This site and its content are under development.

Erica vagans L., Cornish Heath

Account Summary

Protected as native, very rare and vulnerable, but could possibly be an introduction. Oceanic southern-temperate.

1936; Dickie, Major; blanket bog near Black Bridge, above Belcoo.

June to December.

Growth form and preferred habitats

An evergreen, dwarf or tall shrub up to 80 cm tall with numerous ascending, erect branches without short axillary shoots. The glabrous leaves, 7-10 mm in length, are borne in whorls of four or five. The sub-terminal inflorescence is a dense, cylindrical raceme often terminated by leaves. The four-lobed flower corolla is bell-shaped, either pink or white and the purple anthers are fully exerted beyond the petals. E. vagans is a very rare, sometimes locally abundant and co-dominant shrub of heathland over ultrabasic rocks (serpentine or gabbro), or moist gleyed peat flushed with alkaline, base-rich groundwater.

E. vagans is long regarded as native on the Lizard Peninsula in Cornwall where John Ray found it 350 years ago, and it has also been recorded in most of the southern and midland counties of Britain, plus in a handful of widely scattered sites further NW into N Scotland and further east in Holland, generally appearing in these other areas as isolated bushes representing casual garden escapes (McClintock 1971; New Atlas).

Fermanagh occurrence

This really is a puzzling plant since it was first reported in Ireland in 1834, not in its extant remote W Fermanagh station, but rather from a coastal site in Co Waterford (H6), about which there arose an error in communication with J.D. Hooker, who had first published the find in his British Flora of 1834. Eventually this station was given in More and Moore (1866) as, 'Cliffs in Islandicane townland, west of Tramore'. The station was never confirmed and the original voucher specimen apparently disappeared after it was sent to Dr Ball, the referee of the time. However the location in the far south of the island is precisely where the species might most feasibly appear, mirroring the native occurrence of the plant on the Lizard peninsula in S Cornwall and minimising the disjunct distribution between the two stations. The Waterford population or individual plant was never refound so the report became regarded with scepticism (eg in Colgan & Scully 1898). The assumption grew that the record was a garden escape, and gradually the record was discounted. At the turn of the century Praeger totally ignored the Waterford record in his influential and important contribution to distribution studies, Irish Topographical Botany (Praeger 1901).

Two other subsequent Irish appearances of E. vagans have occurred: one plant found in 1899 on sandhills at Dundrum, Co Down (H 38) (Praeger & Megaw 1938), which was still present in 1978 (Nelson 1979); and a stony hillside west of Lough Swilly, Co Donegal (H35), where it was definitely planted and only fleeting grew (Browning 1928; McClintock & Rose 1970).

Imagine the surprise when a substantial established population of entirely white-flowered plants was discovered by an army Major from Enniskillen when out shooting! The shrub was found on a remote and at first glance apparently undistinguished moorland hollow and on a slope above a small nearby stream, referred to locally as 'the Black River', near Belcoo in W Fermanagh (Praeger 1938; Webb 1954b; McClintock & Rose 1970). It remains the solitary Irish site where this species appears well established, and it might possibly be indigenous at this site (but see later argument below on this topic).

Contrary to first impressions, the Fermanagh E. vagans site is ecologically quite distinctive. Investigations carried out between 1952 and 1970 by ecologists and taxonomists has shown that almost all the ground now known as Carrickbrawn ASSI is flushed by a spring that seeps alkaline, lime-rich water through the rather sandy mineral substrate underlying the shallow moorland peat. The soil here is derived from sandstone and shale, overlain with a thin layer of sedge and moss peat (Webb 1954b). The flushed vegetation here is comparatively species-rich with 56 flowering plants species and 26 bryophytes recorded by McClintock & Rose (1970).

Within its very restricted, constantly moist, but not waterlogged site, E. vagans was previously abundant and co-dominant with Molinia caerulea (Purple

Moor-grass), Carex spp. (Sedges) and Juncus acutiflorus (Sharp-flowered Rush). Beyond the species-rich, lime-flushed ground, apart from a few outlying clumps of the shrub, some of which are on the stony 'cliff' above the stream within a short distance of the main E. vagans area, this unusual heather species is surrounded by completely ordinary Calluna-dominated moorland vegetation containing just 22 flowering plants species and 19 non-flowering species (McClintock & Rose 1970).

Growth in garden cultivation

The fact that Cornish Heath is such a widely appreciated garden subject, with 15 horticultural varieties, pink, white or cream flowered being listed in the RHS Index of Garden Plants (Griffiths 1994), highlights the reality that the species does not require or prefer base- or lime-rich flushed soil conditions in order to grow, but simply that it tolerates such conditions better than other heathers. In the wild, it is essentially this property which enables Cornish Heath to compete and persist on shallow, relatively dry or well-drained, somewhat lime-flushed peatland surrounded by Calluna vulgaris and its associated species. Basically, E. vagans avoids competition with these more aggressive species by tolerating conditions which they cannot. In garden cultivation, E. vagans actually prefers and grows best on a heavier, peaty soil and Flora Europaea also describes it as a calcifuge (ie lime hating or lime avoiding) species (D.A. Webb & E.M. Rix, in: Tutin et al. (eds), 1972, p. 7)). E. vagans also tolerates shade better in the garden setting than most other heathers, continuing to flower profusely, even under significantly darkened conditions (Grey-Wilson 1989).

Growth and ecology at the Fermanagh site

The Carrickbrawn E. vagans colony has definitely contracted, perhaps by 75% during the last 60 years, very possibly due to changes in the grazing regime. Sheep replaced cattle at the site 30 years ago and the thus a modified trampling and grazing regime could well be responsible for the observed losses. Early reports of the colony by Praeger and by Webb both refer to vigorous regeneration after fire and grazing, involving both re-sprouting of stem bases and the appearance of seedlings between the somewhat elevated clumps of mature Cornish Heath (Praeger 1938; Webb 1954b). Both these visitors recognised the limited population variation at Carrickbrawn and absence of pink flowers − unlike at the Lizard in Cornwall, where there is an almost equal mix of lilac to pink coloured flowers and white ones; the entire Fermanagh population is white flowered. The white flowers suggest that the Irish colony may be derived from one (or more) original homozygous plants, which would mean that if sexual reproduction has taken place (and elsewhere the species is considered self-fertile), the colony would retain its uniformity (Nelson & Coker 1974).

At present there is no evidence of recruitment from seedlings at the Fermanagh site. Prof Webb, in particular, would have been very alert to the possibility of population sterility, since at the time of his visit he had just discovered this to be the case in Erica mackaiana (Mackay's Heath) in both Connemara and Donegal (Webb 1954a). At Carrickbrawn, he described finding seedlings of Cornish Heath, "frequent on the intervening stretches of cattle-grazed turf", ie between the peaty hummocks on which the larger bushes of E. vagans then grew (Webb 1954b, p. 216).

In terms of genetics, small isolated populations inevitably lose their vigour over extended periods of time as they gradually lose genetically variability (a process sometimes referred to as 'genetic erosion'). This is a consequence of their restricted gene flow and 'genetic drift' (ie the tendency for gene allelles to fix within small populations at random, or even somewhat against selective forces)(Richards 1997, pp. 46-49). In small populations, these processes associated with inbreeding, lead to the accumulation of both homozygous gene alleles and deleterious recessive mutants, but in polyploid species, such as e.g., most Pteridophytes, these weakening processes limiting sexual reproductive success, may be very slow indeed to operate.

The lack of seed germination in the limited samples collected by Nelson and Coker (1974) at Carrickbrawn does not provide sufficient evidence to enable us to conclude that the population is totally sterile. These authors themselves remarked that small plants growing on the cliffs overhanging the stream could have been produced only from seed, although just to be safe they qualified this verdict by commenting, "unless very small fragments of parent plants had successfully produced adventitious roots and become established''. We believe the improbability of vegetative reproduction by fragments rather than seedling establishment in the river 'cliff' environment merits very little serious consideration, and in the light of Praeger's and Webb's observations, and with our own knowledge of the occasional nature of heather seedlings - typically associated only with fire or with other major habitat disturbance -, we believe that a low frequency of successful sexual reproduction probably still is involved in the maintenance of this population.

This opinion is strengthened and coloured by the fact that dwarf heath plants of any ericaceous species are generally not long-lived, at least in comparison with other woody plants. Dwarf heather shrubs normally survive only 15, 20 or 25 years, or very exceptionally up to 30, perhaps 50 years maximum. Regular grazing and occasional burning helps keep woody crowns and underground parts of ericaceous subshrubs rejuvenated, but in any heath, older individuals will inevitably die from time to time, and they must be replaced if the population is to maintain itself. The contraction of the overall colony size might well be a reflection of such inevitable deaths, but the density of the remaining shrubs argues otherwise. Close study of our autumn 1978 photographs of E. vagans at Carrickbrawn shows that there is some die-back of tall, leggy stems within otherwise vigorous, actively flowering clumps of the heather.

Another consideration is the possibility of a genetic reservoir of Cornish Heath seed persisting in the soil. However, while seed of other heather species is known to remain viable for 30 years or more, unfortunately we have not been able to trace any mention of E. vagans seed longevity in the literature (Thompson et al. 1997).

A detailed investigation is urgently required of the reproductive ability of E. vagans and its current powers of regeneration at this site, perhaps including a controlled burn of a portion of the colony.

Contraction in size of the Fermanagh colony

When examined by Praeger in August 1937, shortly after its discovery, the main area of the E. vagans colony in Fermanagh was closely grazed, apparently chiefly by cattle, which must also have damaged it to some extent through trampling, although sheep, goats and donkeys have also been seen at times in the vicinity (Nelson & Coker 1974). The owner of the site has stated that cattle grazed the area up until 35 years ago, when sheep were substituted and only they have browsed the ground since.

During his visit Praeger (1938) reckoned the patch of E. vagans covered an area of 75 × 50 yards [69 × 46 m] plus a few outliers. In the early 1950s, the size of the colony was roughly estimated by Prof. Webb to cover a 50 m square (Webb 1954b), but by 1966 and 1970 it had contracted to occupy just 50 × 30 m plus the small outliers (McClintock & Rose 1970; Nelson & Coker 1974). Curtis & McGough (1988) in The Irish Red Data book of Vascular Plants estimated that there were then around 500 plants covering 45 × 30 m.

A visit by the current author and RHN in 2003 found that the plot where E. vagans was still densely present measured approximately 40 × 30 m, but the individual clumps of the plant were smaller in size than the two of us had previously noticed them, and Molinia caerulea (Purple Moor-grass) and Myrica gale (Bog-myrtle) appeared to be more obviously dominant and overgrowing them (Northridge & Northridge 2004). The fact that the colony appears to have shrunk by around half its reported size during the 70 years it has been known is undoubtedly significant. The reduction is too large to be imaginary or the result of careless estimation and we can be certain that eminent naturalists like Praeger and Webb made no such mistakes. A survey conducted in September 2004 by conservation staff of EHS, using satellite positioning (GPS), found the colony covered 1577 m2, slightly larger than a previous estimate they had made and very similar to the autumn 1970 estimate made by Nelson & Coker (Nelson & Coker 1974), although the plants currently appear very overgrown and less vigorous and, indeed, had to be searched for (P. Corbett, pers. comm., December 2004).

The importance of herbivory and trampling

It appears to from the above observations that the role of herbivory in maintaining E. vagans population competitiveness and co-dominance with Molinia caerulea, Juncus acutiflorus and Myrica gale on the Fermanagh site has probably been underestimated (Hulme 1996). The replacement of cattle by sheep could well be responsible for the observed decline of the E. vagans population. Direct effects such as changes in selective browsing and in the level and timing of overall grazing pressure, could lead to increased competition from more unpalatable shrubs and graminoid associates. The heavyweight trampling disturbance which previously created vegetation gaps and thus assisted E. vagans seedling establishment has been greatly reduced by the change from cattle to sheep. The vegetation gap deficit would certainly limit the species' successful sexual reproduction. In addition, cattle are more likely to graze new spring shoots of the deciduous grass Molinia caerulea, a serious competitor of Cornish Heath on this stretch of bogland. On the other hand, sheep are more likely to graze young heather seedlings than cattle, and it is at the seed and seedling stages that herbivory principally influences plant mortality, both directly and indirectly (Harper 1977; Watt & Gibson 1988).

A variety of animals, including birds, insects, molluscs and mammals forage on seeds and seedlings in a frequency-dependent manner and, indirectly, this may affect interspecific competitiveness (Hulme 1996). Cattle were probably also responsible for more breakages of older woody tissues, which might both stimulate and assist vegetative regeneration of Cornish Heath on the site, by fragmentation and by layering of surviving trampled stems. Undoubtedly there are further indirect effects of a major change in herbivory involving differences in soil nutrient cycling, the significance of which would be more difficult to quantify, but one only has to consider cow and sheep manure for a moment to realise that such differences really can be significant (Crawley 1983; Hulme 1996).

Status of the Fermanagh colony

Nobody knows for certain how long the E. vagans colony has been at its Fermanagh site. It might be less than a century, although this would be greatly stretching the imagination in view of the observed low competitive ability of the species, the narrow range of ecological circumstances where it achieves establishment and persistence, and its probable, consequent slow growth rate. Alternatively, like other ericaceous subshrubs in Ireland, it might prove to be an ancient, genetically attenuated remnant of a larger prehistoric population of wider variability, ecological amplitude and geographical range, currently surviving in this one very isolated locality. If this is the case, it has become very disjunct from its Cornish nearest neighbours. Whether it arrived naturally or with the assistance of man, cannot yet be decided. What is certain is that Carrickbrawn is an extremely unlikely place for anyone to choose to deliberately plant any species.

British occurrence

Although E. vagans has been recorded at around 20 scattered sites in Britain in the wild, either as a garden escape or deliberately planted out, it occurs as an established species only on the Lizard peninsula in S Cornwall. Due to widespread losses of lowland heaths in England, nowadays it is gone from two of the Cornish hectads and has become restricted to just four of such squares and a total of less than ten sites (Perring & Farrell 1977; D.E. Coombe, in: Wiggington 1999; Preston et al. 2002). Despite the reduced distribution, it is still co-dominant here with Molinia caerulea and Schoenus nigricans (Black Bog-rush) on silty clay and on neutral to slightly acid gley soils derived from serpentine and gabbro rocks. The vegetation it occupies is described locally as 'Tall Heath', since on the Lizard Peninsula shrubs of Cornish Heath often grow up to 80 cm high − at least 20 cm more than is normally the case in Fermanagh.

European occurrence

E. vagans is a member of the small group of plants (15 at most), with disjunct geographical distributions strictly confined to mild oceanic climatic conditions spanning some or all of N Spain, W & NW France, SW Britain and S & W Ireland. The Latin specific epithet 'vagans' is the present participle of 'vagor' and translates as 'wandering' or 'of wide distribution', and the former seems a lot more appropriate in this case (Gilbert-Carter 1964; Stearn 1992). From an Irish perspective, this group is often referred to as either the 'Hiberno-Lusitanian', the 'Hiberno-Cantabrian' (Praeger 1934, paragraphs 35, 37), or sometimes simply as the 'Atlantic' element in the Irish flora, although the latter encompasses a more northerly extension than is appropriate to this discussion and is therefore best avoided (Perring 1962; Webb 1983).

Alternatively, like other disjunct ericaceous species such as Arbutus unedo (Strawberry Tree) in Kerry and Sligo (Mitchell 1993), Daboecia cantabrica (St Dabeoc's Heath) and E. erigena (Mediterranean Heath) in Connemara, E. mackaiana (Mackay's Heath) in Connemara and Donegal, Erica ciliaris (Dorset Heath) in Connemara and Cornwall (Webb 1966; Webb & Scannell 1983; Rose et al. 1996), E. vagans might prove to be a fairly ancient, genetically attenuated remnant of a larger population of wider variability, ecological amplitude and geographical range, currently surviving in Ireland in just this one highly disjunct Fermanagh locality.

The extremely restricted NW Irish distribution of E. vagans is mirrored to various extents by all the other ericaceous species in the W & N of Ireland listed above, and indeed the population of E. ciliaris near Clifden in Connemara is very much smaller than that of E. vagans at Carrickbrawn. It consists of just five entirely vegetatively reproducing plants, which did, however, recover very well after a severe bog-fire in May 1966 (Webb 1966). This tiny Irish population of E. ciliaris, so small that it was entirely lost for 119 years, is distinct from native English populations of E. ciliaris in Dorset and Cornwall (vcs 1, 2 & 9), in that it never has glandular tips on the stout marginal hairs (the cilia) of its leaves, a feature which only about 10 % of English plants display (Webb 1966; McClintock 1968; Nelson 1989). However the fact that this tiny bog roadside population of E. ciliaris has neither increased nor decreased over the 160 years since its first discovery, does suggest to some botanists at least, that it probably is a human introduction, transported by unknown means (Matthew Jebb, pers. comm., Dec. 2004).

In common with all these other 'Lusitanian' plant species except Arbutus, there is no fossil evidence to prove, indicate, or even as much as suggest the presence of E. vagans anywhere in Ireland in the Late-glacial or the early Post-glacial periods (Mitchell & Watts 1970). This is not terribly surprising since the flowers are chiefly - but probably not exclusively - insect-pollinated, and although the anthers are exerted from the corolla tube in E. vagans, and the plant flowers profusely, the chances of pollen or fruit and seed preservation and recognition of them at species level, are not very great, even if it were once a much more common species than it is at present. While not of course conclusive in itself, the lack of fossil evidence tends to suggest that at least the more frost-sensitive members of the Lusitanian plant group such as Arbutus and Daboecia could not have survived peri-glacial conditions in the vicinity of Ireland during the last cold spell, and that species perhaps including E. vagans must have survived further south and arrived here more recently during the current warm interglacial period, called the 'Littletonian' in Ireland (Mitchell 1965).

In addition to E. vagans, five other members of the Heather Family belong to the 'Lusitanian' group, Erica ciliaris (Dorset Heath), E. erigena (Irish Heath), E. mackaiana (Mackay’s Heath), Daboecia cantabrica (St Dabeoc’s Heath) and Arbutus unedo (Strawberry-tree), the last four of which are oddly absent from Britain. The 'Lusitanian' distribution pattern is followed by E. vagans in all the regions listed above, except that in Ireland it occurs in Fermanagh, unlike the other Ericaceous members that are mainly or entirely found in Connemara or Kerry. It is also the case that in N Spain and in France, E. vagans extends further inland than do most other species of the Lusitanian group (Praeger 1938; Webb 1983; for a map see Nelson & Coker 1974, Fig. 1).

Threats

The population is vulnerable, especially to fire, but also to overgrazing by sheep and to any changes reducing an already shrinking population with only limited genetic variation. Myrica gale, Molinia caerulea and other common graminoid plants appear to be exerting considerable competitive pressure upon E. vagans, severely limiting its reproduction, or perhaps completely preventing it. Careful monitoring and active conservation management directed at encouraging the species survival, based on detailed scientific research of the local population, is urgently required.