Woodsiaceae. Athyrium distentifolium, Alpine lady-fern. Fronds bi-pinnate, ovate-lanceolate, yellow-green 20 cm to over 100 cm, arising in a ‘shuttlecock’ from a central crown. Stipe straight, one fifth to one quarter the length of the blade with pale brown scales, usually broad, but occasionally narrow. Pinnae widely spaced near base of the frond, and sometimes deflexed, crowded near the tip, meeting the rachis at an angle of 90° at the midpoint of the blade, or slightly ascending. Pinnules broadest at the base, tapering to a point. Sori circular, with irregular filaments making up a rudimentary indusium when young, soon obscured as the sori grow. Sori concentrated in the upper part of the frond, becoming sparse towards the base. Number of sori varies with size of frond. Plants frequently infertile (McHaffie et al. 2002).
Athyrium distentifolium var. flexile Newman's lady-fern. Fronds bi-pinnate, narrow, blue-green, 10–40 cm, arising from a central crown, erect or sharply angled near the base of the rachis to lie nearly flat against the substrate. Stipe short, one sixth to one eighth the length of the blade, often densely covered with broad, pale scales that may continue beyond the midpoint of the blade. Blade broadest near the base, or nearer the mid-point. Pinnae close together near base of frond, often strongly deflexed to at least half way up the frond. Uppermost pinnae widely spaced. Pinnules taper towards the base. Sori circular, sometimes with only a few sporangia; rudimentary indusium visible while the sorus is still immature. Sori concentrated at base of the frond, becoming less frequent towards the tip. Plants usually fertile (McHaffie et al. 2002).
Athyrium distentifolium is a chionophilous fern that grows at altitudes or latitudes which ensure long-lying winter snow cover. Deep snow provides insulation from severe winters, and prevents early growth in a mild spring. Only a limited number of species can tolerate these conditions and this reduces competition. Although previously known as a Scandinavian and continental European species, A. distentifolium was not recognized in Britain until 1844 (Newman 1851). Within Britain it is found only in Scotland where it is confined to the remote Highlands. Soon after it was identified, another similar taxon of uncertain taxonomic status was described (Backhouse 1852). Subsequent research has determined that this is a recessive variety which is now known as A. distentifolium var. flexile. Both taxa are covered in this account.
I. Geographical and altitudinal distribution
Within the British Isles, Athyrium distentifolium is found only in Scotland (Fig. 1) from 455 m to over 1220 m a.s.l. (Preston et al. 2002) on open screes with extended snow cover in the central and northern Highlands. Growing in the same habitat, but found only at a few sites, is A. distentifolium var. flexile (Fig. 2). This taxon has only ever been recorded in Scotland and occurs across the central Highlands from 600 m to over 900 m a.s.l. (Preston et al. 2002). A. distentifolium has a disjunct circumpolar arctic-montane distribution (Preston & Hill 1997), which is clearly displayed in its European ranges (Fig. 3). In south-east Greenland it is found on the coast (Devold & Scholander 1933) but in Austria it is found up to 2770 m a.s.l. (Polatschek 1997) (Fig. 3). The North American plants with longer pinnules are sufficiently distinct to have been named A. distentifolium ssp. americanum (Butters) Hultén (Flora of North America editorial committee 1993).
(a) climate and topographical limitations
In Scotland, Athyrium distentifolium grows only in sites where snow lies for extended periods during the winter. Snow can lie on the fern beds from October until May or even June. Precipitation on Athyrium distentifolium snowbed vegetation in the period 1941–70 varied from 1600–3200 mm in the west to 1200–1600 mm in the east Grampians. During the same period in the Central and North-west Highlands the January mean minimum temperature (screened at 1.25 m), was −5.0 °C, with a mean maximum of 1.0 °C (corrected by 0.5 °C and 0.7 °C, respectively, for each 100 m to give approximate temperatures at altitudes of around 800 m). The July mean minimum for the same period was corrected to nearer 6.0 °C, with a mean maximum corrected to 12.4 °C. The mean number of days with snow lying was 60–100 days in the Central Highlands. The snow did not lie for so long in the North-west Highlands, with a mean of 40–60 days (Meteorological Office 1979). These conditions combine to give cool summers, and extended snow cover in winter. In a mild spring at Glen Prosen in 1997, the ferns started to grow prematurely and were severely frosted (McHaffie 1998a). This illustrated the need for the protection of snow cover and the role it plays in controlling excessively early growth. Below snow cover an even temperature is maintained. Marchand (1987) established that beneath 40–50 cm of snow, small changes in density of the snow are unimportant and the temperature is almost constant at zero.
In the Cairngorms plants of both taxa grow on granite. Elsewhere the rocks are part of the Moine series, which are mostly acidic, or of Dalradian metamorphosed rocks of sedimentary origin, which can include limestone (Institute of Geological Sciences 1977, 1979; British Geological Survey 1987). Pl. Comm. Scot. gives a surface pH range of 4.8–5.4 from 22 Athyrium distentifolium sites in the central and north-west Highlands. The screes are free-draining but melt-water and seepage maintain an adequate level of moisture. Soil samples from field sites in the Central Highlands ranged from pH 3.2–4.5 (McHaffie 1999).
Two types of A. distentifolium communities were distinguished. The first is similar to McVean's Tall herb association (McVean 1964) on calcareous soil with rare Cicerbita alpina, frequent Dryopteris expansa, D. filix-mas, Polystichum lonchitis, Rumex acetosa and Sedum rosea. Grazing pressure has restricted this association to ledges so there are few natural montane meadows and all populations are above the present limit of woodland (McVean 1964). Quadrats sampled between Meall Buidhe and Beinn Achallader near Bridge of Orchy, together with the grass-rich quadrats from Creag Meagaidh, most closely resemble this association. Athyrium distentifolium var. flexile was frequent, together with Alchemilla alpina, Campanula rotundifolia, Deschampsia cespitosa, Dicranum scoparium, Gymnocarpium dryopteris, Hypnum cupressiforme, Luzula sylvatica, Plagiomnium undulatum, Pleurozium schreberi, and Rumex acetosa which distinguished this group. Cryptogramma crispa was present but less frequent. There was more Dryopteris expansa and Gymnocarpium phegopteris than had been observed elsewhere. Ferns were the major component of the vegetation. The closest NVC community to this (Rodwell 1992) is U16 the Luzula sylvatica–Vaccinium myrtillus tall-herb community in the U16a classification, the Dryopteris dilatata–Dicranum majus subcommunity, but it has close affinities with U18 and many of the species overlap. There is a significant proportion of Deschampsia cespitosa which was most marked in two of the Creag Meagaidh quadrats, one of which had up to 50% cover. McVean (1964) linked the presence of this species to high grazing levels and suggested that it replaced the tall herb association. Odland (1995) stated that the majority of Norwegian A. distentifolium stands are found among the Lactucion alpinae alliance in an association which he considered to be one of the least anthropogenically disturbed habitats. As the nearest approximation to this alpine meadow in Scotland is confined by grazing to cliff ledges, there is a greatly reduced representation of this type of vegetation, although it can be locally extensive. The second of McVean's communities (McVean 1964) is linked to poor soil and low fertility. The NVC classification (Rodwell 1992) related this second type of Cryptogramma crispa–Athyrium distentifolium snowbed vegetation to U18. The constant species are listed as Alchemilla alpina, Athyrium distentifolium, Barbilophozia floerkii, Cladonia bellidiflora. Cryptogramma crispa, Deschampsia cespitosa, D. flexuosa, Galium saxatile, Hylocomium splendens, Hypnum callichroum, Kiaeria starkei, Polytrichum alpinum, Rhytidiadelphus loreus, Rumex acetosa, Saxifraga stellaris and Viola palustris. This community is described as occurring among boulders around the steeper areas behind snowbeds. The Ben Alder and Beinn Eibhinn vegetation with A. distentifolium var. flexile corresponded most closely to U18. This group possibly indicates longer snow lie than the other subdivisions. Ben Alder has large populations of Cryptogramma crispa on the floor of the corrie. The Athyrium tends to grow higher up the side of the corrie but is intermixed with Cryptogramma. Some A. distentifolium also occurs in U11, the Polytrichum sexangulare–Kiaeria starkei snowbed community (Rodwell 1992).
Three quadrats from the margin of the scree at Bridge of Orchy together with two from Glen Prosen give an association particularly marked by Anemone nemorosa, Blechnum spicant, Festuca vivipara, Oreopteris limbosperma, Pellia epiphylla, and Viola palustris. This association was again similar to U18, but distinguished areas which were among the earliest to emerge from the snow, but also consistently moist. In view of the species present, this association suggested the Herb-rich birchwood described in Pl. Comm. Scot. with species indicating former woodland. These particular quadrats are in the lower-altitude sites, and the occasional cliff-bound Sorbus aucuparia and the Salix lapponum at Bridge of Orchy and Glen Prosen indicate the potential for montane scrub. In many other countries in which it is found, A. distentifolium grows in woodland extending to the upper limit of the tree line. Examples of these woodland communities are found, for example, in Scandinavian, Central European, Carpathian, and Siberian sites (Davis 1965; Odland 1995; Malyschev 2000; & Dolezal & Strutek 2002), but such divisions cannot be made in Scotland, other than from relict vegetation. In this respect Scottish populations differ from those elsewhere. Many of the open, rocky, acidic Scottish sites with low vegetation cover provide a specific habitat of a type that appears to be less frequent elsewhere.
IV. Response to biotic factors
Several authors reported the susceptibility of A. distentifolium to grazing (Moore 1859; Britten 1881; Cowan 1911; Adams 1930). Red deer are present at all the sites. At Ben Alder 50 clumps of each taxon were scored in a traverse across the corrie to assess the percentage of whole clumps which had been eaten. Separate populations were identified at intervals of 30–200 m and the first five clumps encountered were scored. Grazing was very variable depending on the terrain. Accessible plants among stable rocks were often grazed; those among large unstable boulders were not. Of the 20–30% of either taxon which had been grazed, up to 90% of the foliage had been removed (McHaffie 1999). This might represent a significant impact over time as these plants would have to use their reserves to produce new fronds.
Overwinter snow provides a specialized habitat which restricts competition (McVean 1958). Vigorous clumps of A. distentifolium shade smaller plants and Gjaerevoll (1950) described the prodigious amounts of litter which suppressed other plant growth. The dead fronds are slow to decay and form dense litter layers. Munther & Fairbrother (1980) found that fern fronds produce a toxic compound which can be leached out of the fronds by rain. The leachate from A. distentifolium helps to have similar effects. In a dense population this could also inhibit the growth of gametophytes near the parents and helps to explain the lack of young plants in large established colonies. As a taller plant, A. distentifolium could shade the smaller var. flexile.
(c) human influence
Land management has affected the grazing pressure through control of sheep stocking levels and the amount of deer culling. The habitat is less affected than lowland sites, although without grazing A. distentifolium would probably have been in the upper zone of woodland (Pl. Comm. Scot.); the ungrazed remnants among boulder screes occupy a semi-natural environment. Remote location, deep winter snow cover and high altitude give protection from recreational disturbance. The upland habitat is, however, very vulnerable to climate change and enhanced deposition of pollutants. Wilson et al. (1989) found that soils in the range pH 4.2–4.6 are vulnerable to further acidification. Fractionation of ions within a snowbed concentrates the acidic pollutants in the lower layers. Fifty to 80% of the ion load is released in the first 20% of snowmelt to give an acid flush, and bryophytes such as Kiaeria starkei, which occurs in A. distentifolium snowbed vegetation, have been damaged (Woolgrove & Woodin 1996). Melt waters have been found to contain high levels of nitrate and sulphate resulting in a pH as low as 3.2 (Lee et al. 1989). Acidification can lead to slower growth. Hill cloud forms where rising air, which may be polluted, cools at the condensation level. With increasing wind speed and droplet size these pollutants are transferred to plant surfaces (Grace & Unsworth 1988). This gives a greater acid deposition in upland rain as the higher precipitation in montane areas gives an increased input of pollutants. Nitrogen deposition in the Southern Uplands and south-west Scotland is 25–50 kg N ha−1 year −1 compared with 5–10 kg N ha−1 year−1 in lowland Scotland (Cannell et al. 1997) illustrating the higher deposition on high ground. Additional nitrogen could also cause a flush of premature growth and make the plants more vulnerable to frost damage (Lee et al. 1989).
V. Response to environment
Athyrium distentifolium is found in populations varying from a few plants to thousands. Other species, particularly ferns, are present but frequently A. distentifolium is the dominant species. The A. distentifolium var. flexile populations are generally smaller. At Glen Prosen there were fewer than 100 crowns of var. flexile. At Ben Alder there were several hundred clumps of var. flexile, as at Beinn Eibhinn and the main corrie at Bridge of Orchy. Four clumps were found at Creag Meagaidh. In Glen Einich, there was only one large clump and a single crown 1 m away. Other single clumps have also been recorded. There are 17 possible areas where plants of A. distentifolium var. flexile have been recorded, although sometimes only as single plants (McHaffie 1998a).
(b) performance in various habitats
There is a wide range in the reported height of A. distentifolium. Odland (1995) recorded frond sizes from 11 cm to over 150 cm. Schaminée et al. (1992) measured ferns up to 1 m high in the Massif Central, France. Davis (1965) noted southern European A. distentifolium that was 20–50 cm high. The North American A. distentifolium ssp. americanum has fronds up to 80 cm (Cody & Britton 1989).
In Norway, Odland (1995) found a correlation between frond size and fertility of A. distentifolium. The percentage of fertility showed a steady increase with frond size and beyond 71 cm all fronds were fully fertile. The tallest fronds which he recorded were over 150 cm. There was decreasing fertility with increasing altitude and in water stressed areas. The highest percentage of fertile fronds was in subalpine rich talus meadow at 500–900 m a.s.l. In very late snow beds the fronds were not fertile (Odland 1991).
A sample of Scottish fronds from sites where A. distentifolium var. flexile has not been found gave a mean of 46.7 cm. This mean is slightly larger than the mean for A. distentifolium at any of the sites where var. flexile has been found, the nearest being Bridge of Orchy with a mean of 45.8 cm. This is unusually large for A. distentifolium in var. flexile habitats but the var. flexile here is also larger than anywhere else, at 30.4 cm. The Beinn Eibhinn var. flexile was on average the shortest, 14.6 cm, together with the plants from Glen Prosen, 16.1 cm (McHaffie 1999).
Herbarium specimens of A. distentifolium from Continental Europe, Scandinavia and North America (McHaffie 1998a) had the largest mean, 56.9 cm, but the range of 28–139 cm included the size of fronds found in Scotland and does not exclude the possibility of coexistence with A. distentifolium var. flexile-sized plants. In Norway, Odland (1995) measured fronds from 11 cm to 150 cm high covering the full range of A. distentifolium sizes that occur elsewhere, with a potential niche for var. flexile, but although there might be suitable habitats for var. flexile in other countries, it has been found only in Scotland (McHaffie 1999).
(c) effect of frost, drought, etc.
During the growing season temperatures recorded at the Glen Prosen site did not fall below zero (Table 1). As many of the fern populations are located on steep slopes this gives added protection as cold air will drain away to a lower level.
Table 1. Means of the maximum, minimum and mean temperatures at Glen Prosen for the period 19 June to 11 September 1996 for the screened thermistors a-d placed on the ground at points around the rocks (± SE)
(a) Front of rocks
11.0 ± 0.2
16.9 ± 0.5
8.2 ± 0.3
(b) Below big rock
10.7 ± 0.2
18.7 ± 0.7
8.2 ± 0.3
(c) Back of rocks
9.3 ± 0.2
11.5 ± 0.2
7.6 ± 0.3
(d) Clump of ferns
10.3 ± 0.2
13.6 ± 0.3
8.2 ± 0.3
A maximum and minimum thermometer at Bridge of Orchy showed temperatures above zero in the growing season but lower during the winter period (Table 2). The minimum recorded in the period from the end of October to the end of April was −4 °C. This implied that snow would have covered the site for much of the time as a lower temperature would have been recorded otherwise.
Table 2. Temperatures at Bridge of Orchy 1996–97, recorded on a maximum and minimum screened thermometer
28 October 1996 to 28 April 1997
12 May 1997
3 June 1997
17 June 1997
1 July 1997
15 July 1997
29 July 1997
17 October 1997
Plants might be frosted in late spring or early in autumn but there is usually no frost during the short growing season (McVean 1958). Sato & Saki (1981a) used sporophytes of A. distentifolium in freezing experiments and found that they could withstand freezing to −15 °C for one day but were killed at −20 °C. This indicated that over-winter snow cover is necessary to protect from extreme temperatures. Large sporophytes of both taxa in plastic pots survived over winter 1994–95 in a cold frame with temperatures as low as −7 °C. This is more extreme than the lowest over-winter temperature at Bridge of Orchy of −4 °C (Table 2).
When gametophytes grown from Scottish spores were frozen the results were variable. Gametophytes of A. distentifolium from Bridge of Orchy regenerated after 1 day and 1 week at −6 °C. Ben Alder A. distentifolium var. flexile gametophytes regenerated after 1 day at −20 °C and 4 weeks at −10 °C. One gametophyte in a set of var. flexile gametophytes from Bridge of Orchy regenerated after 4 weeks at −20 °C. The −10 °C and −20 °C examples could have come from a thickened part of the thallus, which was frequently observed in cultivation (McHaffie 1998a).
Snowbeds above the fern beds that melted during the course of the summer were observed to provide seeping moisture at Ben Alder and Meall Buidhe near Bridge of Orchy. The lack of adequate over-winter snow in 1996–97 resulted in plants at Ben Alder suffering from a shortage of melt water.
In an investigation into the effects of drought, four sets of gametophytes and four boxes of senescent sporophytes of both taxa were allowed to dry out and given no water for periods from 4 weeks to 16 weeks. After the period of desiccation they were well watered. The gametophytes all died, even after only 4 weeks. The sporophytes recovered well after 4 and 8 weeks, but only three A. distentifolium and two A. distentifolium var. flexile sporophytes recovered after 12 weeks, and one A. distentifolium after 16 weeks. This suggested that the sporophytes can withstand periods of drought better than the gametophytes, but only for a few months (McHaffie 1998a).
Gametophytes on a temperature gradient bar were accidentally subjected to high temperatures for 48 h. The temperature at the 25 °C point on the gradient rose to 35 °C. Of the eight dishes of gametophytes, all but three were killed. These surviving three, Glen Doll A. distentifolium, Ben Alder A. distentifolium and Glen Prosen A. distentifolium var. flexile, were badly browned but produced new prothallial growth from the few cells which had survived. On the same occasion, the gametophytes that were normally maintained at approximately 20 °C experienced 29.2 °C and seemed unaffected. This suggested that extreme temperatures could be briefly tolerated in the wild and demonstrated the ability to regenerate from a few surviving cells (McHaffie 1998a).
VI. Structure and physiology
Athyrium distentifolium and A. distentifolium var. flexile fronds arise in irregular shuttlecocks from a central crown. Several authors refer to large branching rhizomes (Gjaerevoll 1950; Odland 1991; Page 1997). Rocks frequently slip downhill over the rhizomes and on excavation the plants are found to originate further up the slope. At Creag Meagaidh, one particular clump of A. distentifolium var. flexile took the form of a dispersed series of 11 crowns within a radius of 50 cm. Isozyme evidence failed to detect any differences suggesting that these plants were all the same clone. As it is so uncommon to find single crowns, this suggested that most populations are composed of long-established plants and that colonization by spores is an infrequent event. A plant of A. distentifolium var. flexile exposed in a rock fall was cut longitudinally. One frond base had a bud with two croziers. This illustrated the potential for offshoots from the rhizome and would also enable new growth in the event of injury to the actual crown, through severe frosting or mechanical damage. At least one root is produced at the base of each frond and this assists in anchoring the rhizome into the mobile scree (McHaffie 1998a).
A morphometric analysis of the two taxa gave two distinct groups. Apart from differences in the shape of the fronds a major difference is in the location of the sori. Athyrium distentifolium is less frequently fertile and sori are located at the tip of the frond. Athyrium distentifolium var. flexile is usually fertile and the sori are most dense at the base of the frond. When the sporangia were examined A. distentifolium var. flexile was found to have higher numbers of indurated cells in the annulus than A. distentifolium (McHaffie et al. 2002).
Van Cotthem (1970) classified fern stomata into five types; Athyrium stomata correspond to the polocytic type where the stoma is attached to the side of a single cell that is often horse-shoe shaped. The A. distentifolium stomata had a mean length of 49.4 µm (SE 1.07) and a range from 42 µm to 54 µm that was very similar to A. distentifolium var. flexile 48.8 µm (SE 1.05) with a range from 44 µm to 54 µm (McHaffie 1998a).
Stomatal frequency scores on individual plants were very variable. The mean basal counts of the Glen Prosen A. distentifolium ranged from 27 mm−2 to 86 mm−2 (mean 54.7, SE 6.1) and indicated how difficult it is to give an accurate generalized mean. A. distentifolium var. flexile from the same site had a mean range from 27 to 56 (Mean 43.5, SE 3.0). The position of a frond within the clump could give a greater degree of shading and many ferns, but not all, grew among the rocks. Athyrium distentifolium var. flexile, as a smaller plant, was more likely to be sheltered by rocks and sometimes only the tips of the frond were exposed. The horizontal habit of var. flexile would also shelter the underside more than an upright frond. Although there appears to be a difference in the stomatal frequency between taxa this is the inevitable result of the growth form of var. flexile and plants of A. distentifolium growing in similar situations can range down to these scores.
Ludlow & Wolf (1975) found fern species that always grow in the shade have higher chlorophyll content than ferns from sunny habitats, compensating for the lower photosynthetic rates. There is a marked colour difference between A. distentifolium, which tends to be more yellow-green, and A. distentifolium var. flexile which is usually a blue-green, indicating it is more a shade fern. This colour difference is maintained in cultivation where both taxa receive the same light levels.
Nespiak (1953) found no mycorrhiza in Athyrium alpestre (A. distentifolium) in a Polish Oxyrieto-Saxifragetum association, but this community suggests a high nutrient environment where mycorrhiza might be less beneficial (Read et al. 1976). Dominik & Nespiak (1953) recorded arbuscular mycorrhizas from A. distentifolium in Pinus mugo and Adenostyletum-alliariae associations. Dominik et al. (1954) also noted A. distentifolium roots colonized by arbuscular mycorrhizas in a Picetum excelsae myrtilletosum association. Samples of both taxa were examined from several sites in Scotland for arbuscular mycorrhiza and all showed some colonization ranging from 2% to 74% root length colonized. There was some evidence for increasing levels of colonization in the summer months and in less fertile soil (McHaffie 1998a).
(c) perennation: reproduction
Reproduction must be infrequent from spores as most plants have a multiple rhizome and gametophytes are rarely found. The plants appear to be long-lived and spread by budding sideways. It was found that when grown in conditions of high humidity in a glasshouse, some plants of A. distentifolium var. flexile from Creag Meagaidh and a site near Bridge of Orchy produced bulbils. These were usually confined to the axils of the lowest pair of pinnae and soon extended roots (McHaffie 1998b). This has not been observed in the field.
Manton (1950) counted the chromosomes of A. distentifolium, A. distentifolium var. flexile and A. filix-femina and for all of them found 2n = 80. This was confirmed by McHaffie (1998a).
(e) physiological data
The response to drought and freezing has already been described in V(C). Athyrium distentifolium and A. distentifolium var. flexile were found to behave differently in response to nutrients. Both gametophytes and sporophytes were grown at different nutrient levels (McHaffie et al. 2001). Under low-nutrient conditions, especially those with a pH of 3.8, or with a calcium, phosphorus or potassium deficiency, gametophytes of A. distentifolium var. flexile grew faster than A. distentifolium gametophytes. Similarly, sporophytes grown at different levels of nutrient showed different responses. Athyrium distentifolium required high levels of nutrient to be even partially fertile, while var. flexile was fertile with fewer nutrients (McHaffie et al. 2001). Where the two taxa grow together in the wild A. distentifolium is typically taller than var. flexile but frequently is not fertile. Only in areas with a high pH does A. distentifolium attain the full height, often in excess of 1 m, and become fully fertile. Where both taxa grew together the pH was low, ranging from 3.3 to 4.5 (McHaffie 1998a). It appears that A. distentifolium var. flexile is found only where the competition from A. distentifolium is reduced by the low-nutrient environment.
(f) biochemical data
Frond samples for allozyme electrophoresis were collected from 12 Scottish sites and sent to the research team at the Natural History Museum in London. The Scottish material included six sites where both taxa were present and six sites where A. distentifolium var. flexile had never been recorded. A few samples of A. filix-femina were also included. Samples of A. distentifolium were also sent from the Pyrenees in France, Southern Switzerland and Erzgebirge in Germany making a total of 23 populations with over 600 samples. Allozyme electrophoresis was used to test the species identity of A. distentifolium var. flexile in relation to A. distentifolium. Thirteen enzyme systems with 21 loci could be analysed and of these allelic variation was recorded for nine loci. There were no alleles specific to A. distentifolium var. flexile and there was more variation in the allele frequency between localities than between the two taxa within each locality. This showed that populations in separate corries have been isolated for some time. A cluster analysis of all the populations included in the study indicated three groupings. Plants from the Pyrenees were very distinct from all the others. All the Scottish populations with both A. distentifolium and var. flexile had a similar allele frequency together with a population with a doubtful old record of var. flexile and another population with no previous record. The final grouping included the remaining Scottish populations of A. distentifolium and the populations from Germany and Switzerland (McHaffie 1998a).
Athyrium distentifolium was also used as part of a study for the development of EST–SSRs using plant material from around the northern hemisphere. Polymorphic markers were found, suitable for biodiversity research (Woodhead et al. 2003; Squirrell et al. 2004) and a clear separation was seen between European plants and those collected in North America (Woodhead et al. 2005).
In Norway, where the snow could last until August, Odland (1995) found that A. distentifolium at 750 m a.s.l. usually commenced growth in June Frond expansion did not start until the soil temperatures at a depth of 5 cm reached 6–7 °C. It then took 24–27 days for the fronds to expand fully. This was faster than Oreopteris limbosperma or Matteucia struthiopteris which grew nearby (Odland 1991). In Central Europe, in the areas of longest snow cover, A. distentifolium did not normally emerge until the end of June and growth soon commenced (Schaminée et al. 1992).
Sato et al. (1989) found that ‘summer green’ ferns such as A. distentifolium and A. filix-femina in Austria were damaged by a late frost or snow fall. The fronds started to die back before sporangia could mature. The spores of Central European A. distentifolium are ripe in July or August (Davis 1965). Sato & Saki (1981b) reported that the spore dispersal period of A. distentifolium in Japan was in the first part of September only.
Scottish plants were monitored during 1996 and 1997 at a site below Meall Buidhe near Bridge of Orchy and on the slopes of Mayer in Glen Prosen, being north- and south-facing sites, respectively. In 1996 at both sites some ferns were released from snow cover several weeks before others, but frond expansion was delayed until June. In 1997 there was no snow on the sites in April and the fronds did not start to expand until the end of May at the north-facing site, but early May in the south-facing one. Within 2 weeks these early fronds had been frosted and there were many blackened croziers and growth started again by the beginning of June. These observations showed that fronds of A. distentifolium required 6 or 7 weeks to expand fully and the first fronds of A. distentifolium var. flexile followed a similar time-scale. Most of the season's fronds started to expand in the first 2 or 3 weeks with only a gradual addition of new fronds thereafter. In Norway, Odland (1991) measured individual marked fronds and found that they were completely expanded within 24–27 days, which is comparatively faster than the Scottish observations. The major part of the expansion in Scottish sites had occurred by 4 weeks, but Odland's specimens might have had more consistent, higher, continental temperatures. He gave a mean July temperature at an A. distentifolium site at 500 m as 12 °C, when the Glen Prosen mean was nearer 10 °C (Table 1), more typical of the variable, oceanic climate. In the areas of longest snow cover in Central Europe, A. distentifolium did not normally emerge until the end of June and growth began soon after (Schaminée et al. 1992).
VIII. Reproductive characters
(a) reproduction of sporophyte
Scottish plants have been reported to be frequently infertile (Backhouse 1852) and A. distentifolium var. flexile is fertile at a smaller size than A. distentifolium (Boswell et al. 1886; Page 1997). Some of the monitored plants did not produce any sporangia. One of 10 monitored A. distentifolium plants at Bridge of Orchy had been fertile in 1995, but was not fertile in either 1996 or 1997. Otherwise both taxa near Bridge of Orchy were more fertile than at other localities visited. At Ben Alder and Beinn Eibhinn A. distentifolium var. flexile was usually observed to be at least partially fertile, but A. distentifolium frequently was not fertile at all. The Glen Prosen A. distentifolium var. flexile plants were less frequently fertile than in many sites; eight were fertile in 1996 and seven in 1997. Five Glen Prosen A. distentifolium plants were fertile in 1996 and seven in 1997.
(b) discharge and dispersal of spores
In 1996 at Bridge of Orchy, 9 out of 10 A. distentifolium plants began to shed ripe spores at 10 weeks from first expansion (62% mean germination) rising to 93% at 12 weeks. The dead fronds yielded spores with a mean germination percentage of 92% at the end of October. Plants of A. distentifolium var. flexile at the same site had one plant with 28% germination at 8 weeks, 9 out of 10 plants with 54% mean germination at 10 weeks rising to 94% from 10 plants at 12 weeks and 95% germination at the end of October. At Glen Prosen only 5 out of 10 A. distentifolium plants produced spores with 31% mean germination from four plants at 10 weeks, all five had 97% germination at 12 weeks but only three plants yielded 3% mean germination at the end of October. Eight out of 10 of the monitored A. distentifolium var. flexile plants produced ripe spores at 10 weeks (62% mean germination) rising to 83% at 12 weeks and 84% at the end of October
The time of spore-shedding is determined by the length of growing season. The spores of American A. distentifolium were ripe in July or August (Davis 1965), as in Scotland. Sato & Saki (1981b) reported that the spore dispersal period of A. distentifolium in northern Japan was only in the first part of September which suggested late emergence and a short growing season. In very late snow beds in Norway the fronds were not fertile (Odland 1991). Presumably too short a season does not allow sufficient reserves to be built up for the production of sori. Wardlaw & Sharma (1963) experimented with Dryopteris dilatata plants and found the nutrient availability the previous year determined the following year's fertility.
Stunted late-season fronds of A. distentifolium var. flexile were able to produce spores as the sporangia at the base matured rapidly. Athyrium distentifolium was less likely to produce fertile fronds and was more vulnerable to grazing at the tips. There were many stages of maturity observed in sporangia on any one var. flexile frond. This gave the capacity to keep growing and producing fresh spores until finally frosted. The Glen Prosen var. flexile plants were good examples of this, with the continuous production of new fronds, and variable states of maturity along the length of the narrow fronds. The A. distentifolium fronds tended to mature more simultaneously and such a wide range of maturity was not observed (McHaffie 1998a).
(c) germination of spores
Samples of spores for each taxon showed that most plants bearing fertile fronds can produce at least some viable spores, ranging from only 5% to nearly 100%. Some spores were immature even though collected at the beginning of August, which gave lower germination. Athyrium distentifolium var. flexile spores showed a higher percentage of germination which is partly due to earlier maturity. It is not known whether spores germinate in the late summer immediately after they have been shed or whether the spores would meet with greater success by delaying germination until the following spring. Spores were exposed to various day lengths of 6, 12, 18 and 24 h. There was a difference in the rate of germination for spores from different sources but this accounted for more variation than the difference between various photoperiods. Only at 6 hours was a marked difference becoming apparent. No dark germination was found after 3 months (McHaffie 1998a).
The failure of dark germination in A. distentifolium and var. flexile indicated that they might be able to maintain a spore bank for some time, unlike Pteridium aquilinum which can germinate in the dark (Conway 1949). Soil samples were collected from Bridge of Orchy, Ben Alder and Glen Prosen before the current season's spores had been released. Nearly all produced gametophytes from which sporophytes of both A. distentifolium and var. flexile grew, in addition to other pteridophytes and mosses (McHaffie 1998a).
Spores from different localities were germinated on a temperature gradient bar. Spores from the Glen Prosen and Bridge of Orchy area reached their maximum germination faster than spores from Beinn Eibhinn. Spores germinated successfully and grew well from 10 °C to 25 °C. A set of spores was maintained at 5 °C for 16 weeks with no germination. The temperature was then raised to 6–7 °C and the spores germinated although growth was extremely slow. At 30 °C spores germinated but soon died and above 31 °C there was no germination. After seven days these spores were reduced to a lower temperature, but after 2 weeks they were presumed dead (McHaffie 1998a). Temperatures that had been suitable for germination were too high for sustained growth, and some of the gametophytes grown at 25 °C died. Most of the gametophytes grew faster at 20 °C, except the Glen Doll A. distentifolium and Ben Alder var. flexile that responded better to a lower temperature.
Hill (1971) found that species of ferns which inhabited open areas could germinate over a higher range of temperatures 10–35 °C than the 10–30 °C of woodland ferns. Although the A. distentifolium habitat in Scotland is open scree, many populations in other countries are at the upper limit of woodland. Germination up to, but not exceeding, temperatures of 30 °C suggest that A. distentifolium and var. flexile and might be more adapted to woodland conditions.
The temperatures recorded by a data-logger at Glen Prosen (McHaffie 1998a) showed that for parts of the day, temperatures could rise to 30 °C, but on only two occasions reached towards the lethal 35 °C (McHaffie 1998a) and then only in the one location at the south-facing front of the rocks. While germination can occur at 6 or 7 °C there is the potential to exploit higher temperatures if available.
(d) gametophyte morphology
The spores have a folded and ridged perispore giving the appearance of a reticulate surface (Page 1997). There is no difference between the two taxa in the shape of the spores (Fig. 4). Athyrium distentifolium var. flexile has a mean spore size of 34.1 µm (SE ± 0.4) with a mean maximum and mean minimum range from 25 to 48 µm. Athyrium distentifolium has a mean spore size of 32.3 µm (SE ± 0.4) with a mean maximum and mean minimum range from 25 µm to 45 µm. Their size is significantly different. (t76 = 3.3, P ≤ 0.001) although the measurements are very close. The mode for A. distentifolium is 32.8 µm and for var. flexile is 36 µm (McHaffie 1998a).
(e) reproduction of gametophyte
Cultivation experiments with A. distentifolium showed that apparently typical sporophytes of A. distentifolium can produce progeny with the morphology of either A. distentifolium or var. flexile. Sporophytes were grown from 37 wild-collected individual plants of A. distentifolium var. flexile and all but four resulting sporophytes were typical var. flexile. The four A. distentifolium were probably due to occasional contamination. Spores from 34 wild-collected A. distentifolium plants were sown and of these 14 produced only A. distentifolium sporophytes and 20 plants (all collected from sites where var. flexile was present), produced a mixture of 381 A distentifolium and 250 var. flexile. This suggested that var. flexile is a recessive form of A. distentifolium (McHaffie et al. 2001). Although there were occasional aberrant varieties of each taxon (McHaffie 1998b), sporophytes were usually either typical narrow-fronded var. flexile or the broader A. distentifolium. Gametophytes from heterozygous plants of A. distentifolium were isolated in single cells and an approximately 50 : 50 ratio of the two taxa was produced suggesting segregation at meiosis. Athyrium distentifolium var. flexile is due to variation at a single gene with pleiotropic effects which are expressed both at the gametophyte and sporophyte stage (McHaffie et al. 2001).
In a further experiment, 12 gametophytes of known homozygous A. distentifolium were paired with 12 gametophytes of var. flexile. Some of the gametophytes died, some might have self-fertilized as they produced the expected taxon, but three var. flexile gametophytes produced distentifolium sporophytes. This showed that A. distentifolium is the dominant taxon (McHaffie et al. 2001).
The distinctive frond morphology of both A. distentifolium and var. flexile was seen from the very first frond produced on a gametophyte. Athyrium distentifolium var. flexile was markedly more dissected (Figs 5 & 6).
(f) ecology of gametophyte
Gametophytes of A. distentifolium were rarely found in the wild. Sato (1982) observed that summer green ferns had prothalli that overwinter as gametophytes and might not be fertilized in the first full season. He found that sporophytes were produced towards the end of the second season and over wintered as very small plants. This is possibly the pattern followed by A. distentifolium and is supported by the discovery of large gametophytes found emerging from late snow in July 1986 at Caenlochan (H.S. McHaffie, unpublished data).
The hybrid of A. distentifolium and A. filix-femina (Athyrium×reichsteinii Schneller & Rasbach) has been recorded in Europe but not in Britain (Schneller & Rasbach 1984). The hybrid exists as a diploid similar in size to the parents, occurring as dense clones which compete well. It is morphologically intermediate with a round sorus and a small, but visible indusium. Most of the spores were abortive and none germinated. Triploid hybrids can also occur, with 120 chromosomes, counted from the root tip. These have hybrid vigour and two types were found, resembling one parent or the other. While wild hybrids are not uncommon in Switzerland, an attempt to synthesize hybrids was not successful. There was no evidence of introgression (Schneller & Rasbach 1984).
No parasites were recorded but several fungi were found on decaying fronds of A. distentifolium. Mycena cinerella Karsten, Mycena metata (Fr: Fr) Kummer, Galerina calyptrata Orton, and Ramariopsis subtilis (Pers: Fr) Corner sensu lato were found (McHaffie 1998a). These fungi are widely distributed and have no alpine affinities, also being found in broad-leaved woodland (Courtecuisse & Duhem 1995).
In the immediate postglacial period, from 14 000 years ago, snowbed communities would have been widespread at lower altitudes. Populations would gradually have become more isolated as the plants became restricted to a few high altitude sites. Increasing grazing pressure has further restricted the distribution.
The cold period known as the ‘Little Ice Age’ in the 17th and 18th centuries might have marked a return to more suitable conditions for A. distentifolium. The range could have been extended and at the end of this period new habitats might have been available on a scale which has not been equalled since the end of the most recent phase of glaciation. Ballantyne (1986) described a comparatively fast-moving lobe of solifluction which he linked to this period and indicated a disruption of the landscape. Present screes have become relatively stable and new habitats are not normally exposed very frequently.
This taxon was first recognized as a British species in 1844 at Caenlochan (Newman 1851). Moore (1859) gave alternative names for A. distentifolium, frequently without giving a source. Names quoted for A. distentifolium include: Aspidium distentifolium Tausch ex Opiz (according to Steudel 1841); Aspidium rhaeticum Swartz, 1800 (Moore 1859); Aspidium alpestre Hoppe, Tashenb. 1805 (Moore 1859); Polypodium alpestre Pl. Exc. Hoppe, 1829 (Moore 1859); Athyrium alpestre Nylander ex Ledebour (Moore 1859); Pseudathyrium alpestreNewman (1851) (Newman 1851); Athyrium alpestre (Hoppe) Rylands (Moore 1859); Athyrium distentifolium Tausch ex Opiz (Jermy et al. 1978).
Athyrium distentifolium var. flexile was first found in Glen Prosen in 1852 (Backhouse 1852; McHaffie 1997b). It was collected so heavily that it was no longer found there in the 1880s (Boswell et al. 1886) but by that time a large population had been found on Ben Alder which remains the best known and most visited locality. Two further populations were found near Bridge of Orchy and Beinn Eibhinn. As this variety has been found only in Scotland it might have arisen in the postglacial period during the last 14 000 years.
Previous names of the flexile taxon: Polypodium alpestre var. flexile Moore (Moore 1859); Athyrium alpestre var. flexile (Moore 1859); Polypodium flexile (Newman) T. Moore, non-Fée (Dandy 1958); Athyrium alpestre var. flexile Milde (Milde 1867); Athyrium alpestre ssp. flexile (Boswell et al. 1886); Athyrium flexile Syme (G. C. Druce 1932); Athyrium alpestre var. flexile (Newman) Druce (Fl. Br. Isl. 1952); Athyrium alpestre var. flexile (Newman) Milde (Fl. Br. Isl. 1962); Athyrium flexile (Newman) Druce (Page 1997); Athyrium distentifolium Tausch ex Opiz var. flexile (Newman) Jermy (Jermy et al. 1978).
Athyrium distentifolium is a scarce species (Stewart et al. 1994) and A. distentifolium var. flexile has been the subject of a UK Biodiversity Action Plan. Once the latter was proved to be a variety there has been less concern for its preservation. The single plants of var. flexile reported from large populations of A. distentifolium indicate that the recessive gene is present in more populations than the restricted distribution suggests. In the few large populations, at Ben Alder, Beinn Eibhinn and near Bridge of Orchy a combination of factors provide suitable conditions in favour of this variety. The habitat is vulnerable to climate change as less snow cover would increase the risk of the crowns being frosted. Some populations could migrate to a higher altitude but they tend to occupy the rocky ground at the foot of cliffs and would have a smaller area available to colonize if the altitudinal range was higher. Also, without the stress of snowbed cover other species would colonize the habitat and the competitive advantage would be lost.
I thank my PhD supervisors Colin Legg and Chris Sydes and also Mary Gibby. Thanks also to Johannes Vogel for the isozyme analysis and to the many helpful suggestions from the referees for this paper. The fungi were identified by Adrian Newton, Rosemary Smith, Mary Clarkson and Roy Watling.