the disturbance that is more important for the establishment of the weed than its [powers of] mobility." (Cook 1985).

It is very clear that disturbance by man and his animals, and pioneer colonisation of bare muddy ground are two very important ecological features influencing the occurrence of many Batrachian Ranunculus forms. In many cases, these species could not successfully reproduce sexually and establish seedlings without the habitats they occupy temporarily drying out. Following Cook's advice, and maintaining a healthy scepticism regarding the real distribution of many difficult-to-identify Batrachian Ranunculi, it appears to follows that his remarks regarding introduced aquatics must apply quite closely to R. trichophyllus, since it does not appear any better equipped for long-distance jump dispersal than R. peltatus (Pond Water-crowfoot), R. penicillatus (Stream Water-crowfoot) or, indeed, any of the other seed- and vegetative propagule-producing members of this complex subgenus.

Names

The Latin specific epithet 'trichophyllus' is derived from the two Greek words for 'hair' and 'leaves', and thus translates as, "having leaves divided into hair-like segments" (Gilbert-Carter 1964). The English common name, 'Thread-leaved Water-crowfoot', is a straightforward invented 'book name', following the same line of thought.

Threats

None.

Growth form and preferred habitats

An annual or short-lived perennial that grows in shallow water in marshes, ponds and ditches and at the edges of slow-moving streams and margins of sheltered lakes. It prefers lowland sites with water that is eutrophic, somewhat base-rich and subject to a moderate degree of disturbance that helps keep the habitat open. These conditions can occur where cattle graze lakeshore margins and water meadows, or pastures around muddy ponds.

Fermanagh occurrence

As emphasised in much of the literature, the heterophyllous and large-flowered water-crowfoot species, R. aquatilis, R. peltatus (Pond Water-crowfoot) and R. penicillatus subsp. penicillatus (Stream Water-crowfoot) are not easily distinguished. This fact should be borne in mind when studying distribution maps of the three species, either ours or anyone else's for that matter matter (Cook 1966a; Preston & Croft 1997; Preston et al. 2002).

Undoubtedly identification difficulty leads to under-recording, and we have only eleven records for R. aquatilis. Since a degree of reasonable doubt hangs over them, we have combined them as s.l. We do not mean to suggest by this that we are considering R. aquatilis s.l. as the old pre-Cook (1966a) species aggregate. However, it is probable that most of the Fermanagh identifications were determined using An Irish Flora (1977) as a field guide. This text used only two features to discriminate this species: flowers between 12 and 17 mm in diameter and floating leaves present in summer. As Cook (1966a) and Webster (1991) have shown, flower size in this species group varies considerably with environmental conditions, and we now appreciate that it is not a reliable distinguishing species character.

Keeping in mind the mentioned reservations, as the tetrad map indicates, the Fermanagh records of R. aquatilis are thinly and widely scattered across ten squares in lowland Fermanagh. Only one record for R. aquatilis pre-dates 1975, that of Barrington (1884), from a wet ditch on Devenish Island. Soon after this record was reported (as R. heterophyllus Fries), Tetley wrote into his copy of Irish Topographical Botany in pencil, "I have no doubt this should be R. peltatus". While only two of the records (Barrington's and one by the NI Lakes Survey) are supported by vouchers in BEL, the eleven records involve at least seven determiners, increasing the probability that some of them are correctly identified. The habitats listed are typical for this species, ie shallow, eutrophic, somewhat base-rich water in lakes, turloughs, rivers, streams and ditches. Further investigation of Batrachian Ranunculi in Fermanagh is very obviously required.

British and Irish occurrence

The distribution map in Preston & Croft (1997) took a more conservative approach to R. aquatilis records than did the editors of the New Atlas (Preston et al. 2002). Recognising that uncertainty still clings to the identification and status of the taxon, the earlier approach is undoubtedly the wiser one. This shows modern records of R. aquatilis extremely thinly scattered throughout Ireland, while the Great Britain distribution displays the plant as more frequent and there are four or more discrete areas of concentration in England and Wales. At the same time, R. aquatilis becomes much rarer and almost entirely coastal in Scotland (Preston & Croft 1997).

European occurrence

R. aquatilis is widespread in Europe from S Scandinavia southwards to the Mediterranean, but while it reaches Sicily, it does not penetrate far into the Iberian Peninsula, or into Greece (Jalas & Suominen 1989, Map 1894).

World occurrence

R. aquatilis s.l. is also found in N Africa, and in the boreal zone it extends across Asia except in the south, plus W and C North America where it also occurs in the alpine zone of the Middle Rockies and southwards into California and New Mexico (Scott 1995, p. 657). It also extends into the southern hemisphere, being present in both western South America (the Andes between 2,700-4,000 m) and SE Australia (Cook 1966a; Jonsell et al. 2001). The world distribution of R. aquatilis is probably second only to that of the closely related species R. trichophyllus (Thread-leaved Water-crowfoot). However, one has to remember that here we are dealing with species which are difficult to identify, and since they are plants which live in disturbed muddy habitats, they are undoubtedly and unwittingly spread by man, so that we do not have much of an idea as to their native ranges (Cook 1985).

Name

The Latin specific epithet 'aquatilis', means 'growing in water' (Gilbert-Carter 1964).

Threats

Shading from competitors and overgrowth by algae and/or other macrophytes due to excessive cultural eutrophication.

Growth form and preferred habitats

Like several other species and forms within the Ranunculus subgenus Batrachium, in permanently wet habitats R. peltatus is entirely perennial provided it is not overgrown or shaded by larger or more robust species. When growing in water, the plant normally produces both floating lobed laminar leaves and finely-dissected submerged capillary ones. However, if the site R. peltatus occupies dries out in summer, the species behaves as an annual and becomes both dwarf and prostrate, or even somewhat tufted in appearance (Cook 1966a; Webster 1988). As far as we know the annual form of the plant has never been reported as such in Fermanagh, but at least in grazed ground in and around turloughs (ie vanishing lakes rarely occurring in limestone districts), it could very well be the normal mode of the species.

In Ireland, R. peltatus is most typically found in more or less unshaded stretches of slow to moderately-flowing streams and rivers in calcareous districts. However, it can also occur in similar flow rates in much more acidic situations over peat or clay substrata and it appears to occupy a rather broad range of habitats in terms of nutrient status.

It should be noted that despite the species epithet 'peltatus', none of the organs of the plant, including the floating leaves, are ever peltate in form. The botanical name is misleading and a definite misnomer, arising as an 18th century error made by Prof. Carl Linnaeus himself! Further remarks on this topic are made below in the section 'Names'.

Occurrence in Fermanagh

In Fermanagh, this perennial or sometimes annual aquatic has been recorded in a total of just 14 lowland tetrads, 2.7% of those in the VC. It grows in shallow water in calcareous rivers, streams, lakes and ponds. The main site in Fermanagh for R. peltatus appears to be the Colebrooke River, where both R. peltatus and R. penicillatus subsp. penicillatus (Stream Water-crowfoot) occur. R. peltatus is also found at the Green Lough turloughs (ie limestone lakes or ponds that drain vertically – often referred to as 'vanishing lakes', since they dry out quite regularly), and it is found in an additional five or six rather scattered locations in the VC.

Although locally R. circinatus (Fan-leaved Water-crowfoot) appears in four more tetrads than R. peltatus, the latter is probably the commonest aquatic species of Ranunculus subgenus Batrachium in the VC, provided that is, R. hederaceus (Ivy-leaved Crowfoot) is discounted, since at least in our survey the latter is more often a terrestrial rather than an aquatic species.

Variation and identification

According to An Irish Flora (1977), specimens of R. peltatus with semi-circular to circular floating leaves produced in spring and summer (ie in addition to the constantly present finely divided, submerged, capillary leaves) and with large flowers (at least 18 mm across), should be reasonably easy to recognise as this species. In reality, however, there is considerable environmentally and seasonally induced variation in R. peltatus and it overlaps with R. penicillatus subsp. penicillatus (Stream Water-crowfoot), both in terms of morphological characteristics and ecology. Thus these two forms, in particular, are very readily confused (Cook 1966a; Holmes 1979; Webster 1988, 1991; Rich & Jermy 1998, p. 60).

Prior to Cook's 1966 monograph on Batrachian Ranunculi and for many years afterwards, due in part to the inadequacies of the Irish field Flora in widespread use here (An Irish Flora 1977), the few records we have of this closely related group of species were regularly lumped as R. aquatilis agg. or R. aquatilis s.l. (Common Water-crowfoot). Three truly aquatic members of the subgenus Batrachium, namely R. peltatus, R. aquatilis and R. trichophyllus (Thread-leaved Water-crowfoot), all exhibit very similar patterns of variation, both genetic and in response to the environment (ie genotypic and phenotypic variation), which Cook (1966a) believes indicates their close parental affinity. When these three species are in a submerged or terrestrial, vegetative, wholly divided-leaf state, they are in fact morphologically indistinguishable. More work is definitely required to sort out the differences and clarify the distribution of the water-crowfoot species in Fermanagh and, since the problem is universal, the same is true everywhere in the British Isles and, indeed, throughout Europe (Preston & Croft 1997; New Atlas).

Nutrient ecology and tolerance of disturbance

In the Sheffield area of England, Grime et al. (1988) believed R. peltatus was biased towards mildly acidic (often peaty) habitats, and that it was more or less absent from calcareous waters, where it was replaced by R. trichophyllus. This is definitely NOT the case in Fermanagh. Like R. hederaceus and R. lingua (Greater Spearwort), R. peltatus also appears to favour lowland aquatic sites which are subject to a degree of disturbance, appearing for instance in drainage ditches, quarry pools, and streams and ponds in areas that are regularly grazed and trampled by cattle or other stock animals, or in waters polluted from time-to-time with manure and agricultural chemical run-off (Cook 1966a; Hong 1991). One is here reminded of Grigson's aside in his essay on alien plants entitled 'The Wandering Flower', where he writes, "You may not be able to say so to a botanist, but nature interfered with by man is rather more fascinating than nature left to itself." (Grigson 1952, p. 114).

Cook (1966a) regarded R. peltatus and its two very closely relatives R. aquatilis and R. trichophyllus, as being characteristic pioneer plants of freshly dug or recently cleaned pools and ditches. All three rarely occur in water deeper than 100 cm and they do not tolerate deep shade or very swiftly flowing water. Cook also regarded R. peltatus and R. aquatilis as being confined to eutrophic waters although in Fermanagh we regard R. peltatus as a plant of more or less mesotrophic conditions. Cook never found these two species sharing a habitat, while both of them may frequently be intermingled with R. trichophyllus (Cook 1966a, p. 134).

In addition to the above habitats, in low-lying coastal areas of Britain and Ireland R. peltatus also occurs in dune-slacks and in lagoons. In S England, it frequents the uppermost reaches of calcareous rivers, which like the Irish turloughs, tend to dry out in summer (C.D. Preston, in: Preston et al. 2002).

The effect of drought on wetland plants

While numerous studies exist on the tolerance of wetland plants to flooding, in comparison very little research has looked at the effects of drought periods on such species. A worthwhile exception is Volder et al. (1997), who carried out a pot experiment on R. peltatus from wetlands in the French Camargue. This study found that the stage of development at which drought onset occurred had a major impact on the growth form and seed output of the species. R. peltatus plants subjected to the most severe experimental drought treatment (ie air drying of the soil with no added water), changed their vegetative growth form and survived for up to two weeks. Despite the very severe drought conditions, their flower, fruit and seed production, while very low, was not zero. The ability of R. peltatus to survive and produce viable seed under a wide range of hydrological conditions ensures a regular input to the seed bank. This fact helps explain the success of the species in temporary marshes and other wetland habitats, where exposure to drought stress is a regular or even an irregular occurrence (Volder et al. 1997).

Flowering and pollination

How the breeding system of R. peltatus operates is a matter of minor dispute. Cook (1966a) describes pretty well all of the Batrachian Ranunculi as being protogynous (ie female organs develop first) and self-pollinated, with a tendency to cleistogamy (a term literally meaning 'closed marriage', ie flowers self-pollinate in the bud). As he describes it, there is a variable length of time during which un-pollinated stigmas are exposed and outbreeding might occur. In R. peltatus, Cook reckons this duration can be up to 48 hours, but it is dependent upon the prevailing weather conditions being warm and bright. The flowers produce nectar and do have a scent similar to Cratageus monogyna (Hawthorn) blossom (Cook 1966a, p. 183).

In a study of Batrachian Ranunculi flowering behaviour in S Sweden, Hong (1991) agreed with Cook's findings for R. aquatilis and R. trichophyllus, but found no open anthers in the flower buds of R. peltatus. When the flowers opened, the stamens spread horizontally and gave good separation (c 5 mm) from the stigmas. In other Batrachian species Hong examined, the stigmas withered within a few days, but in isolated flowers of R. peltatus they remained fresh, swollen and looked receptive for periods up to 15 days. Beetles were observed visiting the flowers and Hong considered that they were the real agents of pollination in this species (Hong (1991) cited in Proctor et al. 1996). Other observations led Hong to suggest there might be a measure of self-incompatibility, such as definitely exists in R. acris (Meadow Buttercup). Jonsell et al. (2001) go further in this matter, stating that in Scandinavia R. peltatus is an obligate outbreeder, but somehow it retains a low degree of self-compatibility.

Overall reproductive strategy

As is often the case, nothing is known regarding the relative significance or frequency of successful establishment of sexual versus vegetative reproduction in R. peltatus, but as with R. penicillatus, a successful vegetative process is certainly easier to imagine happening. Certainly the species forms large stands and these probably consist of one or just a few clones, propagating simply by fragmentation (Jonsell et al. 2001).

British and Irish occurrence

The New Atlas map shows R. peltatus having a widely scattered, decidedly patchy, although quite frequent British and Irish distribution (Preston et al. 2002). However, the data must be qualified with an appreciation of the fact that all the Batrachian Ranunculi species are difficult to recognise and distinguish, and errors occur at an unknown frequency. We must therefore approach all maps of their distribution with caution and do not regard the presented pattern as totally reliable.

Having said this, the New Atlas hexad map for Ireland shows more frequent recording of R. peltatus in NE Ireland than elsewhere, a pattern that probably reflects recorder effort. In Great Britain, the species presence appears to thin out and become increasingly confined to coastal situations in both SW England and to the north of the Glasgow-Edinburgh conurbations in Scotland. In Scotland, at least, this probably reflects increasing altitude, but it will also mirror the number of available expert field recorders capable of identifying the species of this difficult plant group.

European occurrence

The maps in Cook (1966a, Fig. 15), Jalas & Suominen (1989, Map 1888) and Jonsell et al. (2001, p. 261), show R. peltatus widespread throughout Europe from the Mediterranean to the far N of Scandinavia, but with an overall distinct western concentration of records.

World occurrence

Beyond Europe, R. peltatus occurs only in the coastal area of N Africa and in parts of Asia Minor adjacent to SE Europe, but a closely related form called R. sphaerospermus, occurs in SW Asia (Cook 1966a; Hultén & Fries 1986, Map 872; Preston & Croft 1997; Jonsell et al. 2001).

Names

The Latin specific epithet 'peltatus' is derived from the Greek 'pelta' meaning 'a shield'. In plants, the technical term 'peltate' most often refers to the attachment of the stalk or petiole directly to the under-surface of the leaf blade, rather than to the blade margin in the normal manner (Gilbert-Carter 1964; Gledhill 1985). As mentioned above, the name 'peltatus' was given to this species in error by Carl Linnaeus in 1751, since no member of the Batrachian Ranunculi ever develops peltate leaves or other organs (Cook 1966a, p. 123). However, since the species name complies with the international rules of botanical nomenclature and is correct in every other respect, it has priority and must be retained. In this instance, the name is best considered merely a label, rather than as is more usually the case, descriptive, reflecting a specific character of the plant.

Threats

Shading and overgrowth by more vigorous plants encouraged by eutrophication.

Identification difficulties and under-recording of subgenus Batrachium

Although there has been a huge amount of survey work on the aquatic flora of Fermanagh since 1986, almost all of it has concentrated on the lakes and comparatively little focus and attention has been given to rivers and streams in the county. The resultant plant records could give the impression that field workers appear to have avoided Ranunculus subgenus Batrachium. If this really is the case and the group is under-recorded, it will largely be on account of the difficult taxonomic nature of the subgenus, the range of variation and the confusion engendered by frequently changing names within the complex. Another contributory factor is the fact that the Irish field Flora in use for much of the period 1975-97 (ie An Irish Flora 1977) did not handle this plant group at all well. For example, it failed to mention R. penicillatus and completely ignored the detailed monographic treatment of the subgenus carried out by Cook (1966a).

Growth form and habitat preferences

R. penicillatus s.l. is by far the most common and one of the most robust perennial species of water-crowfoot in the flora of Britain and Ireland. It frequently forms large, conspicuous rafts of dense branching stems rooted in the bed of moderately flowing rivers and streams, but it can also occur in waterways that are liable to occasional, torrential spates (Haslam 1978, p. 31; Holmes 1980; Webster 1988). The waters that subsp. penicillatus frequents range from oligotrophic to mesotrophic nutrient status (ie they are capable of supporting poor to medium levels of plant and animal growth).

Under favourable water conditions with stable bottom substrates of well-cemented rounded gravel and silt, R. penicillatus subsp. penicillatus can become the dominant aquatic vascular plant, its dense floating mats representing a significant structural feature of the water channels it occupies. It has been observed covering around 70% of the stream bed and modifying water flow, promoting silt deposition, and providing both shelter for other plants and animals, and food for invertebrate animals and fish (Cook 1966a; Haslam 1978, p. 40; Ham et al. 1982; Preston et al. 2001).

In an Irish study of water quality indicator species, Caffrey (1985) placed all of the Batrachian Ranunculi in his group of macrophytes that are most sensitive to increased nutrient forms of pollution, ie especially high levels of nitrogen and phosphates. We believe that all or most of the previously luxuriant rafts of R. penicillatus in the Colebrooke River have disappeared in recent times, almost certainly due to increased nutrient levels in the water originating from run-off of agricultural chemical fertilisers, slurry and manure applied to adjacent field swards.

Heterophylly and identification features

Throughout the year, plants of R. pencillatus bear dissected, thread-like submerged leaves and, during the summer months, in response to longer day-length, they develop additional floating leaves that are 3-5 lobed and either distinctly or bluntly toothed. Thus the plants become heterophyllous, ie bearing leaves of two quite different types. The teeth on the laminar floating leaves may end in long slender filaments, which actually are the leaf veins extending beyond the margin. The submerged leaves of R. penicillatus subsp. penicillatus are always limp or flaccid, and when removed from the water they collapse to form a very slender tassel. When fully developed these leaves are generally longer than the adjacent stem internode (Webster 1991; Webb et al. 1996). This feature, used with care, helps separate the species from R. peltatus (Webster 1991).

Fermanagh occurrence

Whatever is or was the situation regarding the difficulty of identifying Batrachian Ranunculi, relatively few records have been made of the water-crowfoot group as a whole in the Fermanagh survey (Cook 1966a; Webster 1991). There are a total of just eleven records for R. penicillatus in the Fermanagh Flora Database, the species being present in eight post-1975 tetrads in the VC. Ten of the records need checking since only one voucher specimen has been made by a field worker (R.M. Barrington). A few finds were originally determined as R. peltatus subsp. pseudofluitans using An Irish Flora (1977), but we can reassign them to R. penicillatus since they carried with them information that floating leaves were absent (which could be a seasonal effect), and that the submerged leaves were longer than the internodes.

Webster (1991) looked at a wide range of subgenus Batrachium material throughout Ireland and, with the exception of one site on the River Roe in Co Londonderry (H40), she recognised all the R. penicillatus found in Ireland as belonging to subsp. penicillatus. Our Fermanagh records undoubtedly are of this subspecies and therefore we have transferred all our R. penicillatus records to this taxon. The Fermanagh tetrad distribution map shows the subspecies well represented along the Colebrooke River and scattered in several other rivers feeding into Lough Erne, most notably the Swanlinbar River, also referred to as the Claddagh River discharging into Upper Lough Erne, and the Ballycassidy River entering Lower Lough Erne.

Flowering

The white or very pale pink flowers are frequently produced in large numbers and R. penicillatus is arguably the most conspicuously beautiful of all the Batrachian Ranunculi. Although the flowers of subsp. penicillatus are usually slightly larger than those of R. peltatus, like many other characters in this subgenus, the flower size varies considerably with the environmental conditions, and in this case they do not allow reliable distinction between the two species (Cook 1966a).

The flowers are self-compatible and there is a tendency towards pollination in the bud (ie cleistogamy), so that seed is regularly set and can be independent of external pollinating agents (Cook 1966a). Flowering material of subsp. penicillatus from Ireland and Wales is reported to be very fertile, but it is usually far less so in the rivers of the English Lake District and SW England (Holmes 1980).

Anchorage in swift-flowing waters

Like other Batrachian Ranunculi species living in more or less rapidly flowing water, R. penicillatus anchors itself firmly in stream bed shingle or lightly silted gravel by means of profuse clusters of fine branching roots produced at the lower stem nodes (Sculthorpe 1967, p. 155). Unlike R. fluitans (River Water-crowfoot), which also frequents moderate- to fast-flowing waters (although this latter species is only very rarely recorded in Ireland), R. penicillatus does not form a prostrate compact overwintering stem firmly rooted to the bottom substrate. Despite lacking this specialised structure, R. penicillatus is capable of developing fresh roots from numerous lower stem internodes throughout the year, and thus anchors and maintains itself very satisfactorily, even on somewhat unstable calcareous substrata (Cook 1966a).

Finding a 'safe site' and surviving long enough to become established in a suitable habitat is generally the most hazardous phase in the life history of any plant species, terrestrial or otherwise. It is almost impossible to imagine just how minute the odds of achieving successful anchorage must be for a plant propagule in the bed of a moderate or swift-flowing stream. It is similar to the stretch in imagination required to comprehend the improbability of long-distance jump dispersal to remote oceanic islands (Carlquist 1974). Nevertheless, in both instances, somehow the extremely unlikely event does happen! In the case of aquatic macrophytes, downstream dispersal is never a problem since the water flow supplies the transporting vector. Rather the difficulty occurs in the plant propagule settling out from rapid or even moderately swift flowing water, making the initial anchorage, and rooting securely without being dislodged again by the current.

Reproductive strategy

In contrast to the fruits and seeds of emergent waterside plants, the propagules of submerged and floating-leaved flowering plants generally have no great powers of flotation. They either sink immediately after release from the parent plant, or after a few hours, or a day (Sculthorpe 1967). Clearly there must be some method of establishment, since aquatic Ranunculus species do occur and survive in relatively fast-flowing streams. Do propagules find sheltered sites in eddies behind boulders, or do they temporarily lodge in the slower flow amongst established plants or other debris lodged in the riverbed? Or do they grow out from the relative shelter of the riverbank?

Perhaps for aquatic species found in faster flowing waters, the major or maybe the only role for seed has to do with long-range or jump-dispersal from one water system to another. Most probably this involves animal vectors and, in particular, water birds (Sculthorpe 1967, pp. 331-2; Cook 1988). In swift or even moderately flowing waters, colonisation and establishment within the same (ie original) river system or water body by vegetative means involving fragmentation of an existing clone, appears very much more likely, straightforward and achievable, than from seed. The stems of R. penicillatus break at quite low forces (Haslam 1978, p. 49) and it is easy to imagine detached rafts snagging downstream on rocks, firmly 'planted' sticks or detached branches of trees, thus re-establishing the macrophyte in a fresh site. According to Haslam (1978, p. 163), such trailing fragments must be held sufficiently close to the stream bed for a month or two in order for them to root and become anchored.

In comparison with this vegetative process, it is impossible to conceive of a suitable site for seed germination occurring in a moderate or swift flow of water. Indeed, except in almost perfectly still water, it is difficult to imagine how a Ranunculus or any other seedling might produce sufficient roots to achieve anchorage.

Clonal longevity

Well established aquatic clones probably survive for many years, but they undergo vegetative cycling, the clump increasing and decreasing on an annual basis with the seasons, a process referred to by ecologists as 'wash out' (Haslam 1978; Ham et al. 1982).

Really basic research is still required on the reproductive biology and ecology of aquatic Ranunculus species in order to definitively answer the sort of questions posed here. Let us hope that airing the topics here, will help stimulate the necessary work.

British and Irish occurrence

Subspecies penicillatus is rather local and has a decidedly western distribution in Britain and Ireland. It generally occurs in Britain in rivers over base-poor, acidic igneous rocks, but in Ireland it tolerates a much wider range of alkalinity, both base-poor, pH 6.1, and base-rich, up to pH 8.5 (Webster 1988). In Britain, on the other hand, in base-rich waters subsp. penicillatus is replaced by subsp. pseudofluitans and the latter is much more widespread and abundant than the former (Webster 1991; Preston et al. 2002).

European occurrence

The distribution of R. penicillatus in Europe is even less well known than is the case in Britain and Ireland for exactly the same taxonomic and identification reasons. However, it appears to be quite widely scattered in middle European latitudes, the distribution thinning out considerably to both north and south. For instance, apart from two provinces in Denmark, it is entirely absent from the Nordic countries (Cook 1966a, Fig. 29; Jalas & Suominen 1989, Map 1892; Jonsell et al. 2001).

Names

The Latin specific epithet 'penicillatus' means, 'furnished with a tuft of hairs, like a paintbrush' (Gilbert-Carter 1964), presumably a reference to the tassel-like submerged leaves.

Population control measures

Where aquatic macrophytes including Batrachian Ranunculi become too productive and begin to form extensive mats obstructing river and stream drainage, the best method of achieving control is not to physically cut the vegetation, but to shade the banks by planting trees. If cutting does become essential, then where aquatic Ranunculus species are dominant, the most successful modification of subsequent growth is achieved when the mats are cut in full flower. This usually means cutting and dredging in June, but the timing of flowering is dependent on both site altitude and latitude (Dawson 1980; Ham et al. 1982).

Threats

Pollution of rivers and streams and over-zealous drain clearance operations.

Growth form and habitat preferences

Unlike most other aquatic Ranunculus species, R. circinatus is an easy species to recognise and ought to be well recorded. R. circinatus is a long-lived perennial of clear, base-rich, meso-eutrophic to eutrophic, still or slow-moving waters. It has no floating leaves, producing only submerged, finely dissected ones, the short, rigid segments of which are arranged in a very distinctive flat fan shape 1-2.5 cm across (Parnell & Curtis 2012). Typically, it lives rooted in mud or clay substrates, submerged in fairly deep (around 1-3 m), sheltered lake water, or in ditches in grazing marshes or fens. It is only found in shallower situations provided that the water is permanently present and never dries out. This is because unlike many other British and Irish species of Batrachian Ranunculi, it has no annual, terrestrial life-form, or only very rarely forms one on regularly flooded shorelines. Thus normally it cannot tolerate even a brief period of desiccation and therefore cannot colonise and exploit exposed bare mud (Cook 1966a; Preston & Croft 1997; Jonsell et al. 2001).

In other parts of Britain and Ireland, R. circinatus has colonised man-made still-water habitats such as reservoirs, quarry pools, gravel pits and even garden ponds (Preston & Croft 1997), but it does not appear to have done so in Fermanagh. In Baltic Scandinavia, R. circinatus is quite frequently found in brackish conditions, where it can occur down to depths of 5 m (Cook 1966a; Jonsell et al. 2001).

Fermanagh occurrence

Preston & Croft (1997) note that R. circinatus often fails to flower, or does so only sparingly, which may at least in part account for the paucity of records prior to the very thorough 1988-91 NI Lakes Survey. There were no records at all for this species in Fermanagh prior to 1977, and 22 of the 34 existing records were made by the NI Lakes Survey. This water-crowfoot has been found in a total of 22 tetrads, 4.2% of those in the VC. As the tetrad map indicates, most of the records are in Lower Lough Erne with outliers on the shores of Upper Lough Erne and at Rossgole Lough on the outskirts of Enniskillen.

Reproduction and dispersal

R. circinatus is known to flower only occasionally in Britain and Ireland, typically doing so between May and August. When flowering does occur, it appears to produce very little or no seed. With this background we may safely assume that plants or clonal colonies of R. circinatus reproduce almost exclusively by vegetative means, involving fragmentation and re-rooting of stems (Hong 1991). When autumn arrives and the main shoot is in the process of dying off at the end of the growing season, shoot tips consisting of a short length of the stem apex plus a few green capillary leaves are formed and released from the parent plant. These vegetative fragments float away and have the potential to disperse and propagate the plant (Jonsell et al. 2001). Similar stem fragments may be accidentally transported, for instance by attachment to boats or to the gear of anglers, and of course they may also disperse by attachment to water-living animals including birds (Cook 1966a; Preston & Croft 1997).

Colonising behaviour

In an interesting experiment, recolonisation by macrophytes of a stretch of river in France previously completely cleared of such plants, showed that early colonisation by R. circinatus combined two distinct strategies of invasion. Firstly there was the 'border effect', involving vegetative outgrowth from adjacent intact vegetation, while the second mechanism (without implying secondary importance to it), comprised random recruitment from detached propagules (ie viable vegetative fragments or seed diaspores). The survey technique used could not distinguish between these two forms of propagule, making it impossible to distinguish sexual from asexual reproduction (Barrat-Segretain & Amoros 1996). However, we know that in Britain and Ireland, reproduction of R. circinatus is almost exclusively vegetative.

The two strategies of aquatic colonisation that the French workers detected, mirror the terrestrial 'phalanx' and 'guerrilla' approaches described by Lovett-Doust (1981) for clonal terrestrial Ranunculus species such as R. repens (Creeping Buttercup).

Overwintering

Plants of Fan-leaved Water-crowfoot overwinter submerged as a prostrate length of branching stem lying along the bottom substrate attached by roots formed at numerous stem nodes. In the summer, the prostrate stems develop erect, unbranched stems that grow up towards the surface of the water. (Cook 1966a).

Irish occurrence

In the Republic of Ireland, R. circinatus is very thinly and widely scattered in the N and C of the country (New Atlas). North of the border in Northern Ireland, it is concentrated in three discrete areas: the limestones of Fermanagh and S Tyrone, around Lough Neagh and the adjacent Upper reaches of the River Bann, and finally on the River Quoile in E Down (NI Vascular Plant Database 2015).

British occurrence

In Great Britain, R. circinatus is quite frequent S of a line between Hull and Bristol, but it rapidly declines to rarity towards the N and W of the island.

European occurrence

R. circinatus occurs chiefly in W, C and E areas of Europe, but is very rare in the Mediterranean basin and reaches its northern extremity in both Sweden and Finland at a latitude of 63^oN in the Gulf of Bothnia (Jalas & Suominen 1989, Map 1898).

World occurrence

Elsewhere, R. circinatus is found in C and E Asia, and it, or more likely a closely related form, also occurs at similar latitudes in N America (Cook 1966a; Preston & Croft 1997; Jonsell et al. 2001).

Names

The Latin specific epithet 'circinatus' means 'circular' or 'rounded', presumably an inaccurate reference to the fan-like shape of the dissected submerged leaves (Gilbert-Carter 1964). The English common name 'Fan-leaved Water-crowfoot' gives a rather better description of the most characteristic feature of the plant, but is a pure 'book name' of no folklore merit.

Threats

R. circinatus is intolerant of both drainage and excessive cultural eutrophication and has been in decline in both Britain and Ireland, probably for 80 or more years (Grime et al. 1988; Preston & Croft 1997; C.D. Preston, in: Preston et al. 2002). In Fermanagh, further investigation will be required to discover whether the species is actually spreading at present, particularly in Lower Lough Erne, in the way that Potamogeton pectinatus (Fennel Pondweed) did, perhaps due to slow but progressive effects of cultural eutrophication. It is also possible that shortly after the population was recorded it began to decline, following the pattern that appears to be the case elsewhere.

Aquilegia vulgaris L., Columbine

Possibly both native and an introduced, neophyte, garden escape; occasional. European temperate, but introduced in N America and New Zealand and part of a circumpolar species complex.

1899; Tetley, W.N.; Dunbar House, Fintonagh Td.

January to September.

Growth form

Aquilegia vulgaris is a tall, erect rosette-forming perennial with a short, stout, erect, rather woody, perennating rootstock. According to Clapham et al. (1962), the rootstock is often branched, yet the plant appears to produce only one rosette of basal leaves and a solitary flowering stem which can reach 100 cm, but more usually measures about 60 cm tall. The plant does not possess any powers of vegetative reproduction, relying entirely on seed for population increase and dispersal.

Native and garden forms and their preferred habitats

Throughout Britain and Ireland, native populations typically grow in woodland glades and fairly open scrub, by woodland rides and streamsides, in damp stony grassland and fen and on scree slopes. Garden escapes tend to be naturalised in more disturbed or man-made sites, eg in quarries, on roadsides and railway banks and on or near old walls. Sometimes, the colour, size or hybrid form of the plant makes it obviously a garden escape, but on other occasions in Fermanagh, as for instance on Knockninny or at Carrickreagh, the typical blue or deep purple form of the species occurs and it appears to be a well established member of a stable, semi-natural plant community. It is then probably best to consider it native.

Columbine is generally associated with limestone rock and calcium- or base-rich soils, though at Lisblake in Fermanagh it grows on acidic peaty roadside banks. It is frost tolerant down to around -25C and can be thought of as a plant of cool conditions, generally preferring deciduous woodland or hedgerow shade and a moisture retentive, though not wet, soil. In gardens, it does not tolerate heavy clay soils (Huxley et al. 1992).

As with Hyacinthoides non-scripta (Bluebell) and other species, it is generally assumed to tolerate and prefer shade conditions; under the frequent grey, cloudy skies that are the norm in western Ireland during much of the growing season, A. vulgaris also grows both in the open on rocky pastures and in disturbed habitats, eg on roadsides and around buildings, as well as occupying partial shade in more natural relatively undisturbed habitats, such as Hazel scrub and lakeshore woodland.

Fermanagh occurrence

A. vulgaris is rare or occasional in Fermanagh, typically occurring either in very sparse populations or as isolated individuals. It has been recorded in a total of 26 tetrads, 4.9% of those in the VC. As the distribution map indicates, it is widely and thinly scattered in the lowlands. Local habitats include woods, shaded banks and lakeshores, open limestone rocks including quarries, roadsides, a churchyard and a graveyard.

Longevity and persistence of garden escapes

Observation shows that in Britain and Ireland some plants are ephemeral while others definitely persist for many years. Garden forms are generally considered short-lived, perhaps surviving two or three years in cultivation (Grey-Wilson 1989). The longest persistence known to us locally in Fermanagh is the stand at Carrickreagh woods on the shore of Lower Lough Erne, which was first recorded by Carruthers over 55 years ago and is still there (Forbes & Northridge 2012).

Increasing occurrence in Britain and Ireland

The New Atlas survey completed in 2000 found that A. vulgaris had increased in Great Britain and Ireland in the 40 years since the original Botanical Society of the British Isles Atlas (Walters & Perring 1962), with a calculated Change Index of +1.70. The change in species presence may be attributed to a combination of better recording and a presumed increase in the frequency of the plant escaping from garden cultivation (R.A. Fitzgerald, in: Preston et al. 2002). The editors of the New Atlas recognise that the native distribution of the species is now totally obscured, and they therefore mapped all A. vulgaris records as if they were native.

Flowering

Flowers are produced in May and June and copious nectar, secreted in five long hooked spurs, attracts bumblebees with sufficiently long tongues as pollinators (Garrard & Streeter 1983). In the absence of pollinators, the plant carries out self-pollination. The wild form of the plant (often blue or purple in colour) has no fragrance, but white forms and other garden varieties are often sweetly perfumed, the scent being reminiscent of cloves. This is probably due to the presence of the genes of A. pyrenaica and other perfumed species, being incorporated through the efforts of plant breeders.

Although the flowers are pendulous, after fertilisation the follicular fruits are held stiffly erect. When ripe, the four or five follicles split open at the top of their inner sides so that the many black, shiny, smooth seeds are shaken out when the wind is strong enough to sway and jerk the tall fruiting branch like a censer (Melderis & Bangerter 1955).

Seed and germination

As is often the case, reproductive performance statistics do not appear to be available in the literature for the species, but the individual seeds are quite large (2-2.5 mm), their average weight varying from 0.78 mg (Grime et al. 1981) to 1.95 mg (Salisbury 1942). In germination tests, only 4% of freshly collected seed germinated, indicating that such seed is dormant, perhaps immature, and requires stratification. After dry storage for six months chilled at 5C, mimicking the overwintering stratification process, only 30% of seed germinated (Grime et al. 1981).

Dispersal

Given the limited height of the fruit above ground and the passive release of the wind dispersed seeds which have no appendage to assist flight, nor any form of edible secondary dispersal lure to entice ants or other animal vectors, one would not expect the seeds to travel from the parent plant more than a very few metres at most. However, A. vulgaris regularly manages to escape into the wild from gardens, indicating that it does possess considerable powers of mobility.

Further studies required

An obvious need exists for population and reproductive ecology studies of this familiar species, since a deep well of ignorance persists regarding the behaviour of the plant. Even the question of survival in the soil seed bank does not appear to have been investigated as the species does not feature in the major survey of this topic in NW Europe (Thompson et al. 1997).

Irish and British occurrence

The distribution of A. vulgaris in the New Atlas shows the species (wild and introduced records combined) widely but fairly thinly scattered in Ireland and nowhere common (Webb et al. 1996). In Great Britain, the map indicates a much greater presence in the S and W, although again it is very widely scattered throughout England and Wales, but thinning markedly towards the north and especially so in Scotland (Preston et al. 2002).

European and World occurrence

Beyond the British Isles, opinions on the taxonomic circumscription of A. vulgaris sens lat. are far from settled, with some authorities including A. atrata and A. nigricans, and others not (Jalas & Suominen 1989). Taking it in the broad sense, A. vulgaris s.l. is considered native in much of W and C Europe, but is regarded as a naturalised introduction in Holland, N Germany and the Nordic countries (Jalas & Suominen 1989, Map 1907). A. vulgaris s.l. is also regarded as native in N Africa, and it is naturalised widely in N America and in temperate areas of the S Hemisphere including New Zealand (Jonsell et al. 2001, p. 312). Together with closely related species, sometimes considered subspecies or varieties, it forms a circumpolar polymorphic complex (Hultén & Fries 1986, Map 879).

Toxins

Aquilegia species are reported to contain toxic alkaloids similar to those in the genus Aconitum (Monk's-hood), ie the cyanogenic glucoside triglochinin (Tjon Sie Fat 1979). However, there is no evidence of recent poisoning by any form of Columbine in Britain and Ireland (Cooper & Johnson 1998). The toxins contained in A. vulgaris are destroyed by heating or drying.

Medicinal and folk uses

The species was formerly employed in herbal medicine, mainly for its antiscorbutic effect. It has fallen out of favour and is little used nowadays, although the root and leaf sap are still sometimes used externally in poultices to treat ulcers, swellings and the commoner skin diseases (Grieve 1931; Vickery 1995). However, a homeopathic remedy is made from the plant and is used in the treatment of nervous afflictions including hysteria. In the past, dried and crushed Columbine seed was used as a parasiticide to rid the hair of lice and to kill external body parasites (Plants for a Future Database Website http://www.pfaf.org/ accessed 2016).

The flowers are so well supplied with nectar they were used to make a tea and were also eaten in salads. However, since the plant contains poisons it is not recommended for any such purpose (Mabey 1972).

Folklore reports that in the Middle Ages, Columbine was believed to be the food of lions, and those who rubbed the sap of the plant on their hands became gifted with the courage of a lion (Grigson 1987).

Names

The origin of the genus name 'Aquilegia' is doubtful but is sometimes said to be derived from the Latin 'aquila' meaning 'an eagle', the flower spur supposedly resembling an eagle's claw (Johnson & Smith 1946). The Latin specific epithet 'vulgaris' means 'common', which the species is not, at least in the wild in Ireland. The English common name 'Columbine' is derived from the Latin 'columba' meaning 'doves'. It alludes to the appearance (especially of the short-spurred form of the plant) which looks like five doves drinking (Stearn 1992).

A large number of alternative common names exist in folklore. Grigson (1987) lists 35 of these names, many of which compare the flower shape to that of a style of bonnet worn by elderly ladies, for instance 'Granny's bonnet' and 'Old Woman's nightcap'. Other folk names refer to various items of ladies' wear, including shoes, slippers and petticoats, and a few of male apparel, eg 'Batchelor's buttons' and 'Rags and Tatters' (Grigson 1987).

Threats

None.

Growth form and preferred habitats

T. flavum is a rhizomatous perennial of low lying, moist to wet, near-neutral, moderately fertile meso- to eutrophic, base-rich, generally calcareous grasslands. In Fermanagh, some of these particular grasslands were previously used as hay-meadows although they really are rather damp for this purpose. However, the species is also capable of colonising fen sedge peat soils, ie peat formed from the partial decomposition of sedge and herb litter. This type of organic soil has a significantly higher nutrient and mineral content than the much more acid moss peat found in bogland.

The plant possesses a long, creeping, branching yellowish rhizome from which stout, annual, unbranched, furrowed stems arise, and clonal patches can develop if good growing conditions are met (Melderis & Bangerter 1955; Clapham et al. 1962; Hill et al. 1999). Common Meadow-rue generally grows in full sun, but occasionally it is found in light shade.

The erect habit and the fact that its ultimate leaflets are much longer than wide readily distinguish T. flavum from T. minus. In Fermanagh, T. flavum is often on the small side of its size range, only around 50-60 cm tall unless it is forced to compete with taller, ranker vegetation, when it can become extremely lax and straggly. The plant can become so enmeshed with the surrounding, competing, grassy herbage, that it becomes difficult to spot, even when one knows it is present.

Flowering and pollination

The cream to yellowish flowers, which are produced between June and August, are largely composed of the slender filaments of the stamens. In Fermanagh, the stamens are sometimes much more sparse than normal, which again makes the plant easily overlooked even when in flower. The bunches of erect yellowish stamens attract small flies and bees (both bumble-bees and honey-bees). They manage to attract insects without the advantages of nectar, perfume or petals to advertise and reward the visitors; they simply collect and feed on the plentiful supply of nutritious pollen (Proctor & Yeo 1973, p. 60; Garrard & Streeter 1983). Some authorities regard the flowers as fragrant, others not, so there might well be a measure of geographical variation in this particular respect. But then every nose is different! (Proctor &Yeo 1973; Jonsell et al. 2001).

In addition to pollination induced by movements of the insect visitors, very probably some degree of wind pollination of the reduced flowers also takes place (Clapham et al. 1962).

Fruit

Each small flower of the more or less dense branched panicle inflorescence contains 3–12 free carpels (some of which are often abortive). The fruit is a small cluster of single-seeded, almost spherical, ribbed achenes (ie single seeded dry fruits) (Butcher 1961; Hutchinson 1972).

Seed survival

There is a solitary estimate of the soil seed bank longevity of the species in the NW Europe survey (Thompson et al. 1997). This suggests the seed is long-persistent (ie buried seed may persist for at least five years).

Conservation field studies

Information is lacking or difficult to locate on such matters as the scale of seed production, population dynamics, the balance of vegetative versus sexual reproduction, or the competitive ability of the species, studies of which would assist any serious attempts to conserve this declining plant species. The various bodies involved in species conservation in both Ireland and Britain give the impression (deserved or not) that attending meetings, drawing up reports, endlessly revising species lists, drawing boundary lines on maps and arguing for resources are all that they are required to do in order to effectively manage threatened wild populations of plants and animals – work they regard as a business like any other (Marren 2002, p. 307). One wonders when the basic natural history studies, let alone the detailed ecological field studies and localised site observations and analysis that should inform and direct active conservation management are going to be carried out. The conservation industry, if we can call it that, appears much more concerned with words than deeds. Hands and boots need to get dirty to really make a change in species fortunes on the ground.

Fermanagh occurrence

T. flavum occurs very rarely in fens, ditches in wet meadows and in marshy grasslands on lake shores. It has been recorded in a total of just nine Fermanagh tetrads, seven of which have post-1975 records. As the tetrad distribution map indicates, it currently occupies six stations on both shores of the southern half of Upper Lough Erne, but has not been refound at either of the two earliest sites in the VC which lie closer to Enniskillen. In 1996, an interesting new station, quite remote from the seven previous ones, was discovered by a survey team from the EHS on the shore of another of the larger lakes in the county, at the NW end of Lough Melvin. In July 2010, another new site was discovered by RHN and HJN at Derrymacrow Lough, where a solitary plant grew alongside Thelypteris palustris (Marsh Fern).

The other record details are: in damp meadow, W of Lough Digh, 1950, MCM & D; Lough Corby, 3 km W of Newtownbutler, 1983, RHN, also 6 June 2002, RHN, RSF, J.S. Faulkner & I. McNeill, and 11 July 2010, RHN; Drummully Td and Croostan Td shores, 2 km NE of Derrylin, 23 July 1986, P. Corbett & P.J.T. Brain; Derryad Td shore, 13 August 1986, S.A. Wolfe-Murphy & L.W. Austin; S shore of Trasna Island, 19 August 1986, T. Waterman & P.J.T. Brain; SE corner of Corraharra Lough, Mullynacoagh Td, 30 June 1991, RHN, also 15 August 1994, RHN & RSF, 25 July 2004, RHN, and 8 July 2010, RHN & HJN; S shore of Derrymacrow Lough, 11 July 2010, RHN & HJN; shore of Lough Melvin, NW of Garvos, opposite the end of Bilberry Island, 26 June 1996, EHS Habitat Survey Team.

Irish occurrence

The Fermanagh records represent the only nine stations (seven extant) for this rare perennial plant in Northern Ireland outside the Lough Neagh and Lough Beg basin and the connected River Bann drainage system. Even in the latter areas, where once it was frequent, T. flavum is now regarded as being on the decline due to drainage and widespread changes in and intensification of agricultural practices (Harron 1986; Hackney et al. 1992).

In the Republic of Ireland, T. flavum is widely but very thinly scattered, (mainly along the course of the River Shannon) and it is locally frequent in a few central areas of the island (Webb et al. 1996; Preston et al. 2002).

British occurrence

In Britain, the distribution is greatly concentrated to the SE of a line between Swansea and Whitby with a few rare outliers, chiefly coastal, stretching as far north as Edinburgh in the east and Kintyre on the west coast (Preston et al. 2002). With the decline of the species on agricultural land, in many areas of Great Britain, Common Meadow-rue is no longer common. Rather, it has become restricted to relict land, often occurring along linear landscape features such as river banks and roadside ditches (R.A. Fitzgerald, in: Preston et al. 2002).

Significant local factors affecting the widespread decline of Common Meadow-rue in Britain and Ireland include the almost universal move away from hay to silage production, the widespread use of herbicides and the ploughing and reseeding of pastures, very often subsequently fertilised by slurry spraying and/or agrochemical application. These measures have fundamentally altered grassland ecology, increasing production and heightening interspecific competition in ways that force some plants towards local extinction.

Fossil record

The fossil record of T. flavum is almost exclusively based on identification of their characteristic and abundant achenes. Pollen grains can really only be distinguished to the genus level, but the fruits do prove that T. flavum has persisted in the British Isles throughout the last four interglacial periods and the last two glacial periods − at least in the southern part of England and Wales (Godwin 1975).

European occurrence

In Europe, T. flavum is widespread throughout N, W and C regions, stretching N to within the Scandinavian Arctic Circle and S to the toe of Italy and to the Spanish pre-Pyrenees, but the distribution thins markedly further E and it is absent from Greece and all the Mediterranean islands (Jalas & Suominen 1989; Jonsell et al. 2001).

Uses

T. flavum is quite a popular garden border perennial, especially a very fine, large, robust variety called 'Illuminator' (Griffiths 1994). Despite this, there is very little evidence of it as a garden escape in the British Isles, certainly in comparison with several other members of the family (Preston et al. 2002).

The leaves of T. flavum in the past were used to allay fevers, and the root was sometimes used in dyeing (Melderis & Bangerter 1955); supporting evidence for these uses of the plant has not been located. Rather surprisingly, Grieve (1931) makes no mention of the species whatsoever in her very comprehensive herbal.

Toxins

In addition to containing the toxin protoanemonin, common to all members of the Ranunculaceae (Cooper & Johnson 1998), T. flavum contains at least six alkaloids, the principal one being berberine (Velcheva et al. 1992). There is a tremendous amount of current pharmaceutical research going on involving European species of Thalictrum. RSF found that 80 out of 81 scientific papers that were located by a major Internet database searching back as far as 1981 were on this topic alone.

Names and folklore

The genus name 'Thalictrum' is an ancient Classical Greek name first given by Dioscorides to a member of the genus, possibly in the T. minus species aggregate (Gilbert-Carter 1964). The name may be linked to the Greek 'thallo', meaning 'to flourish' or 'to abound in', perhaps a reference to the numerous bundles of stamens in the inflorescence (Johnson & Smith 1946). The Latin specific epithet 'flavum' is straightforward by comparison, translating as 'yellow'.

The plant has four English common names in Britten & Holland (1886), of which 'Meadow-rue' is the most well known, being given on account of the finely divided rue-like leaves. The suggestion made by Gerard (1633, p. 1252) is that some old herbalists confused this species with Ruta graveolens (Rue) which had medicinal uses. Indeed, T. flavum was previously called 'Ruta palustris', or 'Fen Rue', from the appearance of its leaves and its typical place of growth (Prior 1879, p. 77).

The two remaining names in Britten & Holland (1886), refer to T. flavum as 'False Rhubarb' and 'Meadow Rhubarb'. According to Lyte (1578), this was because like Rhubarb the plant has laxative properties and also because it has roots that are yellow like those of Rhubarb! It was also called 'Great Bastard Rhubarb', or 'English Rhubarb' (Gerard 1597, 1633), and it was supposed to cure old ulcers, be good for the belly and act as an astringent 'without biting' (Gerard 1633, p. 1252).

Threats

The Corraharra site could well become overgrown by conifers and the Lough Corby site has already been damaged by drainage.

Growth form and habitat preferences

T. minus is a extremely variable, clump-forming stoloniferous perennial species or species aggregate, the differing growth forms being produced by the orientation and length of its rhizome, which can be either ascending and short, or horizontal and long (Clapham et al. 1987). The plant produces erect stems, variable in form and colour and up to 120 cm tall, but in Fermanagh often much less, most typically around 50-60 cm in height. Variation in the species also extends to the wide variety of habitats it occupies. Again, in a local context, these range from fairly dry to quite moist, lime- or base-rich conditions. The three or four times pinnately divided leaves are also very variable, but the ultimate leaflets are generally about as long as broad, a fact which clearly distinguishes the species from T. flavum, which is much less finely divided and generally grows under quite different ecological conditions. In comparison, T. flavum is a plant of very much wetter and more open sun-lit, marshy grasslands and fen-like habitats.

In Great Britain, T. minus is also represented in at least three rather different types of habitat that are identified by Clapham et al. (1962), as: i. limestone rocks and grassland; ii. coastal dunes; and iii. streamside or lakeshore gravel and shingle. To the first of these we should also add cliff ledges, screes and the gryke (or grike) crevices of limestone pavement (Halliday 1997).

Fermanagh occurrence

In Fermanagh, T. minus has been recorded in a total of eleven tetrads (2.1%), eight of which have post-1975 records. It is chiefly distributed around Lough Erne and Lough Melvin and along stretches of the Colebrooke River. In Fermanagh, T. minus typically occurs as very small populations of just a handful of scattered individuals, or as isolated clumps growing in fairly dry, rocky or sandy lakeshore situations, or in similar soils on river banks under the shelter and shade of scrub. Unless active conservation measures are taken to reverse the present decline, these small, fragmented populations represent the end of the line for T. minus in Fermanagh. Small, isolated populations inevitably suffer from poor cross-pollination, consequent inbreeding depression and a loss of fertility. Undoubtedly all such rare and scattered species populations are on the downhill slope towards local extinction at some future date. In the case of T. minus, the existence of some degree of apomixis may delay matters for a while, but eventually fragmented species like it are fated to fade out and disappear.

Variation

The New Flora of the British Isles (Stace 1997) accurately describes the species or species aggregate as, "very variable and little understood". Stace also points out that while up to eight subspecies or other rank of taxa have been described on the basis of fruit, habit and hairs, the variation has not been properly studied. Until it is, his opinion chimes with my own, "these variants are not worth recognising". Matters are complicated by the fact that some races within the species or species aggregate appear to be apomictic and can thus set asexual seed, which is a rather sophisticated form of vegetative reproduction.

By whatever means the plant manages to reproduce, once it becomes established in a site T. minus becomes very persistent. It still survives, for instance, on the banks of the Colebrooke River where Smith found it and at Gubbaroe Point in crevices in the limestone pavement shore of Lower Lough Erne, two sites where it was originally found in the second half of the 19th century.

Irish occurrence

T. minus used to have other inland stations in Northern Ireland, but with the exception of the Fermanagh records and a few stations in the Mourne Mountains, Co Down (H38), practically all the remaining Northern Irish stations for the species are on coastal sand dunes, where they are sometimes distinguished as subsp. arenarium (Clapham et al. 1987; FNEI 3).

In the Republic of Ireland, T. minus is confined either to the coast or to the shores of larger lakes in the west and in the Central Plain, with a few additional stations, mainly in the mountains of Cos Sligo (H28), Leitrim (H29) and the Connemara region in the mid-west of the country. In these mountain stations, Lesser Meadow-rue occupies rocky ground and mountain ledges where it can avoid intensive plant competition; at the same time it is less accessible to browsing sheep (Walters & Perring 1962; Preston et al. 2002).

Toxicity and medicinal uses

Browsing this herb must be a mixed pleasure for stock animals, since like other members of the Ranunculaceae T. minus contains the bitter tasting and highly irritant toxin, protoanemonin (Cooper & Johnson 1998). In common with T. flavum, T. minus is the subject of a great deal of current pharmaceutical research since both species contain several additional poisonous alkaloids, including berberine, although thalactamine is the main ingredient in the toxin cocktail in T. minus. Berberine is used in modern medicine to treat stomach and gall bladder ailments (Urmantseva et al. 2000).

Garden cultivation and occurrence in Great Britain

T. minus has also had a long career in Britain and Ireland as a garden border perennial. It is certainly not grown for its flowers, but rather for its attractive finely dissected foliage which is greatly appreciated by florists and flower arrangers (Robinson 1909, p. 888). Due to these widespread garden populations, T. minus also regularly jumps the garden wall and occurs as an established introduction. In Great Britain, these garden escapes sometimes become intermixed with supposed native plants, mainly in the Midlands, S & SE England and in Wales, although it can perhaps more rarely occur in this way as far north as Glasgow (Preston et al. 2002).

European occurrence

T. minus is widely but unevenly spread throughout continental Europe, being rather less frequent on the Atlantic coastline than the extent of the species or aggregate distribution in the British Isles would suggest. It occurs up to 70N in Scandinavia but is absent from much of the rest of Norway, Sweden and Finland, and, indeed, it appears to have declined in the latter two countries. On the Iberian Peninsula, T. minus penetrates to the southern Spanish Sierras and it also occurs rarely on Corsica, Sardinia and Sicily, yet is absent from the Peloponnese, Crete and the smaller Mediterranean islands (Jalas & Suominen 1989).

World occurrence

Beyond Europe, the distribution of T. minus stretches to N, E and S Asia, Ethiopia, S Africa and Alaska (Jonsell et al. 2001).

Names

The Latin specific epithet 'minus' is a comparative of 'parvus', and means 'smaller than', presumably smaller than T. flavum, which generally is the case (Gilbert-Carter 1964). The plant appears to have been insufficiently known or was ignored as being of little or no interest or use, so it does not appear to have accumulated much in the way of English common or local names, apart from 'Lesser Meadow-rue' (see T. flavum species account). Gerard (1633, pp. 1251-2), however, distinguishes it from T. flavum as, 'Small Bastard Rhubarb', and for an explanation please again see the T. flavum account.

Threats

The Colebrooke River sites could easily be destroyed by river bank improvements and the Muckross site by tourist developments.

Growth form and preferred habitats

The deciduous, sharp spiny shrub B. vulgaris typically grows in hedgerows, scrub banks, coppice woodland and on waste ground. Being a very spiny shrub, at quite a late stage in history, Barberry was widely planted for hedging.

Secondary host of plant pathogen

Barberry happens to be the secondary host plant of Puccinia graminis, the pathogenic fungus which causes black stem rust of wheat and other grasses. When this was realised in 1865, Barberry was eradicated in some countries (eg Denmark and in parts of N America), thus providing some degree of cereal disease control. According to Rackham (1986, pp. 42-3), Barberry is not necessary to the life-cycle of the fungus in England (and presumably elsewhere in Britain and Ireland). Although farmers have from time-to-time removed the shrub from field hedgerows, Rackham (1986), who appears to think chiefly or entirely in terms of S England, considers it unlikely that B. vulgaris was ever common, nor was it much diminished by these intermittent eradication measures.

Flowering

The drooping bunches of small lemon-yellow flowers which appear in May and June are sweetly and delicately scented, and their abundant nectar readily attracts bees and flies as pollinators (Genders 1971; Lang 1987). Interestingly, the six anther filaments in each flower display an instantaneous reaction, moving inwards when touched by an insect visitor and thus dabbing pollen on to it, ie they are not just elastic and under tension, but rather they are tactile or 'touchy', a feature we associate very much more with animals than with plants! In his book The action plant, Simons (1992) deals with the phenomenon of flower 'irritability' in great detail. He describes the mechanism of the reversible touch-sensitive stamens of Berberis involving biochemistry and transferred electrical impulses and motor cells, exactly comparable to these processes in animal nerve tissue but, of course, minus the actual nerve cells (Simons 1992, pp. 35-39 & 240).

Fruit dispersal

The red berries attract birds and at least nine species have been recorded eating them (Lang 1987), chiefly members of the thrush and crow families (Ridley 1930). They are, however, very acid, and apparently are not popular with them for this reason, but cattle, sheep and goats are known to eat them, and in N America after the shrub was introduced by settlers, stock animals are reputed to have dispersed the seed with their dung (Ridley 1930, p. 365).

Culinary, medicinal and other uses

Barberry was a much cultivated shrub in medieval times, the ripe red, egg-shaped berries borne in clusters of about ten, making a delicious, if somewhat tart jam or dessert jelly. The berries were also candied and eaten as sweets and they were used to flavour punch. The bark has an unforgettable bright yellow 'blaze' on its interior when cut with a knife. It was used to tan leather and to colour it yellow. Above all else, in terms of uses for the plant, the yellow inner bark of both stem and root was medicinally prized as a cure for jaundice following the well known 'Doctrine of Signs' or 'Signatures' (Grigson 1987; Vickery 1995). The Berberis jaundice cure was also extended in unspecified areas of Ireland from humans to cattle (Allen & Hatfield 2004).

Toxicity

While the ripe berries are edible and contain high levels of vitamin C, unripe they are distinctly poisonous, containing the alkaloid berberine and at least three other toxins. The highest content of berberine is actually present in the bark, but the seed (especially that of related cultivated species and varieties), should also be avoided. Berberine (which is also present in species of Thalictrum) is an antibacterial gastric and mucosal irritant, causing vomiting and diarrhoea if large quantities are ingested. Berberine can depress respiration by acting on the central nervous system, and it can also cause uterine contractions. THUS IT IS DANGEROUS TO EAT THE BERRIES, AND ESPECIALLY SO FOR PREGNANT WOMEN (Lang 1987). Having said this, no authenticated cases of either animal or human poisoning by any part of the plant were uncovered by Cooper & Johnson (1998).

Barberry it is still very much used in herbal medicine and homeopathy for a wide range of complaints (see below). In its favour, Cooper & Johnson (1998) report that the pharmacological activity of berberine to treat some cancer tumours is currently being investigated.

Fermanagh occurrence

There have only ever been four stations recorded for the plant in Fermanagh. The one found by Meikle and co-workers at Gubbaroe Point on the shore of Lower Lough Erne very probably fell victim to forestry plantation operations in the vicinity soon after its discovery. The details of other three more recent stations are: hedge E of Clonagore Td, near the old Ulster Canal, 23 August 1996, RHN; hedge at Clonelly, NW of Kesh, 17 March 1999, RHN & HJN; and several large bushes in front of old castle at Castle Caldwell, 11 November 2006, RHN.

British and Irish occurrence and status

For all of the above reasons, in Ireland, B. vulgaris is regarded (at least probably) as a deliberately planted introduction, whereas in Britain it might either be a native or an archaeophyte (T.D. Dines, in: Preston et al. 2002). In Britain, the 1962 BSBI Atlas and its 1976 revised edition, recorded B. vulgaris as being quite a frequent and widespread species. Indeed, the New Atlas displays the plant as even more widespread in Britain than the earlier BSBI Atlas map, thinning out now only N of Glasgow and Edinburgh (BSBI Atlas; New Atlas).

Formerly in Ireland, Barberry was much more frequent, but it is now very rare and widely scattered, having been most successfully extirpated for disease prevention during the 19th century (Cat Alien Pl Ir).

European and World occurrence

Beyond the British Isles, B. vulgaris is considered native in C Europe by Jonsell et al. (2001), but it also extends over most other temperate parts of the continent, although absent in the far north and around the Mediterranean (Lang 1987). Other botanists consider it a native of Asia's middle and western mountains (Royer & Dickenson 1999). I believe it is best considered an archaeophyte in most of W Europe (ie it is an ancient or very long-standing introduction). Grieve (1931) regarded the range of the species included N Africa, but the N Hemisphere distribution published by Hultén & Fries (1986) does not support this idea (Hultén & Fries 1986, Map 886). In all these regions, as in Britain and Ireland, it is difficult to distinguish where the shrub is naturalised, from areas where it is supposed or imagined to be native. Barberry is introduced in N America, and it has spread there from coast to coast. It has also been recorded in New Zealand.

Names

The genus name 'Berberis' is thought by some to be the Latinised form of the Arabic word for the fruit of the plant, 'berberys' (Chicheley Plowden 1972; Stearn 1992), in some way signifying a shell (Grieve 1931). Other writers consider it simply a geographical reference to the Barbery coast of N Africa (eg Gledhill 1985). Gilbert-Carter (1964) plays it safe, regarding the name as, "a medieval Latin word of doubtful origin". The Latin specific epithet, 'vulgaris', meaning 'common' is certainly inappropriate in Ireland, and perhaps this is becoming the case elsewhere since the species is regarded as declining in Great Britain (Change Index calculated as -0.61) (T.D. Dines, in: Preston et al. 2002).

There are at least nine English common names for the plant, several referring to its supposed jaundice cure mentioned above, eg 'Jaundice Tree', 'Jaunders Tree' and 'Jaunders Berry' (Grigson 1987). There are records of this jaundice cure from all over Britain and Ireland (Allen & Hatfield 2004).

The name 'Barberry', which we have already mentioned in connection with its Latinised genus name, has several local variants around the British Isles, for example, 'Berberry', Barbaryn', 'Barboranne' and 'Berber'. The remarkable yellow inner bark and wood is undoubtedly the origin of the name 'Guild' or 'Guild Tree', while 'Woodsour', 'Woodsore' and 'Woodsower' are applied to both B. vulgaris and Oxalis acetosella (Wood-sorrel) on account of their woodland habitat and sour taste (Britten & Holland 1886).

The remaining suite of about eight variant names centres on 'Piperidge' and includes 'Pipricks', 'Piprage' and 'Pepperidge'. Prior (1879) reckons that this is derived from the French 'pepin', a pip and 'rouge', red, a reference to the fruit which he considers rather small and lacking in juice, and hence a pip rather than a berry!

Threats

Deliberate destruction in the mistaken belief that it is necessary to do so to limit the spread of cereal rust.

Growth form and preferred habitat

This is the only poppy, other than the Welsh Poppy (Meconopsis cambrica), which is in any way frequent in Fermanagh. It is a distinctive, tall, blue-green, fleshy summer annual of lowland, ruderal habits, most typical of roadsides, waste ground and rubbish tips, but occasional also in disturbed soil in gardens. In the latter case, it may have been accidently imported with container plants purchased from garden centres. It flowers in July and August, but even when not bearing its large white, lilac or red terminal flowers, often with a basal purple blotch on each petal, this poppy is very easily identifiable. The only plant it might possibly be confused with is Glaucium flavum (Yellow Horned Poppy) but since it is a strictly maritime plant (certainly with no inland stations in Ireland (Preston et al. 2002)), and Fermanagh has no coastline, we do not meet with any such problem.

Like the other 'cornfield' red poppies, P. somniferum is principally found in light, dry to moist but typically well-drained soils, always in disturbed habitats. In Britain and Ireland, it is always a weed of cultivation, a garden escape or discard (sometimes double-petalled), or a bird-seed introduction (Clement & Foster 1994; Cat Alien Pl Ir).

There are two forms of opium poppy, often but not always given subspecies rank as subsp. somniferum and subsp. setigerum (DC.) Arcang. The former is the common, usual form, while subsp. setigerum, which is more conspicuously bristly on its stems, leaves and sepals, is a rare casual only in Britain and Ireland (Stace 1997).

Fermanagh occurrence

In Fermanagh, P. somniferum has been recorded in a total of 28 tetrads, 5.3% of those in the VC. Twenty-three Fermanagh tetrads contain post-1975 records. It is confined to the lowlands of the county and is found especially in and around the Enniskillen and Maguiresbridge 'conurbations', although both are hardly large enough to be called such. Previously, when Fermanagh still had railway transport connections, it grew along lines and in stations, but now Opium Poppy is only found as a weedy escape along roadsides, in or near gardens and on disturbed or waste ground.

Irish occurrence

While P. somniferum has been recorded at least once in 30 of the 40 Irish VCs (Cen Cat Fl Ir 2; Cat Alien Pl Ir), the New Atlas hectad map shows it is much more frequently recorded in N Ireland than further south. Apart from a concentration around Co Dublin (H21), it appears very thinly scattered across the Republic of Ireland (New Atlas). I regard this Irish distribution pattern purely as a reflection of recorder hours in the field. The restricted distribution in northern Britain and to lower ground everywhere is undoubtedly due to the susceptibility of young plants to late season frost.

British occurrence

The New Altas distribution shows P. somniferum as a frequent and widespread casual in lowland areas of Great Britain, mainly occurring S of a line between Anglesey and Hull, and becoming more confined

to coastal habitats further north. The restricted distribution in the north of these isles, and to lower ground everywhere, is undoubtedly due to young plants being susceptible to late season frost.

Flowering

The flowers are self-compatible and pollination probably takes place in the bud. Again like other widespread poppies, P. somniferum is phenotypically very plastic and it has a high seed output (a mean of between 6,000 and 7,000 seeds/capsule). The seed is long-persistent in the soil seed bank (Salisbury 1964; Jonsell et al. 2001).

History of cultivation and uses

P. somniferum has been in cultivation for medicinal, culinary and decorative purposes since ancient prehistory, and images have been preserved from the civilisation of ancient Sumeria (4000 BC) depicting the poppy along side other images indicating euphoria. Sumerian knowledge of the plant was passed via the Assyrians and Babylonians to the Egyptians, who around 1300 BC began cultivating P. somniferum var. album, referred to as the 'Opium Thebaicum', 'The White Lotus'. The Late Bronze Age Mediterranean trade routes of the Phoenicians and Minoans moved this profitable item from Egypt to Greece, Carthage and Europe. Figurines of Late Minoan and Mycenaean goddesses dating from about 1350 BC, show them wearing poppy decorated crowns (Hood 1978, p. 109; Baumann 1993, p. 69).

Writing sometime between c 460-377 BC, Hippocrates, the Greek father of medicine, dismissed the magical attributes of opium. However, he acknowledged its usefulness as a narcotic and a styptic in treating internal diseases, the ailments of women and epidemics. By 300 BC, however, opium was being widely used by Arabs, Greeks and Romans, both as a sedative and as a soporific.

For many years, the Christian church and the associated state authorities banned opium, probably in part because of its 'devilish' eastern origins and, rather surprisingly, it fell into disuse. In 1527 AD, during the height of the European Reformation, opium was reintroduced into medical literature by the Swiss alchemist and physician Paracelsus (1493-1541). He very cleverly renamed the drug 'Laudanum' (Tincture of Opium in alcohol), and thus circumvented church and state disapproval. The black pills or 'Stones of Immortality' Paracelsus prescribed were a compound of opium, citrus juice and essence of gold, and they were used as painkillers (Webpage 'Opium throughout history' at: http://www.pbs.org/wgbh/pages/frontline/shows/heroin/etc/history.html, accessed April 2016).

The Romans are reputed to have sprinkled the top of their bread with the nutty flavoured poppy seeds, the only part of the plant which contains no, or very little alkaloid (Le Strange 1977).

Grieve (1931) provides much information on both cultivating the plant and the history of its herbal medicinal uses. Rather surprisingly there does not appear to be any great fund of plant lore associated with this interesting, long cultivated and heavily used plant (Vickery 1995).

Nowadays, P. somniferum is a frequent escape from cultivation throughout its range, where it is usually grown either for oil-seed, or as a pharmaceutical crop plant for the latex from its unripe capsules. In past years, most recently in the period between 1880 and 1930 when agriculture was suffering one of its periodic cycles of decline in the British Isles, a variety with pale lilac flowers and whitish seeds was grown for opium in England, chiefly in Lincolnshire (Grieve 1931; Thirsk 1997).

However, and in spite of this, in most of the British Isles P. somniferum contains very little alkaloid due to our less than ideal climate (too little sunshine) for the species. It is mainly grown, therefore, as a garden ornamental of some minor culinary and herbal medicinal use. Nine garden varieties are listed, and these and others illustrated in Phillips & Rix (1999) book, Annuals and biennials. The garden or cultivated form is subsp. setigerum, occasionally referred to as subsp. hortense (Hussenot) Corb., but not so by Griffiths (1994) in his RHS Index.

Geographical origin of the two subspecies

Having been in cultivation so long, the native territory of P. somniferum naturally has become obscured, but it is generally agreed to have originated somewhere in SW Asia, probably in Anatolia (Kadereit 1986a). It has been suggested that subsp. setigerum may be a native of W & C Mediterranean, Cyprus and probably or at least possibly the Atlantic Islands (Kadereit 1986 a, b).

Fossil history

The earliest fossil remains of apparently cultivated P. somniferum in Europe were found in S Holland and in adjacent regions of W Germany, but according to Kadereit (1990) there does not appear to be any reliable information about where cultivation first began on the continent. I suggest that the most obvious place to look would be around the coastal area of the Tyrrhenian Sea, ie S Italy and E Sicily, and especially in Campania, since this was the most important early trading area in southern Europe of the ancient Greeks and Phoenicians. These traders certainly brought the product from Egypt and very probably introduced the plant along with it (Boardman 1980; Grant 1987).

In Britain and Ireland, P. somniferum first appeared as seeds in the Bronze Age archaeological digs and it is common from the Iron Age onwards (Godwin 1975; P.J. Wilson, in: Preston et al. 2002).

Toxins

A total of about 25 alkaloids are obtained by the pharmacist from the opium latex, but six of these account for about 98% of the total. The six are: morphine (named after Morpheus, one of the sons of the god of sleep, Hypnos (Radice 1973)), codeine, papaverine, narcotine, narceine and thebaine (Stodola et al. 1992). Morphine is the most potent and the most present of these alkaloids. Although nowadays opium is chemically purified and separated, none of the drugs obtained have been superseded by any synthetic product.

Stock animals are very unlikely to graze the plant, but Opium Poppy IS EXTREMELY AND FATALLY POISONOUS, AND IT SHOULD NEVER BE USED FOR SELF-MEDICATION (Cooper & Johnson 1998).

Names

The origin of the genus name 'Papaver' appears to be obscure but it is a Latin name for a poppy, probably the medicinal P. somniferum, Opium or White Poppy (Gilbert-Carter 1964). It may have been derived from the Celtic 'papa', meaning 'thick milk', or Latin 'pappa' meaning 'milk', either way an obvious reference to the milky latex members of the genus contain (Melderis & Bangerter 1955; Hyam & Pankhurst 1995).

The Latin specific epithet 'somniferum' is a combination meaning 'sleep-bringing' or 'soporific' from 'somnos', sleep and 'fero', to bear (Gilbert-Carter 1964).

Apart from the English common name 'Opium Poppy' which refers to the origin of the drug, Britten & Holland (1886) list eight additional common names including 'Balewort', said to be derived from 'bealo' or Old English 'bealu', meaning 'bale', as in evil, mischief, woe, injury or pain (Cockayne 1864-6). The names 'Cheesebowl', 'Chasbol', 'Cheesebouls' and 'Chasse' all derive from the shape of the ripe capsule being considered as resembling round cheeses (Prior 1879). The name 'Mawseed' has some connection with the fact that the seed is provided for garden birds in winter. In German, it is called 'Magsamen', and in Polish, 'mak' (Prior 1979).

Threats

None.

Growth form and preferred habitats

Common Poppy is a summer annual which in Britain and Ireland has always been tightly restricted to open, disturbed, artificial habitats. Before the advent of selective herbicides, which over the last 30 years or so have almost exterminated the species from farmland, the distinctive red poppy flower was closely associated with fallow field margins around arable land, especially where the ground was under cereal cultivation. Nowadays, P. rhoeas is confined to other unsprayed forms of open or disturbed, artificial habitats, such as waysides, quarries, building sites, soil heaps, waste land and on walls (McNaughton & Harper 1964; Grime et al. 1988). The soil which suits Common Poppy best is a fairly light, warm, sandy and free draining one with a pH between 6 and 8. The ideal soil is of medium fertility and moderate to rich in calcium or other bases (Sinker et al. 1985).

Status as an ancient introduction

This ruderal annual species has an interesting and much studied biology, ecology and history. Along with numerous other arable weeds, populations of which were once considered stable and resilient, P. rhoeas populations have rapidly become impoverished and in Co Fermanagh the species has declined to rarity. Thus Papaver rhoeas has become of greater interest and concern to some conservation-minded people, while other botanists consider it a mere accidental contaminant of crop seed, albeit that of migrant Neolithic or Bronze Age farmers, 5500-2500 B.P. (Grieg 1988; Kadereit 1990; Wilson 1992). Plants of this origin in the British Isles have recently and belatedly (see eg Stace 1997) been recognised as 'archaeophytes', ie ancient, pre-1500 AD 'synanthropic' introductions (ie associated with man). A total of 149 such species are now acknowledged in the flora of Britain & Ireland, including P. rhoeas, by the editors of the New Atlas (Preston et al. 2002), and this figure is likely to grow.

Unlike other archaeophyte arable weeds, for example, Scandix pecten-veneris (Shepherd's-needle) and Centaurea cyanus (Cornflower), P. rhoeas is not yet a rare plant on an entire British Isles basis, but it has already become a rare casual in Fermanagh, there being a total of only twelve records in the Flora Database, three of which are derived from sown seed mixtures (see details below). On account of its local rarity, the 'Common Poppy' is given more consideration here than might otherwise be considered appropriate. After all, as Wilson (1992) reminds us, "weeds have accompanied these cultivated annual grasses [he is here referring to cereals] since their wild days, and in a very real sense, the history of weeds is the history of mankind".

Geographical and cultural origin

Along with three other poppies which are also archaeophytes (ie P. dubium (Long-headed Poppy), P. lecoquii (Yellow-juiced Poppy) and P. argemone (Prickly Poppy)), Common Poppy probably originated in SE Europe. The fossil history of P. rhoeas stretches back to middle and late Neolithic times in C Europe, and to the late Bronze Age in the British Isles (Godwin 1975). Kadereit (1990) examined the distributions of these poppies and compared them with their nearest 'wild' (ie non-synanthropic) relatives. His study suggests that P. rhoeas originated in the eastern coastline of the Mediterranean, where it appears to have differentiated from one of three close relatives, or from a hybrid between them (although without involving any change in chromosome number). It probably did so only after disturbed and cultivated ground in sufficient extent was provided by man as a habitat. Thus it seems possible that not only was P. rhoeas widely distributed by man with his cereals, but that as a taxonomic entity, the species may also be of synanthropic origin (Kadereit 1990).

P. rhoeas is chiefly a summer annual, a lifestyle which in itself might suggest a circum-Mediterranean origin, although in mild areas of the British Isles or in exceptional winters with little severe frost, autumn germinated seedlings may sometimes survive (Greig 1988). Summer annuals like P. rhoeas are capable of maturing and flowering after a short spurt of rapid vegetative growth in the spring, or failing this, flowering even when the plant is very small. Having this short, or relatively short, growing season, poppy plants are certainly favoured by our mild, damp oceanic climate, rather than a hot, dry continental one, which curtails their growth and development during the summer months (Greig 1988).

Fermanagh occurrence

For this weedy species, there have only ever been a total of twelve records spread over seven Fermanagh tetrads, four of which have post-1975 records. Until RHN discovered an individual plant flowering on roadside waste ground near Gublusk Bay, Lower Lough Erne in September 2003, Common Poppy had not been seen 'wild' anywhere in the county since 1950. However, there were two instances of it being deliberately sown in so-called 'wild-flower' seed mixtures in 1996 and 2003 in garden and in roadside settings. Since this very colourful and conspicuous species is well known to be capable of persisting buried and dormant in the soil seed bank for at least 80 years, and probably for more than a century (Salisbury 1964), it is always worth looking out for it on any freshly cultivated or recently disturbed patch of ground.

Before RHN, the current author and our contemporary botanical associates began recording in the early- to mid-1970s, agricultural weeds and plant introductions of any type or origin tended to be completely ignored by previous field workers in Fermanagh. Apparently earlier recorders considered them unimportant, or at least a waste of time in comparison with the need to record more obviously significant native plants, ie definite or perceived native species.

The details of the other eleven existing Fermanagh records are: football club car park, Newtownbutler, 1930-40, R.C. Faris & J.M. Cole; old railway station, Florencecourt, 1950, Meikle et al.; planted wild flower seed mix in garden near Levally House, 1 km SE of Roosky, October 1996 & October 1997, RHN; roadside waste ground, Gublusk Bay, Lower Lough Erne, September 2003, RHN; sown wild flower mix, at road junction of Sligo and Ballinaleck Roads, Enniskillen, September 2003, RHN; one plant on Queen Elizabeth Road, Enniskillen, 17 July 2006, RHN; one plant in a garden at Goblusk Bay, Lower Lough Erne, 14 August 2009, RHN; Clabby village, 26 June 2011, RHN; and Riversdale Forest, 29 June 2011, RHN.

Factors limiting distribution

Unsuitable soils

Apart from coastal situations, warm, sandy soils with the particular micro-environment that suits this poppy are scarce in the north and west of Ireland, in Scotland and on higher ground anywhere throughout the British Isles. Most of these geographical areas combine a geological structure that produces strongly acidic, high-silica soils that are frequently covered with a layer of nearly impervious glacial till of similar rock mineral origin and chemistry. In addition, lying close to the Atlantic seaboard, many soils in the more westerly parts of Britain and Ireland have suffered high levels of precipitation throughout the entire 11,000 years of the post-glacial period. The subsequent percolation of rainwater has resulted in soil mineral leaching, acidification, gleying and podsolisation, sometimes leading to the formation of huge swathes of organic blanket bog peat. The resultant lack of suitable soils makes it unsurprising that P. rhoeas is a scarce or casual species in the most northerly and western parts of Britain and Ireland.

Decline of tillage

The scarcity of this previously common weedy species in the north and west of Britain and Ireland was produced and perpetuated when the low intensity arable agriculture that previously allowed the Common Poppy's annual seed population renewal, gradually became increasingly uneconomical. Arable cultivation, especially of cereals, continued to decline further after the Second World War, eventually reaching a very low level, or near extinction.

A measure of disturbance

All poppy species, including P. rhoeas, require open, sunny habitat conditions, with bare patches of soil for seedling establishment. Thus poppies are rarely found in closed, grazed turf. If they are, then disturbance such as grazing and trampling must be sufficient to minimise competition from neighbouring plants, and provide sufficient openings in the soil surface for this ruderal species to germinate and establish. At the same time, the foraging disturbance must be limited to allow the species sufficient time to grow, flower and set seed successfully.

Toxicity

In order to limit grazing pressure, P.rhoeas contains a number of toxic or narcotic alkaloids, eg rhoeadine and rhoeagenine, which give off an unpleasant odour and make the plant unpalatable or sickening to browsing animals. There are no recent cases of animals being poisoned by P. rhoeas in the British Isles (Grime et al. 1988; Cooper & Johnson 1998).

Phenotypic variation and reproductive capacity

All Papaver species in Britain and Ireland are very plastic in their phenotypic response to growing conditions. They are particularly sensitive to shading, nutrient deficiencies, lack of water, or an excess of it, and interference from non-related species. All of or any of these variables can drastically reduce the poppies growth and reproductive capacity. P. rhoeas is exceptionally plastic in these respects, and McNaughton & Harper (1964) described finding depauperate plants in soils of very low fertility that were only 5 cm tall, bearing a single capsule containing four or five seeds. At the other extreme, under favourable conditions, an isolated individual might bear 300 to 400 large capsules, each containing more than 1,000 seeds (McNaughton & Harper 1964; Harper 1966).

An unusual feature of P. rhoeas variation is that flower size is also among the characters which are very plastic. Nutrient or water deficiency results in reduction of petal size, number of stamens, ovary locules and seeds. When species are as variable as poppies are, mean figures for properties such as seeds/capsule, or per plant unfortunately become rather meaningless, so I will refrain from quoting any such figures (Salisbury 1942; McNaughton & Harper 1964).

Flowering and pollination

The flowering season of P. rhoeas is longer than for other poppies that occur in Britain and Ireland. Although flowering begins rather late in the season in mid-June, and it peaks in early July, flower production still straggles on intermittently until October. The scarlet blooms attract a wide range of insect visitors, offering no nectar but easily accessible pollen in vast quantity. Honey bees are probably the most significant cross-pollinators. The flowers are outbreeders and are almost entirely self-sterile. Self-sterility is unusual in poppies and, indeed, this is true for successful annual weeds in general (Baker 1974). For further comparisons and discussion of this matter, please see the following P. dubium species account. Self-incompatibility in P. rhoeas is controlled by a complex multi-allelic gametophytic system; those wishing to know more about this topic are recommended to consult Richards (1997), Chapter 6, for a technical explanation of this rather involved genetic mechanism.

Seed dispersal

Seed ripens rapidly and is shed within four weeks of flowering, being quickly shaken out of the large pores at the top of the capsule which is borne on a long, swaying elastic stalk. This is often referred to in textbooks as the 'censer mechanism' of dispersal (McNaughton & Harper 1964). Dispersal distance was measured in a German experiment which purported to represent "near-natural conditions". However, the experimental setup could be regarded as rather artificial, since it involved dispersal over carefully maintained constantly bare ground. In any event, measurement showed that the great majority of seed travelled less than 100 cm, and the maximum distance travelled was 3.5 m (Blattner & Kadereit 1991).

Competitive ability

A range of competition experiments within and between different species populations has clearly shown that P. rhoeas reacts differently and in a complex manner to different sorts of neighbours − sometimes by mortality, sometimes by plasticity (Harper 1966). Generally, P. rhoeas is a weak aggressor when it comes to plant competition, but it is a "magnificent opportunist" (Harper 1966, p. 34). Common Poppy reacts rapidly to fill any vacancy in the environmental 'space', should some factor, eg disease, or selective browsing by an herbivore, happen to weaken neighbouring plant species. It is especially exploitive of conditions providing such openings when high nutrient levels are available (Jonsell et al. 2001). The competitive situation should be visualised, however, as a form of balance between coexisting species, rather than as one species replacing another, since mortality and competitive exclusion are not always involved. From this standpoint, plant species are often more sophisticated than we tend to allow. Biological scientists of most disciplines nearly always want to simplify and reduce the subtle patterns we find in nature. Plant (and animal) species do not belong in any fixed order of merit; "common species have their peculiar biology satisfied in frequent places; rare species may be superbly adapted to rarer sites" (Harper 1966, p. 35).

Irish occurrence

The New Atlas map shows that in Ireland, P. rhoeas is much more locally frequent to abundant and continuously distributed SE of a curve linking Newry in the north with Cavan, Longford and Athlone in the Midlands, and running south to Cork City. Elsewhere in the Republic of Ireland (RoI), P. rhoeas is now a very rare and scattered casual, while in Northern Ireland (NI) it is currently very thinly and widely distributed. The number of pre-1970 symbols on the hectad map indicates that it has declined from being a rather local plant and become a rare casual species (NI Flora Website, accessed 2002; Preston et al. 2002). It might well be argued that P. rhoeas is well down the road towards local extinction in NI, as Hackney et al. (1992) have suggested for the three counties of NE Ireland.

The pattern of occurrence in SE RoI does not appear to correlate all that well with geology, soils or climate (except perhaps a lesser number of rain days than elsewhere). However, it does reflect rather well the distribution map of 1970 tillage, and even better the distribution of wheat fields at the same date. Both of these features are mapped in the Royal Irish Academy's Atlas of Ireland (Haughton et al. 1979). Unfortunately, the same connection cannot be made in NI, where the species distribution pattern is really very fragmentary and does not appear to correlate or compare with any relevant factor, even if we include figures of rural land valuation.

Opposing the observed marked decline in the species occurrence across Ireland, however, two recent P. rhoeas Fermanagh sites arose from a sowing of 'wild flower seed'. This mode of origin is a phenomenon that in recent years has become rather common throughout the British Isles, and it is distorting the previous 'natural' distribution achieved by the species, derived from agricultural disturbance and previous crop-associated sowings. As both forms of the plant's occurrence arise entirely as a result of human activity, perhaps we should not get too agitated about this new pattern of introduction, of what is, after all, a beautiful, easily eradicated, weedy species.

British occurrence

In Britain, P. rhoeas is shown in the New Atlas to be much more common and widespread in the Midlands and SE England, extending more or less continuously up the E coast to around Dundee. Elsewhere in Britain, it is more scattered and occasional, becoming very rare and local in N Scotland. Being largely associated with arable agriculture, Common Poppy is essentially a lowland species, generally occurring below 200 m. Above this altitude, it is probably more casual than established (Grime et al. 1988; Preston et al. 2002).

European occurrence

In Europe, P. rhoeas is a common and widespread archaeophyte in W and C regions from the Mediterranean north to around 55N. However, it rapidly declines to casual status beyond this in many parts of Scandinavia. The species just reaches 60N on the Baltic coast of Sweden (Jonsell et al. 2001). P. rhoeas is largely absent from E Europe (Jalas & Suominen 1991). As elsewhere, populations rapidly declined in Europe from the 1950s onward due to use of selective herbicides, but since the 1980s it has recovered somewhat due to fallowing or set-aside (Jonsell et al. 2001).

World occurrence

Worldwide the distribution continues into SE Asia, N Africa, N & S America, Australia and New Zealand (Jonsell et al. 2001).

Uses

Corn Poppy carries with it numerous historical, folklore and medicinal traditions and uses. The seeds have a pleasant nutty flavour, are processed for Poppy seed oil, and are commonly sprinkled on cakes and bread for decoration and flavouring. Parts of the plant have also been used as a source of black dye ( Vickery 1995; Grieve 1931, p.651; Le Strange 1977, p.199-200; Mabey 1996).

Names

The origin of the genus name 'Papaver' appears to be obscure but it is a Latin name for a poppy, probably the medicinal P. somniferum, Opium or White Poppy (Gilbert-Carter 1964). It may have been derived from the Celtic 'papa', meaning 'thick milk', or Latin 'pappa' meaning 'milk', either way an obvious reference to the milky latex members of the genus contain (Melderis & Bangerter 1955; Hyam & Pankhurst 1995). The Latin specific epithet 'rhoeas' used to be the genus name. It is possibly derived from the Greek 'rhoia', meaning or referring to the 'pomegranate', Punica granatum, which both in the colour of the flower and shape of the fruit, P. rhoeas does indeed resemble (Johnson & Smith 1946; Gilbert-Carter 1964).

Grigson (1987) lists an overwhelming 50 English common names, many of which refer to the red flower colour or warn of the supposed bad luck or ill health (for example, thunder and headache, or blindness), consequent upon touching the plant. These myths were possibly used to keep children away from the poisonous plant, or out of the field to prevent trampling the corn (Vickery 1995). The gardeners' 'Shirley Poppy', which has a wide range of colour shades from scarlet to white, grey and lavender and a double form, was bred and selected from a sport of P. rhoeas (McNaughton & Harper 1964; Mabey 1996; Phillips & Rix 1999).

Threats

None.

Growth form and preferred habitats

As with P. rhoeas (Common Poppy) dealt with above, nowadays the weedy annual P. dubium and its two subspecies only merit the status of rare casuals in the flora of Fermanagh. The same is true in many other areas of Britain and Ireland where this once common plant has markedly declined, being almost eliminated by the combination of modern agricultural seed cleaning and the use of herbicides. The stations of P. dubium now tend to be rare or scarce, and generally the plant is transitory and confined to recently disturbed soils on roadsides, urban or residential areas, disused farmland, lough shores, calcareous screes and gravel pits.

The most recent edition (8th) of the Irish field botanist's An Irish Flora (2012), lists subsp. lecoqii (also regularly spelt lecoquii) on the basis that it differs from P. dubium in containing yellow rather than white latex, and bearing brownish to bluish anthers rather than yellow anthers. However, Parnell & Curtis (2012) remark that "other supposedly distinguishing characteristics appear to be unreliable". Since the BSBI's New Atlas maps P. dubium species and its two subspecies separately, we will follow it and list the record details of the eleven Fermanagh records in that manner. There are sufficient similarities between the subspecies, however, to make it sensible to discuss them together in this one species account.

Soil requirements

The substrate requirements of P. dubium and its subspecies are fairly dry, or at least free-draining, light, sandy or clayey soils, which again, as for P. rhoeas, absolutely must be subject to some degree of disturbance. The brief species accounts in the New Atlas suggest that subsp. dubium favours light calcareous soils and that subsp. lecoquii differs in colonising heavier ones (P.J. Wilson, in: Preston et al. 2002). It is uncertain just how reliable a feature this is, but it might possibly assist identification decisions based on the morphological characters of latex and anther colour mentioned above.

On account of the essential pioneer colonising behaviour of the species, and the necessity of soil disturbance and open ecological conditions for its growth, P. dubium, like Common Poppy, used to occur frequently in arable land before the advent of herbicides, and previously these two Papaver species regularly coexisted.

Variation and identification difficulties

Whatever the ranking of the taxa, the morphological similarity between P. dubium and P. lecoquii is such that most recorders experience great difficulty in distinguishing them, despite the fact that they do have different chromosome numbers – ie subsp. dubium is hexaploid while subsp. lecoquii is tetraploid. Kadereit (1990) advocates their separation mainly by the colour of the latex, "This is white or cream when fresh, and brown to black when dry in subsp. dubium, but mostly yellow or turning yellow when fresh and red when dry in subsp. lecoquii." However, another study undermines this identification character and echoes Stewart's (1888) doubts on the matter in N Ireland. Some forms or races of otherwise P. dubium morphology have been encountered in Europe that contain yellow latex (Koopmans 1970). The separation of P. dubium and related taxa is certainly difficult and, from the above, clearly reliance on the colour of the contained latex colour alone is unwise.

Kadereit (1990) also points out that despite the difference in ploidy levels, hybrids between the two forms can easily be produced. Furthermore, the hybrid form carries out regular but unequal meiosis, indicating that chromosome homology and pairing does occur. It therefore seems likely that subsp. dubium originated as a hexaploid directly from tetraploid subsp. lecoquii without the participation of any other taxonomic form.

Other distributional differences between the forms in C Europe suggested to Kadereit (1990) that subsp. dubium might be a native of SE Central Europe, and subsp. lecoquii of W. Anatolia and SE Europe. This idea is very much built upon his assumption, however, that Papaver species in this taxonomic section of the genus predominantly have vicarious distribution patterns (ie they occupy separate areas that do not overlap to any great extent). On the other hand, it is very clear from the studies of McNaughton & Harper (1964) and of Harper (1966), that the five archaeophyte (ancient, introduced) forms of 'cornfield' poppy which are found in the British Isles, "represent a progression of more and more northerly distribution, but that there is nothing in the distribution of these species which suggests any geographical or a regional displacement of one species by another." (Harper 1966, p. 27).

Wild and crop weed forms

Is it the case that the wild relatives of our cornfield poppies have separate, distinct species ranges, while the several forms closely associated with man and with cultivation typically co-exist and overlap in distribution with one another? As a measure of the distributional overlap and association between poppy species in S and SE England, McNaughton & Harper (1964) tabulated the number of stations in the total sample of 75 in their study, where the species co-existed; P. rhoeas and P. dubium did so in no less than 33 of these stations, while in an additional four stations they were joined by P. lecoquii (McNaughton & Harper 1964, p. 771, Table 3).

Fermanagh occurrence

As the coincidence tetrad map indicates, both poppy species, P. dubium and P. rhoeas, are widely scattered across Fermanagh, but they display a slight concentration in the SE of the county. There are just six records of P. dubium s.l. in the Fermanagh Flora Database, and a total of seven for the two subspecies. The details of the other four records of P. dubium s.l. are: pile of topsoil beside a house, Lisnagole, 2 km S of Lisnaskea, 3 July 1995, RHN & RSF; disused farm to SE of Clonmackan Bridge, 25 August 1995, RHN; roadside at Whitepark, 2 km NW of Lisnaskea, 8 September 1995, RHN & HJN; roadside N of Killymackan, near Teemore, 25 July 2000 and 11 July 2004, RHN.

The details of the three records of Papaver dubium subsp. dubium L. are:

• scree slope above Doagh Lough, 1986, S.J. Leach & S.A. Wolfe-Murphy;
• NW shore of Derryvore Td, Upper Lough Erne, 11 September 1986, T. Waterman;
• Gortaree gravel pit, Slieve Rushen, 11 September 1988, RHN, voucher in DBN.

The details of the four records of Papaver dubium subsp. lecoquii (Lamotte) Syme (Yellow-juiced Poppy) are:

• Belleek village, 1950, MCM & D;
• about twelve plants in a field of rye, halfway between Newtownbutler and Galloon Bridge, 1950, MCM & D;
• Enniskillen town, 16 May 1988, RHN;
• roadside E of Killymacken Lough, September 2004, N.F. Stewart.

Northern Ireland occurrence

In the pre-1940 period, in the 2nd edition of the Flora of the North-east of Ireland which covers Cos Down (H38), Antrim (H39) and Londonderry (H40) – an area of Northern Ireland at that time very much better recorded than Fermanagh – P. dubium and P. lecoquii were regarded as being about equally frequent (Praeger & Megaw 1938). Fifty years previous to this, in 1888, dealing with the same NE region of Ireland, S.A. Stewart wrote of P. dubium occurring, "on sandy cultivated fields, and sandy or gravelly waste ground; abundant in Down, less common in [Cos] Antrim and Derry." Sagely, he also commented, "Our plant seems to be var. lecoquii of Lamotte, but scarcely distinguishable. The colour of the sap does not appear to be a reliable guide." (Stewart & Corry 1888). In the latest, 3rd edition of the same Flora, Hackney et al. (1992) wrote of P. dubium agg., noting that it was, "less frequent than before", and was, "recorded from only eight 10 Km squares [ie hectads] in Co Antrim."

Interestingly, in view of the slight tendency for Fermanagh sites of P. dubium to occur in towns and villages, in their survey of the Urban Flora of Belfast, Beesley & Wilde (1997) found P. dubium was widespread on disturbed and waste ground in the city.

The effect of competition

In the late 1950s, when McNaughton and Harper's pioneering work on plant populations was carried out, it required very detailed experimental study and careful analysis to discover that density dependent self-thinning of poppy populations was more intense within a single species than between species (Harper 1966). This selection pressure, added to the range and degrees of plastic response which poppy species produce in many individual characters (including some normally quite conservative floral characters) when they are reacting to various forms of environmental stress, taken together often enable or allow plant survival in the population, rather than necessarily always leading to thinning mortality. This in turn then helps us to account for the frequency with which mixed stands of poppy species are found in the field.

The broad similarity in range of habitats and the weedy characteristics shared by all poppy species, contrasted with the subtle differences in their requirements, their ways of using the environment and in their breeding systems, is also very striking. This is very clearly reflected in the distributions in Britain and Ireland of P. rhoeas (a strongly self-incompatible outbreeder, extremely plastic in phenotype) and P. dubium subsp. dubium – which is facultatively self-compatible and almost exclusively inbred, but not quite so phenotypically variable as P. rhoeas (Rogers 1971).

These two poppy taxa are remarkably alike in their ecology, both being opportunist colonisers of relatively dry, bare ground, and prior to the use of selective herbicides, both plants were widespread major weeds of annual arable crops in Britain and Ireland. For these reasons, they extensively overlap in their distribution, but P. dubium is the better represented of the two in Wales and further north, and especially so in Scotland. It is also better able to transcend the altitude limit of arable agriculture (Preston et al. 2002).

Flowering

P. dubium flowers from May to July, and the flowers are fully self-compatible, although in an experimental natural selfing, seed set was measured at only between 20 and 30%. The anthers are borne well below the level of the stigmatic disk in the flower, an arrangement which clearly assists the possibility of outbreeding. Even so, Rogers (1969) found that the seed set possible by assisted self-pollination was 41% of the average set in normal 'open pollination' involving insect visitors which, in view of the relative positions of anthers and stigma already mentioned, is a surprisingly high figure. There was, however, great variation in the numbers of seed set as a result of selfing between trials, and this may indicate population differences in the degree of inbreeding success that are likely to be encountered within this species.

In P. lecoquii, the flower's sexual organs are at the same level and self-pollination, commencing in the dangling unopened flower bud, is more likely to be the norm. In this case, Rogers (1969) found the average number of seed set by self-pollination in P. lecoquii was 61% of the open pollinated average (Rogers 1969, p. 56, Fig. 1 and p. 59, Table 2b).

Reproductive strategies and genetics of colonising weed species

Success as an opportunist coloniser of bare, disturbed or cultivated ground requires a high degree of adaptation, and the gene combinations which confer this on plant species are generally maintained by three inter-related mechanisms: self-compatibility, rigorous inbreeding and low levels of genetic recombination (Baker 1959). Certainly P. rhoeas does not conform to this pattern of breeding behaviour, and to a much lesser extent P. dubium and P. lecoquii do not fit it well either. Rogers (1971, p. 274) argues that while the premium for the opportunist coloniser is on adaptations such as rapid germination and establishment, a short vegetative phase and a high reproductive capacity (ie large seed output), the most fundamental characteristic enabling success in this particular form of weedy lifestyle is flexibility. This especially means the ability to mature successfully and reproduce, irrespective of seasonal, climatic or soil variations, or indeed the effects of interspecific competition. Baker (1974) describes this combination of characteristics as, "releasing the weedy species from restrictions on its range of tolerance of abiotic environmental variation", and he refers to the ideal weed characteristics as the development of, "a general purpose genotype (or genotypes)".

Some degree of genetic heterozygosity permitting the flexibility found in these poppy species (and which is particularly marked in the strongly outbreeding and ecologically diverse P. rhoeas) would also be of great advantage to opportunistic colonising species. Research indicates that the balance between inbreeding and outbreeding and how it is achieved is different in each species. For instance, in the case of P. dubium, the degree of inbreeding has been shown to vary from one population to another (Rogers 1971). It is likely that in some cases, only an occasional outcrossing event is required in order to maintain the required level of heterozygosity that enables adequate phenotypic flexibility. This is particularly so if the species has both a very high seed output and a long-lived dormant soil seed bank as a fallback genetic resource (Snaydon 1980). All poppy species appear to meet these two criteria. P. rhoeas is rigorously outbreeding and so far there is no evidence of any breakdown in its incompatibility mechanism. The fact that it has a high degree of genetic heterozygocity is obvious from the wide range of variation shown in many different morphological characters. The really significant question we must ask is, How does P. rhoeas maintain a genotype, or more likely, a range of genotypes which confer adaptive advantage in open, disturbed habitats typical of crop situations, without having to resort to inbreeding? (Rogers 1971; Baker 1974).

Firstly, it should be clear that the characteristics of the ideal weedy species, as delineated by Baker (1974), do not require that the weed population should contain only a single genotype, and individuals certainly do not need to be completely homozygous, even when selfing and inbreeding prevails and selection pressure is at its most severe (Baker 1974). If this is the case, then the same thing may apply in reverse, ie outbreeding, and the associated chromosomal heterozygosity, need not imply absolute gene heterozygosity. We can then postulate that over time, selection pressure would gradually remove less well adapted recessive homozygotes, reducing the frequency of recessive alleles in the populations of the particular weed species. In turn, this would also favour a low mutation rate in the dominant habitat-adapted alleles. Most phenotypes would then as a consequence approximate to the optimum for strongly selected characters, while still maintaining a higher level of heterozygosity in other characters that are subjected to less rigorous selection pressure (Rogers 1971).

Survival of buried seed

In an experiment studying seedling emergence and seed survival of four poppy species (P. rhoeas, P.dubium, P. lecoquii and P. argemone), seedling numbers of all four were greatest in the first year following their production and burial in the soil. Thereafter, germination decreased exponentially from year-to-year and while all species had viable seed after five years, P. dubium showed the slowest rate of decline in seedling emergence. The numbers of seed which remained dormant and viable tended to be greatest for P. lecoquii and P. dubium (Roberts & Boddrell 1984). It is probably safe to conclude from this, that seeds of P. dubium and P. lecoquii are capable of remaining viable in the soil seed bank for a long period, certainly for more than five years, and possibly as long as has been shown for P. rhoeas, ie 80 years or more.

Genetics, numbers and survival

The seed output of poppies such as P. rhoeas and P. dubium is so vast, and their seed longevity in the soil is sufficiently long, that these poppy species generally have an enormous surplus of genotypes available to provide the next generation. The proportion of the seed which produces the next generation is correspondingly small, and a small proportion of ill-adapted phenotypes, even if they were somehow to survive and breed, would be unlikely to have any significant effect on the ultimate population size, or on the success of the species as a weed (Rogers 1971).

Pollination and seed dispersal

The flowers of P. dubium and P. lecoquii attract mainly bees and hover-flies which feed on the plentiful pollen mostly in the early morning. After pollination, the capsule requires five to six weeks to ripen before the pores open to shake out the seed. The pores are quite large in relation to seed size, but even so dispersal distances are likely to be only a few metres even under optimum wind conditions for release and transport. The seed show marked dormancy, with few or none germinating within the first months after shedding (McNaughton & Harper 1964).

Toxicity

As with other poppy species, plants of P. dubium and P. lecoquii are undoubtedly able to deter some browsing herbivores by means of their toxic or distasteful alkaloid content.

European occurrence

In Europe, P. dubium is widespread over much the same area as P. rhoeas, from the Mediterranean, through W & C regions and with more of a presence in the north, extending on both shores of the Baltic to 60N and slightly higher than this in on the Swedish shore (Jalas & Suominen 1991, Map 1989). According to McNaughton & Harper (1964), the northerly spread of P. dubium in Scandinavia is due to the discharge of ships' ballast containing the seed.

World occurrence

Beyond Europe, again like P. rhoeas, P. dubium s.l. stretches from N Africa, through SE Asia and eastwards to reach Nepal. As a previously very common crop seed weed contaminant, it has also been widely introduced by agriculture to both N and S America, Fennoscandia, Australia, New Zealand and Tasmania (Hultén & Fries 1986, Map 888; Jonsell et al. 2001).

Names

The Latin specific epithet 'dubium' means 'doubtful' or 'uncertain' which in view of the identification problems associated with it and P. lecoquii, is all too apt (Gledhill 1985). The name 'lecoquii' is clearly called after someone, and the most likely candidate appears to be Henri Lecoq (1802-1871), a French botanist or plant geographer. There is a genus of the Apiaceae (= Umbelliferae) called 'Lecoqia', 'Lecoquia', or 'Lecokia' found in the eastern Mediterranean, which perhaps is another such memorial (Willis 1973).

Threats

While both P. rhoeas and P. dubium have clearly declined over their entire synanthropic range following the advent of modern seed cleaning procedures and the use of selective herbicides, it still remains possible that these poppy species may produce herbicide resistant forms, as has already happened in over one hundred other crop weeds (Warwick 1991; Briggs & Walters 1997).

Growth form and preferred habitats

M. cambrica is a slender, hairy, long-lived montane perennial with a deeply penetrating tap-root or tapering root-stock (Butcher 1961). Sometimes, as at the Hanging Rock NR, Welsh Poppy occurs in great abundance, forming a yellow swathe on rocky slopes under trees, but more usually it occurs as isolated clumps in slightly damp areas. In Fermanagh and elsewhere in these islands, native occurrences are typically found on moist, shady, rocky slopes, cliffs or screes, sometimes under trees in woodland or scrub and usually on limestone or base-rich soils (Stewart et al. 1994). The species also occurs quite frequently as a garden escape in somewhat disturbed ground sites, and since it is very capable of naturalising itself in the wild, distinguishing the 'escapees' from native plants is not always an easy matter.

Flowering

The solitary, fragrant, lemon-yellow flowers are produced from June to August. Like other poppies, the blossom is nectar-less but the flowers offer visiting insects a plentiful supply of protein-rich pollen as an alternative food source. Large numbers of small, black, finely pitted seeds are shaken out of the capsule and are the only means of reproduction open to the species.

The facts that M. cambrica can become a weed in the garden and that it regularly escapes into the wild, both attest to its great fecundity and manifestly efficient powers of dispersal. It is the easiest Meconopsis species to grow in the garden: all the other members of the genus are Asiatic in origin and, in the experience of the current author (Ralph Forbes), demand a degree of skill and an element of luck to cultivate successfully in our climate.

Fossil record

There are no fossil records for Meconopsis cambrica since its pollen is not distinguishable from that of Papaver (Godwin 1975).

Fermanagh occurrence

M. cambrica has been recorded from a total of 16 Fermanagh tetrads (3.0%), 15 of which have post-1975 dates. There are fairly frequent records from localities on the upland limestones, but just four questionable lowland stations in the VC that might be non-native occurrences. These records from suspect sites (eg by roadsides, on waste ground or tips, or at Boho Church) are omitted from the Fermanagh tetrad map. However, three of the four sites are in the near vicinity of upland limestones where the species occurs naturally, and it is certainly possible that they may have spread to these sites naturally. This leaves only the record on waste ground near a bridge on the Ballycassidy River as a very probable garden escape.

N Ireland and the Republic of Ireland occurrence

Elsewhere in N Ireland, M. cambrica is generally considered a rare, native plant. Apart from its Fermanagh occurrence, it has only about nine listed indigenous stations in N Ireland − one in Co Londonderry (H40), two in Co Down (H38) and six in Co Antrim (H39) (FNEI 3). In the Republic of Ireland, M. cambrica has isolated stations on the hills in ten scattered VCs at altitudes from 15-500 m, but it is strangely absent from likely sites in counties such as Kerry (H1 & H2), Mayo (H26 & H27) and Donegal (H34 & H35) (The Botanist in Ireland; BSBI Atlas; Cen Cat Fl Ir 2). It is also surprisingly rare in the limestones of the Burren, Co Clare (H9), where one would readily imagine the terrain to be ideally suited to it. In this exceptionally well-explored karst limestone region on the shores of Galway Bay, M. cambrica was first discovered as recently as the mid-1970s. It appears perfectly indigenous in the two known sites there (Webb & Scannell 1983; Nelson & Walsh 1991). Elsewhere in the Republic, the New Atlas map indicated the species is thinly scattered mainly across the Midlands and the east in non-native stations that presumably are of garden origin (Preston et al. 2002).

British occurrence

In Britain, this species is regarded as a scarce native in Wales, N Devon (VC 4) and N & S Somerset (VCs 5 & 6) (Stewart et al. 1994). Everywhere else in Great Britain (from the south coast to Shetland), it is considered a very common and widespread garden derived introduction.

European occurrence

M. cambrica is endemic to W Europe and is distributed from W Ireland southwards to the French and Spanish Pyrenees (Jalas & Suominen 1991, Map 2023). It is an established garden escape in other parts of the British Isles, France and southern Scandinavia, but not truly established in Germany, Netherlands and Switzerland (Jalas & Suominen 1991; Jonsell et al. 2001).

Names

The genus name 'Meconopsis' is derived from the Greek 'mekon', meaning 'poppy' and 'opsis', meaning 'like' or 'appearance' (Johnson & Smith 1946; Gilbert-Carter 1964). The Latin specific epithet 'cambrica' means 'of Wales', since the first site reference for the plant in the British Isles was, "in many places of Wales", in Parkinson (1640), a fact also commemorated in the English common name.

Threats

Clearance of the woodland under which it occurs.

Growth form and preferred habitats

This conspicuous bright-yellow flowered, short-lived perennial or biennial species has a branching, short, woody rootstock and brittle stems that contain acrid and very poisonous orange latex which burns and stains the skin when handled (Cooper & Johnson 1998). Possession of latex is a feature shared with other members of the family Papaveraceae, but the four-petalled yellow flower, the elongated pod-like fruit capsule and the almost-pinnately lobed leaves all rather strongly resemble members of the closely related Cabbage Family, the Cruciferae (or the Brassicaceae as we must now learn to call it under new taxonomic rules!). Nevertheless, because of other details such as the one-celled capsule, C. majus really is a poppy (Webb et al. 1996).

The widely used English common name 'Greater Celandine' is another source of possible confusion, but in truth the plant is completely distinct from 'Lesser Celandine', Ranunculus ficaria of the already much too large and varied Ranunculaceae! Adoption here of the common name widely used in N America, 'Celandine Poppy', would be an easy way to avoid any confusion (Kang & Primack 1991).

C. majus is typically found in sunny, sheltered, lowland sites, which are kept open by some level of human-related disturbance. It is always found close to habitation and the gardens from where it originates, which suggests that it has only rather limited powers of dispersal. The type of sites inhabited by C. majus includes roadside verges, hedgebanks, waste ground and rubbish tips. It also occurs on the tops of old walls, or in crevices in them and even more frequently in the sparse blown soil and litter accumulated along their base. C. majus is essentially a species of moderate soils with respect to such factors as moisture, fertility, acidic-alkaline reaction (ie pH), but it definitely prefers base-medium to base-rich or calcareous conditions (Sinker et al. 1985).

Distribution and variation

The native distribution of Greater Celandine has been obscured by innumerable escapes from cultivation and it is now found in gardens mostly as a weed. C. majus does not show great variation throughout its total range, but a small number of subdivisions have been made, most notably separate subspecies in Europe (subsp. majus), C and E Asia (subsp. grandiflorum) and in coastal China and Japan (subsp. asiaticum) (Jonsell et al. 2001).

Fermanagh occurrence

Greater Celandine is only an occasional or rare species in Fermanagh, there being reports from a total of 20 tetrads, 14 of them with post-1975 records. As the distribution map indicates, it is rather thinly scattered around the county, but chiefly occurs to the east of Lough Erne. While its populations are usually not very large, they may seed themselves freely and sometimes (perhaps rarely) they become well established and long persistent. For instance, the population at Old Crom Castle appears to have survived for well over a century.

N Ireland occurrence

Considering the position of the plant in the three VCs of the NE of Ireland (Cos Down, Antrim and Londonderry), Hackney et al. (1992) more definitely stated that Greater Celandine, "rarely persisted to become truly naturalised".

Fossil record

The few fossil finds of the species in the current interglacial warm period are confined to sites of human settlement and their dates suggest that it was first present in the British Isles in Roman times. We should therefore regard it as an ancient introduction by man (ie a pre-1500 AD 'archaeophyte') (P.J. Wilson, in: Preston et al. 2002). The fact that it has been in use in herbal medicine (and more recently in homeopathy) for a long time and has been cultivated for this purpose, strongly reinforces this argument. However, the fossil record also indicates that Greater Celandine was present in Britain in the previous Ipswichian interglacial, when it must have arrived by entirely natural means of dispersal without the assistance of man. In turn, this suggests that it might have been able to disperse to Britain, if not to Ireland, under its own power during the current Flandrian interglacial period (in Ireland known as the Littletonian) (Godwin 1975, p. 129; Mitchell 1986).

Flowering reproduction

Flowers are produced all summer from May to August. In addition to being visited and pollinated by bees and flies, both selfing and cleistogamy occur (the latter process involves selfing within the flower bud) (Jonsell et al. 2001). An interesting American study of C. majus flowering behaviour found that most of the variation in the sizes of reproductive characters occurred within individual plants, instead of among plants, or between populations. Flower and fruit sizes as well as seed number per fruit declined significantly during the flowering season, while mean seed size per fruit remained more stable (although even this normally extremely conservative characteristic dropped significantly during the season in one population studied). This suggests that maternal plants may have a strategy for conserving resources by a gradual reduction in the size of some reproductive characters, possibly in order to prolong the period of seed production (Kang & Primack 1991).

This study also showed that (as expected) larger plants not only produced more flowers and fruits than smaller plants, but that these organs were also consistently larger. Thus an important general conclusion was that, in ecological and evolutionary studies, characters such as seed size and number that are commonly taken to reflect plant fitness, should not be viewed in isolation from vegetative characters. Flower and fruit sizes matter, if the goal is to understand the mechanism of natural selection in wild populations (Kang & Primank 1991). Those contemplating a study of flowering performance would do well to read this important paper carefully.

Seed form and dispersal

The elongated capsule contains white-tipped black seeds. The white outgrowth on the seed coat is a nutritive elaiosome oil-body which attracts ants (and possibly also snails and birds), which help disperse the seed to some unmeasured extent (Ridley 1930, p. 522).

Toxicity and medicinal uses

Greater Celandine contains several toxic alkaloids including chelidonine, homochelidonine, chelerythrine, sanguinarine and protopine. Cooper & Johnson (1998) remark that the plant rarely causes poisoning because it is so unpalatable, having an acrid taste, a pungent foetid smell and caustic sap. Despite this, cattle and horses have been known to browse the plant, and it is particularly dangerous, or even lethal, when it is in seed (Cooper & Johnson 1998, p. 169).

NONE OF THE FOLLOWING HERBAL USES SHOULD EVER BE PRACTICED WITHOUT QUALIFIED MEDICAL SUPERVISION. A dilution of C. majus sap was formerly widely used for the treatment of sore or cloudy eyes. External application of the sap was also used to treat warts, corns and ringworm, although being caustic, it is reputed to damage any skin that it touches. However, Cooper & Johnson (1998) reported that repeated external application of the sap failed to cause any skin damage, and they concluded that the reputation of Greater Celandine as a severely toxic plant is dubious. A tincture is used in homeopathy and an herbal ointment is used to treat chronic eczemas. The alkaloid chelidonine affects cell division (like colchicine does) and in Russia it has been investigated for possible use as an anti-cancer drug.

European occurrence

C. majus subsp. majus is widespread in temperate areas of W, C and E Europe, thinning to both N & S, although it does occur on all the islands of the W Mediterranean, and it just reaches 65N (Jalas & Suominen 1991, Map 2028). As is the case in Britain and Ireland, this subspecies is not always considered native on the continent. In Scandinavian countries, it is regarded by most botanists as an archaeophyte.

World occurrence

It is also present in N Africa (status undetermined), while it definitely is an introduction in both N America (where it stretches across the temperate zone) and in New Zealand (Jonsell et al. 2001).

Names

The somewhat unusually prolonged flowering period typical of the species is said to be reflected in the genus name, 'Chelidonium', which is derived from the Greek 'khelidon', meaning 'a swallow' (ie the summer visiting bird), supposedly because the arrival and departure of swallows is thought to coincide with the flowering period of the plant (Johnson & Smith 1946; Gilbert-Carter 1964).

C. majus has quite a number of English common names: Grigson (1987) lists 14, and Britten & Holland (1886) no less than 19. Several names as usual are merely different dialect spellings, for example, 'Celandine' and 'Celidony', plus 'Saladine', 'Seladine' and 'Sollendine' (Britten & Holland 1886). The name is derived similarly to the Latinised genus name, either from the original Greek or

from Old French, 'celidoine', referring again to the swallow. One legend had it that the mother bird used the orange sap of the plant to restore sight to her blinded young (Grigson 1987).

The use of C. majus for cleansing the eyes is certainly older than the explanation or the civilization of the Greeks, since it was equally a part of ancient Chinese medicine (Grigson 1974). The juice of Celandine took away cloudy spots on the eyeball which were called 'kennings' and the plant was introduced to NE America under the name 'Kenning-wort' (Grigson 1987). The Americans also call the plant 'Swallow Wort', a name known to both Gerard and Lyte (Grigson 1987). The name 'Yellow spit' obviously refers to the sap, and names involving 'wart' remind us of another major use of it, for example, 'Wartflower', 'Wart plant', 'Wartweed', 'Wretweed' and 'Kill Wart' (Grigson 1987). Another name in the old herbals is 'Tetter-wort', tetters being some form of skin disease involving a running sore or wound, which the sap was reputed to heal (Prior 1879).

Threats

None.

Growth form and preferred habitat

A rhizomatous perennial, endemic to the lower slopes of calcareous parts of the Swiss and Italian Southern Alps, in its native ground P. lutea is a plant of shaded rocks and screes (Landolt & Urbanska 1989; Grey-Wilson & Blamey 1995). In Britain and Ireland, it typically grows in the mortar of old walls and prefers half-shade. Dainty and undemanding, Yellow Corydalis has been a popular horticultural choice for walls and beds in shady corners of gardens in the British Isles since the end of the 16th century (Grey-Wilson 1989). The only problem with the plant is its ability once established to seed itself prolifically or indeed excessively and become a weedy nuisance (Hansen & Stahl 1993).

Fermanagh occurrence

This conspicuous garden escape has been recorded just six times in Fermanagh and only twice in the past 50 years – the last observation having been made by RHN in 1989. The details of the other five Fermanagh records are as follows – the first four being made by Meikle and co-workers, the last two by RHN: on west bridge, Enniskillen town, 1946; ruined house near Sand Lough, 1949; bridge between Belcoo and Blacklion, 1952; waste ground beside Ballycassidy River, 100 m N of road bridge, 9 May 1988, RHN; wall of old Railway Station, Maguiresbridge, 3 September 1989, RHN.

Flowering

P. lutea has an extremely prolonged flowering period (March to November). The flower is self-compatible and is pollinated by bees (Jonsell et al. 2001). Initially the flower is closed, and it opens with a once-only 'explosive' mechanism, similar to that of Gorse, Broom and other members of the Pea family. All of these plants are pollinated by vigorous bees of sufficient weight to trigger the flower opening mechanism (Proctor & Yeo 1973, Fig. 58, p. 205).

Seed and dispersal

The seed of P. lutea is rather too large and heavy to be readily wind-borne, but it has an attached white elaiosome oil food body which attracts ants and possibly birds. This probably explains how it manages to colonise walls (and cliffs in its native area), at heights well above ground level (Salisbury 1964). Ridley (1930) noted it growing approximately 5 m up a wall, and he suggested that ants were responsible for transporting the seed!

Fossil record

I am not aware of any fossil record for this species. I believe that there are very few fossil records for the related Fumaria species, some only at the generic level, and most or all of the records are associated with sites of human settlement or cultivation (Godwin 1975).

British occurrence

In England and Wales, P. lutea is fairly widespread and commonly naturalised S of a line between Carlisle and Berwick-upon-Tweed. However, N of this region Yellow Corydalis becomes somewhat more coastal and diffuse, except around the Edinburgh conurbation where it is again more frequent. Since the 1962 BSBI Atlas, P. lutea has increased in some areas in Great Britain (calculated Change Index = + 0.59), eg in Wales and SW England (P.J. Wilson, in: Preston et al. 2002).

Irish occurrence

The New Atlas shows the species in Ireland is very much more rare and widely dispersed than in Britain, but as with Papaver somniferum (Opium Poppy), in N Ireland it appears somewhat more frequent, although still decidedly sparse (Preston et al. 2002). Between them the Irish Flora Census Catalogue 2^nd edition and Reynolds's list of alien plants Cat Alien Pl Ir indicate that P. lutea has been recorded at least once from a total of 18 Irish VCs (excluding Fermanagh, which has not had records published until now). The great majority of the Irish P. lutea records emanate from the north. I believe that this scattered pattern of occurrence simply reflects the greater number of recorder hours spent per hexad in the northern part of the island, since this level of recording enables casual species and infrequent garden escapes to be discovered along with the more common plant species.

European occurrence

In W Europe, P. lutea is widely naturalised beyond its narrow endemic range, but apart from near Narbonne in S France, it is completely absent from the Mediterranean, as well as from the entire Iberian Peninsula. Yellow Corydalis is better represented in Scandinavia than shown by Jalas & Suominen (1991, Map 2059) and, indeed, Jonsell et al. (2001) records that it is spreading northwards in coastal areas of both Norway and Sweden where it was first recorded as late as 1922.

Names

The genus name 'Pseudofumaria' combines the Greek 'pseudo' meaning 'false' and the genus name 'Fumaria', which is ultimately derived from the Mediaeval Latin 'fumus terrae', meaning 'smoke of the earth', ie arising from the ground like smoke. This is considered a reference to the diffuse foliage of some species of the genus (Gilbert-Carter 1964; Hyam & Pankhurst 1995). The Latin specific epithet 'lutea' means 'yellow', being derived from 'lutum', the ancient name of Reseda luteola (Weld), which produces a yellow dye (Gilbert-Carter 1964).

Threats

None.

Growth form and preferred habitat

C. claviculata is a delicately stemmed, much branching annual, whose branched leaf-tendrils give it a decidedly limited climbing, scrambling and trailing ability since the plant usually only extends up to around 80 cm or so in height. In terms of soil preference, it is markedly calcifuge in character. Most unusually for an annual species, its most natural habitat in the British Isles is in shady or semi-shade sites in rocky hillside glens, or sometimes along drier areas of scrub-lined river or stream banks. In such 'wild' situations, acidic mineral or peaty soils are well-drained by being on fairly steep slopes. Here, the plant typically scrambles amongst rocks and over other plants in half-shade or in the better light of clearings in deciduous woods, scrub or conifer plantations, especially when these are protected from grazing and disturbance. Sometimes it may also occur in gorse or other scrubby heath vegetation (again, particularly when they are protected from browsing), and it may occasionally be found on heathy hillsides or roadside banks that have become infested with bracken colonies or bramble patches (McMullan 1972; Chater 2010). C. claviculata becomes especially conspicuous after the associated woody species have been cut or burnt in order to harvest, control or regenerate heath or woodland vegetation.

Flowering

The plant flowers from June to September; the small flowers, borne in racemes of five to ten, are pale yellow or cream to almost white in colour and they produce nectar which attracts bees. They are, however, self-compatible and probably mainly self-fertilising (ie they are autogamous) (Clapham et al. 1962; Jonsell et al. 2001).

Seed dispersal

The fruit capsule produces between two and four seeds which (depending on the authority cited) may (Stace 1997), or may not (Jonsell et al. 2001) possess a fleshy aril or elaiosome food body. The latter attracts ants and possibly other animals and appears adapted to effect some degree of seed dispersal. An interesting project would be to clarify the true position regarding the existence of a seed oil body and, if one does exist, to arrive at an estimate of the distance ants (or possibly snails, or birds) transport the seed (Ridley 1930).

Fossil record

Again, as with Pseudofumaria lutea, I am not aware of any fossil record for this species. I believe that there are very few fossil records for the related Fumaria species, some only at the generic level only, and most or all of these are associated with sites of human settlement or cultivation (Godwin 1975).

Fermanagh occurrence

There is a solitary recent record of this rare casual in the Fermanagh Flora Database as listed above. The overgrown derelict garden of an abandoned dwelling of architectural interest contained an interesting collection of around 30 weeds, some of which, like this species, are rare, unusual or seldom found and recorded in Fermanagh, or indeed elsewhere in most of Ireland. Since the Fermanagh record was made, the house has been restored and reoccupied and therefore it is very likely that the garden has also been brought into cultivation and the species ousted.

Irish occurrence

Mackay (1836) described the species in his book Flora Hibernica as occurring, "on rocks, walls, and on tops of old thatched houses". He mentions stations occurring for instance, "on thatched cabins between Ballinteer and the little Dargle (river)", and by "the wayside" (ie under hedges or on hedgerow banks), "between Dundrum and the Dublin mountains".

At the end of the 19th century, Colgan & Scully (1898) described C. claviculata as occurring, "chiefly in the counties of Dublin and Wicklow, with two out-stations in Waterford and Donegal". Three years later, Praeger (1901), who had meantime found an additional station for the plant upstream from Waterford on the River Barrow in Co Kilkenny (H11), was writing of C. claviculata as being, "a plant of the SE, with one outlying station in Donegal". With the addition of two further localities in the south-centre of the island in N Tipperary (H10) (McMullan 1972), this still remains very much the position taken in the Census Catalogue of the Flora of Ireland by Scannell & Synnott (1987). The latter reference lists the plant as Corydalis claviculata, which the authors regard as indigenous and recorded at least once in seven Irish VCs, not at that time including Fermanagh.

The couple of records of C. claviculata that exist in the north of Ireland, including the old Culmore, Co Donegal record of Hart (1874), and another northern 1988 record in a cleared plantation wood in Co Monaghan (Alan Hill, pers. comm., January, 1990), on examination appear decidedly synanthropic. They were found growing either close to or upon dwellings, or occurring as garden weeds or in estate plantation woods. Since the earliest mentioned Cos Dublin (H21) and Wicklow (H20) stations also fit this pattern closely, I believe there definitely is a case for describing C. claviculata as a naturalised, neophyte alien throughout Ireland. It was probably introduced with soil on the roots of garden plants and plantation tree stocks, most likely during the late 18th and early 19th centuries.

British occurrence

In Great Britain, C. claviculata is widely dispersed throughout the island and is locally common in the west. Although it is chiefly a lowland plant, it does reach 430 m in S Northumberland (VC 67). Despite the amount of forestry plantation that has gone on across Britain since 1970, the distribution has remained stable when compared to that in the 1962 BSBI Atlas. It is, however, better recorded now than previously was the case (Walters & Perring 1962; P.J. Wilson, in: Preston et al. 2002).

European occurrence

Beyond Britain and Ireland, the distribution of C. claviculata is restricted to W Europe, ie it is endemic to that region (Hultén & Fries 1986, Map 894). The species extends in a slightly disjunct, discontinuous manner from N Portugal along the Atlantic coastline to W Norway, reaching its northern limit just beyond 60N. The European distribution closely overlaps that of heathland in W Europe (Gimingham 1972) and the perimeter of both these distributions fits well with Koppen's (1918) 'Cfb', ie a constantly moist and mild 'Oceanic' type of climate.

C. claviculata is regarded as indigenous in Denmark and populations are expanding, encouraged by the plantation of conifers. The species is also considered indigenous in coastal parts of S Norway. Elsewhere in Norway, however, established alien populations are known to occur. Furthermore, in S Sweden, Climbing Corydalis arrived in the late 1950s and is now recognised as an established alien variously transported along with imported timber, with Rubus shrubs, or as a weed amongst other horticultural stock. It already occurs in S Sweden as a garden escape (Jonsell et al. 2001).

Names

The genus name 'Ceratocapnos' is a difficult combination to decipher, and it does not really appear to make much sense. 'Keras' is Greek for 'a horn', and 'capnos' (or in Greek, 'kapnos'), means 'smoke'. The latter is possibly a reference to its related genus (and family name), 'Fumaria', which is reputed to comes from the Mediaeval Latin 'fumus terrae', meaning 'smoke of the earth', a supposed reference to the diffuse (perhaps, smoke-like), foliage of some members of the family. Another possibility is reference to a supposedly smoky, irritating smell the plant is said to give off (Stodola et al. 1992). The Latin specific epithet 'claviculata' means 'having tendrils', the reference here being to the tendrils of the Vine and also to the branched tendril resembling a small key (Gilbert-Carter 1964).

Threats

None.

Growth form and preferred habitats

This fairly tall, diffuse or scrambling summer or occasionally winter annual has large purple-tipped creamy white flowers. The New Atlas map and the species account by David Pearman and Chris Preston in Scarce plants in Britain both indicate that in Britain and Ireland, F. capreolata tends to be much more prevalent at the coast, on open scrub, hedge banks, old walls and cliffs. It is only occasionally found inland in lowland situations, where the habitats then include arable farmland, gardens and open waste ground. It is, however, less frequently found as a weed of disturbed ground than other species of fumitory (Stewart et al. 1994).

Variation

F. capreolata has been split into two subspecies, the one commonly found throughout Britain and Ireland is the endemic form, F. capreolata subsp. babingtonii (Pugsley) Sell. The more southerly, continental form of the species is subsp. capreolata, which is found on the Channel Isles and possibly also on the south coast of England, as well as on the European mainland.

Flowering

The flowers of F. capreolata are self-compatible and they habitually and automatically self-pollinate in the bud before they open (ie they are cleistogamous). This floral mechanism ensures an abundant seed set (D.A. Pearman & C.D. Preston, in: Stewart et al. 1994).

Seed survival in the soil

There is very little evidence available on the seed bank potential of the species, but Thompson et al. (1997) list one German reference which suggests that seed survival is merely ephemeral.

Fermanagh occurrence

F. capreolata has been recorded on three occasions in recent years. It was first recorded along with F. muralis on disturbed waste ground by the roadside as listed above. The details of the other two records are: in the grounds of the Carlton Hotel, Erne Bridge, Belleek, 12 July 2006, RHN; and S shore of Rossigh Bay near the Point, 2007, I. McNeill.

Status in Britain and Ireland

In Britain, this weedy species is regarded as indigenous, but in Ireland it has long been regarded as a probable accidental introduction. The beautifully illustrated Fumitories of Britain and Ireland (BSBI Handbook, no. 12) did not appear until the great bulk of the current Fermanagh survey was finished, but in it Murphy (2009) refers to this plant as F. capreolata subsp. babingtonii (Pugsley) P.D. Sell.

Changes in distribution

Comparison of the hectad maps of F. capreolata in the BSBI Atlas and the New Atlas indicates that this fumitory's distribution is stable at its maritime sites. However, the same maps and the prevalence of older record date classes on them suggest that the species is declining at its inland sites, to the extent indeed that it is probably only an ephemeral casual in the latter (BSBI Atlas; P.J. Wilson, in: Preston et al. 2002). The frequency of occurrence of F. capreolata in Northern Ireland compared to the Republic of Ireland is quite marked in the New Atlas map, but this probably is an artefact simply reflecting recorder effort.

Fumitories of Britain and Ireland (Murphy 2009) features hectad distribution maps with increasing dot size for more recent records. This visual representation makes it even more obvious that fumitories are much more actively sought in some areas of Britain and Ireland than elsewhere.

Names

The genus name 'Fumaria' is derived from the apothecaries' Mediaeval Latin 'fumus terrae', meaning 'smoke from the earth', a poetic allegory of the way F. officinalis spreads its pale blue-green, diffuse foliage across the soil surface supposedly like smoke when seen from a distance (Grigson 1974). The Latin specific epithet 'capreolata' means 'with tendrils', although in fact the plant has no tendrils, scrambling and climbing instead using its twisting or coiling leaf stalks or petioles (Murphy 2009; Parnell & Curtis 2012). There are no specific English common names for this species in Britten & Holland (1886), but see entry under Fumaria officinalis. The name 'White Ramping-fumatory' is a book name of recent origin.

Threats

None.

Growth form and preferred habitats

This very variable, diffuse to robust scrambling annual is closely associated with disturbed arable or horticultural sites, most frequently appearing alongside other Fumaria species among spring-sown crops, generally on acidic, free-draining soils. Much more rarely, it may be found on hedge banks also offering a well-drained habitat, or on waste ground or roadside verges providing suitable, recently disturbed substrate conditions (P.J. Wilson In: Preston et al. 2002). It can sometimes also occur in dry soils at the base of walls. Everywhere it is confined to the lowlands.

Fermanagh occurrence

Until Ian McNeill recorded it in 2000 and 2002, this very variable weedy annual had been distinguished just six times in Fermanagh and it had not been listed by anyone since 1949 (Revised Typescript Flora). This could possibly be due to the marked local decline of arable farming almost to extinction in our survey area or, perhaps more likely, to difficulties faced by field recorders in its proper identification. This problem is commented on here under F. muralis (Common Ramping-fumitory), the plant with which F. bastardii is most often confused by everyone including myself. The publication in 1989 of Peter Sell's account of the Fumaria bastardii/F. muralis complex in BSBI News 51, and his subsequent key to the British and Irish members of the genus in the BSBI Plant Crib (Rich & Jermy 1998), has made identifications much more reliable. The publication of the BSBI Handbook Fumitories of Britain and Ireland with its excellent illustrations will doubtless prove invaluable for future identification (Murphy 2009). The latter recognises three distinguishable varieties, one of which is endemic in western Britain and Ireland, var. hibernica (Pugsley ex Praeger) Pugsley (Murphy 2009). This variety has not yet been recorded in Fermanagh.

There are a total of eight Fermanagh records scattered across the lowlands of the VC. The record details additional to the first finding listed above (which incidentally was originally identified as F. confusa Jord.) are as follows – in all cases the sites are only loosely characterised: near Lisnarrick, 1942, R. Mackechnie; Belleek, 1948, MCM & D; Lisbellaw, 1948, MCM & D; Donagh Crossroads near Lisnaskea, 1948, MCM & D; sandy fields below Gortaree, Slieve Rushen, 1949, MCM & D; Killee Lough, Ballinamallard, 2000, I. McNeill; Necarne estate, near Irvinestown, 2002, I. McNeill.

On account of taxonomic confusion and identification difficulties described below in the F. muralis subsp. boraei account, I have decided to amalgamate the records for these two taxa and simply map them as F. muralis agg.

Fumaria muralis subsp. boraei (Jord.) Pugsley, Common Ramping-fumitory

Probable introduction, rare. Oceanic southern-temperate, also widely naturalised.

1942; Mackechnie, R.; near Lisnarrick.

April to August.

Fermanagh occurrence

In the 1962 BSBI Atlas, Fermanagh was shown as being close to the NW limit in Ireland of F. muralis and the New Atlas confirms this impression. It is significant and should be noted that this species (or subspecies when listed as subsp. boraei) was recorded only once before 1960 in Fermanagh, yet 13 times subsequently, whereas F. bastardii (Tall Ramping-fumitory) was recorded six times before 1960, and just twice since (Revised Typescript Flora).

In the most recent (7th) edition of An Irish Flora, the editors remarked that in their view it is difficult, if not impossible, to separate Irish material of F. muralis from F. bastardii. Irish material of this species aggregate (ie the F. muralis agg.) is extremely variable and it does not readily conform to earlier descriptions based entirely on material from Britain. This is particularly so with the several subspecies of F. muralis and in the light of this expert taxonomic opinion and of my own difficulties experienced when attempting to key out the two forms (especially when using the earlier 6th edition of An Irish Flora 1977 as field guide). I am certain there has been confusion in the identification of the Fermanagh records for these two common large-flowered (corolla length over 9 mm) fumitories (ie F. muralis subsp. boraei and F. bastardii).

Nevertheless, for what it is worth, in these circumstances, RHN and the current author (Ralph Forbes) are listing in this work the relatively few records we currently have in the Fermanagh Flora Database for each of these taxa separately. With the above qualification the 13 records for F. muralis subsp. boraei, additional to the first one listed above, are: garden weed, Ballinamallard, July 1984, RSF & RHN; Glenross Td/Rossgweer Td, 3 September 1988, RHN; waste ground at Doon Td, SW of Derrylin, 18 July 1991, RHN; Lackaghboy, near Killyvilly, 10 May 1994, RHN; garden at Magheranageeragh, 1 October 1994, RHN; potato field 300 m W of quarry, Aghakillymand Td, 17 August 1995, RHN & HJN; Drumbrughas North, NW of Lisnaskea, 26 April 1996, RHN & HJN; roadside N of Killymackan, 25 July 2000, RHN; waste ground at Tullycleagh, 7 October 2000, RHN; gravel track at Corraslough Point, 17 August 2004, RHN & HJN; roadside waste ground at Tamlaght village, 3 July 2005, RHN; Enniskillen town, 18 August 2008, RHN; and disturbed roadside at Tattynuckle Td, 21 June 2010, RHN.

Obviously the situation regarding these fumitories is very unsatisfactory, and the whole genus requires detailed local re-investigation. The revision of the genus by Peter Sell in the BSBI Plant Crib (1998) will, I hope, provide a better basis upon which to work, and Murphy's Fumitories of Britain and Ireland (2009) handbook will certainly prove essential in future. The publication of Volume 1 of the critical Flora of Great Britain and Ireland by Sell & Murrell (2018) should also greatly enable better discrimination of the members of this genus.

Because of the taxonomic confusion and identification difficulties described above I have decided to amalgamate the records for this and the previous species and to map them as F. muralis agg.

Growth form and preferred habitats

F. purpurea has a loose inflorescence of pinkish-purple flowers, the petals and wings tipped with a darker purple, and it differs from the other large-flowered species, F. capreolata, in the fact that the raceme of flowers is as long as the stalk that carries them, not shorter. Like all the other species of Fumaria, it frequents disturbed, acidic, free-draining soils, and in this case it has a tendency to appear sporadically at some sites.

Fermanagh occurrence

The above is the only Fermanagh record of this plant (Revised Typescript Flora) and despite its date, it does not appear in the 1951 Typescript Flora.

Irish occurrence

As is the case with the slightly more common F. capreolata (White Ramping-fumitory), most records of F. purpurea in Ireland tend to be located at or near the coast. It is perhaps best thought of as merely a casual introduction in all or most of Ireland, occurring chiefly at E coast ports. In many instances, it is synanthrophic (ie closely associated with man and with artificial habitats) (New Atlas).

In Northern Ireland, F. purpurea is the rarest fumitory species by quite a long margin, having post-1970 records from just seven hectads in Cos Antrim and Londonderry (H39 & H40) (NI Vascular Plant Database).

British occurrence

F. purpurea is confined to Great Britain, Ireland and the Channel Isles (Guernsey), ie it is a scarce endemic species, but at a majority of sites it is merely a casual introduction (Jalas & Suominen 1991, Map 2089; Stewart et al. 1994; Murphy 2009). It is thinly distributed across Britain in the whole range of latitude from the Channel Isles to Orkney, but it does not reach Shetland. Again, as in Ireland, it is chiefly but not entirely coastal in its British sites. However, all comments on distribution have to be made guardedly, since very probably this species is overlooked or mistaken by many recorders and is therefore under-recorded (Preston et al. 2002).

Conservation status

Being endemic, F. purpurea is given a high priority in terms of conservation despite its weedy habit and its association with agricultural or industrially disturbed ground (Stewart et al. 1994).

Threats

None.

Growth form and preferred habitats

An erect to diffuse, much branched annual, germination of F. officinalis occurs in the spring and vegetative development and flowering is rapid in reasonably fertile, preferably light mesic soils, given open, sunny, warm growing conditions (Salisbury 1964; Sinker et al. 1985). Having said this, one of the English common names F. officinalis is 'Beggary', a name similar to 'Beggar Weed', that is often applied to such plants as Polygonum aviculare (Knotgrass), Heracleum spondylium (Hogweed), Spergula arvensis (Corn Spurrey) and Galium aparine (Cleavers), either because their presence denotes poor soil, or because they are such noxious weeds they manage to beggar the farmer (Britten & Holland 1886; Grigson 1987).

Fermanagh occurrence

Although this is indeed the commonest fumitory in Fermanagh, it is still a decidedly rare plant in this part of the world. It has only ever been recorded in 13 widely scattered Fermanagh tetrads, 2.5% of those in the VC. Just eleven tetrads contain post-1975 records. Apart from the occasional potato or turnip field, there really is no arable farming carried on now in Fermanagh. Annual ruderal weeds like this one have become confined to other forms of disturbed soil, eg in gardens, on piles of earth dumped beside roadside workings, on waste ground or building sites where they temporarily appear as casuals. As Sell (1991) pointed out in connection with both Fumaria and Papaver (Poppy) species, in the past agricultural methods allowed weeds to flourish in almost any crop. Nowadays, however, since the advent of modern arable management techniques, weeds appear only in crops that do not require heavy application of herbicides, eg onions and potatoes.

Variation

F. officinalis is variable and plastic with respect to the environment in which it grows, and it is also genetically variable – two subspecies comprising four varieties having been described by Sell (P. Sell, in: Rich & Rich 1988; Sell & Murrell 2018). Stace in New Flora of the BI (1997 & 2010) recognises just the two subspecies, as does Murphy (2009) in her Fumaria handbook. Our local Fermanagh form of the plant is probably subsp. officinalis, the most widespread form in both Britain and Ireland, but subsp. wirtgenii (W.D. Koch) Arcang. may also be present here. A plant collected at Lisnaskea was keyed down to this by RHN in 1988 but, unfortunately, no voucher was kept. This form has also been found at least once in Co Londonderry (H40) (FNEI 3).