S J Leach yPeytoureau M N Sanford wigBAlJr ';' - ;/;i mi u • c ♦ BOTANICAL LINKS IN THE ATLANTIC ARC Proceedings of an Anglo-Hiberno-French meeting arranged by the Botanical Society of the British Isles 8th-12th May 2003 Camborne, Cornwall EDITED BY S. J. Leach, C. N. Page, Y. Peytoureau & M. N. Sanford Botanical Society of the British Isles with financial support from English Nature 2006 HISTORY 2 2 SEP 2806 PURC ENGLISH NATURE Conference Report No. 24 , a s' rs< 'rttCcr( * * Contents Preface . Dedication . Resumes franyais . The geological background to the landscape of the Atlantic Arc. C. M. Bristow Phytogeography of Ireland, past and present. S. Waldren, P. Coxon & N. KINGSTON The Mediterranean-Atlantic and Atlantic elements in the Cornish flora. C. D. Preston & H. R. Arnold . 1 5 7 19 31 41 Fern range determination within the Atlantic Arc by an environment of complex and interacting factors. C. N. Page . 59 An introduction to the flora of the Isles of Scilly. R. Parslow . 65 Maritime communities as habitats for Ophioglossum ferns in the Isles of Scilly. R. Parslow 73 Atlantic Arc grasslands: the example of the rhos pastures of south-west Wales and their conservation. R. D. Pryce . 83 Celtic Hedges as refuges for fern diversity in predominantly agriculturalised landscapes. C. N. Page . ■ . . 93 The task of the Conservatoire Botanique National of Brest in the knowledge and conservation of the Armorican flora as illustrated by Trichomanes speciosum Willd. S. Loriot & S. Magnanon . 103 The Isle of Oleron orchids. M. Breret . 113 An introduction to the flora of the Lizard peninsula, Cornwall. J. Hopkins . 119 The Lizard trackways project. A. J. Byfield & B. R. Wheeler . 127 Semi-natural vegetation on National Trust Land in Cornwall, with an emphasis on that of the Lizard peninsula. J. Bruce, L. Cordrey, J. Harvey & A. Cameron . 133 Genetic variation in Irish threatened plant species: a European perspective. R. J. Smith & S. Waldren . . 137 Biogeography of the Irish ‘Lusitanian’ heathers. N. Kingston & S. Waldren . 147 Ecological profiling as a means of improving mud plant conservation. R. V. Lansdown ... 157 Cornish bryophytes in the Atlantic Arc: cell biology, culturing, conservation and climate change. J. G. Duckett, S. Pressel, J. Burch & R. Ligrone . 165 Erica erigena R. Ross (Irish Heath) in the Medoc region of S.W. France. C. Lahondere ... 177 Euphrasia vigursii Davey (Vigurs’ Eyebright) in Devon. R. HODGSON & N. BALDOCK ... 181 The integration of wildlife information in Cornwall. S. Myles . 185 Cystopteris diaphana (Bory) Blasdell (C. viridula (Desv.) Desv.) new to Britain. R. J. Murphy & F. J. Rumsey . 191 Database developments and biological recording in Cornwall. C. N. French . 195 Assessment of threats to populations of Rumex rupestris Le Gall (Shore Dock) in Britain and France. F. Bioret & R. Daniels . 201 Rumex rupestris Le Gall (Shore Dock) in S.W. England: review of recent surveys and assessment of current status. E. J. McDonnell & M. P. King . 211 A review of recent work on the Limonium binervosum aggregate (rock sea-lavenders) in the British Isles. S. J. Leach & D. A. Pearman . 217 Recent surveys of endemic Limonium (Rock Sea-lavender) taxa in S. Devon (v.c. 3). S. J. Leach . 225 Vegetation and habitats of the western European endemic Asparagus prostratus Dumort. (Liliaceae), Wild Asparagus. T. C. G. Rich, L. K. Rich, S. B. Evans & A. E. Evans ... 231 A strategy for the conservation of Genista pilosa L. (Hairy Greenweed) in Picardie, France. P. Saliou & J-C. Hauguel . 243 Romulea columnae Sebast. & Mauri (Sand Crocus) refound in Cornwall after 121 years and Jnncus capitatus Weigel (Dwarf Rush) new to E. Cornwall (v.c. 2). I. J. Bennallick . 247 Hypericum linariifolium x humifusum: a hybrid adapted to the Atlantic coast. R. A. Jones & M. D. Sutton . 251 Lotus angustissimus L. (Slender Bird’s-foot-trefoil) in S.W. England. S. J. Leach . 255 Puccinellia foucaudi (Coste) Hackel on the French mid-Atlantic coast: taxonomic status, morphology and distribution. C. Lahondere . . 257 Status of Teucrium scordium L. (Water Germander) in north-west France. B. Toussaint, B. Destine & F. Hendoux . 261 Restoration and conservation management for the habitat of Mibora minima (L.) Desv. (Early Sand-grass) in northern France. F. Basso & F. Hendoux . 269 Viola kitaibeliana Schult. (Dwarf Pansy) in the Isles of Scilly. R. E. Randall . 277 Polygonum maritimum L. (Sea Knotgrass) in Cornwall. R. J. MURPHY . 281 Mentha pulegium L. (Pennyroyal) in Britain and Ireland. R. D. PORLEY . 285 Hypericum undulatum Schousboe ex Willd. (Wavy St John’s-wort) in S.W. England and Wales. R. D. Porley . 289 Conservation of Spiranthes romanzoffiana Cham. (Irish Lady’s-tresses) in Scotland - the role of twin lateral bud production. R. Gulliver, M. Gulliver, E. Grant, J. Robarts, M. Keirnen, S. Jonch Moller, A. Beare & C. Sydes . 291 The turlough form of Ranunculus repens L. (Creeping Buttercup). S. Waldren, D. Lynn & S. Murphy . 301 Creeping Water-plantain ( Dyfr lyriad ymlusgawl), Baldellia ranunculoides subsp. repens (Lam.) A. Love & D. Love in Wales. R. A. Jones . 311 Isoetes histrix L. (Land Quillwort) on the Lizard peninsula. R. J. Murphy & C. N. Page ... 321 Pilularia globulifera L. (Pillwort) in Cornwall. I. J. BENNALLICK . 325 Ranunculus ophioglossifolius Vill. (Adder’ s-tongue Spearwort) in the Nord/Pas-de-Calais region of northern France. C. Blondel . 327 Echium plantagineum L. (Purple Viper’ s-bugloss) at Boscregan, St Just - its conservation and management. L. Butterfield . . List of participants . , . Colour Plates . 331 334 337 1 PREFACE This volume comprises a series of papers and posters presented at the symposium ‘Botanical Links in the Atlantic Arc’, organised by the Botanical Society of the British Isles (B.S.B.I.) and held in May 2003, in Camborne, Cornwall. This meeting - the brainchild of the late Dr Franklyn Perring - gave botanists an opportunity to come together to celebrate the wild plants of the Atlantic coastal regions of Europe, and to consider a whole range of topics concerning their distribution, autecology, conservation and management. The meeting was a resounding success, attracting more than one hundred delegates from Ireland, Scotland, Wales, England and France - and even one from Sweden! Participants learnt a great deal from one another, and everyone involved contributed to the friendly and informal atmosphere that allowed new links to be made between botanists working, often in considerable isolation, along the Atlantic seaboard of Western Europe. Even for those of us who already knew each other, it was a pleasure to replace the bland anonymity of e-mails and the Internet with a few days of face-to-face contact. The Atlantic Arc, as adopted here, forms a cohesive biogeographic region from Britain and Ireland in the north and west, through Nord/Pas-de-Calais, Normandy, Brittany and around the Bay of Biscay as far south as the northern and western Iberian peninsula. This might appear to be unduly restrictive, and others might well feel that the southern limit could legitimately extend southwards as far as the extreme north-west tip of Africa, as well as embracing the Canary Islands, Madeira and the Azores. Equally, the European Union’s ‘Atlantic Biogeographical Region' extends northwards as far as Scotland and western Denmark, and takes in much of northern and western France as well as the coastal regions of Spain and the extreme north of Portugal. The meeting was, however, principally concerned with studies of a range of plant species and species- assemblages occurring on or close to the Atlantic-facing coasts of Ireland, Britain and France (and, to a lesser extent, northern Spain). This is an extensive and diverse area, it is true, but one united by the weather, where climatic extremes - especially winter cold - are ameliorated by the close proximity of the Atlantic Ocean. Under such mild conditions, growing seasons tend to be long, supported by typically light and frequent rainfall, and often a high humidity and extensive cloud- cover. The geography of the Atlantic Arc is dominated by the sea, into which its land areas intrude mostly as disparate and geologically varied assemblages of islands, archipelagos, coastal strips and peninsulas. The fact that these land areas fall within many separate countries - with their peoples speaking several different languages - may help to explain why there have been so few attempts to bring people together to consider them as a single cohesive phytogeographic unit. Yet, as we discovered, the botanical similarities between these areas far outweigh their differences. Indeed, they have a great many species and plant communities in common, and there are two that can serve as examples to illustrate this botanical cohesiveness: Rumex rupestris (Shore dock), which has a world distribution that corresponds closely with the area embraced by the Atlantic Arc, from Galicia in northern Spain up the west coast of France, and then in the Channel Islands, S.W. England and Wales; and Erica erigena (Irish heath), which has a markedly disjunct distribution extending from central and southern Portugal and Galicia in the south to western Ireland in the north. Throughout the symposium, we were able to share valuable insights into the biology and conservation of a host of such species, including quite a few rarities. Many papers and poster displays presented the results of studies undertaken to answer quite specific questions relating to the conservation of rare taxa - and these included a number of species that are legally protected or for which there are regional or national conservation plans (e.g. in Britain, species listed on Schedule 8 of the Wildlife & Countryside Act and/or included in the U.K. Biodiversity Action Plan). Indeed, contributors often laid considerable emphasis on the opportunities and challenges associated with the conservation of these rare species and the habitats in which they are found. For a large number of species the Atlantic Arc is clearly an important refuge - not just an ‘arc’, but also an Atlantic Ark. 2 ATLANTIC ARC As editors, we wish to thank all those who helped in the organisation of the symposium and the field meetings that followed it, including Ian Bennallick, Jane Croft, Rosaline Murphy, Alistair Cameron, Rachel Holder, Ray Lawman and various members of the B.S.B.I. in Cornwall. We also thank the authors of papers and poster presentations for their contributions, and for their patience during the editorial process. In particular, we wish to congratulate our young French guests from Conservatoire Botanique National de Bailleul and Conservatoire Botanique National de Brest who were bold enough to present their papers in English to a predominantly English-speaking audience, thus illustrating the obvious need for mutual understanding between European peoples. They contributed with friendly daring to the dialogue between British and Irish botanists and those from ‘the Continent’ - an excellent example of Entente Cordialel We are grateful to colleagues who helped to referee papers; particular thanks to Chris Preston for providing comments at very short notice on the ‘final drafts’ of so many papers. We are especially grateful to Yves Peytoureau (Secretary, Societe Botanique du Centre-Ouest) for translating the English abstracts and this preface into French, and for undertaking this task with such efficiency and good humour. Our thanks, too, to Jill Sutcliffe (English Nature), for her help with editorial matters, especially in the early stages. English Nature assisted financially when the symposium was being planned, and provided a substantial grant towards the cost of publishing these proceedings. The editorial process has been long and complicated, and we apologise to all concerned for the length of time it has taken us to bring the present volume to fruition. As for our individual roles, Chris Page edited the scene-setting and ‘habitat’ papers, while Simon Leach was responsible for those focussing on individual species. Yves Peytoureau coordinated submission of the French manuscripts, helped several authors with English translations, and checked edited versions of these papers and all the French abstracts. Martin Sanford picked up the baton in late 2004, taking on the task of seeing the edited papers through to final publication: without his considerable input over the last few months, this volume would not now be seeing the light of day. Our final acknowledgement must surely be to Franklyn Perring. Without his vision and dogged persistence, the ‘Atlantic Arc’ symposium would never have happened. In May 2003, he was already in poor health, but his enthusiasm still shone through. He saw this meeting as one step of a journey that would bring those of us from Britain and Ireland into an ever closer working relationship with botanists in France and elsewhere in mainland Europe. In this connection, he would have been delighted to know that colleagues from Bailleul, Brest and the Societe Botanique du Centre-Ouest are now only too eager to reciprocate the Cornish symposium. Franklyn was particularly keen that the proceedings of the meeting should be published, and we are pleased that this has now been done - in hardback and with a great many illustrations, just as he had wished. We hope that Botanical Links in the Atlantic Arc will serve as a fitting and lasting tribute to a man who worked so tirelessly - and over so many decades - to bring plants and people together. It is, therefore, to Franklyn Perring that this volume is most respectfully dedicated. SIMON LEACH, CHRIS PAGE, YVES PEYTOUREAU & MARTIN SANFORD May 2005 3 PREFACE Ce volume comprend une serie de communications et d’affiches presentees au Symposium « Liens Botaniques dans l’Arc atlantique » organise par la Societe Botanique des lies Britanniques (B.S.B. I.) qui s’est tenu en mai 2003 a Camborne en Comouailles. Cette rencontre - l’idee de feu le Dr Franklyn Perring - offrit F occasion aux botanistes de se retrouver pour celebrer les plantes sauvages des regions coheres atlantiques d’ Europe et de se pencher sur toute une serie de sujets concemant leur distribution, autecologie, conservation et gestion. Elle connut un succes retentissant, attirant plus d'une centaine de delegues d’lrlande, d’ Ecosse, du Pays de Galles, d’Angleterre et de France - et meme un de Suede ! Les participants apprirent beaucoup les uns des autres, et toutes les personnes concemees contribuerent a F ambiance amicale et conviviale grace a laquelle de nouveaux liens s’etablirent entre des botanistes travaillant, sou vent tres isoles, tout le long du littoral atlantique de F Europe occidentale. Meme pour ceux d’ entre nous qui se connaissaient deja, ce fut un plaisir de remplacer le teme anonymat des E-mails et de F Internet par quelques joumees de contact en face a face. L’Arc atlantique, tel qu’il est defini ici, forme une region biogeographique coherente de la Grande Bretagne et de l’lrlande au nord et a l’ouest, en passant par le Nord/Pas-de- Calais, la Normandie, la Bretagne, pour contoumer le Golfe de Gascogne a sa limite meridionale au nord- ouest de la Peninsule iberique. Ceci pourrait sembler excessivement restrictif, et certains ont peut- etre bien l’impression que la limite meridionale pourrait s’etendre logiquement vers le sud jusqu’a l’extreme pointe nord-ouest de FAfrique, et aussi inclure les lies Canaries, Madere et les Azores. De plus, la « Region Atlantique Biogeographique » de FUnion Europeenne s’etend vers le nord jusqu’a l’Ecosse et l’ouest du Danemark, et inclut une grande partie de la France du nord et de l’ouest ainsi que les regions littorales de l’Espagne et de F extreme nord du Portugal. Cependant, la rencontre portait surtout sur F etude d’une variete d’especes vegetales et de groupements vegetaux que l’on rencontre sur les cotes atlantiques d’lrlande, de Grande Bretagne et de France (et, a un moindre degre du nord de l’Espagne) ou a proximite. II est vrai que c’est la une zone vaste et diversifiee, mais qui a une unite climatique, avec des extremes meteorologiques - en particular des hivers froids - radoucis par la grande proximite de F Ocean atlantique. Avec des conditions aussi clementes, les saisons de croissance ont tendance a etre longues, favorisees par des chutes de pluie legeres et frequentes, et souvent une forte humidite et une importante couverture nuageuse. La geographie de l’Arc atlantique est dominee par la mer, que les terres penetrent sous forme d’ assemblages disparates et de geologie variee d’tles, d’archipels, de bandes coheres et de peninsules. Le fait que ces zones de terre se trouvent dans de nombreux pays differents - aux peuples parlant diverses langues - permet peut-etre d’expliquer pourquoi il y a eu si peu de tentatives de rapprocher les gens pour les considerer comme une unite phytogeographique coherente. Pourtant, comme nous l’avons decouvert, les similitudes botaniques entre ces regions l’emportent de loin sur leurs differences. En verite, elles ont en commun de tres nombreuses especes et communautes vegetales, et il y en a deux qui peu vent servir d’exemples pour illustrer cette cohesion botanique : Rurnex rupestris (Rumex des rochers), qui a une distribution mondiale correspondant etroitement a la zone couverte par l’Arc atlantique, de la Galice au nord-ouest de l’Espagne jusqu’a la cote occidentale de la France au nord, puis aux lies anglo-normandes, le sud- ouest de l’Angleterre et le Pays de Galles ; et Erica erigena (Bruyere mediterraneenne), qui presente un aire de distribution fortement disjointe s’etendant du centre et du sud du Portugal et de la Galice au sud a l’ouest de l’lrlande au nord. Tout au long du Symposium, nous avons pu partager des eclairages precieux sur la biologie et la conservation d’une foule de telles especes, y compris un certain nombre de plantes rares. Nombreuses furent les communications et les affiches a presenter les resultats d’ etudes entreprises pour repondre a des questions vraiment specifiques en rapport avec la conservation d' especes rares, qui incluaient pas mal d’especes legalement protegees ou pour lesquelles existent des programmes regionaux ou nationaux de conservation (en Grande Bretagne par exemple, les especes inscrites au Programme 8 du « Wildlife & Countryside Act » et/ou comprises dans le « Biodiversity Action Plan » du Royaume Uni). En fait, les orateurs ont souvent mis F accent avec force sur les opportunites et les defis lies a la conservation de ces especes rares et aux habitats dans 4 ATLANTIC ARC lesquels on les trouve. Pour grand nombre d’especes, l’Arc atlantique est clairement un refuge d' importance - non pas simplement un « arc », mais aussi une Arche atlantique. En tant qu’editeurs, nous souhaitons remercier tous ceux qui ant apporte leur aide a L organisation du Symposium et des sorties sur le terrain qui l'accompagnerent, entre autres Ian Bennallick, Jane Croft, Rosaline Murphy, Alistair Cameron, Rachel Holder, Ray Lawman et differents membres de la B.S.B.I. en Comouailles. Nous remercions egalement les auteurs des communications et des presentations d’affiches pour leurs contributions et pour leur patience pendant le processus d’edition. Nous desirons en particulier feliciter nos jeunes invites fran9ais du Conservatoire Botanique National de Bailleul et du Conservatoire Botanique National de Brest qui ont eu la hardiesse de presenter leurs communications en anglais devant un auditoire majoritairement anglophone, illustrant ainsi le besoin flagrant de comprehension mutuelle entre les peuples d’ Europe. Par leur audace amicale, ils ont contribue au dialogue entre botanistes britanniques et irlandais et ceux du « Continent » - excellent exemple d 'Entente Cordiale ! Nous sommes reconnaissants aux collegues qui ont aide a valider les communications ; remerciements speciaux a Chris Preston qui a foumi ses commentaires dans les plus brefs delais sur les « demieres epreuves » de tant de communications. Nos remerciements aussi a Jill Sutcliffe (« English Nature ») pour son aide en matiere d’edition, en particulier dans les etapes preliminaries. « English Nature » offrit son aide financiere en cours de preparation du Symposium et accorda une subvention consequente pour financer la publication de ces reunions. Le processus d’edition a ete long et complexe, et nous presentons nos excuses a toutes les personnes concemees pour le temps qu’il nous a fallu pour realiser le present volume. En ce qui conceme nos roles individuels : Chris Page s’ est occupe des communications sur l’environnement et 1’ habitat, alors que Simon Leach etait responsable de celles mettant 1’ accent sur des especes individuelles. Yves Peytoureau (Secretaire, Societe Botanique du Centre-Ouest) a coordonne 1’ envoi des communications fran^aises, a aide plusieurs auteurs a rediger leurs traductions en anglais, a verifie les versions editees de ces communications et a traduit tous les resumes anglais. Martin Sanford a pris la releve fin 2004, assumant la tache de suivre les communications editees jusqu'a leur publication finale : sans sa contribution considerable ces demiers mois, ce volume ne paraitrait pas maintenant. Notre reconnaissance finale va necessairement a Franklyn Perring. Sans sa vision et sa perseverance obstinee, le Symposium sur « l'Arc atlantique » n’aurait jamais vu le jour. En mai 2003, sa sante declinait deja, mais son enthousiasme l’emportait toujours. II considerait cette rencontre comme une etape dans un voyage qui conduirait ceux d' entre nous originates de Grande Bretagne et d'lrlande a des relations de travail toujours plus resserrees avec les botanistes de France et d'ailleurs en Europe continentale. A ce propos, il aurait ete ravi de savoir que les collegues de Bailleul, Brest et de la Societe Botanique du,Centre-Ouest sont maintenant tres desireux de retoumer 1’ invitation apres le Symposium de Comouailles. Franklyn souhaitait particulierement que le contenu des seances de la rencontre soit publie, et nous nous rejouissons que cela soit desOrmais realise - volume cartonne et avec de nombreuses illustrations - tout comme il l’avait souhaite. Nous esperons que les « Liens Botaniques dans l’Arc Atlantique » serviront de tribut approprie et durable a un homme qui travailla si infatigablement - et pendant tant de decennies - a rapprocher plantes et hommes. C’est par consequent a Franklyn Perring que ce volume est respectueusement dedie. SIMON LEACH, CHRIS PAGE, YVES PEYTOUREAU & MARTIN SANFORD Mai 2005 5 Dedication to Dr F. H. Perring This conference report is dedicated to the memory of Frank Perring (1927-2003). Frank was instrumental in setting up the conference, and its predecessor in France. It turned out to be the last major botanical meeting which he attended, for he was already ill in May 2003 and died five months later. It is fitting that Frank’s last BSBI conference should have been one at which botanists from several western European countries were so well represented. Throughout his career Frank was keenly aware that the British flora could not be understood without reference to neighbouring islands and to the European landmass. Even his Ph.D. thesis (1956) on chalk grassland plant communities interprets the behaviour of species in Britain in relation to their ecology at sites in France and Germany. In his retirement he led wildlife holidays to many of the botanically rich areas of Europe, from the Arctic to the Mediterranean. He had a special interest in the Irish flora, perhaps exceeding that of any other British botanist of his generation with the sole exception of those who have actually taken up residence in Ireland. This arose from his fieldwork in Ireland for the Atlas of the British Flora (1962), and subsequent visits to improve the coverage of under¬ recorded species. It led to papers on the phytogeography of Ireland, published in 1963 and 1967 under the title The Irish problem (the pun is characteristic). He returned to this theme for a major paper in the D. A. Webb memorial volume of Watsonia (1996), a publication commemorating the life of his companion on many Irish field trips. The most enduring legacy of Frank's desire to forge links between European botanists is likely to be Atlas Florae Europaeae. This project represented a fusion of the taxonomic work of Flora Europaea (1964-1980) with the methods developed by Frank Perring and Max Walters (the latter himself a Flora Europaea editor) for the Atlas of the British Flora , with the additional inspiration of Hulten’s Atlas of the distribution of vascular plants in N.W. Europe (1950). As the first volume of Atlas Florae Europaeae stated, “a map presented by Dr Perring at the Tenth International Botanical Congress in Edinburgh, 1964, which showed the 50-km square distribution of Silene acaulis in Europe, can be regarded as the starting-point of the present work”. Frank acted as secretary to the nascent committee, visiting a number of European countries in 1965 and producing a further nine maps at Monks Wood before responsibility for the Secretariat was handed over to Helsinki in December 1965. He remained an active Advisor to the Committee for Mapping the Flora of Europe until his death. If the subject of Frank’s final BSBI conference was appropriate, so too was the spirit in which it was held. The intense buzz of conversation which characterised the proceedings from start to finish was evidence of friendships being developed and renewed, and must have delighted such a convivial botanist. It is with sadness but also with gratitude for his many contributions to British and European botany that we dedicate this volume to the memory of Frank Perring. 6 ATLANTIC ARC Hommage au Dr F. H. Perring Ce compte rendu de la conference est dedie a la memoire de Frank Perring (1927-2003). Frank contribua a organiser cette conference ainsi que celle qui la preceda en France. Elle s’avera etre la demiere reunion botanique d’ importance a laquelle il assista, car il etait deja malade en mai 2003 et mourut cinq mois plus tard. Ce n’est que justice que la demiere conference B.S.B.I. de Frank ait ete un evenement auquel des botanistes de plusieurs pays d’ Europe occidentale ont assiste en nombre. Tout au long de sa carriere, Frank fut convaincu que la flore britannique ne pouvait pas etre comprise sans faire reference aux lies voisines et au continent europeen. Meme sa these de doctorat (1956) sur les communautes vegetales des prairies calcaires interprete le comportement des especes de Grande Bretagne par rapport a leur ecologie dans des stations de France et d’Allemagne. Pendant sa retraite, il conduisit des Sessions naturalistes dans de nombreuses regions d’ Europe riches botaniquement, de FArctique a la Mediterranee. Il s’interessait particulierement a la flore irlandaise, peut-etre plus que tout autre botaniste britannique de sa generation a P exception de ceux qui se sont reellement installes en Irlande. Ceci etait du a son travail de terrain en Irlande pour F Atlas de la Flore de Grande Bretagne (1962) et aux visites qui s’ensuivirent pour approfondir la connaissance d’ especes insuffisamment repertoriees. Cela donna lieu a des communications sur la phytogeographie de F Irlande, publiees en 1963 et 1967 sous le titre La question irlandaise (le jeu de mot est caracteristique). Il aborda de nouveau ce theme pour une communication d’ importance majeure dans le volume de Watsonia a la memoire de D. A. Webb (1996), publication commemorant la vie de son compagnon de nombreux voyages d’ etude en Irlande. Le legs le plus durable du desir de Frank de forger des liens entre les botanistes europeens est sans doute V Atlas Florae Europaeae. Ce projet representait la fusion du travail taxonomique de Flora Europaea (1964-1980) et des methodes mises au point par Frank Perring et Max Walters (ce dernier etant lui-meme Pun des editeurs de Flora Europaea ) pour V Atlas de la Flore britannique , qui s’inspirait en plus de V Atlas de distribution des plantes vasculaires du nord-ouest de V Europe de Hulten (1950). Comme le declarait le premier volume de P Atlas Florae Europaeae , « On peut considerer une carte presentee par le Dr Perring au Dixieme Congres International de Botanique d’Edimbourg en 1964 qui montrait la distribution de Silene acaulis en Europe sur un quadrat de 50 km de cote comme etant a Porigine du present ouvrage ». Frank fut Secretaire du comite juste cree, se rendant dans un certain nombre de pays europeens en 1965 et presentant neuf cartes supplementaires a Monks Wood avant que la responsabilite "du Secretariat ne fut transmise a Helsinki en decembre 1965. Il demeura actif comme Conseiller du Comite de Cartographic de la Flore d’ Europe jusqu’a son deces. Si le theme de Pultime conference B.S.B.I. de Frank etait approprie, Pesprit dans lequel elle se deroula le fut aussi. L’ intense bourdonnement des conversations qui caracterisa les seances du debut a la fin etait la preuve que des amities naissaient et se renouaient : cela ravit a coup sur un botaniste si plein d’ entrain. C’est avec tristesse mais egalement avec gratitude pour ses nombreuses contributions a la botanique britannique et europeenne que nous dedions cet ouvrage a la memoire de Frank Perring. 7 RESUMES FRAN^AIS 1. Le contexte geologique du paysage de l’Arc atlantique / Les territoires mari times sur la frange occidentale de F Europe, qui s’etendent de l’Ecosse vers le sud et passent par Flrlande, le Pays de Galles, le sud-ouest de FAngleterre, la Bretagne et continuent jusqu’au nord-ouest de la Peninsule iberique, consistent en une serie de massifs anciens formes pendant la Periode hercyniennne (durant le Cycle varisque) et les orogeneses anterieures. A F exception de certaines roches plus anciennes en Ecosse, Irlande septentrionale, le nord du Pays de Galles et quelques parties de la Bretagne, les roches furent a Forigine deposees dans les latitudes equatoriales pendant les Periodes du devonien et du carbonifere sous forme de boues, de sables et de roches volcaniques dans une depression entre deux plaques tectoniques convergentes. La collision entre ces plaques comprima, plia et crea des failles dans ces materiaux de la cuvette pour former les roches dures des massifs. Les granites furent introduits de force et occupent de grandes surfaces des massifs. A la suite d’une periode de conditions tres chaudes et arides au permien et au triasique, les conditions d’humidite revinrent pendant le reste du mesozoique, et durerent jusqu’au debut du tertiaire, periode pendant laquelle ,un intense nivellement entama le processus de formation du paysage actuel. Une vegetation sub-tropicale et une forte degradation lateritique caracterisent cette periode. Le refroidissement du climat a la fin du tertiaire menerent aux eres glaciaires du quatemaire. La recherche palynologique sur F existence importante de sediments appartenant a la fin du tertiaire en Comouailles nous apprend beaucoup sur le climat et la vegetation de la periode conduisant au commencement des conditions de froid. Pendant les eres glaciaires, la Comouailles et les regions au sud ne furent pas envahies par les glaciers mais subirent des temperatures extremement basses ; les caracteristiques periglaciaires nous apprennent beaucoup sur la rigueur du climat. Le Pays de Galles, Flrlande et les regions au nord furent recouvertes par les glaciers, ce qui fit que la majeure partie du manteau de desagregation pre- glaciaire fut enlevee et que des formations terrestres glaciaires se formerent. Vers la fin de la demiere glaciation (devensien), le climat devint non seulement froid, mais aussi extremement aride, ce qui fit voler de la poussiere formee par les sediments glaciaires entraines par les eaux sur le fond de la Mer d’ Irlande ; le niveau de la mer a Fepoque se trouvait 100 m plus bas qu’a present. Cette poussiere forme une couche de loess recouvrant certaines parties de la Comouailles et est probablement une composante importante des sols de la Comouailles et des regions voisines. L’ amelioration du climat au cours des 10 000 demieres annees depuis la fin de la demiere manifestation de glaciation a donne naissance a la couverture actuelle en sol et en vegetation. Cette communication porte surtout F accent sur la partie centrale de l’Arc atlantique, notamment la Comouailles, bien qu’il soit fait mention de regions au nord et au sud ; elle decrit aussi brievement quelques unes des caracteristiques botaniques de Fhistoire geologique. Mots-cles : paleobotanique, geologie, granites, exposition aux intemperies, processus periglaciaires, armoricain. 2. Phytogeographie passee et presente de Flrlande Par comparaison avec la Grande Bretagne toute proche et F Europe continentale, la flore d’ Irlande est pauvre. Pourtant, certains elements biogeographiques sont particulierement bien representes en Irlande ; ils incluent des especes a la distribution a predominance atlantique ou oceanique europeenne. Nous suggerons qu’il y a de bonnes raisons ecologiques sous-tendant ces distributions, et de plus que les changements de distribution des especes irlandaises depuis le pliocene refletent des tendances similaires : la perte d’ especes ligneuses continentales, mais la presence repetee de plantes herbacees et d’arbrisseaux atlantiques, dont beaucoup sont des especes des zones humides au sens le plus large. Determiner Fepoque de F isolation entre les populations irlandaises et continentales, et, dans le cas des especes amphi-atlantiques, des populations d’Amerique du Nord, sera cmcial pour mieux comprendre la phytogeographie irlandaise. Nous prevoyons que les etudes phylogenetiques moleculaires et phylogeographiques contribueront grandement a cette connaissance. Mots-cles : biogeographie, taxons de plantes vasculaires, elements atlantiques, histoire quatemaire. 8 ATLANTIC ARC 3. Les elements mediterraneo-atlantiques et atlantiques de la flore de Comouailles En Grande Bretagne, les especes a distribution mediterraneo-atlantique sont concentrees dans le sud et celles a distribution atlantique (oceanique) sont concentrees dans l’ouest. Les especes appartenant aux deux sont done bien representees en Comouailles, le comte britannique le plus au sud-ouest, ou 70% des plantes vasculaires mediterraneo-atlantiques britanniques et 70% des taxons oceaniques sont presentes comme especes indigenes. La majorite des especes mediterraneo- atlantiques du comte sont des plantes aux habitats cotiers, mais le groupe comprend aussi quelques especes d’autres sols sujets a la secheresse ou perturbes. Cette variete relativement faible d’ habitats contraste avec la grande variete d' habitats oil Ton trouve les especes oceaniques dans le comte. La taille des populations de plusieurs especes mediterraneo-atlantiques varie fortement. Les conditions les plus favorables sont produites par des etes de secheresse suivis d’ hivers doux et de printemps ensoleilles mais pas trop secs ; de telles annees, la taille de la population d’ especes comme Trifolium incarnatum peut etre 20 fois celle de saisons moins favorables. Certaines especes peuvent egalement reagir positivement aux incendies et a des perturbations humaines limitees. En plus des especes indigenes, une serie d’ especes originaires de la region mediterraneenne s’est naturalisee en Comouailles, et la plante grasse Carpobrotus edulis d’Afrique du Sud a acquis une distribution mediterraneo-atlantique en Europe. Mots-cles : Comouailles, flore vasculaire, elements mediterraneo-atlantiques, caracteristiques ecologiques. 4. La determination de la zone des Fougeres dans 1’Arc atlantique par un environnement de facteurs complexes et interactifs Nous avons montre des cartes d’un choix de zones a Fougeres de Grande Bretagne et d’lrlande s’etendant assez loin dans l’Arc atlantique ; seule une selection en est reproduite ici. L’ analyse des conditions propices au succes de ces Fougeres de climat tres oceanique montre a la fois une concordance proche et des variations circonstanciees des conditions a travers l’Arc atlantique. Un compte rendu plus complet de ce montage de donnees d’ ensemble dans l’Arc atlantique et son integration en particulier avec les donnees des Sciences de la Terre constitue l’objectif eventuellement plus vaste de cette etude. Mots-cles : Pteridophytes, determination de la repartition, donnees climatiques. 5. Introduction a la flore des lies Sorlingues (= Scilly) /V Les lies Sorlingues ont ete formees par la submersion d’une masse terrestre plus grande qui fut habitee pour la premiere fois il y a 4000 ans. Situees au large des terres continentales au bord de L Atlantique, V influence de la Derive nord-atlantique (un bras du Gulf Stream) qui les rechauffe se reflete dans la flore et la faune des lies. Nombre des especes vegetales de Scilly ont ete introduites, soit accidentellement, soit intentionnellement, au fil des siecles et sont maintenant des membres a part entiere de la flore. Sinon, la flore indigene est limitee par la faible surface de terre, les habitats, les conditions climatiques et l’isolement. II y a une forte influence « lusitanienne » dans la flore, avec de nombreuses plantes de distribution d’ Europe meridionale y compris certaines d’origine mediterraneenne. La culture des Narcisses odorants pendant l’hiver pour le marche de la fleur coupee etait jusqu’a une date recente l’industrie principale des fermiers. Ceci a donne lieu a une flore associee de plantes sauvages d’annuelles hivemales et d’autres, printanieres precoces. Le climat hi vernal tres doux permet a beaucoup de plantes de continuer a pousser pendant tout l’hiver. La flore s’est trouvee augmentee par un grand nombre d’adventices echappees, la plupart en provenance des Jardins de Tresco, et ces demieres se sont sou vent installees dans la nature en raison du climat doux. /V Mots-cles : Comouailles, lies Sorlingues, floristique, habitats. /*v 6. Les communautes maritimes comme habitats pour les Fougeres Ophioglossum dans les lies Sorlingues (= Scilly) Les communautes de plantes herbacees maritimes et des landes, influencees par la proximite de 1’ Atlantique et du Gulf Stream foumissent des habitats aux Fougeres du genre Ophioglossum dans les lies Sorlingues. On trouve dans ces lies les trois taxons indigenes d’ Ophioglossum. Mots-cles : Ophioglosse, Ophioglossum lusitanicum , O. azoricum, O. vulgatum. RESUMES 9 7. Les prairies de l’Arc atlantique : Eexemple des Rhos Pastures du sud-ouest du Pays de Galles et leur conservation Les communautes vegetales et les especes typiquement associees aux Rhos Pastures - prairies humides acidophiles, souvent des landes, qui se rencontrent encore frequemment au sud-ouest du Pays de Galles sont decrites : leur gestion, leur signification sociale et leur conservation sont resumees. Carum verticillatum (Carum verticille) est hautement caracteristique des Rhos Pastures et fait partie d'une suite d' especes « atlantiques » qui contribuent au caractere particulier de ces prairies. II est fait reference a Ehistoire du recensement de Carum dans le sud-ouest du Pays de Galles ; son caractere distinctif et sa survie souvent abondante dans le bassin houiller du Carmarthenshire sont expliques. II est aussi fait reference aux pertes majeures qui sont advenues ces demieres annees, dues au charbonnage a ciel ouvert, au developpement commercial et domestique et aux changements agricoles. Etant donne Pinefficacite des mesures de protection actuelles, un appel est fait a toute personne concemee en vue d’un effort concerte pour proteger et maintenir ce qui subsiste. Mots-cles : Pays de Galles, prairies de l'Arc atlantique, jongaies, classification de la vegetation, defense de Penvironnement. 8. Les haies celtiques comme refuges pour la diversite des Fougeres dans des paysages a predominance agricole Ce compte rendu s'interesse brievement aux principaux aspects des haies celtiques de terre et de pierre qui caracterisent une grande partie de l'Arc atlantique. S’appuyant sur le contexte de leur structure, histoire, distribution et signification botanique dans des paysages a predominance fortement agricole, il met specialement P accent sur la composante en Fougeres souvent abondantes de ces habitats faits par Phomme, et sur la valeur de nos jours de ces habitats comme refuges particulierement importants pour ce groupe de plantes. II est foumi des raisons pour suggerer, depuis les temps neolithiques jusqu'a present, a cause des processus de changement des paysages de forets originelles non entretenues depuis des environnements pre-agricoles jusqu'aux conversions de P agriculture modeme, pourquoi les haies celtiques en sont venues a foumir des habitats pteridologiques particulierement vitaux. Deux elements specialement importants de ce probleme sont que ces haies ont fonctionne pour foumir des refuges iliens vitaux, chaque individu ayant sa localite specifique, pour la survie d'une tres ancienne diversite de base des Fougeres, alors que - de la meme importance et egalement semblables a des lies - a partir de cette base, elles ont aussi continue a foumir d'importantes stations pour que les processus naturels de micro-evolution des Fougeres continuent a se reproduire de fagon constante. La valeur permanente des haies pour la persistance de la diversite resultante des Fougeres dans des environnements modemes est presentee par une etude detaillee des haies dans trois communes de l’ouest de la Comouailles. L' accent est ainsi mis sur P importance de la conservation des haies celtiques dans cette perspective, et quelques unes des structures politiques deja en place en Comouailles sont citees a titre d’exemples de reussite de planification botanique, archeologique, agricole et de cooperation gouvemementale a ces fins environnementales collectives. Mots-cles : Comouailles, diversite des Pteridophytes, paysages modifies par P agriculture, refuges, protection de Penvironnement. 9. La tache du Conservatoire Botanique National de Brest dans la connaissance et la conservation de la flore armoricaine illustree par Trichomanes speciosum Willd. Le Conservatoire Botanique National de Brest a mis en place un vaste programme d'inventaire permanent de la flore du Massif armoricain, dans le but d'ameliorer P identification de la repartition des taxons, mais aussi afin de permettre P elaboration de strategies de preservation de la flore menacee. L’ analyse des donnees recueillies au travers des differents inventaires, et en particulier du programme “Atlas de la flore armoricaine’’ a deja permis la constitution de diverses rouges regionales d’ especes regionales « rouges » menacees, et la definition de listes de taxons prioritaires en terme de conservation. Trichomanes speciosum Willd., fougere figurant a P annexe 2 de la directive Habitats, fait partie de ces taxons dont la conservation est jugee prioritaire par le Conservatoire Botanique National de Brest. Un plan regional d' action a done ete elabore en 2001 pour cette espece, qui fait desormais Pobjet d’ analyses ecologiques, biologiques, physiologiques et genetiques approfondies, ainsi que de mesures de protection particulieres. Mots-cles : Trichomanes, plan d' action, Massif armoricain, Bretagne. 10 ATLANTIC ARC 10. Les Orchidees de rile d’Oleron L’lle d’Oleron, situee sur la cote atlantique fran^aise, beneficie grace au Gulf Stream d’un climat doux rappelant celui de la Mediterranee. Ces conditions exceptionnelles alliees a des biotopes varies permettent la croissance de plantes meridionales parmi lesquelles bon nombre d’ orchidees. Ces demieres trouvent ici leur limite nord et s’ajoutent aux especes de distribution atlantique. Mots-cles : mediterraneo-atlantique, deprise agricole, dunes forestieres. 1 1 . Introduction a la flore de la Peninsule du Lizard, Comouailles La Peninsule du Lizard comprend la partie la plus meridionale du sud-ouest de l’Angleterre et est generalement reconnue comme ayant un caractere floristique inhabituel. Une florule distincte est identifiee ; elle inclut des especes qui dans un contexte national ou regional sont rares. La presence d’un climat oceanique doux avec quelques caracteristiques mediterraneennes et la geologie inhabituelle de la partie meridionale de la Peninsule, y compris des roches ultrabasiques (serpentine et gabbro) sont souvent considerees comme autant de raisons pour expliquer le cote distinctif de la flore de cette zone. Des preuves sont avancees pour soutenir l’hypothese qu’a la fois la geologie et le climat expliquent en partie la geographic floristique de cette region, alors que la survie d’une grande proportion de la zone d’habitat diversifie a demi-naturel presente au debut du 20eme siecle est consideree comme etant un troisieme facteur important. Mots-cles : Comouailles, le cap Lizard, flore vasculaire, ecologie des plantes. 12. Le Projet des Sentiers de la Pointe du Lizard L’ importance des sentiers inondes de fa^on saisonniere dans la zone du Lizard (ouest- Comouailles) pour la science botanique et la conservation de la nature est decrite brievement. Ils sont d’une importance particuliere quant a la presence de six especes de plantes vasculaires ou de Characees rares ou peu abondantes au niveau national. Une etude de trois ans pour evaluer l’etat actuel du systeme de sentiers et la situation des especes rares est en cours. Les decouvertes initiales confortent 1’ opinion que les sentiers sont particulierement menaces par la fermeture ou le manque d’entretien, avec seulement 18% qui sont dans un etat convenable pour la croissance des especes de valeur conservatoire. Mots-cles : Comouailles, le cap Lizard, sentiers, mares temporaires. 13. Vegetation semi-naturelle sur des terrains du National Trust en Comouailles, en insistant sur celle de la Peninsules du Lizard Le National Trust possede 9000 ha de terrain en Comouailles, dont 50% sont occupes par des communautes vegetales semi-naturelles. Cet article presente un bref panorama de quelques unes des communautes vegetales concemees et des problemes de leur gestion, avec un accent particular sur les habitats de la Peninsule du Lizard. Mots-cles : Comouailles, le cap Lizard, vegetation semi-naturelle, gestion des habitats. 14. Variation genetique chez les especes vegetales menacees en Irlande : perspective europeenne La variability genetique des populations de plantes peut avoir des implications a court terme et a long terme sur la persistance de ces populations, et 1’ etude de la genetique des populations est par consequent devenue partie integrante de la recherche dans la conservation des especes vegetales menacees. La technique d’empreinte genetique A.F.L.P. a ete utilisee pour evaluer le statut de conservation d’un certain nombre d’ especes vegetales menacees en Irlande dans un contexte europeen ; elles incluent Campanula trachelium L. (Campanule gantelee), espece hautement fragmentee montrant des signes possibles de recession par consanguinite et Colchicum autumnale L., (Colchique d’automne), espece irlandaise menacee a la distribution fortement localisee et au statut indigene douteux. Nous presentons des resultats pour C. autumnale qui indiquent que les populations irlandaises contiennent de grands niveaux de diversite genetique et qu’il est done improbable qu’ elles aient ete introduites et qu’ elles soient sujettes aux effets nuisibles de la recession par consanguinite. Les comparaisons de populations montrent que la perspective europeenne est fondamentale pour 1’ evaluation du statut de conservation des especes vegetales irlandaises menacees. Mots-cles : Genetique de conservation, Colchicum autumnale , Campanula trachelium, recession par croisement, biogeographie. RESUMES 11 15. Biogeographie des Bruyeres « lusitaniennes » en Irlande L’ element « portugais » (ou hibemo-cantabrien) de la flore de Grande Bretagne et d’ Irlande consiste en une suite d’especes rencontrees principalement dans l’ouest de l’lrlande et dans l’ouest de la Peninsule iberique, comprenant six especes d’ Ericaceae : Arbutus unedo, Daboecia cantabrica, Erica ciliaris, E. erigena, E. mackaiana et E. vagans. On s’ est pose la question de savoir si ces taxons se sont recolonises en provenance de stations relictuelles europeennes a la suite d’ episodes glaciaires, ou bien si elles ont survecu a des cycles glaciaires complets dans des stations relictuelles plus proches. Ce groupe d' especes est d’une importance considerable pour la conservation en Europe, et son etude est en rapport avec les scenarios de changements climatiques passes et presents. La technique d'empreinte genetique de polymorphismes de longueur de fragments amplifiee (A.F.L.P.) a ete appliquee a trois taxons, Erica mackaiana, Erica erigena et Daboecia cantabrica , pour evaluer la differentiation demographique des populations irlandaises et continentales, dans le but de comprendre si cela peut aider a expliquer leur origine. Alors que l’A.F.L.P. ne parvint pas a donner dedications quant a l’origine de ces taxons dans la flore irlandaise, elle suggere bel et bien qu'au moins Daboecia cantabrica et Erica mackaiana sont peut-etre arrivees grace a une dispersion sur une longue distance pour faire de la sorte partie integrante de la flore d' Irlande. Ainsi, alors que plusieurs modeles et coincidences peuvent etre reperes dans les ecologies et la distribution a aires disjointes des taxons « portugais », ils devraient etre etudies a part, plutot que de supposer une origine commune, lorsqu’on considere comment ces taxons se trouvent faire partie de la flore des lies Britanniques. Mots-cles : A.F.L.P., conservation, Daboecia cantabrica , Ericaceae, Erica erigena , Erica mackaiana , glaciation. 16. Le profilage ecologique comme moyen d’ameliorer la conservation des plantes des vasieres De nombreuses plantes des vasieres ont des exigences d’ habitat tres specialises. L’ habitat adequat est souvent fragmente, disperse et ephemere. Ces conditions signifient que la plupart des populations font partie de metapopulations plus vastes et que, bien que certaines populations soient peut-etre perdues, beaucoup de metapopulations semblent etre tres saines, par opposition a rimpression donnee par revaluation portant sur une station. Dans cette communication, je suggere que P application du profilage ecologique a la conservation des plantes des vasieres permet une evaluation conservatoire et un suivi efficaces et portant sur une espece. Ce precede est relativement peu couteux et donne souvent naissance a des moyens de traiter la conservation des plantes des vasieres qui sont plus efficaces que la plupart des approches courantes. Les benefices a long terme du profilage ecologique depassent de loin le prix a court terme de la preparation du profilage. Mots-cles : profilage ecologique, plantes des vasieres, metapopulations. 17. Les Bryophytes de Comouailles dans l’Arc atlantique : biologie cellulaire, culture, conservation et changement climatique Ce compte rendu met 1’ accent sur les habitats-cles naturels et faits par l'homme des Bryophytes et sur les taxons que Ton y trouve en Comouailles et dans les lies Scilly et explore les priorites pour leur conservation. Les changements recents de l’abondance, soit des augmentations, soit des diminutions, de certaines especes de Comouailles sont sans doute davantage dus aux changements des habitats, par exemple a des pratiques de gestion modifiees et a 1’ abandon des landes, avec de plus des changements climatiques plus subtiles, plutot qu’a la pollution au SO2 et a 1’ augmentation des depots de N2. Deux Hepatiques recemment introduites aux lies Scilly envahissent maintenant la Grande Bretagne continentale, alors qu’une autre Mousse adventice anterieure, Campylopus introflexus, semble supplanter les especes indigenes sur les landes de Comouailles. Les etudes au microscope electronique ont revele differentes symbioses chez les Hepatiques de Comouailles impliquant des champignons (Ascomycetes, Basidiomycetes, Glomeromycetes) ; elles ont aussi montre que des oleocorps subsistent tout au long de periodes de dessiccation dans des conditions naturelles. La survie de Weissia controversa var. densifolia et de Cephaloziella spp. sur des sols contamines par le plomb et le cuivre peut etre en rapport avec leurs rhizoides aux parois extremement epaisses. La culture in vitro de Mousses de Comouailles, autant qu'elle clarifie les problemes taxonomiques, est employee pour conserver des taxons rares, specialement ceux se trouvant sur des sols instables. La manipulation du milieu induit la production abondante de gemmae ou de tubercules qui peuvent etre utilises pour foumir en quantite les cultures pour des 12 ATLANTIC ARC essais de reintroduction et pour des analyses de diversite genetique a Finterieur de populations et entre elles. Mots-cles : cultures axeniques, conservation, changement environnemental, metaux lourds, introductions, protonemata, symbioses. 18 Erica erigena R. Ross (Bruyere mediterraneenne) dans la region du Medoc au sud-ouest de la France En France, Erica erigena R. Ross n’est presente que dans une zone tres limitee du Medoc (Sud- Ouest de ce pays). Le substratum sur lequel se developpe cette bruyere est essentiellement constitue par des formations detritiques quatemaires humides ou souvent mouillees pendant toute l’annee. Sur le plan phytosociologique, elle caracterise une association faiblement acide de Falliance de YUlici minoris - Ericion tetralicis J.- M. Gehu : YErico scopariae-erigenae F. Bioret et C. Lahondere. Mots-cles : Erica erigena R. Ross, Medoc, aquifere, phytosociologie, landes humides. 19. Euphrasia vigursii Davey (Euphraise de Vigurs) dans le Devon Euphrasia vigursii Davey (Scrophulariaceae) est une Euphraise rare endemique de la Comouailles et du Devon. C’est un des taxons d’Euphraises les plus facilement reconnaissables, plante tout a fait saisissante aux fleurs d’un violet fonce et aux poils glanduleux. Elle se rencontre typiquement sur des pelouses ouvertes assez courtes dans des landes a Ulex gallii-Agrostis curtisii broutees ou pietinees ; on la trouve aussi bien a Finterieur des terres qu’en haut des falaises littorales en Comouailles. Nous decrivons la distribution actuelle de cette Euphraise dans ses quelques stations du Devon, du cote occidental du Dartmoor, et insistons sur F importance de la gestion de site - surtout paturage et pietinement - pour conserver les conditions ouvertes qu’elle requiert. Mots-cles : Lande a Ulex gallii - Agrostis curtisii, Euphraise, endemisme, distribution, conservation, comptages de population, paturage. 20. Faune et flore en Comouailles Cette communication etudie une partie du travail entrepris a F Environmental Records Centre for Cornwall and the Isles of Scilly (E.R.C.C.I.S.) pour ce qui est de F integration des donnees sur la faune et la flore pour leur emploi par les naturalistes faisant des pointages. Son but est de demontrer comment les donnees peuvent etre integrees en employant la technologie informatique pour presenter des representations spatiales claires et comprehensibles de la faune et de la flore dans des zones de Comouailles et des lies Scilly. Mots-cles : inventaires biologiques, Centre local des inventaipes, information environnementale, mise sur ordinateur. 21. Cystopteris diaphana (Bory) Blasdell (C. viridula (Desv.) Desv.) plante nouvelle pour la Grande Bretagne La decouverte de Cystopteris diaphana en Comouailles, nouvelle pour la Grande Bretagne, est decrite. Nous donnons un resume de son habitat, avec des notes sur sa determination sur le terrain, sa taxonomie et sa plus grande distribution en Europe. La question de savoir si cette espece est indigene ou introduite en Grande Bretagne est consideree. Mots-cles : Cystopteris, Cystopteris diaphane, distribution, statut indigene ou etranger. 22. Developpements de bases de donnees et enregistrement biologique en Comouailles Ces demieres annees, le systeme de Recensement Biologique en Comouailles a mis nettement F accent sur Faide a la communaute de recensement volontaire. En particulier, les ameliorations technologiques ( aussi bien le materiel que les ameliorations permanentes de la base de donnees E.R.I.C.A.) ont fait que les « recenseurs » individuels peuvent tous avoir chez eux la totalite les donnees disponibles pour la Comouailles sur leur P.C. portable, utiliser ces donnees pour planifier d’autres recensements, ajouter a ces donnees, et par la contribuer a F effort collectif da la communaute de recensement biologique de Comouailles, et finalement de plus loin. Ce niveau de capacite individuelle a stimule Factivite des recensements, et ce processus a ete grandement aide par les efforts combines de la Cornwall and Isles of Scilly Federation for Biological Recorders (C.I.S.F.B.R.) et de F Environmental Records Centre for Cornwall and the Isles of Scilly (E.R.C.C.I.S.) et autres groupes locaux qui foumissent ateliers, reunions sur le terrain et seminaries et publient toute une gamme de materiel utile. Mots-cles : Centre local des inventaires, information environnementale, mise sur ordinateur. RESUMES 13 \ 23. Evaluation des menaces des populations d’Oseille des rochers ( Rumex rupestris Le Gall) au Royaume-Uni et en France L’Oseille des rochers ( Rumex rupestris Le Gall), a ete etudiee dans un programme de recherche Alliance « A conservation strategy for threatened botanical resources of the English Channel- Atlantic coastal environment ». Apres avoir precise la repartition geographique et l’ecologie de l’espece, 1’ evaluation du niveau de menaces des populations d'Oseille des rochers echantillonnees en 1998 et en 1999 au Royaume-Uni et en France est presentee. Les possibility de recherches complementaires sont discutees. Mots-cles : conservation, distribution. 24. Rumex rupestris Le Gall (Rumex des rochers) dans le sud-ouest de l’Angleterre : analyse d' etudes recentes et evaluation du statut actuel Des etudes recentes sur Rumex rupestris Le Gall dans le sud-ouest de rAngleterre sont analysees et son statut actuel resume. Le nombre de populations connues a augmente de fagon spectaculaire depuis le debut des annees 1990. Cet accroissement reflete un etat de connaissances en progression a la suite d’ etudes ciblees effectuees du milieu a la fin des annees 1990, y compris une etude d' importance faite dans le sud-Devon et le long de la cote sud de la Comouailles. Les populations existantes sont a present suivies de pres grace au projet « Flora Guardian » mis sur pied par Plantlife. Depuis 1999 R. rupestris est connu avec certitude de 36 stations dans le Devon et en Comouailles (les lies Scilly comprises). Mots-cles : Rumex des rochers, Rumex rupestris , etude, surveillance, distribution, conservation. 25. Compte rendu du travail recent sur Lagregat Limonium binervosum (Limonium a deux nervures) dans les lies Britanniques -A. Limonium binervosum agg. dans les lies Britanniques comprend un ensemble de 9 especes, 15 sous-especes decrites et 16 varietes decrites. On pense que huit de ces especes ( et toutes les sous- especes et varietes sauf une) sont des endemiques britanniques-et-irlandaises. Dans le Royaume- Uni, les taxons endemiques sont inclus comme especes prioritaires dans le Biodiversity Action Plan national ; ils sont egalement inscrits dans le Livre Rouge. Comme tels, ils commencent a susciter un interet considerable depuis quelques annees : des etudes sont actuellement effectuees pour definir la distribution nationale, regionale et locale des divers taxons ; et la taxonomie du groupe est en cours d’ etude selon une approche moleculaire de V ensemble de Lespece en employant le polymorphisme de longueur de fragments amplifiee (A.F.L.P.). Une fois resolus les divers problemes taxonomiques, on espere que la B.S.B.I. publiera un manuel des Limonium britanniques et irlandais. Mots-cles : Limonium , distribution, taxonomy, A.F.L.P., Biodiversity Action Plan. 26. Enquetes recentes sur des taxons endemiques de Limonium ( Limonium a deux nervures) dans le sud-Devon (v.c. 3) Des etudes recentes sur Limonium binervosum subsp. mutatum et L. britannicum subsp. coombense dans le sud-Devon (v.c. 3) ont ajoute de fagon marquante a notre connaissance de ces taxons endemiques. Tous deux semblent etre plus frequents le long du littoral du sud-Hampshire qu'on ne le pensait auparavant. Un troisieme taxon, n’ayant pas encore de nom, est signale sur la rive est de l’estuaire de la Salcombe. Mots-cles : apomixie, conservation, distribution, recensements de populations. 27. Vegetation et habitats d' Asparagus pro stratus Dumort. (Liliaceae), (Asperge prostree), endemique de V Europe occidentale Les types de vegetation d 'Asparagus prostratus Dumort. (Asperge prostree) sont decrits a partir de 94 quadrats rassembles dans toute sa zone en Europe. 226 especes ont ete enregistrees associees a A. prostratus, avec une moyenne de 13 especes par quadrat. Elle est presente en general a proximite de la mer (altitude moyenne 18 m), sur les pentes de toutes sortes (19° en moyenne), dans une vegetation rase (hauteur moyenne 23 cm) et ouverte (couverture moyenne 79%). Les plantes associees les plus frequentes etaient Festuca rubra et Daucus carota. Huit groupements vegetaux principaux ont ete denombres par analyse T.W.I.N.S.P.A.N. II y avait huit communautes 14 ATLANTIC ARC des falaises littorales : des plaques de Carpobrotus edulis, des prairies maritimes du Crithmo maritimi-Armerion maritimae, des zones d’embruns a vegetation du Crithmo maritimi-Limonion biner\’osi, et une vegetation de falaises rocheuses sur sols peu profonds du Thero-Airion ou peut- etre du Saginion maritimo. Elle est presente dans trois types de vegetation des dunes : le classique et tres repandu Ammophilion arenariae, la combinaison dynamique Ammophlion et Euphorbio portlandicae-Helichrysion staechadis, et des dunes perturbees quelque peu nitrophiles appartenant probablement au Corynephorion canescentis. Elle se rencontre aussi dans les broussailles dunaires de L Hippophaeo-Ligustretum. En general, elle se trouve dans une variete etonnamment large de types de vegetation cotiere et sa rarete n’est probablement pas la consequence du manque d’ habitats ou de restriction a un type de vegetation specifique. Mots-cles : Asparagus officinalis subsp. prostratus , communautes de plantes britanniques, Belgique, Angleterre, Irlande, France, Pays-Bas, Espagne, Pays de Galles. 28. Strategic de conservation de Genista pilosa L. (Genet poilu) en Picardie, France Le Genet poilu ( Genista pilosa) est une espece d’ Europe mediane. La limite nord-ouest de sa repartition passe dans le sud de V Angleterre et en Picardie, une des trois regions franchises ou la plante est protegee. Le Centre Regional de Phytosociologie / Conservatoire Botanique National de Bailleul et le Conservatoire des Sites Naturels de Picardie travaillent en partenariat pour developper une strategic de sauvegarde de Genista pilosa en Picardie. Mots-cles : distribution (Europe, France, Grande Bretagne), conservation. 29. Romulea columnae Sebast. & Mauri (Romulee de Columna) retrouvee en Comouailles apres 121 ans et Juncus capitatus Weigel (Jonc capite) nouveau pour l’est-Comouailles (v.c. 2) Cette communication donne les details de deux decouvertes botaniques majeures dans l’est- Comouailles en 2002. D’abord, la redecouverte de Romulea columnae a Polruan, dans une prairie ouverte dessechee, en haut d’une falaise, tres differente de la prairie dunaire ou elle se trouve a Dawlish Warren, sud-Devon (sa seule autre station britannique), mais semblable a ses habitats de haut de falaise dans les lies anglo-normandes. Ensuite, la decouverte de Juncus capitatus a The Blouth, premiere mention de cette espece en est-Comouailles ; espece d’ habitude limitee dans les lies Britanniques aux lies anglo-normandes, la Peninsule du Lizard (ouest-Comouailles) et Anglesey. Mots-cles : enquete de vegetation, National Vegetation Classification, prairie de haut de falaise littorale, decouverte botanique, distribution. 30. Hypericum linariifolium x humifusum : hybride adapte a la cote atlantique Les hybrides entre Hypericum linariifolium Vahl. (Millepertuis a feuilles lineaires), rare a l’echelle nationale et intemationale, et H. humifusum L. ( Millepertuis couche), plus repandu, sont decrits sur une lande cotiere de l’ouest du Pays de Galles. La combinaison de leur tolerance a la secheresse et de leur habituel port couche a peut-etre un avantage selectif dans cette situation manifestement ventee. Mots-cles : Hypericum , Millepertuis, hybridation, adaptation evolutionniste, landes coheres, Pembrokeshire, Llyn peninsula. 31. Lotus angustissimus L. (Lotier tres etroit) dans le sud-ouest de L Angleterre Au debut des annees 1990, pendant que l’on rassemblait de nouvelles donnees pour la troisieme edition du Livre Rouge de Grande Bretagne, des recherches furent entreprises pour etablir la distribution et le statut de conservation de Lotus angustissimus L. dans le sud-ouest de L Angleterre. Cette communication donne un bref resume des resultats de ces etudes et de travaux plus recents sur cette espece. Mots-cles : distribution, conservation, gestion, prairie de haut de falaise, broussailles, Livre Rouge. 32. Puccinellia foucaudi (Coste) Hackel sur la cote atlantique frangaise : situation taxonomique, morphologie et distribution Puccinellia foucaudi (Coste) Hackel (Poaceae) est un taxon signale sur les cotes centre-atlantiques frangaises de la Vilaine a la Gironde, le plus souvent dans des marais saumatres estuariens. La plante se distingue des especes voisines par son developpement en touffes atteignant et pouvant depasser un metre de hauteur et par le fait qu’elle peut produire des stolons apres Lanthese. Mots-cles : Puccinellia , [taxonomie], estuaire, stolon, rhizome. RESUMES 15 33. Situation de Teucrium scordium L. (Germandree d’eau) dans le nord-ouest de la France Teucrium scordium (Germandree d’eau) est une espece polymorphe : deux sous-especes sont connues en France, subsp. scordium et subsp. scordioides. Dans cette communication, nous discutons brievement les resultats d’une etude biometrique de populations du nord-ouest de la France et de Bretagne ; et nous suggerons qu’une population dans les depressions dunaires de la Reserve Naturelle de Merlimont (Pas-de-Calais) se refere peut-etre a la subsp. scordioides mediterraneo-atlantique, alors que d’autres populations de cette station semblent etre intermediaries entre subsp. scordioides et subsp. scordium. Nous donnons un resume des biologie, autecologie et distribution de T. scordium dans la region du Nord/Pas-de-Calais. Une etude de toutes les populations connues de la region nous a permis de mettre en evidence les priorites pour la conservation a long terme de cette espece menacee. Mots-cles : autecologie, distribution, conservation. 34. Gestion de la restauration et de la conservation de 1' habitat de Mibora minima (L.) Desv. (Mibora nain) dans le nord de la France Mibora minima (L.) Desv. (Poaceae) est une espece de vaste distribution en Europe, mais elle est tres rare dans le nord de la France. Dans cette communication, nous decrivons la distribution et l’ecologie de M. minima sur les dunes de la region du Nord/Pas-de-Calais en France et sur les systemes voisins de dunes en Belgique et aux Pays-Bas. Cette espece se trouve sur les pelouses au sol sablonneux pauvre, dans une vegetation rase ayant une forte couverture de Bryophytes et une couverture basse de plantes vasculaires associees. Le systeme de gestion des dunes dans le Nord/ Pas-de-Calais differe dans chaque station, mais des mesures sont desormais en place pour restaurer et gerer les pelouses dunaires de fa^on a preserver des zones d’ habitat convenable pour M. minima. Les grandes potentialites de ces stations et la capacite de dispersion de Mibora suggerent que nous n’avons pas a creer une gestion specifique pour cette espece. D’un autre cote, la suppression de broussailles doit etre encouragee par la coupe des buissons et le broutage des pelouses dunaires par les ovins, le betail ou meme les chevaux pour maintenir ces zones a l’abri des broussailles. Couper ou faucher pour faire du foin peut egalement etre rentable. Cependant, le paturage par les lapins semblerait etre le meilleur mo yen pour conserver les conditions d’ habitat correctes pour M. minima. Mots-cles : ecologie, conservation et gestion des habitats, dunes, Nord/Pas-de-Calais. /V 35. Viola kitaibeliana Schult. (Pensee de Kitaibel) dans les lies Sorlingues (= Scilly) Cette communication resume les donnees passees et recentes sur Viola kitaibeliana Schult. (Pensee de Kitaibel) dans les lies Sorlingues. On l’y trouve, comme ailleurs sur les cotes atlantiques de 1’ Europe occidentale, pres de la mer, sur des pelouses ouvertes d’herbe courte sur des sols sablonneux sujets a la secheresse. Broutage par les lapins, pietinement humain et autres formes de perturbation du sol sont importants pour la survie des pelouses fortement broutees, de texture ouverte ou par plaques ou pousse cette espece. La gestion des senders et la taille de la vegetation sur les dunes ont amene P apparition de plusieurs nouvelles colonies de V. kitaibeliana. Mots-cles : dunes, prairie de dune fixe a Festuca rubra-Galium verum (SD 8), broutage par les lapins, pietinement. 36. Polygonum maritimum L. (Renouee maritime) en Comouailles Cette communication resume les donnees passees et recentes sur Polygonum maritimum en Comouailles. Cette espece apparait irregulierement, mais des donnees recentes en Comouailles a Gunwalloe Church Cove et Par Beach, avec de plus sa decouverte dans plusieurs nouvelles stations ailleurs le long de la cote sud de l’Angleterre, suggerent que des recherches plus approfondies sur les plages de Comouailles sujettes aux tempetes peuvent s’averer fmctueuses. Mots-cles : conservation, distribution, comptage de populations. 37. Mentha pulegium L. (Menthe pouliot) en Grande Bretagne et en Irlande Cette communication discute le statut actuel de Mentha pulegium (Menthe pouliot) en Grande Bretagne et en Irlande. Son declin en tant que plante indigene, principalement du a la perte de son habitat, a ete compensee ces demieres annees par son accroissement du fait de son introduction comme polluant des melanges de graines d’herbes. Etablir si des populations particulieres sont indigenes ou introduites est une condition prealable importante pour une action de conservation 16 ATLANTIC ARC efficace selon le Biodiversity Action Plan du Royaume-Uni. Les populations etrangeres averees ou suspectees semblent comprendre des plantes qui sont plus hautes et plus robustes que leurs contreparties indigenes. Les resultats preliminaries d’ etudes moleculaires font penser que les populations introduites peuvent etre mises a part genetiquement. Mots-cles : conservation, populations indigenes et introduites, empreintes genetiques. 38. Hypericum undulatum Schousboe ex Willd. (Millepertuis ondule) dans le sud-ouest de l’Angleterre et au Pays de Galles Cette communication examine le statut actuel d’ Hypericum undulatum (Millepertuis ondule) en Angleterre et au Pays de Galles. On pense que cette espece a decline les demieres decennies, surtout en raison de la pression ou de la deprise agricoles, mais il y a desormais diverses initiatives conservatoires en place pour proteger V espece et la vegetation des prairies marecageuses occidentales et des paturages de jonc ou on la trouve habituellement. Mots-cles : distribution, conservation, prairies de Culm, paturages de Rhos. ✓ 39. La conservation de Spiranthes romanzoffiana Cham. (Spiranthe de Romanzoff) en Ecosse : le role de la production de bourgeons lateraux jumeaux Le developpement de capsules mures chez Spiranthes romanzoffiana Cham. (Orchidaceae) n'a jamais ete observe en Ecosse, ce qui laisse a penser que la production de graines dans les populations ecossaises est rare, extremement rare ou non existante de nos jours ; et qu’elle peut n’impliquer que de petites quantites de graines. II s’ensuit que la reproduction vegetative a des chances d’ avoir une importance considerable. Les pieds de S. romanzoffiana se reproduisent vegetativement par la production de bourgeons lateraux jumeaux. Sur Tile de Colonsay en 2000 et 2001, les taux de production de bourgeons jumeaux dans les populations de reference etaient de 2,4% et 1,8% respectivement, alors que la frequence des plantes jumelles etait de 4,3% et de 3% (basee sur les donnees d’une seule visite). La perte de Tun des bourgeons jumeaux et des bourgeons produits par les plantes jumelles etablies se produisent au fil du temps. Sur 1’Ile de Coll, les taux de production de bourgeons jumeaux en 2000 et 2001 etaient tres semblables a ceux de Colonsay (2,1%, 3,7%, et 0% ; par echantillonnages), mais la frequence de plantes jumelles dans les echantillons de 2000 etait de 26,3% et de 22% (donnees d’une seule visite). Ces resultats indiquent que 1’ importance relative de cette forme de reproduction vegetative peut varier d’une lie a 1' autre. Nos resultats sont compares et mis en contraste avec ceux obtenus par analyse genetique. II est essentiel de comprendre les differentes manieres dont cette espece se reproduit en Ecosse pour que des strategies conservatoires d’ ensemble et des plans de gestion detailles puissent etre mis en place pour sa survie a long terme. Mots-cles : orchidee, conservation, reproduction vegetative, paturage, bourgeons. 40. La forme de Ranunculus repens L. (Renoncule rampante) des « turloughs » (mares exondables) Les « turloughs » sont des mares temporaires dans des zones de calcaire karstique. On ne les trouve presque exclusivement qu’en Irlande, et bien que le regime d’inondation soit largement saisonnier, l’inondation peut se produire a tout moment de l’annee a la suite de fortes chutes d’eau. Ce regime hydrologique unique a donne lieu a une flore appauvrie mais hautement interessante, incluant une forme morphologiquement differente de Ranunculus repens aux feuilles fortement decoupees. Cette morphologie foliaire determinee genetiquement est accompagnee d’autres caracteristiques physiologiques, comprenant differents modes d’ accumulation saisonniere d’ hydrate de carbone, une photosynthese et une respiration sous-marines plus importantes, et des modes de distribution stomatale alteres par rapport aux populations ne se trouvant pas dans les mares temporaires. De plus, alors que les populations hors des mares temporaires montrent une elongation du petiole typique de la reponse a « l’accoutumance a la profondeur » quand elles sont submergees, cette elongation du petiole ne se produit pas chez les populations des « turloughs » qui survivent a plusieurs metres de submersion tous les ans sous forme de plantes tout a fait vertes et metaboliquement actives. L’ analyse A.F.L.P. a montre que les populations de Ranunculus repens des « turloughs » peuvent montrer des affinites genetiques plus proches les unes par rapport aux autres qu'aux populations ne poussant pas dans les « turloughs », ce qui indique une ascendance evolutive commune. Nos etudes detaillees d’un choix de « turloughs » ont montre que malgre la capacite de R. repens a une croissance clonale abondante, un nombre relativement grand de genotypes distincts se trouvent dans ces habitats. Nous suggerons que cette diversite genetique RESUMES . 17 est le resultat d’un heterogeneite de micro-habitat au coeur du « turlough ». II semble done probable que de fortes pressions selectives aient maintenu diverses caracteristiques morphologiques et physiologiques chez les populations de Ranunculus repens pour permettre sa survie dans un environnement aussi difficile. Mots-cles : Irlande, turlough. Ranunculus repens , morphologie, adaptation physiologique a la submersion, biologie de population, A.F.L.P. 41. Baldellie fausse-renoncule ( Dyfr lyriad ymlusgawl), Baldellia ranunculoides subsp. repens (Lam.) A. Love & D. Love au Pays de Galles Les populations de Baldellia ranunculoides du Pays de Galles peuvent etre separees en deux types : une variante rare et stolonifere croissant dans des eaux peu basiques et la plante plus typique, dressee, des habitats faiblement saumatres ou calcaires. Elies sont toutes deux davantage caracterisees par une serie de caracteres morphologiques et phenologiques qui demeurent constants en culture. Ces differences, que le Reverend Hugh Davies fut le premier a observer a Anglesey il y a presque deux cents ans, correspondent a la distinction modeme de la subsp. repens et de la subsp. ranunculoides dans les taxonomies d’ Europe continentale. Mots-cles : Baldellia repens, Baldellia ranunculoides subsp. ranunculoides, caracteres morphologiques, stolons, taxonomie, conservation. 42. Isoetes histrix L. (Isoete epineux) sur la Peninsule du Lizard Cette communication resume ce qui est connu de la distribution et de l’ecologie d 'Isoetes histrix L. (Isoete epineux) a la limite septentrionale de sa distribution mondiale sur la Peninsule du Lizard dans l'ouest de la Comouailles. Mots-cles : Isoete epineux, ecologie, gestion conservatoire, mares provisoires, cuvette d‘ erosion. 43. Pilularia globulifera L. (Pilulaire a globules) en Comouailles Cette communication resume les donnees passees et recentes sur Pilularia globulifera L. (Pilulaire a globules) en Comouailles. C’est une espece sur le declin en Grande Bretagne, et elle a certainement dispam d’un certain nombre de ses stations isolees en Comouailles ; mais elle est encore presente dans de nombreuses stations dans les zones de son bastion sur la Peninsule du Lizard et les landes du centre de la Comouailles. Une plus grande appreciation des exigences de son habitat et une meilleure capacite chez les botanistes a la reconnaitre sur le terrain ont conduit a la decouverte de plusieurs nouvelles populations et a la redecouverte de plusieurs vieilles stations que Ton pensait dispames. Mots-cles : conservation, distribution, plans d'eau ephemeres, lacs, etangs, Peninsule du Lizard, landes du centre de la Comouailles. 44. Ranunculus ophioglossifolius Vill. (Renoncule a feuilles d’ophioglosse) dans la region Nord/ Pas-de-Calais du nord de la France Cette communication resume les travaux recents sur la distribution, les preferences d' habitat et la situation actuelle de Ranunculus ophioglossifolius Vill. (Renoncule a feuilles d'ophioglosse) dans la region Nord/Pas-de-Calais. Cette espece se cantonne dans seulement trois stations (toutes littorales) et elle est a l’heure actuelle l’objet d’un plan de conservation detaille. Les plus grandes populations existantes se trouvent dans la vallee inferieure de la Slack, pres de Boulogne-sur-Mer. Mots-cles : comptages de population, plan de conservation, projets agro-environnementaux. 45. Echium plantagineum L. (Viperine faux-plaintain) a Boscregan, St-Just : sa conservation et sa gestion Cette communication resume le succes des tentatives par le National Trust depuis 1995 pour preserver E. plantagineum (Viperine faux-plantain) a Boscregan, St-Just (ouest-Comouailles). Le retablissement d'un regime de culture a faible apport a conduit a une augmentation rapide de la population de cette espece et d’autres especes cultivables ; maintenant, sa reintroduction est proposee sur une ferme voisine appartenant au National Trust. Mots-cles : archeophyte, plantes herbacees cultivables, conservation, National Trust. 19 The geological background to the landscape of the Atlantic Arc C. M. BRISTOW 46 Chatsworth Way, Carlyon Bay, St Austell, Cornwall PL25 3SN, England ABSTRACT The maritime territories on the western fringe of Europe, stretching from Scotland southwards through Ireland, Wales, south-west England, Brittany, and on to north-western Iberia, consist of a series of massifs of older rocks formed during the Variscan and earlier orogenies. With the exception of some older rocks in Scotland, northern Ireland, north Wales and parts of Brittany, the rocks were originally laid down in equatorial latitudes during the Devonian and Carboniferous Periods as muds, sands and volcanic rocks in a trough between two converging tectonic plates. The collision between these plates compressed, folded and faulted these trough materials to form the hard rocks of the massifs. Granites were intruded and occupy large areas of the massifs. Following a period of hot and arid conditions in the Permian and Triassic, humid conditions returned during the rest of the Mesozoic and continued into the early part of the Tertiary, when extensive planation began the process of forming the present landscape. Sub-tropical vegetation and deep lateritic weathering characterise this period. Climate cooling in the later part of the Tertiary led to the Ice Ages of the Quaternary. Palynological research on important occurrences of sediment belonging to the late Tertiary in Cornwall tells us much about the climate and vegetation of the period leading up to the onset of cold conditions. During the glacial periods Cornwall and areas to the south were not glaciated, but suffered extremely low temperatures and periglacial features tell us much about the severity of the climate. Wales, Ireland and areas to the north were glaciated, which resulted in most of the pre-glacial weathering mantle being removed and glacial landforms produced. Towards the end of the last (Devensian) glaciation the climate became not only cold, but also extremely arid, resulting in dust being blown from the glacial outwash sediments on the floor of the Irish Sea; sea level at that time was over 100 m lower than present. This dust forms a loessic layer over parts of Cornwall and is probably an important component of soils in Cornwall and adjoining regions. Climate amelioration over the last 10,000 years, since the end of the last pulse of glaciation, has led to the present soil and vegetation cover. The paper concentrates particularly on the central part of the Atlantic Arc, notably Cornwall, although mention is made of regions to the north and south; it also briefly describes some of the botanical features of the geological history. KEYWORDS: palaeobotany, geology, granites, weathering, periglacial processes, Armorican. INTRODUCTION The geological histories of the territories of the Atlantic Arc have much in common, although there are some important differences in the most recent stages of the geological history, during the Ice Ages, arising from the considerable spread of latitudinal positions. The development of the landscape, climate and soils has had a fundamental influence on the present day flora of the territories and this paper aims to explain the geological evolution of the area with this in mind. Where applicable, references to past floras are also mentioned. THE FORMATION OF THE VARISCAN MASSIFS Figure 1 shows the distribution of the older massifs in Western Europe. Most of the Atlantic Arc territories are underlain by rocks formed during the Palaeozoic Era although there are some older rocks in Scotland, Ireland, Wales and Brittany which were involved in earlier mountain building events. We can further narrow down the period when most of the Atlantic Arc rocks were formed, to the younger part of the Palaeozoic Era - the Devonian Period which began 410 million years ago and the Carboniferous Period which ended 290 million years ago. In south-west England and many other parts of the Atlantic Arc, many of the rocks of the Devonian and Carboniferous Periods were formed as sediments and volcanics laid down under the sea (Bristow 2004). 20 ATLANTIC ARC Mainly formed during the Variscan Orogeny 350-290 million years ago. Mainly formed during the earlier orogenies (pre 400 million years ago) In Cornwall, most of the rocks are of Devonian age and Figure 2 shows typical Devonian sediments, consisting of greywacke sandstones and mudstones, laid down under marine conditions, in Pendower Bay, on the south coast of Cornwall, about 9 km south-east of Truro (Leveridge et al. 1990). Land cannot have been far away for the fossilised remains of wood and other vegetation are found in these rocks. The age of these rocks was first determined by the discovery of fossil wood of Dadoxylon hendriksi by Miss E. M .L. Hendriks (1949) (a species first described by Lang (1929)), which indicates the rocks cannot have been laid down earlier than the Middle Devonian. More recently palynology has been used to further refine the dating of these rocks, and miospores and acritarchs extracted and identified by the British Geological Survey are now known to indicate a late-Devonian age (Leveridge et al. 1990). In the early part of the following Carboniferous Period the Carboniferous Limestone was deposited, which underlies large areas of Wales and southern and central Ireland (Fig. 3). Mountain building had already started by the early Carboniferous in south Cornwall, although deposition of muds and sands continued in north Cornwall and central Devon, together with much volcanicity. Later in the Carboniferous Period, great swamps developed over much of the British Isles, which would, after burial, become the Coal Measures. GEOLOGICAL BACKGROUND 21 Figure 2. Greywacke sandstones and slates of Middle/Upper Devonian age, originally laid down as marine turbiditic sands and muds, seen on the foreshore of the western part of Pendower Bay. Figure 3. Scarp of Carboniferous Limestone near Llangollen, North Wales. This limestone was laid down in the early part of the Carboniferous Period in a shallow warm sea. 22 ATLANTIC ARC At this time there were two giant tectonic plates: Gondwanaland in the south and Euramerica in the north with, in between, a narrow ocean which geologists call the Rheic Ocean. These two great plates were slowly converging and the sediments and volcanic rocks laid down in the trough between them were being compressed. By the beginning of the Carboniferous the crumpled mass of sediments in the collision zone began to be piled up to form the Variscan mountain chain. This collision of tectonic plates is known as the Variscan Orogeny and many of the ancient massifs of western Europe were formed by this orogeny. The Lizard, which forms the southernmost part of Cornwall (Plate 2), is of great interest to geologists as it represents a battered mass of oceanic crust which somehow was incorporated into the pile of rocks forming the Variscan Mountains. Oceanic crust is high in magnesium and iron and low in silica, and this has had a marked effect on the flora of the Lizard. As the crumpled mass of sediments was piled up ever higher to form the mountain chain, so the roots of the mountain chain were forced ever deeper by the weight and one plate began to descend below the other. As a result some continental crust became so hot that melting began and granite magma was formed. Some of the molten granite masses rose up to within a few kilometres of the surface, notably in south-west England, and volcanicity broke out above where the granite masses are now. The accompanying heat sweated out metals to form orefields such as the Cornubian orefield of south-west England. POST-VARISCAN OROGENY EVENTS AND THE MESOZOIC The climate then became hot and arid, as most of the area we now know as the Atlantic Arc lay only just north of the equator and was situated in the centre of an enormous continental mass. Occasional storms sent rivers down the normally dry valleys taking great quantities of stones and sand with them. Gradually erosion reduced the mountains to gentle hills and, about 210 million years ago, the sea invaded the lower lying areas. However, it is probable that for long periods of the Mesozoic Era, the time of the dinosaurs, Cornwall and the other Variscan massifs were large low-lying islands. This area was much further south than we are now and the whole planet was warmer than present, so the climate would have been tropical, probably with a humid climate not unlike present day Jamaica or Singapore. This would have caused intense deep chemical weathering. This type of weathering, given time, will gradually reduce a mountain range to a low relief planation surface and it is at this stage that the landscape began to take on its present form. In west Cornwall there is a widespread development of a flat plateau, cut in the slaty Devonian rocks, known as the Reskajeage surface. The earliest stage in the formation of this planation surface may well have taken place in Mesozoic times. It forms a gently undulating surface, which is cut by deep river valleys formed much later. It slopes gently away from the spine of the peninsula and is truncated by the cliffs at the coast, which are also a much later feature. It probably originally continued sloping gently towards a distant shoreline. In Mesozoic times the planation surface would have been covered with a deep weathering mantle, perhaps up to 100 m thick. Occasional deeper pockets of this weathering mantle are still found in Cornwall, but most of the Mesozoic weathering mantle has been swept away by later erosion. Similar planation surfaces are developed in Brittany, although at a lower level, so the deep valleys and coastal cliffs are much less prominent, especially on the south coast. Again, the earliest phase in the development of this surface may well have taken place in Mesozoic times. There is widespread development of a kaolinitic weathering mantle in south and east Brittany and many shallow excavations reveal this material; although unravelling whether this was formed in the Mesozoic or in the later Tertiary is not easy. China clay deposits were partially formed by this deep weathering and are found not only in Cornwall, but also in Brittany, Galicia and northern Portugal. Figure 4 shows a china clay pit in southern Brittany, near Lorient. At the end of the Mesozoic, during the Cretaceous Period, rift valleys formed to the west of our area and split up the old supercontinent. As the rift valleys widened they became filled by the sea, to begin the formation of the Atlantic Ocean, which has been widening ever since. Towards the end of the Cretaceous, sea levels briefly rose to unprecedented heights. This caused much of western Europe to be submerged beneath the chalk sea for a short time. GEOLOGICAL BACKGROUND 23 Figure 4. The china clay pit of Kaolins d'Arvor at Ploemeur, near Lorient, southern Brittany. One of the earliest phases in the development of this kaolin deposit may well have been chemical weathering in the Mesozoic. Figure 5. Cliff near Piriac, southern Brittany showing chemically weathered schist (pale, above) overlying darker coloured unweathered schist. The boundary between the weathered and unweathered material is remarkably sharp. 24 ATLANTIC ARC TERTIARY EVENTS Planation development returned in the early Tertiary and weathering attributable to this time is widespread in Brittany, where there has been less erosion of the weathering profile than in Cornwall because of the lower relief. Indeed, this widespread weathering (Fig. 5) is such a nuisance to French hard rock geologists that they call it “La Maladie Tertaire”. Similar planation surfaces are developed in northern Portugal and Galicia; some of the planation surfaces in this area are, however, later and may be mainly Quaternary features. We now enter a particularly interesting time from the palaeobotanical point of view. Something strange seems to have happened to the climate about 35 million years ago, when a climate developed unlike any on the planet today. We know a great deal about this time, as the valuable ball clay deposits in Devon at Newton Abbot and Petrockstow are sedimentary kaolinitic clays laid down in Eocene and Oligocene times (about 35 million years ago) in fault formed basins. The clays are often stained dark brown or black, due to the presence of much organic material and in south Devon there are extensive seams of lignite (Plate 6). Much woody debris is preserved and palynological work by a number of researchers, reported in the BGS Newton Abbot Memoir (Selwood et al. 1984) revealed two distinct floras: “Upland flora” Sequoia couttsiae (woody debris with well defined growth rings abundant) Osmunda (Fern) Pityosporites (Conifer) “Swamp flora” True aquatics: Stratiotes , Potamogeton & Brasenia Climbing plants: Vines and Rubus Plants and trees overhanging the water: Nyssa, the Lauraceae, Symplocos, Carpinus, Magnolia and Meliosma Fern: Polpodiaceaesporites The overall impression is of a warm humid sub-tropical climate which was frost free with a flood plain or swamp flora dominated by water loving plants. The “Upland flora” dominated by Sequoia couttsiae was probably swept into the basin by periodic floods. The presence of growth rings indicates seasonality. In the North Devon Basin at Petrockstow a 35 million year old weathering mantle is preserved beneath the ball clay sediments in Woolladon ball clay pit. Studies (Bristow 1968; Bristow & Robson 1994) have shown that this weathering mantle is similar to weathering mantles which one finds nowadays in places like south-east Asia. One of the most curious features of ball clays is that throughout much of Europe and North America the best ball clays are of similar age (Bristow 1990). This suggests some kind of climatic control. Possibly, if the whole planet is warmed up, as happened in the early part of the Tertiary, at about 40-45° north a zone of deep lateritic weathering developed, in addition to the equatorial zone which we are familiar with today, which could have provided the right kind of weathered material to form ball clays. Possibly a feint relic of this northern zone of deep lateritic weathering will be found in north¬ west Spain, where there is a warm temperate, nearly frost free humid climate. Confirmation of this may be found in the occurrence of ball clays, laid down in geologically very recent times in the Porrino Basin, south of Vigo (Fig. 6). One particularly interesting feature is the steep drop in global temperatures at the end of the Eocene. It is tempting to believe that this drop, possibly associated with increased precipitation, was somehow involved in the erosion and transport of material into the ball clay basins. In the later part of the Tertiary, known as the Neogene, the temperature had fallen to a level where the climate was in between present day climatic conditions and that of the early Tertiary, which meant that, in Cornwall, the climate was similar to the present day Mediterranean. A small outlier of Tertiary sands and clays is found in Cornwall near St Agnes and in 1983 a pollen assemblage was obtained from a lignitic bed in one of the sand pits. It is dated to the Miocene and to quote Walsh et al. (1987): GEOLOGICAL BACKGROUND 25 Figure 6. Quaternary ball clays being exploited in the Porrino Basin south of Vigo, Spain. Weathering of the rocks of the Galician Massif under humid warm temperate conditions has produced kaolinite rich material which has been washed into the Porrino Basin to form some localised deposits of ball clay. This may be a feint contemporary relic of the type of weathering which produced the ball clays of South-west England at 40-45°N in the Palaeogene. Figure 7. Photograph of the St Erth sand/clay workings in the early part of the 20th century. These sediments were deposited in a shallow sea which formerly separated the Penwith ‘island’ from the mainland. They were deposited just before the Ice Ages and provide a valuable insight into climatic conditions at that time. Reproduced with permission from a photograph supplied by the Royal Institution of Cornwall. 26 ATLANTIC ARC “The Neogene pollen assemblage suggests a subtropical Mediterranean climate with a typical European vegetation of conifer forest and mixed woodland and shrub. The assemblage is similar to inland conifer forest (Pinus- like species, Taxodiaceae, etc.) from upland vegetation and mixed angiosperm and conifer forest perhaps from less elevated surfaces (tricolpates and conifers). In addition, shrubland and heathland type vegetation may be represented by forms such as P. vestibulum, P. versus and Erecipites. The absence of palm pollen may mean that winter temperatures fell below 0°C, and the absence of dinocysts confirms that marine influences were absent.” A further insight into climatic conditions just before the onset of the Ice Ages is provided by some small outliers of marine late Pliocene sands and clays near St Erth in Cornwall, formerly exploited for moulding sand and clay (Fig. 7). Head (1999) provided a valuable insight into the botany of the terrestrial environment in the QRA Field Guide based on the palynology: “Five plant communities have been recognised: Estuary and salt marsh Freshwater Damp soil Woodland and dry soil Coniferous woodland, heath and bog The dominant pollen is Pinus (47%) and Ericales (21%), characteristic of coniferous woodland and heath, and Chenopodiaceae (17%) which is characteristic of estuaries and saltmarsh. Seeds of the eelgrass Zostera noltii are present, this plant growing between the tide marks in the British Isles today. The spores and pollen accumulated in a marine environment fairly close to shore: hence any reconstruction of the regional vegetation must allow for edaphic and taphonomic factors. Two aspects of the assemblage are clearly significant: Firstly, thermophillic deciduous forest elements are poorly represented; Nyssa or Fagus, Sciadopitys and Quercus occur only rarely, and Tilia, Acer, Liquidambar and Juglans are absent. Secondly, the presence of Ilex indicates summer temperatures of not less than 15°C and winters no cooler than 0°C, and the presence of the frost-sensitive conifer Sciadopitys likewise implies mild winters. These aspects point to a cool- or mild-temperate climate.” THE QUATERNARY - THE ‘ICE AGES’ Britain now succumbed to the rigours of the Ice Ages. Any idea that any of the warm-temperate or sub-tropical species already mentioned could have survived the cold conditions must be dismissed as, at times, conditions were very cold indeed, perhaps comparable to northern Canada or Siberia today, even in Cornwall. The Ice Ages consisted of at least three major cold periods when, because of the amount of water contained in the huge continental ice sheets, sea levels fell by up to 120 m, so the coast was a considerable distance offshore and most of the Irish and Celtic Seas were dry land. During the glaciations the rivers cut their beds down well below present sea level, in response to the lowered sea level; with the mud-laden spring floods having huge erosive capacity. Plate 5 shows an estuary at Looe which would have had a river rushing along its bed, perhaps about 30 m below present sea level, during the coldest periods of the Quaternary. In Wales and Ireland thick ice sheets developed, which swept away nearly all of the earlier weathering mantles and produced new landforms due to the action of the glaciers. Moraines and other glacigenic sediments covered the lower ground with a thick mantle of what geologists call ‘drift’, notably in Ireland. As this drift contains material derived from a wide variety of rock types, the chemistry of the present day surface soil no longer reflects the underlying ‘solid’ geology. Plate 3 demonstrates the sequence of events before, during and after the last glaciation in a section just west of Spit Beach, near Par in South Cornwall (Bristow & James 2002). On the underlying Devonian slates, about 5-8 m above present O.D., there is a raised beach deposit about 2-3 m thick, which was probably formed in the last interglacial, when conditions were slightly warmer than present. It is a raised beach, because all the pebbles are rounded by marine action. On top of this is a deposit of what is known geologically as ‘Head’. This contains angular fragments of slate, which indicates that they have been split by frost action. GEOLOGICAL BACKGROUND 27 During the coldest periods the frost penetrated to a considerable depth and the quantity of ice built up so that the whole frozen mass of soil and weathered material was able to move very gently downslope rather like a very dirty glacier, causing the accumulation of ‘Head’, particularly at the foot of north facing slopes. Loessic material may also have been incorporated into the head, see below. Plate 7 shows another periglacial feature called an ‘ice wedge’ which is also to be seen at Spit Beach near Par. The horizontal bedding of the raised beach is interrupted by a vertical feature, containing pebbles with their long axes vertical. This represents a contraction crack formed under exceptionally cold conditions. These cracks form a polygonal pattern and under present day conditions do not form unless the mean annual temperature is less than -6 °C. The coldest point of the last glacial was about 20,000 years ago, which could be when this feature formed. At one stage a tongue of ice extended across the Irish Sea and deposited moraine in the northern isles of the Isles of Scilly. However, Cornwall and Devon, as well as Brittany and Galicia, were unglaciated. Because sea level had fallen by up to 120 m, the coast was way out near the edge of the continental shelf during the coldest periods. At the end of the penultimate glacial, temperatures at first rose for a brief interglacial, followed by a long and rather erratic decline in temperatures and sea level, to the coldest point of the last glacial, about 20,000 years ago. There was then a rapid warming, briefly interrupted by a return of cold conditions about 1 1,000 years ago, followed by further warming to present day conditions. In Plate 3 a layer of fine silty material about 0-5 m thick is seen under the present day soil and turf. Particle size analysis shows that this is likely to be loess, which is a wind blown dust formed under cold but extremely dry conditions. This is a common feature in south Cornwall and much of the fine matrix of the Head may also have a loessic origin. Because of the lowered sea level, a large area of the floor of the Irish Sea was dry land composed of fine glacial outwash material and this was picked up by strong winds under the ‘dust bowl’ conditions of between 20,000 and 15,000 years ago and blown towards what is now Cornwall, to be deposited on lee slopes to form a layer of loess. If the theory that this loess was derived from dust blown into this area between 20,000 and 10,000 years ago, ultimately derived from glacial outwash sediments in the present day Irish Sea area, then this could have implications for the early post-glacial soils in adjoining areas. The sediments on the floor of the Irish and Celtic seas have been shown to have an appreciable carbonate content (Scourse et al. 1990) and the glaciers feeding material into the Irish Sea basin would have come from areas where there are extensive limestone outcrops, especially Carboniferous Limestone, so it is reasonable to suppose that the dust would have had an appreciable carbonate content, which would have been reflected in the vegetation growing on the early post-glacial soils. The abundance of hazel nuts in the vegetable layer overlying the tin ground (see below), described by many writers when these alluvial deposits were worked in the late 18th and early 19th centuries, may be a reflection of the alkaline nature of these soils (Bristow 2005). Leaching by rain over the last 10-15,000 years would have gradually removed this carbonate to transform the soils into the acid soils we are familiar with today. From about 15,000 years ago the climate began to ameliorate, but between 11,000 and 10,000 years ago there was a period of renewed cold, known as the ‘Younger Dry as’. At this time glaciers reformed briefly in locations such as Snowdonia and the Scottish highlands, which means that the glacier-cut landforms are still remarkably fresh (Fig. 8). Following the Younger Dryas cold period the climate warmed rapidly and the record of the Flandrian transgression in Cornwall is superbly preserved in the sediments laid down in the overdeepened valleys formed during the glacials, such as the Pentewan valley near St Austell, site of the famous ‘Happy Union’ tin working, and the ‘Poth’ workings at Par described by Philip Rashleigh in 1792 (Rashleigh 1802). The basal layer at Poth was a poorly sorted gravel, known as the ‘tin ground’, as it contained considerable quantities of cassiterite, the ore of tin, and was much sought after by tin streamers, even to the extent of working it underground, under the sea. It was probably formed from high density flows of muddy, sandy gravelly material which flowed down the valleys in the closing stages of the last glacial period. This is then followed by a layer of peaty and woody material containing hazel and oak, sometimes in the position of growth; radiocarbon dates from Mount’s Bay suggest a date of around 10,000 years ago for this early peat layer. In the Pentewan valley this layer contained undecayed oaks in their position of growth, with oysters 28 ATLANTIC ARC Figure 8. Looking towards Llyn Gwynant in Snowdonia. Glaciers briefly reformed in Snowdonia during the Younger Dryas cold period 1 1-10,000 years ago, so the ice-formed landforms are relatively fresh. attached to the boles of the oaks. This would seem to imply that the initial rise in sea level was extremely rapid, perhaps by as much as 6 m in 100 years. This is in turn overlain by sand containing sea shells, indicating that the sea level rise had overtaken the rate of sedimentation. A further peat layer is then found at Poth, indicating a return to land conditions and then further silts and sands, which may represent debris brought down by the stream after alluvial tin working had started in Bronze age times. This upper peat layer in Rashleigh’s Poth section corresponds with the many occurrences of ‘submerged forest’ found around Cornwall at, or just below, sea level. These would seem to indicate that Cornwall was surrounded by a flat forested coastal plain in the period from about 8000 to about 3-4000 years ago, with the coast possibly marked by a low line of dunes and a sandy beach. As the sea level rose, this coastal plain was inundated, until the sea again re-occupied its temperate ‘interglacial’ position at the base of the cliffs around the peninsulas of Galicia, Brittany, Cornwall, south Wales and southern Ireland. One of the features of the north Cornwall coast, as at Perranporth (Plate 4), is the presence of extensive sandy beaches and dune fields. These sands contain between 30% and 90% calcium carbonate, derived from marine organisms’ shells, blown onto the shore by the prevailing westerly winds. The strong Atlantic gales are capable of driving this carbonate rich sand some distance onshore and ‘plumes’ of alkali tolerant plant communities are found extending some distance inland as a consequence. Finally, one of the main lessons to learn from recent geological history is that it is normal for the climate to change, sometimes with alarming speed. To the geologist, there is no such thing as ‘normal climate’, or ‘normal sea level’ or, for that matter, ‘normal vegetation’. Continual change is normal. REFERENCES Bristow, C. M. (1968). The derivation of the Tertiary sediments in the Petrockstow Basin, North Devon. Proceedings of the Ussher Society ; 2: 29-35. BRISTOW, C. M. (1990). Ball clays, Weathering and Climate, in ZUPAN, A. W. & MAYBIN, A. H. Ill, (Eds.) Proceedings 24th Forum on the geology of Industrial Minerals, May 2-5, 1988, Greenville, South Carolina, pp. 24-37 . South Carolina Geological Survey, Columbia, SC. GEOLOGICAL BACKGROUND 29 BRISTOW, C. M. (2004). Cornwall’s Geology and Scenery, 2nd Edn. Cornish Hillside Publications, St Austell. BRISTOW, C. M. (2005) (In Press) Hazel nuts and the evolution of Holocene soils in Cornwall. Geoscience in south-west England. Bristow, C. M. & James, H. C. L. (2002). Field Excursion to the area of St Austell Bay between Carlyon Bay and Par, 2nd January, 2002. Geoscience in south-west England 10: 373-376. BRISTOW, C. M. & Robson, J. (1994). Palaeogene basin development in Devon. Transactions of the Institution of Mining and Metallurgy 1023: 8-13. Head, M. J. (1999). The Late Pliocene St. Erth Beds of Cornwall: a review of the palynology and reappraisal of the dinoflagellates, in SCOURSE, J. D. & FURZE, M .F. A. (eds.) The Quaternary of West Cornwall - Field Guide, pp. 88-92. Quaternary Research Association, London. HENDRIKS, E. M. L. (1949). The Gramscatho Series. Transactions of the Royal Geological Society of Cornwall 18: 50-61. Lang, W. H. (1929). On fossil wood ( Dadoxylon hendriksi , n.sp.) and other plant remains from the clay-slates of S. Cornwall. Annals of Botany 43: 663-683. Leveridge, B. E., Holder, M. T. and Goode, A. J. J. (1990). Geology of the country around Falmouth. Memoir of the British Geological Survey, Sheet 352 (England and Wales), pp. 7-14 RASHLEIGH, P. (1802). Section of the Stream Works at Poth, St Blazey (in 1792). Description of British Minerals, Part II, Plate xxi. Scourse, J. D., Austin, W. E. N., Bateman, R. M., Catt, J. A., Evans, C. D. R., Robinson, J. E. & Young, J. R. (1990). Sedimentology and micropalaeontology of glacimarine sediments from the Central and Southwestern Celtic Sea, in Dowdeswell, J. A. & Scourse, J. D. eds. Glacimarine environments: processes and sediments. Geol. Soc. Spec. Pub., 53, 329-347. SELWOOD, E. B. and others. (1984). Geology of the country around Newton Abbot. Memoir of the British Geological Sur\’ey, Sheet 339, 133-136. Walsh, P. T., Atkinson, K., Boulter, M. C. & Shakesby, R. A. (1987). The Oligocene and Miocene Outliers of West Cornwall and their bearing on the geomorphological evolution of Oldland Britain. Philosophical Transactions of the Royal Society of London 323: 21 1-245. 31 Phytogeography of Ireland, past and present S. WALDREN* Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6, Ireland P. COXON Department of Geography, Trinity College, Dublin 2, Ireland and N. KINGSTON National Parks & Wildlife Service, Department of the Environment, Heritage & Local Government, 7 Ely Place, Dublin 2, Ireland ABSTRACT Compared to nearby Britain and continental Europe, the flora of Ireland is depauperate in species. However, some biogeographic elements are particularly well represented in Ireland, and these include species with predominantly Atlantic or European oceanic distributions. We suggest that there are good ecological reasons underlying these distributions, and moreover that changes in distribution of Irish species since the Pliocene reflect similar trends - a loss of woody continental species but the repeated presence of Atlantic herbs and small shrubs, many of which are wetland species in the broadest sense. Determining the timing of isolation between Irish and continental populations, and, in the case of amphi-Atlantic species, of North American populations, will be crucial to a better understanding of Irish phytogeography. We anticipate that molecular phylogenetic and phylogeographic studies will contribute greatly to this knowledge. Keywords: biogeography, vascular plant taxa, Atlantic elements, quaternary history. INTRODUCTION - THE PRESENT FLORA OF IRELAND The island of Ireland, lying at the western extremity of Europe, has a depauperate but none the less interesting flora in comparison to the nearby large island of Britain, and to continental Europe. The present flora of Ireland has been shaped by historical, ecological, climatic and biotic events. The Pleistocene glaciations and the insularity of Ireland have undoubtedly had a large influence in determining the flora, while particular ecological and climatic conditions - especially that of oceanicity - have altered the types of species able to inhabit Ireland. More recently, human influence through the spread of agriculture, and by the introduction of some 920 non-native species that are now considered to some degree established in Ireland (Reynolds 2003), have also had dramatic effects. This review aims to draw together several themes regarding the phytogeography of Ireland. It has been strongly influenced by two major contributions to the biogeography of Ireland. Firstly, David Webb’s review considered the flora of Ireland in a broader European context (Webb 1983). Secondly, Harry Godwin’s review of the history of the flora of the British Isles gave an historical context to the origins of the flora, particularly with respect to Pleistocene glaciations (Godwin 1975). We have previously developed some of the ideas of both Webb and Godwin, by reviewing information on interglacial records from the Irish flora (Coxon & Waldren 1995, 1997). Subsequently, Chris Preston and Mark Hill’s monumental review of the biogeography of the flora of the British Isles (Preston & Hill 1997) provides a framework with which to discuss these data in *e-mail: swaldren@tcd.ie 32 ATLANTIC ARC TABLE 1. SELECTED BIOGEOGRAPHIC AFFINITIES OF THE IRISH AND BRITISH FLORAS; DATA DERIVED FROM PRESTON & HILL (1997) Element British Isles % British Isles total Ireland % Irish total % British Isles Wetland species Arctic-montane 79 5-33 22 2-32 28 1 Oceanic Boreal-montane 7 0-47 4 0-42 57 4 Oceanic Boreo-temperate 8 0-54 6 0-63 75 3 Oceanic Temperate 47 3-17 39 411 83 22 Oceanic S-temperate 25 1-69 10-5 Ml 42 7 Mediterranean- Atlantic 69 4-66 38 4-00 55 8 Total 1481 950 64 The columns indicate the numbers of taxa in each element in the British Isles, these values expressed as a percentage of the total flora of the British Isles, the numbers of taxa in each element in the Irish flora, these values expressed as a percentage of the total flora of Ireland, the percentage of the total flora of the British Isles for each element that occurs in Ireland, and the numbers of wetland (including aquatic) taxa in each category. greater detail. We will concentrate particularly on those species which are restricted in Europe (though not necessarily so elsewhere) to the western extremity of the continent, the so-called Atlantic species. No list of species native to a given area is ever likely to meet with universal approval. Reasons for disagreement include differences in taxonomic opinion, such as recognition of narrowly defined species of Euphrasia , or the validity of including apomictic ‘species’, and the difficult decisions regarding native or introduced status. Taking the numbers of vascular plant taxa listed in the Biodiversity Data Sourcebook (WCMC 1994) as a validly published comparison, the general trends in the floristic diversity of European countries becomes apparent: Ireland (950 species) is much more floristically depauperate than nearby Britain (1623), France (4630), Belgium (1452) and the Netherlands (1221). Webb (1983) argued that insularity from both Britain and the continent explains much of the impoverishment, but also cited ecological factors (smaller climatic range, more limited geology) as contributing to the floristic impoverishment of Ireland. Ecological impoverishment may well be particularly important, as Bennett (1995) showed that sea barriers of 10-100 km are unlikely to present dispersal barriers to the majority of the tree and shrub flora of the British Isles. ECOLOGICAL IMPOVERISHMENT - THE ARCTIC-MONTANE ELEMENT Webb (1983) pointed out the relative paucity of arctic-montane species in the Irish flora, listing 16 species as being ‘arctic-alpine’ in distribution. More recently, Preston & Hill (1997) list 1481 vascular plants as being native to the British Isles (c.f. WCMC figure above), 79 of which they place in the arctic-montane element, some 5-33% of the total (Table 1). Only 22 of these 79 occur in Ireland, and, given the estimated number of Irish native species as 950 (see above), this arctic- montane element of the Irish flora represents only 2-32% of the total Irish flora (Table 1). A glance at Preston & Hill’s Figure 14 shows the far more limited distribution of the arctic montane element in Ireland compared to Britain; the same is also true of the boreo-arctic montane element (Preston & Hill, Fig. 15, 197). Webb (1983) described the ecological differences that determine this particular impoverishment of arctic-montane species, and the main ecological constraint is certainly oceanicity (Crawford 2000; Crawford et al. 2003). With mild, wet winters and mild wet summers, annual climatic variation is far lower in Ireland than in Britain, and lower still than much of continental Europe. But it is the prevailing weather systems with depressions from the south west bringing mild but generally wet weather to Ireland which are likely to provide conditions which are not conducive to the growth of arctic-montane species. Snow rarely lingers on Irish mountains, and moist, mild conditions mean that carbohydrate reserves are depleted when arctic- montane species are normally dormant (Crawford 2000). PHYTOGEOGRAPHY OF IRELAND 33 Figure 1. Distribution of Erica cinerea , a widespread oceanic temperate species. THE ATLANTIC ELEMENT IN THE IRISH FLORA If the prevailing climatic and ecological conditions in Ireland have restricted the presence of arctic-montane species in Ireland, has it been beneficial to others? The 950-odd species considered native to Ireland represent 64% of Preston and Hill’s total of 1481 species native to the British Isles. If all biogeographic elements occurred in the same relative proportions in both Britain and Ireland, the expectation would be that Ireland contains 64% of all floristic elements occurring in the British Isles. Table 1 shows that this is not the case; as mentioned above, only 28% of the arctic-montane species occurring in the whole British Isles occur in Ireland. However, Ireland has a much higher percentage of the total numbers of boreo-temperate oceanic (75%) and temperate oceanic (83%) species found in the whole of the British Isles. These species, which tend to be restricted in range to western Europe, have often been considered ‘Atlantic’ in their distribution. Erica cinerea (Fig. 1) provides an example extending to the Netherlands, Belgium and eastern France; other taxa are more restricted to the western periphery of Europe, such as Saxifraga spathularis and S. hirsuta (Fig. 2). The latter form part of an interesting suite of species that have often been termed ‘Lusitanian’ in distribution, occurring in the northern part of the Iberian 34 ATLANTIC ARC FIGURE 2. Distributions of Saxifraga spathulciris and S. hirsuta , oceanic temperate species with disjunct, localised ranges. Heavy shading - S. spathularis, light shading - S. hirsuta , intermediate shading - both species. peninsula, and the western part of the British Isles. Several of these, including the saxifrages just mentioned, Daboecia cantabrica and Erica mackaiana, are not native to Britain. Ecological and climatic influences are also undoubtedly important in determining this high proportion of Atlantic species, many of which can be considered wetland or aquatic species. However, it is also possible that historical events, particularly in relation to the location of glacial refugia, may also play a role in their presence (see below). Some of these Atlantic species have their global distribution centered on Ireland (e.g. Dryopteris aemula ) or the British Isles (e.g. Hymenophyllum wilsonii, Saxifraga hypnoides ). Webb (1983) points out that many of these Atlantic species are commoner in Ireland than in Britain; the same argument also holds for continental Europe. It has sometimes been stated that Ireland contains a high number of Mediterranean-Atlantic and Oceanic southern-temperate species. While Table 1 shows that these make up a moderate proportion of the Irish total, proportionally fewer of these species occur in Ireland than in Britain. Again, the cooler, moister climate of Ireland probably limits their distribution; for example, many of the Cornish specialities in the British flora are absent from Ireland. PHYTOGEOGRAPHY Op IRELAND 35 AMPHI- ATLANTIC SPECIES IN IRELAND Several European species also occur in North America, the so-called amphi-Atlantic element. Of particular interest here are those amphi-Atlantic species which, although reasonably widespread in North America, are highly restricted in Europe. Most of these species with markedly uneven amphi-Atlantic distributions are in Europe entirely restricted to western Ireland and in some cases to the Hebrides of Scotland. Good examples include Sisyrinchium bermudiana (Ireland only), Eriocaulon aquaticum and Spiranthes romanzojfiana (Ireland & Scotland). The occurrence of these taxa with such restricted European distributions is puzzling, and several theories have been put forward to explain their anomalous distribution. Some of these species have been suggested as introductions, including Hypericum canadense (Preston et al. 2002), Sisyrinchium bermudiana (Preston et al. 2002), and Spiranthes romanzojfiana (see Plate 44) (T. Curtis, pers. comm.), though others have doubted this (Webb 1983; N. Robson, pers. comm.). Other suggestions include introduction as seed adhering to birds (e.g. Hypericum canadense ; N. Robson, pers. comm.), and even survival of glacial conditions in or proximal to Ireland (Webb 1983; Coxon & Waldren 1995). Clearly, in addition to the study of the present distributions of these and other taxa, some investigation of the history of these taxa in Ireland is needed in order to comprehend fully their biogeography. QUATERNARY HISTORY OF THE IRISH FLORA The onset of the Quaternary marked the end of relatively stable climatic conditions in the northern hemisphere, heralding the start of a series of cycles of cold glacial climates interspersed with warmer inter-glacial periods, and shorter warm periods termed stadials. Evidence for these climatic oscillations in the Quaternary comes from various sources: from ocean and arctic/ Antarctic ice cores, from sediment stratigraphy, and from analysis of fossil and sub-fossil plant and animal remains trapped in sediment (Coxon 2001). It is clear that the Tertiary flora existing at the end of the Pliocene period was radically different from the modem day vegetation of Ireland. Few Irish late Pliocene sites have been looked at in detail, but white sand and associated lignite infilling karst limestone from Pollnahallia in Co. Galway has revealed a flora consisting of numerous woody taxa (Table 2; Coxon & Flegg 1987; Coxon et al. 1998). Many of these are not only absent from the modem Irish flora, but are also absent from Europe; several occur in classic ‘tertiary refugia’ in eastern Asia (e.g. Metasequoia , Sciadopitys ) or North America (e.g. Sequoia, Taxodium ), and some, such as the classically disjunct genera Carya, Liriodendron and Thuja , occur in both regions. It has been suggested (R. Milne, unpublished data) that the reason for the persistence of such taxa in what are now widely separated refugia is due to the fragmentation of former very wide temperate ranges of these taxa during the Quaternary, coupled with relative stability of climate and ecology within the refugia, thereby reducing both speciation and extinction driven by selection. We concur with this view, but note that some of the characteristic Irish Tertiary species persisted until the middle Pleistocene, many of which have rather broader geographical distributions than those alluded to above. Compared to the late Pliocene, there are many more Irish Quaternary sites, particularly from interglacial stages, though Ireland still lacks the range of sites known, for example, in the Netherlands or Britain, thus hampering full determination of Irish Pleistocene biostratigraphy (Coxon 1993). Despite this, those sites where middle Pleistocene deposits have been discovered in Ireland reveal much useful information on the history of the Irish flora. We have previously reviewed much of the published available information on the Irish interglacial floras (Coxon & Waldren 1995, 1997), much of which centres on the ‘Gortian’ period, named from the type interglacial site at Gort investigated by lessen, Andersen & Farrington (1959). Recent studies at Cork Harbour suggest that the ‘Gortian’ may in fact represent several, non-synchronous interglacials (Dowling & Coxon 2001), thus making dating of the Gortian uncertain and hampering correlation with British and Continental interglacials. The following discussion is based largely on data assembled by Coxon & Waldren (1995, 1997), with additional evidence from selected sites (Dowling & Coxon 2001). 36 ATLANTIC ARC TABLE 2. LATE TERTIARY/EARLY PLEISTOCENE TAXA RECORDED FROM POLLNAHALLIA, IRELAND, AND THEIR PRESENCE THROUGH TO GORTIAN (G), HOLOCENE (H) OR CONTEMPORARY (M) IRISH FLORA. A DASH INDICATES NO RECORD AFTER THE LATE PLIOCENE/EARLY PLEISTOCENE Taxon Interglacial/modern status Abies G Acer G Aesculus G Alnus M AzoIIa G Betula M Calluna M Cary a G Castanea G Corylus Coryloid indet. Cupressaceae M Cyperaceae M Empetrum M Ericaceae M Fagus G Gleicheniidites senonicus - Gramineae M Juglans G Juniperus M Larix G Liriodendron - Liquidambar - Myrica M Nyssa - Osmunda regalis M Osmunda cf. claytoniana - Ostrya - Picea G Pinus (subgenus Diploxylon ) Pinus (subgenus Haploxylon ) H Polypodiaceae M Pterocarya - Quercus M Salix M Sequoia - Sphagnum M Taxodium - Tax us M Tsuga - Despite the uncertainties surrounding the dating of the Gortian, the data available on Irish interglacials suggest several trends. An obvious point is the progressive reduction in the diversity of the tree flora. Some of the species listed in Table 2 persisted until the mid to late Pleistocene, examples include Abies sp., Picea sp. and Pterocarya. From the relatively few Irish interglacial sites thought to be later than the Gortian, several other tree species are recorded which are no longer native, including Carpinus betulus, and Larix sp. Hence the history of the Pleistocene in Ireland is one of continued and progressive loss of tree taxa (Coxon & Waldren 1995, 1997). These species presumably failed to recolonise from refugia following full glacial periods, and the PHYTOGEOGRAPHY OF IRELAND 37 species generally seem to have retreated east (or been lost from Eurasia) during successive glaciations. The tree flora of Ireland today remains impoverished, even compared to that of Britain and the adjacent continent. In contrast to the trees, many of the herbs and shrubs with present day Atlantic distributions that have been recorded in the Irish interglacial floras are still present. Coxon & Waldren (1995) listed 34 taxa with modem Atlantic or amphi-Atlantic (sensu Hulten 1958) distributions recorded in Irish interglacial floras; among this number are several with extreme oceanic distributions in Europe that are absent from Britain, including Daboecia cantabrica , Erica mackaiana and Saxifraga section Gymnopera (i.e. probably S. spathularis or S. hirsuta). Some of the fossil determinations of these taxa are extremely difficult, and may be open to some doubt. Pollen of Ericaceae is notoriously difficult to identify to species, leaf macrofossils of Erica mackaiana may be confused with the more widespread E. tetralix, and seed of the Saxifraga species may also prove problematic in determination. Despite these difficulties, the identifications of leaf macrofossils of Daboecia and of Eriocaulon pollen and leaves are generally indisputable, and provide strong evidence for the long history of these taxa in Ireland. Even so, many doubts have been raised about the native status of some of these Atlantic species. Foss et al. (1987) suggested that the absence of fossil records of Erica erigena prior to approximately 500 b.p. and the possible trade and pilgrimage links between north-west Spain and Ireland provide evidence that this species at least is a Holocene introduction. Doubts have subsequently been raised about the native status of several other of this group of species, particularly those with disjunct modem distributions, including the Atlantic Daboecia cantabrica and the Atlantic/Mediterranean Arbutus unedo. However, lack of a fossil or sub-fossil record does not imply lack of a species occurrence - many of our commonest native species also lack fossil evidence for their historical existence in Ireland. Several possible theories therefore exist to explain the occurrence of these species in Ireland: • Deliberate or accidental human introduction in the Holocene • Repeated recolonisation from distant glacial refugia following full glacial cycles • Survival in situ or in more proximal refugia. For some of these species, deliberate human introduction seems highly unlikely, the aquatic Eriocaulon aquaticum provides a good example. Repeated introduction by migratory birds may be more likely, but in the case of Eriocaulon the presumed distant refugia would be North America, and Webb (1983) considered this repeated dispersal by birds as unlikely. Repeated colonization from distant refugia by other species has often been suggested as unlikely due to the time involved, but Preece (1997) has shown that even land molluscs, traditionally considered as somewhat sedentary, can show rapid dispersal and large shifts in distribution in response to Quaternary climate changes. Equally unlikely is in situ survival during full glacial cycles, but survival in proximal refugia is a possibility, perhaps in areas of continental shelf that are now submerged, but which may have been emergent during lower sea levels in full glacial periods. A difficulty here is lack of knowledge of rates of change of sea level, and of paleoclimates during glacial cycles. Were conditions during glacial cycles on the western seaboard of Ireland as highly oceanic as they are currently? Evidence suggests that even in northern Norway, markedly increased oceanicity occurs close to the coast (R. M. M. Crawford, pers. comm.). A second line of evidence comes from Eriocaulon. British and Irish material differs from at least the majority, and possibly all. North American material in chromosome number and certain morphological characteristics (Godwin 1975; Webb 1983). This implies a lengthy period of isolation between European and American populations, allowing sufficient time for evolutionary change. However, we again have problems in simply not knowing how much time is required to bring about any such change. Clearly, the Eriocaulon case deserves much closer study, in terms of both population differentiation and the taxonomic distinctness of the European material. If one taxon can be plausibly shown to have persisted through full glacial cycles in the western extremity of Europe, there is no good reason to doubt that other species may also have survived in similar locations. Modem molecular techniques may help to shed more light on these uncertain cases (e.g. Kingston & Waldren 2005), but as yet this approach has not been applied to Eriocaulon. The Eriocaulon example referred to above provides an extreme case. There are other taxa whose modem distribution centres on the British Isles, although they are more widespread there than Eriocaulon. Examples such as Dryopteris aemula and Hymenophyllum wilsonii were mentioned 38 ATLANTIC ARC earlier, though it is possible that these ferns, with a relatively high potential for long distance dispersal, shifted to their current distributions post-glacially from more southern or south-western refugia. Their current distribution would therefore reflect ecological conditions favourable for their growth. D. aemula currently occurs in the northern Iberian peninsula, Brittany and Macaronesia (Jalas et al. 1972) and the species may have recolonised the British Isles from these regions, tracking climatic change. Less certain is H. wilsonii , which is absent from Iberia but occurs sparsely in Brittany and western Norway (Jalas et al. 1972), though the highly disjunct distribution of the related and still predominantly Atlantic H. tunbridgense may suggest fragmentation of a former wider range. But even less certain is the case of Saxifraga hypnoides , restricted to the British Isles, Iceland and very rare in western Norway (Webb & Gomall 1989). Exactly where did it occur during the last full glacial? Related species occur in the mountains of northern Iberia, but it seems unlikely that S. hypnoides was ousted from any such more southern refugium by these species. More likely is that the taxon has either relatively recently evolved (i.e. in the Holocene) from related European taxa, or that it has survived the last glaciation somewhere within a limited European range. A detailed molecular study of population genetic diversity and the evolutionary lineage of this taxon may well prove rewarding. CONCLUSIONS The current distributions of Irish plants have been governed by a wide range of factors, including climate changes, ecological tolerances, anthropogenic effects, and the historical range of the taxa concerned during and subsequent to the Pleistocene. Clearly, the location of glacial refugia, the dispersal of taxa from these refugia during subsequent warm periods, and the possible post-glacial fragmentation of ranges through local extinction are the major biogeographical problems with the Irish flora. As yet there are insufficient data to determine which of the hypotheses mentioned above is likely to be the most relevant. As Webb stated (1983), in some cases all of these hypotheses - long distance colonisation from distant refugia, survival in proximal refugia, deliberate or accidental human introduction - seem unlikely, particularly where repeated biogeographic patterns appear to occur in different interglacial periods, as in the case of Eriocaulon aquaticum. The present day distributions of several taxa suggest that survival during glacial periods in Ireland or on land now submerged offshore may be the least unlikely of the hypotheses presented. However, considerable biological data (population structure, genetic diversity and migration pathways) and geographical information (sea level changes, palaeo- climates) are required to test these hypotheses further. We have recently begun investigations on the genetic diversity and biogeographic affinities of Irish populations of several species using molecular tools (see Kingston & Waldren 2005; Smith & Waldren 2005) and we hope that these studies may shed further light on the origins of the Irish flora. REFERENCES Bennett, K. D. (1995). Insularity and the Quaternary tree and shrub flora of the British Isles, in Preece, R. C. ed. Island Britain: a Quaternary perspective. Geological Society Special Publication no. 96, pp. 173— 180. COXON, P. (1993). Irish Pleistocene biostratigraphy. Irish Journal of Earth Science, 12: 83-105. Coxon, P. (2001). Cenozoic: Tertiary and Quaternary (until 10,000 years before present), in Holland,C. H. The Geology of Ireland. Dunedin Academic Press, Edinburgh, pp. 387^-27. Coxon, P. & Flegg, A. M. (1987). A Late Pliocene/Early Pleistocene deposit at Pollnahallia, near Headford, Co. Galway. Proceedings of the Royal Irish Academy 87B: 1 5^42. Coxon, P., McMorrow, S. & COXON, C., (1998). Pollnahallia: a tertiary palaeosurface and a glimpse of Ireland’s pre-Pleistocene landscape, in BARTON, K. & MOLLY, K. eds., South Central Mayo. Field Guide No. 22. Irish Association for Quaternary Studies (IQUA), Dublin, pp. 63-70. COXON, P. & Waldren, S. (1995). The floristic record of Ireland’s Pleistocene temperate stages, in PREECE, R. C. ed. Island Britain: a Quaternary perspective. Geological Society Special Pub. no. 96, pp. 243-268. Coxon, P. & Waldren, S. (1997). Flora and vegetation of the Quaternary temperate stages of NW Europe: Evidence for large scale range changes, in Huntley B., Cramer W., Morgan A. V., Prentice H. C. & ALLEN J.R.M. eds. Past and Future Rapid Environmental Changes: The Spatial and Evolutionary Responses of Terrestrial Biota , pp. 104-117. NATO ASI Series I: Global Environmental Change, vol. 47. Springer-Verlag, Berlin. PHYTOGEOGRAPHY OF IRELAND 39 Crawford, R. M. M. (2000). Ecological hazards of oceanic environments. New Phytologist 147: 257-281. Crawford, R. M. M„ Jeffree, C. E & Rees, W. G. (2003). Paludification and forest retreat in northern oceanic environments. Annals of Botany 91: 213-226. Dowling, L. A. & COXON, P. (2001). Current understanding of the Pleistocene temperate stages in Ireland. Quaternary Science Reviews 20: 1631-1642. FOSS, P. J., DOYLE, G. J. & Nelson, E. C. (1987). The distribution of Erica erigena R. Ross in Ireland. Watsonia 16: 311-327. GODWIN, H. (1975). The History of the British Flora. 2nd edn. Cambridge University Press, Cambridge. HULTEN, E. (1958). The amphi -Atlantic plants and their phytogeographical connections. Kungl Svenska Vtenskapsakademiensis Handlingar Series 4: 1-340. JALAS, J & SUOMINEN, J. (1972). Atlas Florae Europaeae, I. Pteridophyta. Cambridge University Press, Cambridge. Jessen, K„ Andersen, S. T. & Farrington, A. (1959). The interglacial deposit near Gort, Co. Galway. Proceedings of the Royal Irish Academy 60B: 1-77. Kingston, N. & Waldren, S. (2005). Biogeography of the Irish ‘Lusitaniam heathers, in Leach, S. J., Page, C. N., Peytoreau, Y. & Sanford, M. N. eds. Botanical Finks in the Atlantic Arc , pp. 147-156. Botanical Society of the British Isles, London. Preece, R. C. (1997). The spatial response of non-marine Mollusca to past climate change, in Huntley B., Cramer W., Morgan A. V., Prentice H. C. & Allen J.R.M. eds. Past and Future Rapid Environmental Changes: The Spatial and Evolutionary Responses of Terrestrial Biota , pp. 163-177. NATO ASI Series I: Global Environmental Change, vol 47. Springer- Verlag, Berlin. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Biological Journal of the Linnean Society 124: 1-120. PRESTON, C. D., Pearman, D. A. & Dines, T. D. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. Smith, R. J. & Waldren, S. (2005). Genetic variation in Irish threatened plant species: a European perspective, in Leach, S. J., Page, C. N„ Peytoreau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc, pp. 137-145. Botanical Society of the British Isles, London. Webb, D. A. (1983). The flora of Ireland in its European context. Journal of Life Sciences, Royal Dublin Society 4: 143-160. Webb, D. A. & Gornall, R. J. (1989). Saxifrages of Europe. Christopher Helm, London. WCMC (1994). Biodiversity Data Sourcebook. World Conservation Press, Cambridge. 41 The Mediterranean-Atlantic and Atlantic elements in the Cornish flora C. D. PRESTON & H. R. ARNOLD C EH Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE28 2LS, England ABSTRACT In Britain species with a Mediterranean-Atlantic distribution are concentrated in the south and those with an Atlantic (Oceanic) distribution are concentrated in the west. Species in both elements are therefore well represented in Cornwall, the most south-westerly of the British counties, where 70% of the British Mediterranean-Atlantic vascular plants and 70% of the Oceanic taxa occur as natives. The majority of Mediterranean-Atlantic species in the county are plants of coastal habitats, but the group also includes some species of other draughted or disturbed soils. This relatively narrow range of habitats contrasts with the wide range of habitats in which Oceanic species are found in the county. The population size of several annual Mediterranean-Atlantic species fluctuates widely. The most favourable conditions are produced by summer droughts followed by mild winters and springs which are sunny but not too dry; in such years the population size of species such as Trifolium incarnation may be 20 times that in less favourable seasons. Some species may also respond positively to fire and to limited human disturbance. In addition to the native species, a suite of species native to the Mediterranean region has become naturalised in Cornwall, and the South African succulent Carpobrotus edulis has attained a Mediterranean-Atlantic distribution in Europe. KEYWORDS: Cornwall, vascular flora, Mediterranean-Atlantic elements, ecological characteristics. INTRODUCTION Cornwall has long been renowned for its notable flora. At its heart, metaphorically if not geographically, is the Lizard peninsula, to which most British botanists will no doubt be drawn at some stage in their life. Along with Breckland, Upper Teesdale, Ben Lawers and the Burren, The Lizard is one of a quintet of botanical localities in Britain and Ireland which possess a special romance derived from concentrations of rare species and a long tradition of botanical pilgrimage. However, the attraction of the Cornish flora does not lie in its sheer diversity - there are many areas of England and Wales where the density of native species is matched or exceeded (Preston, Pearman & Dines 2002, p. 28). It rests, rather, in a rich assemblage of species with restricted distributions in the British Isles. The purpose of this paper is to examine the distribution and ecology of two groups of species which contribute towards this assemblage, the Mediterranean- Atlantic and Atlantic elements. How well represented are the native members of these groups in Cornwall, and what are their ecological characteristics? To what extent have these native plants been joined by alien species of Mediterranean affinities? DEFINITION OF THE MEDITERRANEAN-ATLANTIC AND ATLANTIC ELEMENTS The Mediterranean-Atlantic element comprises plants which are rather strictly confined to the Mediterranean region, extending northwards only along the Atlantic fringe of Europe (Preston & Hill 1997). Umbilicus rupestris provides an example (Fig. 1). Preston, Pearman & Dines (2002) accept 57 species as British natives, including two extinct species, Euphorbia peplis and Otanthus maritimus. There are in addition three doubtfully native Mediterranean-Atlantic plants, Matthiola sinuata, Polycarpon tetraphyllum and Serapias parviflora. To these 60 species we have added one well-marked subspecies, Raphanus raphanistrum subsp. maritimus. These 61 taxa are the group analysed in this paper, we have not included the four Mediterranean-Atlantic species restricted to the Channel Islands {Ano gramma leptophylla. Milium vemale, Myosotis sicula , Ranunculus 42 ATLANTIC ARC 18* E SC£ — / / - 40° • v • • 30 w Figure 1. The distribution in Europe of a species with a typical Mediterranean- Atlantic distribution. Umbilicus rupestris. The distribution is mapped in 50 x 50 km UTM grid squares (Jalas et al. 1999). paludosus ) and three found as natives only in Ireland {Arbutus unedo, Asplenium onopteris, Neotinea maculata, the latter formerly recorded in addition from the Isle of Man). The Atlantic species are those described as ‘Oceanic’ by Preston & Hill (1997). They are confined, in Europe, to the Atlantic zone and do not extend into the Mediterranean region, or are very rare there (Fig. 2). Unlike the Mediterranean-Atlantic plants, which by definition have predominantly southerly distributions, the Atlantic taxa range from plants which have northern distributions in the Oceanic Boreo-arctic Montane or Boreal-montane elements (e.g. Euphrasia ostenfeldii, Saxifraga hypnoides) to those with southerly ranges which are classified in the Oceanic Southern-temperate element (e.g. Agrostis curtisii , Herniaria ciliolata). However, the majority have Oceanic Temperate ranges, a typical example being Spergularia rupicola (Fig. 3). Some are endemic to Europe but others, although confined to the Atlantic zone within Europe, also occur in other continents. Preston, Pearman & Dines (2002) accept 77 native species in the ‘Oceanic’ elements, to which Asparagus prostratus (Plate 28) should be added following its recent restoration to specific status (Kay, Davies & Rich 2001). To these we have added two doubtfully native species ( Festuca lemanii, Pyrus cordata ) to give a total of 80 species analysed in this paper. There are in addition four Atlantic species restricted to the Channel Islands {Festuca armoricana, F. huonii, Limonium auriculae-ursifolium, L. normannicum ) and seven found as natives only in Ireland {Daboecia cantabrica. Erica erigena , E. mackaiana, Pinguicula grandiflora , Saxifraga hirsuta , S. spathularis, Sisyrinchium bermudiana). ATLANTIC ELEMENTS IN THE CORNISH FLORA 43 Figure 2. The Atlantic (A) and Subatlantic (B) zones of Europe (from Preston & Hill 1997). The two elements considered in this paper are, of course, separated rather arbitrarily from species with rather wider distributions. Submediterranean-subatlantic and Suboceanic (subatlantic) species have rather wider distributions than those in the Mediterranean-Atlantic and Atlantic elements respectively (Fig. 2). In addition there is an interesting group of species which are intermediate between the Mediterranean-Atlantic and Atlantic species, as they are found along the Atlantic seaboard and in the western (but not eastern) Mediterranean. This group, included in the Suboceanic Southern-temperate element by Preston & Hill (1997), include species such as Asplenium marinum, Oenanthe lachenalii and Parapholis strigosa. Species endemic to the British Isles are not considered in this paper. However, some endemics have distributions with clear affinities to rather more widespread species in the Atlantic element, including Fumaria occidentalis and several Euphrasia and Limonium species. DISTRIBUTION IN THE BRITISH ISLES AND CORNWALL The number of Mediterranean-Atlantic and Atlantic species in the 10-km squares of the British Isles are mapped in Figs 4 and 5. The Mediterranean-Atlantic taxa are concentrated along the coasts of southern England and Wales, whereas the Atlantic taxa are concentrated in western England and Wales, becoming rather less numerous in western Scotland. Cornwall, the most south-westerly of British counties, lies in the area where these distributions overlap, and both elements are very well represented. A similar proportion of each group is native to the county: 43 of the 61 British Mediterranean-Atlantic plants (70%) and 56 of the 80 Atlantic species (70%). 44 ATLANTIC ARC FIGURE 3. The distribution in Europe of a species with a typical Atlantic distribution, Spergularia rupicola. The distribution is mapped in 50 x 50 km UTM grid squares (Jalas & Suominen 1983). The absence from southern France is typical of many such species with Temperate distributions. In Cornwall, as in Britain as a whole, the Mediterranean-Atlantic and Atlantic elements are concentrated in different areas. The Mediterranean-Atlantic species are coastal, even at the tetrad scale (Fig. 7). There are particular concentrations in the Isles of Scilly, The Lizard and at the Hayle estuary and Newquay on the north coast. By contrast, the Atlantic species are much more evenly spread (Fig. 8), although the Lizard peninsula is again a hotspot, as is an area N.W. of St Austell. ATLANTIC ELEMENTS IN THE CORNISH FLORA 45 Figure 4. The number of the 61 British Mediterranean- Atlantic species in the 10-km squares of the British Isles. Based on native records made from 1970 onwards and published by Preston, Pearman & Dines (2002). In addition to the Cornish natives, two Mediterranean-Atlantic species {Arbutus unedo, Vicia bithynica ) and two Atlantic species ( Daboecia cantabhca, Sedum forsterianum) are naturalised in Cornwall but native elsewhere in the British Isles. Only S. forsterianum is at all frequent in the county; the others are rare (French, Murphy & Atkinson 1999). Most of the 17 Mediterranean-Atlantic species which are native to Britain but are neither native nor established in Cornwall are coastal species which are relatively widespread in S.E. England (e.g. Hordeum marinum , Parapholis incurva, Sarcocornia perennis ); others are plants with very restricted distributions and ecological niches in Britain (e.g. Bupleurum baldense, Centaurium tenuiflorum, Galium constriction) . Only four occur in Ireland. The most surprising of these 17 absentees is perhaps Tuberaria guttata , which occurs in the Channel Islands, North Wales and western Ireland and seems unaccountably missing from Cornwall. 46 ATLANTIC ARC >••••••••••* ••••• •• • < •«•••••»* ••••••• ••**•4 »••••••< . )***•••»< »H( >••••••••4 _ , »•«•«••* •••••••« (••••••••••••••a l >•••*••••••«••••! >••«••••••••*•••••< F«*«* *••••• • • a • »••••••••••••«••••< »•••••••••«•••••••< >••••••••• •••••• 4 (*••••••••••••••• 4 »••*•••••••••••« «4I**« •« )•••«•»«*••#•••»***•»• < (••••••••••••••••••a •••< IIIMHIttlHMallMMa • a < >•••••••••• a •• a a a a a • a i l«ff*4****»M*«t*t*»IM •••« >•*••«•••••••» a altaatK MtM«IM»Ht»«aM•••••• •«••• a • a a • a • a 4 !«•••••* •••••••••••< .. JMHMMtaattaiiall »«•••••••• a ••••••• a >••••••« a •••••••• a a >•#••••••••••••••« a • a ( •••4 «••••••••! • ••«•••••! • ••••• a afT* ••••••••4 ••••••< a a • a a 4 **••##( •••J ^ m _ aaaialatIMliattIMIMI aallatMaltlHaHMi _ a ••• allMtta •«•••••< »»«>ataaaM>MM>MM aa< _ ••••« . _ HHf*M4iaa|MMl«IMHtaaM*»lta«i •••#*•••••••••• a •••••••••••••••«•••« *•»««•*••• •••••4 - «M***iM«f»*M*i MNTa FIGURE 5. The number of the 80 British Atlantic species in the 10-km squares of the British Isles. Based on native records made from 1970 onwards and published by Preston, Pearman & Dines (2002). The 24 Atlantic species present as natives in Britain but absent from Cornwall include 10 Boreo- arctic, Boreal or B oreo-temperate species (e.g. Dactylorhiza purpurella , Euphrasia foulaensis , Myosotis stolonifera). Many of the Temperate and Southern-temperate absentees are species with predominantly eastern distributions in England that contrast with their restriction to the Atlantic zone of Europe (Oenanthe fluviatilis, Salicomia pusilla , Thesium humifusum). Festuca longifolia is present in Devon and the Channel Islands, and is perhaps worth looking for in Cornwall. Nine of the 24 absentees occur in Ireland. ATLANTIC ELEMENTS IN THE CORNISH FLORA 47 Figure 6. The number of the 141 British Mediterranean- Atlantic and Atlantic species in the 10-km squares of the British Isles. Based on native records made from 1970 onwards and published by Preston, Pearman & Dines (2002). ECOLOGICAL CHARACTERISTICS LIFE-FORM Although there is a significant representation of annual species in the Cornish Mediterranean- Atlantic element, most of the species are perennials. Of the 43 native species, 16 (37%) are annuals, 4 (9%) are usually biennials (sometimes behaving as perennials) and 23 (54%) are usually perennial (Table 1). The Atlantic element includes only 10 annuals (18%) and 2 (3%) normally biennial species compared to 44 (79%) perennials (Table 2). 48 ATLANTIC ARC TABLE 1. THE HABITAT OF MEDITERRANEAN-ATLANTIC SPECIES IN CORNWALL Rocks, banks, hedges and scrub Arum italicum Polypodium cambricum Asplenium obovatum Rubia peregrina Geranium purpureum* Umbilicus rupestris Coastal habitats - sandy beaches Calystegia soldanella Matthiola sinuata f Euphorbia paralias Otanthus maritimus Euphorbia peplis* Polygonum maritimum Glaucium flavum Vulpia fasciculata * Coastal habitats - consolidated dunes Juncus acutus Salvia verbenacea Coastal habitats - salt marshes and sea-cliffs Adiantum capillus-veneris Inula crithmoides Atriplex portulacoides Lavatera arboreal Beta vulgaris Raphanus raphanistrum subsp. Crithmum maritimum maritimus f Grassland, especially in coastal sites Oenanthe pimpinelloides Serapias par\’iflora Coastal habitats - summer-droughted soils near the sea Catapodium marinum* Trifolium bocconei* Isoetes histrix T. glomeratum* Ophioglossum lusitanicum T. incarnatum * Romulea columnae T. suffocatum* Scilla autumnalis Disturbed, sometimes winter-wet and/or summer-droughted soils Fumaria bastardii* Parentucellia viscosa* Gastridium ventricosum * Poa infirma* lsolepis cemua * Polycarpon tetraphyllum * Juncus pygmaeus* Torilis nodosa * Linum bienne f Annuals are asterisked (*) and biennials marked with a dagger (t) HABITAT REQUIREMENTS The habitat of the Mediterranean-Atlantic and Atlantic species in Cornwall are set out in Tables 1 and 2. The majority of the Mediterranean-Atlantic plants are found in a rather narrow range of habitats, which are predominantly coastal with a smaller group of species of draughted or disturbed soils. By contrast, the Atlantic species are remarkably heterogeneous: the coastal species, although the largest single group, constitute only a minority of the element. The differences between the ecological requirements of the two groups of species was first commented on by Stapf (1914, 1917), who was the first to recognise the difference between the “Atlantic” and “Mediterranean” components of the British flora. He described “a deep and very general contrast between the hygrophilous species which inhabit bogs or boggy places, wet meadows or wet rocks and heaths on one side, and the more xerophilous tenants of light, open soil” (Stapf 1917). ATLANTIC ELEMENTS IN THE CORNISH FLORA 49 TABLE 2. THE HABITAT OF ATLANTIC SPECIES IN CORNWALL Woodland, scrub, hedges and shaded rocks Care x laevigata Meconopsis cambrica Ceratocapnos claviculata* Pyrus cordata Dryopteris aemula Sibthorpia europaea Hyacinthoides non-scripta Trichomanes speciosum Hymenophyllum tunbrigense Hymenopyllum wilsonii Ulex europaeus Heathland Agrostis curtisii Euphrasia confusa* Carex binervis Genista anglica Descliampsia setacea Lobelia urens Erica ci Haris Ulex gallii E. cinerea Viola lactea E. vagans Wahlenbergia hederacea Grassland, sometimes also heathland Conopodium majus Euphrasia arctica* Euphrasia tetraquetra* Damp grassland and wetland Anagallis tenella Hypericum elodes Carum verticillatum H. undulatum Cirsium dissectum Myosotis secunda Dactylorhiza praetermissa Narthecium ossifragum Eleogiton fluitans Pinguicula lusitanica Coastal habitats Asparagus prostratus Lepidium heterophyllumf Atriplex laciniata* Puccinellia maritima Centaurium scilloides P. rupestris* Cochlearia anglica't Rumex rupestris C. danica* Scilla verna Euphorbia portlandica Spergularia rupicola Eestuca arenaria Trifolium occidental Summer-droughted soils Herniaria ciliolata Hypericum linariifolium Sedum anglicum Disturbed, sometimes winter-wet and/or summer-droughted soils Fumaria muralis * Fumaria purpurea * Ranunculus tripartitus* Annuals are asterisked (*) and biennals marked with a dagger (f) 50 ATLANTIC ARC Figure 7 The number of the 43 Mediterranean-Atlantic species native to Cornwall in the tetrads (2x2 km squares) of the county, based on native records made from 1970 onwards. Based primarily on data collected for French et ah (1999). Figure 8. The number of the 56 Atlantic species native to Cornwall in the tetrads (2x2 km squares) of the county, based on native records made from 1970 onwards. Based primarily on data collected for French et ah (1999). ATLANTIC ELEMENTS IN THE CORNISH FLORA 51 Figure 9. The number of 99 Mediterranean- Atlantic and Atlantic species native to Cornwall in the tetrads (2 x 2 km squares) of the county, based on native records made from 1970 onwards. Based primarily on data collected for French et al. (1999). POPULATION FLUCTUATIONS Many of the Mediterranean-Atlantic species in Cornwall are at or near the northern edge of their range. Climatic extremes such as extremely hot, dry summers or cold winters have marked effects on the populations of some of these species, and provide insight into the ecological factors which enable these species to grow in Cornwall, so far from their core range, but restrict their distribution elsewhere. Much of our knowledge of these fluctuations results from the detailed studies of the Lizard flora carried out by D. E. Coombe and L. C. Frost from 1950 until the 1990s, in collaboration with M. H. Martin and a number of research students. The best summary of this aspect of their work is unfortunately in an unpublished report (Frost et al. 1983). EXTREME DROUGHT The fluctuation in the numbers of the rare Lizard winter- annual Trifolium incarnation subsp. molinerii (Plate 18) has been documented in detail by Martin & Frost (1980). This plant is confined to open cliff slopes and cliff edges where the soil is sufficiently deep to sustain the plant during spring droughts but kept open by summer droughts which prevent the formation of a closed perennial sward. Even this limited niche does not always provide favourable conditions for the species, and the total number of plants varies from year to year by a factor of over 20. The species is favoured by warm, moist autumns (which promote seedling establishment), mild winters, and springs and early summers which are sunny but not too dry. Exceptionally hot summers kill competitive species, especially perennial grasses, and allow the annual clover to exploit the resulting areas of bare soil or areas covered only by dead plant remains. In subsequent seasons the perennial grasses and herbs gradually recover, and the rare annuals decline in abundance. The best year for Trifolium incarnation recorded by Martin & Frost (1980) was 1977, following the summer droughts of 1975 and 1976. They counted 36,614 plants in the 9 populations known that year, a number which fell to 5,669 in 1978 and 1,558 in 1979. 52 ATLANTIC ARC FIGURE 10. The number of 55 alien species of Mediterranean origin in the 10-km squares of the British Isles, based on records made from 1970 onwards. Species qualifying for inclusion are those recorded from at least 50 10-km squares in the British Isles, and therefore mapped by Preston, Pearman & Dines (2002). Similar ‘Coombe/Frost cycles’ characterise the ecology of other species in Cornwall, including the Mediterranean- Atlantic Trifolium bocconei and Vicia lutea, the Submediterranean-subatlantic Trifolium scabrum, T. strictum and T. subterraneum and the Southern-temperate Juncus capitatus, Lotus angustissimus (Plate 34) and L. subbiflorus. Other more widespread Temperate annuals show the same pattern, including Cerastium diffusion , Myosotis discolor and Trifolium dubium (Frost et al. 1983; Martin & Frost 1980; Stewart et al. 1994; Wigginton 1999). In the legumes hot summer temperatures not only knock back competitive species but probably also stimulate the dormant seed to germinate by their effect on the hard seed coat. In the Avon Gorge, Bristol, the same cycle has been documented for the annual grass Gastridium ventricosum (Lovatt 1981), which historically has tended to be recorded in years following hot summers and also benefited greatly from the summer droughts of 1975 and 1976 followed by wet autumns. ATLANTIC ELEMENTS IN THE CORNISH FLORA 53 Many Mediterranean-Atlantic perennials, by contrast, are notably drought tolerant: these include geophytes such as Isoetes histrix, Ophioglossum lusitanicum, Romulea columnae and Scilla autumnalis, as well as slightly succulent plants of coastal rocks such as Crithmum maritimum and Inula crithmoides. It is not only the adult plants of I. histrix which are drought tolerant - the spores have also survived 34 years’ dry storage (Wigginton 1999). The cycles of the Mediterranean-Atlantic annuals contrast with the fluctuations in abundance of the more northern, Boreo-temperate legume Anthyllis vulneraria, which is conspicuously abundant in some years but much less abundant in others on both south and north Cornish coasts (Plate 47). The good years for this species do not follow drought years; indeed, Martin & Frost (1980) suggest that there is less T. incamatum than usual in good years for Anthyllis. The reasons for the seasonal fluctuations in this species require further research. Although population fluctuations are less apparent in the predominantly perennial Atlantic flora, they are a feature of the ecology of at least one species, Trifolium occidental . Although perennial, this is a plant of relatively shallow, drought-prone soils in coastal sites, and often in places exposed to salt spray. It is absent from deeper, less exposed sites where the closely related T. repens is frequent (Coombe 1961). Despite its relative drought-tolerance, most plants of T. occidental e at the Lizard were killed in the exceptionally hot summers of 1959 (Coombe 1961) and 1975-1976. In February and June 1977 C. Preston could only find one plant of T. occidental which had survived these droughts at Caerthillian Cove, Lizard, although he spent over three hours searching there. Seedlings were, however, frequent by February and some even had one or two inflorescences by June. In contrast to T. occidental, other Atlantic perennials are extremely drought-tolerant, including the succulent Sedum anglicum and the less obviously xeromorphic Hemiaria ciliolata. FIRE Some Mediterranean-Atlantic annuals and perennials may be favoured by fire. In the early 1980s we recorded the gradual demise of a subpopulation of Vicia lutea at Church Cove, Landewednack, which appeared in abundance in 1979 in an area of burnt gorse scrub and gradually declined as the bushes regenerated from the burnt stumps and from seed. A similar efflorescence and subsequent decline after burning has been observed for the Southern-temperate Ornithopus pinnatus in Scilly (Wigginton 1999). Deliberate burning of Ulex europaeus, and accidental fires, provide temporarily open sites in which Trifolium bocconei and T. incamatum can grow (Martin & Frost 1980; Wigginton 1991). There is a clear parallel between the effects of summer drought and those of burning on these species. It is interesting to speculate whether fire had any place in their ecology before man and grazing animals became such a dominant component of the ecosystem. Heath and scrub fires may destroy shallow rooted plants but leave geophytes unaffected, and able to benefit from the temporary lack of competition. On the small islet of Hommet Benes, Guernsey, in April 1979 Isoetes histrix, Romulea columnae and Scilla autumnalis were abundant in a 15 m2 area of burnt Ulex europaeus scrub; other species were very infrequent. Ryan (1989) counted over 1,000 I. histrix plants on the islet in 1979, a population which, as the burnt area became recolonised by gorse and coarse grasses, fell to 40 plants in 1988. The Atlantic geophyte Scilla verna may also benefit in similar manner from the controlled or uncontrolled burning of coastal heath: in May 2003 a large area of burnt heath on the coast of Holy Island, Anglesey, was coloured pale, slate blue by the massed flowers of this species and visible for hundreds of metres; plants were both more numerous and more vigorous than in adjacent areas of unbumt heathland. WINTER COLD Although many Mediterranean-Atlantic and Atlantic species occur in areas of mild winters, the direct effects of winter cold are not often as obvious in the field as those of summer drought. However, there is experimental evidence to show that the Mediterranean-Atlantic biennial Lavatera arborea is particularly susceptible to winter cold: Okusanya (1979) found that 60% of the seedings he studied were killed by 1 hour’s exposure to a temperature of -5°C and all of them died after 3 hours’ exposure; 50% of young plants died after 3 hours at this temperature. Seedlings of the Mediterranean-Atlantic Inula crithmoides and the Atlantic Spergularia rupicola also suffered 50-60% mortality after 3 hours at -5°C, although those of the related but more widespread Inula conyzae and S. rubra were unaffected. Frost damage to Lavatera arborea can 54 ATLANTIC ARC sometimes be seen in the field: most plants on the cliff-top at the most southerly point of the Lizard were killed outright in the 1978/79 winter, for example, although some plants further down the cliff survived. It would be interesting to examine populations of the other susceptible species after particularly cold winters. Winter-browning of Ulex europcieus bushes can often be observed after cold winters, but I have the impression that the woody bushes of this Atlantic species usually recover. In many cases, of course, the effects of temperature in limiting the distribution of Mediterranean-Atlantic species are likely to be more subtle than simply those of direct mortality from winter cold. Crithmum maritimum and Inula crithmoides, for example, both fruit late in the season (September-November) in Britain, and in cooler summers may fail to set seed at the northern fringe of their range. RESPONSE TO HUMAN DISTURBANCE The ecological requirements of the Mediterranean-Atlantic annuals, with their growth period in autumn, winter and early spring followed by a requirement for the disruption of the perennial vegetation in high summer, might ‘pre-adapt’ them for the pressures of human disturbance, including those resulting from tourism, on the Cornish coast. It is perhaps not surprising that the Mediterranean-Atlantic species in the British flora did relatively well in the 20th century (Preston, Pearman & Dines 2002, p. 41). Martin & Frost (1980) note that Trifolium incarnation can grow along the mown or disturbed sides of maintained paths, or on unofficial subsidiary paths which are used chiefly in summer. ,In such places it may even grow in sites with a northerly aspect, something it is unable to do in less disturbed vegetation. Poa infirma is a more characteristic species of trampled places and now appears to be spreading rapidly in Britain. The arrival in Cornwall in 1988 of the much more conspicuous (although even smaller) Submediterranean- subatlantic Crassula tillaea , a species of disturbed, compacted and often shallowly winter-flooded sites, was one of the first indications of a major expansion of its national range. Juncus pygmaeus is a species which appears to be completely dependent on the disturbance it receives on ancient trackways, although it does not appear to be able to colonise new localities in the same way that C. tillaea has done. Clearly climate change may be playing a role in these changes - tourist disturbance in Cornwall must, after all, have been a feature of the county for several decades. One group of Mediterranean species which appear to have declined in Cornwall are the Mediterranean-Atlantic perennials of sandy beaches. All three of the Mediterranean-Atlantic species which have become extinct in Cornwall (and in two cases nationally), Euphorbia peplis , Matthiola sinuata and Otanthus maritimus, belong to this group, and a fourth member, Glaucium flavum, has declined in Cornwall. E. peplis and O. maritimus can be seen at their most vigorous on the upper, rather undisturbed sands of the tideless Mediterranean, and are perhaps less suited to life on the stormy Atlantic coast. (The only surviving site for O. maritimus in Ireland is an atypically large expanse of coastal sand.) Human disturbance of sandy beaches may also have played a major part in their decline. It is possible that the extinct species may recolonise Cornwall from sea-borne propagules, just as Polygonum maritimum , another member of this group, appears to have spread along the south coast of England in recent years. ALIEN SPECIES OF MEDITERRANEAN ORIGIN There can be few parts of the British Isles where alien species contribute more to the sense of place than S.W. England. The long history of sea-faring trade, and more importantly the more recent importation of plants from abroad for the so-called ‘subtropical’ gardens of Cornwall, have led to the establishment of a rather rich and characteristic alien flora. Many of these naturalised species are perhaps appreciated more keenly by ordinary tourists than by botanists. As David McClintock wrote of Allium triquetrum , “People who come in July will see none of it, for soon after it has finished flowering it vanishes ... but earlier visitors should not fail to savour the chintzy charm of its waxen bells” (Bichard & McClintock 1975). Some 55 alien species of Mediterranean origin have been recorded in at least 50 10-km squares in the British Isles, of which 48 are ATLANTIC ELEMENTS IN THE CORNISH FLORA * 55 Figure 11. The number of the 48 alien species of Mediterranean origin listed in Table 3 in the tetrads (2x2 km squares) of Cornwall, based on records made from 1970 onwards. Based primarily on data collected for French et al. (1999). TABLE 3. SPECIES ORIGINATING IN THE LOWLAND MEDITERRANEAN REGION ESTABLISED IN CORNWALL Acanthus mollis Lavatera cretica Alliwn roseum Linaria purpurea Allium triquetrum Lobularia maritima Anisantha diandra Malcolmia maritima Anisantha madritensis Matthiola incana Anisantha rigida Mentha requienii Antirrhinum majus Nigella damascena Arenaria balearica Petasites fragrans Arum italicum subsp. italicum Phalaris minor Aubrieta deltoidea Phalaris paradoxa Borago officinalis Pinus pinaster Briza maxima Quercus ilex Bupleurum subovatum Rapistrum rugosum Calystegia silvatica Reseda alba Centranthus ruber Rosmarinus officinalis Cerastium tomentosum Salvia viridis Crepis vesicaria San tol ina chamaecyparissus Ficus carica Senecio cineraria Genista hispanica Soleirolia soleirolii Gladiolus communis Spergularia bocconei Hirschfeldia incana Tamarix gallica Hypericum hircinum Tragopogon porrifolius Lagurus ovatus Viburnum tinus Laurus nobilis Vinca major Additional species of Mediterranean origin naturalised in the British Isles but not in Cornwall: Euphorbia characias, Juncus subulatus, Lathyrus grandiflorus, Linaria pelisseriana, Malva parviflora, Phalaris aquatica, Spartium junceum. 56 ATLANTIC ARC established in Cornwall (Table 3). These include such widespread plants as Calystegia silvatica, Centranthus ruber , Cerastium tomentosum and Soleirolia soleirolii, as well as species with more south-western ranges in Britain, including Acanthus mollis , Allium roseum, A. triquetrum , Briza maxima , Laurus nobilis and Gladiolus communis. The coincidence map for these species (Fig. 10) shows that they are concentrated in the same areas as the native Mediterranean-Atlantic plants. However, their particular concentration on the south-west coast of England is more westerly than that of the native species, and the ‘hotspot’ in the London area is unmatched by the natives. In Cornwall (Fig. 11), there is a surprisingly good correspondence between the distributions of native and alien species of Mediterranean distributions, although the concentration of natives in the Lizard is not matched by a corresponding concentration of aliens. The Mediterranean aliens do not include many especially troublesome species, although the spread of Quercus ilex in some areas has caused concern. Allium triquetrum grows in dense masses on Cornish hedgebanks, although in no greater abundance than the native equivalents A. ursinum and Hyacinthoides non-scripta. It appears to grow in mixed stands or mosaics with other species rather than in large monospecific stands, perhaps because its competitive ability is limited by its short growing season. A much more competitive species is Carpobrotus edulis , a South African species which has assumed a convincingly Mediterranean-Atlantic range in Europe (Jalas & Suominen 1980). In the narrow coastal belt in which it grows it can form dense masses which override and exclude most native plants. It has become such a feature of the Cornish coast that it is not infrequently chosen to feature on postcard views. Like some of the native Mediterranean-Atlantic taxa, it is clearly limited by cold. Optimistic gardeners fail to grow it ‘up country’ and the winter of 1978/79 which killed many plants of Lavatera arborea at the Lizard also resulted in much damage to C. edulis. Severe spells of winter weather in January and February 1987 also killed (at least temporarily) large mats of the species on coastal cliffs in south-west England and the Channel Islands (Preston 1987). However, the current British climate is unlikely to provide severe winters sufficiently frequently to reduce to any great extent the populations of this resilient species. AC KNO WLEDGMENTS We thank the botanists who contributed to the Flora of Cornwall and New Atlas of the British and Irish Flora mapping projects, the results of which have been used to generate most of the figures. The late David Coombe introduced C.D.P. to the Cornish flora, and to many of the ideas discussed in this paper. REFERENCES BlCHARD. J. D. & Mcclintock, D. (1975). Wild flowers of the Channel Islands. Chatto & Windus, London. COOMBE, D. E. (1961). Trifolium occidental , a new species related to T. repens L. Watsonia 5: 68-87. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. FROST, L. C., Martin, M. H. & Coombe, D. E. (1983). The effects of the cessation of grazing and of drought on the grasslands in the Caerthillian valley. University of Bristol Lizard Project report no. 9. The University, Bristol. HILL, M. O., Mountford, J. O., Roy, D. B. & Bunce, R. G. H. (1999). Ellenberg' s indicator values for British plants. Institute of Terrestrial Ecology, Huntingdon. JALAS, J. & SUOMINEN, J., eds. (1980). Atlas Florae Europaeae, 5. Chenopodiaceae to Basellaceae. Committee for Mapping the Flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. JALAS, J. & SUOMINEN, J., eds. (1983). Atlas Florae Europaeae, 6. Caryophyllaceae (Alsinoideae and Paronychioideae). Committee for Mapping the Flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. JALAS, J., SUOMINEN, J., Lampinen, R. & Kurtto, A., eds. (1999). Atlas Florae Europaeae, 12. Resedaceae to Platanaceae. Committee for Mapping the Flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. Kay, Q. O. N., Davies, E. W. & Rich, T. C. G. (2001). Taxonomy of the western European endemic Asparagus prostratus (A. officinalis subsp. prostratus ) (Asparagaceae). Botanical Journal of the Linnean Society 137: 127-137. Lovatt, C. M. (1981). The history, ecology and status of Gastridium ventricosum (Gouan) Schinz & Thell. in the Avon Gorge. Bristol. Watsonia 13: 287-298. ATLANTIC ELEMENTS IN THE CORNISH FLORA 57 MARTIN, M. H. & Frost, L. C. (1980). Autecological studies of Trifolium molinerii at the Lizard Peninsula, Cornwall. New Phytologist 86: 329-344. OKUSANYA, O. T. (1979). An experimental investigation into the ecology of some maritime cliff species IV. Cold sensitivity and competition studies. Journal of Ecology 67: 591-600. PRESTON, C. D. (1988). Hottentot Fig, Carpobrotus edulis (L.) N.E.Br., in Guernsey. Report and Transactions La Societe Guernesiaise 22: 296-302. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal oftheLinnean Society 124: 1-120. PRESTON, C. D., Pearman, D. A. & Dines, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. Ryan, P. (1989). Taking a look at Isoetes histrix. Report and Transactions La Societe Guernesiaise 22: All- 419. STAPF, O. (1914). The southern element in the British flora. Botanische Jahrbiicher 50, suppl.: 509-525. STAPF, O. (1917). A cartographic study of the southern element in the British flora. Proceedings of the Linnean Society of London 1916-17: 81-92 STEWART, A., Pearman, D. A. & Preston, C. D. eds (1994). Scarce Plants in Britain. Joint Nature Conservation Committee, Peterborough. WlGGINTON, M. J. ed. (1999). British Red Data Books. 1. Vascular plants , edn 3. Joint Nature Conservation Committee, Peterborough. 59 Fern range determination within the Atlantic Arc by an environment of complex and interacting factors C. N. PAGE Halgarrick Lodge, Quenchwell Road, Carrion Downs, Truro, Cornwall TR3 6LN, England ABSTRACT Maps of a selection of fern ranges of species native to Britain and Ireland with extensions well into the Atlantic Arc were displayed (a selection only of which are reproduced here). Analysis of conditions promotional to the success of these highly oceanic-climate ferns shows both close concordance and detailed variation in conditions across the Atlantic Arc. A fuller account of this overall data assemblage across the Atlantic Arc and its integration especially with earth science data for this region forms a wider eventual objective of this study. Keywords: Ferns, range determination, climatic data. INTRODUCTION Importance is attached here to understanding four main sources of influence in determining and sustaining the prescription of these fern species ranges: directly recordable climatic data; derived climatic and other mainly compound data; additional pterido-biotic complexity; and historical factors. DIRECTLY RECORDABLE CLIMATIC DATA Directly recordable conditions promotional to establishment of the ranges of these Atlantic ferns include: • the existence of high levels of humidity, high and well-distributed rainfall, high frequency of cloud-cover, and mild winter temperatures (mappable e.g. as Winter Minimum or February Minimum Temperatures - see Figs 1 & 2 and Page (1997)). Collectively these conditions are supported and buffered across the whole Atlantic Arc by the great volume of ocean water and its directions and constancy of drift, and promote the length, disposition and success of growing seasons achieved in adjacent land areas. DERIVED CLIMATIC AND OTHER MAINLY COMPOUND DATA Further (‘derived’) climatic variables also involved in determining these fern species ranges include: • running day-degree totals of winter thermal input (mappable as Winter Accumulated Temperatures) and the special affinity of these conditions with areas closely influenced by ocean water (Fig. 3), • absolute levels of both thermal and drought extremes (and the special scarcity of these in oceanic areas) and the frequencies, seasonal disposition and durations of these and other short¬ term events. Additionally, complex interplays of all of the above then occur with one another and with other independent physical habitat factors, including rock/soil type, altitude/aspect/exposure and hydrological status of each site, and the year-to-year constancy versus inconstancy of all of these multiple interacting conditions. 60 ATLANTIC ARC Figure 2 FIGURES 1-2. Winter minimum temperatures by 5°C increments (as horticultural 'hardiness zones’ for central and western Europe above, and, in greater detail below, as February Minimum Temperature °C), shows the contrasts between the mild winter temperatures (zone H5) maintained by oceanic influence in the Atlantic arc (dotted areas, above) compared with those of central Europe (H3-H1), and the more detailed increase in long¬ term average minimum temperatures close to the coasts (below, uncoloured areas averaging +2 to +5°C, using Britain and Ireland as examples). Figure 1 compiled from Walters et al. (1986). Figure 2 adapted from Page (1997), data kindly provided by the British Meteorological Office and Irish Meterological Service. FERN RANGE DETERMINATION 61 Figure 3. Winter Accumulated Temperature (as day degrees °C/°F, recorded meteorologically December- March, for Britain and Ireland), provides an invaluable basic framework against which to begin to compare overall ranges of especially Atlantic plant species ranges, indicating areas where more winter plant growth is possible. Compiled from data kindly provided by the British Meteorological Office and Irish Meteorological Service (after Page 1997). 62 ATLANTIC ARC Figure 4. Atlantic ranges of three species of pteridophyte native to Britain Figure 5. Atlantic ranges of three species of pteridophyte native to Britain and Ireland: Dryopteris aemula (dotted line), Hymenophyllum wilsonii and Ireland: Asplenium marinum (solid line), Equisetum telmateia (dashed (dashed line) and Trichomanes speciosum (sporophyte only, solid line). line) and Cystopteris diaphana (dotted line) FERN RANGE DETERMINATION 63 ADDITIONAL PTERIDO-BIOTIC COMPLEXITY Imposed upon the above bases, additional biotic factors consequent on the ecological strategies of ferns (Page 1986. 2002a) and sheer taxonomic diversity (Page 1979a), collectively recognised here as pterido-biotic complexity, typically further modify consequent life-cycle success and thus contribute to further detailed range delimitation. These include: • subtle variations of some of the above conditions operating differently on different individual genotypes, provenances, ploidy levels and species in determining both vegetative and reproductive success, as well as on different life-cycle stages, such as between gametophytes and sporophytes (Page 1973; Cousens et al. 1985; Sheffield 1994: Rumsey & Sheffield 1997). • time-dependent habitat-dynamic factors (including stages of serai state of plant communities, presence of disturbance regimes, niche opportunities, safe-sites, patch dynamics) may need to be satisfied in order to determine the ability of species to either persist or open opportunities for new ones to establish (Gureyeva 2001; Page 2002b). • affinity with or intolerance of other biota: the total of the above conditions also influence the presence or absence of other potentially competing biota, and thus may indirectly modify detailed suitability of sites through this additional subtle and probably biotically extremely complex route (Grime 1985). Importantly in ferns, superimposed upon this diverse range of criteria, high airborne-spore mobility is a further biological asset w’hich gives them high potential to maintain dynamic range boundaries and respond rapidly, as shifting climatic constraints occur and new site opportunities and challenges constantly arise (Page 1979b, 2002a). HISTORICAL FACTORS Historical factors, frequently embracing variations in multiple conditions sometimes far greater than present, may have already played major roles in determining overall species and current gene pools, and have influenced species ranges differently from those of today in both positive and negative ways (Loriot & Magnanon 2005). Some indications of these may be represented by distributions of sub-fossil data e.g. Godwin (1975). Additionally, amongst living pteridophyte taxa, past ranges are also sometimes still indicated today by the presence of unusual pteridophyte hybrids which, through vigour, can sometimes out- persist one or both pteridophyte parents in specific areas, and thus still 'ghost' important past ranges, of which there may be no other evidence (Page 1986; Page & Barker 1985). CONCLUSIONS Fern range determination within the Atlantic Arc is thus achieved by an environment of complex and interacting factors. These include compound physical determinants interacting with both environmental constraints plus multiple innate abilities of the fern taxa themselves (Page 1977, 1979. 1986. 1997; Cousens et al. 1985; Tryon 1987; Gureyeva 2001; Murphy & Page 2005). Especially, because of the high airborne-spore mobility of ferns and other pteridophytes, details of fern range boundaries can thus assume a particular fluidity and dynamism with time, responding swiftly (perhaps often far more so than do most flowering plants) to shifting opportunities as elements of the environment, both past and present, constantly change (Page & McHaffie 1991; Page 2001, 2002b). Beginning to understand the individual sensitivities of ferns to these complex determinants and constraints thus allows their roles as highly responsive indicators to be determined, with an ability to provide widespread and naturally sensitive ever-running early-warning response monitors to subtle fluctuations of environmental conditions, both past, present and future, within the whole of the Atlantic Arc and beyond (Figs 4 and 5) 64 ATLANTIC ARC REFERENCES COUSENS, M. I., Lacey, D. G. & Kelly, E. M. (1985). Life-history studies of ferns: a consideration of perspective. Proceedings of the Royal Society of Edinburgh 86B: 371-380. GODWIN, H. (1975). Tlte History of the British Flora. Cambridge University Press, Cambridge. GRIME, J. P. (1985). Factors limiting the contribution of pteridophytes to a local flora. Proceedings of the Royal Society of Edinburgh 86B: 403-421. GUREYEVA, I. I. (2001). Homosporous ferns of South Siberia. Taxonomy. Origin, Biomorphology, Population Biology. Tomsk State University Publishers, Tomsk (in Russian). Loriot, S. & Magnanon, S. (2005). The task of the Conservatoire Botanique National of Brest in the knowledge and conservation of the Armorican flora as illustrated by Trichomanes speciosum Willd., in LEACH S. J., page, C. N., Peytourau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc, pp. 103-1 12. Botanical Society of the British Isles, London. MURPHY, R. J. & PAGE, C. N. (2005). Isoetes histrix (Land Quillwort) on the Lizard peninsula, in LEACH S. J., PAGE, C. N„ Peytourau, Y. & Sanford, M. N. (Eds.) Botanical Links in the Atlantic Arc , pp. 321-323. Botanical Society of the British Isles, London. PAGE, C. N. (1973). Ferns, polyploids and their bearing on the evolution of the Canarian flora. Monograph Biologicae Canariense 4: 83-88. PAGE, C. N. (1977). An ecological survey of the ferns of the Canary Islands. Fern Gazette 11: 297-312. PAGE, C. N. (1979a). The diversity of ferns: an ecological perspective, in Dyer, A. F. The Experimental Biology of Ferns, pp. 9-56. Academic Press, London. PAGE, C. N. (1979b). Experimental aspects of fern ecology, in Dyer, A. F. The Experimental Biology of Ferns, pp. 551-589. Academic Press, London. Page, C. N. (1986). The strategies of bracken as a permanent ecological opportunist, in Smith, R. T. & TAYLOR, J. A. Bracken. Ecology, Land Use and Control Technology , pp. 173-181. Parthenon Publishing, Camforth. PAGE, C. N. (1997). The Ferns of Britain and Ireland. 2nd edn. Cambridge University Press, Cambridge. PAGE, C. N. (2001). Ferns and allied plants, in HAWKSWORTH, D. L. The Changing Wildlife of Great Britain and Ireland , pp. 50-77. Taylor & Francis, London. PAGE, C. N. (2002a). Ecological strategies in fern evolution: a neopteridological overview. Review of Palaeobotany and Palynology 119: 1-33. PAGE, C. N. (2002b). The role of natural disturbance regimes in pteridophyte conservation management. Fern Gazette 16: 284-289. Page, C. N. & Barker, M. A. (1985). Ecology and geography of hybridisation in British and Irish horsetails. Proceedings of the Royal Society of Edinburgh 86B: 265-272. PAGE, C. N. & McHaffte, H. S. (1991). Pteridophytes as indicators of landscape changes in the British Isles in the last hundred years, in CAMUS, J. M. (Ed. ) The History of British Pteridology 1891-1991 , pp 25^40. British Pteridological Society, London. RUMSEY, F. J. & SHEFFIELD, E. (1997). Inter-generational ecological niche separation and the 'independent gametophyte' phenomenon, in CAMUS J. M., GlBBY, M. & JOHNS, R. J. Pteridology in Perspective, pp. 563-570. Royal Botanic Gardens, Kew, London. SHEFFIELD, E. (1994). Alternation of generation in ferns: mechanisms and significance. Biological Reviews 69: 333-343. TRYON, R. (1987). Ferns speciation and biogeography. Proceedings of the Royal Society of Edinburgh 86B: 353-360. Walters, S. M., Brady, A., Brickell, C. D., Cullen, J., Green, P. S., Lewis J., Matthews, V. A., Webb, D. A., Yeo, P. F. & Alexander, J. C. M. (1986). The European Garden Flora, Vol. 1. Cambridge University Press, Cambridge. 65 An introduction to the flora of the Isles of Scilly R. PARSLOW 17 St Michaels Road, Ponsanooth, Truro, Cornwall, TR3 7ED, England ABSTRACT The Isles of Scilly were formed by the submergence of a larger landmass that was first inhabited 4.000 years ago. Situated away from land on the edge of the Atlantic the warming influence of the North Atlantic Drift (a branch of the Gulf Stream) is reflected in the flora and fauna of the islands. Many of the plant species in Scilly have been introduced, either accidentally or intentionally, over many centuries and are now established members of the flora. Otherwise the native flora is limited by the small land area, habitats, exposure and isolation. There is a strong Lusitanian influence in the flora, with many plants of southern European distribution including some of Mediterranean origin. The cultivation of scented narcissus during winter for the cut-flower trade had, until recently, been the major industry for the farmers. This has resulted in an associated weed flora of winter annuals and other early vemals. The very mild winter climate enables many plants to continue growing throughout the winter. The flora has been augmented by a large number of escaped aliens, most originating from the Tresco Gardens, and these have frequently become established in the wild due to the mild climate. Keywords: Cornwall, Isles of Scilly, floristics, habitats INTRODUCTION The Isles of Scilly is an archipelago of more that two hundred rocks and islands situated 45 km south-west of Land’s End on the Cornish mainland. The islands are part of the granite batholith that extends under the peninsula of S.W. England as a series of bosses or domes, one of which Figure 1 . The Isles of Scilly. 66 ATLANTIC ARC formed Scilly (Edmonds et al. 1975). The overlying soils of the island are derived from degraded granite (Selwood et al. 1998) either as blown sand or ‘ram\ Ram is a cement-like material formed of periglacially frost-shattered granite fragments in a sandy matrix that forms a platform under the sands, the edge of cliffs and under beaches around the coasts. This ram ‘shelf’ is particularly important as one of the main habitats where Rumex rupestris grows, where freshwater seepages trickle onto the shore (Neil et al. 2001). In the north of the islands, on St Martin’s and the adjoining White Island there are small areas of glacial deposits that mark the southern edge of a tongue of ice from the southern edge of the ice sheet that probably just reached Scilly 18,000 years ago (Scourse et al. 1990). Although it has been recognised for at least 250 years that the islands are gradually submerging (Borlase 1756), it was Professor Charles Thomas (1985) who produced his model for Scilly as a drowned landscape. At the end of the last Ice Age the present islands formed one large landmass, then as the ice melted and the seas rose the land became separated into three islands. One island became the present main group of islands, St Mary’s, Tresco, Bryher, Samson and the Norrad and Eastern Isles. Separated from the main mass by a deep channel was another island, that is now the present day islands of St Agnes and Gugh, still further to the south-west was what is now the Western Rocks. When the shallow seas between the main islands dry out at low spring tides, remains of stone walls and hut circles of former occupations are revealed on the extensive sand flats. ISLAND SIZE AND POPULATION v The total land area of the Isles of Scilly is only about 1603 ha. Today only the five largest islands are inhabited. The largest island, St Mary’s (Fig. 2) is 629 ha with a resident population of FIGURE 2. Aerial photograph of St. Mary’s. FLORA OF THE ISLES OF SCILLY 67 approximately 2,000 which becomes inflated to about double that in the summer holiday season. Tresco, the next largest, is 297 ha, with a population of about 130 people and is run as a private estate; it includes the famous Tresco Abbey Gardens, a farm, and holiday accommodation. The other three inhabited islands, St Martin's, St Agnes and Bryher (Fig. 3) are much smaller, each with resident populations of only about 100. All the rest of the islands are now uninhabited although some were occupied, or used to pasture stock, up to the 19th century (Ratcliffe & Straker 1996). All the islands are owned by the Duchy of Cornwall and with the exception of properties in Hugh Town (St Mary’s), the ‘capital' of Scilly, all the farms and other land are leasehold. Much of the unfarmed land, including the uninhabited islands, is now leased to the recently formed Isles of Scilly Wildlife Trust, and managed for nature conservation. EARLY SCILLY When Bronze Age people first settled Scilly during the second millennium BC they would have found a very different landscape from today. At that time the land was well wooded and the present islands were then gentle hills with low-lying land between, some of which may have been marshy. Later the trees were cleared and the land broken up for agriculture by the early farmers. There was much commerce between Scilly and the mainland of Britain, but very significantly the trade routes to France, the Atlantic coasts of Europe and the Mediterranean may have been the source of many plants that are now part of the flora of the islands (Ratcliffe 1992). INHABITED ISLANDS The main habitats on the inhabited islands are farmland, both arable and pasture, maritime heath, coastal dunes, and coastal grasslands; additionally there are wetlands, including open water, mire habitats as well as man-made structures such as ditches and stone hedges. There are numerous small hamlets and farmsteads but the only really urban area is Hugh Town on St Mary’s. UNINHABITED ISLANDS Of the approximately 200 rocks and islands (the number depends on what constitutes an island and the state of the tide) only 30^40 have any kind of higher vegetation on them. Several of the islands are joined together by shingle bars so that at high water they appear as separate islands, but at low water it is possible to walk from one to another. The island of Samson (37 ha) was still inhabited until relatively recently, reaching a population of thirty-seven people in 1829. The island was abandoned a few years later, in 1855. The remains of the houses and fields of the former inhabitants are still evident and also some of the plants they introduced, for example Sambucus nigra, Tamarix gallica and Primula vulgaris are still growing there (Grigson 1948; Lousley 1971). The island consists of two rounded hills joined by a sandy neck. In the north of Samson is a low dune system, much of the rest of the island is covered in dense Pteridium aquilinum communities with heathland on the top of the hills. Samson now holds the most important colonies of Rumex rupestris in the Isles of Scilly around its shores. Tean (16 ha) was also populated for many years and still had a family farming there in 1822. Again there are field boundaries, ruined buildings and some plants that are a legacy of the former occupation. Cattle were still pastured on the island in the 1940s and Lousley (1971) records a number of plants that reflect the agricultural past although most have now completely disappeared. Tean has a very indented coastline composed of sandy bays and dunes between several large hills, at least one of which has a heathland summit. Among a number of interesting species present on the island are the rarities Rumex rupestris, Ornithopus pinnatus and Viola kitaibeliana (Plate 42). Annet (21 ha) is a low island, rather exposed and frequently swept by salt-laden winds. There are tall rock cams at the extremities and the shore is made up of rocky outcrops surrounded by extensive boulder beaches. The island is important for seabird populations whose long occupation has modified the vegetation. Most of the north of the island and the coastal fringe is covered in an extreme version of the Armeria sub-community of the Festuca rubra-Armeria maritima plant community dominated by enormous Armeria maritima tussocks (Rodwell 2000). In May the whole island glows pink from end to end when the Armeria flowers. Until 1983 there were at least five colonies of Rumex rupestris on boulder beaches on the south of the island before heavy seas washed all but one away. 68 ATLANTIC ARC There are several other large, uninhabited islands. St Helen's and Round Island in the north of the archipelago have both been inhabited in the past. St Helen’s still has the ruins of a hermitage and a pest house (a quarantine place for sick sailors in the days of sailing ships). The rest of the island is a high heathy hill and in places Carpobrotus edulis has spread from gulls’ Larns spp. nests (gulls frequently take plant material to the islands to build their nests). Round Island is a dome-shaped island with high cliffs, topped by a lighthouse, now automatic. Established on the top of the island are several species of succulents, possibly taken there by the lighthouse-keepers for their garden. The most exposed group of islands in Scilly are the Western Rocks, situated as they are on the south western edge of the archipelago. Virtually bare, jagged rocks and cams, surrounded by rough seas, many reefs and deep water, they support hardly any vegetation. Few higher plants can grow on the islands, at most a few strandline species, Atriplex spp., Cochlearia officinalis and Spergularia rupicola and sometimes a stand of Lavatera arborea. Rosevear is the only island in the group ever to be inhabited, although it is only two hectares across. It was occupied in the summers between 1851-1858 by the workmen building the Bishop Rock lighthouse. Although these hardy characters are reputed to have made a garden on the island there is no sign today and as the whole island gets drenched in salt spray during storms the conditions could not have been ideal. Another similar group of islands are the Norrad Rocks on the west of the Isles of Scilly. In contrast to the extreme exposure of the Western Rocks, the Eastern Isles are relatively gentle in contour and somewhat sheltered by the ring of the main group of islands and shallow water over the sand flats between them. There are extensive Zostera marina beds between the islands. Most of the Eastern Isles are well vegetated with Pteridium/Rubus communities, Festuca rubra grassland around the edge and heathland on the top of the hills. Some of the islands have sandy bays with Crambe maritima, Glaucium flavum and dune plants. All of the smaller islands are occupied by gulls Larus spp. in spring and summer, so have typical seabird communities with Cochlearia officinalis, Tripleurospermum spp., Holcus lanatus and Atriplex spp., all with very large and fleshy leaves. Among the more unusual aspects of the Eastern Isles are the presence of rarities such as Ornithopus pinnatus and Scrophularia scorodonia, also Stachys officinalis at one of only two places in Scilly. Some of the plants that are found on the Eastern Isles, for example Euphorbia amygdaloides, Ruscus aculeatus and Calamagrostis epigejos, are believed to be a link with the former forest that covered the islands 4,000 years ago (Thomas 1985). The only apparently native Quercus robur tree outside the inhabited islands also grows on Great Ganinick in the Eastern Isles - still a stunted tree, just one metre high, as it has been for more than 70 years. THE BASIS OF THE FLORA The small size of the islands, their long separation from mainland Britain and the limited range of habitats have, as would be expected, led to a paucity of native flora and fauna. Compared with the Lizard and the rest of Cornwall, Scilly has many fewer native species, although there are many archaeophytes (plants that became naturalised before 1500 AD) that have greatly augmented the island flora. Many common mainland plants are either very rare or absent from Scilly, and the distance and the prevailing wind direction and currents make colonisation from that direction unlikely. Scilly shares many similarities with the flora of the Channel Islands, where there are also many southern species that are unable to survive the colder climate of mainland Britain. The Isles of Scilly were visited by people from Neolithic times and became settled during the Bronze Age. From the time they were first inhabited the islands were on the early trade routes to the Mediterranean from where plants could have been introduced to the islands either accidentally or deliberately. The first colonists also brought in seeds for their crops, kept livestock and traded, some plants would have arisen as seed contaminants or accidentally among goods or wrappings. Early Christian settlements on the islands, for example the Benedictine priory on Tresco, were also probable sources of introduction for medicinal plants and herbs, Tanacetum parthenium for example, and possibly the Tazetta Narcissus that later became the basis of flower farming (Lousley 1971). The gardens on Tresco have also been the source of many of the recent introductions to the flora. In 1834 the Isles of Scilly were leased to Augustus Smith, a wealthy FLORA OF THE ISLES OFSCILLY 69 Englishman with an interest in horticulture. Smith settled on Tresco, on the site of the former priory, and laid out a garden around his house. He found that by planting deep shelterbelts of deciduous trees and later Cupressus macrocarpa and Pinus radiata around the garden he was able to grow many tender plants from Australia, New Zealand and South Africa and he was soon exchanging plants with the botanical gardens at Kew. The Tresco Abbey Gardens have been maintained and expanded by Augustus Smith's successors the Dorrien Smiths, and are now world famous. Not satisfied with growing exotic species in the Gardens some plants were deliberately spread around the islands, notably into the dune systems in the south of Tresco where bizarre mixtures of alien plants have become established alongside the native plants. These include Agapanthus praecox, Fascicularia bicolor, Ochagavia earned, Muehlenbeckia complexa and many other aliens, giving the dunes a very exotic appearance. Carpobrotus edulis was also planted around the islands to stabilise the sand dunes, blown sand being frequently driven right across the islands by the strong winds right into the houses and into the fields, causing a great nuisance to the inhabitants. This has resulted in Carpobrotus spp., and some other succulent species, becoming so successfully naturalised that in places they swamp the native flora. Carpobrotus has also become established on some of the uninhabited islands from pieces of the plant taken there by gulls. Many of the hedging plants now universally planted in Scilly, Pittosporum crassifolium, Olearia traversii, Escallonia macrantha and several others, as well as wall plants such as Lampranthus spp. and other succulents, and Oxalis megalorrhiza, all originated from Tresco gardens. Every year more species escape into the countryside, assisted frequently by human means or sometimes by birds. Birds, especially water-birds, are also probably the agents for spreading native species, presumably from seeds stuck to their feathers, for example the arrival of Bidens tripartita by the Abbey Pool, Tresco, in 1994. Other plants may be windblown, the appearance of Anacamptis pyramidalis on Samson about ten years ago, or waterborne, such as with the rare appearances of Lathyrus japonicus , and recently both Polygonum maritimum and P. oxyspermum subsp. raii. MAIN HABITATS COASTAL HABITATS The coastal flora is varied and quite rich. The configuration of the coasts around the Isles of Scilly allows much variation between sandy shores, dunes, rocky shores, boulder beaches and low cliffs. There is a gradation from sandy beach to dunes, dune grassland and heathland on stabilised dune in a number of places. Carex arenaria is a common pioneering species in these habitats, growing away from the shore in heathlands often right to the summit of the hills. It also grows along the tops of the stone 'hedges’ around the fields, under Pteridium and as dense, talk ‘grassland' on the island of Annet. Scilly was once considered to be the stronghold of Rumex rupestris in S.W. Britain. Over the past few decades the population on Scilly has declined markedly, with colonies being washed off the beaches during storms and high tides and not replaced. Although the plants produce plenty of viable seed it does not appear to be finding suitable germination sites to enable it to re-colonise (Neil et al. 2001). Poa infirma appears to spreading around the islands since it was first recorded in 1950 (Lousley 1971). This tiny annual grass of sandy pathways, short turf and bare ground is found where there is either vehicle or pedestrian traffic. It was even discovered growing under tables in the tea gardens on St Agnes, presumably carried there on the shoes of holiday visitors. Another plant, now increasingly rare in mainland Britain, is Chamaemelum nobile , which in Scilly is a feature both of coastal grasslands and of heathland/grassland mosaics. Scrophularia scorodonia is typically found in Pteridium-Rubus communities at the back of the dunes and around the coast where the soils are more acidic, but also occurs in farmland and on waste ground especially on Tresco. Another rarity, the miniscule annual Viola kitaibeliana is found in dune grasslands (Randall 2004, 2005) on several islands. 70 ATLANTIC ARC HEATHLAND There are extensive areas of heathland on all the larger islands and many of the smaller islands have patches of heathland plants on the tops of hills. In places where the soils are deeper Pteridium aquilinum has invaded and become locally dominant. In former times it would have been kept in check by the islanders who cut it for bedding for cattle, thatching their cottages and even for fuel. Similarly Ulex europaeus has spread hugely since the 1970s when myxomatosis began to decimate the rabbit population. As the loss of rabbits coincided with the reduction in grazing by farm stock and as large tracts of former cultivated land were being abandoned, this led to an explosion in the growth of invasive species. No longer kept in check, U. europaeus soon formed dense thickets over large areas. A mixture of cutting, burning and rolling is slowly reversing this process. The intention is to introduce conservation grazing to many of the heathland areas in the next decade. Fortunately, there are still substantial areas of Calluna-Erica and Calluna-Ulex gallii heathland on the islands. Where they are exposed to the fierce westerly winds the heathlands have developed into wind-eroded heath or ‘waved heath’ with its characteristic rippled appearance (Rodwell 1991) caused by the plants being tumbled over by the prevailing wind (Plate 8). Ornithopus pinnatus is one of the rare species usually associated with heathland, although it was also found in disturbed sandy habitats including arable fields and even sometimes as a weed in the Tresco Abbey Gardens! Several of the rare clovers, Trifolium spp., grow in grassy patches of the heathland mosaics; one that is very abundant in early spring, especially around the more maritime edges and cliffs is Trifolium occidental. Where the rock pavements are very close to the surface the tiny ferns Ophioglossum azoricum and O. lusitanicum may occur (Parslow 2005). FIGURE 3. Bryher, one of the smaller islands. FLORA OF THE ISLES OF, SCILLY 71 CULTIVATED HABITATS The flower industry that has been the mainstay of island life for over a century is currently in decline and many farms have had to look to tourism to supplement their livelihood. Scented narcissus are still grown for the cut-flower market between October and April, but bulb production is also now becoming more important. Some bulb fields are now being taken out of production, either turned over to alternative crops such as summer cut flowers, left fallow or sown with grass. Frequently the colourful bulb fields attract the interest of passing holiday makers, who may admire the Oxalis pes-caprea, Chrysanthemum segetum, Lavatera cretica or Fumaria spp. growing along with the cultivated narcissus, without recognising how unpopular they are with the farmers. Most of the successful species found in the bulb fields are winter annuals or species that reproduce vegetatively by bulbils or corms, making them difficult to eradicate. Some of the species that are now weeds such as Gladiolus communis subsp. byzantinus were originally introduced as potential cut flowers but later abandoned (Lousley 1971) (Plate 9). An apparent reduction in the amount of herbicide use on the farms in recent years has favoured the bulb field weeds and Lavatera cretica, Polycarpon tetraphyllum, Fumaria occidentalis and Silene gallica are all spreading. Other species, for example the buttercups Ranunculus muricatus and R. parviflorus, seem less common than they were 20 years ago. Allium ampeloprasum var. babingtonii is still common in arable fields and in the Pteridium-Rubus communities around the coast. Silene gallica is still abundant, as are Erodium moschatum , Briza minor, Anisantha diandra, Fumaria capreolata and F. occidentalis. The latter has apparently increased its range on St Mary’s although it is mainly a scrambler over walls rather than a weed amid the crop. Wind-breaks are an important element in providing the mild microclimate of the bulb fields. Initially Tamarix gallica and Ulmus spp. were planted to supplement the stone hedging and reed screens to shelter the fields. Later these were supplanted by evergreen shrubs with shiny, waxy leaves able to resist salt spray and filter the winds. Many of these shrubs were originally from New Zealand or other oceanic climes: Pittosporum crassifolium, Escallonia macrantha, Hebe x franciscana, Olearia traversii and other shrubs were found to grow fast and provide the necessary shelter in the winter to protect the narcissus. WETLANDS The islands have very little natural freshwater, the only stream of any size is on St Mary’s and this is extracted to feed into the public water supply before flowing through the marshy Higher Moors into Porthellick Pool and then into the sea. The two main wetlands on St Mary’ s are Lower Moors and Higher Moors (Moors in this case denoting marshes or mires). These are areas of Phragmites swamp and Juncus maritimus (var. atlanticus according to Lousley 1971) surrounding freshwater pools (these still have controlled links with the sea so at times may become slightly brackish). The Great Pool on Tresco extends almost from one side of Tresco to the other and is edged by large Phragmites beds. The nearby Abbey Pool supports a number of different species of wetland plants around the drawdown zone of the lake, some that are not found elsewhere in Scilly, including Elatine hexandra, Littorella uniflora and a tiny form of Ranunculus flammula. There is another, occasionally slightly brackish pool on St Agnes with Ruppia maritima and Potamogeton pectinatus, surrounded by Bolboschoenus maritimus and Juncus gerardii. Many of the smaller pools on the islands have floating populations of Ranunuculus baudotii and Juncus bulbosus and occasionally Glaux maritima around the margins. The most important true brackish lagoon is Great Pool on Bryher with a leat connecting it to the sea; it is inhabited by marine invertebrates and fish, and a fluctuating population of Ruppia maritima. URBAN HABITATS For many plants the buildings and walls are just an extension of the coastal rocks, so Euphorbia portlandica, Asplenium marinum and several other fern species are as well established on the town walls as on the coast. Polycarpon tetraphyllum is everywhere in cracks in the pavements and many of the dune and bulb field species seem to grow as weeds in town gardens. ACKNOWLEDGMENTS So many people have sent records and given of their time and expertise that they are too many to acknowledge here. They know who they are and I thank them for their generosity. 72 ATLANTIC ARC REFERENCES Climatic Tables for the Isles of Scilly 1931-1993, Brackley: National Meteorological Library and Archive (in litt.). EDMONDS, E. A., Mckeown, M. C. & Williams, M. (1975). British Regional Geology: South-west England. H.M.S.O., London. GRIGSON, G. (1948). The Scilly Isles. Paul Elek, London. LOUSLEY, J. E. (1971). The Flora of the Isles of Scilly. David & Charles, Newton Abbott. Neil, C. J., King, M., Evans, B. B„ Parslow, R. E., Bennallick, I. B. & Mcdonnell, E. J. (2001). Shore Dock Rumex rupestris. Report on work undertaken in 2000. Unpublished report to Plantlife & English Nature. PARSLOW, R. E. (2005). Maritime communities as habitats for Ophioglossum ferns in the Isles of Scilly, in the Isles of Scilly, in Leach S. J., Page, C. N., Peytoureau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc, pp. 73-81. Botanical Society of the British Isles, London. RANDALL, R. E. (2004). Biological Flora of The British Isles: Viola kitaibeliana Schlt(es). Journal of Ecology 92: 361-369. Randall, R. E. (2005). Viola kitaibeliana Schult. (Dwarf Pansy) in the Isles of Scilly, in Leach S. J., Page, C. N., PEYTOUREAU, Y. & SANFORD, M. N. eds. Botanical Links in the Atlantic Arc, pp. 277-280. Botanical Society of the British Isles, London. Ratcliffe, J. & STRAKER, V. (1996). The Early Environment of Scilly. Cornwall Archaeological Unit, Cornwall County Council, Truro. RATCLIFFE, J. (1992). Scilly ’s Archaeological Heritage. Cornwall Archaeological Unit, Twelveheads Press, Truro. RODWELL, J. S. ed. (1991). British Plant Communities. Volume 2: Mires and heaths. Cambridge University Press, Cambridge. RODWELL, J. S. ed. (2000). British Plant Communities. Volume 5: Maritime communities and vegetation of open habitats. Cambridge University Press, Cambridge. Selwood, E. B., Durrance, E. M. & Bristow, C. M. eds. (1998). The Geology of Cornwall. Exeter University Press, Exeter. Thomas, C. (1985). Exploration of a drowned Landscape - Archaeology and History of the Isles of Scilly. Batsford, London. 73 Maritime communities as habitats for Ophioglossum ferns in the Isles of Scilly R. PARSLOW 17 St Michaels Road, Ponsanooth, Truro, Cornwall TR3 7ED, England ABSTRACT Maritime grass and heathland communities, influenced by the proximity of the Atlantic and the Gulf Stream, provide habitats for ferns of the genus Ophioglossum in the Isles of Scilly. All three native taxa of Ophioglossum are found in the islands. KEYWORDS: Adder’s tongue, Moon wort, Ophioglossum lusitanicum , O. azoricum, O. vulgatum. INTRODUCTION The Isles of Scilly are an archipelago of between 150 and 200 rocks and islands of which five are inhabited and about 45 have higher vegetation of some kind. The largest island is St Mary’s, where most of the human population live; the other four inhabited islands are Bryher, Tresco, St Martin’s and St Agnes (including Gugh, joined to it by a sandbar). All four ferns in the Ophioglossaceae were found in the Isles of Scilly until recently. Botrychium lunaria was present in the dunes at Bar Point on St Mary’s, but despite a number of searches, it has not been seen since 1982. Of the three British taxa of Ophioglossum , O. lusitanicum and O. vulgatum are only found on the island of St Agnes. O. azoricum is found on all the five inhabited islands, plus a small colony on Toll’s Island just off the coast of St Mary’s. MARITIME PLANT HABITATS Immediately inland from the coastal fringe on the Isles of Scilly are broad bands of habitats among the granite outcrops and tors. These are made up of mosaics of maritime grasslands and heathland, which vary according to soil depth and exposure to salt-laden winds. Where the vegetation grows on shallow soils over rock pavements, as on Wingletang Downs on St Agnes, is where many of the colonies of O. lusitanicum and O. azoricum are found. O. lusitanicum frequently forms patches of fronds in the crust of thin turf around the rim of flat rocks, although it also grows in more therophyte associations as well as in damp hollows dominated by mosses. Characteristically these places are usually saturated with rainwater in winter, even temporarily submerging the fronds at times, later becoming baked in summer when the fronds have died down. O. azoricum grows in a wider range of habitats than O. lusitanicum, ranging from the edges of rocks alongside O. lusitanicum, to characteristic therophyte communities on rocky outcrops, damp areas on heathland, also under bracken, in seasonal pools and in dune grassland. The only colony of O. vulgatum is in a damp comer of a meadow, growing under bracken. The mild climate enables O. lusitanicum to grow throughout winter and complete its life cycle by early spring when O. azoricum appears, sometimes in the same localities. Although O. azoricum is the most common species, it is less widespread than formerly. Lack of grazing has led to scmb encroachment on many of its former sites and at least one has been ploughed up. As there has been considerable confusion between the different taxa in Scilly in the past, this has sometimes resulted in mistakes in identification and in consequence, some erroneous records. CLIMATE The climate of the Isles of Scilly is usually characterised as oceanic, with wet, mild winters with very little frost and rarely any snow; mild, sunny summers and frequent strong winds and gales 74 ATLANTIC ARC and also sea fog. It is warmer in Scilly in winter than on the mainland and relatively cooler in summer, the average monthly mean is 11-7 °C (National Meteorological Library). A major influence on the climate is the North Atlantic Drift, an arm of the Gulf Stream. The most important influence on the plants is the temperature, with c. 350 days a year usually above 5 °C, the growing temperature for most plants. This mild climate is believed to be the key to the survival of O. lusitanicum on Scilly as it enables the fern to grow throughout winter and to have shed its spores before spring. Although frosts are few and snow is rare, on the occasions when there have been frosts the fertile fronds have been damaged, blackening and disintegrating. GENERAL OBSERVATIONS Although Ophioglossum lusitanicum had been known on Guernsey in the Channel Islands since 1854, it was not recorded from the Isles of Scilly until John Raven found it on St Agnes in March 1950 (Raven 1950). In 1953 Lousley described the colony as consisting of about one hundred fronds covering just over a square metre at the foot of a large boulder. The colony was visited from time to time by botanists and frequently photographed. Eventually it was discovered that O. azoricum grew in the same locality and, with its fronds emerging in April, just as those of O. lusitanicum disintegrate, it had often been mistaken for the earlier species. Similar confusion has also dogged the presence of O. vulgatum. Lousley clearly did not see the fern in Scilly, but examined material collected by J. E. Dallas labelled as from a damp area at the north end of Wingletang Bay on St Agnes. Today there is a colony of O. azoricum in this locality, but the only O. vulgatum identified recently is from another damp area at the north end of the island where it has been known for many years. It is intriguing to wonder whether Dallas had put the ‘wrong’ locality, or whether O. vulgatum may have been more widespread in the past. According to Paul (1987) O. azoricum may not be intermediate between O. lusitanicum and O. vulgatum, but one end of a spectrum of variation in O. vulgatum. But she does acknowledge there are populations that are very distinctive. The Isles of Scilly plants seem to fall into this category. „ iri ,. , „ . . . , Differences between the three species rIGURE 1. Raven s site where Ophioelossum lusitanicum was . . first discovered. are given in Table 1. TABLE 1. THREE OPHIOGLOSSUM TAXA, BASED ON SCILLY MATERIAL 0. lusitanicum O. azoricum 0. vulgatum Size 1-2 cm 2-4 (-5) cm 5-12 cm Season (October) November-March April-July May-August Sterile fronds 1-2 1-2 1 Sporangia 3-8 6-13 11 + (pairs) OPHIOGLOSSUM IN THE ISL^S OF SCILLY 75 FIGURE 2. Sterile fronds of Ophioglossum lusitanicum (scale in cm). OPHIOGLOSSUM LUSITANICUM Over the past thirty years several botanists, including the author, have surveyed St Agnes to map the fern, both on the heathland where it had originally been found, as well as looking for it elsewhere on St Agnes, the adjoining island of Gugh and in suitable habitat on other islands. Most searches have been in early spring or October, but visits during the winter months have extended the search period at the most advantageous time to find the fronds. Ophioglossum lusitanicum usually appears above ground in October or November with the sporangia ripening in January or February, before the fronds gradually turn yellow and disintegrate in April. Some particular clones appear earlier than others, with the fronds appearing as early as September in some years and a few sterile fronds may still be found, again at certain sites, at the end of April. From probing of the ground it has been found that the fern usually grows in substrate only 2-3 cm deep. It does occur in deeper soils, but is then liable to be out-competed by grasses, Armeria maritima and more robust plants. O. lusitanicum grows in short turf over granite platforms; several sites clearly hold water during the winter and at times the fronds may be completely submerged. At one site the fronds emerge directly from among mosses, notably Polytrichum sp. The associated species (Table 2) vary from site and from year to year, for example Radiola linoides may be more common in wet years, and in dry years the parched turf may be dominated by Plantago coronopus, usually very tiny plants, sometimes with virtually entire leaves. It is sometimes difficult to distinguish between very small fronds of O. lusitanicum and the leaves of Plantago spp. 76 ATLANTIC ARC 7 6 5 4 3 2 1 □ 9 8 7 6 456789012345 FIGURE 3. Distribution of Ophioglossum lusitanicum and O. azoricum in the Isles of Scilly. OPHIOGLOSSUM IN THE ISLES OF SCILLY 77 FIGURE 4. Fertile fronds of Ophioglossum azoricum. OPHIOGLOSSUM AZORICUM O. azoricum is found on all the inhabited islands and in a greater range of habitats than O. lusitanicum. Associated species for eight sites are given Table 3. In several places on St Agnes it grows directly beside O. lusitanicum. Where both grow side by side, O. azoricum is found in the deeper soils. This is by far the commonest of the three Ophioglossum taxa. It appears in some respects to be intermediate in character between O. lusitanicum and O. vulgatum , although quite distinct in size, sporangia and season (Table 1). The distribution of O. azoricum appears to have contracted over the past twenty years. Some colonies have disappeared on St Agnes and St Martin’s and possibly elsewhere. In one case a pasture was ploughed up and put down to arable crops, another field was treated with herbicide to clear bracken and the Ophioglossum was not seen for many years although a few fronds have been reported recently. Several former sites have become submerged by taller vegetation, grasses for example, and often Pteridium or Ulex due to lack of grazing. This latter is a particular hazard in the case of O. azoricum : where it is growing in deeper soil there can be direct competition with Ulex europaeus or dense Pteridium aquilinum and Rubus spp. (some Pteridium appears to be acceptable as light cover). Page (1988) noted that Ophioglossum species can survive for a time as underground rhizomes without producing fronds, this seems to happen quite frequently in Scilly. On heathland sites such as Wingletang Down, there was some re-emergence of the fern when firebreaks were cut through the gorse,but how long it can remain dormant is not known. OPHIOGLOSSUM VULGATUM O. vulgatum has been recorded from several of the islands in Scilly but most of these records are now believed to have been errors for O. azoricum , possibly because it had not been recognised how variable the latter fern can be in Scilly. The O. vulgatum site on St Agnes is in a patch of Pteridium in a damp comer beside a stone hedge. The ground is sandy and the O. vulgatum emerges in May as the Pteridium is at the ‘crazier’ stage. The other species in the area include Rumex acetosa, Viola riviniana and Rubus sp. The Ophioglossum had not been seen in this area for several years since very dense Rubus invaded the site. Over the past two years (2002 & 2003) the site has been cut to reduce the Rubus and, it is hoped the Ophioglossum may reappear. 78 ATLANTIC ARC MONITORING OPHIOGLOSSUM LUSITANICUM In early April 1980 sketch maps were made of the known O. lusitanicum sites, partly as an aid to re-finding the colonies later and also to record their size. This was followed by a baseline survey in December 1981/January 1982, including making sketch maps and taking photographs to identify the sites. Soil depths were recorded at some sites, also soil temperature. Since then the colonies have been recorded very simply by counting fronds and photographing the sites. It has not been possible to visit St Agnes every winter so counts have had to be made whenever possible, although this means counts at either end of winter tend to be low. Several methods were tried, but the simplest and most practical proved to be visiting each colony in turn, recording the number of both fertile and sterile fronds and relating this to a site map. Photographs and site descriptions were made if time allowed. Searches for new colonies were also made both on St Agnes and elsewhere. Up until 2004 more than 15 ‘colonies’ of O. lusitanicum have been located, all on St Agnes. Most are restricted to the heathland on Wingletang Down, the number of fronds varying from less than ten to hundreds. From counting fronds at each site, in some cases over more than twenty years, the numbers of sterile fronds do not seem to vary a great deal, although the number of fertile fronds is very variable. This may be because the sporangia break off very easily (many of the sites are on pathways), and some may even be eaten by rabbits or drop off in frosty conditions or heavy rain. An attempt to revisit each colony is made every year, especially as it has been noted that no fronds may be seen at some stations for several seasons before reappearing later. Also the time of emergence varies from colony to colony; this may reflect the effect of rock pavement and open situations causing some sites to warm up earlier than others. One colony consistently emerges some weeks before most of the others. OPHIOGLOSSUM O. AZORICUM AND O. VULGATUM. It has not been practical to monitor these species in the same detail. The colonies of O. azoricum in places are very large, covering several square metres. One colony near Cam Adnis, St Agnes had more than 1000 fronds per square metre. Although some counts were made, it was only found practical to record localities due to lack of time, the large number of colonies and the several islands involved. TABLE 2. SPECIES ASSOCIATED WITH OPHIOGLOSSUM LUSITANICUM ON WINGLETANG DOWN, ST AGNES Species No. sites Species No. sites Sedum anglicum 6 Hypochaeris radicata 1 Plantago coronopus 10 Holcus lanatus 1 Radiola linoides 2 Ulex europaeus 1 Festuca rubra 10 Agrostis capillaris 2 Plantago lanceolata 3 Rumex acetosa 1 Ornithopus perpusillus 3 Danthonia decumbens 2 Armeria maritima 8 Erica cinerea 2 Cochlearia danica 1 Centaurium erythraea 1 Spiranthes spiralis 1 Anagallis arvensis 1 Trifolium sp. 1 Chamaemelum nobile 2 Dactylis glome rata 1 Calluna vulgaris 1 Ophioglossum azoricum 1 Agrostis stolonifera 1 Lotus comiculatus 2 Bare ground 3 Carex arenaria 2 Mosses 5 Leontodon saxatilis 2 Polytrichum sp. 1 Rubus sp. 2 Lichens 1 OPHIOGLOSSUM IN THE ISLES OF SCILLY 79 TABLE 3. SPECIES ASSOCIATED WITH OPHIOGLOSSUM AZORICUM IN ISLES OF SCILLY Species A B C D E F G H No. Sites Plantago coronopus + + + + + + 6 Radiola linoides + + + + + 5 Festuca rubra + + + + 4 Omithopus pinnatus + + + + 4 Sedum anglicum + + + + 4 Aira praecox + + + 3 Armeria maritima + + + 3 Calluna vulgaris + + + 3 Polytrichum sp ./ mosses + + + 3 Rumex acetosa + + + 3 Aira caryophyllea + + 2 Danthonia decumbens + + 2 Erica cinerea + + 2 Hydrocotyle vulgaris + + 2 Isolepis setacea/cemua + + 2 Juncus bulbosus + + 2 Lichens + + 2 Lotus comiculatus + + 2 Lotus subbiflorus + + 2 Omithopus perpusillus + + 2 Pedicularis sylvatica + + 2 Potentilla erecta + + 2 Pteridium aquilinum + + 2 Ranunculus flammul a + + 2 Sagina apetala/procumbens + + 2 Ulex europaeus + + 2 Agrostis capillaris + 1 Agrostis stolonifera + 1 Anagallis tenella + 1 Beilis perennis + 1 Carex arenaria + 1 Carex viridula subsp. oedocarpa + 1 Centaurium erythraea + 1 Chamaemelum nobile + 1 Eleocharis sp. + 1 Erodium cicutarium + 1 Euphorbia paralias + 1 Euphorbia portlandica + 1 Galium saxatile + 1 Hypericum humi fusum + 1 Juncus bufonius + 1 Juncus effusus + 1 Poly gala sp. + 1 Prunella vulgaris + 1 Taraxacum section laevigatum + 1 Trifolium omithopodioides + 1 Trifolium repens + 1 Vulpia bromoides + 1 Sites on Isles of Scilly are coded - A: Chapel Down, St Martin’s; B: Heathy Hill, Bryher; C: Giant’s Castle, St Mary’s; D: Beady Pool, St Agnes; E: Merchant’s Point, F: Rushy Bank, Tresco; G: Browarth Point, St Agnes; H: Wingletang, St Agnes. 80 ATLANTIC ARC HEATHLAND MANAGEMENT AND OPHIOGLOSSUM FERNS Many of the heathland areas on the Isles of Scilly have become over-run by Pteridium aquilinum and Ulex europaeus due to lack of grazing. Over the past forty or more years there has been a considerable reduction in the numbers of cattle and horses on the islands. This has coincided with the impact of myxomatosis on the rabbit population with a consequent explosive expansion of the Ulex. The islanders have usually reacted to this by burning off the Ulex at intervals, mainly to minimise the danger of uncontrolled fires and to restore the heathland. Without follow-on grazing this has not been very successful, as the areas have quickly reverted to scrub. The main danger to the Ophioglossum ferns is from uncontrolled bums that can get into the thin peaty soils and can bum down to the bedrock. Plans by the Isles of Scilly Wildlife Trust and others to regulate burning in future, as part of heathland restoration, should benefit the Ophioglossum ferns. Controlling Pteridium is more problematic, but some cutting and rolling of bracken is proving effective, and if combined with grazing it could restore some of the heathland and maritime grassland where O. azoricum formerly occurred. CONCLUSIONS More colonies of O. lusitanicum are now known on St Agnes than when counts started more than twenty years ago. This could be the result of more concentrated searches and greater familiarity with the fern and its habitats, or spores may have spread to form colonies in new localities. Some ‘sub-colonies’ appear to have arisen from the spread of the rhizomes under the turf. All three taxa have suffered from competition with scrub and rank vegetation resulting from lack of grazing due to reductions in stock grazing and the decimation of the rabbit population by myxomatosis. Some active management, e.g. strimming the sites, has been effective at keeping them open, but in some places even this has not managed to maintain the open, very short sward necessary for O. lusitanicum. FIGURE 5. Ophioglossum lusitanicum site in 1987 FIGURE 6. Recent photograph of the same site showing close cropped vegetation. showing rank vegetation due to lack of grazing. OPHIOGLOSSUM IN THE ISLES OF SCILLY 81 AC KNOWLEDGMENTS A number of botanists over the years have spent many hours searching for Ophioglossum ferns, notably J. Bevan, R. Finch and A. Colston. I am grateful to them and to all the others who have contributed their time and records. D. Mawer of the Isles of Scilly Wildlife Trust has both helped with monitoring sites as well as carrying out practical management work on the O. lusitanicum sites. I am particularly indebted to many tolerant friends in the Isles of Scilly. Maps were produced using DMap, by Alan Morton. REFERENCES LOUSLEY, J. E. (1971). Flora of the Isles of Scilly. David & Charles, Newton Abbot. PAGE, C. N. (1988). Ferns. Collins New Naturalist, London. PARSLOW, R. (1999). Ophioglossum Species in the Isles of Scilly: Species Action Plans , Management Prescriptions & Inventory of sites. Unpublished report to English Nature. PAUL, A. M. (1987). The status of Ophioglossum azoricum (Ophioglossaceae: Pteridophyta) in the British Isles. Fern Gazette 13 (3): 173-187. Raven, J. E. (1950). Note on the flora of the Scilly Isles and the Lizard Head. Watsonia 1: 356-358. 83 Atlantic Arc grasslands: the example of the rhos pastures of south-west Wales and their conservation R. D. PRYCE Trevethin, School Road, Pwll, Llanelli, Carmarthenshire SA15 4AL, Wales * ABSTRACT The vegetation communities and species typically associated with rhos pastures - the damp, acidic, often heathy grasslands that still occur frequently in south-west Wales - are described; their management, social significance and conservation is summarised. Carum verticillatum (Whorled Caraway) is highly characteristic of rhos pastures and is part of a suite of ‘Atlantic’ species that contributes to the particular character of these grasslands. The history of the recording of Carum in south-west Wales is referred to, and its distinctive character, together with its often abundant survival on the Carmarthenshire coalfield, is explained. Reference is made to the major losses which have occurred in recent years, due to opencast coal-mining, commercial and domestic development and agricultural changes. In view of the apparent ineffectiveness of current protection measures, a plea is made to all concerned for a concerted effort to conserve and maintain what remains. Keywords; Wales, Atlantic Arc grasslands, rush-pastures, vegetation classification, conservation. INTRODUCTION This paper is based upon the B.S.B.I. Presidential Address delivered during the Atlantic Arc Conference on 10 May 2003 and which has already been published in Wat sonia (Pryce 2004). Although based upon the Address, this paper includes and emphasises certain aspects more fully than time allowed at the Conference. The term rhos pasture was originally coined in the mid 1970s by Chris Fuller of the then Nature Conservancy Council. He used it to describe the unimproved, species-rich, damp, acid meadows or pastures of south-west Wales which typically included an element of dwarf-shrub heath (maybe 10-20% cover of Calluna vulgaris (Heather)) which reflects the fact that they are derived from the overgrazing and periodic burning of heathland, as has been demonstrated at Dowrog Common in Pembrokeshire (S. B. Evans, pers. comm.). Rhos is, of course, not an invented word, but is derived from the Welsh rhos meaning moorland and was intended to convey the impression, particularly amongst the Welsh farming community, the custodians of these grasslands, of unimproved, heathy pastureland, worthy of conservation for its floral and faunal diversity. The term seems since to have come to be used in a wider sense, a sign of the characteristically diverse nature of these grasslands which has resulted from small-scale variations in such factors as hydrology, soil origin, soil chemistry and physical characters: variations which are destroyed by cultivation and intensive farming practices. The majority of sites are well on the acid side of neutral, having a pH of 4-5 to 5-5 with soils being damp and clayey and slow to warm-up in the spring. Vegetation reminiscent of rhos pasture is, of course, not confined to south-west Wales and comparable communities are found along the Western European seaboard from the uplands of northern Portugal and north-west Spain, through Brittany, Devon and Cornwall, north to south¬ west Scotland and east to Belgium and Germany. This vegetation is, however, surprisingly restricted in most of these areas, with only the Scottish stands approaching the extent of those in south-west Wales. In all areas it remains under extreme pressure from agricultural improvement or neglect, as well as being lost to commercial development. *e-mail: PryceEco@aol.com 84 ATLANTIC ARC OVERVIEW OF RHOS PASTURE VEGETATION Grassland cannot be considered as climax vegetation in South Wales and depends upon man’s intervention for its maintenance. Over the last 5000-6000 years the gradual removal of the native broad-leaved forest and the introduction of domestic grazing animals has created and maintained a more open face to the landscape. The local soils are mostly derived from sticky, impervious, acid, glacial boulder-clay, which generally blankets the bedrock and often exceeds 10 m in thickness. Drainage is often further impeded by the formation of a ferruginous pan in the soil at around 20- BO cm depth. Variations in topography and the inherent poor drainage also provide an opportunity for the deposition of thin peaty layers locally and these often form a humic component in the predominantly mineral soils. Where topography and other physical factors permit, blanket peat and small raised-bogs have developed locally. These features have historically been the most difficult areas to drain and improve in agricultural terms and several remain today, although they have often been subject to conifer plantings in recent years and are increasingly susceptible to drainage facilitated by modem powerful machinery. Additional diversity is provided where, for example, seepages and flushes have developed their own characteristic vegetation. Flushes are generally acid and may support small areas of swampy vegetation dominated by bryophytes, Carex spp. (sedges) and Juncus spp. (rushes). However, they may be base-rich locally, particularly where water arises from the Carboniferous Limestone or from base-rich strata in the Old Red Sandstone or Farewell Rock (the uppermost member of the Millstone Grit formation). Base enriched water may also arise from pockets of glacial material derived from these rocks. Better-draining banks and slopes also occur locally and result from the occurrence of sandy glacial deposits. These different physical conditions have contributed to the variety of distinctive vegetation types found. Although rhos pastures would have been considerably more extensive in the past, modem agricultural (so-called) ‘improvement’ has virtually eliminated them from all the currently intensively-farmed ‘better land’ except, for example, in valley bottoms where draining the land has so far been uneconomic. However, traditional farming practices, to some extent, even now remain a way of life in the Carmarthenshire coalfield and also still continue on parts of the northern flank of Mynydd Du (the Black Mountain) and on clayey drift-covered land on the southern flank of the Tywi valley south-east of Carmarthen. The most extensive unimproved or little-improved grasslands remaining in South Wales today therefore include concentrations in these areas (Fig. 1). The reason for their survival in the coalfield is because land holdings were invariably small with production only needing to provide a supplement to the main source of income from the mines or associated industries. On the flanks of Mynydd Du, however, their survival may be because the farm-holdings are linked with extensive common mountain-grazing rights which allowed less intensive use of the in-by land in summer (Bevan 1999). CHARACTERISTIC NATIONAL VEGETATION CLASSIFICATION TYPES The National Vegetation Classification (N.V.C.) (Rodwell 1991 et seq .) has provided a means of characterising the range of vegetation communities typical of rhos pastures. The high rainfall of the region strongly favours the formation of damp and wet grasslands on the acid, clayey soils and it is not therefore surprising that the most extensive N.V.C. communities are rush pastures and Purple Moor-grass mires. The M23 Juncus effusus/acutiflorus - Galium palustre rush pasture community is the most extensive and is characteristically dominated by either Juncus ejfusus (Soft Rush) or J. aciitiflorus (Sharp-flowered Rush), the latter (indicative of the M23a Juncus acutiflorus sub-community) often characteristic of more diverse swards of higher conservation value. The M25 Molinia caerulea - Potentilla erecta mire tends to be very variable and the degree of dominance of Molinia depends to a great extent on its management. TLe M25a Erica tetralix sub¬ community is frequent in the area, and where grazing has been neglected, Molinia can become overwhelmingly dominant and very tussocky but the wettest stands may include significant amounts of Sphagnum and Polytrichum commune mosses and support extensive populations of Narthecium ossifragum (Bog Asphodel). More heavily grazed sites are generally more diverse and RHOS PASTURES OF SOUTH-WEST WALES \ 85 FIGURE 1. Areas of ‘Marshy Grassland' in Carmarthenshire identified during the Countryside Council for Wales Phase 1 Habitat Survey of 1989-91, shown in black. Two of the principal concentrations of rhos pastures are in the Carmarthenshire coalfield (shown in grey) and in the area north of Mynydd Du (shown in light grey). The Phase 1 definition of ‘Marshy Grassland' not only includes rhos pastures but also several other grassland types. This is seen most clearly, for instance, in the concentration of ‘Marshy Grassland' in the north of the county which, whilst including some areas of rhos pasture, also supports large tracts of species- poor Molinia- dominated grassland which falls outside the generally accepted meaning of the term rhos pasture. may include the M25b Anthoxanthum odoratum sub-community on drier ground or the M25c Angelica sylvestris sub-community where the water table is high in more nutrient-rich substrates. Where the dwarf-shrub element increases, M25 mire can grade into M15 Scirpus cespitosus - Erica tetralix wet heath. Flushed stands may commonly grade into M6 Carex echinata - Sphagnum recurvum/auriculatum mire or, rarely, where the seepages are more base-rich, MlOa Carex dioica - Pinguicula vulgaris mire, Carex demissa - Juncus bulbosus/kochii sub-community, a vegetation type more characteristic of the north of Britain. Where a degree of base-enrichment is present in the soil, the M24c Molinia caerulea - Cirsium dissection fen meadow, Juncus acutiflorus - Erica tetralix sub-community may occur. This is a less common community in the dominantly acid, boulder-clay derived soils of the region but where present, Cirsium dissection (Meadow Thistle) can form dense stands with an additional suite of characteristic base-demanding species. In better-draining soils which, for example, may occur on the crests or slopes of drumlins or other glacial features or where the drift is less clayey, typical wet rhos vegetation may give way to MG5 Cynosurus cristatus - Centaurea nigra grassland. The MG5c Danthonia decumbens sub¬ community is most characteristic of typically acid rhos pasture vegetation but where a minor degree of agricultural improvement has occurred, such as the application of lime or basic-slag, conversion to the MG5a Lathyrus pratensis sub-community is usual. 86 ATLANTIC ARC Figure 2. Some characteristic rhos pasture species showing their Atlantic distributions (from Meusel et al. 1965, 1978, 1992).. RHOS PASTURES OF SOUTH-WEST WALES 87 CHARACTERISTIC SPECIES OF RHOS PASTURES Several plant species of rhos pastures can be regarded as having purely Atlantic distributions although many principal species, such as Molinia caerulea (Purple Moor-grass), Calluna vulgaris (Heather) and Succisa pratensis (Devil’ s-bit Scabious), tend to be more widespread. It is, however, the association of the Atlantic species which contributes to the character of rhos grassland vegetation. These include, for instance, Cirsium dissectum (Meadow Thistle), Erica tetralix (Cross-leaved Heath), Genista anglica (Petty-whin), Scutellaria minor (Lesser Scullcap), Vicia orobus (Wood Bitter-vetch) and Wahlenbergia hederacea (Ivy-leaved Bellflower) (Fig. 2), but, without doubt, the most characteristic species in south-west Wales is Carum verticillatum (Whorled Caraway) which, when growing abundantly, may turn fields white with its flowers in summer (Plate 11). Carum verticillatum was first recorded in the region in 1773 by Rev. John Lightfoot and Sir Joseph Banks, during their botanical tour of Wales. They noted “ Sison verticillatum [C. verticillatum] in a low moist meadow on the left hand of the Road adjoining to a small bridge call’d Pelcombe Bridge Wi miles from Haverford West in the way to St. Davids, in Abundance in Flower”. On their return journey, they again noted the abundance of the species at Narberth, Lanreed near St. Clears, and then “the same afterwards for 6 miles together in almost all the low moist meadows between St. Clears and Carmarthen”. They made notes on the abundance of the plant between Llandeilo and Llandovery and also in the area around Edwinsford. They concluded that “it is a common plant in moist and boggy meadows in the 2 Counties of Pembroke and Carmarthen”. A specimen collected by them at the time is now in the herbarium of the Natural History Museum in London (BM). Recording of C. verticillatum and its associates has continued over the succeeding years and its most up-to-date distribution is shown in Figure 3, which has been derived from Meusel et al. (1978), Preston et al. (2002) and Pryce (1999). Carmarthenshire CARMARTHENSHIRE Carum verticillatum Figure 3. Distribution maps of Carum verticillatum (Whorled Caraway) (from Meusel et al. 1978, Preston et al. 2002 and Pryce 1999). 88 ATLANTIC ARC Carum verticillatum is a member of the Apiaceae, generally grows to about 0-5 m in height and has very distinctive, simply divided, pinnate leaves which give the impression of being whorled (Plate 10). The plant is a biennial or short-lived perennial and can be recognised at most times of the year, except in the dead of winter, as the leaf rosette appears early, being usually recognisable by late February. The main flowering season is from the end of June and through July but some flowers are often produced though into October. C. verticillatum reaches its greatest abundance in the traditionally managed fields of the Carmarthenshire coalfield and although it is often frequent into east and north Pembrokeshire, Ceredigion, west Breconshire and west Glamorgan it seldom attains such profusion. C. verticillatum occurs in all the N.V.C. communities cited above, with the possible exception of M10, but reaches its greatest abundance in M23 rush pasture. Most prolific flowering occurs in hay meadows cut in July or August where grazing stock has been excluded prior to mowing (Plate 11). C. verticillatum is strongly Atlantic in its distribution and predominantly western in Britain (Fig. 3), but only occurs in a few areas in Ireland. However, the Carmarthenshire distribution as shown by the tetrad map is misleading in as much as, although the species occurs in virtually all tetrads in the coalfield area (i.e. the south-east of the county), the map-dots would indicate a similar occurrence of the plant elsewhere. This, in fact, is not the case, as although there are meadows away from the coalfield where C. verticillatum is abundant, they are much fewer in number and the majority of populations are confined to relatively few plants in relict valley mires or field comers or to even fewer plants in flushes at stream heads and water-collects. C. verticillatum will tolerate a degree of agricultural improvement and can persist in the otherwise rather monotonous swards resulting from moderate amounts of fertiliser input. THE DECLINE OF THE RHOS PASTURE RESOURCE South-west Wales is fortunate to have retained a relatively extensive unimproved grassland resource which has resulted from traditional farming practices, as mentioned above. In particular, the Amman and Gwendraeth Fawr valleys of the Carmarthenshire coalfield still support a relatively large area of unimproved rhos grassland although in recent years much has been lost. Despite the wholesale despoliation of land caused by the coal-mining industry over the last 200 years, the practice of miners augmenting their wages by keeping a few livestock ensured that the land not affected by mining activities was managed in a way which maintained the characteristic vegetation communities. It must, however, have been a hard life and, in recent years, many of those land-holdings which had survived development in one form or another have been lost or degraded by neglect or have been subject to small-scale agricultural intensification, often by the ‘horsiculture’ fraternity. Of course, the mines and their associated spoil tips, together with railways and other infrastructure, must have destroyed large areas of unimproved grassland which, in contrast to today’s improved agricultural swards, would have been commonplace and their loss was of little concern to local inhabitants to whom the greater prosperity afforded by employment was a great benefit. But associated with the collieries, extensive areas of land were required for hay production in order to supply fodder for the ponies which, at the time, provided the primary haulage-power underground. This therefore, further encouraged grassland management to continue. The result was a mosaic of rhos pastures, some degraded (and often diversified) by alteration of drainage patterns caused by mining subsidence or the placement of spoil-tips in the vicinity, juxtaposed with despoiled land and growth of housing to accommodate the expanding population. Furthermore, pit-heads were generally sited where the seams outcropped, which tended to concentrate them in relatively small areas. Mine developers generally avoided areas where thick barren strata or glacial drift overlaid the productive coal-measures, or where the geological structure was so complex as to cause uncertainty over their investment. These factors can be illustrated, for instance, in the lower Gwendraeth valley, where areas of drift and alluvium up to 40 m thick mask the solid geology. The geological complexity here, where two major faults converge (the Trimsaran Disturbance and the Llannon Disturbance) (Frodsham et al. 1993), made RHOS PASTURES OF SOUTH-WEST WALES 89 it impossible for the seams to be followed by the miners underground. For this reason, the area remained largely unexploited, allowing the survival of the small, extensively managed fields which supported all the characteristic rhos pasture vegetation types until it was opencasted in the 1980s and 1990s (see below). Up until the Aberfan disaster in 1966, the land-management situation remained more-or-less unchanged throughout British coalfields, but the resulting shock to the nation caused a public outcry, the outcome of which was the call for all colliery tips to be made safe. The wholesale reclamation of colliery tips commenced and gathered pace as the coal industry’s decline accelerated. Not only did this result in the loss of the tips and their wildlife interest per se , but the spoil was invariably spread, often covering areas of otherwise unspoiled rhos grassland, the end result being topographically uniform expanses of bland, cultivated grass and clover mixes, interspersed with occasional alien tree plantings. The cost of carrying out such reclamation work was considerable, and opencasting the coal that remained in situ in the vicinity of closed collieries offered a way to achieve the goal without cost to the public purse. Opencast had started as a means of cheaply winning coal in the 1940s with a few small sites helping the War Effort but economies of scale were soon realised and the magnitude of sites increased exponentially. Many former collieries were targeted and in the 1970s and 1980s were excavated and restored to agriculture. But the whole exercise was primarily driven by the enormous profits gained by maximising the tonnage of coal recovered, so that the amount of previously undisturbed land also included, in the name of land reclamation, was very considerable. At that time, all emphasis was on increasing agricultural production and opencast restorations were hailed for their considerable success in converting poor agricultural land into land capable of greater production. No consideration was given, at that time, to the damage being inflicted on the natural flora and fauna. Many of the restorations carried out at that time remain of little biodiversity value today. Restoration on the completion of mining invariably consisted of sowing the site with cultivated grass and clover leys of generally low ecological value. These were difficult to manage because of the lack of soil structure which resulted in waterlogging in winter and excessive drying-out in summer. Restored land requires large agricultural inputs to maintain its productivity and those restorations which have not received such intensive management, or have been neglected, have quickly been subject to infestation by J uncus effnsus (Soft Rush), often accompanied by the natural colonisation of a few common forbs such as Ranunculus repens (Creeping Buttercup), Rumex acetosa (Common Sorrel) and Cardamine pratensis (Lady’s-smock), the result after twenty or thirty years being an N.V.C. MG 10-like coarse, rush-dominated grassland. Agricultural productivity is also minimal. Experiments to restore some species diversity by the introduction of turves, containing Carum verticillatum and other rhos pasture species, to restored land were unsuccessful. The plants did not grow well and soon died out, and the little seed which was produced would not easily germinate in the alien conditions (Medcalf 1989, 1990). Where restored land is more intensively farmed, there is also little likelihood of traditional rhos pasture vegetation becoming re-established due to the changed hydrology and topography, frequent high nutrient inputs and the high sheep-grazing intensity needed to control the constant threat of rush infestation. Furthermore, the restored fields, being of a size to facilitate more efficient agricultural working, are completely alien to the traditional small field pattern of the pre-opencasted land. The paucity of hedgerows and mature trees reduce the former rich farmland bird, mammal and invertebrate fauna as well as having an adverse landscape impact. Recent restorations have included the reinstatement of small areas of vegetation types reflective of those which existed prior to opencast exploitation. The Gilfach Iago site near Cross Hands includes an example of a small area of rhos grassland successfully restored following opencast (Humphries & Benyon 1999). But the overall losses far exceed any recent gains. Another major cause of rhos pasture destruction over the last ten to twenty years, has been the large-scale development of business parks, often joint ventures between Local Authorities and the Welsh Development Agency (WDA). Again the emphasis has been given to the clearance of derelict land and, for example, at Cross Hands the old colliery tips have been re-contoured to provide the development plateaux for business and retail developments. In common with the opencast operations of former decades, large areas of adjacent unimproved grassland have also been destroyed as part of these schemes. 90 ATLANTIC ARC A few, as yet, unrestored colliery sites remain, often associated with rhos grassland and areas of equally valuable semi-natural vegetation. It is notable, however, that most are proposed for designation as land suitable for reclamation or development in the emerging Carmarthenshire Unitary Development Plan, although a few local communities have had the vision to make sure that such sites are listed for public recreation and heritage. Locally, house building has, in recent years, also caused considerable losses, especially along road-frontages where the Local Planning Authority has been notorious for allowing ribbon- development, seemingly unchecked, which now links most of the previously discrete villages. This practice has resulted in many examples of unimproved grassland being lost or damaged, including a rhos pasture S.S.S.I. This problem continues, not least because much of this land has been designated in the past as development land in the Local Structure Plan, and therefore might expose the local authority to compensation payments to the land owner should its designation be reversed. The small coalfield farms were always barely viable or merely provided a supplement to the main income and now that their owners are getting older or passing-on, they are too work¬ demanding and not large enough or sufficiently productive to sustain a viable living. The neglect of appropriate management has now, therefore, taken over as possibly the most damaging factor causing rhos pasture losses. Absence of grazing allows scrub colonisation and Molinia domination. Reinstatement of this neglected land is expensive and requires many years’ commitment as well as conservation-management awareness and even then may not be entirely successful. Mechanical trashing of tussocky Molinia generally requires the employment of specialist equipment, and scrub clearance is equally expensive, time-consuming and laborious. CONSERVATION? ‘ Molinia meadows on chalk and clay’ is a biome included in the E.U. Habitats Directive (Anon. 1992), and Purple moor-grass rush pastures are listed as a ‘Habitat of Principal Importance for Conservation of Biological Diversity’ under Section 74 of the Countryside and Rights of Way Act 2000; they are also a Priority Habitat in the U.K. Biodiversity Action Plan and a habitat for which action plans have been implemented in all county Local Biodiversity Action Plans in Wales. Although a few rhos grasslands have been designated as S.S.S.I.s in Carmarthenshire and elsewhere and should, in consequence, be relatively safe from degradation, it might be expected that Tir Gofal, the whole-farm, Wales-wide countryside stewardship scheme, would be instrumental in conserving rhos pastures and unimproved grassland by encouraging the farming community to do just that. It is hoped that the days are gone when countryside stewardship meant excavating a pond or tree-planting on the most species-rich wet areas of a farm. Nowadays, the Tir Gofal budget is £16-8m per year for spending on such things as fencing, increasing access, woodland maintenance and planting, set-aside and less intensive grassland management. However, it is feared that the most important targets for conservation - the areas which need the most care in their management and protection - are not being thoroughly addressed. There certainly appears to be little feedback on how such payments have actually benefited the conservation or re-instatement of valuable grasslands. Contrast this £16-8m per year with the mere £2-6m being spent in Wales to maintain Sites of Special Scientific Interest, the sites which are constantly being hailed as the ‘Jewels in the Crown’ of our wildlife heritage. Even worse, only £15000 is being spent on S.S.S.I.s in Carmarthenshire, most of which will be accounted for by existing management agreements, i.e. annual payments, and many of these will not be grassland sites. This leaves very little scope for new management work which is so desperately needed to arrest the decline of sites through neglect. It also raises the question as to whether the little available money is being channelled towards the most urgent and effective conservation targets. Furthermore, with the impending reorganisation of C.C.W., it is proposed that there will be an embargo on the notification of new S.S.S.I.s for three years, except in exceptional circumstances, due to the insufficiency of funding. Nature conservation continues to be a very poor relation to any enterprise which might make or potentially make money or bring in jobs or votes. South-west Wales may well be better off than many areas where habitat losses have been more severe, but this fact alone reinforces the RHOS PASTURES OF SOUTH-WEST WALES 91 responsibility of those in authority to retain and conserve what is left. Local Authorities should, as a matter of urgency, introduce the designation of second tier nature conservation sites in their Unitary Development Plans. They should not, as in the case of Carmarthenshire, delete from the Draft Unitary Development Plan the few Sites of Importance for Nature Conservation which are already identified and accepted in a current, adopted Local Plan which relates to part of the county. Developers, local authorities and the WDA should ensure that basic ecological assessment is undertaken prior to site acquisition and areas found to be of ecological importance should be left untouched. Both central and local government should ensure that basic education at all levels includes teaching on the appreciation and fundamentals of the natural environment to ensure that the general public is aware, and a proportion, at least, might therefore care, about the well-being of the environment and biodiversity in general. CONCLUSION In south-west Wales a major part of the rhos pasture resource (as well as that of other habitats of high conservation value) has been lost. However, the extent of that which remains is reasonably well known and the urgency of protecting it is constantly being emphasised (if often ignored by those in authority). Substantially more money has been directed towards nature conservation, and wildlife legislation is much stronger than in the past, but the habitat resource continues to diminish. The uphill struggle to retain and manage at least a proportion of those rhos pastures which remain depends upon the importance given to their value by those who have to balance the limited resources available to fund all aspects of spending. Only the constant and increased lobbying of these decision-makers will continue to draw attention to the current dire situation of rhos pastures and may encourage them to take the appropriate remedial action. ACKNOWLEDGMENTS I acknowledge C.C.W. in permitting me to use various data in this paper and thank my many friends and colleagues within B.S.B.I., C.C.W. and elsewhere for their help and encouragement in its drafting. REFERENCES Anon. (1992). Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official Journal of the European Communities: L 206. Bevan, J. M. S. (1999). Semi-natural Grassland and its Conservation in Carmarthenshire. Bulletin of Llanelli Naturalists. 4: 27-32. FRODSHAM, K., Gayer, R. A., James, J. E. & Pryce, R. D. (1993). Variscan Thrust Deformation in the South Wales Coalfield - a Case Study from Ff os-Las Opencast Coal Site, in GAYER. R. A.. GREILING R. O. & VOGEL, A. K. (eds.) The Rhenohercynian and Sub Variscan Fold Belt., pp. 315-348. Vieweg, Wiesbaden. HUMPHRIES, R. N. & Benyon, P. R. (1999). Standard criteria & methods to assess the reclamation of native vegetation on mined land. Paper presented at the 16th National Meeting of the American Society for Surface Mining and Reclamation, Scottsdale, Arizona, August 13-19, 1999. MEDCALF, K. A. (1989). The Ecology and Reinstatement of Carum verticillatum. Brief Summary. Ph.D Thesis, University of Aberystwyth. MEDCALF, K. A. (1990). Recreating Carum verticillatum meadows by use of species-rich turfs. Interim Results. Ph.D Thesis, University ofAberystwyth. MEUSEL, H., JAGER, E. & WEINERT, E. (1965). Vergleichende Chorologie der Zentraleuropaischen Flora. Vol. 1. 2 vols. Gustav Fischer, Jena.. Meusel, H., Jager, E., Rauschert, S. & WEINERT, E. (1978). Vergleichende Chorologie der Zentraleuropaischen Flora. Vol. 2. 2 vols. Gustav Fischer, Jena.. Meusel, H., Jager, E., Brautigam, S, Knapp, H.-D., Rauschert, S. & Weinert, E. (1992). Vergleichende Chorologie der Zentraleuropaischen Flora. Vol. 3. 2 vols. Gustav Fischer, Jena. PRESTON, C. D., Pearman, D. A. & Dines, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. PRYCE, R. D. (1999). Carmarthenshire Rare Plant Register. Llanelli. Pryce, R. D. (2004). The rhos pastures of south-west Wales and their conservation. Presidential Address, 10 May 2003. Wat sonia 25: 1-16. RODWELL, J. S. ed.(1991 et seq.). British Plant Communities. Cambridge University Press, Cambridge. 93 Celtic hedges as refuges for fern diversity in predominantly agriculturalised landscapes C. N. PAGE Halgarrick Lodge, Quenchwell Road, Camon Downs, Truro, Cornwall TR3 6LN, England* ABSTRACT This account touches briefly upon the principal aspects of Celtic hedges of earth-and-stone, which characterise much of the northern area of the Atlantic Arc. Against a background of their structure, history, distribution and botanical significance within the predominantly highly agricultural modem landscapes, it focuses especially on the large and often abundant fern component of these man-made habitats, and the value of these habitats today as particularly important refuges for this plant group. Reasons are advanced for suggesting, from Neolithic times onward, through the processes of change of the original wildwood landscapes from pre- agricultural settings to modern agricultural conversions, why Celtic hedges have come to provide particularly vital pteridological habitats. Two especially important elements of this are that these hedges have functioned to provide vital linear but island-like refuges, each individual to the specific locality, for survival of a basis of ancient fem diversity, while also, equally importantly and also like islands, from this basis have continued to provide important sites for the natural processes of fern micro-evolution to continue to constantly recur. The continuing value of hedges for persistence of the resultant fem diversity in modern settings is presented by a detailed survey of hedges across three west-Cornwall parishes. The importance of conservation of Celtic hedges is thus stressed, and some of the political structures already in place in Cornwall are cited as examples of successful botanical, archaeological, agricultural, planning and local government co-operation to these collective conservation ends. Keywords: Cornwall, fern diversity, agriculturalised landscapes, refuges, conservation. THE CELTIC HEDGE - ITS STRUCTURE AND OVERALL OCCURRENCE The earth-and-stone ‘hedge' which we are familiar with in the south-western parts of Britain, is, in Cornwall, referred to as the ‘Cornish Hedge’. It is a linear boundary-structure with an earth interior and faced with stones of varying sizes, which are typically initially structured to become tightly- interlocked with time. Such hedges may be free standing (two-sided hedges), or act as ground- retaining structures (one-sided hedges). Often, across the slopes which so abound within the county, combinations of these are involved, while the latter also often form bases to support the former. Commonly around 1 -5-2-0 m or more in height and around at least 1-25-1 -50m wide at the base, these are particularly substantial structures. Near to the sea, they may be bare of further woody growth, whilst inland, their crests are often the sites of woody plant growth entanglements and often trees, which add further both to their functional role and to their habitats for biological diversity (Fig. 1) (Page 1988; Meneer 1994; Meneer & Page 2000). In Cornwall at least, such hedges abound everywhere, along almost every roadside, lane and bridleway, divide fields from one another, and form the steep-set banks of many sunken ‘water- lanes’ which follow original stream courses down valleys to the sea. Such structures are thus extensive features of the fabric of the overall landscape, and form conspicuous elements of virtually every inland rural view. Today such hedges are often particularly wildlife-rich, and host abundant plants, lichens, insects, isopods, amphibia and reptiles, smaller mammals and native land snails. Amongst their plant content, they are often particularly fern-rich (Plate 12). The differences of this Celtic hedge structure from the ‘English hedge’, especially of more inland England (Pollard et al. 1974; Rackham 1986), formed merely of a line of shrubs and small trees, and usually fern- poor, are thus profound. *e-mail: pterido@hotmail.com 94 ATLANTIC ARC Figure 1. Diagrammatic representation of cross-sectional structure of three typical Celtic hedges as widely represented in Cornwall. LEFT: typical free-standing Celtic hedge beside a track or bridleway in an exposed location, structured to descend the contours and capped with turf, on which a scattered windswept growth of occasional shrubs and small trees such as Blackthorn ( Prunus spinosa) may be present; MIDDLE: typical free-standing Celtic hedge between fields in an inland location, constructed to follow the contours and thus cross the slope, capped with turf from which typically medium-sized to quite large trees of species including Oak ( Quercus ), Ash ( Fraxinus ), Cornish Elm ( Ulmus ), Hawthorn ( Crataegus ) and Hazel ( Corylus ) may be present; RIGHT: typical mainly ground-retaining Celtic hedge alongside a sunken (and probably ancient) bridleway (set to the right in the diagram) descending the contours near to the sea, retaining appreciably higher soil levels in the field behind, and capped with turf and a shrubby growth typically mainly of the above tree species or of Gorse ( Ulex ). Note the success of fern growth (here stylised) in all of these and the tendency of different fern life-forms to be zoned according to shade, shelter, drainage and soil moisture, thus reflecting several different original wild habitat sources, for which the hedges continue to substitute today. Cornwall is not alone in having such boundary structures, similarly constructed fern-rich hedges have been identified in: • Devon and parts of Dorset • south-west and west Wales, the Llyn Peninsula, Anglesey and the Isle of Man • south-west and central southern Ireland • Guernsey, Jersey and the smaller Channel Islands • many parts of Brittany. The hedges of all of these areas have the following features in common: • they occur within generally similar mild, moist oceanic climates • they occur through regions of similarly ‘clean’ air • they have originated through largely the same construction methods • they have served the same agricultural needs of providing shelter for stock, especially from prevailing westerly winds • they have come into existence through similarly long periods of history and pre-history • they form rather similar landscape features across the Atlantic Arc • they have far more in common with one another biotically than they have differences. CELTIC HEDGES AND FERN .DIVERSITY 95 The Celtic hedge is thus quite widespread within the Atlantic Arc. It is a habitat type which is particularly well-represented in the landscape of Cornwall, where over 30,000 miles (50,000 km) of such hedges occur within this county alone (P. McCartney, Cornwall Wildlife Trust, pers. comm.). This makes the Celtic hedge the most widespread semi-natural habitat in Cornwall, and this may be the case in the Atlantic Arc region as a whole, where its intrinsic high biotic, evolutional and conservational significance contrasts greatly with that of the ‘green deserts’ of generally intensively-managed agricultural pastures which these Celtic hedges enclose. For the above reasons, these hedges are thus here genetically christened Celtic hedges , to reflect this overall similarity across much of the Atlantic Arc region. This also distinguishes them collectively from the English hedgerow (which is a very much later addition to the environment of especially the English shire counties - e.g. Pollard et al. 1974; Rackham 1986). SIGNIFICANCE OF CELTIC HEDGES TO FERN POPULATIONS Ferns are one group of biota which have succeeded and persisted especially well within these ancient Celtic hedges, to a far greater degree than they have in the traditional English hedgerow. The biodiversity significance of Celtic hedges to fern populations is one which can be expressed both collectively and individually: Collectively, Celtic hedges: • occur through a wide range of altitudes, aspects, and degrees of exposure or shelter • form conspicuous and functional elements within a wide range of landscapes, from moorland uplands, across undulating landscapes and sheltered valleys to sea coasts • offer a range of habitats from shady woodland banks to exposed cliff slopes • have a linear structure which provides not only inter-linking wildlife corridors, but also has extensive ‘edge effects’ because of the juxtaposition of hedges to other habitats • are surprisingly permanent and largely unchanging structures through time. Individually, Celtic hedges: • achieve overall enhanced local shelter, even in exposed coastal sites • offer a multitude of habitat niches within the stone-earth matrix • across each individual hedge profile provide habitats with many contrasting exposures and degrees of drainage (varying from dry and well-illuminated to moist and shady ones) • reflect closely the geology specific to each locality (for their construction stone was seldom moved further than it had to be) • have often a further individuality of local construction methods, with details dictated in-part both by local tradition and by the nature of the materials. Celtic hedges, of earth-and-stone construction, thus occur widely in the modem man-made habitats of at least the northern part of the Atlantic Arc. They are a widespread consequence of the agricultural systems which caused their construction, have provided a uniquely effective substitute habitat for a whole cross-section of former more tmly wild habitats that existed within the original wildwood (Rackham 1986). Built out of natural materials which originate locally, the matrix of each hedge thus differs in type from place to place, and reflects the indigenous geological materials. Many of the elements contributing to the structure of the Celtic hedge, especially shelter, shade, combinations of enhanced moisture and drainage regimes, generally low-nutrient status and direct mineral contact of plants with sometimes complex edaphic mosaics, combine to maintain habitats which, in the mild moist climates of the Atlantic Arc, are especially suitable for establishment of fem species. These include several of south-western Mediterranean-Atlantic and Atlantic overall range (Jalas & Suominen 1972; Jermy et al. 1978; Page 2005a), often in luxuriance, as well as a number of other species of similar range, including, for example, Rubia peregrina and Umbilicus rupestris (Preston & Hill 1997; Preston & Arnold 2005), and an extensive growth of ferns (Ivimey-Cook 1984; Page 1988; French et al. 1999; Murphy, Page & Parslow in press.). Together with the known ancient history of many of the hedges bearing them, this makes Celtic hedges, at least in Cornwall, of unusually high biodiversity importance - a picture which is reflected across the whole of the northern part of the Atlantic Arc, to a degree which is seldom fully appreciated. 96 ATLANTIC ARC FERN INDICATOR-VALUE IN CELTIC HEDGES In the British Isles, ferns have been shown to also have high potential indicator value in terms of interpretation of landscape history (Page 1988, 2001; Page & McHaffie 1991) as well as in terms of providing important tools for study of elements of the processes of continuing evolutionary progress (Page 1978, 1997a & b, 2001, 2002a, 2003). Ferns are one group of Celtic hedge species which would thus appear to be candidates for having high current indicator value with respect to these structures (other groups must also exist - including various insects and other invertebrates, possibly snails). Ferns are useful indicator species within Celtic hedges (as well as in other habitats) because: • there are multiple species (but not too many to be daunting) • most species range widely (thus valid comparative studies between localities and sites can be made) • most are large, do not move, can be accurately identified, and enjoy a long recording season (many can be recorded on a year round basis - thus extensive and objective surveys are possible) • are surprisingly tolerant of variation in natural edaphic features such as low nutrient surfaces and exposure to unusual rock content and soil types • are highly intolerant of many man-induced influences (e.g. airborne pollutants) • their species embrace a range of subtly differing habitat preferences and responses, many of which are different from (and often opposite to) those of flowering plants (to which conclusions based on their study are thus independent and stand in their own right) Collectively, these aspects help to provide information about: • natural climatic and edaphic extremes of habitats and the fluctuations of these • the influences causing and the processes involved in change • man-made influences Thus, ferns can be used to improve our understanding of the derivation of habitats in the past and their conservation management both now and in the future. For in these perspectives, Celtic hedges specifically are biologically in an ideal position to exploit two extremes of evolutionary opportunity: • They are in an ideal position to preserve elements of ancient fern diversity • These will have almost certainly provided recurring microsite opportunity for continuing evolutionary microprocesses to constantly and actively recur. In terms of survival of ancient fern diversity, a unique aspect of Celtic hedges is that they have been in an ideal position to preserve at least some of the diversity originating more or less directly from the pre-agricultural landscapes of the locations in which the hedges occur. For many Celtic hedges date in their origins from Medieval times, others earlier. Many such hedges were constructed around the time that the adjacent land was cleared from original cover (especially woodland) to fields, there was almost certainly abundant opportunity for fern immigration into the newly-made hedges to take place from the surrounding wild vegetation of the time - thus from the landscapes which themselves have since become extensively agricultural around them. Because of the high mobility of the airborne spores of ferns, this would have allowed original wild species genotypes to establish anew as local hedge populations on the basis of multi-individual immigrants directly from the former wild habitats. Observations from newly-constructed hedges shows that this would happen faster than for most flowering plants, whose subsequent populations would usually have had to build from a smaller number of initial immigrant propagules. By contrast, my observations show further that virtually a full complement of fern species naturally immigrate and can be detected as young sporophytes within as little as three to five years from construction, and that it is usually in this period that the greatest pace of recruitment occurs. It is thus likely that a large part of the fern flora of such ancient hedges today is likely to have persisted more or less directly from the pre-agricultural landscapes of the regions and specific locations within which the hedges still occur. CELTIC HEDGES AND FERN .DIVERSITY 97 In terms of provision of sites for the natural processes of fern evolution, such Celtic hedge structure itself appears to provide mosaics of particularly favourable microsite opportunities for continuing evolutionary microprocesses. For this, the conditions of close juxtaposition of rock and soil matrices with vertical mineral surfaces produce a detailed structure of locally enhanced dynamics in which a myriad of small erosion-mediated disturbance-patches more or less constantly recur. Evolutionary exploitation of opportunities arising will have likely achieved some further refinement of genotype adaptation to particular local conditions, as well as yielding opportunities for origins of new hybrid combinations (which may have occurred elsewhere in the past, but now occur in the hedges themselves) (Page 2002 a & b, 2003, in press a & b). In Celtic hedges, both of these two extremes of evolutionary opportunity are almost certainly widely present and probably deeply interwoven in their rich pteridophytic tapestries. Consequently today within many Cornish Celtic hedge structures, it is possible to identify local morphological variations from place-to-place within the same species, and also local enclaves of especially high within-species variation in certain hedge sites. How much of this variation is relictual and how much the result of subsequent evolutionary processes will provide an area for valuable future research. The occurrence of hybrid formation in many Cornish Celtic hedges is however exemplified today by the presence of substantial numbers of inter-specific hybrids in particular (notably, for example, in Polypodium - see Page & Murphy 2001) and occasional ones even of intergeneric origins (such as in x Asplenophyllitis - Page 1991, 2005b). FERNS AS INDICATORS OF THE PRESENT HEALTH- OF CELTIC HEDGES To appreciate the degree of intensity of agriculture in our present landscape, the full landscape tapestry has to be seen from the air today (Figs 2-A). Such perspectives make clear the importance of these hedges as some of the only refuges available over surprisingly extensive areas of some of the most intensively managed landscapes anywhere. It is also clear to see, in many rural views, where previous Celtic hedges have been removed wholesale in recent years in order to make ever larger field sizes. Such hedges may have originally once taken many men several months to build, and existed subsequently for many centuries. But their removal and utter destruction of their distinctive and sometimes highly detailed structure is today scarcely a day or two’s work for one man driving a modem tractor-mounted diesel-driven digger. In addition to total loss of such Celtic hedges, there are more insidious causes of losses of biodiversity value too, which can deteriorate rapidly even within persisting hedgerows. A detailed survey of roadside hedges to assess their value as sites for fern diversity in the parishes surrounding the villages of Stithians, Perranwell and Ponsanooth south of Truro, Cornwall was undertaken 1996-2003. Ferns were adopted as particularly good biomonitors of the overall ‘health’ of the hedges concerned. ‘High pteridological value’ was assessed as ten or more individuals of ferns present per 10 m of hedge (Plate 13). TABLE 1. GROUND-LEVEL SURVEY OF FERN DIVERSITY VALUE ACROSS CELTIC HEDGES THROUGHOUT THREE PARISHES IN MID-CORNWALL, 1996-2003 Class length surveyed (m) length found to be of high pteridological value (m) % A Roads (Heavy Passing Traffic) 33,400 1,050 3-14 B Roads (Moderate passing Traffic) 21,200 4,200 19-81 Minor and Unclassified Roads, Bridle Tracks and paths 117,200 65,400 55-80 The diversity of ferns on those hedges with least passing traffic may reflect the levels of diversity once found generally in the country. There can thus be little doubt from this survey that along well-trafficked roads, there is already a substantial and significant decrease in pteridological biodiversity value in the adjacent hedges. In the latter, such hedges change particularly to ones more often dominated by irregular growths of grass and often much bare mud-splashed stone and earth (where more illuminated) or similar with intermittent blanketing curtains of ivy ( Hedera helix ) and green alga-covered earth surfaces (where more shaded). These latter plants tolerate 98 ATLANTIC ARC Figure 2 {above) and Figure 3 {opposite). The modern agricultural tapestry of the land showing coastal field systems: oblique aerial views of the extensively rural and typically highly agriculturalised landscape of the Lizard Peninsula, south-west Cornwall, showing the general abundance of small field systems separated by numerous Celtic Hedges of earth-and-stone (many of which here may date from Mesolithic/Bronze Age origins): Above : the area surrounding Poldowrian settlement SW7416 c. 8 km NE of Lizard Point, as seen from from c. 12,000 ft. altitude looking NE, 8 June 1995; opposite : same area adjacent to Borgwitha Farm SW7516, as seen from from c. 5,000 ft. altitude looking NNW, 8 June 1995. Note the frequency of hedges surrounding fields and along tracks, their overall mostly tree-bare character in this windswept location (cf. Fig. 4), the infrequent occurrence of patches of surviving woodland and scrub (usually limited to steeper- ground along valley sides), bracken-and-gorse covered clifftops on which many fields extend nearly to the edge, and coastal cliffs with rather precariously-suspended cliff-path for scale. Photos : Steve Hartgroves, Historic Environment Sendee, Environment and Heritage, Cornwall County Council: negatives F44/123 (above) and F44/1 15 (opposite). airborne pollutants, whereas ferns do not, and collectively show the degree to which this profound floristic change is a direct consequence of the effects of passing traffic through a combination of direct vehicle-exhaust pollution, mud-hurl, and passing-vehicle air-turbulence. It is particularly noteworthy, in these respects, that where such fern-decimated main-road Celtic hedges follow, in places, slightly wandering courses which bring them occasionally further from their modem road edge, then where such hedges achieve distances of as little as 3 m away from the edge of the road, then their fem component typically begins to show local better survival again. Where distances of as little as 5-6m from the immediate road-edge occur, ferns can often become appropriately more extensive and sometimes fairly luxuriant again, approaching that of the original diversity. Ferns can thus be used as effective traffic biomonitors. I occasionally see Highways Department teams in fluorescent jackets and hard-hats laboriously scoring the number of vehicles passing on various Cornish roads to assess traffic density when they happen to be there, and say to myself, all you have to do is count the number of ferns as an existing accurate recurrent year-round bio¬ assessment! CELTIC HEDGES AND FERN DIVERSITY 99 CONCLUSIONS The Celtic hedge as presented here is certainly an ancient feature of the landscape of so many areas of the Atlantic Arc. These structures, although initially man-made, were constructed largely simultaneously with the clearance of the landscape around them for agriculture, and have persisted with an unbroken history of continuity of environmental stability as well as agricultural purpose to the present day. From Neolithic times onward, through the processes of change of the original wildwood landscapes from pre-agricultural settings to modem agricultural conversions, Celtic hedges have come to provide particularly vital pteridological habitats. Two especially important elements of this are that these hedges have extensively functioned to provide vital island-like refuges for survival of probable ancient fem diversity while also, equally importantly and also like islands, have continued to provide sites for the natural processes of fem evolution to constantly and steadily recur. In the case of agriculture in Cornwall, this history has been over the course of at least the last 4,000 years. Today we need to minimise for the future the steady process of wholesale losses of these Celtic hedges which have especially taken place over the last 50 years, mainly by complete mechanical removal of those between many fields. This has been driven by the traditional small size of Celtic fields and a perceived need for a steady increase in field-size for ever larger machinery access in the name of enhancing agricultural efficiency (the ‘prairie’ syndrome). But prairies do not work in the characteristically steeply-graded landscape of Cornwall, for these processes also lose much of the natural protection of the fabric of the landscape for which the hedges were laboriously constructed to preserve in the first place. Re-opened to exposure and especially when ploughed, such sites can be prone to wholesale surface soil-erosion combined with downhill soil-creep, particularly during heavy winter rains. Consequences are not just of soil loss from the fields (as readily demonstrated by the colour of the run-off water under such circumstances), but also of 100 ATLANTIC ARC Figure 4. Analysis of likely ages of origin of some typical inland field systems: oblique aerial view of the landscape of inland mid-Cornwall near Carrine Common c. 3 km SW of Truro and 2 km NNW of the village of Carnon Downs, as seen from from c. 3,000 ft. altitude looking NNW, 8 June 1995. It shows the general abundance and different sizes of small field systems with ages of landscape use which can be established from known dates of origins of settlements plus additional archaeological features: the three farm settlements (1. to r. Hugus, Chygoose and Gooderne) are all first known to have been mentioned between 1300-1500 AD, while the unusual circular hedges in the middle foreground enclose a barrow of prehistoric origin (Gooderne Round, 17 m wide, and one of a group of 10 nearby barrows) lying on a platform which has been enclosed. The hedges thus date from many different periods, but are probably at least Medieval and may be much earlier, and have almost certainly a continuity of vegetation cover since original construction. Note the much more tree-clad aspect of the hedges in this more sheltered region of Cornwall, which is reflected in different fern components, here dominated especially by Polystichum setiferum and Phyllitis scolopendrium , with some Dryopteris filix-mas, Athyrium filix-femina and Blechnum spicant , while several of the damper adjacent woodland hollows also contain Osmunda regalis. By contrast of dates, running nearly east-west through the background of the picture is the main London-Penzance railway line dating from 1869, on the southerly embankment of which scrub of mainly gorse and oak has established. Photo : Steve Hartgroves, Historic Environment Service, Environment and Heritage, Cornwall County Council: negative F44/31. steady downstream build-up of earth, soil and especially stones in stream and river beds causing increased repetition of downstream flood-spread episodes during subsequent storm-surge discharges. To the end of protecting the traditional structure of the landscape and its tapestry of hedged fields, efforts have been made, at least both in Devon and Cornwall, to officially recognise and celebrate the important function as well as established biodiversity value of the traditional hedge structures. This has been achieved largely through the establishment of various dedicated bodies and working groups. Such a group in Cornwall is that of the Cornwall County Council hedge working group. Unofficially known as the ‘sitting-on-the-hedge committee’, this group is, however, an active one, which has particularly benefited from the outset by containing a broad CELTIC HEDGES AND FERN DIVERSITY 101 TABLE 2. TOTAL NUMBERS OF PTERIDOPHYTE SPECIES RECORDED ON OR IN ASSOCIATION WITH CORNISH CELTIC HEDGE HABITATS Species and Hybrids Hedge Habitats Equisetum an’ense Equisetum palustre Equisetum fluvicitile Osmunda regalis *Polypodium vulgare *Polypodium interjectum Polypodium x mantoniae Pteridium ciquilinum *Phyllitis scolopendrium * Asplenium adiantum-nigrum Asplenium obovatum Athyrium filix-femina *Polystichum setiferum Polystichum aculeatum Polystichum x bicbiellii Dryopteris filix-mas Dryopteris x complexa Dryopteris ciffinis Dryopteris aemula Dryopteris dilatata Blechnum spicant Occasional at bases of hedges Occasional in adjacent ditches Sometimes in local water impoundments by hedges Water impoundments, especially where stream-fed Widespread hedge top species, especially above 100m Abundant hedge-top species, especially near coasts Frequent around contact-zone of parents Frequent in hedges beside arable fields Abundant and widespread on shadier faces of hedges Abundant colonist of better-lit faces of hedges On well-vegetated hedges, more locally than the last Widespread on more shaded hedges near damper bases Abundant in association with Phyllitis scolopendrium Local on a few hedges in extreme east Cornwall Very local only with its P. aculeatum parent Widespread especially on more shaded hedges Local on hedges but perhaps under-recorded Widespread, especially where sunny and sheltered Moist shady hedges locally, especially mid-Cornwall Widespread in more sheltered locations Widespread especially on more shaded hedges * Especially widespread members of the typical ‘guild’ of Cornish Celtic hedge ferns cross section of farmers, highways, landscape and regional consultants, district planners, archaeologists and biologists. Keeping farmers involved in discussions has proved important, while significantly, in landscapes as old as those of Cornwall, biological and archaeological value of these features may closely coincide. Recently I have seen many hedgeless fields in Brittany, and I am told that here large-scale hedge removal has been driven by agricultural need to increase field size and by funding made available for this purpose. Within Cornwall, much discussion has focussed on establishing desirable hedge management regimes (Meneer 1994 & pers. comm.) for maximisation of their wildlife value. In response to the demonstrated wildlife value of these hedges, a locally-available colour-brochure has now also been produced by Cornwall County Council (Meneer & Page 2000). Today, Celtic hedges have come to provide vital pteridological habitats, which are important continuing research sites and provide highly under-researched natural field laboratories throughout much of the Atlantic Arc. At least in Cornwall, they are now largely protected (at least from wholesale removal) and are collectively well respected elements of our ever-evolving landscape. ACKNOWLEDGMENTS I am grateful for useful discussions and input of many fragments of ideas with many local colleagues in Cornwall, particularly Mary & Tony Atkinson, Ian Benallick, Jacqueline Davey, Dr Colin French, Paul Gainey, Steve Hartgroves, Colin Hawke, Dr David Holyoak, Dr Loveday Jenkin, the late Major Walter Magor, Paul McCartney, Robin Meneer, Rose Murphy, Sarah Myles, Catriona Neil, Clare Page, Rosemary Parslow, Adrian Spalding, Matt Stribley and Yvonne Vivian. I am also grateful to Cornwall Archaeological Unit, Cornwall Wildlife Trust, Cornwall and the Isles of Scilly federation of Biological Recorders, the Royal Institute of Cornwall and the Royal Geological Society of Cornwall for help with background information, and to Cornwall County Council for support of the objectives of this account via their Cornwall Hedge Group. 102 ATLANTIC ARC REFERENCES French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. IviMEY -COOK, R. B. (1984). Atlas of the Devon Flora. The Devonshire Association, Exeter. JALAS, J. & SUOMINEN, J. eds. (1972). Atlas Florae Europaeae. Distribution of Vascular Plants in Europe. 1 Pteridophyta. Committee for Mapping the Flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. Jermy, A. C., Arnold, H. R, Farrell, L. & Perring, F. H. (1978). Atlas of Ferns of the British Isles. The Botanical Society of the British Isles and The British Pteridological Society, London. MENEER, R. (1994). Wildlife Revival in Cornish Hedges. Dyllansow Truran Publications, Redruth. MENEER, R. & PAGE, C. N. (2000). Wildlife in Cornish Hedges. (Brochure). Cornwall County Council, Truro. MURPHY, R. J., Page, C. N. & Parslow, R. (in press). An Atlas of the Ferns of Cornwall and the Isles of Scilly. Cornwall and the Isles of Scilly Federation of Biological Recorders, Truro. PAGE, C. N. (1978). Ferns as taxonomic tools and the future of pteridology. Transactions of the Botanical Society of Edinburgh 42: 37-41. PAGE, C. N. (1988). Ferns: Their Habitats in the British and Irish Landscape. Collins New Naturalist No. 74, London & Glasgow. Page, C. N. (1991). Past and present possible occurrence of the rare fern xAsplenophyllitis microdon (T. Moore) Alston in Cornish hedgebanks. Botanical Cornwall 5: 31-33. PAGE, C. N. (1997a), The Ferns of Britain and Ireland. Cambridge University Press, Cambridge. Page, C. N. (1997b). Pteridophytes as field indicators of natural biodiversity restoration in the Scottish flora. Botanical Journal of Scotland 49: 405-414. PAGE, C. N. (2001). Ferns and allied plants, in HAWKSWORTH, D.L. The Changing Wildlife of Great Britain and Ireland , pp. 50-77. Taylor & Francis, London. Page, C. N. (2002a). Ecological strategies in fern evolution: a neopteridological overview. Review of Palaeobotany and Palynology 119: 1-33. PAGE, C. N. (2002b). The role of natural disturbance regimes in pteridophyte conservation management. Fern Gazette 16: 284-289. Page, C. N. (2003). Adaptive ancientness of vascular plants to exploitation of low-nutrient substrates - a neopteridological overview, in HEMSLEY, A. R. & POOLE, I. The Evolution of Plant Physiology, pp. 447- 465. Elsevier, Amsterdam & Oxford. Page, C. N. (2005a) Fern range determination within the Atlantic arc by an environment of complex and interacting factors, in Leach S. J., Page, C. N., Peytourau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc., pp. 59-64. Botanical Society of the British Isles, London. PAGE, C. N. (2005b) Re-appearance of the rare intergeneric hybrid fern xAsplenophyllitis jacksonii Alston (Aspleniaceae) in the flora of Cornwall. Watsonia 25: 331-338. Page, C. N. (in press a). The ecology of long-term evolutionary survival - what can be learned from Pteridophyta ? Fern Gazette (in press a). Page C. N. (in press b). Why record ferns? Pteridologist (in press). Page, C. N. & Mchaffie, H. S. (1991). Pteridophytes as indicators of landscape changes in the British Isles in the last hundred years, in CAMUS, J. M. (Ed.). The History of British Pteridology 1891-1991 . pp. 25- 40. British Pteridological Society, London. PAGE, C. N. & Murphy, R. J. (2001 ). Polypodium x shivasiae - a rare hybrid. Botanical Cornwall 10: 3 1 — 40. Pollard, E., Hooper, M. D. & Moore, N. W. (1974). Hedges. Collins New Naturalist, London & Glasgow. Preston, C. D. & Arnold, H. R. (2005). The Mediterranean-Atlantic and Atlantic elements in the Cornish flora, in Leach S. J., Page, C. N., Peytourau, Y. & Sanford, M. N. (Eds.) Botanical Links in the Atlantic Arc, pp. 41-57. Botanical Society of the British Isles, London. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. RACKHAM, O. (1986). The History of the Countryside. J. M. Dent, London. 103 The task of the Conservatoire Botanique National de Brest in the knowledge and conservation of the Armorican flora as illustrated by Trichomanes speciosum Willd. S. LORIOT*1 and S. MAGNANON*2 Consen’atoire Botanique National de Brest, 52 allee du Bot 29200 Brest, France ABSTRACT The Conservatoire Botanique National de Brest has established a large programme to create an inventory of the flora of the Armoricain massif, with the aim of improving the knowledge of the distribution of taxa and to establish methods to preserve the endangered flora. Analysis of the data gathered has given rise to various approaches, and in particular the programme ‘Atlas of the Armorican flora’ has already allowed the identification of diverse regionally threatened ‘red’ species and the establishment of prioritised lists of taxa for conservation. Trichomanes speciosum Willd., a fern listed in Annex 2 of the Habitats Directive, is one of the taxa whose conservation is judged to have high priority by the Conservatoire Botanique National de Brest. A regional Action Plan was prepared in 2001 for this species, which identifies the need for ecological, biological, physiological and genetic in-depth studies, from which to establish appropriate measures for protection. Keywords: Killarney Fern, action plan, Armorican flora, Brittany. INTRODUCTION The Conservatoire Botanique National de Brest was established in 1975 with the aim of saving the most endangered plant species throughout the world. This original mission has widened, especially since 1990, when the Conservatoire Botanique received the status of Conservatoire National from the French Ministry of Environment. From then on, the Brest C.B.N. has been working on two levels: on an international scale for the preservation of threatened species on oceanic islands, and on a local scale, in an area covering the bio geographic entity of the Massif Armoricain. This area is made up of three administrative regions (Brittany, Pays-de-Loire and basse-Normandie, see Fig. 1). Within this area, four missions have been assigned to the Brest C.B.N.: • to know the flora of the Massif Armoricain, • to protect the threatened species, • to inform public authorities, and carry out specific studies for them • to educate and increase the awareness of the general public. This paper aims at showing how the Brest C.B.N. connects the first mission (to know plants) with the second mission (to protect plants) in the particular case of Trichomanes speciosum , a species mentioned in Annex 2 of the “Habitats Directive’Mt is in this context that a conservation plan has been presented to the regional authorities of Brittany and accepted by them in 2001 (Magnanon & Foriot 2001). In this conservation plan, several actions were proposed to improve the knowledge concerning the biology of this species and to protect the sites where the fern lives, especially in old wells. ^e-mail: loriot@univ-brest.fr *~e-mail: s.magnanon@cbnbrest.com 104 ATLANTIC ARC Figure 1. The assigned area of the Conservatoire Botanique National de Brest : the Massif Armoricain. BACKGROUND Trichomanes speciosum (Killarney Fern) is restricted to Europe and the islands of Macaronesia (Madeira, Azores and the Canary islands). It extends from Portugal in the south to Poland in the east, where the species has recently been found (Krukowski et al 2002), and to the British Isles, where the species reaches its north-western limit. In France, T. speciosum has been recorded from three main disjointed regions: the Basque Country (Jovet 1933), the Vosges mountains (Jerome et al. 1994) and Brittany (Louis-Arsene 1953). T. speciosum grows in moist habitats sheltered from draughts and generally shaded. For example, in the British Isles the fern occurs in deep wooded valleys near running water or small waterfalls, in rocky overhangs and caves (Ratcliffe et al 1993). Like most ferns, T. speciosum has a life cycle with alternately two successive generations (Fig. 2). The sporophyte, with 2n chromosomes, is the generation with fronds producing sporangia which release spores after meiosis. The germination of the spores results in the formation of a filamentous gametophyte, n chromosomes, which bears the gametangia: antheridia and archegonia. Fertilisation between an antherozoid produced by the antheridia and the oosphere of the archegonium produces a zygote, 2n, the first cell of a new sporophyte individual. An enigmatic phenomenon seems to occur in Brittany: populations with fronds (sporophytes) have been found only in old wells, but never in natural habitats (Prelli 1992). The insides of these wells are characterised by obvious dampness, freshness and also darkness. In such artificial habitats, the prothallian stage has also been found associated with the sporophyte or alone as independent gametophyte populations. Unlike the fronds, gametophytes of T. speciosum are found CONSERVATION OF TRICHOMANES SPECIOSUM IN BRITTANY 105 FIGURE 2. Trichomanes speciosum Willd.: life cycle showing the two successive generations. in natural habitats. Indeed, in Brittany, independent gametophytes have been recorded in rocky crags of fresh and damp forested valleys and woodland outcrops. Gametophytes of T. speciosum , in the absence of sporophytes, were also observed in numerous coastal caves or cracks never reached by the sea at high tide but characterised by fresh water, temperate temperatures and low light levels. The occurrence of “independent gametophytes” is not peculiar to T. speciosum. It was defined by Farrar (1967) who described the phenomenon in three fern families, Grammitidaceae, Vittariaceae and Hymenophyllaceae. In such cases, the gametophytic populations are perennial, rather than only existing for the time needed to produce another sporophytic generation. So. there is no more alternation from the gametophyte to the sporophyte because populations of gametophytes have been able to reproduce and thus persist only vegetatively by releasing gemmae (Farrar 1967). Such populations were first described for T. speciosum in England in 1989 (Rumsey et al. 1990) and in Brittany in 1990 (Prelli 1992). In Brittany the species is now threatened by habitat disturbance, that is to say massive change in the natural habitats for populations of gametophytes, like the destruction of vegetation cover and destruction or closing up of old wells with populations of sporophytes (Boudrie, in Olivier et al. 1995). The Brest C.B.N. had problems setting up a realistic conservation plan for the fern in Brittany because its life cycle was too poorly known in the region. More information on the ecology, reproductive biology and physiology of T. speciosum in wells and in natural habitats will provide better understanding of the factors involved in the apparent life cycle disruption and in the separation of niches. 106 ATLANTIC ARC FIGURE 3. Distribution of Trichomanes speciosum Willd. in Brittany. A localities with at least one population of independent gametophytes, • localities with at least one population of sporophytes. RESULTS First, field work was performed consisting of descriptive site forms, compiling information about the presence of T. speciosum gametophytes or sporophytes, the abundance of both generations (small cushion of gametophytes or large carpet, number of fronds), habitat typology (wells, outcrops and rocky crags, caves), ecological data (climate data, geology, ventilation, plant communities) and the present or expected threats. Thirteen sites were selected covering a large geographic distribution area and most of the habitat typology. These sites have now been regularly visited each month since last summer. Three factors that seem to be of importance for the development of sporophytes (Ratcliffe et al. 1993) were measured at each visit: temperature, relative humidity of the atmosphere and illumination. The possibility of ecophysiological differences between the wells with sporophytes and the natural habitats where only independent gametophytes grow in Brittany was investigated. Samples were collected for observations and laboratory studies at the Universite de Bretagne Occidentale (Brest). Field work to check known sites for independent gametophyte populations has resulted in 56 sites known in Brittany today (Fig. 3). There are nearly as many independent gametophyte populations known in wells as there are from natural habitats (Table 1). But above all, gametophytes have always been found in wells where sporophytes are present; this contradicts literature accounts of the absence of gametophytes in wells where sporophytes occur. The CONSERVATION OF TRICHOMANES SPECIOSUM IN BRITTANY 107 TABLE 1. HABITAT TYPES FOR GAMETOPHYTES AND SPOROPHYTES OF TRICHOMANES SPECIOSUM RECORDED IN BRITTANY IN 2003 Habitats Independent gametophyte Sporophyte Old wells 25 wells 32 wells Natural habitats boulders: 14 sites coastal caves: 15 sites 2 sites Others artificial habitats 1 former quarry 1 small cellar - Total 56 sites 34 sites abundance of gametophytes is particularly variable ranging from a cushion less than 1 cm wide at the bottom of fissures in rocky outcrops to a carpet more than 1 m2 in most of the coastal caves. There has been a dramatic decrease in the number of sites with fronds (sporophytes): there are now only 32 wells sheltering them out of 178 in the 1950s. Populations are composed of ten to, more rarely, a thousand fronds. But, very recently, two sporophytic populations have been found in natural habitats (Poux et al. 2003). This may be an exciting discovery for the flora of Brittany, but the identity of the species remains questioned because of the non-typical fronds, smaller and less denticulate, compared to those found in wells. Regular observations have enabled the Brest C.B.N. to see that spore production occurs in the wells. But it seems that most of the increase in the number of fronds occurs due to vegetative growth of the rhizome with the emergence of new fronds. Biochemical viability tests with Trypan, Blue Evans and TTC of the spores have been successfully undertaken and further tests to obtain the germination and gametophyte development from spores are planned. Numerous gemmae were always observed on the gametophytes (Fig. 4a). This shows intensive vegetative reproduction. Sexual reproduction structures were also occasionally observed in the independent gametophytic sites (Fig. 4b, 4c). Additionally, we found some structures resembling sporophytic lamina in the independent gametophyte populations (Fig. 4d). These were comparable to those found in the British Isles (Sheffield 1994). These sporophyte-like structures were less than 1 cm long and may be juvenile individuals. Their scarcity could result from the rarity of the sexual reproduction structures or to a short lifetime caused by some physiological deficiency. During previous field work, temperature and relative humidity of the atmosphere were shown to be important factors, but not as limiting as light seems to be (results not presented here): some differences were found between the available light in wells where sporophytes grow and that available in the independent gametophytes sites. Recordings in one of the two Breton natural sites where sporophytes have been found, showed that in winter, when there is less cover from vegetation, the available light increased to reach a maximum in March. Concurrently the length of the fronds increased to 5 to 12 cm on average. The atypical fronds described above were confirmed to be juvenile. Laboratory cultivation tests are in progress to study the influence of different light levels on sporophyte emergence and growth. KNOWLEDGE OF THE ARMORICAN FLORA AND STRATEGIES FOR ITS PROTECTION The programme of survey of the Armorican flora, of which this study on Trichomanes speciosum forms a part, was launched in 1991, with the objective of better recording the distribution of the flora and building up a tool for the protection of endangered species. This programme involves the collaboration of a network of 300 volunteer field recorders, and a standardised method for recording field data. Botanists collect information on specific field recording sheets, with maps attached, on which they record the location of the sites of every species they can identify, as precisely as possible. The data are then gathered at the Brest C.B.N. and compiled into a Geographic Information System. Using these data it is possible to produce maps of the flora for each “department”: draft atlases for several of them (Finistere, Morbihan, Loire- Atl antique, Vendee, Cotes d’ Armor, Ille-et-Vilaine) have already been produced. In exchange for their contribution, the volunteer field recorders 108 ATLANTIC ARC Figure 4. Trichomanes speciosum Willd. in Brittany. 4a: Gemma still attached to the gemmiferous cell of the gametophyte SEM (x 900). FIGURE 4b Male gametangia: antheridium (a): SEM (x 600). CONSERVATION OF TRICHOMANES SPECIOSUM IN BRITTANY 109 Figure 4c Female gametangia: archegoniophore (ar), bearing several archegonia (<-) SEM (x 600). Figure 4d Tiny sporophyte (s) less than 1 cm long found within a gametophytic population (g) (xl2). 110 ATLANTIC ARC FIGURE 5. Well sheltering the fern Trichomanes speciosum secured by a grate. {Petit Pelan, Hellean in Brittany, France). receive an annual field meeting programme, and the journal E.R.I.C.A. issued by the Brest C.B.N. These activities and publications improve the cohesion of the network. While they contribute to the continuous training of members, they also come as a compensation for the time spent on the project. In addition to the atlases at different “department” levels, a general atlas of the flora for the whole Massif Armoricain will be published soon. Information about the status of species present in the region will allow: • better knowledge about rare and threatened species, • priority setting, and thus, • determination of the most appropriate actions for the preservation of these species. However, it is not possible to wait for perfect and complete knowledge of the status of all species before starting work on protection, especially if a species seems to be under threat. So, even before the publication of the atlas of the Armorican flora, the Brest C.B.N. has initiated a process to determine priorities for protection: based on a red list of 502 Armorican endangered species produced in 1993 (Magnanon 1993), a second list of priority taxa was established in 1997 (Magnanon & Hardy 1999). This priority red list includes 250 species which : • are either extinct or currently present in a single “department” of the Massif Armoricain, • have disappeared from over half of the “departments” where they had been present. CONSERVATION OF TRICHOMANES SPECIOSUM IN BRITTANY 111 • are currently present in less than 10 sites of the Massif Armoricain. For all these species, the Brest C.B.N. aims at drawing up a conservation plan for the long term preservation of the sites where they are present. This cannot be done without technical and financial support from administrative public authorities and organisations. We must then proceed step by step, and collect the necessary subsidies to implement these conservation plans, which is not always an easy thing! PROTECTION POLICIES ALREADY IMPLEMENTED FOR TRICHOMANES SPECIOSUM In the case of T. speciosum , because we now know the dramatic consequences of closing wells, the Brest C.B.N. has to take urgent measures to protect such sites. In wells that have been hermetically closed with board or corrugated iron sheets, a withering process occurs, followed by the disappearance of fronds (sporophytes), with only the maintenance of the rhizome as a sorry proof of the past presence of the species. It appears that lush sporophytic populations can only grow in wells made safe by a grate. So the Brest C.B.N. has set up a campaign to heighten well owners’ awareness by personal meetings followed up by an informative leaflet about the fern and actions needed to preserve it in Brittany. Grates were ordered for each well and fixed by the engineering department of the town (Fig. 5). The grates will secure the wells (for people and animals) while allowing water and light supplies for the maintenance of T. speciosum in Brittany. CONCLUSIONS Studies carried out on T. speciosum have shown the factors involved in the formation of sexual reproduction structures and in the development of adult sporophyte populations. The genetic variability of T. speciosum in Brittany has also been studied: this knowledge will guide the Brest C.B.N. for the production of a long-term conservation strategy. This example illustrates the way the Brest C.B.N. wants to carry out its duties on the conservation of the threatened flora in the Massif Armoricain, by linking work to improve knowledge about species with the implementation of means of protection. In Brittany, the Brest C.B.N. has received funds from local authorities and from the government for a 6-year programme to look at the protection of 37 rare or endangered taxa of the region. For each of these priority species, preservation plans based on the following schemes are drawn up: In the first part, the global situation of the species is described, especially: • the current distribution of the plant in the Massif Armoricain, in France, and throughout the world • the ecology and biology of the plant • its status of protection • threats existing on the sites where it is present • protection programmes already in force The second part of the plan consists of proposing actions for protection: • improvement of data collection concerning the plant (distribution, biology, ecology) • suggestions for actions to protect existing populations (information for the public, legal protection and management of sites) The Brest C.B.N. believes that the progress made so far on Trichomanes speciosum illustrates the success of this programme very well. ACKNOWLEDGMENTS We want to thank MM. Pierre Mens, Yves Peytoureau and Alain Cottignies for their help in writing this paper. 112 ATLANTIC ARC REFERENCES Farrar, D. R. (1967). Gametophytes of four tropical fern genera reproducing independently of their sporophytes in the Southern Appalachian. Science 455:1266-1267. JEROME, C., RASBACH, H. & RASBACH, K. (1994). Decouverte de la fougere Trichomanes speciosum ( Hymenophyllaceae ) dans le Massif Vosgien. Le Monde des Plantes 450. JOVET, P. (1933). Le Trichomanes radicans (Sw.) et V Hymenophyllum tunbridgense (Sm.) au Pays Basque fran^ais. Bulletin de la Societe Botanique de France. Krukowski, M., & Swierkosz, K. (2002). Sandstones, landscapes: diversity, ecology and conservation.. IUCN European Program Newsletter Central and Eastern Europe. Dec 2002/32C45. LOUIS-ARSENE, Fr. (1953). Les stations de Trichomanes speciosum dans la region de Ploermel. Bulletin de la Societe Botanique de France. Magnanon, S. (1993). Liste rouge des especes vegetales rares et menacees du Massif Armoricain; Conservatoire Botanique National de Brest. Erica 4: 1-22. Magnanon, S. & Hardy, F. (1999). Strategic integree de conservation des taxons les plus rares et les plus menaces du Massif Armoricain. Bulletin de la Societe Botanique de Centre-Ouest , No. special: Les plantes menacees de France, Actes du Colloque de Brest 15-17 octobre 1997; 355-378. MAGNANON, S. & Loriot, S. (2001). Strategie d’actions prioritaires pour la presentation d'une espece vegetate a tres forte valeur patrimoniale, Trichomanes speciosum. Rapport Conservatoire Botanique National de Brest, Conseil Regional et D.I.R.E.N. de Bretagne. OLIVIER, L., Galland, J. P. & Maurin, H. (1995). Livre Rouge de la Flore Menacee de France Tome I: Les especes prioritaires. Ministere de l’Environnement. Museum National d’Histoire Naturelle. Conservatoires Botaniques Nationaux. Paris. Poux, L, Philippon, D. & Prelli, R., (2003). Bilan des decouvertes interessantes de l’annee 2002 dans les Cotes d’ Armor. Erica 17: 81. PRELLI, R., (1992). Atlas ecologique des fougeres et plantes alliees. Illustration et repartition des Pteridophytes de France, ed Lechevallier. Ratcliffe, D. A., Birks, H. J. B., Birks, H. H. (1993). The ecology and conservation of the Killamey Fern Trichomanes speciosum Willd. in Britain and Ireland. Biological Consen’ation 66: 231-247. Rumsey, F. J., Sheffield E. & Farrar D. R. (1990). British filmy-fern gametophytes. Pteridologist 2: 40- 42. SHEFFIELD, E. (1994). Alternation of generation in ferns: mechanisms and significance. Biological Reviews 69: 333-343. 113 The Isle of Oleron orchids M. BRERET 8 rue P. Cezanne 17138 Saint-Xandre, France ABSTRACT Oleron is an island with many assets: it is a migratory bird sanctuary in winter; it has oyster farms galore; it is bustling with tourists in summer, especially now it is connected to the mainland by a bridge; and for botanists, it has a diversified flora to offer, in particular orchids and several Mediterranean plants. KEYWORDS: Mediterranean- Atlantic, agricultural decline, forest dunes. INTRODUCTION The Isle of Oleron is situated off the Atlantic shore of the department of Charente-Maritime (Fig. 1), between the mouth of the Seudre River in the south and that of the Charente River in the North. Separated from the Charentais coast by the Pertuis d’Antioche in the north-east, the Figure 1. Location of lie d’ Oleron. 114 ATLANTIC ARC Figure 2. Simplified geological map. Coureau d’Oleron in the east and the Pertuis de Maumusson in the south, the Isle is the second biggest French island by size (175 km2) after Corsica. Oriented north-west to south-east along the Gemozac anticlinal line - an anticline formed by the post-Cretaceous tectonics contemporary with the Pyrenean fold - it stretches 32 km long and 10 km wide (highest point: 34 m) and is only 1 km from the mainland at the Pertuis de Maumusson. Its substratum is made up of chalky soils of the Mesozoic which a fault splits into two periods: 1 . Upper Jurassic in the north-east. 2. Lower Cretaceous in the south-west. Quaternary alluvia form the gley (bluish-grey sticky clay) and the dune bars. The strong erosion which the island was submitted to caused the disappearance of Tertiary soils, except some traces of Pliocene soils (Fig. 2). Two biggish forests grow on the former stable dunes: the Saint-Trojan State forest in the outh- est and the Saumonards State forest in the north-east, mostly planted with Maritime Pine ( Pinus pinaster ) and Holm Oak ( Quercus ilex). Numerous small woods growing on the limestone substratum in the centre of the island are composed of Holm Oak and deciduous trees. The island also has lots of more-or-less salty marshes, representing the remains of former salt marshes in the lower clayey areas. Lastly, the decline of farming zones creates fallow lands that suit the growth of orchids. Benefiting from the currents of the Gulf Stream, the island has an Atlantic climate mild in winter, just fairly damp in summer, and a marked period of sunshine. These conditions are ISLE OF OLERON ORCHIDS 115 reminiscent of the Mediterranean climate and the flora is unexpectedly rich and unique. Many species bear witness to the mild climate: four Cistus species, - C. salviifolius (and its parasite Cytinus hypocistis), Cistus monspeliensis, C. laurifolius, C. psilosepalus, and two hybrids: C. salviifolius x C. psilosepalus = C. x obtusifolius , and C. salviifolius x C. monspeliensis = C. x florentinus , Daphne gnidium, Osyris alba, Scorpiurus muricatus subsp. subvillosus, Quercus ilex, Pinus halepensis and Pinus pinea to mention but a few Mediterranean ones. Paradoxically, the presence of mountain plants such as Pyrola chlorantha and Arctostaphylos uva-ursi remains unexplained! All these conditions and biotopes make lie d’ Oleron an exceptional island for many orchids. FALLOW LANDS AND MOORS When you reach lie d' Oleron, the countryside in front of you is made up of numerous fields lying fallow. This is the result of a decline in agriculture, particularly vineyards in the southern part of the island, and of the disappearance of cattle breeding as a result of tourist pressure. Little by little, the fallow lands are becoming overgrown with bushes, especially Blackthorn ( Primus spinosa ), Wayfaring Tree ( Viburnum lantana), Privet ( Ligustrum vulgare ) and Dog-rose bushes ( Rosa canina) which form moors. Those areas located on clayey-chalky soils, damp in winter and dry in summer, provide good habitat for numerous species of Ophrys , mostly Mediterranean, which occur here at the northern limit of their range in Europe. Bee Orchid Ophrys apifera Hudson, frequent in France, especially in our region, is also to be found on short lawns. Woodcock Orchid Ophrys scolopax subsp. scolopax Cavanilles, only to be found in the southern half of France. Ophrys passionis J. and P. Devillers-Terschuren. a taxon badly known for a long time, is only present around the Mediterranean and some departments in the West of France. This Mediterranean species may be confused with Early Spider Orchid Ophrys sphegodes Miller and can be distinguished from it by its wider, more colourful petals and its darker labellum and lower lip (Plate 14). Ophrys sulcata J. and P. Devillers-Terschuren. looks like a small Brown Bee Orchid Ophrys fusca Link. It is only to be found in a few departments of southern and south-western France and on the Atlantic coast. Present in Italy and Spain, it is also near its northernmost limit although there is one station in Brittany. To finish with Ophrys, the presence of Ophrys vasconica (O. and E. Danesch) P. Delforge (Plate 15), rare in France except in a few south-western departments, remains doubtful. As for Fly Orchid Ophrys insectifera L., frequent in France but for Brittany, it can be found occasionally in Oleron. It has only be seen once, on a recently installed car park. CLAYEY TO CLAYEY-SANDY SOILS Going further north on the island, we reach lower areas of clayey soils containing more or less sand. They are said to be ‘heavy’, and are often flooded in winter and very dry in summer. As early as April, Green-winged Orchid Anacamptis morio (L.) R.M. Bateman. Pridgeon & M. W. Chase subsp. morio (= Orchis morio ) is the first to flower, making beautiful purple displays. It is a very common species. It can also be found on lighter soils (but always very damp in winter). Often found with the former, Tongue Orchid Serapias lingua L. opens a few weeks later. Like most Serapias species, it has a Mediterranean and south-Atlantic distribution. It reproduces vegetatively by producing thin tubers to form compact groups. The colour of the labellum varies greatly from red to pale yellow. Loose-flowered Orchid, or Jersey Orchid Anacamptis laxiflora (Lam.) R.M. Bateman. Pridgeon & M.W. Chase (= Orchis laxiflora), grows on very damp or soaked soils. It is quite frequent in France, but is protected in some regions. Marsh Orchid Anacamptis palustris (Jacq.) R.M. Bateman, Pridgeon & M.W. Chase (= Orchis palustris ), with regional protection, is often found in the same places as A. laxiflora, and hybrids of these two similar species are frequent. Sometimes on its own, it is found in stations of several dozen individuals, as is the case on the commune of Saint-Denis. 116 ATLANTIC ARC Small-flowered Tongue Orchid Serapias parviflora Parlatore, with national protection, is a rare species. It is present in a few southern and south-western departments and is found at two sites on the island. One site, on the Saint-Trojan commune, was partly destroyed by the building of a housing estate, despite its protection! Autumn Lady’s Tresses Spiranthes spiralis (L.) Chevalier, is the last of our orchids to flower (end of August or even early September). It is relatively frequent along the French Atlantic coast, yet it is protected in some regions. It can also be found on short lawns. On the island, it can remain unnoticed, almost solitary or else in quite large populations, like that in Dolus which totalled nearly 100 plants in 2002. WOODED DUNES We now arrive in one of the two State forests, the Saumonards forest. These forests were planted with Maritime Pine in the 19th century, firstly to stabilise the sand and also later as a source of resin. This type of forest gets a lot of sunshine, which makes it possible for several species of orchids to grow. As the forests quickly dry up, springtime is flowering time. Early Spider Orchid Ophrys sphegodes Miller is a widespread Ophrys in France and relatively numerous in some stations on the island. Small Spider Orchid Ophrys araneola Reichenbach is akin to O. sphegodes. It can be distinguished from it by its earlier flowering and its labellum lined with yellow. Man Orchid Orchis anthropophora (L.) All. (= Aceras anthropophorum ) is also common and flowers soon after the previous species. Narrow-leaved Helleborine Cephalanthera longifolia (L.) K. Fritsch can form really white "meadows’ some years. Bird’s-nest Orchid Neottia nidus-avis (L.) L.C.M. Richard, a saprophytic species, is found locally, sometimes in large numbers in the same station. Lastly, in early summer, Green-Flowered Helleborine Epipactis phyllanthes G. E. Smith (Plate 16), with regional protection, can be found in several stations in both forests. It is quite rare in France, growing mainly on the south-western Atlantic coast. The type is slightly variable, according to P. Delforge, the Oleron stations shelter a particular variety, with flowers more widely open than those of the type, which he has named ‘ olarionensis\ In the cooler forests where the Holm Oak grows and where humus is more abundant, can be found: Red Helleborine Cephalanthera rubra (L.) L.C.M. Richard, this pretty pink orchid is present mostly in the south of the island but never in any numbers. It can also be seen on stable dunes. More or less widespread in France, it is absent in a large portion of the west. Greater Butterfly Orchid Platanthera chlorantha (Custer) Reichenbach, is often found on its own. It is also found on moors on the edge of copses. It is common all over France. This is also the case with Pyramidal Orchid Anacamptis pyramidalis subsp. pyramidalis (L.) L. C.M. Richard, a ubiquitous species, locally present in forests, but also in fallow lands and on the edge of woods. STABLE DUNE SANDS Walking out of the forest toward the sea, we reach the stable dune sands. Lizard Orchid Himanthoglossum hircinum (L.) Sprengel, is fond of this dry substratum, especially when rich in chalk and in bright sunlight. It is generally found in clusters wherever this kind of environment is present. It is chiefly in damp depressions, sometimes flooded in winter, that the rarest and most beautiful orchids of the island are to be found. Marsh Helleborine Epipactis palustris (L.) Crantz, regionally protected, is scattered all over France but rare. It is present in several stations on the island, mostly in the south. This species, with a lacy labellum, can be pink or white and often found in large numbers in the same station. Bug Orchid Anacamptis coriophora (L.) R.M. Bateman, Pridgeon & M.W. Chase subsp. fragrans (= Orchis coriophora subsp. fragrans ), nationally protected, is an infrequent southern ISLE OF OLERON ORCHIDS 117 ■ species. So far, it has been found only in two stations: one in the south of the island in Saint- Trojan, part of which has been destroyed when the housing estate mentioned above was built; the other in the North, on the Saint-Denis commune. This species is very dependent on rainfall: the latter station was discovered in 2001, a very rainy year, when 100+ plants were counted. In 2002, when the winter was drier, it was not seen anywhere. Summer Lady’s Tresses Spiranlhes aestivalis (Poiret) L.C.M. Richard, nationally protected, is a Mediterraneo-Atlantic species, always rare. This small orchid with pure white flowers in a spiral on the stem, is found in a few depressions in the Grand-Village commune. As it finds it more and more difficult to grow in high dense vegetation, an arrangement was made between S.F.O. (Societe Frangaise d’Orchidophilie) and some members of O.N.F. (Office National des Forets = the Forestry Commission) aware of the problems of the protection of Nature. It was decided to clear a few depressions as from 1995. Epipactis palustris and Spiranthes aestivalis soon prospered. But above all, Fen Orchid Liparis loeselii (L.) L.C.M. Richard (Plate 17), nationally protected and never before found on the island, was discovered at the bottom a depression in the dunes. It is very rare in France and present in only one station in the Poitou-Charentes region. This entirely green- yellow species is a pioneer plant which cannot bear competition for long. This is a fine example of sensible, reasoned management of the environment. Lastly, in 2002, a new species was discovered by an O.N.F. ranger, between the stable dune and the Saint-Trojan forest: Dense-flowered Orchid Neotinea maculata (Desfontaine) Steam. This Mediterranean orchid creeps up the western coast of France, as witnessed by its recent discovery in Brittany. If its presence in south-western Ireland seems to be reminiscent of pre-ice Period relics, the present global warming affords it with favourable conditions for developing on our coasts. It is the first station in Poitou-Charentes. A FEW HYBRIDS As a finishing touch, let us mention the presence of numerous hybrids, some of them very common like: Anacamptis x lloydiana Rouy (A. laxiflora x A. palustris) Anacamptis x alata Fleury (A. morio x A. laxiflora ) Ophrys sphegodes x O. passionis Ophrys x jarigei Soca ( O . sulcata x O. passionis ) And above all two intergeneric hybrids: xOrchiserapias capitata De Larambergue ex E.G. Camus ( Anacamptis morio x Serapias lingua) xOrchiserapias complicata E.G. Camus (. Anacamptis laxiflora x Serapias lingua). * French editorial note : Owing to recent nomenclatural upheavals in the names of Orchids, it has been necessary to change most binomials... Our most sincere apologies for the inconvenience ! CONCLUSION To date, 25 species of orchids, among which eight have regional or national protection, have been recorded on the Isle of Oleron. Compared with other French islands off the Atlantic coast, Oleron is by far the richest (for example, the Isle of Re, close by, only has ten or so species). In contrast with some regions, in which species are regressing, Oleron has a rich, varied environment which orchids seem to enjoy. Thanks to the climate and its different biotopes, the island has suitable conditions for several Mediterranean species at their northern limits, thus adding to the species normally found in the Atlantic area. Such exceptional richness and bounty delight orchid-lovers; yet, care is needed to avoid damage to this fragile environment, especially from tourist pressure. ACKNOWLEDGMENTS The author is indebted to Michele Dupain, Jean-Michel Mathe, Dominique Pattier for helpful 118 ATLANTIC ARC suggestions on an earlier draft of this paper and to Yves Peytoureau for his suggestions and translation. BIBLIOGRAPHY /\ BOTINEAU, M. et al. (1989). Mini-session dans Pile d’Oleron. Bulletin Societe Botanique du Centre-Ouest 20: 370-377. BOURNERIAS, M. et al. (1998). Les orchidees de France, Belgique et Luxembourg. Parthenope Collection, Paris. BUTTLER, K. P. (1991). Field Guide to Orchids of Britain and Europe. The Crowood Press, Swindon. Champagne, P. (1986). Compte rendu de la sortie botanique dans file d’Oleron: 12 mai 1985. Bulletin Societe Botanique du Centre-Ouest 17: 321-323 Champagne, P. (1988). Compte rendu de l’excursion du 7 juin 1987: Orchidees de Pile d’Oleron. Bulletin Societe Botanique du Centre-Ouest 19: 475^478. Champagne, P. (1995). Orchidees de Pile d’Oleron. L’Orchidophile 68: 880-886. CHAMPAGNE, P. (1997). Liparis loeselii en Oleron. Bulletin Societe Botanique du Centre-Ouest 28: 107-108. Champagne, P. (1998). Compte rendu de la sortie du 4 mai 1997 a Pile d’Oleron (Charente Maritime). Bulletin Societe Botanique du Centre-Ouest 29: 177-178. DAUNAS, S. (1984). These sur la contribution a V etude des plantes mediterraneennes en Charente-Maritime. Faculte de Medecine et de Pharmacie, Universite de Poitiers. DELFORGE, P. (2001). Guide des orchidees d' Europe, d'Afrique du Nord et du Proche-Orient. Delachaux et Niestle, Lausanne. DUPONT, P. (1986). Index synonymique de la flore des regions occidentals de la France. Societe Botanique du Centre-Ouest, Saint-Sulpice-de-Royan. Numero special 8. GUERIN, J. C. , Mathe, J. M. & Merlet, A. (1995) Orchidees de Poitou-Charentes et Vendee. Meloe, Niort. TERRISSE, A. (1994) Inventaire des plantes vasculaires presentes dans File de Re. Societe Botanique du Centre-Ouest, Saint-Sulpice-de-Royan. Numero special 13. 119 An introduction to the flora of the Lizard peninsula, Cornwall J. HOPKINS English Nature, Northminster House, Peterborough PEI 1 UA, England ABSTRACT The Lizard peninsula comprises the most southerly part of south-west England and is widely recognised as having an unusual floristic character. A distinct florule is identified which includes those species which in a national or regional context are rare. The presence of a mild oceanic climate with some Mediterranean characteristics and the unusual geology of the southern part of the peninsula, including ultrabasic serpentine and gabbro rocks, are often seen as reasons for the floristic distinctiveness of the area. Evidence is put forward to support the hypothesis that both geology and climate partially explain the unusual plant geography of the area, whilst the survival of a large proportion of the diverse semi-natural habitat area present at the start of the 20th century is seen as a third important factor. Keywords: Cornwall, The Lizard, vascular flora, plant ecology. INTRODUCTION The Lizard peninsula, Cornwall (Fig. 1) includes the most southerly part of mainland Britain at Lizard Point, and is bounded to the north by the tidal reaches of the Helford River and the freshwater Looe Pool (formerly the sea outlet of the River Cober but cut off from the sea in the late middle ages). Due to the unusual geological and ecological character of the area, it is identified by English Nature as a distinct Natural Area with its own specific conservation objectives (Prosser 1995). Figure 1 . Location map of the Lizard peninsula, Cornwall. 120 ATLANTIC ARC The area is well known due to the richness and distinctiveness of its flora. In the 19th century, Johns (1848) perhaps did more than any other botanist to bring the area to wider attention, but many notable figures in the history of British botany, including Ray and Babington have visited the area. In the second half of the 20th century The Lizard flora was intensively recorded by Dr D. E. Coombe, Dr L. C. Frost, Dr M. H. Martin, the current author, Dr M. Hughes and A. Byfield. Much of this documentation, and that of others, is published by Margetts and David (1981), Hopkins (1983), Margetts (1988), French et al. (1999) and Lawman (1994), although there has been relatively little study of why this distinctive flora exists. An attempt is made here to elucidate some of its causes. CLIMATE AND GEOLOGY Due to its extreme south westerly location The Lizard is notable for the extreme mildness and oceanicity of its climate. The incidence of frost is low, with mean January temperatures of 6-8°C at Lizard village, close to the most southerly point. Summer temperatures are warm but equable due to the influence of the sea with mean July temperature of 15-9°C (Staines 1984). Rainfall at The Lizard is relatively low, in part due to the lack of relief rainfall, the peninsula being a low lying plateau, the highest point only 112 m above sea level. The rainfall pattern is strongly seasonal with January, November and December the wettest, and April and June the driest months (Staines 1984). Particularly on the west coast, The Lizard Peninsula is exposed to strong westerly winds and maritime influence in the form of increased salt deposition can be detected even in the interior of the peninsula (Malloch 1972). Low rainfall, high temperatures and windiness create conditions of high evapo-transpiration in the spring and early summer and this causes frequent local die back of vegetation on shallow soils of steep south facing slopes of valleys and around rock outcrops, due to microclimatic conditions with distinct Mediterranean characteristics. The northern part of the Lizard peninsula, known as The Meneage, is underlain by a complex of largely sedimentary rocks (Bristow 1996; Selwood et al. 1998) which have weathered to give a range of soils also found in other parts of Cornwall and Devon where there is similar geology and climate. Much of the area of the Meneage is given over to pastoral farming (Staines 1984). In contrast the southern part of the peninsula is known to geologists as The Lizard District or Lizard Complex and is one of the most geologically unusual parts of Britain. Modem consensus is that this is an ophiolite, that is a section of ocean floor which was pushed to the surface in the Devonian period, as two tectonic plates collided (Bristow 1996; Selwood et al. 1998). The major part of The Lizard District is underlain by serpentine (also known as peridotite). Also called ultrabasic rock, serpentine is largely magnesium ferro-silicate and weathers to give soils which are base-rich due to high magnesium content, but which have negligible calcium content, an unusual situation as most other base-rich soils are high in calcium and low in magnesium. Serpentine also contains unusually high levels of the heavy metals chromium, vanadium, nickel and cobalt and weathers slowly to give shallow soils. The eastern part of The Lizard District is underlain by gabbro, also an ultrabasic rock type, with high magnesium but richer in calcium than serpentine and lacking the heavy metal enrichment. Gabbro weathers to give soils which are rich in bases but typically deeper than serpentine soils. Both rock types are poor in plant nutrients and have retained a large part of their heathland cover during the 20th century (Hopkins 1983; Staines 1984). Surrounding the ultrabasic rocks is a discontinuous ring of schists formed by metamorphosis of sedimentary rocks and igneous rocks associated with ophiolite development. They lack the high base status of the serpentine and gabbro but have slightly higher potassium status, and are largely exploited by grassland or arable agriculture except where there is steep topography in small valleys and on coastal slopes where semi-natural vegetation survives. In addition to these major rock types there are a range of other igneous and metamorphic rocks of more restricted extent which have only a very small effect upon land use and ecology, including Kennack Gneiss, various dyke complexes, and pillow lavas. In places deposits of loess, and on the gabbro Crousa gravels, overlie the ultrabasic rocks giving rise to patches of acid soils. FLORA OF THE LIZARD PENINSULA 121 THE LIZARD FLORA AND PLANT ECOLOGY Lawman (1994) has compiled a comprehensive list of the Lizard flora based upon the work of A. Byfield. Following the treatment of the British flora by Preston et al. (2002a), and excluding: i. alien species ii. native species outside their natural range iii. critical taxa of the genera Taraxacum and Rubus The list for the Lizard includes 667 putative native taxa. A high proportion of these species are widespread in Britain. There are however a number of features of the Lizard flora which have attracted special interest: a. The presence of a significant number of nationally rare and scare species, including several taxa which are confined to The Lizard on mainland Britain (e.g. Juncus pygmaeus, Herniaria ciliolata, Trifolium incarnation ssp. molinerii (Plate 18)), whilst others have few records elsewhere and are relatively abundant at The Lizard (Wiggington 1999; Stewart et al. 1994). b. The presence of taxa which are abundant at The Lizard but rare or absent in the rest of Cornwall. c. The occurrence of taxa which have unusual ecological characteristics at The Lizard, as for example Juncus maritimus which occurs as a wet heathland plant and Minuartia verna which occurs on shallow serpentine soils (Hopkins 1983). d. Vegetation in which species with quite different soil preferences in the rest of Britain occur in close association, particularly in Erica vagans - Ulex europaeus heath and Erica vagans - Schoenus nigricans heath of the National Vegetation Classification, which cover the largest part of the heathlands (Hopkins 1983; Rodwell 1991). e. A rich assemblage of critical taxa (Margetts 1988), including unusual phenotypes, ecotypes and caryotypes of many relatively common species, such a genetically dwarf forms of Leucanthemum vulgare , populations of mixed prostrate and upright forms of Cytisus scoparius (Hopkins 1983), and tetrapolid Scilla autumnalis (Parker, pers. comm.). THE LIZARD FLORULE An important question to be addressed is “Why is the Lizard flora so distinctive?” I propose that this question is most clearly answered if the suite of species which give The Lizard a distinctive biogeographical character are identified. Here I refer to this sub-set of species as the Lizard florid e (Table 1). In order to identify this group with relative objectivity I here select them using the following decision rules: a. All Red Data Book vascular plants listed by Wiggington (1999). For the most part these species and subspecies are those found in Britain in 15 or fewer 10 km squares of the national grid. b. All Scarce plant species listed by Stewart et al. (1994), that is species recorded in Britain in 16-100 10 km squares of the national grid. c. Notable Cornish Rarities, that is those species which are recorded as rare or absent as native species in the rest of Cornwall but are relatively abundant at The Lizard (Hopkins 1983; French et al. 1999). In a number of cases plants are only recorded from The Lizard in Cornwall but are also rare at The Lizard (e.g. Carex pseudocyperus, Cirsium disectum) and these are judged to have limited value in biogeographical interpretation and have been excluded. d. Other than subspecies listed by Wigginton (1999) and Stewart et al. (1994), micro-species of the genera Taraxacum and Rubus, along with taxa below species level, genotypes and caryotypes have been omitted. Two complex factors can be put forward to explain the existence of The Lizard florule: a. Climate. The south-westerly location of The Lizard results in an exceptionally mild oceanic climate with distinct Mediterranean features. If it is accepted that climate plays the primary role in plant and vegetation distribution the Lizard florule might be expected to consist of taxa which are typical of such climatic and microclimatic conditions in other parts of Europe. b. Geology. It might be expected that the taxa of the Lizard florule are restricted to The Lizard District, and perhaps specifically to the unusual serpentine and gabbro. 122 ATLANTIC ARC TABLE 1 THE LIZARD FLORULE Species Biogeographical Element (Preston & Hill 1997) Habitat Preference Red Data Book species (Wiggington 1999) Allium ampeloprasum subsp. babingtonii Mediterranean Atlantic Hedgerow/Scrub Asparagus prostratus * Oceanic Temperate Cliffs Cytisus scoparius subsp. maritimus * European Temperate Cliffs/Cove valleys Erica ciliaris* Oceanic Temperate Heath Erica vagans* Suboceanic Southern Temperate Heath Fumaria occidentalis Oceanic Temperate Ruderal Genista pilosa* European Temperate Heath Herniaria ciliolata Oceanic Temperate Grassland/Cliff/Cove valleys Hypochaeris maculata * Eurosiberian Temperate Heath Isoetes histrix* Mediterranean Atlantic Seasonal pools/Grassland Juncus capitatus* European Southern Temperate Heath/Grassland/Seasonal Pools J uncus pygmaeus * Mediterranean Atlantic Seasonal Pools Polycarpon tetraphyllum* Mediterranean Atlantic Ruderal Polygonum maritimum Mediterranean Atlantic Beach Ranunculus tripartitus* Oceanic Southern Temperate Seasonal Pools Rumex rupestris Oceanic Temperate Beach Trifolium bocconei* Mediterranean Atlantic Heath/Grassland Trifolium incarnation subsp. molinerii Mediterranean Atlantic Cliffs/Cove valleys Trifolium strictum* Submeditteranean Subatlantic Cliff/Cove valleys Scarce (Stewart et al. 1994) Adiantum capillus-veneris* Mediterranean Atlantic Buildings Allium schoenoprasum* Circumpolar Boreo- Arctic Montane Seasonal pools Arum italicum Mediterranean Atlantic Ruderal Asplenium obovatum Mediterranean Atlantic Walls Bromus hordeaceus subsp. ferronii Eurosiberian Southern Temperate Cliff slopes Chamaemelum nobile Suboceanic Southern Temperate Ruderal/Grasslands Cicendia filifo rm is * Submediterranean Subatlantic Seasonal pools Cyperus longus European Southern Temperate Reed beds Deschampsia setacea* Oceanic Temperate Pool margin Elatine hexandra European Temperate Pool margin Erodium moschatum Mediterranean Atlantic Ruderal/Grassland Euphorbia portlandica Oceanic Southern Temperate Dune Fumaria purpurea Oceanic Temperate Ruderal Genista tinctoria subsp. littoralis European Temperate Cliff slope/Heath Hypericum undulatum Oceanic Southern Temperate Flush Lathy rus japonicus European Boreo Arctic Montane Beach Lotus subbiflorus Suboceanic Southern Temperate Cliff slope Medicago polymorpha Submediterranean Subatlantic Ruderal Melittis melissophyllum European Temperate Scrub Minuartia verna* European Boreal Montane Cliff/Heath/Grassland Orobanche alba European Temperate Cliff/Heath Orobanche hederae* Submediterraean Subatlantic Scrub Orobanche rapum-genistae Suboceanic Southern Temperate Scrub Poa infirma Mediterranean Atlantic Ruderal Scilla autumnalis Mediterranean Atlantic Cliff/Heath/Grassland Trifolium glomeratum* Mediterranean Atlantic Heath Trifolium occidentale Oceanic Temperate Cliff Trifolium suffocation* Mediterranean Atlantic Ruderal/Grassland Vicia lutea Submediterranean Subatlantic Scrub Viola lactea* Oceanic Temperate Heath Zostera marina Circumpolar Wide Temperate Intertidal/Subtidal FLORA OF THE LIZARD PENINSULA 123 TABLE 1 CONTINUED Species Status Biogeographical Element Habitat Preference Notable Cornish Rarities Cladium mariscus* Eurosiberian Southern Temperate Reed beds Filipendula vulgaris * Eurosiberian Temperate Heath/Grassland Genista anglica* Oceanic Temperate Heath Geranium sanguineum* European Temperate Heath Orchis morio* European Temperate Heath/Grassland Sanguisorba officinalis* Circumpolar B oreo-Temperate Heath *Taxa confined to the ophiolite on the Lizard District on the Lizard peninsula (Hopkins 1983; French et al. 1999) 45 40 - 35 -- X 30 03 £ 25 § 20 o 0 15 CL IO 10 5 0 □ Britain & Ireland (n=1 ,481 ) ■ The Lizard (n=667) □ The Lizard Florule (n=56) nj~l rm _ -n 0 C 0 c o E i o o < o 'o ® 0 03 I -4— • o c CD O o ^ DQ 03 0 L- O _Q I 0 ■D 0 C 0 C o E i 0 0 O CQ Figure 2. Biogeographical groupings of the native plants of Britain and Ireland, The Lizard and The Lizard Florule. PLANT GEOGRAPHY AND CLIMATE Preston and Hill (1997) have prepared a geographical classification of the British and Irish flora in terms of their Eurasian distribution. This includes 27 individual biogeographical elements. The biogeographical elements to which the taxa of the Lizard florule belong are listed at Table 1. Many of these elements contain only small number of taxa both in the British and Irish flora and therefore at The Lizard, and are not readily amenable to numeric analysis. However Preston and Hill have grouped these into 9 higher categories using numeric codings, which I here call biogeographical groupings (Fig. 2). These groupings broadly reflect latitudinal patterns of distribution and can be expected to be associated primarily with temperature aspects of climate (in their full listing of biogeographical elements Preston and Hill identify within each grouping further divisions which broadly reflect degree of oceanicity-continentality but this aspect has not been investigated here). 124 ATLANTIC ARC TABLE 2. PATTERNS OF GEOLOGICAL RESTRICTION IN THE LIZARD FLORULE Serpentine Serpentine and gabbro Serpentine and schist Allium schoenoprasum1 Deschampsia setacea2 Erica ciliaris2 Genista anglica1 Geranium sanguineum 2 Hypochaeris maculata2' Juncus capitatus 3 Minuartia verna2 Orchis morio2 Cicendia filiformis1 Cladium mariscus 2 Erica vagans Filipendula vulgaris2 Juncus pygmaeus Sanguisorba officinalis2 Viola lactea1 Asparagus prostratus Genista pilosa Isoetes histrix Trifolium bocconei Trifolium strictum 1 Species usually associated with acid rocks and soils in the rest of Britain and Ireland. 2 Species usually associated with calcareous or base rich soils in the rest of Britain and Ireland . 3 There is a single transitory record on schist. It can be seen from Figure 2 that comparison of the Lizard flora with that of the rest of Britain and Ireland shows only minor variations in the representation of the biogeographical groupings, mainly due to the absence or poor representation of northern elements in the Lizard flora. However, comparison of the Lizard florule with the British and Irish flora shows more striking discrepancy mainly due to the larger proportion of the Lizard florule comprised of Mediterranean and Sub-Mediterranean species. Clearly climate has a highly significant influence upon the distinctive elements within the Lizard florule. However, it can be seen that climate does not fully account for the occurrence of the Lizard florule. Particularly notable is the occurrence of four species of Boreo-Arctic Montane, Boreal Montane and Boreo-Temperate distribution, which in the core of their European distribution show no affinity to areas climatically comparable to The Lizard and are likely to occur there due to non-climatic factors. CORRELATION BETWEEN GEOLOGY AND DISTRIBUTION OF THE LIZARD FLORULE Detailed mapping of the Lizard flora at 1 km grid square scale (Hopkins 1983; Margetts 1988) and 2 km square scale (French et al. 1999) reveals that the species exhibit a wide range of patterns of distribution. However correlating these patterns with underlying geology is confounded in many cases because the boundaries between rock types are not always clearly defined in the field and soil, some derived from overlying acid loess and Crousa gravels, masks the effects of solid geology. It is therefore only in cases where species occur on shallow soils largely derived from rock parent material that the correlations between plant distribution and geology can be determined with confidence. Despite this there is significant evidence that the unusual character of the Lizard flora is influenced by geology. Of the 56 species of the Lizard florule, 28 are found only in the ophiolite complex of The Lizard District and it is notable that 13 of the 19 Red Data Book species are confined to the The Lizard District, as are all of the Notable Cornish Rarities. Table 2 lists the geological restrictions for 21 species of the Lizard florule where the pattern is distinct. All but Hypochaeris maculata and Deschampsia setacea occur at multiple stations and it is therefore reasonable to assume that these patterns reflect ecological characteristics of the species, rather than chance events. It can be seen that those species confined to serpentine, and those confined to serpentine and gabbro, comprise mixtures of taxa which in other parts of Britain are found on either acid or calcareous soils, and not usually in combination, reflecting the unusual characteristics of these ultrabasic rock types; an intermixing of acidophile and basophile species is also found at smaller spatial scale in analyses of the Lizard heathland vegetation (Hopkins 1983; Rodwell 1991). However a number of taxa show no close association with the serpentine or gabbro. Five species are found on both serpentine and schist, including Isoetes histrix and Trifolium bocconei (found only at The Lizard in mainland Britain), and T. strictum, a species otherwise only known on igneous rocks at Stannar Rocks, Radnorshire in mainland Britain. FLORA OF THE LIZARD PENINSULA 125 Although for many taxa of the Lizard florule it is not possible to determine patterns of geological restriction, two species stand out. Herniaria ciliolata is confined to The Lizard in mainland Britain, which hosts the majority of its world population (Wiggington 1999), but is extremely catholic, occurring on a wide range of sedimentary, metamorphic and igneous rocks near the coast and at one inland station on serpentine. Similarly Trifolium incarnatum subsp. molinerii (Plate 18) is only found at The Lizard in mainland Britain1, but is confined to schist (Martin & Frost 1980), other than at one station north of Gunwalloe on sedimentary rocks of The Meneage, where it has been recorded very infrequently. DISCUSSION AND CONCLUSION From the above it can be seen that, by identifying a discrete florule which particularly characterises the Lizard Natural Area it becomes a more tractable proposition to answer the question “Why is The Lizard distinctive in terms of its plant geography?” It is unlikely that this question has a simple answer and only two complex variables have been focussed upon here, climate and geology. Clearly, if it is accepted that broad patterns of geographical distribution at national and continental scale are controlled principally by climate, a central premise of much research into climate change being carried out, the high proportion of the Lizard florule which is of Mediterranean and Submediterranean affinity indicates that climate has a highly significant influence. However this is only a partial explanation, given that not all taxa of the Lizard florule have such a distribution and several taxa are of northern geographical distribution. Similarly study of the influence of geology on the Lizard flora indicates that whilst it appears significant it does not offer a complete explanation, particularly as arguably the most important species at The Lizard, Herniaria ciliolata , shows no clear pattern of geological restriction. In practice, patterns of geological restriction are not imposed directly by geology, but reflect the soil environment, and studies of factors such as soil structure, depth and chemistry are likely to provide more meaningful insights into the ecological basis of plant distribution at The Lizard. A wider range of other factors might also be considered in an attempt to understand the ecological basis of The Lizard flora. Haines-Young et al. (2000) and Preston et al. (2002b) have demonstrated the use of trait analysis, to understand the factors which are driving change in the flora of Britain and Ireland. Similar approaches offer opportunities to explore the factors which control the floristic character of biogeographically distinct areas such as The Lizard and may be of great value in defining the ecological requirements for plant conservation. A notable feature of the Lizard flora is that it comprises a wide range of life forms, although none are trees and shrubs. Table 1 includes an indicative analysis of the habitats occupied by species of the Lizard florule. These habitat associations are clearly complex and the diversity of the Lizard flora and the Lizard florule in no small part reflects the wide range of habitats present. One striking feature of The Lizard is the survival of this diversity of habitats. By comparison with other parts of Britain The Lizard has retained a high proportion of the semi-natural habitat present at the beginning of the 20th century. For example approximately 69% of the heathland present in 1908 survived in 1980 (Hopkins 1983), a significantly higher proportion than in other lowland heathland areas in England (Farrell 1993), and there has been almost no built development on the coast away from the fishing coves. In part this is due to the relative geographical isolation of The Lizard, which has inhibited economic development, and the difficulty of carrying out agricultural improvement on relatively infertile and poorly drained soils of the gabbro and serpentine. In the past 50 years a high proportion of The Lizard has become the subject of statutory designation, as Site of Special Scientific Interest, Special Area of Conservation and Area of Outstanding Natural Beauty, which has further limited the extent of adverse land use change. A significant proportion of this area is now managed as National Nature Reserve by English Nature or is owned and managed by the National Trust and Cornwall Wildlife Trust. The prospect for survival of the Lizard flora is therefore increasingly favourable, provided appropriate management and protection is in place. 'This taxon has recently (2005) been discovered as a presumed native on coastal schists near Bolt Tail, S. Devon (eds.) 126 ATLANTIC ARC REFERENCES BRISTOW, C. M. (1996). Cornwall’s Geology and Scenery: An Introduction. Cornish Hillside Publications, St Austell. FARRELL, L. (1993). Lowland Heathland: the extent of habitat change. English Nature Science Report No. 12. English Nature, Peterborough. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall : Atlas of the Flowering Plants and Ferns of Cornwall. Wheal Seton Press, Camborne. HAINES- YOUNG, R. H., et al. (24 authors) (2000). Accounting for nature: assessing habitats in the UK countryside. Department of the Environment Transport and the Regions, London. HOPKINS, J. J. (1983). Studies of the historical ecology, vegetation and flora of The Lizard, Cornwall, with particular reference to heathland. PhD Thesis. University of Bristol. Johns, C. A. (1848). A Week at The Lizard. Society for the Promotion of Christian Knowledge, London. LAWMAN, J. (1994). A Natural History of The Lizard Peninsula. Institute of Cornish Studies, Redruth. MALLOCH, A. J. C. (1972). Salt spray deposition on the maritime cliffs of The Lizard Peninsula. Journal of Ecology 60: 103-1 12. MARGETTS, L. J. (1988). The Difficult and Critical Plants of The Lizard District of Cornwall. Grenfell Publications, Bristol. Margetts, L. J. & David, R. W. (1981). A Review of the Cornish Flora 1980. Insititute of Cornish Studies, Redruth.. MARTIN, M. H. & Frost, L. C. (1980). Autoecological studies of Trifolium molinerii at The Lizard Peninsula. New Phytologist 86: 329-344. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnaean Society 124: 1-120. PRESTON, C. D., Pearman, D. A. & Dines, T. D. eds. (2002a). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. Preston, C. D., Telfer, M. G., Arnold, H. R. & Rothery, P. (2002b). The changing flora of Britain, in PRESTON, C. D., Pearman, D. A. & Dines, T. D. New Atlas of the British and Irish Flora, pp. 35 — 46. Oxford University Press, Oxford. Prosser, C. D. (1995). Conserving our Earth heritage through Natural Areas. Geoscientist 5: 19-20. RODWELL, J. S. ed. (1991). British Plant Communities Volume 2: Mires and Heaths. Cambridge University Press, Cambridge. SELWOOD, E. B., DURANCE, E. M. & Bristow, C. M. (1998). The Geology of Cornwall. University of Exeter Press, Exeter. STAINES, S. J. (1984). Soils in Cornwall III (The Lizard). Soil Survey Record No. 79. Soil Survey of England and Wales, Harpenden. STEWART, A., Pearman, D. A. & Preston, C. D. eds. (1994). Scarce Plants in Britain. Joint Nature Conservation Committee, Peterborough. WlGGINGTON, M. J. ed. (1999). British Red Data Books 1, Vascular Plants ( 3rd edition). Joint Nature, Conservation Committee, Peterborough. 127 The Lizard trackways project A. J. BYFIELD Plantlife International, 14 Rollestone Street, Salisbury, Wiltshire SP1 1DX, England and B. R. WHEELER Peterfield House, 55 Station Road, Horrabridge, Yelverton, Devon, PL20 7ST, England ABSTRACT The importance of the seasonally-flooded trackways of The Lizard District (West Cornwall) to botanical science and nature conservation is briefly described. They are of particular importance for the occurrence of six nationally rare or scarce vascular plant and stonewort species. A three year survey to assess the current condition of the trackway system and the status of the rare species, is being carried out. Initial findings support the belief that the trackways are under particular threat from infilling or neglect, with just 18% in a state suitable for the growth of the species of nature conservation value. Keywords: Cornwall, The Lizard, trackways, temporary pools. INTRODUCTION The cliffs and heathlands of The Lizard have long been famed amongst botanists and plant conservationists for their exceptional diversity and richness of vegetation and flora. The vascular plant flora includes 20 species that are regarded as Nationally Rare, and a further 53 species are considered Nationally Scarce. In addition, The Lizard is of exceptional importance for its diversity of stoneworts, lichens and bryophytes, and collectively it is regarded as one of the five richest and most distinctive botanical ‘sites’ in the Great Britain and Ireland archipelago. The trackways of The Lizard are of particular nature conservation importance for their assemblage of local and rare species associated with the seasonally-flooded nature of this habitat. A total of six Nationally Rare and Nationally Scarce vascular plants and stoneworts occur within this habitat ( Cicendia filiformis, Juncus pygmaeus, Mentha pulegium, Ranunculus tripartitus, Pilularia globulifera and Chara fragifera), together with a range of other localised species apparently declining rapidly across the heathland areas of Britain (e.g. Anagallis minima , Chamaemelum nobile and Radiola linoides ). These rare and local species roughly fall into three groups: (i) winter annual species that germinate under water in autumn and early winter, ‘fruiting’ in early spring as water levels drop {Chara fragifera. Ranunculus tripartitus)', (ii) spring annual species that most probably germinate on damp, but not flooded, substrate as the trackways dry out in late spring/early summer {Anagallis minima, Cicendia filiformis, Juncus pygmaeus, Radiola linoides)', and (iii) perennial mat forming species that are typically low-growing with a low competitive ability, which favour the open, rutted and often heavily grazed conditions associated with trackways {Chamaemelum nobile, Mentha pulegium, Pilularia globulifera). Table 1 gives details of the occurrence of these plants in the Lizard District, together with comments on their current status. The Lizard trackway system, together with shallow, seasonal heathland pools, has been formally recognised as the only outstanding example (in a European context) of the Habitats Directive priority habitat Mediterranean temporary ponds in the UK. The occurrence of this habitat here is one of the contributing reasons why The Lizard has been selected as a Special Area of Conservation (SAC), as part of the UK’s contribution to the Natura 2000 network of key sites across Europe. The Joint Nature Conservation Committee notes (www.jncc.gov.uk/ProtectedSites/ SACselection): TABLE 1: RARE AND LOCAL SPECIES ASSOCIATED WITH SEASONALLY-FLOODED TRACKWAYS IN LIZARD DISTRICT 128 ATLANTIC ARC s aD I H-H 3 X -a v T3 3 o cd 0) ca 3 o o ^ c« aD £ o ^ . $-4 r i . -a - aD ,3 X) £3 3 TD o c CD 3 CD X * 3 3 O O ,Fi £ o 3 X X* >> 3 aD fc 3 u CD 'C m o o — I ^ (U |?s2 G « dJ Cd Vh 03 T3 03 N J aD 3: H m 3D 3 3 3 O' c/d £ i > 3 3 3 3 CD u £ o O >, "3 ^3 U U oo •3 R- 2 B, ^ rS ON — m 3 13 ■0 u* 3 CJ cd 3 ^0 N O c • ^4 J u OJ > OD LC O 3 T3 C 3 0 PL -o aD T3 3- 00 CL 3 C/D OJJ cd aJ O — C^ 4— » 13 • C/5 aD CD aD 0£j *C 0 ■3 CL X 3 X D 0 X C/5 O jD 4—* £ C/D .2 s £. < 3 CL Pai 4- O ’cd aD ■0 *3- CQ "O 3 D cd C OJ r- b 13 aD > cl 00 >> cd N • ^4 j aD D CD 3 N J aD aD X _3 3 _ CD -3 o — m aD 3 3 oa: 3 3 _o w 3 z D3 c« 3 m / — ' S 8 3 2 3D aD 3 jo aD 3 .ZZ O PL m w 3 O -3 On ON 3 ^ X) CD *3 * m 00 O HD ^ 3 5 aD o 3 o LJ 00 CD -3 3 aD C< fc ^ zs iT1 (J 1 « o .2 •JO CD 3 CD Z ^ ^ 00 3 aD -a -La o -o 03 3 .H o J i+c "O H 0 e 0 £ 3 O "O C/D o ao Lh •J-h •3 3 £ 2 o 3 ^ , CL 3 J2 E L- 3 »i S 5 *c 3 OO 3 3, 3 0 C A s Cl CL 2 b C/5 s CL N 1/3 b X aD <; Cl - L - OQ aD X 4—* O- X CL CL 3 - _o *3 Oh CL < CQ Z> aD *0 CD CD 3 • — 3 £3 CD c/D 00 Q -q — 3 3 3 C N .2 j 3 aD Z oc ^ H c« aD 3 3 3 cr C/D J3 CD H £ 3 o .3 CO ^ ■>-, cd jL* 3 ; — 1 3 3 3 3 cr O C/D •—! C« ^ .2 ■3 ‘u 2 ^ '3 C/D CL C/D aD 3 3 3 aD 3 3 N3 rv • 3 o C/D 3 3 CL O CL 3 CD 3 z rt- O na ^ E % 2 2 3* CD "O 3 aD aD 73 ts 3 0 ^ CD ^ 2 -O DC aD . -3 3- O -o — 1 aD A aD ^ Z 2 -3 -3 < H TZ 3 -P J 3- O aD , J= ^ H § 3 ’ZZ •— c3 c« "£ aD 5 3 CL 3 O 3 CL cr (4^ ^ o E s -La o 1 'ZZ ^ 3 ° £ ^ g b 13 3 o N O •2 aD a 3 aD -3 £ O ,3 CD 3 O CD H x .£ a3 § o Cr aD ■~a 2 s- ^ C/D 2 £ *3 C X) . QC JZ 3h O O 3 cn 00 rt- CN in cn oc ON ON in on in r- — »».*/ ^ § 4n. A * 3 CL K So * ■««* GO -2 Cl "s -Ci ~Go cs ^ s •5 J 2~ s C3 • *44 -2 L *4- -s: 3 44 -2 •2 3 CD * CJ ^5 1 • *»•* =L 33 3 3 3 QC CC THE LIZARD TRACKWAYS PROJECT 129 “Mediterranean temporary ponds consist of winter-flooded areas, which dry out to give vegetation rich in annuals; many of these are nationally rare species of southern European distribution, which are principally confined to this habitat type, for example pygmy rush Juncus pygmaeus, pennyroyal Mentha pulegium and yellow centaury Cicendia filiformis. There are two main pool types: a more acid pool community of trampled and grazed areas, often found on flooded trackways, and a basic pool type on serpentine rock found only at The Lizard, Cornwall.” In spite of the exceptional botanical importance of these trackways, they have long been under threat from the conflicting requirements of farmers, the general public (both of whom use the trackways for access) and nature conservationists. Threats to the trackways largely stem from changes in the way that the tracks are used. The serpentine quarrying, which created and maintained many of the unsurfaced trackways, ceased around 1970. In parallel, the economics of farming on The Lizard changed radically, rendering many traditional practices unprofitable. A decline in stock keeping has led to tracks and paths being almost entirely used for recreation. Consequently some tracks are used less, which results in successional changes to the habitat and the loss of open ground supporting the ephemeral pool species. Other tracks are still used for access, but due to modem demands of vehicles, horses and walkers, the ‘quality’ of the track is deemed poor and some tracks have been surfaced to remedy this and improve access. In short, the trackway system has either suffered from overuse (infilling, drainage) or under use (becoming overgrown with coarse vegetation). The history of The Lizard tracks reached its nadir in the 1980s when one of the two most important stretches of trackway - on Kynance Downs - was infilled with rubble by the Nature Conservancy Council (the predecessors to English Nature) in the months following acquisition of the Kynance Downs area as a National Nature Reserve, with the loss of nationally important populations of Juncus pygmaeus, Pilularia globulifera , and Ranunculus tripartitus. Acts like this focused the attention of plant conservationists on the plight, and conservation requirements, of the trackways, but even so, the decline in botanical interest of the trackways continues. Plantlife International became involved with the conservation of this highly specialist flora in its role as Lead Partner for Pillwort ( Pilularia globulifera) and Three-lobed Water-crowfoot (. Ranunculus tripartitus ), both priority species identified as requiring urgent conservation action through the UK Government’s Biodiversity Action Plan. Plantlife acts as Lead Partner for 77 species of vascular and lower plant, and their conservation is achieved through its Back from the Brink species recovery programme, which receives generous financial support from English Nature. THE 2002 LIZARD TRACKWAY SURVEY In March 2001, recognising the continued decline of the internationally important trackway habitats on The Lizard, Ruth Davis and Ro FitzGerald of Plantlife drew up a proposal for surveying the plant biodiversity interest of The Lizard trackways. As a result, Plantlife commissioned Dr Belinda Wheeler to undertake a survey of The Lizard trackways over a two-year period starting 2002, with the following aims: • to locate areas of important plant diversity; • to assess the current status of the rare and threatened plant species, and to determine the current state of vegetation of the trackway system (and its potential for restoration); and • to target areas suitable for conservation management and restoration, and to identify optimum management within the terms of the EU Habitats Directive legislation, insofar as it relates to Special Areas of Conservation (SACs). To assess the value of the trackway system to plant conservation, six ‘target’ species ( Chara fragifera, Cicendia filiformis, Juncus pygmaeus, Mentha pulegium, Pilularia globulifera and Ranunculus tripartitus ) were identified as indicator species of high quality habitat, and the survey work focused heavily - although not exclusively - on these species. A desk study was carried out jointly by Plantlife International and the Environmental Record Centre for Cornwall and the Isles of Scilly (ERCCIS - housed at the Cornwall Wildlife Trust), who are kindly supporting this project. All known records of each of the target species were located within The Lizard survey area. ERCCIS has recently extracted all records from Colin French’s (vice county 1 recorder) ERICA Database and incorporated them into the ERCCIS Microsoft Access database. This is part 130 ATLANTIC ARC of an ongoing process at ERCCIS to move over to using Recorder 2000 for all county records. All records were checked for accuracy by Miss Rosaline Murphy (recorded for v.c. 2 and local expert). A full list of records for each target species is now available to Plantlife for the survey. ERCCIS provided further support to Plantlife by linking the records for the target species held in the database with the ArcView GIS system. This resulted in the production of an overview map at 1:60,000, and a series of more detailed maps at 1:10,000, showing the locations of all records for the target species in the survey area. These data were combined on the map with habitat data already held by ERCCIS (based on the Cornwall Wildlife Trust’s 1995 Life Project - an interpretation of habitats from aerial photographs resulting in digitised boundaries of habitat data). For the purposes of this survey only heathland and wetland communities were mapped, and it became immediately apparent from this combined mapping that the records for the target species are largely confined to areas where semi-natural vegetation survives. The trackways were resurveyed systematically, and making a particular point of trying to relocate the historic populations of target species, recording associated taxa, and documenting the current state of each trackway section. The routes were walked at two times of year: in March/ April to locate winter annual species and in June to resurvey for spring-germinating annuals. In the first year (2002) the survey work concentrated on Lizard and Predannack Downs, which make up the principal heathland blocks in the western half of The Lizard District (within the parishes of Landewednack and Mullion), but since then the survey has been extended to cover the most easterly blocks of heathland at The Lizard (notably the Goonhilly Downs and Crousa Common and Downs). RESULTS The first year (of two) of survey work is complete, and populations of all six target species have been located. Many recently recorded populations have been resurveyed whilst a number of older records have been confirmed as extant. Additionally, new populations of several of the target species have been located on little-used and abandoned trackways, and in quieter areas of The Lizard NNR where recording effort has not been focused in the past. Whilst the current presence of a population of one or more target species was welcomed, information about the state of the trackway lengths was perhaps more important from a nature conservation point of view. Table 2 gives an overview of the state of the 36-8km of trackway walked. Just 6-5 km (18% of the total) remained in a good state, and thus still suitable for one or more of the rarer species. 23-2 km of the trackways survived in a poor state (63%), unsuitable at the present time for the target species (albeit easily recoverable), whilst 7T km of trackway formerly known to be of importance for rarer species had been destroyed - 19% of the overall total. DISCUSSION The full findings of the trackways survey will be published in report form during 2004, on completion of the full survey during spring 2004. However, the initial findings suggest that the final report will be useful both in (i) directing track improvement schemes (particularly for tourism purposes) away from stretches of trackway that either support rarer species or at least have the potential to support such species should they be managed correctly; and (ii) prioritising conservation action on trackway systems that are currently in a poor state, but have the potential to be easily restored through conservation management. TABLE 2: THE CURRENT STATE OF TRACKWAYS IN LANDEWEDNACK & MULLION PARISHES IN 2002 State of trackways Length of trackway (km) % of total Good condition (i.e. currently suitable for growth of rarer species) 6*5 18 Poor condition (i.e. currently unsuitable for growth of rarer species) 23-2 63 Destroyed 7'1 19 Total 36*8 THE LIZARD TRACKWAYS PROJECT 131 ACKNOWLEDGMENTS We are grateful to Sarah Myles and Ian Bennallick at ERCCIS for their considerable help in providing the background data and maps that formed the basis of this survey. Plantlife International wishes to acknowledge the financial support of English Nature. REFERENCES Davis, R. & Fitzgerald, R. (2001). Proposal For Survey of Biodiversity Interest of The Lizard Trackways. Unpublished report, Plantlife, London. Fitzgerald, R. (2001). Pygmy rush Juncus pygmaeus : National status in 2000, Lizard Peninsula, Cornwall. English Nature Research Reports No. 412. English Nature, Peterborough. STEWART, N. F. (2003). Review of the records for Chara fragifera Durieu in Britain. Botanical Cornwall 12: 51-72. 133 Semi-natural vegetation on National Trust land in Cornwall, with an emphasis on that of the Lizard peninsula J. BRUCE, L. CORDREY, H. J. HARVEY The National Trust, 33 Sheep Street, Cirencester, Gloucestershire GL7 IRQ, England and A. CAMERON The National Trust, The Lizard, Cornwall, England ABSTRACT The National Trust owns about 9,000 ha of land in Cornwall, of which about 50% is occupied by semi-natural plant communities or by habitats little influenced by man (e.g. mudflats, ‘bare’ rock). This article presents a brief overview of some of the plant communities involved and of issues of their management, with special emphasis on the habitats of the Lizard peninsula. Keywords: Cornwall. The Lizard, semi-natural vegetation, habitat management. INTRODUCTION The two most abundant types of semi-natural vegetation on National Trust land in Cornwall are terrestrial coastal communities and lowland heath. For these habitats the National Trust is a major owner: for example it owns about 41% of the of coastal grassland in Cornwall (Plate 19). Presented below is a brief overview of some of the plant communities involved and of issues of their management, with special emphasis on the habitats of the Lizard peninsula, the most southerly area of The British mainland. GENERAL CORNWALL BOTANICAL SURVEYS Between 1979 and 2000, the National Trust employed a permanent Biological Survey Team. Most Trust Land in Cornwall was surveyed at least once during this period, and the vegetation mapped to the UK’s Phase 1 classification. Vegetation maps were transferred to GIS (Map Info) and data tables were created in Excel. Raw data were vetted to seek to ensure a uniform allocation to Phase 1 types. Although boundaries between communities were mapped by eye and are therefore subject to some error, measurements of mapped areas generally have a high precision. Data derive from various dates and their summation may not completely reflect the present situation. The analysis of data reported here provides a basis for unbiased judgements about relative value of sites, can contribute to decisions about priorities for management, and might guide habitat re-creation schemes. Overviews of the data produced from these surveys are presented in Table 1 and in graphical form below (Figs 1-4). LIZARD PENINSULA HABITAT MANAGEMENT The Lizard peninsula proved a good management case study, combining a range of unique habitats with busy visitor sites. Much of its land is in the ownership of a variety of organisations sympathetic to maintaining its wildlife value and landscape diversity: the National Trust, English Nature and Cornwall Wildlife Trust. 134 ATLANTIC ARC TABLE 1. HABITATS ON NATIONAL TRUST LAND IN CORNWALL - SUMMARY DATA Hectares % of Total Total Area Surveyed 8662-40 Improved Grassland 2201-20 25 Arable 1378-30 16 Lowland Heathland 910-20 10 Coastal (terrestrial) 762-20 9 Woodland 674-30 8 Data source: National Trust Central habitats database 2003 Figure 1 Extent of selected Phase 1 lowland heathland habitat types in Cornwall (% of National Trust total). ■ Dry acid □ Dry basic □ Dry heath/grass □ Wet heath 160 140 'm 120 o S 100 o a 80 ro 60 a> < 40 20 0 Zennor: Lizard Chapel Crackington Rosemergy & Peninsula: Porth Haven Trevean Cliffs Predannack Figure 2. Lowland Heathland habitat types at selected National Trust properties in Cornwall. VEGETATION ON NATIONAL TRUST LAND IN CORNWALL 135 Boulders/rocks 3% Soft cliff <1 % grassland 44% Figure 3. Extent of Selected Phase 1 Coastal terrestrial habitat types (% of National Trust total). □ Coastal grassland □ Coastal heathland DDune grassland □ Open dune ■ Hard cliff □Boulders/rocks □ Shingle/gravel (Plate 19) _ 160 140 _ 120 I 100 1 80 60 < 40 20 0 Figure 4. Coastal terrestrial habitat types at selected National Trust properties. Agricultural field systems: The coast of Cornwall is essentially an agricultural landscape. Over the last twenty years much work has been done with farmers to revert the improved agricultural fields to a less intensive pasture or arable system providing a buffer to the unimproved cliff edges. Over time, the once improved fields have often become more species-rich themselves. At Predannack, a National Trust farm on the west coast of the Lizard, a period of over twenty years of conservation restrictions (involving agreed limits to fertiliser and pesticide use and agreed stocking levels and grazing times on the farmlands) has allowed a rich flora to develop in the coastal pasture fields. Coastal scrub management : Much of our coastline has suffered from scrub encroachment as agricultural economies discouraged grazing of the cliff land outside the enclosed fields. On this land, management by conservation organisations over the past twenty years has been aimed at ■ _ • Pentire Head Zennor: Crackington Cubert Penrose Estate & Portuin Bay Rosemergy & Haven Trevean Cliffs 136 ATLANTIC ARC reducing the level of scrub on the cliffs. We must, of course, still be aware of the wildlife interest that can be present in the scrub. For example Dormice were recently found on south Cornwall cliffs. In terms of practical management of scrub, a variety of methods are regularly employed to tackle scrub encroachment, from physical labour (e.g. by volunteers or Community Service workers) to use of brushcutters, hand-operated mowers, tractor-mounted flails or converted silage- cutters. The latter are capable of clearing large areas of scrub. Less commonly used on the coast, but employed regularly by English Nature on the heathland inland at The Lizard, is the use of controlled burning to manage heathland or to quickly remove large areas of scrub. Heath and scrub fires also occur regularly in an uncontrolled way, especially in the dry summer months. These can cause significant damage. Grazing regimes adopted : Scrub removal is only the first part of the restoration of the habitat after years of neglect. A range of hardy animals are being used for grazing on the coastal fringe. Where possible, cattle are used such as Friesians at Predannack or Highland Cattle on English Nature’s land at Kynance, now the largest herd of the breed outside Scotland. Elsewhere on some of the smaller strips of land, or steeply sloping coastal slopes where cattle are harder to keep, ponies owned by English Nature or the National Trust, such as Shetland ponies, have been used to good effect over twenty years. These hardy breeds are easy to keep and will tackle gorse and blackthorn as part of their diet. Alternatively sheep, such as Soays, are used, their surefootedness meaning that even the steepest of slopes receive some grazing benefit. Recent incentives to farmers in the form of the rare breeds scheme to use traditional breeds will hopefully mean that more of these breeds will be able to graze the cliffs, and this will certainly be a welcome addition to the variety of animals available to manage the land. Visitor impact. The impact of visitors on sensitive sites is not easy to control. At Kynance Cove (Plate 2) the National Trust have successfully managed to control the flow of 150,000 visitors each year to the beach to minimise the impact on surrounding habitats. This has been achieved through careful positioning of paths and car parking to encourage people to use the least-damaging route. Where irreparable damage has already taken place, we have recently trialled, with some success, the use of turf-cutters to re-vegetate eroded paths. Invasive plants: On The Lizard, Hottentot Fig ( Carpobrotus edulis ) from South Africa, is extensive, and has become a significant threat in many places. Loved by many locals and visitors for its gaudy flower display, it forms curtains of hanging vegetation which smothers the natural cliff flora and is extremely labour intensive (and difficult of access) to remove. We have concentrated principally on the areas where it threatens rare plant populations and have recently used teams of climbers to pull it from very steep sections of cliff. Elsewhere, Japanese Knotweed (. Fallopia japonica ) has been tackled using a cut-and-inject method of treatment that has proved extremely effective. Managing for other wildlife: The Lizard recently welcomed the natural return of the Chough, a bird formerly more widely known in this area (and which has long been featured on the Cornish coat of arms), but, until recently, had not seen here for about fifty years. The establishment of close-cropped coastal pastures that has been one aim of management on The Lizard, to maintain and improve the diversity of plant species, has also created the right habitat for this bird. The successful rearing of the first brood of young birds is thus a local conservation success story, that has captured the imagination of professionals and public alike. 137 Genetic variation in Irish threatened plant species: a European perspective R. J. SMITH Department of Botany, Trinity College, Dublin 2, Ireland and S. WALDREN Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6, Ireland* ABSTRACT The genetic variability of plant populations can have both short-term and long-term implications for population persistence and the study of population genetics has therefore become an integral part of research into the conservation of threatened plant species. The AFLP genetic fingerprinting technique has been used to assess the conservation status of a number of Irish threatened plant species in a European context and these include Campanula trachelium L. (Nettle-leaved Bellflower), a highly fragmented species showing possible signs of inbreeding depression, and Colchicum autumnale L. (Meadow Saffron), an endangered Irish species with a highly localised distribution and questionable native status. Results are presented for C. autumnale , and these indicate that the Irish populations contain high levels of genetic diversity and are therefore unlikely to have been introduced, and unlikely to be susceptible to the deleterious effects of inbreeding depression. Population comparisons illustrate that the European perspective is fundamental to the assessment of the conservation status of Irish threatened plant species. KEYWORDS: Conservation genetics, Colchicum autumnale, Campanula trachelium, inbreeding depression, biogeography. INTRODUCTION The conservation of plant diversity is dependent on the conservation of individual plant species and their inherent genetic variation. It is therefore necessary for conservation strategies to be based upon a thorough scientific knowledge of the species they aim to conserve (Falk & Holsinger 1991). As genetic diversity is the base level of biodiversity it is of vital importance to the concerns of conservation biology and should be a prime consideration in plant conservation research. The relevance of the concepts of population genetics to the field of conservation management was stressed in Berry’s paper to the British Ecological Society Symposium of 1970 (Berry 1971) and the genetic perspective has since become an integral part of conservation biology. While the demographic and ecological analyses that form the basis of many plant studies are vital in assessing short-term threats to populations, ultimately it is the genetic structure of a population that will govern its ability to persist on an evolutionary time-scale (Frankel 1970). In addition to this long-term consideration, genetic processes can also have more immediate effects on population viability through interaction with other factors, and through processes such as inbreeding depression that can have direct effects on the reproductive fitness of individuals. Concerns about the genetic status of threatened species have led to the incorporation of population genetics into a large number of plant conservation studies (e.g. Delgado et al. 1999; Godt & Hamrick 1996; Schmidt & Jensen, 2000; Zawko et al, 2001), but such information is only useful if viewed in the context of other populations throughout the range of the species. This is * e-mail: swaldren@tcd.ie 138 ATLANTIC ARC particularly pertinent in the case of species at their distributional limits, as species are usually more fragmented in this part of their range (Griggs 1914; Hengeveld & Haeck 1982; Carter & Prince 1987). In all cases, an assessment of the conservation status of a species can only be relevant if it has been undertaken in a wider context involving comparisons of levels of genetic variation from a wide spread of populations. GENETIC THREATS TO SMALL POPULATIONS The genetic processes that occur in plant populations are stochastic in nature and therefore become less stable the smaller the population size. Genetic drift, the random change in allele frequencies from generation to generation, becomes more pronounced in small populations where it is more likely to lead to the loss of alleles (Futuyma 1986) and to the accumulation and fixation of deleterious mutations (Ellstrand & Elam 1993). Things can be confounded further in small and fragmented populations due to the increased incidence of inbreeding and the associated loss of fitness known as inbreeding depression. This phenomenon is often manifested through the loss of reproductive fitness and can therefore have a significant impact on the breeding population (Demauro 1993). There are many studies that document the detrimental effects of low genetic variability on plant populations (Newman & Pilson 1997) and the effects include lower reproductive output, lower germination rates, lower seedling survival rates and lower estimated mean fitness (Demauro 1993; Buza et al. 2000; Sevems 2003). Thus, while maintenance of genetic variability is a long-term conservation goal it clearly can also have a marked effect on short-term viability (Reed & Frankham 2003). Another issue for consideration is that strongly outbreeding species are particularly susceptible to inbreeding depression (Wright 1977), and the fragmentation of populations of such species can therefore have far-reaching genetic consequences. This situation has occurred in Irish populations of Campanula trachelium (Nettle-leaved Bellflower). It occurs predominantly in the Nore valley in south-east Ireland, where its main habitat is base-rich woodland on the sides of river valleys. It was formerly more widespread in Ireland and has undergone a decline since records began in the sixteenth century, probably due to fragmentation of its woodland habitat. As a largely outbreeding species with protandrous floral development, it is particularly susceptible to the deleterious genetic effects of population decline and fragmentation and could therefore have experienced loss of genetic variation, inbreeding depression and associated loss of fitness. Seed output was found to be significantly lower in the smaller populations, possibly as a result of inbreeding depression (Smith & Waldren, unpublished data). An alternative hypothesis would be that due to the low density of individuals in the smaller populations, pollen limitation is causing the effect. However, when supplemental pollen was added experimentally in the field, there was no significant increase in seed output. The genetic screening of Irish, and other European populations, will help to clarify this issue of potential genetic erosion. The potential for small populations to deteriorate as a result of a lack of genetic variation is clear. Even without inbreeding depression, a lack of genetic diversity restricts the potential for evolutionary change thus limiting the capacity for a species to adapt to a changing environment. The smaller the gene pool, the higher the probability of further allele loss and accumulation of deleterious alleles, and the lower the adaptability of the population. THE EUROPEAN PERSPECTIVE TO CONSERVATION GENETICS There are a number of reasons why the population genetics of threatened plant species should not be restricted to populations in one province or country. Firstly there is the question of biogeography. Unless the study species is a restricted endemic, the reasons for its present European distribution may be important in the determination of its conservation value. This is particularly relevant if there is a question regarding the native status of the species. From a conservation perspective, the native flora of a region is considered to take priority over the introduced and naturalised flora and it may therefore be important to verify the native status of a species to ensure the efficient use of conservation resources. The colonisation routes and GENETIC VARIATION IN THREATENED IRISH PLANTS 139 movements of a plant species are recorded in the genetic make-up of its constituent populations and a comparison of their genetic characteristics can reveal their biogeographical history. The level of population differentiation can also only be assessed in this broader context. If populations in one locality have differentiated significantly enough to be regarded as genetically distinct, they may be of high conservation priority in a European context. For example, Irish Asparagus prostratus Dumort. is at the north-western limit of its range, and may therefore represent a unique component of the European-wide genetic diversity. A. prostratus is considered by Curtis and McGough (1988) to be Vulnerable in Ireland (IUCN ‘Country’ category), although what is probably the largest European population occurs in a sand dune system along the south coast in Co. Wexford (v.c. HI 2). To look at the genetic characteristics of the Irish populations would be meaningless without a European comparison. The comparative level of genetic diversity will also be an important consideration in terms of the long-term survival of plant species as it gives an indication of evolutionary potential. It may also have relevance to the short-term survival of the population if genetic variation is too low, with the possibility of inbreeding depression and further genetic deterioration. Once again, this European context is vital because if Irish populations were examined in isolation and low levels of genetic variation were found, there would be no indication of whether this was a phenomenon of the species or simply a characteristic of the Irish populations. The genetic distance between populations also has implications for conservation management if reinforcement is considered. This involves the introduction of individuals into a population to boost overall numbers, numbers in a specific demographic class or to boost genetic variability. If genetic information is ignored in any of these cases, there is a possibility that outbreeding depression could occur. This is the converse of inbreeding depression, and is a loss of fitness caused by the break-up of locally adapted gene complexes through the mixing of individuals from different localities (Templeton 1986). It is clear that genetic factors are a necessary part of the process to assess the conservation status of species and that this cannot be done through a blinkered snapshot of the genetic variation in one small part of the distribution of a species. AFLP ANALYSIS OF IRISH THREATENED PLANTS The genetic fingerprinting technique of amplified fragment length polymorphisms (AFLP) is a multi-locus technique that yields a large amount of highly reproducible genetic information. It was first developed by Vos and co-workers (Vos et al. 1996) and has since become the preferred method for many studies of the genetics of rare plant populations (e.g. Travis et al 1996; Palacios et al. 1999; Gaudeul et al 2000; Zawko et al. 2001; Ronikier 2002; Coart et al. 2003). It is particularly useful in this field due to the relatively small amount of DNA required to carry out the reactions (~250ng). The sampling of leaf material can therefore be carried out without adversely affecting the growth or reproduction of the plants, an important consideration in the sampling of threatened plant species, which often exist in small populations. This technique is therefore being used to evaluate the genetic characteristics of a number of plant species that have limited distributions in Ireland and are considered to be under threat. Campanula trachelium has a highly localised distribution in Ireland and exists in a series of fragmented populations that may be acting as a metapopulation. The restriction of its range may have reduced genetic variation in individual populations, causing inbreeding depression, and may have led to population differentiation. Irish, Welsh, English and French populations are currently being screened using AFLP fingerprinting to provide answers to these important questions. Similarly, European populations of Asparagus prostratus have been sampled extensively and are currently awaiting analysis using AFLPs. The species is dioecious and all individual males and females have been mapped accurately in two of the Irish populations to allow a full genetic analysis. In addition to these works in progress, the AFLP technique has already been successfully used on other Irish threatened plant species including Colchicum autumnale, which will now be discussed in more detail. 140 ATLANTIC ARC CASE STUDY: COLCHICUM AUTUMN ALE There are six vascular plant species that are considered to be Endangered in Ireland (IUCN ‘Country’ category) on the basis of threat assessments detailed in the Irish Red Data Book (Curtis & McGough 1988). These are given legal protection under the Flora Protection Order, 1999 (S.I. No. 94 of 1999), along with 67 other threatened plant species. One of the most interesting of the six Endangered plants is Colchicum autumnale (Meadow Saffron). In Ireland this species is found predominantly in damp, base-rich riverside meadows, but throughout much of its range it is also commonly found at the edges of base-rich woodland. Its present distribution in Ireland is restricted to five sites along the valley of the River Nore, Co. Kilkenny (v.c. HI 1) and a further site in Co. Wexford (v.c. HI 2). Other sites in Co. Limerick (v.c. H8) have recently disappeared and the reason for this, and for its decline in the Nore valley where it was once more common, is thought to be the intensification of agricultural practices. The majority of its Irish sites are riverside meadows, and many of its former sites have been ploughed, drained or re-seeded, with excessive fertiliser application. However, the five populations occurring along the valley of the River Nore all occur within the Nore and Barrow candidate S.A.C., and so will be fully protected under European Law. The Nore populations lie within 14 kilometres of one another along a stretch of the river valley that is particularly rich in abbeys and monasteries. This monastic association, coupled with the restricted distribution of the species in Ireland, has led to suggestions that Colchicum was introduced with the spread of monasteries into the area. The reason for this suspected association is that the seeds and underground corm of the plant contain the alkaloid colchicine, which has historically been used in the treatment of rheumatism and gout, and it is possible that it may have been cultivated in monastic grounds for this purpose. Further support was lent to this theory when a sixth Irish site was recently discovered adjacent to Tintem abbey in Co. Wexford. Despite the possibility that the species may have been introduced into Ireland, it is certainly native to the British Isles, is non-invasive, forms an integral part of the Irish natural and cultural heritage and is therefore worthy of conservation. However, it is questionable whether scarce conservation resources should be allocated to a non-native species, and for this reason Colchicum autumnale is currently the subject of scientific study to evaluate its genetic conservation status in Ireland through comparison with other European populations. This is being carried out in conjunction with demographic and reproductive monitoring of the species. The objective of the population genetic study was to characterise the genetic variation present in Irish populations of Colchicum autumnale and to evaluate this information in a European context. Nineteen populations were sampled from Ireland, Wales, England, France, and the Iberian Peninsula and these are shown in Fig. 1. Preliminary AFLP analysis was carried out using two primer pairs to screen one hundred and eighty two individuals. Ten individuals were screened from all populations with the exception of population 19, where only three samples were taken due to the low number of individuals at the site. PAUP version 4, a genetic analysis software package, was used to construct a neighbour-joining tree of the samples, using mean character difference (Fig. 2). The samples are labelled with population number and it can be seen that there is little structuring of individuals from British and Irish populations, most of which appear to be undifferentiated. However, all the samples from the French populations are grouping together in the same cluster and are also closely related to samples from populations 15 and 16, which are located in the eastern Pyrenees. This group of populations is obviously significantly differentiated from the other European populations. A bootstrap consensus of the data showed significant support for this grouping but the structure of the remainder of the tree is not well-supported. To look more closely at the population groupings, a UPGMA dendrogram constructed using Nei’s (1978) measure of genetic distance showed that no significant groupings occurred among Irish, English and Welsh populations but these are distinctly separated from all the French populations and the two eastern Pyrenean populations (Fig. 3). However, the other three Iberian populations have clustered together and are also included in the same major group as the British and Irish populations. These three populations are located in the Basque country and Navarre, near the northern coast of Spain and are separated from the other Spanish populations by a significant area of the Pyrenees. This indicates an interesting biogeographic link between northern Iberia and the British Isles, which may be indicative of a post-glacial colonisation route that progressed from an Iberian refugium, along the Atlantic coast of France and into Britain and Ireland. Further evidence to support this may be provided by the addition into the analysis of recently-collected samples from the lower Loire Valley in France. GENETIC VARIATION IN THREATENED IRISH PLANTS 141 Figure 1. Approximate locations of Colchicum autumnale populations sampled for genetic analysis. For Ireland and Britain, the location and County are given, for France, location. Province and Departement are given and for Spain, location. Province and Autonomous Community are given. 1. Jerpoint, Co. Kilkenny, Ireland 2. Thomastown, Co. Kilkenny, Ireland 3. Fourpenny Rock, Co. Kilkenny, Ireland 4. Inistioge, Co. Kilkenny, Ireland 5. Inistioge Lower, Co. Kilkenny, Ireland 6. Llanmerewig Glebe, Powys, Wales 7. Cwm Bwchel, Monmouthshire, Wales 8. West Wood, Gloucestershire, England 9. Asham Wood, Somerset, England 10. Fonds l’Olivier, Meuse, Lorraine, France 1 1. Val des Nonnes, Meurthe-et-Moselle, Lorraine, France 12. Moncley, Doubs, Franche-Comte, France 13. Avanne, Doubs, Franche-Comte, France 14. Valleraugue, Herault, Languedoc Roussillon, France 15. Campo-Esera, Huesca, Aragon, Spain 16. Campo-Rialbo, Huesca, Aragon, Spain 17. Lapizea, Navarre, Spain 18. Gasteiz- Vitoria, Alava, Basque Country (Euskadi), Spain 19. Sendadiano, Alava, Basque Country (Euskadi), Spain 142 ATLANTIC ARC 11 FIGURE 2. Neighbour joining tree of Colchicum autumnale samples from 19 European populations. The numbers correspond to population numbers, the locations of which are shown in Figure 1 (Boostrap values over 80% are shown on the relevant branches). GENETIC VARIATION IN THREATENED IRISH PLANTS 143 — pop 1 pop? — pop8 — pop9 — pop2 — pop3 - pop5 L pop6 — pop4 — popi6 — pop!8 — pop!9 — poplO — popi3 — popl 1 — pop!2 — popi4 — popl5 — pop!7 Figure 3. UPGMA dendrogram of Colchicum autumnale populations. To assess levels of genetic variability, genetic diversity statistics for the populations were generated using POPGENE analysis software. These were obtained using Nei’s analysis of gene diversity in sub-divided populations (Nei 1987). Fig. 4 shows average gene diversity for each country, obtained from averaging Nei’s gene diversity for the populations in each country grouping. Although diversity varied considerably, the Irish populations contain high diversity when compared to other populations throughout the range of the species. The Irish populations are therefore not genetically depauperate and as a result, demographic and environmental stochasticity may be of more relevance to the survival of the populations. The comparative levels of genetic variation found within the Irish populations also indicate that introduction is unlikely. Although there is evidence for a monastic link, the monks may have been utilising a native plant resource rather than introducing the species themselves. If introduction has occurred, the presence of diversity in the Irish data-set would indicate the introduction of multiple corms and, subsequently, substantial outbreeding. Nevertheless, the species is mapped as ‘native’ in Ireland in the New Atlas of the British and Irish Flora (Preston et al. 2002), and the genetic studies that we have undertaken would tend to support this view. The Irish populations are not genetically differentiated enough to be considered a unique part of the overall diversity of the species. However, combined with the British populations they form a group that is genetically distinct from the mainland European populations and the British and Irish populations should therefore be considered together as an important conservation unit. 144 ATLANTIC ARC Figure 4. Comparison of average gene diversity Hs (Nei 1973) in the four sampled countries. CONCLUSIONS It is clear that the genetic viability of plant populations can have short-term and long-term implications for population persistence and must therefore be considered as an important component of conservation research. In addition to this, the study of the genetic variability of Irish threatened plant species has illustrated that to effectively evaluate the genetic conservation status of these species, a European perspective is vital. Values for genetic diversity are meaningless unless placed in the context of population statistics from other parts of the range of that species and, in a biogeographical context, the relationships between populations can also be important to the evaluation of conservation status. This wider perspective ensures that species conservation is not restricted by the assumptions concerning ‘normal’ levels of variability that must be made if populations are viewed in the isolation of a within-country study. This same principle applies equally to all threatened species that have a Europe-wide distribution and the European perspective should therefore form the basis of all conservation genetic research on such species. ACKNOWLEDGMENTS This work forms part of a PhD funded by the National Parks and Wildlife Service, Ireland (formerly Duchas, the Heritage Service) and Trinity College Dublin. The European sampling of populations was made possible with a grant from the B.S.B.I. Thanks also to Tom Curtis, Clive Jermy, Rob Canning, Julie Smith, Hugh Smith, Sally McSweeney, Padraic Comerford, Lorcan Scott, Michael Tennyson, Joseph Teesdale, George Dyer, Kevin and Louise Barnes, Walter Walsh, Isabelle Diana (Jardin Botanique de l’Universite et de la Ville, Besangon), Jose Daniel Gomez Garcia (Instituto Pirenaico de Ecologfa), Max Debussche (Centre d’Ecologie Fonctionnelle et Evolutive / Centre National de la Recherche Scientifique), Romaric Pierrel (Conservatoire et Jardins Botanique de Nancy), Mikel Lorda, Pedro-Uribe Echebaria, Frederic Bioret (Universite de Bretagne Occidentale), Julian Woodman (Countryside Council for Wales), Clive Falkner (Montgomeryshire Wildlife Trust), Gwent Wildlife Trust, Somerset Wildlife Trust, Hanson Aggregates and B.S.B.I. V.C. recorders: H. J. M. Bowen, Trevor Evans, Stephen Evans, Dave Green, Jean Green, Quentin Kay, Mark Kitchen, Clare Kitchen and Marjorie Wainwright. GENETIC VARIATION IN THREATENED IRISH PLANTS 145 REFERENCES BERRY, R. J. (1971). Conserx’ation aspects of the genetical constitution of populations , in DUFFEY, E. & Watt, A. S. eds. The Scientific Management of Animal and Plant Communities for Conserx’ation , pp. 177-206. Symposium of the British Ecological Society, II., Blackwell Scientific Publications, Oxford. Buza, L., Young, A. & Thrall, P. (2000). Genetic erosion, inbreeding and reduced fitness in fragmented populations of the endangered tetraploid pea Swainsona recta. Biological Conserx’ation 93: 177-186. CARTER, R. N. & Prince, S. D. (1987). Distribution limits from a demographic viewpoint, in Davy, A. J., HUTCHINGS, M. J. & Watkinson, A. R. eds. Plant population ecology. Blackwell Scientific Publications, Oxford. Coart, E., Vekemans, X., Smulders, M. J. M„ Wagner, I., Van Huylenbroeck, J., Van Bockstaele, E. & ROLDAN-RUIZ, I. (2003). Genetic variation in the wild apple ( Mains sylvestris (L.) Mill.) in Belgium as revealed by amplified fragment length polymorphism and microsatellite markers. Molecular Ecology 12(4): 845-857. CURTIS, T. G. F. & McGough, H. N. (1988). The Irish Red Data Book. 1 Vascular Plants. Wildlife Service, Office of Public Works, Dublin. DELGADO, P., Pinero. D.. Chaos A. et al. (1999). High population differentiation and genetic variation in the endangered Mexican pine Pinus rzedoxvskii (Pinaceae). American Journal of Botany 86: 669-676. DEMAURO, M. M. (1993). Relationship of breeding system to rarity in the lakeside daisy ( Hymenoxys acaulis var. glabra). Conserx’ation Biology 7 (3): 542-550. Ellstrand, N. C. & Elam, D. R. (1993). Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics 24: 217-242. FALK, D. A. & Holsinger, K. E. (1991). Genetics and Conserx’ation of Rare Plants. Oxford University Press, New York. FRANKEL, O. H. (1970). Genetic Conserx’ation in Perspective, in FRANKEL, O. H. & BENNETT, E. Genetic Resources in Plants - Their Exploration and Conserx’ation, pp. 469-89. IBP Handbook No. 11, Blackwell Scientific Publications, Oxford. Futuyma, D. J. (1986). Evolutionary' Biology. Sinauer, Massachussetts. Gaudel M, Taberlet, P. & Till-Bouttraud, I. (2000). Genetic diversity in an endangered alpine plant, Eryngium alpinum L. (Apiaceae), inferred from amplified fragment length polymorphism markers. Molecular Ecology 9: 1625-1637. GODT, M. J. W. & HAMRICK, J. L. (1996). Genetic structure of two endangered pitcher plants, Sarracenia jonesii and Sarracenia oreophila (Sarraceniaceae). Am. J. Bot. 83: 1016-1023. Griggs, R. F. (1914). Observations on the behaviour of some species at the edges of their ranges. Bulletin of the Torrey Botanical Club 41: 25^-9. HENGEVELD, R. & Haeck, J. (1982). The distribution of abundance. I: Measurements. Journal of Biogeography 9: 303-316. Newman, D. & Pilson, D. (1997). Increased probability of extinction due to decreased genetic effective population size: Experimental populations of Clarkia pulchella. Evolution 51: 354-362. Palacios, C., Kresovich, S. & Gonzalez-Candelas, F. (1999). A population genetic study of the endangered plant species Limonium dufourii (Plumbaginaceae) based on amplified fragment length polymorphism (AFLP). Molecular Ecology 8: 645-657. Reed, D. H. & Frankham, R. (2003). Correlation between fitness and genetic diversity. Conserx’ation Biology 17(1): 230-237. Ronikier, M. (2002). The use of AFLP markers in conservation genetics - a case study on Pulsatilla vemalis in the Polish lowlands. Cellular & Molecular Biology Letters 7(2B): 677-684. SCHMIDT, K. & Jensen, K. (2000). Genetic structure and AFLP variation of remnant populations in the rare plant Pedicularis palustris (Scrophulariaceae) and its relation to population size and reproductive components. American Journal of Botany 87(5): 678-689. Severns, P. (2003). Inbreeding and small population size reduce seed set in a threatened and fragmented plant species, Lupinus sulphureus ssp. kincaidii (Fabaceae). Biological Conserx’ation 110: 221-229. TEMPLETON, A. R. (1986). Coadaptation and outbreeding depression, in SOULE, M. E. ed. Conserx’ation Biology: The Science of Scarcity and Diversity, pp. 105-116. Sinauer Associates, Sunderland. Massachusett, . Travis, S. E, Maschinski, J. & Keim, P. (1996). An analysis of genetic variation in Astragalus cremnophvlax var. cremnophvlax , a critically endangered plant, using AFLP markers. Molecular Ecology 5: 735-745. Wright, S. (1977). Evolution and the Genetics of Populations. Vol. 3. Experimental Results and Evolutionary Deductions. University of Chicago Press, Chicago. Zawko, G., Krauss, S. L., Dixon, K. W. & Sivasithamparam, K. (2001). Conservation genetics of the rare and endangered Leucopogon obtectus (Ericaceae). Molecular Ecology 10: 2389-2396. 147 Biogeography of the Irish ‘Lusitanian’ heathers N. KINGSTON National Parks & Wildlife Service, Department of the Environment, Heritage and Local Government, 7 Ely Place, Dublin 2, Ireland* and S. WALDREN Trinity College Botanic Gardens, Palmerston Park, Dartry, Dublin 6, Ireland ABSTRACT The ‘Lusitanian’ (or Hibemo-Cantabrian) element of the flora of Britain and Ireland consists of a suite of species found mainly in the west of Ireland and the western Iberian peninsula, including six species from the Ericaceae: Arbutus unedo , Daboecia cantabrica. Erica ciliaris , E. erigena , E. mackaiana and E. vagans. Questions have been raised about whether these taxa recolonised from southern European refugia following glacial episodes, are anthropenic introductions or survived full glacial cycles in more proximal refugia. This group of species is of considerable European conservation importance, and its study is relevant to past and present climate change scenarios. The genetic fingerprinting technique of amplified fragment length polymorphisms (AFLP) was applied to three of the taxa, Erica mackaiana , Erica erigena and Daboecia cantabrica , to assess the population differentiation of the Irish and Continental populations, with a view to understanding whether this can help explain their origin. While AFLP was not found to provide any indications as to the origin of these taxa in the Irish flora, it does suggest that at least Daboecia cantabrica and Erica mackaiana may have arrived through long distance dispersal and so constitute part of the native flora of Ireland. Thus, while several patterns and coincidences can be seen in the ecologies and disjunct distributions of the ‘Lusitanian’ taxa, they should be examined separately, rather than assuming a common origin, when considering how these taxa came to be in the flora of the British Isles. KEYWORDS: AFLP, conservation, Daboecia cantabrica, Ericaceae, Erica erigena, Erica mackaiana, glaciation. INTRODUCTION In biogeographical studies, the term ‘Lusitanian’ is used to describe species whose distribution centres on Portugal and northern Spain, though they frequently extend northwards from there along the Atlantic coast of Europe into France, Britain and Ireland. The origin of this group of species in the flora of the British Isles has been the subject of much discussion as it shows some curious features, most notably in the fact that nine of the 16 species are limited to Ireland and not found in Britain (the Hibemo-Cantabrian element), with the ones that are found in Britain mainly limited to the south-westerly comer. Also, it is noteworthy that six of the species are from the Ericaceae: Arbutus unedo L. (Strawberry-tree), Daboecia cantabrica K. Koch (St Dabeoc’s Heath), Erica ciliaris L. (Dorset Heath), E. erigena R. Ross (Irish Heath, Plate 20), E. mackaiana Bab. (Mackay’s Heath, Plate 21) and E. vagans L. (Cornish Heath). Several different theories have been suggested to account for this Hibemo-Cantabrian distribution (Sealy 1949; Godwin 1975; Webb 1983; Foss & Doyle 1988; Coxon & Waldren 1995; Waldren et al. 2005): • Repeated recolonisation of Ireland from southern European refugia following glacial episodes • Survival of full glacial cycles in proximal refugia • Introduction in historical times by trade and human travel between Ireland and SW Europe *e-mail: Nkingston@duchas.ie 148 ATLANTIC ARC Although these differing theories have circumstantial evidence to support them, they all also have considerable problems, which will be discussed here. While on examination the first theory would seem the most plausible, if the plants repeatedly recolonised, why are several not found in Britain? Were they previously more widespread and continuous, but have become extinct over the bulk of their distribution during post-glacial climatic fluctuations? The current 5°C mean January air temperature isotherm (5°C isotherm) corresponds well with the Lusitanian distribution, and ecological factors may also come into play (e.g. E. vagans, which is found on suitable base-rich substrates in Ireland and the Lizard peninsula (Nelson & Coker 1974)). Long range dispersal by wind would be possible for the tiny seeds of Lusitanian species such as the orchid Neotinea maculata , but for larger seeded species (such as several of the Ericaceae) long distance transport would have to be by zoochory or hydrochory, and would still not explain their absence from Britain. The suggestion that the species survived the last glaciation in situ, or in a proximal refugium, has been postulated and addressed by several authors (Praeger 1910, 1932; Webb 1983; Watts 1977; Coxon & Waldren 1995) although the location of this potential refugium has never been determined. The refugium theory originally gained popularity due to Gortian interglacial fossil records of Daboecia cantabrica , Erica ciliaris and E. mackaiana from sites in Ireland (Coxon & Waldren 1995). Foss (1986), however, has shown that most ericaceous pollen can only be accurately determined to the species level using electron microscopy, and suggests that some of the pollen and macrofossil evidence should be reviewed. This refugium theory was previously disputed on the basis that the species are frost-sensitive, but more recently it has been shown by Bannister & Polwart (2001) that E. mackaiana and E. erigena are in fact frost-tolerant if the plants originated from farther north (i.e. there is a correlation between frost sensitivity and day length). However, even in a proximal refugium, or in the areas of Ireland potentially uncovered by ice during the last glaciation, the survival of these plant species through an extended period of permafrost would seem unlikely (Sealy 1949), and in any case evidence that a taxon was present in the Gortian does not mean it is necessarily now native (Coxon & Waldren (1995) list 37 taxa known from Gortian records that are now absent from Ireland). By far the most controversial theory suggests that the species may be recent introductions, carried by trade links between north-west Spain and the west coast of Ireland, which date back to Neolithic times (5500-4000 years BP) (Foss & Doyle 1988). This could explain why the distributions are so disjunct, and why seed-set is unusual for E. ciliaris and E. mackaiana given that they are outside their natural ecological range. Ericaceous species were widely used in Spain as packing for wine casks and as animal bedding, and E. erigena is planted in and around religious sites in Spain (Foss & Doyle 1988). The necessity to determine the status of these species in Ireland has come to the forefront with the recent exclusion of Erica ciliaris from the Irish Flora Protection Order (1999) on the grounds that it is presumed not native (Curtis 2000). As these species are western European endemics with disjunct and narrow distributions, if native in Ireland this would mean that the Irish authorities would have an international obligation to protect them. If they are not native, then conservation of the species and their habitats would be seen by many as a waste of limited conservation resources. The species and their habitats are variously threatened by overgrazing, peat extraction, habitat drainage, disturbance and forestry (Foss & Doyle 1988; Van Doorslaer 1990; Rose et al. 1996). These problems are compounded by the fact that there is comparatively little ecological information available for the species. On a more investigative level, the origin, colonisation, current and historical distributions of these taxa may be relevant to several research questions. This will be relevant to modelling climate change scenarios if it can be determined that the distributions are a product of postglacial climatic fluxes, and whether in past climatic conditions the species were more widespread and in some cases capable of sexual reproduction (i.e. Erica ciliaris and E. mackaiana ). In addition it will add to the body of literature relating to the postglacial colonisation and biogeography of the current European flora. Finally it will help to decide if any credence should be given to the ongoing question of a proximal refugium off the coast of Ireland. The aim of this research was to try to elucidate some of these questions posed as to the origin of the ‘Lusitanian’ element in the Irish Flora, using recently developed techniques in population genetics to examine populations of three of the Lusitanian heathers, Erica mackaiana , E. erigena and Daboecia cantabrica. Genetic distance for Irish and Continental populations was assessed using the genetic fingerprinting technique of Amplified Fragment Length Polymorphism (AFLP). IRISH LUSITANIAN HEATHERS 149 MATERIALS AND METHODS Tissue samples of three species, Daboecia cantabrica, Erica mackaiana and E. erigena, were collected from populations in Ireland, France and Spain (details are given in Appendix 1), and placed in zip-lock bags in the field and rapidly dried in silica gel, as recommended by Chase and Hills (1991). Between six and ten individuals were sampled from each population. DNA was extracted using a protocol developed for each taxon using modifications on the techniques of Escaravage et al. (1998), O’Hara (2000) and Kingston (2001). AFLP was carried out using standard Applied Biosystem protocols, and of a possible 64 combinations tested, three primer combinations, which successfully amplified and showed the highest amount of variation, were chosen for each species: for Daboecia cantabrica the pairs EcoRl ACA - Msel CTC, EcoRl ACC - Msel CTC, and EcoRl AGG - Msel CAG, giving 215 polymorphic bands; for Erica mackaiana the pairs EcoRl ACA - Msel CAG, EcoRl AAC - Msel CAC, and EcoRl AAG - Msel CAG, giving 263 polymorphic bands; for Erica erigena the pairs EcoRl ACA - Msel CAG, EcoRl ACC - Msel CTC, and EcoRl AAG - Msel CAG, giving 66 polymorphic bands. The agglomerative clustering technique known as unweighted pair-group method with arithmetic mean (UPGMA), using Sprensen distance, was carried out to visualise the relationship between populations and to look at the population structure. Nei’s (1978) gene diversity statistics were calculated in order to quantify and compare the levels of within- and between-population genetic variation. RESULTS AND DISCUSSION DABOECIA CANTABRICA UPGMA of the AFLP data for D. cantabrica clearly separated the Irish populations from those in mainland Europe (Fig. 1), but shows that the diversity in the Irish populations may be due to more recent radiation. The UPGMA tree shows a pattern which commonly represents a founder effect, where long range dispersal of a species to a new location results in fewer genotypes and a thinning of diversity (marked by a bar in Fig. 1). A closer look at the Irish populations shows there to be very little population structure, which may suggest that D. cantabrica was once more continuous over its range but has become fragmented, or that the species in Ireland does not have genetically isolated populations. Nei’s gene diversity statistics show the total amount of diversity in the Irish populations to be on a par with the continental populations (Irish populations HT range from 0T1-0T3; continental populations HT range from 0T1-0T4). The partitioning of the diversity in both the Irish and mainland European populations shows most of the diversity to be within populations, but this ‘within-populations’ proportion is larger in the Irish populations, (Irish populations 1-GSt = 0-87; continental populations 1-Gst = 0-70) and is reflected in the higher level of gene flow between the Irish populations (Irish populations Nm = 2-75; Continental populations Nm = IT 7; Nm of <1 suggests that populations are genetically isolated). The fact that there is implied gene flow shown between the Irish and continental populations (Nm = 2-01) suggests that the separation is a result of postglacial migration. ERICA MACKAIANA UPGMA of the AFLP data for E. mackaiana showed an unusual pattern (Fig. 2), with several distinct clusters, three containing Spanish populations and two of Irish populations, one of the latter containing the plants from Donegal, the other comprising the Galway populations. Several hypotheses may account for this separation: • the Irish populations arose from more than one colonisation event • the divergent Donegal populations are a glacial relict • the Donegal site is composed almost entirely of hybrid plants, which has occurred due to a lack of sexual reproduction {E. mackaiana has never been observed to set seed in Ireland, at any of its localities) Information Remaining (%) 150 ATLANTIC ARC □ ■O IN □ □ S3 FIGURE 1. UPGMA dendrogram of AFLP data for Daboecia cantabrica. Individual samples are labelled with the country of origin and a letter to indicate which population they were collected from. The black bar shows the position of a potential dispersal event & separation of the populations. Information Remaining (%) IRISH LUSITANIAN HEATHERS 151 C3 I* C4 □ Wf □ □ < < do oq << CJ ^ X T3 CJ « § z > D cu e" 2 — i- 3 aj O x E H o x X 4" ~0 . ~ c U 2 ■£ A c til o £ - ^ 3 eg c_ 18. 5.1 < » 2 X 3 *- CJ i 2 w> 4. o 2 i_ x ■o . X m 0 — 3 m >. a. x 4* 0 D 3 CL a 3 n >n o O' o V ~o Z z z z Z Z oo X m in CO CO 1-H rJ in CO CJ •*— * O o o o o o C3 m ro m CO CO J Gj- Gt- in in m Z z Z Z Z c- ON t-~ in cl- o o o in in o o o o o CO ol- cf- CO CO oh in in in in Z z z z X CO % CD 3 on on la ~a 03 3 & o D- ■*— > G 3 0D X •— 3 0*0 3 ■3 > ID CD C _o w o 3 3 U c _o G CD c J cd G Q on .CD £ < od G oo 3 > c3 03 >-»■ eu D O CZ) cz) c3 03 > w- oj C/3 CD O O £ C3 £ 13 a u » '£ LJ LU o o o o o o o o o o o o M M M M M M r— oo oo oo o o o o o o NO 00 — -— — — O O -H — • — l 3 a a a a o 3 3 3 3 3 3 s S 2 k e <3 3 3 3 3 3 -33 2 r— 2 3 3 3 -33 -33 2 2 3 3 3 3 3 g >n *G c3 2 S o U o U * G S .2 3 O O 03 O u >< cd 5 & >> £ c S ‘ G 3 CD <. ^ C/3 2 ‘C d 4— » C/3 < 03 on G ’C 3 3 00 3 33 13 C _o ’3 G z on < — 4 ^ ! 6 o u a u. 3 G > G Z 2 ° £ 2 < G fC 3 s- O U G w P3 ~0 _ _ _ v- aj O 33 TD G on o o Z U U Ml bfi £>XJ O O c D3 CQ 3 O 03 y G C £ o o y£ ^ 4—i ^ C/3 C ^ 3 3 3 W) c c 3 3 3 O O O J oi O O O O O O O O O O • < c G G G o 'g 'g CL, Lh CL CL £ <3J £ ul 00 00 , H — l-H s— t— i 3 3 3 3 3 3 3 3 3 3 3 3 3 .3 .3 .3 .3 .3 .3 .3 .3 .3 .3 .3 .3 .3 "C 'l ’0! ’3 ’3 ’3 3 3 3 3 3 3 tq tq tu ttj t*3 hj < CQ u Q u Q U < CQ "O CD CD CD < CQ Q CD CD CD CD c C C c C c c c c C C c C G CL 00 £ G £ G G CL 00 'g CL 00 G CL OO G CL 00 £ £ £ £ £ ■— LL £ H LL aj •— t— ( £ 3 •—1 CD 3 M OD 3 3 CD 3 1— I CD 3 4—1 157 Ecological profiling as a means of improving mud plant conservation R. V. LANSDOWN 45 The Bundle, Stroud, Gloucestershire GL5 45Q, England* ABSTRACT Many mud plants have very specialised habitat requirements. Suitable habitat is often fragmented, scattered and short-lived. These conditions mean that most populations form part of larger metapopulations and that, although some populations may be lost, many metapopulations appear to be very healthy, in contrast to the perception given by site-based assessment. In this article I suggest that application of ecological profiling to mud plant conservation enables taxon-specific, efficient conservation assessment and monitoring. The process is relatively cheap and often identifies means of addressing mud plant conservation that are more efficient than most current approaches. The long-term benefits of ecological profiling far outweigh the short-term costs of profile preparation. Keywords: Ecological profiling, mud plants, metapopulations. INTRODUCTION There is a guild of plants that are characteristic of what may be described as freshwater mudflats and which are becoming known as “mud plants”. Mud plants provide a link between true aquatic plants and arable weeds, they typically occur in temporary pools, small muddy comers of fields, on poached tracks and in the draw-down zones of ponds, lakes and reservoirs. They are rarely truly aquatic and never tmly terrestrial, but are tolerant of, and may require, both extended periods of inundation and long periods of drought, even extreme drought. It is accordingly difficult to assign them to any particular habitat type or group of habitat types and, as a consequence, mud plants have rarely been the subject of detailed ecological research. By and large, their ecological requirements are still poorly known. Mud plants are subject to particular threat within the Atlantic Arc, mainly due to increased urbanisation and a general trend for agricultural improvement, leading to loss of damp, unmanaged areas that are particularly important for the guild. Three-lobed crowfoot ( Ranunculus tripartitus ) is more or less restricted to this region, whilst the region holds the most important populations of many other rare or vulnerable mud plants such as starfruit (. Damasonium alisma) and floating water-plantain ( Luronium natans ). A number of terms that I employ in this article have a broad colloquial meaning but are used here to refer to a precise circumstance. Definitions of these terms are provided below: Mud plants are defined as “plants which depend upon a process or processes that generate bare substrate, which are able to complete their life cycle on soils that are constantly saturated and may be subject to periodic inundation and which are intolerant of competition”. Site refers to a named area, with a defined boundary, supporting one or more populations of a species. Metapopulation refers to a group of populations connected by exchange of genetic material. Population refers to a discrete group of plants separated from other groups of the same species by an area of a habitat type different to that supporting the species. Sub-population refers to a discrete group of plants separated from other groups by an area of habitat-type similar to that supporting a species. *e-mail: rlansdown@ardeola.demon.co.uk 158 ATLANTIC ARC High dispute tolerance is explained as follows: “When interpreting and using uncertain data, attitudes toward risk and uncertainty may play an important role ... Low dispute tolerance ... include[s] all values, thereby increasing the uncertainty, whereas ... high dispute tolerance ... exclude[s] extremes, reducing the uncertainty” (I.U.C.N. 2001). In this article, I propose that ecological profiling is the most efficient means of identifying actions and targeting funding to address the precise requirements both of the guild and of individual taxa. ECOLOGICAL PROFILING Ecological profiling arose through autecological studies which showed that significant progress could be made in the development of species conservation plans if these were informed by comprehensive review of data from throughout the range of the species and by feedback from throughout the range of the species and by feedback from individuals with a particular interest in the species. It was also clear that there are many sources of very useful information outside professional conservation bodies or outside the U.K. Preparation of an ecological profile involves five main actions; the process is iterative so that new information can be fed into the system. 1 . Contact ecologists and botanists throughout the range of the species with a request for up-to- date information and unpublished data 2. Review the available literature, starting with monographic studies and floras, then trace source references for cited data 3. Compile a basic account, separating data into the following main categories: • Distribution and status • Ecology • Population dynamics • Survey, management and conservation • Monitoring • Further research and survey 4. Circulate the basic review to interested parties for comment, parties to include the individuals and organisations that have responded to previous requests for data and others with a specific interest in the species 5. Input new data including any derived from monitoring and research into the profile with potential to update and modify conclusions An important aspect of an Ecological Profile is application of 'high dispute tolerance’ and a precautionary approach to data review. Preparation of an ecological profile generally requires approximately two months work spread over six months to a year. It is not necessarily most effective, and may be counter-productive, for a profile to be prepared by an “expert” on the taxon. Experts are likely to have preconceptions that may preclude their ability to fundamentally revise aspects of data interpretation or conservation strategy development. The profiler must however have a profound understanding of plant ecology and of the processes that operate within the habitats involved. A core benefit of ecological profiling has been that, in each profile prepared to-date, a means of monitoring and assessing the condition of populations has been identified that is very cost- effective; this is because profiling demands that funds are targeted at addressing the specific needs of the species concerned, based on an in-depth assessment of the best quality information available. In a number of cases, it has also been possible to identify ways in which site managers can carry out monitoring without the need for specialist knowledge of each taxon in their care. Ecological profiling has also led to development of a number of new or modified approaches to mud plant conservation. In particular, it has enabled development of a rigorous approach to metapopulation delineation and a precautionary approach to mud plant population monitoring. These are discussed below to indicate the value of this approach. ECOLOGICAL PROFILING OF MUD PLANTS 159 TABLE 1. THE PROPORTION OF MUD PLANTS IN BRITAIN WITHIN EACH CLASS OF CONSERVATION CONCERN number Percentage of UK Flora Scarce 17 6-83 Not scarce 7 12-50 Rare 2 25-00 Near Threatened 8 11-59 Vulnerable 13 11-21 Endangered 6 14-63 Critically Endangered 4 18-18 Extinct in the Wild 1 5-56 Total 58 1002 CONSERVATION ASSESSMENT Mud plants represent approximately 5% of the U.K. flora; this figure depends on whether an inclusive or exclusive approach is taken to consideration of the guild and what proportion of aliens and sub-specific taxa are included within the flora (using the list from the B.S.B.I. Website, www. bsbi.org.uk/ the flora would include 4250 taxa, of which 211 can be considered facultative or obligate mud plants). A total of 579 taxa have been assigned a conservation concern class. Table 1 shows the percentage of obligate or facultative mud plants listed in each class of conservation concern at a national level (derived from Stewart et al. (1994) and Wigginton (1999)). This shows that, although mud plants represent only 5% of the U.K. flora, they represent a greater percentage of each class of conservation concern and over 10% of all plants of conservation concern in the U.K. In theory, these figures should be reflected in concerted moves to protect and conserve these species. However, the reality of the situation is far more than complex than this analysis suggests. Data on Luronium natans, held by the B.S.B.I. Threatened Plants Database and the database of the Conservatoire Botanique National du Bassin Parisien, were subjected to basic analysis (Lansdown & Wade 2000), grouping records by identifiable (or ‘named’) sites. Thus, where a record is given against a site name that can be related to names in current use, the record is recognised. The site records were then grouped by the first and last record as follows: • 1740-1900 = only recorded before 1900 • 1740-1980 = recorded before 1900 and between 1900 and 1980 • 1900-1980 = only recorded between 1900 and 1980 • 1740-1 900-present = recorded before 1900 and since 1980 • 1900-present = recorded between 1900 and 1980 and since 1980 • Post- 1980 = only recorded since 1980 Figure 1 shows the number of sites in each category for the U.K. and for the Parisian Basin. This shows that in the U.K., whilst there have been some losses, the overall number of sites occupied is fairly constant. In particular, in any given recording period, there have been almost as many new sites found as lost. Reference to distribution maps shows that many lowland populations have been lost, but upland populations are relatively stable. Data from the Parisian Basin show that apart from a large number of sites known only before 1900, very few sites were known until increased interest in the species led recently to a dramatic increase. 160 ATLANTIC ARC recording period Figure 1. Number of sites of Luronium natans in each date class (see text), in the U.K. and Parisian Basin. MUD PLANT ECOLOGY AND POPULATION DYNAMICS One of the most important factors in mud plant conservation is recognition that whilst population assessment based on growing plants is the most obvious approach, seed-banks are able to remain viable for long periods and probably dictate the long-term viability of individual populations. To discount seed-banks will always exaggerate the impression of decline. Mud plant distribution is often patchy, with localised concentrations separated by large areas with few or no populations. Ecological profiles of three-lobed crowfoot ( Ranunculus tripartitus ) (Evans & Lansdown 2000), floating water-plantain ( Luronium natans ) (Lansdown & Wade 2000) and ribbon-leaved water-plantain ( Alisma gramineum ) (Pankhurst & Lansdown 2002) suggest that the concentrations are often linked through dispersal to form metapopulations. Each individual site within one of these complexes may support growing plants for only a relatively short period, unless a factor operates which continuously or repeatedly suppresses succession. When a process such as draw-down of ponds during drought restores the suitability of habitat within a site, or a new area of suitable habitat is formed through actions such as increased rainfall inundating poached ground, plants will either germinate from the seed bank, or will colonise as seed is dispersed from other sites. Many, if not all, populations within a metapopulation are therefore likely to only temporarily support growing plants. These metapopulations are therefore highly dynamic, where individual populations are short-lived but the metapopulation as a whole will almost always support growing plants. Metapopulations are not fixed in area or the number of populations that they include. A single isolated event, such as transport of seed by wildfowl, can link two populations that would otherwise have no contact. Therefore, delineation of a metapopulation is necessarily somewhat artificial, but will become less so as more information becomes available. The spatial range of a metapopulation will also increase over time. For example, in a single year genetic exchange is likely to occur between only a small number of populations; in two years, more populations will be linked by genetic exchange, over ten years a large number of populations will be linked and over thousands of years it is likely that all populations will be linked. Our perception of a metapopulation will also depend on the amount of data available to us. This, in turn, depends in part upon the degree of survey coverage and anecdotal information available. Each metapopulation may contain populations that function in different ways: • Some populations will more or less permanently support growing plants • Some populations will support growing plants during all periods when conditions are suitable but there will be periods when no plants grow. ECOLOGICAL PROFILING OF MUD PLANTS 161 • Some populations recorded in any given year will be new •Some mapped populations may represent failed colonisation attempts and will only once play a role in the function of the metapopulation. Analysing records that allow assessment of the condition of metapopulations may give a very different impression to analysis derived from site records. METAPOPULATION DELINEATION The first stage in mud plant conservation is to define the metapopulation in question. For each population in a metapopulation, the means of genetic exchange must be noted. If no potential for genetic exchange can be identified, then the population must be treated as a separate metapopulation; the number of populations within a metapopulation varies considerably and may be as low as one. This process will enable mapping of metapopulation boundaries. The defined metapopulation or population(s) on a single site will be treated as a monitoring unit. In the U.K., it is rare that a metapopulation of a plant species will extend beyond an individual site, as means of genetic exchange tend to be associated with common management over a defined area. For mud plants, genetic exchange is most likely to be achieved by transport of spores ingested by birds or carried in mud by mammals. Given this basis it is possible to use a high dispute tolerance approach to derive a simple key to assess whether or not two populations should be considered part of the same metapopulation. Site- or population-specific factors must always be considered, and the method should not be thought of as ‘foolproof. The ‘true’ relationships between populations can only be assessed through molecular studies and this is not financially possible for all populations. The method does however provide a practical approach to metapopulation delineation that will improve conservation of plant species that exist as dynamic metapopulations. A KEY FOR ASSESSING WHETHER OR NOT TWO POPULATIONS SHOULD BE CONSIDERED PART OF THE SAME METAPOPULATION 1. Populations are connected by a direct surface hydrological link . One metapopulation Populations not connected by a direct surface hydrological link . 2 2. Populations continuously linked by domestic animal movement . One metapopulation Populations not continuously linked by domestic animal movement . 3 3. Populations less than 3 km apart and not separated by any obvious barrier to wild animal or bird movement . One metapopulation Populations more than 3 km apart or there is an obvious barrier to wild animal movement . 4 4. Populations connected by infrequent and irregular domestic animal movement or there is a reason to believe that regular bird or wild animal movement between populations is to be expected . Assessment must be site-specific Populations are not connected by infrequent and irregular domestic animal movement and there is no reason to believe that regular bird or wild animal movement between populations is to be expected . Different metapopulations Metapopulation delineation involves three stages: 1 . Preliminary delineation based on relevant maps involving application of the key to each known subpopulation and population to assess potential for genetic exchange. 2. Each population and subpopulation should then be visited to confirm or revise the potential for genetic exchange identified from maps. 3. Landowners and managers should be consulted to further confirm or revise the assessment of potential for genetic exchange, with particular emphasis on aspects that cannot be assessed from the maps or site visits, such as active movement of livestock between sites and scattered or isolated landholdings. 162 ATLANTIC ARC Work carried out in the Parc Naturel Regional de la Brenne in central western France showed that in spite of the national database suggesting that the area holds populations of Luronium natans at only two sites, it is extremely abundant. The metapopulation definition method described above was applied to records from the park (Landsdown 2004). Using a short timescale for assessment, this analysis found that the park supports 31 discrete metapopulations, involving 52 populations and 77 sub-populations. METAPOPULATION ASSESSMENT Methods for monitoring the condition of mud plant metapopulations form a logical extension of metapopulation delineation. The monitoring cycle is given a length of three years in this case; this is because it is the maximum known dormancy period of Luronium natans seed: A. Every three years, assess whether the means of genetic exchange between populations can still operate. Any apparent change in circumstances leading to a loss of genetic exchange must be seen as potentially detrimental to the favourable conservation status of the metapopulation. B. Once assessments (1-3 below) have been made for each population within a metapopulation its condition may be assessed. A metapopulation may be considered in favourable condition if a high proportion of populations within it are themselves in favourable condition. What proportion is necessary to maintain a healthy metapopulation is unknown so the threshold for assessment must be set at a high level. • If the means of genetic exchange are still valid for all populations in a metapopulation, then the metapopulation may be considered to be in favourable condition. If not, then management should be undertaken to restore genetic exchange. •If 80% (rounded up) of populations are in a favourable condition then the metapopulation may also be considered to be in favourable condition. If less than 80% are considered favourable, then management will be needed. Each year, record the presence or absence of plants in each population and record whether any of them set seed. Each year, review the data for the three year period up to and including that year (so in 2003 review data from the period 2001 to 2003; in 2004, review data for the period 2002 to 2004, etc.) and ask the following questions: 1. Has conservation management been applied to the population in the last three years? If so, do not make a condition assessment. This is because a period corresponding to the minimum expected seed dormancy period of the species must pass before the effectiveness of management can be assessed. 2. Have plants been recorded at least once over the previous three years? 3. Have plants set seed at the location at least once in the last three years? If the answers to the last two questions are positive then the population may be considered to be in favourable condition because there is known to be seed-set in the population within an acceptable seed-dormancy period. If the answers are negative then management should be considered. CONCLUSION There appear to have been significant losses of mud plants in the U.K. and in continental Europe. However, our perception of the nature and extent of these losses may be based on over-simplistic assessment. Application of ecological profiling to individual species or groups of species will not only enable a more accurate assessment, but also more efficient means of conserving and managing remaining populations. ECOLOGICAL PROFILING OF MUD PLANTS 163 ACKNOWLEDGMENTS I am grateful to the following people for providing data used in this article: Angela Darwell, O. Fritz, Vincent Gaudillat, Andy Jones, Alex Lockton, Francois Pinet, Sarah Whild and P. Wind. Tim Pankhurst read through and commented on various drafts of the text. REFERENCES EVANS, S. B. & LANSDOWN, R. V. (2000). A review of available information on the management history of Ranunculus tripartitus populations, in FITZGERALD, R. AND STEWART, N.F. Three-lobed water crowfoot, Ranunculus tripartitus: report for 1999. Plantlife Report No. 157. Plantlife, London. IUCN (2001) IUCN Red List Categories: Version 3.1. Prepared by the IUCN Species Survival Commission. IUCN, Gland, Switzerland and Cambridge, UK. LANDSDOWN, R. V. (2004). Evaluation de I’etat de sante des populations de fluteau nageant (Luronium natans) (L.) Ratinesque) au sein du Parc Naturel Regional de la Brenne. Unpublished report to the Parc Naturel Regional de la Brenne, France. LANDSDOWN, R. V. & WADE, P. M. (2003). Ecology of the floating water-plantain. Consenting Natura 2000 Rivers; Ecology series No. 9. English Nature, Peterborough. PANKHURST, T. J. & LANSDOWN, R. V. (2002). The ecology and consen>ation status of ribbon-leaved water- plantain (Alisma gramineum Lejeune ) in the United Kingdom. Unpublished Report to English Nature, Peterborough. STEWART, A., PEARMAN, D. A. & Preston, C. D. eds. (1994). Scarce Plants in Britain. Joint Nature Conservation Committee, Peterborough. WlGGENTON M. J. ed. (1999). British Red Data Books - 1. Vascular Plants. 3rd ed. Joint Nature Conservation Committee, Peterborough. 165 Cornish bryophytes in the Atlantic Arc: cell biology, culturing, conservation and climate change J. G. DUCKETT, S. PRESSEL School of Biological Sciences, Queen Mary, University of London, Mile End Rd, London, El 4NS, England and R. LIGRONE Diepartimento di Scienze ambientali, Seconda Universitci di Napoli, Via. Vivaldi 43, 81100 Caserta, Italy ABSTRACT This account highlights the key natural and man-made bryophyte habitats, and the taxa therein, in Cornwall and the Scilly Isles and explores priorities for their conservation. Recent changes in the abundance, both increases and decreases, of some Cornish species, are more likely due to habitat changes e.g. altered management practices and neglect of heathlands together with more subtle climate changes, rather than to S02 pollution or to increasing N2 deposition. Two liverworts recently introduced into the Scillies are now spreading widely in mainland Britain whilst another earlier alien moss Campylopus introflexus appears to be overrunning native species on Cornish heaths. Electron microscope studies have revealed diverse symbioses in Cornish liverworts involving ascomycete, basidiomycete and glomeromycete fungi, and that oil bodies persist through long periods of desiccation under natural conditions. The survival of Weissia controversa var. densifolia and Cephaloziella spp. on lead and copper-contaminated soils may be related to their extremely thick-walled rhizoids. In vitro cultivation of Cornish mosses, as well as clarifying taxonomic problems, is being used to preserve rare taxa, especially those on unstable sites. Manipulation of the medium induces the copious production of gemmae or tubers which can be used to bulk up the cultures for reintroduction trials and for molecular analyses of genetic diversity within and between populations. KEYWORDS: axenic cultures, conservation, environmental change, heavy metals, introductions, protonemata, symbioses. INTRODUCTION With its exceptional topographic diversity, ranging from granitic tors, moorlands, heaths and bogs, marshes, woodlands, streams and rivers, sand dunes, salt marshes and sea cliffs, to a veritable surfeit of man-made habitats, including arable fields, roadside banks, quarries, china-clay works and metalliferous mine wastes, coupled with the most strongly Atlantic-Mediterranean climate in the British Isles, Cornwall and the Isles of Scilly have long been a mecca for bryologists. Not surprisingly these have a rich bryophytic flora, with about 445 taxa of mosses (approximately 58% of the British total and 36% of that for all of Europe including the Azores) and 165 liverworts (approximately 57% of the British total) (Holyoak 1997). New taxa continue to be added to these impressive lists at one or two per year (Holyoak 1999). Detailed analyses of the geographical relationships of British bryophytes (Ratcliffe 1968; Hill & Preston 1998) reveal that Cornwall holds pride of place along the Atlantic Arc where Atlantic and Mediterranean floristic elements meet. Hyperoceanic southern-temperate and oceanic southern- temperate elements are also well represented in the bryophyte flora. In these categories particularly noteworthy are the only English populations of Cyclodictyon laetevirens , Ulota calvescens and Lejeunea mandonii whilst Campylopus pilifer is restricted to Cornwall and Devon. Also striking, 166 ATLANTIC ARC in comparison with the rest of Britain, is the frequent occurrence of Epipterygium tozeri and Funaria attenuata. By contrast, the rarity or absence in Cornwall of typical Mediterranean- Atlantic taxa like Targionia hypophylla, Bryum canariense, Leptodon smithii and Habrodon perpusillus is yet to be explained. It is widely recognised that the Cornish bryophyte flora’s designation as ‘Exceptionally well covered’ (Hill et al. 1991) owes much to the endeavours of Jean Paton and the publication of the most up-to-date recent county bryophyte Flora (Paton 1969). This exemplary work provides an excellent summary of the topography of the county, a historical survey of bryological activity therein and exceptionally clear habitat details for every species. In the historical context it is noteworthy that several of the early bryologists in Cornwall have taxa named after them, for example Petalophyllum ralfsii after Ralfs (although the type specimen came from Wales), Fissidens curnovii after Cumow, Epipterygium tozeri after Tozer and more recently Cephaloziella nicholsonii, after Nicholson. Since Paton’ s Flora perhaps the most notable contributor to Cornish bryophytes has been Harold Whitehouse with his now legendary excursions to the Lizard peninsula, and Cornish specimens featuring large in his seminal work on the use of axenic cultures for elucidating taxonomic problems and leading to the description of new species such as Dicranella staphylina (Whitehouse 1969) and Ditrichum comubicum (Paton 1976). Most recently David Holyoak’s (1997) accounts of mosses and hepatics in the Cornish Red Data Book (Spalding 1997) highlight the status of rare and scarce taxa and point out priorities for conservation. Several Cornish species also feature in the British Red Data Book of Mosses and Liverworts (Church et al. 2001). Currently the same author’s (Holyoak 1999) tetrad mapping is documenting interesting changes in distributions since Paton’ s work in the 1960s, some of which will need incorporation into future conservation strategies. Rather than simply updating previous accounts of Cornish bryophytes or reiterating previous incisive analyses of their geographical relationships (Hill & Preston 1998) our aims here are: 1 . to highlight some of the key Cornish bryophyte habitats and interesting taxa therein, set in the dual contexts of conservation biology and environmental change. 2. to describe some of our current work on Cornish bryophytes directed towards furthering understanding of their cell and reproductive biology. Nomenclature in this account follows Blocked & Long (1998) and full details of the preparation of the specimens used in the illustrations can be found in Duckett et al. (2001) and Ligrone & Duckett (1998). KEY NATURAL HABITATS THE COASTLINES In terms of national, and indeed international importance, topping the list of Cornish bryological habitats must be those harbouring taxa with south-western, Mediterranean and Atlantic distributions. Coastal cliffs with their plethora of microniches are amongst the richest hunting grounds in Britain for Fossombronia spp., Riccia spp., Cephaloziella spp., Gongylanthus ericetorum (Malloch 1972) and spring-fruiting Pottiaceae. Cephaloziella tumeri and Ditrichum subulatum are a pair of species characteristic of crumbling shaley soil on low cliffs around sheltered Cornish estuaries whilst calcareous sand dunes on the north-west coast provide habitats for calcicoles including Didymodon acutus, D. ferrugineus, Thuidium abietinum, Southbya tophacea and Petalophyllum ralfsii. Damp sandy ground on Tresco was the most isolated British locality for Haplomitrium hookeri , but this population like those on the Cornish mainland now appears to be extinct. For several years after its discovery in 1958 by Coombe and Whitehouse Hennediella stanfordensis (Whitehouse 1961) was only known from the Lizard peninsula. Molecular studies are now required to determine whether the coastal Cornish populations of this species match those discovered in Britain along the River Wye and on Magnesian limestone, and at its other sites in mainland Europe, California and Australia (Hill et al. 1992). Whilst there are general conservation concerns, associated with increasing tourist pressures and intensification of agriculture, the status of most of the species in coastal habitats appears to be CORNISH BRYOPHYTES 167 relatively stable although the extreme rarity of Riccia bifurca and the unexplained recent decline of Tortula cuneifolia and T. canescens are cause for concern. The abandoning of grazing along the cliffs has seen scrub taking over from open coastal grassland, not to mention the spread of Carpobrotus edulis after mild winters. Heavier use of the footpaths and intensification of arable agriculture have further encroached on the niches of coastal bryophytes. Riccia crozalsii , Fossombronia maritima and many Pottiaceae have lost sites as a consequence. However, the decline of Tortula cuneifolia in Cornwall and elsewhere across the Atlantic Arc was already occurring long before the increased tourist pressures and the intensification of agriculture. Probably the greatest imponderable regarding the future of the Cornish coastal bryological gems is climate change; hotter, drier summers would almost certainly favour their increase whereas cooler, wetter summers, leading to the establishment of closed vegetation over formerly broken ground, would make their survival more problematic. Another interesting aspect of the biology of coastal Atlantic Arc taxa is their diverse modes of survival through periods of summer drought. Whilst many of the annual ephemeral taxa survive from one growing season to the next exclusively by spores or vegetative diaspores usually packed with lipids (Fig. 1) (see Longton & Schuster 1983; During 2001, for reviews), several perennial liverworts aestivate as stem or thallus tubers also packed with lipids. The most striking examples are Petalophyllum (Figs 3a,b), Fossombronia maritima, Southbya spp. and particularly Gongylanthus, where gemmae are absent and only female plants are known from Cornwall (Paton 1999). A further desiccation problem that has long troubled liverwort taxonomists is the rapid disappearance of oil bodies in herbarium specimens whereas these are invariably present in wild material even after long periods of drought. A series of desiccation experiments on Southbya (Fig. 2 c,d) confirms that oil bodies remain intact as long as specimens become desiccated under natural conditions but disappear when these are dried out rapidly in the laboratory. SHELTERED VALLEYS As far as Atlantic species are concerned, Cornwall lacks the bryological diversity found in North Wales, the Lake District, western Scotland and western Ireland. Nevertheless, the wooded valleys contain unique assemblages of notable taxa like Dumortiera hirsuta, Fontinalis squamosa var. curnovii, Cryphaea lamyana and numerous Fissidens spp. including F. cumovii, F. rivularis, F. polyphyllus and F. serrulatus and the only British locality for Telaranea nematodes. Dripping cliffs around the Lands End peninsula are a stronghold for Philonotis rigida , sea caves the only English localities for Cyclodictyon laetevirens whilst Lejeunea mandonii still occurs in significant quantity (J. G. Duckett, R. Porley and F. Rumsey 11 May 2003) on serpentine rocks at Kynance where it was probably first discovered by Nicholson in the 1930s. TORS. HEATHS AND BOGS Granite tors and cams are southerly outposts for several bryophytes with sub-alpine and Atlantic affinities including Plagiochila spinulosa, P. punctata, Douinia ovata, Scapania gracilis, Lepidozia cupressina, Cynodontium bruntonii. Polytrichum alpinum, Rhabdoweisia denticulata and Grimmia patens. Antitrichia curtipendula remains locally plentiful on Bodmin Moor in striking contrast to its catastrophic decline in Wistman’s Wood on nearby Dartmoor. Most recently Holyoak (1999) has recorded conspicuous declines in Andreaea rothii and Racomitrium aquaticum throughout Cornwall and R. fasciculare in west Cornwall. Increased nitrogen deposition on these acidic rock taxa has been suggested as a possible explanation but has yet to be tested. Antitrichia was almost certainly eliminated from most of southern England before 1900 by S02 but this cannot explain its decline at Wistman’s Wood. The most notable and striking change since the 1960s on Cornish bogs and heaths (and also elsewhere in Britain) has been the decline of the once abundant Campylopus brevipilus and the spread of the introduced species C. introflexus (see Plates 23 and 24). There have also been many losses of Cornish populations of Sphagnum spp., Scorpidium scorpioides, Calliergon sarmentosum and Wamstorfia spp. due to drainage and/or scrubbing over of small mines and flushes. In terms of losses of sites, though not perhaps surface area, some of these species have declined more drastically than Campylopus brevipilus which occupies much more extensive wet heathlands. The endangered liverwort Jamesoniella undulifolia still persists in small quantity at its two Cornish sites (J. G. Duckett and S. Pressel, 12 May 2003) but Cephaloziella dentata, often collected by Nicholson on the Lizard in the 1930s, has not been seen at its last recorded site since 1967. By extreme good fortune a new population was discovered in 2004 (D. Holyoak, pers. com.). 168 ATLANTIC ARC FIGURE 1. a, b. Ditrichum cornubicum. a. Rhizoidal tubers produced in culture on medium lacking nutrients, b. Slightly squashed preparation exhuding copious lipid droplets (arrowed) characteristic of long-lived diaspores. c. Ditrichum cylindricum\ rhizoidal tuber from freshly collected wild specimen; slightly squashed and exhuding massive lipid droplets (arrowed), d, e. Ditrichum flexicaule. d. Chloronemal filaments producing previously undescribed terminal filamentous gemmae (arrowed) in culture on nutrient medium, e. Detail of a gemma; the abscission zone between two cells at the base is arrowed, f-j. Protonemata of Weissia in culture and in the wild, f, g. Weissia rostellata illustrating typical protonemal morphology in the genus: tufts of determinate chloronemal branches are growing from an undulating caulonemal filament. Note the highly pointed apices to the chloronemal filaments, g. Detail of a caulonema showing the lack of wall pigmentation, oblique septa and elongate chloroplasts. h-j. W. contraversa var. densifolia. h. Tufts of chloronemata, with highly attenuate pointed apices growing from caulonemata with highly pigmented walls, i. Detail of a caulonema showing the thick highly pigmented wall and the prominent cuticle layer, j. Wild protonema, with identical morphology to that produced in culture. k,l. Rhizoids of Cephalozella massalongoi. Note the very thick walls and fungal hypha in k (arrowed). Scale bars = 50 pm (a-c, e, g, i, k, l)./200 pm (d, f, h, j). CORNISH BRYOPHYTES 169 FIGURE 2. a, b. Weissia controversa var. densifolia. Transmission electron micrographs of the walls of the caulonema (a) almost 1-5 pm thick and much thinner (c. 0-3 pm) chloronema (b). In ‘a’ note the cuticle (arrowed) with granular material between this and the electron-dense wall, c, d. Southbya tophacea , oil bodies in fully hydrated (c) and dehydrated (d) specimens. Note the disappearance of the vacuole in the dehydrated state whilst the oil bodies remain virtually unchanged. N, nucleus with condensed chromatin; OB, oil bodies; V, vacuole. Scale bars = 2 pm (a, c, d). 1 pm (b). WOODLAND AND ROADSIDE TREES AND SCRUBS As to be expected from the mild climate and absence of significant atmospheric pollution, Cornish woodlands, scrub and roadside trees support a diverse epiphyte flora. Neckera pumila, Cryphaea heteromalla, Orthotrichum spp. and Ulota spp. are often abundant whilst Cololejeunea minutissima is frequent along the coast and ancient gorse bushes provide an unusual habitat for the Atlantic species Coluro calyptrifolia just as they also do in Brittany. The recent discovery that Colura is widespread, in west and south Wales, on sallows in quarries (Bosanquet 2004) suggests that it may also occur in the same situations in Cornwall. Eastern species like Aulocomnium androgynum and the three Orthodicranum spp., Dicranum montanum, D. tauricum and D. flagellare, are absent but Zygodon conoideus appears to be increasing (Holyoak 1999). The spread of this last species elsewhere in southern England is attributed to a reduction in sulphur dioxide pollution but this cannot be true in west Cornwall. 170 ATLANTIC ARC MAN-MADE HABITATS In terms of both their bryological diversity and frequency of scarce and rare taxa the man-made habitats of Cornwall are amongst the most interesting anywhere in Britain - indeed perhaps the world. Although an increasing number of sites are being designated as S.S.S.I. (Sites of Special Scientific Interest) on the basis of their bryophytes, management strategies (in particular for metalliferous sites) are now being worked out to ensure the survival of the rarities, especially pioneer taxa, against scrub encroachment and competition from more aggressive bryophytes, not to mention large scale landscaping, (Holyoak 2000). CORNISH HEDGES The traditional Cornish hedge, comprising a double row of stones with an earth filling and a capping of turf, supports a variety of scarce and uncommon species including Epipterygium tozeri, Fissidens curvatus, Schistostega pennata , Ditrichum subulatum, Cephaloziella turneri, Tortula solmsii, and several Weissia spp. most notably W. multicap sularis whose extant world distribution is currently reduced to three such sites, all in Cornwall (Holyoak, pers. comm.). QUARRIES Granite quarries mainly support common calcifuge bryophytes with the spoil heaps of the kaolinised derivative, better known as China-clay, colonised by a suite of pioneer species of extremely acidic substrata, most notably Discelium nudum far away from the centre of its world distribution on unstable clay banks in the Pennines (Hill et al. 1994). Recent examination of the Cornish site in May 2003 (Duckett & Pressel, May 2003) reveals that it still persists along erosion gullies despite the ‘Zigguratisation’ (i.e. landscaping into regular terraces) of the former mounds. Damp areas in the old China-clay workings are often dominated by Blasia pusilla together with the two homworts Phaeoceros laevis and Anthoceros punctatus and are the only British sites for Marsupella profunda. Indeed, in terms of abundance at some of its sites, Cornwall is very much the world stronghold for this rare Lusitanian liverwort known elsewhere only from Portugal and Macaronesia (Paton 1999). One of only two British sites for Amblystegium radicale is also an old China-clay working. ARABLE FIELDS In the 1960s the arable fields of Cornwall were renowned for their bryological diversity. Bryologists of the 1960s generation have fond memories of the abundance of Riccia spp., homworts and tuberous mosses in virtually every field. Although the bulb fields of the Isles of Scilly remain special with their carpets of Sphaerocarpos, homworts and Riccia ssp. , particularly R. crystallina in its only English location, on the mainland much has changed. Things may be changing due to the increasing use of fertilisers and annual rather than biennial ploughing which militate against species with two-year life cycles like Weissia longifolia var. longifolia. In this context it would be interesting to investigate whether increasing use of nitrogen fertilizers is leading to a decrease in homworts which contain their own N2-fixing endosymbionts. Current management of the EchiumJ Chrysanthemum fields near Lands End (Butterfield 2005) also offers much scope for furthering understanding of the reproductive strategies of arable bryophytes in general, not to mention the possibility of establishing a “stable” field site for Weissia multicap sularis from the hedge nearby. METALLIFEROUS MINE WASTES For those who love small unprepossessing plants, difficult to identify even under the microscope, the metalliferous mine wastes of Cornwall are an absolute must. Here, copper-contaminated earth is encrusted with blackish mats of Cephaloziella spp. particularly C. massalongi, C. stellulifera, C. integerrima and the endemic C. nicholsonii. A further species, C. calyculata is mainly confined to walls at metalliferous sites. Equally non-charismatic are mine waste Ditrichum spp. including another of Britain’s very few endemic bryophytes D. cornubicum. This is also confined to soils with high concentrations of available copper as is the recently found Scopelophila cataractae (Corley & Perry 1985). Weissia controversa var. densifolia (Plate 25) together with var. controversa, and apparently intermediate plants, are a conspicuous feature of spoil heaps and diggings around Cornish lead mines together with the aptly named Ditrichum plumbicola in much smaller quantities. CORNISH BRYOPHYTES 171 ALIENS AND THE ISLES OF SCILLY Apart from their rich flora of coastal bryophytes like that on the mainland the Scillies boast one of the highest concentrations of bryophytes introduced into Britain, presumably with horticultural plants. Whilst some of these, for example Calyptrochaeta apiculata, Sematophyllum substrumulosum , and Telaranea murphyae, remain largely restricted to their original sites, two others Lophocolea bispinosa and even more so L. semiteres, appear to be spreading widely as judged from recent new vice-county records (Blackstock 1999, 2000, 2001, 2002). These two often become dominant in their new localities and are an obvious threat to native species as has almost certainly been the case with two earlier introductions Orthodontium lineare and Campylopus introflexus. However, attribution of the decline of native bryophyte species solely to invasion of their habitats by alien taxa needs to be regarded with caution. Recent records of Sematophyllum substrumulosum from East Cornwall (v.c. 2) and West Sussex (v.c. 13) suggest that this species may also become more widespread in the future. For many years the decline of the native moss Orthodontium gracile during the 20th century was attributed solely to invasion of its habitats by the more aggressive introduction O. lineare (Duckett et al. 2001). The latter species, first collected in Cheshire in 1910, is now widespread over much of England but significantly rarer in Cornwall. A recent survey and evaluation of the native O. gracile revealed that more likely it declined, possibly as a result of atmospheric pollution, before the invasion by O. lineare which then exploited the unoccupied sites (Porley & Matcham 2003). The rarity of O. lineare in air pollution-free Cornwall (and similarly in Brittany, Wales and Ireland across the Atlantic Arc) may be due to the absence of recently available niches. Similarly, the more recent decline of native Campylopus brevipilus has been attributed to the rapid spread of C. introflexus, first recorded in Sussex in 1941. However, other factors such as increased nitrogen deposition may also be involved. Where the two species grow together on the Lizard heaths C. introflexus is now the more abundant (Plate 23). Fortunately another rare Cornish Campylopus , C. pilifer, is typically found on rocks where C. introflexus does not grow (Holyoak 1997). Less obvious, but perhaps even more sinister, is that C. introflexus now forms a thick sward over the formerly bare ground between the vascular plants (Ericaceae, Schoenus , Molinia ) which was the habitat of small liverworts like Cephaloziella intergerrima and C. dentata. To understand more fully the nature of the interactions between C. introflexus and C. brevipilus experimental plots have now been established to record encroachment at natural interfaces and to discover whether clumps of C. brevipilus survive transplantation into swards of C. introflexus and vice versa (Plate 24). These trials have now been running for nearly three years during which time the natural interfaces have remained unchanged and the two species continue to coexist with undiminished vigour on the transplantation sites. To date therefore there is no direct evidence that C. introflexus out-competes C. brevipilus. However the possibility that competition occurs during initial stages of establishment, rather than between mature plants, now requires investigation. SYMBIOSIS IN LIVERWORTS Though the ubiquitous presence of nitrogen-fixing Nostoc colonies in homworts and Blasia is well known to bryologists (Rai et al. 2000), less well appreciated is the fact that liverworts form a wide range of associations with fungi paralleling vascular plant mycorrhizas (Read et al. 2000). In contrast, fungal symbioses have not been detected in mosses. Electron microscopic studies have revealed the identities of the fungi involved with cross-infection experiments providing details of the specificity of the associations. All the various kinds of association are particularly well represented in the Cornish flora. All the Cornish homworts and Marchantiales, (except for Aneura pinguis and Cryptothallus which contain basidiomycetes and are closely similar to orchidaceous mycorrhizas (Ligrone & Duckett 1993) and Riccia and Riccardia which are fungus-free) and the Metzgerialean genera Petalophyllum and Fossombronia , plus Haplomitrium (Carafa, Duckett & Ligrone 2003) contain aseptate vesicular arbuscular glomeromycotean fungi (Schiipler et al. 2001). As illustrated in Petalophyllum (Fig. 3a, b), the ventral cells of the thallus contain large trunk hyphae with much branched, finer arbuscular hyphae. Similar to VA mycorrhiza in vascular plants the arbuscules are short-lived as evidenced by the masses of collapsed hyphae in many of the infected cells. Axenic thalli of Pellia may be similarly infected by seedlings of Plantago lanceolata indicating a wide host range for these hepatic fungi. 172 ATLANTIC ARC FIGURE 3. Endophytic fungi in liverworts, a, b. Petalophyllum ralfsii. a. Scanning image of arbuscules growing out from a broader trunk hypha in a ventral thallus cell. b. Section through the margin of the lipid¬ laden tuber with fungus in the subjacent cell. C, mass of collapsed degenerating hyphae; L, lipid droplets, c, d. Southbya. a, c. Central stem cell packed with hyphae. b, d. Dolipore septum in the fungal endophyte, e, f. Cephcilozia connivens. e. Swollen rhizoid tip packed with hyphae. f. Simple septum with Woronin bodies (arrowed). Scale bars = 10 pm (a). 5pm (b, c, e). 1 pm (f). 0-5 pm (d). CORNISH BRYOPHYTES 173 Endophytes (Fig. 3c), identified as basidiomycetes from the presence of dolipore septa (Fig. 3d), occupy specific regions of the stems of many Lophoziaceae, Nardia , Saccogyna, Tritomaria , Gongylanthus and Southbya as illustrated in Paton’s flora (1999). Unlike the basidiomycetes in orchid roots, those in the liverworts are not digested by their hosts, and are extremely host specific (Duckett et al. 2005). Molecular studies of Gongylanthus ericetorum, Southbya tophacea and Sacogyna viticulosa embracing both the hosts and their fungal endophytes are likely to provide particularly incisive new information about the origins, and relationships between, the different populations of these taxa across the Atlantic arc. Yet a third kind of association is found in the Lepidoziaceae, including both Cornish species of Telaranea, Calypogeiaceae, Cephaloziaceae and all the Cephaloziellaceae, including all the taxa from metalliferous sites (Duckett et al. 1991). The fungi here are restricted to the rhizoids which often have swollen tips (Fig. 3e). The presence of simple septa and Woronin bodies (Fig. 3f) identifies them as ascomycetes. A series of isolation and resynthesis experiments have revealed a wide host range for the fungus even extending to the Ericaceae (Duckett & Read 1995). This knowledge of the biology of the fungal endophytes has important implications for the ex situ conservation strategies of the liverworts concerned. Reintroduction of Cephaloziella spp. from axenic cultures following ploughing of overgrown metalliferous sites should not pose a problem as their ascomycetes will almost certainly be already present in the Ericaceae. However preservation in culture of the basidiomycete-containing taxa. because of the high specificity of the associations, should also embrace isolation and culturing of the fungi. IN VITRO CULTIVATION - A MANY EDGED SWORD In vitro cultivation has already had significant impacts on bryophyte systematics at both specific (e.g. bulbiferous Pohlias spp. and tuberous Bryum spp. (Smith 2004)) and at higher ranks in the taxonomic hierarchy (e.g. the reclassification of Oedipodium with the Tetraphidales (Newton et al. 2000)), and seems certain to maintain this role in the future. Our present comparative studies on Weissia provide a good example of the taxonomic potential of culturing studies. The protonemata of this genus (to date based on all eleven of the British species and three varieties) are most distinctive (Fig. 1 f-i). They comprise undulating caulonemal main axes with tufts of chloronemal side branches with highly acuminate apices. Protonemata have the same morphology in both culture and in nature and in both instances diaspores are absent, not altogether surprising for a genus where most species are autoecious and all species regularly produce sporophytes. Most interesting is the discovery that W. controversa var. densifolia stands apart from all the other taxa in the presence of very deeply pigmented caulonemata with a prominent cuticle. Electron microscopy (Fig. 2 a.b) reveals that the caulonemata are also extremely thick-walled with finely granular material occupying the area between the wall proper and the cuticle. In contrast the caulonemata of W. controversa var. controversa is much less heavily pigmented like those of other Weissia spp. (Fig. lf-g; Plate 26). These striking differences in the protonemata suggest that var. densifolia could well merit elevation to specific status. However, protonemal morphology in intermediate forms from Cornish mining sites has still to be investigated. In a similar taxonomic vein clear cut differences in growth rates and gross morphology between the protonemata of Weissia longifolia var. longifolia and var. angustifolia are again indicative of elevation to specific status. Likewise, the fact that coastal specimens of Phascum cuspidatum var. piliferum ( -Tonula acaulon ) from Cornwall and Brittany, a taxon highly characteristic of the Atlantic Arc, maintain their diagnostic features when cultured from both spores and gametophore fragments alongside typical P. cuspidatum , makes this variety an obvious candidate for specific recognition. Taxonomic considerations aside, the discovery of remarkably thick-walled protonemata in lead mine Weissia controversa var. densifolia and similar very thick-walled rhizoids in Cephaloziella spp. (Fig. lk,l) from metalliferous sites suggests a possible link with their survival on contaminated substrata. Thus we are now investigating the uptake of heavy metals by x-ray microanalysis and dissecting the composition of the walls by immunocytochemistry (Ligrone et al. 2002) to see if heavy metal tolerance resides in their exclusion via specific wall moieties. 174 ATLANTIC ARC Maintenance of taxa in culture, should they become extinct in nature, is another major goal for the ex situ conservation programme. Fortuitously the Ditrichum spp. from metalliferous sites (Fig. la,b) grow extremely well in culture (Arts 1994). When Whitehouse began to culture tuber- producing mosses he discovered by chance that many of these also produced protonemal gemmae (Whitehouse 1987). These were previously unknown but have now increasingly been found in nature (Duckett et al. 1998; Duckett et al. 2004). Since Whitehouse’ s original observations protocols have been devised to promote propagule production in culture (Goode et al. 1992). Thus culturing on media low in nutrients favours the production of tubers whilst nutrient-rich media promote chloronemal gemmae formation sometimes unexpectedly in taxa where vegetative propagules of any kind were otherwise unknown. Ditrichum flexicaule (Fig. ld,e), a species claimed by Arts (1994) to produce neither tubers nor gemmae, is a recent example. Addition of abscissic acid and desiccation are the cues for dedifferentiation of chloronemal filaments into spherical brood cells (Goode et al. 1993). In the absence of spores, gemmae, tubers, chloronemal fragments and brood cell suspensions are ideal inocula for reintroduction trials. Using these propagules, rare species (e.g. Zygodon gracilis , Didymodon glaucus and Seligeria carniolica) can be established on fragments of their native substrata, as a staging post before their return to the wild. Bulking up cultures of taxa, present only in very small quantities in nature, e.g. the Cornish Weissia multicap sularis , Ditrichum cornubicum and D. plumbicola, using diaspores as inocula is also important. It is an essential prerequisite not only for molecular analyses aimed at elucidating genetic variation within and between populations, but may also prove essential for establishing new colonies on managed metalliferous sites. ACKNOWLEDGMENTS Silvia Pressel thanks NERC and the Royal Botanic Gardens, Kew for the financial support of a CASE studentship. David Holyoak collected (under license) the specimens of Ditrichum cornubicum used in this study. He and Ron Porley kindly provided key information on collecting sites. This article would not have been possible without the knowledge and inspiration imparted to Jeff Duckett by Jean Paton and the late Harold Whitehouse. REFERENCES ARTS, T. (1994). Rhizoidal tubers and protonemal gemmae in European Ditrichum species. Journal of Bryology 18: 43-61. Blackstock, T. H. (1999). New vice-county records and amendments to the census catalogue. Hepaticae. Bulletin of the British Bryological Society 73: 36-39. Blackstock, T. H. (2000). New vice-county records and amendments to the census catalogue. Hepaticae. Bulletin of the British Bryological Society 75: 40-45. Blackstock, T. H. (2001). New vice-county records and amendments to the census catalogue. Hepaticae. Bulletin of the British Bryological Society 77: 33-37. Blackstock, T. H. (2002). New vice-county records and amendments to the census catalogue. Hepaticae. Bulletin of the British Bryological Society 79: 3 8 — 42. BLOCKEEL, T. L. & LONG, D. G. (1998). A check-list and census catalogue of British and Irish bryophytes. British Bryological Society, Cardiff. BOSANQUET, S.D.S. (2004). Colura calyptrifolia in Wales. Field Bryology 82: 3-5. BUTTERFIELD, L. C. (2005). Echium plantagineum L. (Purple Viper’ s-bugloss) at Boscregan, St Just - its conservation and management, in Leach S. J., Page, C. N., Peytourau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc, pp. 331-333. Botanical Society of the British Isles, London. Carafa, A., Duckett, J. G. & Ligrone, R. (2003). Subterranean gametophyte axes in the primitive liverwort Haplomitrium harbour a unique type of endophytic association with aseptate fungi. New Phytologist 160: 185-197 Church, J. M., Hodgetts, N. G., Preston, C. D. & Stewart, N. F. (2001). British Red Data Books: Mosses and liverworts. Joint Nature Conservation Committee, Peterborough. CORLEY, M. F. & Perry, A. R. (1985). Scopelophila cataractae (Mitt.) Broth, in South Wales, new to Europe. Journal of Bryology 13: 323-328. DUCKETT, J. G., Renzaglia, K. S. & Pell, K. (1991). A light and electron microscope study of rhizoid- ascomycete associations and flagelliform axes in British hepatics: with observations on the effects of the fungus on host morphology. New Phytologist 118: 233-257. Duckett, J. G. & Read, D. J. (1995). Ericoid mycorrhizas and rhizoid-ascomycete associations in liverworts share the same mycobiont: isolation of the partners and their resynthesis in vitro. New Phytologist 129: CORNISH BRYOPHYTES 175 43SM-47. Duckett, J. G., Goode, J. A. & Matcham, H. W. (2001). Studies of protonemal morphogenesis in mosses VIII. The gemmiferous protonemata of Dicranoweisia and Orthodontium. Journal of Bryology 23: 1 81— 194. Duckett, J. G., Fletcher, P„ Francis, R., Matcham, H. W., Read, D. J., Russell, A. J. & Pressel, S. (2004). In vitro cultivation of bryophytes: a review of practicalities, progress, problems and promise. Journal of Bryology 26: 3-20. Duckett, J. G., Pocock, K., Read, D. J. & Ligrone, R. (2005). Basidiomycetous endophytes in jungermannialean (leafy) liverworts have novel cytology and species-specific host ranges: a cytological and experimental study. Canadian Journal of Botany (in press). Duckett, J.G. & Pressel, S. (2003). Studies of protonemal morphogenesis in mosses. IX. Discelium nudum : exquisite pioneer of unstable clay banks. Journal of Bryology 25: 241-245. DURING, H. J. (2001). Diaspore banks. Bryologist 104: 92-97 . Goode, J. A., Duckett, J. G. & Stead, A. J. (1992). Towards an understanding of developmental interrelationships between chloronema, caulonema, rhizoids and plates in mosses; a comparative study. Cryptogamic Botany 3: 50-59. Goode, J. A., Stead, A. D. & Duckett, J. G. (1993). Redifferentiation of moss protonemata: an experimental and immunofluorescence study of brood cell formation. Canadian Journal of Botany 71: 1510-1519. HILL, M. O., Preston, C. D. & Smith, A. J. E. (1991). Atlas of the bryophytes of Britain and Ireland. Volume 1. Liverworts (Hepaticae and Anthocerotae). Harley Books, Colchester. HILL, M. O., Preston, C. D. & Smith, A. J. E. (1992). Atlas of the bryophytes of Britain and Ireland. Volume 2. Mosses (except Diplolepidae). Harley Books, Colchester. Hill, M. O., Preston, C. D. & Smith, A. J. E. (1994). Atlas of the bryophytes of Britain and Ireland. Volume 3. Mosses (Diplolepidae). Harley Books, Colchester. Hill, M.O. & Preston, C.D. (1998). Bryological Monograph. The geographical relationships of British and Irish bryophytes. Journal of Bryology 20: 127-226. Holyoak, D. T. (1997). Mosses (Musci), Liverworts (Hepaticae and Anthocerotae), in Spalding, A., (Ed.) Red data book for Cornwall and the Isles of Scilly. Pp. 37-92. Croceago Press, Praze-an-Beeble, Camborne. Holyoak, D. T. (1999). Tetrad recording of bryophytes in Cornwall. Bulletin of the British Biological Society 72: 46^-7. Holyoak, D.T. (2000). Bryophyte interest and conservation importance of former metalliferous mine sites in Cornwall. English Namre Research Reports 328: 1-50. Ligrone, R. & Duckett, J. G. (1993). A comparative ultrastructural study of endophytic basidiomycetes in the parasitic achlorophyllous hepatic Cryptothallus micrabilis Malm, and the closely-allied photosynthetic species Aneura pinguis (L.) Dum, (Metzgeriales). Canadian Journal of Botany 71: 666- 679. Ligrone, R. & Duckett, J. G. (1998). The leafy stems of Sphagnum (Bryophyta) contain highly differentiated polarized cells with axial arrays of endoplasmic microtubules. New Phytologist 140: 567- 579. Ligrone, R. Vaughn, K. C., Renzaglia, K. S., Knox, J. P. & Duckett, J. G. (2002). Diversity in the distribution of polysaccharide and glycoprotein epitopes in the cell walls of bryophytes: new evidence for the multiple evolution of water-conducting cells. New Phytologist 156: 491-508. Longton, R. E. & Schuster, R. M. (1983). Reproductive biology, in Schuster, R. M., (Ed.) New manual of bryology, volume I. pp. 386-462. Hattori Botanical Laboratory, Nichinan. Malloch, A. J. C. (1972). A note on the ecology of Gongylanthus ericetorum in the Lizard peninsula, West Cornwall. Journal of Bryology 7: 81-85. Newton, A. E., Cox, C. J., Duckett, J. G., Wheeler, J. G., Goffinet, J. A. & Mischler, B. D. (2000). Evolution of major moss lineages: phylogenetic analyses based on multiple gene sequences and morphology. Bryologist 103: 187-211. PATON, J. A. (1969). A bryophyte flora of Cornwall. Transactions of the British Bryological Society 5: 669- 755. PATON, J. A. (1976). Ditrichum comubicum , a moss new to Cornwall. Journal of Bryology 9: 171-175. Paton, J. A. (1999). The liverwort flora of the British Isles. Harley Books, Colchester. PORLEY, R. D. & MATCHAM, H. W. (2003). The status of Orthodontium gracile in Britain and Ireland. Journal of Bryology 25: 64-66. Rai, A. N., Soderback, E. & Bergman, B. (2000). Cyanobacterium - plant symbioses. New Phytologist 147: 447^181. Ratcliffe, D.A. (1968). An ecological account of Atlantic bryophytes in the British Isles. New Phytologist 67: 365^139. Read, D. J., Duckett, J. G., Francis, R., Ligrone, R. & Russell, A. J. (2000). Symbiotic fungal 176 ATLANTIC ARC associations in ‘lower’ land plants. Philosophical Transactions of the Royal Society B 355: 815-832. SCHUBLER, A., Schwarzott, D. & Walker, C. (2001). A new fungus phylum, the Glomeromycota: phylogeny and evolution. Mycological Research 105: 1413-1421. SMITH, A. J. E. (2004). The moss flora of Britain and Ireland. Cambridge University Press: Cambridge. SPALDING, A., ed. (1997). Red Data Book for Cornwall and the Isles of Scilly. Croceago Press, Praze-an- Beeble, Camborne. WHITEHOUSE, H. L. K. (1961). Tortula stanfordensis. Steere in Cornwall. Transactions of the British Bryological Society 5: 757-765. WHITEHOUSE, H. L. K. (1969). Dicranella staphylina , a new European species. Transactions of the British Bryological Society 5: 757-765. WHITEHOUSE, H. L. K. (1987). Protonema-gemmae in European mosses. Symposia Biologia Hungarica 35: 227-231. 177 Erica erigena R. Ross (Irish Heath) in the Medoc region of S.W. France C. LAHONDERE Vice-President, Societe Botanique du Centre-Ouest, 94 avenue du Parc, 17200 Royan, France ABSTRACT In France, Erica erigena R. Ross (Ericaceae) is only found in a very limited area of Medoc. in the South-West. The substratum on which this heather grows is mostly made up of detrital quaternary formations - damp or often wet all the year round. From the phytosociological point of view, it characterises the Erico scopariae- erigenae F. Bioret & C. Lahondere, a slightly acid association of the Ulici minoris - Ericion tetralicis J.-M. Gehu. Keywords: Erica erigena, Medoc, aquiferous, phytosociology, damp moors. EUROPEAN DISTRIBUTION Erica erigena R. Ross1 is a bushy heather occurring in several markedly disjunct stations from the Atlantic to the Mediterranean. As its English name suggests, it is indeed found in Ireland, where it is restricted to a handful of stations in the extreme west, in Counties Mayo and Galway (Foss 1986; Foss et al. 1987; Foss & Doyle 1988). In mainland Europe it occurs in Medoc in the south¬ west of France, in the Iberian peninsula in the Asturias and Galicia in Atlantic Spain, in central and southern Portugal, and lastly in Andalusia up to the province of Valencia in Mediterranean Spain (Bayer 1993). THE FRENCH STATIONS In France, Erica erigena is to be found only in Medoc, north of Bordeaux. In fact, the station mentioned with a ‘?* (in the Alberes, Eastern Pyrenees) by Fournier (1961) is a mistake which had. incidentally, been corrected long ago (Gautier 1898). The geographic area of E. erigena in Medoc has been the subject of two publications, one by Besanqon (1978), the other by Laporte-Cru & Richin (2002). The results of our phytosociological investigations have also been published (Lahondere & Bioret 1996). The Medoc stations of E. erigena are situated in a zone lying between the Gironde estuary and the Lake of Hourtin and Carcans. STATION 1: Situated in its eastern part on either side of a forest track 4-5 km north-west of Saint-Laurent- Medoc, west of road N215 (Saint-Laurent-Medoc to Lesparre), north-east of Larousse. This is the most important station and covers over 10 ha. STATION 2: Located in the Charite marsh by the D205 (former D4E) road connecting Pauillac to Hourtin. east of road N215. Some spots quite close to each other are situated about 250 m from Charite. STATION 3: The history of this site (since 1895) is described very well by Laporte-Cru & Richin (2002). It is a more westerly location, not far from a brook, the Berle de Lupian, which flows into the Lake of Hourtin, south of Hourtin and the hamlet of Lachanau. Synonymous with E. mediterranea auct., non L., E. hibemica (Hooker & Amott) Syme and E. camea L. subsp. occidentalis (Bentham) Lainz 178 ATLANTIC ARC STATION 4: Located at Picard, about 3*5 km south-west of Saint-Laurent-Medoc. STATION 5: Situated on the edge of road D205, about 3 km west of the D205/N215 crossroads, just before Musset. Contre (unpubl.) notes that the heather was “more abundant (in 1967) near a firebreak to the south of this road... 60 steps west of the milestone: Pauillac 12 km, Hourtin 13 km”; Laporte- Cru & Richin (2002) found only two plants, near Musset. STATION 6: We had located this station in 1983 along road N215, east of the road up to a lane going to Cambeil (west of the road). This station extended along road D104 towards Saint-Sauveur. We did not see Erica erigena there in March 1985 and put this disappearance down to the very low temperatures of the previous winter. STATION 7: Laporte-Cru & Richin (2002) cite the heather “by the edge of the Carcans-Hourtin road (D3) near the Berle de Garrouyere”, a brook flowing toward the Lake of Hourtin. According to those authors, stations 4, 5 and 7 only comprise a few isolated plants. However, Besangon (1978) estimated the combined area of all the patches of Erica erigena to be about 20 ha. EDAPHIC CONDITIONS The various writers interested in the ecology of this species mention its presence in “sandy moors” (Lloyd 1886), “damp moors on silica” (Guinochet & Vilmorin 1982), “damp moors” (Fournier 1961; Jeanjean 1961), “damp places” (Webb & Rix 1972). Bayer (1993) is more precise: “brezales y terrenos humedos, a menudo en barrancos o junto a corrientes de agua, raramente en dunas subcosteras, a veces en substratos calizos, dolonriticos o ultrabasicos” [heathlands and damp soils, often in ravines or in the vicinity of rivers, seldom on coastal dunes, occasionally on chalky, dolomitic or ultra-basic substrata]. Unlike many other members of the Ericaceae, Erica erigena is not a calcifuge: it grows on a variety of soils that are always damp, and at times wet - at least in winter. In Medoc, stations 1, 4, 5 and 6 correspond to soils called ‘F X a-b’, consisting of gravel, shingle, coarse sand and clay, deposited during the period of time that spreads over the Riss and Wurm and are alluvia of a river system. Station 2 (Plate 22) is located on “sand of the Landes district” deposited during the late Ice Age. Stations 3 and 7 are situated on sand and silt, brought by the wind. Thus, all the Medoc stations of E. erigena are found on detrital Quartemary substrata. The scientists who drew the geological map of this part of Medoc (1:50,000) explain that “west of an axis connecting Lesparre to Saint-Sauveur, quaternary soils constitute a generalized continuous aquiferous system, both vertically and horizontally”. Station 2, situated to the east of the Lesparre - Saint-Sauveur axis is still, according to those responsible for the 1:50,000 geological map, on an “axis of natural drainage” of the aquiferous system towards the Gironde estuary nearby. The substratum of E. erigena in Medoc is always a wet detrital one; indeed, “in winter, most of the area is covered with water... Even during the summer, the moor is never dried up and it is not possible to go there without having boots on” (Besan^on 1978). PHYTOSOCIOLOGY The Medoc stations of Erica erigena - as noted by all authors - look like a high moor where E. erigena is associated with species such as Erica scoparia subsp. scoparia , Ulex minor , Frangula alnus and Pseudarrhenatherum longifolium (a species of Franco-Iberian and north- Moroccan moors) and sometimes Pteridium aquilinum (probably as a consequence of drainage). This brought us to place all the Medoc stations in the Calluno - Ulicetea Class and to identify them as belonging to the Association we had previously named Ericetum scopario-erigenae (Lahondere & Bioret 1996). The presence of a few species of low neutrophile marshes ( Molinia caerulea ERICA ERIGENA IN S.W. FRANCE 179 subsp. caemlea, Schoenus nigricans, Scirpoides holoschoenus ) might lead one to think that, like the Erico scopariae-Molinietwn caemleae B. de Foucault, the Medoc Erica erigena stands can be related to Foucault’s (1984) Caricetea fuscae Class which brings together the vegetation of low- nutrient (oligotrophic) marshes and peaty meadows. Foucault (1984) writes concerning the Erico scopariae-Molinietwn that it is usually “... strongly marked by the dominance of Molinia caerulea ”, which is not the case in our stands containing E. erigena, except for station 3 (Laporte- Cru & Richin 2002) which we consider to represent a stage in the evolution of a Molinia lawn toward a moor. Foucault (1984) goes further in his description! of the Erico scopariae-Molinieninr. ''between the plants of Molinia caemlea , in the lower strata, the Ericaceae can be seen, of reduced size, but which brighten up the lifeless aspect of this mostly graminaceous vegetation thanks to their lush flowering”. This does not agree with our own observations of stands containing E. erigena ; our own view is that the E. erigena-E. scoparia stations in Medoc do not (apart from station 3) look like a 'graminaceous lawn’ - they look like a moor! Besides, let us mention that Diaz Gonzales et al. (1994) placed vegetation containing E. erigena (in the Asturias) in the Calhmo-Ulicetea Class and in the Order of the Pmnetalia spinosae Ttixen forest edges. We therefore conclude that E. erigena characterises the Erico scopariae-erigenae Association, a possibly endemic plant-community of the moors of the Medoc region of S.W. France. This Association belongs to the Ulici minoris-Ericion ciliaris (or Ulici minoris-Ericion tetralicis ) J.-M. Gehu. an Alliance embracing a wide range of communities of damp, mesotrophic Atlantic moors. PROTECTION Erica erigena, a species with national protection in France, is nonetheless threatened by drainage in an area where the cultivation of Pinus pinaster subsp. atlantica is particularly important. It must, however, be noted that at station 3 E. erigena may have benefited from "forest clearing... following the December 1999 hurricane which especially affected the Medoc Pine forests” (Laporte-Cru & Richin 2002). The same phenomenon can be noticed at station 1, where E. erigena is expanding southward and westward toward the Semigran Mill; there occur here a number of protected species with a high patrimonial value, some extremely rare, like Iris sibirica w7hich was discovered in the Larousse moor (in Toudignan) in June 1992. Station 1 is partly protected, and we hope that such protection will be extended to the whole area. We also hope that measures will be taken to protect the other stations, even though populations are mostly reduced to a few clumps. Some of these measures could be taken through inclusion of sites within the Natura 2000 network. Stations 3 and 7, at least, are located within a proposed Natura 2000 site covering the lakes and marshes of the inland sand-dunes of the Gironde coast-line. ACKNOWLEDGMENT We wish to thank Yves Peytoureau, Secretary7, S.B.C.O., for translating for this paper into English. REFERENCES Bayer, Ehr. (1993). Erica, in Flora Iberica. Vol. 4, pp. 485-506. Real Jardfn Botanico, C. S. I. C. Madrid. Bes ANCON, H. (1978). Les bruyeres rares du littoral aquitain (2nd part). Bull. Trim. Soc. My col. Bordeaux Soc. Linn. Bordeaux : 8-11. Dlaz GonzAlez, T. E.. FernAdez Prieto, J. A., Nava Fernandez. H. S. y. Fernandez C as ado. M. A. (1994). Catalogo de la flora vascular de Asturias. Itinera Geobotanica 8. Ass. Esp. de Fitosociologia. Leon. FOUCAULT, B. De. (1984). Systemique, structuralisme et synsystematique des prairies hygrophiles des plaines atlantiques francaises. These Univ. Rouen. 2 Vol. 675 p. + tableaux. FOURNIER. P. (1961). Les Quatre Flores de la France. 1. P. Lechevalier. Paris. FOSS, P. J. (1986). The distribution, phytosociology, autecology and post-glacial history of Erica erigena R. Ross in Ireland. Unpublished Ph.D. thesis, Department of Botany, University College Dublin. Foss, P. J. & Doyle, G. J. (1988). Why has Erica erigena (the Irish heather) such a markedly disjunct European distribution? Plants Today 1: 161-168. 180 ATLANTIC ARC FOSS, P. J., DOYLE & Nelson, E. C. (1987). The distribution of Erica erigena R. Ross in Ireland. Watsonia 16: 311-327. GAUTIER, G. (1898). Flore des Pyrenees-Orientales. P. Klincksieck, Paris. Guinochet, M. & Vilmorin, R. De (1982). Flore de France. Fasc. 4. C. N. R. S. Paris. JEANJEAN, A. F. (1961). Catalogue des plantes vasculaires de la Gironde. Soc. Linn., Bordeaux. Lahondere, C. & Bioret, F. (1996). Contribution a V etude de la vegetation des etangs et des zones humides du Medoc. Bull. Soc. Bot. Centre-Ouest ., N.S. 27: 475-502. Laporte-Cru, J. & Richin, D. (2002). Nouvelles observations floristiques sur la foret medocaine et plus particulierement sur les Forets Domaniales de La Teste de Buch et Lege-Cap-Ferret (Gironde). Bull. Soc. Finn. Bordeaux 30(4): 193-206. Lloyd, J. (1886). Flore de I’Ouest de la France. 4th ed. Th. Veloppe, Nantes. J. B. Bailliere. Paris. Webb, D. A. & Rix, E. M. (1972). Erica , in Tutin, T. G„ Heywood, V. H., Burges, N. A., Moore, D. M., Valentine, D. H., Walters, S. M. & Richardson, I. B. K. eds. Flora Europaea. Vol. 3, pp. 5-8. Cambridge University Press, Cambridge. 181 Euphrasia vigursii Davey (Vigurs’ Eyebright) in Devon R. HODGSON Gorselands, Axtown Lane, Yelverton, Devon PL20 6BU, England * and N. BALDOCK Dartmoor National Park, Parke, Bovey Tracey, Devon TQ13 9JQ, England ABSTRACT Euphrasia vigursii Davey (Scrophulariaceae) is a rare eyebright endemic to Cornwall and Devon. It is one of the more easily distinguished Euphrasia taxa, a quite striking plant with deep purple flowers and glandular hairs. It typically occurs in fairly short, open swards within grazed or trampled Ulex gallii - Agrostis curtisii heathland, and is found both inland and (in Cornwall) on coastal cliff-tops. We describe the current distribution of this eyebright in its few Devon localities, on the western side of Dartmoor, and emphasise the importance of site management - especially grazing and trampling - in maintaining the open conditions it requires. KEYWORDS: Ulex gallii -Agrostis curtisii heathland, endemism, distribution, conservation, population counts, grazing. INTRODUCTION Euphrasia vigursii Davey (Vigurs’ Eyebright) is a rare English endemic Euphrasia , with a world range that is restricted to Devon and Cornwall. It was discovered and formally described at the beginning of the last century (Davey 1907); it is closely allied to E. officinalis sensu Silverside (1991), and is thought to have arisen by hybridisation between Euphrasia micrantha and E. anglica (Yeo 1956; Silverside 1991). It is very much restricted to areas of Ulex gallii - Agrostis curtisii heath (N.V.C. community H4 (Rodwell 1991)) and may be found at both inland and (in Cornwall) coastal sites. On the coast it generally occurs in short species-rich turf on cliff-tops and by foot-paths and tracks; inland, it grows on lightly grazed damp heaths and open moorland. It has recently been the subject of a detailed phytosociological study (Granados 2001). Euphrasia vigursii is a relatively easy species to identify and, as commented on by Silverside (1991, 1999), may be strikingly attractive, especially so at its inland sites. It is a hemi-parasitic annual, distinguished from close relatives by its deep purple flowers, and by its stems, leaves and bracts which have a dense covering of long glandular hairs. The leaves are often suffused with purple. The flowers are quite large (corolla size usually 7-8-5 mm) and the plant is normally well- branched (Plate 27). E. vigursii is a Red Data Book species, listed as Vulnerable (Silverside 1999). In Cornwall it has been found from 1980 onwards in ten coastal tetrads, all on the north coast of W. Cornwall (v.c. 1) between Mussel Point and Cligga Head, and in 15 inland tetrads, mainly in the centre and east of the county (v.c. 2) (French et al. 1999). Unfortunately it appears to have been lost from many inland sites in Cornwall due to habitat destruction (French et al. 1999). Apart from outright habitat destruction a major cause of the decline of this species throughout its range, but especially on inland sites, has been lack of grazing (Rumsey 2002). It is lowland in distribution but two of the Dartmoor sites are at an altitude of 300 m. *e-mail: rhodgson@tinyworld.co.uk 182 ATLANTIC ARC DISTRIBUTION IN DEVON - PAST AND PRESENT In Devon Euphrasia vigursii is confined to a few heathland sites on the western edge of Dartmoor. The Flora of Devon (Martin & Fraser 1939) reported the plant from seven parishes (Okehampton, Bridestowe, Clawton, Manaton, Lydford and Buckland Monachorum). In particular it was said to be plentiful at High Down, Lydford, and was also known at that time to occur on Roborough Down. However, during survey work in the 1990s for the Atlas of the British and Irish Flora we located just two populations, at High Down and Vale Down, Lydford. It was not refound at the old Roborough Down site. Extensive searches have been made more recently of suitable habitats on both sides of Dartmoor, including during a B.S.B.I. field meeting, and these have located two additional small populations: one at a second (new) locality on Roborough Down, and another (about 30 plants) at Femworthy Down, Lydford. A possible new site at Wheal Jewell, Willsworthy, was reported in 2002 but this record has still to be confirmed1. High Down and Vale Down, Lydfiord, are managed under Dartmoor Environmentally Sensitive Area (E.S.A.) agreements, their management (including stocking rates) designed with conservation of the Ulex - Agrostis heathland as one priority. THE IMPORTANCE OF GRAZING Of the four (possibly five) known sites for Euphrasia vigursii in Devon, those at Roborough Down and Femworthy Down have only very small populations. The two most important sites, however, have both been affected by recent changes in the levels of stock grazing. Vale Down, Lydford, had an estimated population of about 100 plants in 1996, but unfortunately the site was not grazed at all in 2001, during the Foot and Mouth crisis. No E. vigursii could be found there when one of us (R. H.) visited the site in June 2002, but later in the year - once grazing had resumed - we did find a single plant. This demonstrates how rapidly a population can be affected by lack of grazing. Areas of Ulex at Vale Down have now been cut, and stock grazing reinstated, and we will be monitoring the eyebright population as it (hopefully) recovers2. By far the best site is at High Down, Lydford. Exceptionally heavy grazing by sheep in 2001 (due to stock being ‘trapped’ on the site during the Foot and Mouth crisis) was followed by a count of more than 20,000 E. vigursii plants in 2002 - possibly the largest extant population of this species in the world. At that time, the site had the look of a heavily grazed H4 Ulex gallii - Agrostis curtisii heath with extensive areas of short grass sward; associated species included Euphrasia anglica, Galium saxatile, Erica tetralix, Thymus polytrichus and Potentilla erecta. We suspected that the abundance of E. vigursii would be affected to some extent by spring/summer rainfall, but by far the most important factor seemed to us to be the intensity of grazing: we considered that over- or under-grazing could have a marked effect on the species, but grazing levels at High Down in 2002 seemed to be about right for E. vigursii. The importance of grazing has been highlighted by subsequent monitoring of the Euphrasia vigursii population at High Down: in 2003 there were about 8,000 plants, while in early summer 2004 only 97 plants could be found - a population crash of 95% in just two years, and directly correlated with a reduction in grazing levels that had been instigated (despite the inclusion of E. vigursii in the local Biodiversity Action Plan) in an attempt to restore over-grazed and ‘degraded' heathland to ‘favourable condition’; over the same time period, the cover of Ulex gallii increased from c. 20% to c. 60%. 3 Discussions are now taking place with Dartmoor National Park Authority, English Nature, Defra and the Lydford Common Association to work out a compromise solution whereby Ulex-Agrostis heath (lacking E. vigursii) on the adjoining S.S.S.I. is left ungrazed or lightly grazed, while areas holding the bulk of the eyebright population can be grazed more heavily once again. 1 This site was searched, without success, in 2003; further investigations are planned for 2005. " Unfortunately, no E. vigursii was found at Vale Down in 2003, though E. micrantha, E. nemorosa and E anglica were all present in good numbers. Fortunately there was some recovery of the E. vigursii population later in 2004, though numbers were clearly still well down on previous years. EUPHRASIA VIGURSII 183 REFERENCES DAVEY, F. H. (1907). Euphrasia vigursii sp. n. Journal of Botany 45: 17-20, Plate 486. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. GRANADOS, L. (2001). A fine scale phytosociological study involving selected environmental and floristic parameters between three populations of Euphrasia vigursii Davey (Cornish eyebright), a rare annual endemic to Devon and Cornwall. Unpubl. Honours Project, University of Plymouth. Martin, Rev. W. Keble & Fraser, G. T. (1939). Flora of Devon. T. Buncle & Co., Arbroath. RODWELL, J. S. ed. (1991). British Plant Communities Volume 2: Mires and heaths. Cambridge University Press, Cambridge. Rumsey, F. J. (2002). Euphrasia vigursii , in PRESTON, C. D„ PEARMAN, C. D. & DINES, T. D. eds. New Atlas of the British and Irish Flora , p. 564. Oxford University Press, Oxford. SlLVERSIDE, A. J. (1991). The identity of Euphrasia officinalis L. and its nomenclatural implications. Watsonia 18: 343-350. SlLVERSIDE, A. J. (1999). Euphrasia vigursii Davey, in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants, pp. 155-156. Joint Nature Conservation Committee, Peterborough. Yeo, P. F. (1956). Hybridisation between diploid and tetraploid species of Euphrasia. Watsonia 3: 253-269. 185 The integration of wildlife information in Cornwall S. MYLES ERCCIS Manager, Environmental Records Centre for Cornwall and the Isles ofScilly, Five Acres, Allet, Truro, Cornwall, TR4 9DJ, England ABSTRACT This paper looks at some of the work being undertaken at the Environmental Records Centre for Cornwall and the Isles of Scilly (E.R.C.C.I.S.) regarding the integration of wildlife information for use by recorders. Its aim is to demonstrate how information can be integrated using computer technology to present clear and understandable spatial representations of the wildlife in areas of Cornwall and the Isles of Scilly. Keywords: Biological recording, Local Record Centre, environmental information, computerisation. WHAT IS WILDLIFE INFORMATION? For the purposes of this paper wildlife information refers to: • species - flora and fauna; terrestrial and marine; • habitats - recorded using various classification systems; • sites - both designated and non-designated sites. Other resources are also useful when interpreting wildlife information; these include geology and soil maps; and aerial photos. WHY INTEGRATE WILDLIFE INFORMATION? Information is only of value if it is used, and preferably used many times. This may seem obvious but all too often species records are used once and never see the light of day again! Information collected, such as records of Pilularia globulifera (Pillwort) as part of a survey of nationally scarce plants in Cornwall, whilst being useful for the specific purpose for which they were collected, may also be of interest to someone looking at species colonising shallow pools and wet muddy areas in Cornwall. Integration of wildlife information therefore seeks to bring together data from a variety of sources for many different uses. These uses include the planning of survey effort, by mapping and listing existing species records, habitats in which the species is known to occur, and sites designated for that species or assemblages of similar species. Reports from previous surveys may also provide landowner information as well as site descriptions and other details. The aim of the Environmental Records Centre for Cornwall and the Isles of Scilly (ERCCIS) is to collate, manage and disseminate wildlife and earth science information relating to Cornwall and the Isles of Scilly. The integrating of wildlife information is therefore essential to the efficient working of the Centre. INFORMATION HELD AND MANAGED BY ERCCIS Species information is held within an Access database, whilst habitat and designated site boundaries and information are held in a Geographical Information System (GIS) - a computerised mapping system. The species records, although managed in a database, can easily be brought into the GIS for mapping purposes. As an example Figure 1 shows the distribution of Pilularia globulifera for Cornwall based on the records managed by ERCCIS. 186 ATLANTIC ARC FIGURE 1. Distribution of Pilularia globulifera in Cornwall » »; xu*. U'> ’ i v, : •• - •%*,: : ‘ X- * * S* V » V // • • <■ v v y t;>yf <■ x- vy>xj *• S- Y :• f . v .. .. . SU rv sv ■* * * * V4^~v •• '**>&.*•& ■ l }U . ;l!'; ::u tv \V -fw/ y ■•■:£• X-.v-v.v £ x ^, < .... -V . S. . ■:■ ■: ■ ■ :■:■:■•■. ■ ;■ r ViT J *• g? f > l* * 1 1 Hs* $,i •££*: Ct?'’''’*'* •■. .■ .>. .■ •• .• .* .• a-.\*s , •: •••••• •••'■ * •'••••• ••• < ••• •••••••• . y> . ■ - ' •: ■ O'*.:*-/ ks-^1 %%■ xv'/s-x-ii 4 r :: <••••. •••••• •: . T‘: Xj^x-rv'-x x ;*»«?« •; :' ',■ ■ . v • . •• •r W> f ’ . I C | Q I V' C!s*sSi ilSfsaRas V PBasaria gisfcsISerii a ^ 8 JE iE«:ufe*S V *E« y' 1 &-4-» (top Figure 2. Distribution of species associated with the Lizard trackways in the Kynance Cove area. Lizard WILDLIFE INFORMATION IN CORNWALL 187 (tint JIwkm! p&yi-M, ■ 1b nisi jpJ c ■>* >aXUv * 0 V R ail JfcZMk&i t'^ Jilts* # £&•» ** *.i pi i » * ?fc rr ^ 1 >1 ><,. jS ;\ . y .a y. \ 1 Cn^rdka 8^-^re: 44 ^ S'aifw Fjir^ M. I * _ ✓ sr - V* ^ , • V , s / r*, ~ Mm K -.. 4 %\v :- • ••: ;. ' - - *- - i .Juncas •16 4 iuf {C v& • 1 r | € iuiJC^S &3it SI | .-L - l€ ‘s •': ;%> i ♦ . L1 rr K<* H « *?t»r; Figure 3. Information relating to one point/symbol on the map. •. I *£*§.• , ;W* ife * r.1 1 •. s*.u. i<»Sf y' 1 ■Sjljljl.i H jrt y P ■■ ■ u i a y *- **•»!« y .ijr.rii ? h% T< w vij □ y Ck4r-i i nisg.Tsa •& L' r..4ii lj'js.ii ti r> y [ Wt1Ul^£lS!B) ..n... , ^ ,^r ,x’ M V I *?«x ■ 'j p -i» f>" \^, *"■" , U- -'M> i _ ri' ■ , - - -i •„ mmmfsMMii v- r^L.' - : L ; . 1 ' ^'.vl V ' .* “a.* r,d ^ : .;. ■ ,. ,. lt»t 1 ’ & StiSZik Figure 4. The semi-natural habitats of Kynance Cove area, along with the species records 188 ATLANTIC ARC FIGURE 5. Designated site, semi-natural habitat and species information displayed for the Kynance Cove area. The records shown in Figure 1 have been mapped at the scale at which they were ‘collected’; for example records recorded at a 10 x 10 kilometre grid square resolution have been shown as a 10 x 10 kilometre grid square. This reduces confusion in interpretation as it provides a spatial picture of exactly where, and at what scale, the species was recorded. It is important to take note of what scale/resolution the record was made at and also why it was originally collected, when making use of records. Using the GIS one can then focus in on a particularly geographical area and look more closely at the spatial distribution of Pilularia globulifera. Figure 2 shows the distribution of this species in the Kynance Cove area, on the Lizard peninsula. It also shows records of Cicendia filiformis, Yellow Centaury (Nationally Scarce), J uncus pygmaeus. Pigmy Rush (a Red Data Book species), Ranunculus tripartitus, Three-lobed Crowfoot (Nationally Scarce), and Chara fragifera. Strawberry Stonewort (a Red Data Book species). All of these species are associated with the trackways across The Lizard, inhabiting damp areas and shallow, muddy pools. For Figure 2 the records are shown as symbols, plotted as points in the centre of the 100 metre grid square in which the species was recorded. Chara fragifera is the exception as the only record for this species is at the 10 x 10 kilometre grid square resolution and part of the 10 x 10 kilometre square is shown on the right hand side of Figure 2. On the GIS, one particular point/symbol can be selected; more details concerning the species recorded at that point are then displayed, as shown in Figure 3. In the case of the example shown, three of the species have been recorded at the same six figure grid reference - Cicendia filiformis, Juncus pygmaeus and Ranunculus tripartitus. One record of Juncus pygmaeus has been selected and basic details (year recorded and grid reference) have been displayed. GIS can therefore be used to display distributions of more than one species at once. This is particularly useful when looking at assemblages of species, or species and their food plants or host species. It can therefore help focus fieldwork - targeting areas where associated species have been previously recorded. WILDLIFE INFORMATION IN CORNWALL 189 At the E.R.C.C.I.S. it is not just species data that can be displayed and manipulated on the GIS; there is also habitat information, boundaries of designated wildlife and Earth Science sites; some land ownership boundaries and land management details. Figure 4 shows the main semi-natural habitats for the Kynance Cove area - wetland, heathland and wetland/heathland mosaic, along with the species records. The habitat information was derived for the whole of Cornwall using aerial photographs and field surveys, and relates to 1995 (there is also a dataset for 1988). This integrated information does not only help with refinding species at historical sites but also targeting new areas with similar habitats and species assemblages. Figure 5 takes this one stage further, with designated site information also displayed for the Kynance Cove area. This area has been designated as a Special Area of Conservation (SAC), and is of international importance. It is worth remembering that the information displayed on the GIS (and any other computer database/spreadsheet/map) is only as precise and accurate as the original information collected and entered into the computer. Correct management of all information, including validation and verification, is vital if the information is to be believed and useful in the long term. This integrated approach was used by ERCCIS to provide information to Dr Belinda Wheeler of Plantlife for her survey of the rare and scarce plants of the Lizard trackways (Byfield & Wheeler 2005). Ian Bennallick, joint BSBI Vice-County Recorder for East Cornwall, also used a combination of recent habitat survey information and historical records to refind Romulea columnae, in the Polruan area. This species was refound in 2002, following an interlude of over 120 years - previously recorded in 1879 and 1881 (Bennallick 2005). AC KNO WLEDGMENTS The Environmental Records Centre for Cornwall and the Isles of Scilly is indebted to the hundreds of individual recorders who support us and send us the species records and other information they collect in their own time. Without their support we could not exist. REFERENCES Byfield, A. J. & Wheeler, B. R. (2005). The Lizard Trackways Project, in Leach, S. J., Page, C. N., PEYTOREAU, Y. & SANFORD, M. N. eds. Botanical Links in the Atlantic Arc, pp. 127-131. Botanical Society of the British Isles, London. Bennallick. I. J. (2005). Romulea columnae Sebast. & Mauri (Sand Crocus) refound in Cornwall after 121 years and Juncus capitatus Weigel (Dwarf Rush) new to E. Cornwall (v.c. 2), in Leach, S. J., Page, C. N., PEYTOREAU, Y. & SANFORD. M. N. eds. Botanical Links in the Atlantic Arc, pp. 247-249. Botanical Society of the British Isles, London. 191 Cystopteris diaphana (Bory) Blasdell (C. viridula (Desv.) Desv.) new to Britain R. J. MURPHY Shangri-la, Reskadinnick, Camborne, Cornwall TR14 OBH, England and F. J. RUMSEY Department of Botany, Natural History Museum, Cromwell Road, London SW7 5BD, England ABSTRACT The discovery of Cystopteris diaphana in Cornwall, new to Britain, is described. A summary of its habitat is given, along with notes on its field identification, taxonomy and wider European distribution. The question of whether this species is native or introduced in Britain is considered. Keywords: Brittle Bladder-fern, distribution, native or alien status. THE INITIAL DISCOVERY On 10 February 2000 a few plants, at first identified as Cystopteris fragilis (L.) Bemh. (Brittle Bladder-fern), were found by M. J. Stribley in rocky recesses in a woodland bank by the River Camel at Polbrock Bridge, near Wadebridge, E. Cornwall (v.c. 2) at SX013695. Given the extreme rarity of C. fragilis in Cornwall (French et al. 1999), further visits were made to the site by Mary and Tony Atkinson, Ian Bennallick, Matt Stribley and one of the authors of this note (R.J.M.). As a result, thousands of plants were observed, this time growing on the steep, almost vertical, shaded banks that edged the River Camel. Later, however, doubts were expressed about the correctness of the original identification. In Flora Nordica (Jonsell 2000), the drawings of C. fragilis and C. alpina showed differences in the vein endings between the two. The material from Polbrock could not clearly be assigned to one or the other, some veins ending in tooth-apices like C. fragilis while others ended in U-shaped depressions as in C. alpina. The spores of the Polbrock plants also seemed different, being echinate but with denser spines. A telephone call to the Natural History Museum produced the answer. Christopher Fraser-Jenkins recognised the unknown plant from the description given, a specimen was sent to him and he was able to confirm that the fern was Cystopteris diaphana - the first record of this species not only for Cornwall but for Britain. It was interesting to observe that the habitat at Polbrock was not typical for C. fragilis. It was very shaded, very humid and not all that base-rich, the geology being Staddon Grit which contains only a little thin limestone. Indeed, the habitat was rather reminiscent of that given for C. diaphana in Schafer (2000), “...restricted to the most humid habitats”, and in Tutin et al. (1993), where the plant is described by A. C. Jermy as growing “on shady, mildly-basic rocks”. SUBSEQUENT RESEARCH There still remained the question as to whether the plants along the Camel were all C. diaphana , or whether C. fragilis was also present. More material was collected, all of which was determined by F.J.R. as C. diaphana , and to date we have been unable to find any C. fragilis at Polbrock. There 192 ATLANTIC ARC Figure 1. Cystopteris in Cornwall. O Pre 1987 - former records then listed as Cystopteris fragilis, but no evidence that these we correct. • 1987+ records that may be for garden escapes - both C. fragilis and C. diaphana present (BM). ■ 1987+ C. diaphana in a natural habitat (BM). are, however, old records of C. fragilis in the vicinity of Polbrock and from other areas nearby (Fig. 1), but there are, unfortunately, no herbarium specimens and so we cannot confirm their true identity. It is possible that these old records were C. diaphana rather than C. fragilis. Previous to this, C. diaphana had been recorded in Europe only from Corsica, France (where it is rare), Italy, Portugal, Sicily, Spain and the Azores. In Portugal and the Atlantic areas of Spain it grows in rocky crevices and shady places, and the description of the habitat given in Claves de Flora Iberica , Vol. 1 (2001) - “en las proximidades de cauces de agua, en ambientes oceanicos” - is remarkably similar to the habitat at Polbrock, where the fern extends down river as far as the high tide mark. C. diaphana is clearly growing along the Camel very much ‘in the right place’, but we cannot be absolutely certain whether it is native there or has been accidentally introduced. In this respect, however, it has proved of much interest that C. diaphana has been recently confirmed by Christopher Fraser-Jenkins as being present along a sheltered lane near Mawnan Smith (v.c. 1) in SW73 (Fig. 1), where the lane runs between the two halves of one of Cornwall’s old garden estates, Penjerrick. Old records for C. fragilis have also been confirmed there, but given the proximity of the Penjerrick estate it is possible that both taxa at that site are garden escapes. Bryophytes and flowering plants associated with C. diaphana at Polbrock have now been listed by D. T. Holyoak, but further work is still needed and it is hoped that one of these lines of investigation will involve chromosome counts. CYSTOPTERIS DIAPHANA NEW TO BRITAIN 193 REFERENCES JONSELL, B. ed. (2000). Flora Nordica. Stockholm: Bergius Foundation. MURPHY. R. J. & Rumsey, F. J. (2005). Cystopteris diaphana (Bory) Blasdell (Woodsiaceae) - an overlooked native new to the British Isles? Watsonia 25: 255-263. SCHAFER, H. (2002). Flora of the Azores. Margraf Verlag, Weikersheim, 1: 62-67 . Tutin, T. G„ Burges, N. A., Chater, A. O., Edmondson, J. R„ Heywood, V. H., Moore, D. M., Valentine, D. H„ Walters, S. M. & Webb, D. A. eds (1993). Flora Europaea , 1, 2nd ed. Psilotaceae to Platanaceae. Cambridge University Press, Cambridge. 195 Database developments and biological recording in Cornwall C. N. FRENCH 12 Seton Gardens, Weeth Road, Camborne, Cornwall, TR14 7JS, England ABSTRACT In recent years the biological recording system in Cornwall has become much more focused on supporting the volunteer recording community. In particular, technological improvements (both hardware and the continued development of the ERICA database) have meant that individual recorders can each have all the available data for Cornwall on their own PC computer at home, utilise those records to plan further recording, add to those data, and hence contribute to the collective effort of the biological recording community of Cornwall, and ultimately further afield. This level of individual empowerment has stimulated recording activity, and this process has been greatly assisted by the combined efforts of the Cornwall and Isles of Scilly Federation for Biological Recorders (C.I.S.F.B.R.), and the Environmental Records Centre for Cornwall and the Isles of Scilly (E.R.C.C.I.S.), and other local groups, who provide training workshops, field meetings and seminars and publish a range of helpful material. Keywords: Local Record Centre, environmental information, computerisation. INTRODUCTION In many respects, the biological recording system that has developed in Britain over the last two centuries must be the envy of many other countries. This system is built on the endeavours of many thousands of volunteers who routinely keep records of plants and animals and then pass them onto Local Record Centres, Natural History Groups or organisations that specialise in particular taxonomic groups like B.S.B.I. or Butterfly Conservation. This system has worked extremely well but, for very good reasons, has had two particular failings. 1 . The data flow has been unidirectional from the recorder to the Local Records Centre and on to the B.R.C. at Monks Wood. Data would only travel in the other direction in exceptional circumstances. 2. Although volunteers are the lifeblood of the whole system, they have received very little in return for all their efforts and have had to content themselves with the knowledge that they are contributing to a good cause. The object of this paper is to describe the recent evolution of biological recording in Cornwall, in particular database developments, and to explain how this has addressed the above two failings. BIOLOGICAL RECORDING IN CORNWALL In 1972 the Cornish Biological Records Unit (C.B.R.U.) was formed by the University of Exeter. Manned by volunteers, it began to gather together all the information it could collect about all species that live in Cornwall and offshore. By 1987, through the diligent work of those volunteers, the C.B.R.U. had amassed such a gargantuan databank on card indexes and in filing cabinets, that it became imperative to computerise. The ERICA Project was bom with the remit that all the records had to go on computer without losing any information. A mini-computer at Exeter held the database and this was fed by a network of PCs at the Records Centre in Camborne, Cornwall. For the next six years, six people typed the records into the database, whilst the remainder of the staff continued the normal functions of the Records Centre. It was soon realised that the C.B.R.U. had the technological capability to process additional data to that in the paper archives. This prompted Rosaline Murphy and the author to propose a systematic survey of the vascular plants of Cornwall in order to produce an atlas of the Cornish flora. With the assistance of another small band of volunteers, this project team began surveying every tetrad (2 km square) at different seasons and covering as many habitats as possible. 196 ATLANTIC ARC In the spring of 1996, the CBRU celebrated the computerisation of the millionth biological record - the target that was unofficially set at the start of the ERICA Project. Six months later there were funding problems and the CBRU closed down. The paper archives were transferred to the Cornwall Wildlife Trust and the ERICA database was retained by the University of Exeter. Suddenly, the biological recording community was left without the support of a Records Centre and the ERICA database had become fossilised. The surveying work for the proposed Flora of Cornwall was only part complete and was on the point of abandonment, because the project was effectively left without access to the data already computerised and there was way of adding to those data. The immediate response of the biological recording community was to form the Cornwall and Isles of Scilly Federation for Biological Recorders (CISFBR), and the Cornwall Wildlife Trust set up the Environmental Records Centre for Cornwall and the Isles of Scilly (ERCCIS), which would eventually become a Records Centre. Both these organisations have worked in a complementary way to support the biological recording community in Cornwall. They organise seminars, field meetings, taxonomic workshops and produce a variety of publications, such as the Handbook for Biological Recorders (Atkinson et al. 2000) and the RDB book for Cornwall and the Isles of Scilly (Spalding 1997). However, neither could fill the gap left by the loss of the ERICA Project and the consequent inability to continue processing biological records on computer. To re-float the ERICA project the entire database (then 1,068,000 biological records) was downloaded from the mini-computer in Exeter into text files on a PC. Given the state of PC technology at the time, this took several weeks to accomplish. Unfortunately, it was then discovered that the PC version of the database management system (PRIME INFORMATION PC) had a limit of 420,000 records, and so the ERICA software had to be rewritten to fool the database management system into accepting more than a million records. This accomplished, the database was rebuilt, a process which took a 486 computer three weeks to complete working day and night, leading to the birth of NEW ERICA (Fig. 1). NEW ERICA DATABASE MAIN MENU 1 - Add Some Biological Records 2 - Add Collectors Names 3 - Add References 4 - Add Family Details 5 - Add Species Details 6 - Add Site Details 7 - Printed Reports 8 - Add Parish Details F3 - Edit a Record F4 - Check Records Entered F5 - Other Databases F6 - Edit Consecutive Records F7 - Import/Export 99 - Exit program Figure 1 . The main screen of the NEW ERICA database. The outcome was that by the end of the year biological recording could resume again, this time working from home, and the Flora of Cornwall project could be revived. The result was the creation of “a distributed Records Centre”, whereby each of the volunteers had their own copy of the entire database on their home computer. They then added their own records to their own machines, paper copies went to the Vice-county recorder for validation and periodically the records were sent to the author to be added to the master copy of the NEW ERICA database. The updated whole was then returned to the individual volunteer recorders. Everyone in the scheme, BIOLOGICAL RECORDING IN CORNWALL 197 which included not just vascular plant specialists, had all the data available to them and everyone benefited from the collective effort of all the recorders. Thus in 1997 the recording effort was able to continue with vigour once again, and the atlas of the Cornish flora could be completed. This was published as a book and a CD-ROM (Trench et al. 1999). FLORA OF CORNWALL CD-ROM The CD-ROM (Fig. 2) was designed to take the concept of an Atlas one stage further, and to overcome some of the limitations of publishing in paper format, such as restrictions on the number of maps and photographs, and the cost of colour photographs. The CD-ROM format makes it possible for all the raw data to be made available and be interactively analysed in an accessible, flexible and pleasing manner. Furthermore, CD-ROMs can be periodically updated: there is no need to wait 20 years for the next publication of an atlas in book form. Figure. 2. The CD-ROM Front Screen. To animate this plain screen photographs are randomly displayed every few seconds. The top six option buttons relate to the entire text of the book. The Gazetteer, for instance, is used to generate species lists for any 1 km square or tetrad. The Flora of Cornwall CD-ROM includes all the text from the book plus all the plant records used in the compilation of the book. In addition there are over 3000 photographs. As much information as possible concerning a species was crammed onto this one screen (Fig. 3). A species can be chosen from the dropdown list or by typing in any of its common names, scientific name, or synonyms. The options on the right enable changes to be made to the map background, the sampling scale of the dots and the date classes. The table below' the map show's the selected records and if any of those records are chosen with the mouse, more details are shown 198 ATLANTIC ARC Cirhoriuin intybus L. Chicory Q Choose the hist letter of r the species name i Cichorium in Cicuti virosa Circm luietiana Cixeae* hi tettana forma eordifclia Cnsium uavk Cirjium arvense Cirsium ajvense var setosum Cinivnt iusKtum Cixsiam triop ho rum Cinirn tritfktntm 19 kritajuiien Cinram yaluatre CinntJK vaJgaxe Cireimm ralgart m k^wkwum Print Help Back Chicory has become a rather rare sight and is of sporadic occurrence along road verges and occasionally on hedges. Number of records = 1 43 Number of T etrads = 1 04 % of total tetrads = 1 0.1 % Grid ref. Location | Date Full Name Conf| Coc -*■ j SW75Q LAMB0URNE 1902 F. RILSTONE P923S j SX25R MILLENDREATH 1902 SCHWARTZ P9237 9X3RT .±LJ 9FATFIN RRIDGF 1 90? 9 OH WAR T7 F'9?37,Zj ► j j SW7S51 LA.MBOURNE F. RILSTONE 1902 DAVEY,F.H. 1 902. A tentative list of the flowering plants, ferns, etc., known to occur in the country of Cornwall, including Isles f-* Cornwall f" Roads P CwaH and ScSy F7 Rivers f~ Isles of Scifly P SSSIs 1“ Geology map P Grid p” Dots in centre of grid r Dots by grid reference P Species names f~ Common names P Tetrad P 10 km r 1 km P 5 km P Exclude unconfirmed C AH records C Before a set year C After a set year C From one year (* Before and after a yea C From a range of years 1 Show previous records Enter Year Figure 3. The Atlas Screen beneath, and its location is highlighted on the map. Conversely, if any dot on the map is clicked on with the mouse the corresponding record is shown in the table. Clicking on the photograph will enlarge it to maximum size. Pressing the list button will display a species list for the 1 km square where the record was made. The Crosstabs button displays a series of excel-type pivot tables and the Vector Maps button displays the records on a scaleable vector map. ERICA FOR WINDOWS The Flora of Cornwall CD-ROM is not a working database that everyone can contribute towards. This function is still performed by the NEW ERICA database. However, the NEW ERICA database is built on technology that is 20 years old, is purely text-based, and cannot be developed to meet future needs. So, a Windows version of ERICA has now been written (Figs 4 & 5) to incorporate all the information held within NEW ERICA (1,350,000 biological records covering 22,000 species), to improve on the Flora of Cornwall CD-ROM technology and to embody all the good bits of functionality built into NEW ERICA. Above all, the Windows version of ERICA is designed to keep to the principle of ensuring that all the contributors to the database are true beneficiaries, by each having their own copy of the whole database, which periodically gets updated. Thus everyone can benefit from the collective effort of all the contributors to the database, whether it be records, photographs, drawings, sound recordings or video footage. CONCLUSION Biological recording in Cornwall has been subjected to considerable turmoil as a result of the closure of the C.B.R.U. in 1996. From this calamity has arisen a devolved system of recording, which has focused on embracing the volunteer recording community, making them fully integrated BIOLOGICAL RECORDING IN CORNWALL 199 it. ERICA for Windows Fie E<8 Add Copy Fdrlable Refresh Photographs Help About Options Gazetteer J Taxonomy f GIS Map T Taxonomic Group J Coincidence Maps IChoose species T Whole Record T Analysis A Adiantum capillus-venens, Maidenhair Fem Asplenium adiantum-nigrum, Black Spleenwort Asplenium bulbiferum, Mother Spleenwort I — Asplenium cuneifolium, Serpentine Black Spleenwort Asplenium mannum, Sea Spleenwort Asplenium mannum var acutum, | — Asplenium obovatum ssp lanceolatum, Lanceolate Splee Asplenium onopteris, Irish Spleenwort Asplenium ruta-muraria, Wall-rue I — Asplenium septentrionale, Forked Spleenwort Asplenium trichomanes ssp quadrivalens, Maidenhair Sp Asplenium trichomanes ssp trichomanes, Maidenhair Sp . — Athyrium filix-femina, Lady-fern Athyrium filix-femina var rhaeticum, Azolla filiculoides, Water Fem i £ Enter species name, common name. etc. A MatkmaLy Scarce fem. Lssedafe Spieerssort k fetmd tr: diff-top rock crevices, on hedges srd vtMs dose to the sea. as weS as on ok! Cornish Hedges j=-JL2£l l~ Comwal I- Comwal and Sc*y r Isles of Sc^r I- Geology map R Roads I~ Rivets r sssis r Parish l~~ Grid Vector Maps Comw^ Rivets Comwal and Scsty Districts Isles of Scily _ R Tetrad I- 1 km I 5 km f- 10 km Choose new species: riand Along these hedgebanks it is most easily formd ear^r in the spring before more -rigorous secies have obscured it Although Black Spteersvort mat be foaid r> the game habitats & k very rare to Ida both speeds growing side by side. Good populations occss in West Penwith the lass around Zennot [SW433] and on the hsdgebarks near Lar^ron Quest [SWO-lj. Other good sties are on Tint age! Island iSXOSF-: and down a country lane near Opton Doss sSX27Wj. St Tetradl Easting! Northing 1 Location 1 Date Name Notes Number Code Cj^ SVW3D: 191 37 NARE HEAD 24/JUL/2002 MR I.J. BENNALLICK B52604 — SW43U; 1474 396 W1CCA TO TREVEAL 4/JUN/2002 DR C.N. FRENCH XI 02338 SW44Q 1477 410 CARN NAUN POINT 4/JUN/2002 DR C.N. FRENCH XI 02592 SW33R BOTALLACK 1 1/JUN/2002 M. ATKINSON SITE SURVEY. TRACK TO SW. STARTS IN D1 34729 SXD6Q 20648 6003 SOUTH OF CRIFT 18/AUG/2002 MR I.J. BENNALLICK ON WEST FACING SIDE OF ROAD ON B62243 SVW3J 192 38 THE BLOUTH 9/MAY/2002 MR I.J. BENNALLICK B62020 SW32Y! 139051 26237 ROSPANNEL FARM 10/AUG/2002 R.J. MURPHY BOTANICAL CORNWALL MEETING. OLD XI 04260 u- jU Figure 4. The main screen of ERICA for WINDOWS. Add Biological records 'i'J'? Srnpte ist List and ramber Frequency Aldetak = . .u: ; Add New person AddNewScwce Exit Rrstpage [ _ Second page Enter species name: jlsoetes histrix Enter fn/t/s/s of person: R.J. MURPHY MR I.J. BENNALLICK Year of Enter Authors initials publication Third page Last page Stenatne Grid reference Location SW6851 45 [kynance Day Worth Year Tree of Dap Tune period 3 l3 |2003 | jy Notes GROWING AT BASE OF CORNISH HEDGE| Figure 5. One of the data entry screens showing an incomplete record being entered. 200 ATLANTIC ARC beneficiaries of the biological recording activity in Cornwall. Secondly, through technological improvements this system has overcome the need for the unidirectional flow of data away from the recorder that has characterised biological recording. Instead everyone can have all the available data on their own machine. This has stimulated additional recording effort and has been of real benefit to the individual volunteer recorder. Finally, Cornwall has seen a renaissance in biological recording in the last couple of years thanks to the combined efforts of C.I.S.F.B.R., E.R.C.C.I.S. and the newly formed Botanical Cornwall Group, which meets on a monthly basis and has further stimulated focused recording activity. Combined with the database developments, the future is looking promising for biological recording in Cornwall. REFERENCES Atkinson, M., Bennallick, I., Holyoak, D., Lord, D., & Mccartney, P. (2000). A Handbook for Biological Recorders. A manual for recording plants, animals and their habitats in Cornwall and the Isles ofScilly. CISFBR & ERCCIS, Cornwall. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. SPALDING, A. ed. (1997). Red Data Book for Cornwall and the Isles ofScilly. Croceago Press. 201 Assessment of threats to populations of Rumex rupestris Le Gall (Shore Dock) in Britain and France F. BIORET Geomer UMR 6554, IUEM et Institut de Geoarchitecture, Universite de Bretagne Occidental, CS 93837, 29238 Brest Cedex 3, France *' and R. DANIELS CEH Dorset, Winfrith Technology Centre, Dorchester, Dorset DT2 8ZD, England* ABSTRACT Rumex rupestris Le Gall (Shore Dock) has been the subject of an Anglo-French co-operative study carried out under the Alliance research programme “A conser\'ation strategy for threatened botanical resources of the English Channel-Atlantic coastal environment" . We describe here the geographical distribution and ecology of the species, and assess the level of threat to a number of R. rupestris populations visited in 1998-1999 in France and Britain. We also report on studies of population sizes, and discuss problems associated with the definition of minimum viable population size and ‘favourable conservation status’. It is possible that groups of apparently discrete populations may be inter-connected - acting, in effect, as single metapopulations - and more work on this, and on various other aspects of the biology of R. rupestris , is proposed. KEYWORDS: conservation, distribution. INTRODUCTION BACKGROUND TO THE PROJECT The distribution of coastal zone species is tightly constrained by ecological conditions. Ecological diversity of coastal terrestrial environments has to be considered as the result of the highly dynamic nature of diverse geomorphological processes. Conservation of coastal plants requires us to take into consideration the critical interaction between natural and human-induced processes. Focused on one threatened plant species, Rumex rupestris Le Gall (Shore Dock), this work was carried out under the Alliance research programme “A consen’ation strategy for threatened botanical resources of the English Channel-Atlantic coastal environment ”, and involved cooperation between C.E.H. (Britain) and Geosystemes C.N.R.S. (Brittany), from 1998 to 1999 (Daniels & Bioret 1999). A key objective of the work reported in this paper was to visit a range of sites in Britain and France to establish the nature and degree of threat to R. rupestris populations in the two countries; we have also considered what further investigations might be needed to underpin the conservation of this species. * 1 e-mail : frederic. bioret @ univ-brest.fr *2e-mail: rogerdaniels 1 @ aol.com 202 ATLANTIC ARC TABLE 1. HERITAGE VALUE OF RUMEX RUPESTRIS IN BRITAIN AND FRANCE International National threat status conservation status threat status conservation status France Britain France Britain Rumex rupestris Vulnerable1 Habitats Directive (annex II) Red Data Book Vulnerable Red Data Book Endangered protected protected 1 Council of Europe: List of rare, threatened and endemic plants in Europe (1982) DISTRIBUTION AND STATUS OF RUMEX RUPESTRIS Rumex rupestris is a European endemic littoral eu-atlantic species, only recorded in the following countries, with the core of its range located along the shores of Brittany, from the south of Finistere to the Loire estuary (Fig. 1): • Wales - Anglesey, Pembrokeshire and Glamorgan (Kay 1996; Daniels et al. 1998; King et al. 1999). • England - Devon and Cornwall (formerly also Dorset) (Daniels et al. 1998; King et al. 1999). • Channel Islands (McClintock 1975). • France - Cotentin (Provost 1993), Brittany, Vendee (des Abbayes et al. 1971; Philippon 1991), and north Medoc in Gironde (just a single locality (Dussausois 1996)). • Spain - western Galicia (Dupont 1962; a few localities, S. Ortiz, pers. comm.) Rumex rupestris is of considerable heritage value, and is legally protected at both European (Habitats Directive) and national (U.K. and France) levels; it is listed in the Red Data Books of both countries (Bioret 1995; King et al. 1999) (Table 1). Given its status as a European endemic, it is clear that the U.K., France and Spain each have a major responsibility for the global conservation of R. rupestris. THREAT ASSESSMENT - METHODS The assessment of potential threats was accomplished through a process of progressive synthesis based on both theoretical considerations and field observations of Rumex rupestris on a site-by-site basis. It was felt that more realistic and legitimate solutions could be derived by the use of specific cases and that a comparison with the assessments derived on theoretical grounds would reveal any discrepancies requiring further investigation. A second benefit of using specific site information was that it would reveal differences in susceptibility which could be related to particular factors operating at those sites and so provide more valuable comparative information. In terms of overall distribution patterns and aspects of the biology of R. rupestris, it was necessary to supplement field data with information published elsewhere. This reduced the need for duplication and also provided information that would not have been available to us from direct observation. Data on distribution were collated from a number of published and unpublished sources. This information was used to aid the selection of sites to be visited in order to obtain a reasonable spread of site types and geographical locations, commensurate with the constraints of time and resources. FIELD OBSERVATION Distribution patterns within France and Britain were examined in greater detail in order to choose suitable sampling sites and practical routes to be taken for their examination. 24 localities were sampled in Wales, Cornwall, Brittany and Vendee, in July 1998 and July 1999. Site visits were aimed at recording information on the ecological conditions pertaining to each site, and to assess the nature of any observable threats to individual plants of R. rupestris or to the population as a whole, or to the overall integrity of the plant community/habitat in which the species occurred. THREATS TO R UMEX R UPESTRIS IN B RIT AIN AND FRANCE 203 Habitat conditions were described in general terms in the field, by observation of the main characteristics of each site. Population size was recorded by counting individuals where the populations were small enough for this to be practical. For large, extensive populations, estimates were made, based on local counts to assess density and multiplying up over the area covered. Fecundity was also measured by direct counts of the number of individuals with fruits (or flowers, if the site visit was too early for fruits). For large populations, a series of representative counts covering the whole population was made. The presence of seedling plants (where these could be recognised as such) was noted. 204 ATLANTIC ARC CATEGORISATION OF THREATS We defined ‘threats’ as actual or potential factors that were considered likely to influence the long¬ term survival of R. rupestris populations. These included both natural and artificial/anthropogenic factors, which may be summarised as follows: Natural factors (geological/geomorphological/ecological/climatic) • The inherent instability (or stability) of the substrate, making it more (or less) vulnerable to change as a result of erosion or accretion. This factor can be seen to be ‘double-edged’, since erosion/accretion processes causing the demise of individual R. rupestris plants may also be required to provide patches of new habitat suitable for colonisation. • The level of exposure to the effects of climate-induced change, e.g. the effects of storm surges leading to erosion. • The effects of competition from associated species (e.g. invasion by Phragmites australis or Rubus fruticosus agg.). • The risks arising from small population size, and from isolation of populations or individuals. • The impacts of herbivores and diseases (may be especially important when affecting small or isolated populations). Anthropogenic factors • Irreversible destruction of (or changes to) the habitat as a result of development (e.g. urbanisation) or engineering (e.g. flood defence works). • Trampling (including all forms of direct, localised damage to individual plants or their habitat). • Marine pollution. • Terrestrial pollution. • Eutrophication (elevated levels of N and P) was recorded separately from other forms of terrestrial pollution, as it tends to be associated with different forms of land-use from that resulting in pollution by other organic or inorganic compounds. • Reduction or other changes to the supply of fresh water (e.g. through lowering of the water table, re-direction of surface watercourses or flow rate reduction). • Human-induced instability, when ‘natural’ instability is increased as a result of human disturbance (e.g. accentuated cliff erosion in areas with well-used cliff-top paths). Threats were assessed using a four-point scoring system (high, moderate, low, none) for each of the above factors. Zero values were attributed when there was deemed to be no threat from a particular source. From existing knowledge, information obtained from published and unpublished sources, and data acquired during the course of the field surveys in 1998, estimates were made of the potential overall level of threat to the species on each site and on all sites combined. The final threat assessment value was obtained by assigning a score to each component, adding these together and subtracting the value given to the potentially beneficial effects of local habitat instability (providing the opportunity for recolonisation). The result provided an overall, integrated value defining the vulnerability of the sampled Rumex rupestris populations to damage/extinction as a result of the threats identified (Table 2). We give only a summary of the results in this paper; for further details the reader should refer to Daniels & Bioret (1999). RESULTS ECOLOGY Rumex rupestris grows immediately above the high tide level, in rather sheltered places where there is some accumulation of drift materials, in environments characterised by the following main features: • South or south-west dominant exposure. • A constant supply of fresh water: streams, seepage zones, springs at the junction of superficial deposits (e.g. periglacial head material) and more impervious underlying rock strata. • Often at the base of rocky or clay (head) cliffs, or in rock crevices. • More rarely on the high part of shingle banks, at the back of beaches composed of coarse sand or gravel, or in dune slacks (a few populations, only in U.K.). All populations of R. rupestris are very localised, occupying small patches of suitable habitat, often less than 1 m2 and rarely extending to more than a few square metres (usually less than 10 m2). THREATS TO RUMEX RUPESTRIS IN BRITAIN AND FRANCE 205 TABLE 2. EXAMPLES OF DEGREE OF THREAT FROM DIFFERENT SOURCES ON FOUR POPULATIONS OF RUMEX RUPESTRIS IN FRANCE AND THE U.K., AND OVERALL ASSESSMENT FOR BOTH COUNTRIES COMBINED Threat Pencarrow (rock) Le Conquet (cliff) Score Pendower (beach/rock) La Govelle (cliff) Mean score for France & UK (all sites visited) Substrate instability 1 1 2 2 2 Severity of disturbance 2 1 3 1 2 Competition 1 2 3 2 2 Population size 3 2 3 2 3 Population configuration 3 2 3 2 3 Construction 0 1 2 3 2 Trampling 0 1 2 1 1 Marine pollution 1 1 2 1 1 Terrestrial pollution 0 1 1 1 2 Eutrophication 1 1 2 3 2 Reduction of water supply 1 1 2 3 3 Instability 0 0 3 1 2 Total threat 13 14 28 22 25 Benefits Substrate instability 0 0 1 0 1 Net threat 13 14 27 22 24 0 = no threat: 1 = minor threat: 2 = moderate level of threat: 3 = severe threat (Daniels & Bioret 1999) PHYTOSOCIOLOGY Our surveys indicate that Rumex rupestris is a characteristic species of two geosynvicariant plant associations (Bioret & Gehu 2002) (Fig. 2): • Soncho arvensis-Rumicetum rupestris Bioret & Gehu 2002 - from Wales to Finistere. • Apio graveolens-Rumicetum rupestris Bioret & Gehu 2002 - from Morbihan to Vendee. NUMBER OF PLANTS AND PROPORTION OF FLOWERING PLANTS Among eleven sampled populations (eight in France, three in Cornwall), the population size varied from just a few individuals to more than 50 plants, and the percentage of flowering plants is very variable from one population to another, varying between about 25% and 100% (Fig. 3). It is not known how much year-to-year variation there is in the flowering of individual plants. The small size of colonies has been confirmed by other surveys: for example, in Britain in 1994 there were 22 known sites supporting a total of 290-320 mature plants (average of 13-15 plants/ colony); of these, only three held more than 20 plants (King et al. 1999). A few larger colonies have been located more recently - most notably as a result of a survey carried out in 1999-2000 by Plantlife International - but it is still true to say that the vast majority of colonies hold very few plants, usually less than 20 individuals in each (S. J. Leach pers. comm.). ASSESSMENT OF THREATS Table 2 shows the overall level of threat to Rumex rupestris from the factors identified above. Data are shown for four contrasting sites in France and S.W. England, together with an overall mean assessment for the full set of sites surveyed. As already noted, this species is found in highly dynamic zones of the coast where large-scale changes may occur over a very short time period. Populations are subject to damage linked with normal coastal erosion processes, for example through changes in beach profiles associated with storm surges and cliff falls. Such processes may eliminate individual plants or whole colonies, while on the other hand providing in some cases suitable micro-sites for colonisation. The species possesses seeds capable of surviving inundation in seawater (Daniels et al. 1998), and it is possible that erosion events - or down-washing of seeds 206 ATLANTIC ARC FIGURE 2. Geographic distribution of the two geosynvicariant associations. O Soncho an’ensis - Rumicetum rupestris, • Apio graveolens - Rumicetum rupestris. THREATS TO RUMEX RUPESTRIS IN BRITAIN AND FRANCE 207 □ population size ■ % flowering 1 23456789 10 11 Figure 3 : Population size and flowering % of 1 1 localities in UK and France (1998 and 1999) I : Sainte-Marguerite, 2 : Le Croisic, 3 : Batz-sur-mer, 4 : Le Pouliguen, 5 : Quiberon, Port-Blanc, 6 : Quiberon l'arche de Port-Blanc, 7 : Guidel Plage, 8 : Glenan, Le Loc'h, 9 : Pendower beach, 10 : Pencarrow, I I : Soar Mill Cove. into the top of the littoral zone by fresh water - have positive effects by helping seed dispersal. Whether burial due to accretion of beach material or slumping of cliffs (and subsequent re¬ exposure) provides suitable conditions for seed survival is not known, but if it does then this would constitute, in effect, a form of short-term seed banking. Those populations of R. rupestris situated above the action of the highest tides may be threatened by the natural instability of the cliffs or the sandy or shingle beach communities in which they are found. Among the physically more stable zones in which Rumex is found (e.g. where it encroaches into grassland communities or is present in dune slack areas), there would appear to be more opportunities for the development of a persistent buried seed bank, but competition from other species - and ecological succession - in such habitats may be more severe. DISCUSSION Further studies could be carried out on a whole range of topics within the context of a bilateral co¬ operation programme between France and the U.K. These studies could contribute directly to the development of an international (European) Action Plan for this threatened species. We have identified a range of issues which would merit, in our view, further research. We have noticed that most populations of Rumex rupestris (with one or two notable exceptions) consist predominantly of older individuals and, although seed production is high, the establishment of new seedlings seems to be a comparatively rare event. Existing monitoring studies at some sites may throw more light on the demographic processes within populations; but from our own observations we think that in many sites where the plant is found, especially on cliffs and cliff/ beach junctions, suitable habitat patches for colonisation are uncommon, as well as being highly 208 ATLANTIC ARC FIGURE 4. Distribution of the Shore Dock on the Isle of Ouessant, western France, showing scattered coastal distribution of the species. disjunct/fragmented in distribution. The probability of seeds arriving at suitable germination micro-sites appears to be low. Thus, population size will tend to remain small, with all the consequences - loss of individuals through stochastic events and loss of genetic diversity - that this brings. There has been much discussion in the literature about the concept of minimum viable population size (Shaffer 1981; Gilpin & Soule 1986; Lande & Barrowclough 1987; Nunney & Elam 1994; Montalvo et al. 1997). A number of attempts have been made to define minimum viable population size, on the basis of the findings of McArthur & Wilson (1967) working with island populations, on pragmatic studies by various animal breeders and the discovery of the so- called ‘Allee effect’ (Allee et al. 1949) which suggested adverse consequences (e.g. failure to find a mate) for populations falling below a critical population density. The complexity of the problem has been discussed by other authors, notably Gilpin & Soule (1986), who suggested that the concept of minimum viable population size involved a multi-dimensional set of thresholds, incorporating aspects of population genetics, demography, community processes and distribution ecology. Despite the difficulties attached to any estimate of minimum viable population size. Biodiversity Action Plans still regard such estimates as important when attempting to establish whether species (or individual populations or metapoplations) are achieving ‘favourable conservation status’. The logic here is that the conservation status of a species unable to achieve a viable population size must, by definition, be ‘unfavourable’; and that a major objective of biodiversity action planning must be to maintain populations of sufficient size to make them capable of surviving random damaging events (e.g. storm surges, pollution incidents). If the population is large enough (and widely distributed enough), then it follows that there will be survivors following any such events. All biological communities are dynamic, however, either because relationships between component species change, or because the environment itself is subject to progressive modification. Under such circumstances populations also need to be capable of adaptive evolutionary change in response to the selection pressures imposed. It is therefore important to have some method of assessing the population sizes needed to achieve these objectives. THREATS TO RUMEX RUPESTRIS IN BRITAIN AND FRANCE 209 Habitat conditions, life history and reproductive strategies differ among species within the supra-littoral zone, so the point at which the number of individuals becomes critical to population survival will also be different. The range of typical population sizes among extant populations and the role of the species within the community may provide clues to potential for survival. In the case of our sample populations of Rumex rupestris, the presence of a few tens of individuals of Rumex rupestris at any one site seems to be indicative of a very healthy population. Only two of our sample populations contained more than 50 plants (unless the whole of the north coast of Ouessant (Fig. 4) is regarded as a single population - in which case there would be three such populations). It should be noted that R. rupestris is self-compatible, meaning that even a single individual could, in some senses, be regarded as constituting a ‘viable’ population. On the other hand, some degree of gene interchange and re-assortment is usually regarded as a necessary requirement for the maintenance of a resilient genetic resource; and a single plant will always be more vulnerable to extinction than a group of plants. Because Rumex rupestris fits within the Rabinowitz category of plants having a restricted distribution and small population size, it may always have been rare. This is borne out, at least to some extent, by historical records, since the species has only ever been recorded as consisting of small populations. Thus, in the case of R. rupestris , small population size and geographical restriction should not, perhaps, be automatically interpreted as meaning ‘unfavourable conservation status’; rather, we may need to view this species as a highly adapted plant, having a range of ecological and life history traits which have evolved to enable it to cope successfully with a particularly harsh set of environmental circumstances. We think it likely that neighbouring populations could, in effect, be functioning as components of larger metapopulations. If this is indeed the case, then it is clearly important for the population as a whole that we try to maintain a large proportion of the small, but inter-connected, sub-populations; and to protect areas of suitable but currently unoccupied habitat that may provide the plant with opportunities for future colonisation if other (currently occupied) sites become unsuitable for some reason. A study of the genetics of the Ouessant populations (Fig. 4) would be invaluable in helping determine whether these populations are distinct, or whether they comprise a single metapopulation with some evidence of gene flow between the individual sub-populations. Only long-term monitoring studies can determine the actual longevity of individual plants and populations, though such studies would currently be compromised by the presence of anthropogenic threats. REFERENCES Abbayes, H. Des, Claustres, G., Corillion, R. & Dupont, P. (1971). Flore et vegetation du Massif Armoricain. Vol. 1: Flore vasculaire. Presses Universitaires de Bretagne, Saint-Brieuc. Allee, W. C., Emersen, A. E., Park, O., Park, T„ & Schmitt, K. P. (1949). Principles of animal ecology. Saunders, Philadelphia. BlORET, F. (1995). Rumex rupestris Le Gall, in OLIVIER, L., GALLAND, J.-P., MAURIN, H. eds. Livre rouge de la flore menacee de France. Vol. 1: especes prioritaires. M.N.H.N./Service du Patrimoine naturel, Ministere de l’Environnement, D.N.P., Paris. BlORET, F. & Gehu, J.-M. (2002). Revision phytosociologique des vegetations a Rumex rupestris sur les littoraux atlantiques europeens. Journal de Botanique, Societe Botanique de France 20: 45-54. DANIELS, R. & BlORET, F. (1999). A conservation strategy for threatened botanical resources of the English Channel-Atlantic coastal environment. Alliance Programme no. PN 98.039, I.T.E., Geosystemes U.M.R. 6554 C.N.R.S. DANIELS, R. E., MCDONNELL, E. J. & RAYBOULD, A. F. (1998). The current status of Rumex rupestris Le Gall (Polygonaceae) in England and Wales, and threats to its survival and genetic diversity. Watsonia 22: 33- 39. DUPONT, P. (1962). La flore atlantique europeenne. Introduction a I’etude du secteur ibero-atlantique. 414 pp. Faculte de Sciences, Toulouse. DUSSAUSOIS, G. (1996). Prospections floristiques en Gironde. Bulletin de la Societe Linneenne de Bordeaux 24 (4): 189. Gilpin, M. E. & Soule, M. E. (1986). Minimum viable populations: processes of species extinction, in SOULE, M. E. ed. Conservation biology: the science of scarcity and diversity, pp. 19-34. Sinauer Associates, Sunderland, Massachusetts. Kay, Q. O. N. (1996). The conservation o/Rumex rupestris ( Shore Dock ) in Wales: past, present and future sites and habitats for Rumex rupestris in South and West Wales. Countryside Council for Wales, unpublished report. 210 ATLANTIC ARC King, M. P., Mcdonnell, E. J., Leach, S. J. & Wigginton, M. J. (1999). Rumex rupestris Le Gall, in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants , pp. 320-321. Joint Nature Conservation Committee, Peterborough. Lande, R. & Barrowclough, G. F. (1987). Effective population size, genetic variation, and their use in population management, in SOULE, M. E. ed. Viable populations for consen’ation. Cambridge University Press, New York. McArthur, R. M. & Wilson, E. O. (1967). The theory of island biogeography. Princeton University Press, Princeton. McCLINTOCK, D. (1975). The Wild Flowers of Guernsey. Collins, London. Montalvo, A. M., Williams, S. L., Rice, K. J., Buchmann, S. L., Cory C., Handel, S. N., Cabhan, G. P., Primack, R. & Robichaux, R. H. (1997). Restoration biology: a population biology perspective. Restoration Ecology 5: 277-290. Nunney, L. & Elam, D. R. (1994). Estimating the effective population size of conserved populations. Conservation Biology 8: 175-184. PHILIPPON, D. (1991). Departement des Cotes d' Armor. 76 plantes protegees ou menacees. Prefecture Cotes d’ Armor & DDE, Saint-Brieuc. Unpublished. PROVOST, M. (1993). Atlas de repartition des plantes vasculaires de Basse-Normandie. Centre de recherches sur 1’evolution de la vie rurale & Presses Univ. Normandie, Caen. SHAFFER, M. L. (1981). Minimum population sizes for species conservation. BioScience 31: 131-134. 211 Rumex rupestris Le Gall (Shore Dock) in S.W. England: review of recent surveys and assessment of current status e. j. McDonnell The Old Gaworks Cottage, The Lerbume, Wedmore, Somerset, BS28 4ED, England and M. P. KING 91 South Court Avenue, Dorchester, Dorset DTI IDA, England ABSTRACT Recent surveys of Rumex rupestris Le Gall in S.W. England are reviewed, and its current status is summarised. The number of known populations has increased dramatically since the early 1990s. This increase reflects an improved state of knowledge following targeted surveys during the mid to late 1990s, including a major survey in 1999 in S. Devon and along the south coast of Cornwall. Extant populations are now being monitored under Plantlife’s ‘Flora Guardian’ scheme. Since 1999 R. rupestris has been reliably recorded from 36 sites in Devon and Cornwall (including the Isles of Scilly). Keywords: Shore Dock, Rumex rupestris , survey, monitoring, distribution, conservation. INTRODUCTION Rumex rupestris Le Gall (Polygonaceae) is a European endemic, with a world range restricted to the Atlantic seaboard of western Europe. It is recorded from Spain (Galicia), France (Brittany, Normandy, Vendee), Channel Islands, England (Devon, Cornwall, Isles of Scilly) and Wales (Glamorgan, Pembrokeshire, Anglesey). More than any other species, perhaps, R. rupestris serves as a botanical emblem for the full sweep of the Atlantic Arc - that is, apart from the south of Ireland, where it is mysteriously absent. Emblematic it may be, but for many botanists on the Atlantic fringes of Europe this remains an elusive and unfamiliar species. The genus Rumex is frequently dismissed as rather difficult (even dull); R. rupestris can be readily overlooked as ‘just another dock’, and it is undeniably rare and threatened throughout its range (Bioret & Daniels 2005) - but it is far from dull. R. rupestris is often said to be the ‘rarest dock in the world’, and clearly the English and Welsh strongholds for this species are of global significance. It is listed as ‘endangered’ in the British Red Data Book (King et al. 1999), and is specially protected by being on Schedule 8 of the Wildlife & Countryside Act 1981. It is one of only nine British vascular plant species listed on Annexes II (b) and TV (b) of the European Directive on the Conservation of Natural Habitats and of Wild Fauna and Flora 1992 (the Habitats Directive), and it is also - not surprisingly - a Priority Species in the U.K. Biodiversity Action Plan (H.M.S.O. 1995). In the early 1990s, when R. rupestris was afforded legal protection in Britain and added to English Nature’s Species Recovery Programme, the plant’s distribution and ecology was poorly understood. Apart from a survey carried out by King (1989), we had little to go on. Then, with funding from English Nature - and the subsequent inclusion of R. rupestris on Plantlife’s 'Back from the Brink’ programme - there followed a series of studies to gain a fuller picture of the biology, distribution and status of this species in Britain. A summary of the survey work is given below - including full reference to the considerable volume of ‘grey literature’ on this species - along with an assessment of current status updated from that given in Daniels et al. (1998). 212 ATLANTIC ARC A SUMMARY OF SURVEYS 1994-1999 In 1994 a survey of known sites supporting R. rupestris was undertaken in Devon and on the Cornish mainland (McDonnell 1995). We did not set out, at this stage, to search for ‘new’ populations, and the sites visited were based very largely on the eleven extant colonies recorded in the late 1980s by King (1989). The survey involved recording for each site the size of its R. rupestris population; plus a detailed record of its topography, hydrology, associated species and possible threats. At the same time, a similar survey was carried out in the Isles of Scilly (Parslow & Colston 1994). In 1995, former sites in Dorset, Devon and mainland Cornwall were surveyed to check for the presence (or absence) of R. rupestris , and to assess whether any would make suitable locations for re-introduction attempts (Daniels et al. 1996). Three potentially suitable sites were identified, and seed of R. rupestris was collected under licence from appropriate donor sites. In 1995 we did turn up a few ‘new’ R. rupestris colonies, although none of these coincided with any of the ‘old’ sites being investigated. In 1996 R. rupestris was re-introduced to three sites - one on Devon and two in Cornwall. These sites were subsequently monitored by fieldworkers from the Institute of Terrestrial Ecology (now CEH), who were also involved in carrying out genetic research, and germination, dormancy and dispersal studies (Daniels & Moy 1998). A further survey of R. rupestris populations in the Isles of Scilly was also undertaken (Parslow 1996). Volunteer Plantlife ‘minders’ (or Flora Guardians as they are now called) were recruited in 1997 for most of the R. rupestris sites in Devon and Cornwall. Each volunteer was shown the plants at his/her site, and was introduced to the subtleties of Rumex identification and the special features that are diagnostic of R. rupestris (McDonnell 1998). At about this time, we reviewed the ‘Back from the Brink’ survey work carried out during 1994- 1997 and noted that several previously unrecorded sites had been discovered. Fieldworkers were more experienced now at identifying the species, and were becoming more adept at recognising and homing in on patches of suitable habitat. Often, ‘suitable habitat’ could be reliably predicted by examining the Ordnance Survey map to locate places where seepages, springs and streams debouched onto rocky shores. Thus, in 1998 a small pilot project was undertaken to search for R. rupestris along sections of coastline where the habitat ‘looked right’ but where the plant had not previously been recorded. The survey was outstandingly successful: several new populations were located, and by the end of the 1998 field season we knew of 21 extant sites in Devon and Cornwall (excluding Scilly) - more than twice the number known from that area just four years earlier (McDonnell 1999). The 1998 ‘pilot’ was so successful that Plantlife decided - with financial backing and advice from English Nature - to carry out a more ambitious survey, targeting long stretches of coastline in S. Devon and the south coast of Cornwall that were considered ‘potentially suitable’ for R. rupestris. We examined the British Geological Survey maps and 1:25,000 scale Ordnance Survey maps, and used the presence of coastal exposures of Quaternary ‘head’ deposits and evidence of freshwater seepages, springs or streams to indicate areas of suitable habitat. Most of the selected search areas were reached on foot from the shore or cliff-top, but areas considered to be inaccessible or too dangerous were searched by boat, using binoculars to locate plants of R. rupestris or patches of potentially suitable habitat. Where possible, we checked such areas (and any possible Rumex plants) by scrambling ashore. On several occasions, when the boat could not be taken close to shore, we had to resort to jumping overboard and swimming ! By the end of the 1999 field season the number of extant R. rupestris sites had risen to 32 (McDonnell & King 2000). In addition, several ‘possible’ sites found during 1999 were re-located and confirmed in 2000 (Neil et al. 2001); the most noteworthy of these, at Tregiffian/Boscawen, W. Cornwall - which had been spotted from the boat in 1999 - was confirmed by Ian Bennallick in 2000, and is now thought to be the largest known ‘rocky shore’ colony in Britain (92 plants in 2001). In the field we soon discovered that Rumex rupestris plants can be extremely difficult to refind for monitoring purposes - even when the person attempting to refind them is the same as the person who found them in the first place! R. rupestris is easily overlooked, and if not fruiting can be dismissed as another Rumex , e.g. R. crispus var. littoreus or R. conglomeratus. The drawing of RUMEX RUPESTRIS IN S.W. ENGLAND 213 TABLE 1. DETAILS OF EXTANT RUMEX RUPESTRIS SITES IN DEVON AND CORNWALL (INCLUDING THE ISLES OF SCILLY), AS AT SEPTEMBER 2004 VC SITE NAME GRID REF ABUNDANCE DATE 1 Constantine Bay (reintroduction site) SW8674 2V 2000 1 Penhale Sands SW7656, 7657, 7756 172 (C/V/F) 2002 1 Gear Sands SW7656, 7655 22 (C/V/F) 2001 1 Annet (Isles of Scilly) SV8608 Present 2001 1 Samson (Isle of Scilly) SV8712, 8713 16F, 9V 2001 1 Tean (Isles of Scilly) SV9116 5C, 1 IF, 18V 2001 1 Tresco (Isles of Scilly) SV8913 10F 2001 1 Lamoma Cove SW4524 5F 2000 1 Tregiffian/Boscawen SW4323 1C, 28F, 63V 2001 1 Trebarvah-Stackhouse Cove SW5428-5528 13C, 8F, 6V 2001 2 Great Molunan SW843 1-5528 5C+1F 2002 2 Peter’s Splash SW8734 1C+1F 2003 2 Raven’s Hole SW8734 1C, 6F, 2V 1998 2 Porthbean Beach SW8836 1C 2001 2 Pendower SW9038 2C 2002 2 Pencarrow SX1550 3C, 8V 2001 2 Watchhouse Cove SX1550 IV 2001 2 Conical Hill SX1551 6C 2001 2 Parson’s Cove SX1651 11 (C/V/F) 2001 2 Lansallos Cliff SX1750 1C 2001 2 Colours Cove SX1950 1? 2001 2 Talland Bay SX2351 1C, 4F, 2V 1999 2 Samphire Beach SX2451 1C 2002 2 Portwrinkle (reintroduction site) SX3553 9F, 8V 1999 2 Rame SX4249 13C, 7F, 6V 2001 3 Wembury SX5148 1C 2002 3 Blackstone Point SX5346 6C, 4F, 10V 2002 3 Stoke Beach SX5646 1C 2001 3 Wadham Beach (reintroduction site) SX5746 Absent? 2002 3 Near Keaton Cove SX5947 1C, IF 2000 3 Westcombe Beach SX6345 1C, IV 1999 3 Soar Mill Cove SX6937 1C, 4F, IV 2003 3 Rickham Sands/Gara Rock SX7536 IF 2004 3 Venericks Cove SX7636 1C, 14F, 29V 1999 3 Elender Cove SX7635 IF 1999 3 Great Mattiscombe Sands SX8136 IF, IV? 2004 Abundance: C = Clump; V = Vegetative Plant; F = Flowering/Fruiting Plant. Date: Year of most recent record. sketch maps is difficult, especially due to the fact that so often the terrain is more vertical than horizontal. From the outset, therefore, we decided to keep a detailed photographic record of the colonies. During the first field season (1994), a simple method was devised for producing panoramic views of each site; this was achieved by taking a series of overlapping photographs from the beach, looking towards the back of the shore and the cliffs. The photographic prints were then assembled to form a photomontage of the coastline, on which the exact location of R. rupestris was marked. We also took close-up ’portrait’ shots of individual plants and clumps: often, the successful relocation of a plant would depend on a detailed examination of the land- forms (bed-rock, boulders, etc.) in the immediate vicinity, and the close-up photographs were particularly useful in this respect. 214 ATLANTIC ARC A complete set of panoramas have been made available to Plantlife and English Nature, and several of them have been produced as laminated colour photocopies for use in the field by Plantlife’ s Flora Guardians or others when carrying out monitoring visits. This approach is now being complemented by use of GPS data, which becomes more accurate every year. CURRENT STATUS OF RUMEX RUPESTRIS IN S.W. ENGLAND By the end of 1999 we felt we had a good picture of the distribution of R. rupestris in S.W. England. It was a considerable improvement on that given by Daniels et al. (1998), who listed just 18 extant populations (excluding the three re-introduction sites) compared with 29 known sites following the 1999 survey. But the picture was still incomplete: in particular we were aware of several stretches of suitable coastline that we had been unable to visit in 1999. Thus, surveys from 2000 onwards have concentrated on filling these gaps in coverage, while at the same time ensuring that extant sites are kept under surveillance (King 2002). These further surveys have turned up several new colonies, including 27 plants in W. Cornwall near Marazion (Trebarvah to Stackhouse Cove) in 2001 (King 2002), while single plants were refound at two sites in S. Devon (Great Mattiscombe Sands and Rickam Sands) in 2004 (S. J. Leach pers. comm.). Further work is needed to complete the survey in S.W. England, specifically from Prussia Cove to Porthleven (west of The Lizard) and along the S. Devon coast from Start Point to the Dorset border. Partial surveys on the north coast of Cornwall have so far drawn a blank, but there is much suitable habitat that would merit further investigation - the problem there, as in some other areas, is that so much of the coastline is difficult (or impossible) to reach on foot. Details of all extant R. rupestris sites in S.W. England are given in Table 1: this is based on the latest summary information presented in the species dossier for this species on the Plantlife website (King 2003-2004), updated as necessary to incorporate field data gathered in the eighteen months since that dossier was drafted. ACKNOWLEDGMENTS We would like to thank all those botanists - including many members of B.S.B.I. and Plantlife - who have been involved in the survey and monitoring of Rumex rupestris in Devon and Cornwall. Particular thanks to Ian Bennallick, Roger Daniels, David Holyoak, Simon Leach, Rosaline Murphy, Catriona Neil and Rosemary Parslow; thanks also to the staff of English Nature for their help with various aspects of the survey work, and to all those owners and occupiers who allowed us access to their land. We are grateful to Simon Leach for his extensive comments on an early draft of this paper. Much of the work reported here was carried out on behalf of Plantlife, with financial assistance from Center Parcs and English Nature’s Species Recovery Programme. REFERENCES Bioret, F. & Daniels, R. (2005). Assesment of threats to populations of Rumex rupestris Le Gall (Shore Dock) in Britain and France, in Leach S. J., Page, C. N., Peytoureau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc, pp. 201-209. Botanical Society of the British Isles, London. Daniels, R. E., Mcdonnell, E. J. & Raybould, A. F. (1998). The current status of Rumex rupestris Le Gall (Polygonaceae) in England and Wales, and threats to its survival and genetic diversity. Watsonia 22: 33- 39. DANIELS, R. E. & MOY, I. L. (1998). Species Recovery Programme - Shore Dock (Rumex rupestris Le Gall). Second Report. Report to English Nature, Species Recovery Programme. H.M.S.O. (1995). Biodiversity: the UK steering group report. Volume 2: Action Plans, p. 193. H.M.S.O., London. KING, M. P. (1989). An investigation into the current status and ecology of the shore dock Rumex rupestris in Devon and Cornwall. M.Sc. thesis. University College, London. KING, M. P. (2002). Shore Dock Rumex rupestris in 2001. Plantlife Report No. 196. KING, M. P. (2003-2004). Species Dossier for Rumex rupestris Le Gall. Plantlife. http://www.plantlife.org.uk/ downloads/species_dossier/Rumex_rupestris_dossier RUMEX RUPESTRIS IN S.W. ENGLAND 215 King, M. P., Mcdonnell, E. J., Leach, S. J. & Wigginton, M. J. (1999). Rumex mpestris Le Gall, in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants, pp. 320-321. Joint Nature Conservation Committee, Peterborough. MCDONNELL, E. J. (1995). The status of shore dock (Rumex rupestris Le Gall) in Britain in 1994. Report to English Nature, Species Recovery Programme. MCDONNELL, E. J. (1998). Rumex rupestris (Shore Dock). Report on fieldwork. Plantlife Report No. 101. MCDONNELL, E. J. (1999). Rumex rupestris (Shore Dock). Report on 1998 fieldwork. Plantlife Report No. 128. MCDONNELL, E. J. & KING, M. P. (2000). Rumex rupestris (Shore Dock). Report on fieldwork undertaken in 1999. Plantlife Report No. 140. Neill, C. J., King, M. P., Evans, S. B., Parslow, R. E., Bennallick, I. B. & Mcdonnell, E. J. (2001). Shore Dock (Rumex rupestris). Report on fieldwork undertaken in 2000. Plantlife Report No. 175. PARSLOW, R. (1996). Shore Dock Rumex rupestris Le Gall in the Isles of Scilly. Report to English Nature, Species Recovery Programe. PARSLOW, R. & Colston, A. (1994). The current status of Rumex rupestris Le Gall in the Isles of Scilly. Report to English Nature, Species Recovery Programme. 217 A review of recent work on the Limonium binervosum aggregate (rock sea-lavenders) in the British Isles S. J. LEACH English Nature, Roughmoor, Bishop's Hull, Taunton, Somerset, TA1 5AA, England1 and D. A. PEARMAN Algiers, Feock, Truro, Cornwall TR3 6RA, England 1 ABSTRACT Limonium binervosum agg. in the British Isles comprises a complex of nine species, 17 named subspecies and 16 named varieties. Eight of the species (and all but one of the subspecies and varieties) are thought to be British-and-Irish endemics. In the U.K. the endemic taxa are included as priority species in the national Biodversity Action Plan; they are also listed in the British Red Data Book. As such, they have begun to attract a considerable interest in recent years: surveys are now being carried out to establish the national, regional and local distributions of the various taxa; and the taxonomy of the group is being reviewed following a molecular study of the species-complex using amplified fragment length polymorphism (AFLP). Once various taxonomic issues have been resolved, it is hoped that the B.S.B.I. will publish a handbook of the British and Irish Limonium. Keywords: distribution, taxonomy, AFLP, Biodiversity Action Plan. INTRODUCTION Limonium binervosum agg. (Plumbaginaceae) is a difficult, taxonomically critical group encompassing all the Limonium taxa in the British Isles apart from L. vulgare, L. humile, L. bellidifolium and two species restricted to the Channel Islands (L. auriculae -ursifolium and L. normannicum). The aggregate belongs to the Oceanic Temperate element of the British flora (Preston & Hill 1997), and is supposedly restricted to the Atlantic seaboard of W. Europe, from Portugal and Spain northwards to France, Britain, Ireland, the Channel Islands and the Isle of Man. It reaches the northern limit of its world range in S.W. Scotland and Northern Ireland. Until the mid 1980s the L. binervosum group comprised just four species in Britain and Ireland: L. binervosum ( sensu stricto ), plus the three rare endemics L. paradoxum, L. recurvum and L. transwallianum. Then, an overhaul of the taxonomy by Martin Ingrouille (Ingrouille & Stace 1986) resulted in a re-definition of L. recurvum and L. binervosum s.s., and the delimitation of a further five endemics - L. britannicum, L. dodartiforme, L. loganicum, L. parvum and L. procerum - along with 17 subspecies and numerous varieties (Table 1). Only one species (L. binervosum s.s.) is not endemic to the British Isles - the others, as far as we know, are found nowhere else in the world. All the taxa are tetraploid (2n = 35 or 36) apart from L. recurvum which is triploid (2n = 27). Publication of the revised taxonomy led to a minor upsurge of interest in the British and Irish rock sea-lavenders - although it has to be admitted that, even today, many field botanists are reluctant to tackle such a taxonomically difficult group. All the species were included, with illustrations, in Stace (1991), and distribution maps were published for the first time in Scarce 1 e-mail: simon.leach@english-nature.org.uk 'e-mail: DPearman4@aol.com 218 ATLANTIC ARC TABLE 1. A CHECKLIST OF THE LIMONIUM BINERVOSUM AGG. SPECIES, SUBSPECIES AND VARIETIES OCCURRING IN THE BRITISH ISLES 1 . Limonium binervosum subsp. binervosum (E. Sussex & S.E. Kent) subsp. cantianum (N.E. Kent) subsp. anglicum (Norfolk & Lines) subsp. saxonicum (N. Essex) subsp. mutatum (S. Devon) subsp. sarniense (Channel Islands) 2. Limonium paradoxum (N. Pembs) [vars paradoxum and medium ] 3. Limonium procerum subsp. procerum (widespread - S.W. England, Wales, E. Ireland) [vars procerum, medium, hibernicum, cornubiense, paramedium and wessexense ] subsp. devoniense (S. Devon) subsp. cambrense (S. Pembs) 4. Limonium britannicum subsp. britannicum (N. Cornwall) [vars britannicum and kelseyanum ] subsp. coombense (S. Devon & E. Cornwall) [vars coombense and grandicaule ] subsp. transcanalis (N. Devon & Pembs) subsp. celticum (N. Wales & N.W. England) [vars celticum and pharense] 5. Limonium parvum (S. Pembs) 6. Limonium loganicum (W. Cornwall) 7. Limonium transwallianum (S. Pembs) 8. Limonium dodartiforme (Dorset) 9. Limonium recurvum subsp. recurvum (Dorset) subsp. portlandicum (Dorset & Co. Kerry) [vars portlandicum and recurviforme ] subsp. pseudotranswallianum (Co. Clare) subsp. humile (Donegal, Cumbria, Galloway) Plants in Britain (Leach 1994), and then in the 3rd edition of the British Red Data Book (Wigginton 1999). Suddenly, the rock sea-lavenders were acquiring a measure of respectability (on a par with the whitebeams ( Sorbus spp) and eyebrights ( Euphrasia spp)); and the rare and/or endemic species - and even subspecies - were starting to be taken more seriously by botanists and conservationists. The endemic taxa were included en bloc in the U.K. Biodiversity Action Plan (U.K. Biodiversity Group 1998), and a steering group was formed to oversee implementation of the ‘endemic Limonium’’ Action Plan; and, most recently, all of Ingrouille’s species (and subspecies) were included - unlike the much more numerous dandelions, hawkweeds and brambles - in the New Atlas (Preston et al. 2002). It is extremely difficult, however, to relate the taxonomy of the British rock sea-lavenders to that of the group in the rest of Europe. Llora Iberica (Erben 1993), for instance, recognises no less than 107 Limonium taxa (compared with 45 in Llora Europaea (Pignatti 1972)), but defines no aggregates within the genus. Thus, it is unclear what proportion of the Iberian taxa fall within the L. binervosum group as presently understood. As the taxonomy stands at the moment L. binervosum agg. is an ‘Atlantic’ group; however, taxonomic studies have only been conducted on LIMONIUM BINERVOSUM AGG. IN THE BRITISH ISLES 219 specific geographic ranges, and recent phylogenetic work by Lledo et al. (in press) based on DNA sequence data places L. binervosum with other apomictic taxa from the Mediterranean. It should be noted that this placement was based on the inclusion of only one taxon from the L. binervosum species-complex, but it opens up the possibility that further study may show that the group does indeed extend into the Mediterranean. Certainly, there are taxa that look to be extremely close to ours in the Balearic Islands, and on holiday in 2003 in Syracusa (Sicily) one of us (D.A.P.) collected Limonium specimens that cannot be easily distinguished from those in Britain. We are not aware of any comprehensive review of Limonium in France; indeed, French colleagues at the Atlantic Arc conference confirmed that such a review had yet to be done, and that, for the moment at least, the taxonomic relationships between British and French species are far from clear. DISTRIBUTION MAPPING The BAP process has certainly helped to generate some interest in the Limonium binervosum group, with a ‘core’ of enthusiasts keen to identify gaps in our knowledge and work out ways of getting them filled. We soon realised, for example, that distribution maps of the individual taxa were woefully incomplete and not fully up-to-date; indeed, most dots on the maps (even those in the New Atlas) were for populations determined in the late 1970s/early 1980s by Martin Ingrouille, and few records had been added since that time. In England and Wales, the B.S.B.I. has undertaken a detailed review of the records, both for the individual taxa and for the aggregate as a whole (Leach et al. 2001; Lockton 2003). Essentially, this has involved for each 10-km square taking any records we can find and trying to pinpoint them on a 1:50,000 scale map. The maps have then been supplemented by text, containing information gleaned from the B.R.C. Vascular Plants Database, the Threatened Plants Database, county Floras, local recorders and other sources. The jigsaw is still incomplete, but at least we now have a clear statement of what we know (and don’t know) about the rock sea-lavenders in each 10-km square. In many squares there are taxonomic ambiguities and doubts to be resolved, identifications to be confirmed, and populations of the ‘agg.’ to be critically determined. Even for those squares that have been supposedly well covered, we are usually still lacking information on the size of colonies (both their extent and the number of plants they contain), and often have little idea of associated species and a dearth of habitat data. There are hundreds (possibly thousands) of colonies still needing closer examination; and there is considerable scope for field botanists to add to our knowledge of particular taxa through detailed local studies (e.g. Leach 2005). TAXONOMIC CONSIDERATIONS As already noted, since the mid 1980s our taxonomic view of the rock sea-lavenders has been that of Ingrouille & Stace (1986). This classification was based on a ‘taxometric’ analysis of a sample of Limonium binervosum agg. populations gathered from throughout Britain, Ireland and the Channel Islands, and made use of both morphological and cytological characters. Certainly, with familiarity and plenty of practice, the species described by Martin Ingrouille do appear to make sense morphologically, and are readily identifiable in the field (although distinguishing between the various subspecies and varieties can be troublesome). It would, however, be helpful to know whether there is an underlying genetic basis to the apparent morphological differences between species x and species y, and whether, from a molecular standpoint, x and y are sufficiently different from one another to continue to merit recognition as distinct species. Recent genetic fingerprinting studies (Cowan et al. 1999), part-funded by English Nature, do go some way to confirming that morphologically similar populations are also close genetically. Preliminary results of AFLPs (Fig. 1) - analysed using the UPGMA (Unweighted Pair-Group Method using Arithmetic Averages) method of genetic distance analysis - on a limited number of taxa and samples do show some weak partitioning between the samples, summarised below; it should be borne in mind, however, that the genetic distances between taxa are much smaller than that found in other apomictic groups which, unlike L. binervosum agg., have occasional outcrossing events, e.g. Sorbus (R. Cowan pers. comm.). 220 ATLANTIC ARC bin anglicum bin anglicum Nazare rec port rec loganicum paradoxum parad mut paradoxum rec port kerry La Joseliere bin binervosum bin binervosum paradoxum bin cantianum bin cantianum britannicum transwallianum Lequieto Carteret St Lunaire procerum procerum procerum britannicum parvum transwallianum Noja dodartiforme 2-5% genetic distance FIGURE 1. UPGMA tree showing relationships between taxa within the Limonium biner\>osum group (from: Cowan et al. 1999). • The two samples of L. binervosum subsp. anglicum included in the analysis cluster together, but are not as close to L. binervosum subsp. binervosum as some other taxa that have been ranked as full species. Subsp. saxonicum from neighbouring Essex has yet to be included in the analysis, and it may be that, with further sampling, the genetic data will not uphold the existing taxonomic position (and rank) of these E. Anglian taxa. • Samples of L. recurvum also cluster together. This is a geographically highly disjunct group, comprising the British and Irish triploid taxa (all other taxa in our area are tetraploid). It is reassuring to see them clustered together in a way that suggests that their grouping within a single named morpho-species makes good sense genetically. But note that the northern and western subsp. humile has not yet been sampled - will it, or will it not, fall within this cluster? • L. dodartiforme clusters with French and Spanish samples of L. binervosum agg. However, the genetic separation of these taxa from the other British taxa is weak, and it remains to be seen whether the inclusion of more taxa - and more samples of each taxon - reinforces or negates this grouping. LIMONIUM BINERVOSUM AGG. IN THE BRITISH ISLES 221 • There is a ‘paradoxum/loganicum’ group, encompassing L. loganicum, L. paradoxum and L. binervosum subsp. binervosum. It seems to us unsurprising that L. loganicum and L. paradoxum should belong to the same ‘cluster’, given the similarity in their geographical outposts: L. loganicum on exposed sea-cliffs in the extreme south-west comer of England (West Penwith); L. paradoxum in the same kind of habitat in the south-western comer of Wales (St David’s Head, Pembrokeshire). The apparent genetic closeness between these two taxa and the Kentish populations of L. binervosum subsp. binervosum , however, was unexpected. The latter, incidentally, is supposedly our only non-endemic taxon, being found also in N.W. France (Ingrouille & Stace 1986). • And lastly there is a ‘procerum/britannicum’ group, potentially encompassing a quite large number of the taxa recognised by Ingrouille & Stace (1986) - including the relatively widespread L. procerum and L. britannicum, and the geographically highly restricted L. transwallianum, L. parvum and L. binervosum subsp. cantianum. Samples of several subspecies and varieties within L. procerum and L. britannicum have yet to be analysed, so this picture could change. It must be noted that the genetic support for these groups is very weak, and it is likely that the inclusion of further taxa - and additional samples from each taxon - will tend to make the groups less, rather than more, distinct (R. Cowan pers. comm.). However, on the basis of the preliminary AFLP analyses carried out so far, it appears that the ‘morpho-species’ of Ingrouille & Stace (1986) do show consistent groupings within the UPGMA tree, confirming the genetic closeness of the infraspecific taxa they each encompass - apart from L. binervosum s.s., where populations of the three subspecies so far analysed lie within three separate clusters. Further molecular work is being carried out, and the final UPGMA tree will doubtless look somewhat different to the one presented here. Clearly there is a need to push ahead and complete this research, ensuring that at least all the presently named species and subspecies are covered, including several samples from those that are widely distributed - e.g. L. procerum subsp. procerum , which has several varieties that can be readily distinguished on morphological grounds - and those having highly disjunct distributions (e.g. L. recurvum subsp. humile). Outlying populations of L. binervosum subsp. anglicum (e.g. from Lincolnshire) need to be investigated, and other L. binervosum sub-taxa that have been sampled need to be added to the analysis (i.e. subspp. saxonicum and mutation) - this is especially important given that the analysis to date indicates that the subspecies of L. binervosum s.s. are genetically no more closely related to each other than to other taxa in the aggregate. AN EYE TO THE FUTURE Once the molecular work has been finished, we can expect that a thorough review of the taxonomy will soon follow. This might lead to the disappearance of a few sub-taxa (e.g. varieties, and perhaps some subspecies, of L. procerum and L. britannicum ) and the ‘dismantling’ of L. binervosum s.s. As one of the B.S.B.I.’s referees for the genus Limonium, Martin Ingrouille (Birkbeck College, University of London) is actively engaged in determining specimens being sent to him; he is also supervising the molecular research being done at Kew by Robyn Cowan (Cowan et al. 1999), and is keen - as we are - that this work should ultimately contribute to a taxonomic overhaul of the L. binervosum group. Importantly, he and Robyn are also exploring the relationships between the British taxa and those in France, Spain and Portugal. We understand that Peter Sell (Cambridge) has been determining specimens ready for his new Flora, and that the other B.S.B.I. Limonium referee, Dr L. A. Boorman, has also been naming material. We do wonder whether the three of them share the same taxonomic views on the L. binervosum group - but that's another matter! After the various taxonomic issues have been resolved, we would advocate the publication of a definitive handbook to the sea-lavenders of the British Isles, including not only all those within the L. binervosum agg., but also L. vulgare, L. humile , L. belli difolium, L. auriculae -ursifolium, L. normannicum and the introduced L. hyblaeum. A start has already been made on the illustrations for this: over the last two years Fred Rumsey, with funding from English Nature and the Countryside Council for Wales, has been working on pen-and-ink illustrations of the British 222 ATLANTIC ARC FIGURE 2. Limonium recurvum C. E. Salmon subsp. recur\>um, Portland Bill, Dorset; drawn by F. J. Rumsey (2003). and Irish taxa - including all named species and subspecies, plus a few of the more distinct varieties. An example is given in Fig. 2. It is reassuring to know that, despite the current taxonomic uncertainties and our patchy knowledge of their distribution, our rock sea-lavenders are not under any serious threat - they mostly grow on coastlines that are already well protected, or at any rate well removed, from the pressures of 21st century life. Climate change/sea-level rise/increasing severity of storms may, of course, have a significant impact in the long term, but these fall outside the remit of those of us responsible for implementing the U.K. Biodiversity Action Plan for the ‘endemic Limonium taxa'. Whether we are botanical ‘lumpers’ or ‘splitters’, we can hopefully agree that the genus Limonium encompasses an array of extraordinarily beautiful plants - bringing a dash of colour in high summer to some of our most spectacular coastlines. They also bring a dash of endemism to our relatively impoverished post-glacial island flora, and for that reason alone we think they fully deserve the attention they are now receiving - from botanists and conservationists alike. LIMONIUM BINERVOSUM AGG. IN THE BRITISH ISLES 223 ACKNOWLEDGMENTS We would like to thank all those colleagues and fellow L/racm'wm-enthusiasts who have helped in various ways with this paper. Particular thanks to Martin Ingrouille, for so bravely taking on the Limonium binervosum species-complex back in the late 1970s/early 1980s; to Martin, Robyn Cowan and Mike Fay for permission to refer to the preliminary results of their genetic fingerprinting studies; to Lucy Cordrey, John Harvey and Andy Jones for their input to the UKBAP steering group (especially to Lucy for administering the group and arranging field workshops on Limonium in Pembrokeshire and Dorset); to Alex Lockton for collating the English and Welsh Limonium binervosum records; and to Fred Rumsey for the drawing depicted in Fig. 2. We are grateful to Robyn Cowan, too, for her detailed and helpful comments on an early draft of the taxonomic section. REFERENCES COWAN, R. S., Ingrouille, M. J. & Fay, M. F. (1999). Preliminary studies of rock sea lavenders using genetic fingerprinting. Birkbeck College, University of London, and Royal Botanic Gardens, Kew. Unpublished report to English Nature (Species Recovery Programme). Erben, M. (1993). Limonium Miller, in Castroriejo, S., Aedo, C., Cirujano, S., Lainz. M.. Montserrat, P., Morales. R.. Munoz Garmendia, F., Navarro, C., Paiva, J. & Soriano, C. eds. Llora Iberica. Vol. 2: 2-143. Real Jardfn Botanico, Madrid. INGROUILLE, M. J. & Stace, C. A. (1986). The Limonium biner\’osum aggregate (Plumbaginaceae) in the British Isles. Botanical Journal of the Linnean Society 92: 177-217. Leach, S. J. (1994). Limonium binervosum (G. E. Smith) Salmon sensu lato., in Stewart, A.. Pearman, D. A. Preston, C. D. eds. Scarce Plants in Britain. Joint Nature Conservation Committee. Peterborough. Leach, S. J. (2005). Recent surveys of endemic Limonium (Rock Sea-lavender) taxa in S. Devon (v.c. 3), in Leach, S. J., Page, C. N„ Peytoreau, Y. & Sanford, M. N. eds. Botanical Links in the Atlantic Arc , pp. 225-229. Botanical Society of the British Isles, London. Leach, S. J., Pearman, D. A., Cordrey, L. & Jones, A. (2001). Putting rock sea-lavenders on the map. B. S.B.I. News 88: 49-51. Lledo, M. D.. Crespo, M. B., Fay, M. F. & Chase, M. W. (in press). Molecular phylogenetics of Limonium and related genera (Plumbaginaceae): biogeographical and systematic implications. American Journal of Botany. LOCKTON, A. (2003). Maps of the Limonium binervosum agg. in England. Unpublished B.S.B.I. report to English Nature. Pignatti, S. (1972). Limonium, in T. G. Tutin, V. H. Heywood, N. A. Burges. D. A. Valentine, S. M. Walters. D. A. Webb & D. M. Moore eds. Flora Europaea. Vol. 3: 38-50. Cambridge University Press, Cambridge. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. Preston, C. D., Pearman, D. A. & Dines, T. D. eds. (2002). New Atlas of the British & Irish Flora, pp. 201-208. Oxford University Press, Oxford. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. UK BIODIVERSITY Group (1998). Tranche 2 Action plans. Volume 1. Vertebrates and vascular plants. English Nature. Peterborough. WlGGlNTON, M. J. ed. (1999). British Red Data Books. 1. Vascular Plants. 3rd ed. Joint Nature Conservation Committee, Peterborough. 225 Recent surveys of endemic Limonium (Rock Sea-lavender) taxa in S. Devon (v.c. 3) S. J. LEACH English Nature, Roughmoor, Bishop’s Hull, Taunton, Somerset, TA1 5AA, England * ABSTRACT Recent surveys of Limonium binerx'osum subsp. mutatum and L. britannicum subsp. coombense in S. Devon (v.c. 3) have added considerably to our knowledge of these endemic taxa. Both appear to be more frequent along the South Hams coastline than previously thought. A third, as yet unnamed, taxon is reported from the east side of the Salcombe Estuary. KEYWORDS: apomicts, conservation, distribution, population counts. INTRODUCTION The South Hams coastline of Devon, particularly that between the Salcombe Estuary and Start Point, is a well-known stronghold for several rare Oceanic Temperate species (e.g. Rumex rupestris and Festuca longifolia ), but the importance of its rock sea-lavenders (. Limonium binerx’osum agg.) is perhaps less widely appreciated. Yet two English endemics occur here (Ingrouille & Stace 1986; Wigginton 1999): L. britannicum subsp. coombense, thought to be restricted to a handful of localities along the south coast of E. Cornwall (v.c. 2) and S. Devon (v.c. 3); and L. binervosum subsp. mutatum, known until recently from just a single station, at Lannacombe (GR SX8037), and supposedly nowhere else in the world. Recent survey work - part of an ongoing investigation of this apomictic complex in the U.K. (Leach et al. 2001) - has greatly improved our state of knowledge of these two taxa in S. Devon. Also, we have found what appears to be a third taxon that closely resembles the Welsh endemic L. parvum. Records for the three taxa are summarised below. LIMONIUM BINERVOSUM SUBSP. MUTATUM In 1999 a dozen plants of putative subsp. mutatum (Fig. 1. since confirmed by M. J. Ingrouille, B.S.B.I. referee for the genus Limonium ) were discovered about 7 kms to the west of Lannacombe, at the mouth of the Salcombe Estuary at Limebury Point (SX737376). Then, in 2001-2002 we examined some of the more accessible Limonium colonies along this stretch of coastline, in particular those within the Prawle Point and Start Point Site of Special Scientific Interest (S.S.S.I). Initially, we located three discrete sub-populations of subsp. mutatum at Lannacombe (perhaps 80- 100 plants) - then, working eastwards along Harris’s Beach (SX8037 & SX8136) we turned up at least a further 23 colonies on the head cliffs, several of which contain hundreds of plants, and possibly 1,500-2,000 plants in total. Further east still, at Mattiscombe (SX8 17369), there is another large colony (200-300 plants), along with two small populations between there and Peartree Cove (SX8 18369). With the most recent field data to hand (including that gathered in 2003-2004), we now know that subsp. mutatum is much more abundant (and widespread) than originally thought, with the core area comprising about 30 sub-populations or discrete colonies - and a total of more than 2,500 plants - strung out along at least 2 kms of coast within the S.S.S.I., plus a single tiny outlying population on the Salcombe Estuary. *email: simon.leach@english-nature.org.uk 226 ATLANTIC ARC Srlfcies; Limmium bincrvosum subsp. mttatum Locality: 'Harm's '! Colony ’; some plants very tad and wed-branched (SO- 70cm) Habitat: Head cliffs - all plants growing on open exposures of head, c.S-ISm above ' lie Determiner; FIGURE 1. Limonium binervosum subsp. mutatum. Specimen collected 15 August 2001, cliffs behind Harris’s Beach (SX80573707). £1 coin (22 mm diameter) for scale. ENDEMIC LIMONIUM TAXA IN S. DEVON 227 Species; Limonium Britannicum suSsp. coomBense (var?) Locality: Cjcra 'dgff. VC: S .'Devon (3) Grid Ref: C.SX7S4369 — colony extends from SX7 5423685 west wards as far as SX75153700 Date: 13th Jdugust 200! Recorder: S-J- Leach Population Size: c. 7000-10. 000 plants, mainly gro wing high up the cBffs and so difficult to estimate Habitat: _ ‘Head cliffs - on near-vertical and vertical cBffs, along predominantly south-facing section of coast where the head has suffered several major slips in recent years Associated Species: (Pfkntago mafitima, pd rmeria maritime, Spergularia mpicoCa, Festuca ruSra. Jit this site Cnthmum is not an associate. Being restricted to the underlying schist and stafiii (well- vegetated) head Determiner: S-J. Leach Figure 2. Limonium britannicum subsp. coombense. Specimen collected 13 August 2001, Gara Rock (SX754369). £1 coin (22 mm diameter) for scale. 228 ATLANTIC ARC LIMONIUM BRITANNICUM SUBSP. COOMB ENSE This taxon (Fig. 2) is also now known to be considerably more widespread than previously thought, with an exceptionally large colony of 7,000-10,000 plants found on the head cliffs at Gara Rock (SX7536) in 1999-2000, then several colonies between there and Elender Cove (SX767357), and further populations at Prawle Point (SX774352 - Copstone Cove, Ingrouille’s type locality for his var. grandicaule), Ivy Cove (SX799370), and between Mattiscombe and Peartree Cove with L. biner\>osum subsp. mutatum - ten localities so far, and with several promising sections of coastline still to be searched. All currently known sites lie within the Prawle Point and Start Point S.S.S.I., and are not under any significant threat - although several have been affected by cliff land-slips and slumps due to storms and the recent run of exceptionally wet winters. Figure 3. Voucher specimens of a ‘mystery’ Limonium, collected 20 August 2002, Biddlehead Point, Salcombe Estuary (SX738381). £1 coin (22 mm diameter) for scale. A ‘MYSTERY- LIMONIUM AT BIDDLEHEAD POINT In 2002 we found a second Limonium population on the east side of the Salcombe Estuary, on land owned by the National Trust at Biddlehead Point (SX738381). The plants here are quite unlike any we have seen elsewhere - much smaller than L. britannicum subsp. coombense , with cushions of tiny petiolate leaves (mainly <10 mm long and just c. 3-5 mm wide), slender flowering stems (0-5-1 mm diameter), very few sterile branches, and short spikes with relatively few spikelets per spike. While having the general jizz of a small ‘ britannicum ’ it is not nearly robust enough to be subsp. coombense. M. J. Ingrouille has indicated that voucher specimens (Fig. 3) resemble L. parvum , an endemic species restricted to Carboniferous limestone cliffs in Pembrokeshire. Leaf samples have been collected for Kew, along with samples of L. binervosum subsp. mutatum and L. britannicum subsp. coombense , and it is hoped that molecular analysis now being funded by English Nature will help to determine whether the mystery Limonium is indeed L. parvum , or a new (as yet undescribed) taxon - or merely an odd-looking L. britannicum. ENDEMIC LIMONIUM TAX A IN S. DEVON 229 REFERENCES INGROUILLE, M. J. & STAGE, C. A. (1986). The Limonium binervosum aggregate (Plumbaginaceae) in the British Isles. Botanical Journal of the Linnean Society 92: 177-217. Leach, S. J., Pearman, D. A., Cordrey, L. & Jones, A. (2001). Putting rock sea-lavenders on the map. BSBI News 88: 49-51. WlGGINTON, M. J. ed. (1999). British Red Data Books, 1 Vascular Plants. 3rd ed. Joint Nature Conservation Committee, Peterborough. 231 Vegetation and habitats of the western European endemic Asparagus prostratus Dumort. (Liliaceae), Wild Asparagus T. C. G. RICH Department of Biodiversity and Systematic Biology, National Museums & Galleries of Wales, Cardiff, CF10 3NP, Wales * L. K. RICH 67 Heol Uchaf, Rhiwbina, Cardiff, CF14 6SR, Wales S. B. EVANS, A. E. EVANS Glan-y-mor, Dinas Cross, Newport, Pembrokeshire, SA42 OUQ, Wales and F. HOPKINS Le petit Monthelon, 35 690 Acigne, France ABSTRACT The vegetation types of Asparagus prostratus Dumort. (Wild Asparagus) are described from 94 quadrats collected throughout its range in Europe. 226 species were recorded associated with A. prostratus, with a mean of 13 species per quadrat. It generally occurred close to the sea (mean altitude 18 m), on slopes (mean 19°) of all aspects, in short (mean vegetation height 23 cm), open vegetation (mean cover 79%). The most frequent associates were Festuca rubra and Daucus carota. Eight main vegetation groups were recognised from a TWINSPAN analysis. There were four sea-cliff communities: Carpobrotus edulis patches, maritime grasslands of the Crithmo maritimi-Armerion maritimae, spray-zone Crithmo maritimi-Limonion binervosi vegetation, and rocky cliff vegetation on shallow soils of the Thero-Airion or possibly Saginion maritimae. It occurs in three sand dune vegetation types: classic, widespread Ammophilion arenariae, dynamic combined Ammophilion and Euphorbio portlandicae-Helichrysion staechadis, and somewhat nitrophilous disturbed dunes probably belonging to the Corynephorion canescentis. It also occurs in dune scrub of the Hippophaeo- Ligustretum. Overall it occurs in a surprisingly broad range of coastal vegetation types and its rarity is probably not a consequence of lack of habitat or restriction to a specific vegetation type. Keywords: Asparagus officinalis subsp. prostratus, British plant communities, Belgium, England, Ireland, France, Netherlands, Spain, Wales. INTRODUCTION Asparagus prostratus Dumort. (Asparagus officinalis L. subsp. prostratus (Dumort.) Corbiere), Wild Asparagus (Plate 28), is a western European endemic, scattered on parts of the coasts of Belgium, Britain, Channel Islands, France, Germany, Ireland, Spain and the Netherlands (Kay et al. 2001). Across most of its range, apart from the north coasts of Spain and very locally in France, it is rare or declining. In Spain it is widespread along cliffs on the north coast, more rarely on sand dunes, and in some areas is locally abundant. In France it is scattered from Vendee to Finistere, *e-mail: tim.rich@nmgw.ac.uk 232 ATLANTIC ARC where it is characteristic of dunes and rocks on the islands off the south coast, coastal dunes (especially the older leached grassy grey dunes) and also of rocky granite promontories with blown sand. In Britain it is a Red Data Book species, classified as the I.U.C.N. category ‘Vulnerable’ (Wilkinson 1999), and is a Priority Biodiversity Species (U.K. Biodiversity Group 1998). In England, it occurs on sea cliffs in Cornwall and Dorset, sandy waste ground and also rarely on shingle. In Wales it is found on sandy sea-cliffs in S. Wales, and formerly on dunes on Anglesey (Rich et al. 2002). In the Republic of Ireland it is a protected species, and has been recorded recently at only six sites, all on sand dunes (Curtis & McGough 1988). In Belgium it occurs on dunes and in scrub, and very rarely in more mesotrophic grassland a few kilometres inland. It has declined locally and is being considered for inclusion in a revision of the Belgium Red List (W. van Landuyt, pers. comm. 2002). It is widespread around the coast of the Netherlands on dunes and islands in about 20-30 sites, and again occurs very rarely inland. It is a Red List species regarded as ‘Actually not threatened’ (I.U.C.N. = ‘Low risk/Least concern’) though it has declined on the northern islands (Mennema et al. 1980, van der Meijden et al. 2000). In Germany, it has only been recorded on two North Sea islands (Rich 2002), and may be extinct (H. Kuhbier, pers. comm. 2002). In order to provide background information for the conservation of A. prostratus in Europe as a whole, we have been studying its taxonomy, distribution, genetics, reproductive biology and ecology throughout its range since 1995. In this paper, we characterise its vegetation types and habitats. METHODS A total of 94 representative 2 m x 2 m quadrats centred on plants of Asparagus prostratus were recorded from sites throughout its extant range (Fig. 1). The data were largely collected by T. and L. Rich during family holidays, though data have also been collected during museum fieldwork in Belgium and the Netherlands. Information from 16 quadrats in Spain and France was also collected by S. and A. Evans in 2002. Data from seven quadrats in Cornwall were kindly provided by A. J. C. Malloch from the British Plant Communities database (pers. comm. 1999). The numbers of quadrats from each country, approximately reflecting its local frequency, were as follows: Belgium 6, Netherlands 4, England 21, France 16, Ireland 7, Spain 31 and Wales 8. Copies of the original data have been deposited in the library of the National Museums & Galleries of Wales (Evans & Evans 2003; Rich 1998, 2000; Rich & Rich 1998a, b, c, 1999a, b, 2001, 2003; Rich et al. 2002). Associated species were recorded following standard British recording methods using the 10- point Domin scale of cover- abundance (see Rodwell 1991-2000). Physical variables such as % vegetation cover, height of vegetation, altitude, aspect and slope were also recorded. Bryophytes and lichens were inconsistently recorded, and were excluded from the main descriptions of the vegetation types. A very few vegetative species in Spanish quadrats proved impossible to name and have not been included. Vegetation data were handled in the VESPAN III package (Malloch 1999) using ‘standard’ runs for the TWINSPAN analyses. For the analysis, species groups such as Anthyllis vulneraria, Bromus hordeaceus, Erodium cicutarium, Limonium binervosum and Poa pratensis were treated in a broad sense, although it is known that more than one taxon was involved in each case. European phytosociological names have been updated from the out-dated conspectus in Volume 5 of Rodwell (1991-2000). Soil pH measurements were made on air-dried samples mixed 50:50 with distilled water. RESULTS A total of 226 vascular plants, bryophytes and lichens were recorded associated with A. prostratus in the 94 quadrats. The altitude ranged from 1-75 m (mean 18 m), and most populations occurred on slopes (mean 19°, range 0-80°) of all aspects, though they were marginally more frequent on south-facing slopes. The mean herb height was 23 cm (range 3-100 cm) and the mean vegetation VEGETATION OF ASPARAGUS PROSTRATUS 233 FIGURE 1. Distribution of Asparagus prostratus (O) and location of quadrat samples (•). cover was 79% (range 5-100%) usually with some bare soil, sand or rock. Overall, the most frequent associates were Festuca rubra (73% of quadrats), as also reported by Davies (1961), and Daucus carota (45% of quadrats), with Armeria maritima, Agrostis stolonifera, Lotus corniculatus, Carex arenaria, Anthyllis vulneraria, Dactylis glornerata, Plantago lanceolata, Scilla verna , Sonchus oleraceus and Plantago coronopus all occurring in 20-32% of quadrats. 125 species (55%) were recorded only once. An average of 12*7 + (F54 s.e. species were recorded per quadrat. After revision and evaluation, eight vegetation groups were accepted from the TWINSPAN analysis. The four main divisions are shown in Fig. 2, and the groups are described in more detail below, listing only species in each group occurring at constancy II or above (i.e. occurring in >20% of the quadrats in each group). Group 1 includes two samples from Spain from a cliff site where the alien Carpobrotus edulis formed dense carpets with occasional A. prostratus and Brachypodium pinnatum (Table 1). Carpobrotus often swamps all other species, but the growth form of Asparagus allows it to grow through the mats, albeit at probably reduced frequency. Asparagus has also been observed in this vegetation type near Cadgwith, Cornwall. Group 2 includes a range of general maritime grasslands on cliffs in Spain and England with grassland characterised by Agrostis stolonifera, Armeria maritima, Dactylis glornerata, Daucus carota, Festuca rubra, Lotus corniculatus and Scilla verna (Table 2). This vegetation is usually located above the spray zone and often has calcareous soils derived from limestone or serpentine. This group belongs to the Crithmo maritimi-Armerion maritimae of the Asteretea tripolium. Group 3 comprises the lower rocky cliff vegetation near the sea in Spain, France and England, characterised by Armeria maritima, Crithmum maritimum, Daucus carota, Festuca rubra and Inula crithmoides (Table 3). This group belongs to the Crithmo maritimi-Limonion binervosi of the Crithmo-Staticetea. 234 ATLANTIC ARC With Daucus, Festuca rubra, Armeria, Agrostis stolonifera(ll) With Carex arenaria, Geranium molle(\ 7) WithDc/wcws(54) With Ammophila, Galium, Carex (21) With Lagurus, Eryngium (8) (9) Sea cliffs (4 groups) Dunes (1 group) Brittany dunes Disturbed dunes and ( 1 group) scrub (2 groups) FIGURE 2. Major TWINSPAN divisions of Asparagus prostratus quadrat data, showing indicator species for splits and number of quadrats in brackets. Group 4 comprises rocky cliff vegetation on shallow, free-draining soils, often in crevices and hollows with skeletal soils from Spain, France, England and Wales with Anthyllis vulneraria, Armeria maritima, Bromus hordeaceus, Catapodium marinum, Festuca rubra and Plantago coronopus (Table 4). It belongs either to the Thero-Airion of the Helianthemetea guttati or the Saginion maritimae of the Saginetea maritimae. It often occurs above the Crithmo maritimi- Limonion binemosi and is transitional to the Crithmo maritimi-Armerion maritimae. Group 5 comprises sand dune vegetation from Spain, France, Wales, Ireland, Belgium and the Netherlands, with Ammophila arenaria, Anthyllis vulneraria, Carex arenaria, Festuca rubra, Galium verum, Hypochaeris radicata and Poa humilis (Table 5). It is classic Ammophilion arenariae of the Ammophiletea arenariae. Group 6 forms a tighter group of exposed, often grazed or disturbed coastal and island dunes in Brittany, France, with Aetheorhiza bulbosa, Carex arenaria, Catapodium marinum, Eryngium maritimum, Lagurus ovatus and Phleum arenarium (Table 6). This group is probably transitional between Ammophilion and Euphorbio portlandicae-Helichrysion staechadis , with some Laguro ovati-Bromion rigidi overlaying it (the latter typical of disturbed dunes). In Brittany this group is typical of the Thymo drucei-Helichrysetum staechadis as well as the Roso spinosissima- Ephedretum distachyae. Group 7 occurs on somewhat nitrophilous disturbed or older grey dunes in France, Belgium and the Netherlands with Carex arenaria, Erodium cicutarium, Festuca rubra, Geranium molle, Plantago lanceolata, Sedum acre and Senecio jacobaea (Table 7). These probably relate to the Corynephorion canescentis of the Koelerio-Corynephoretea. Group 8 is a small heterogeneous group characteristic of developing scrub in Belgium and the Netherlands with Rubus caesius , Arrhenatherum elatius and Carex arenaria (Table 8), and with a range of other woody species such as Hippophae rhamnoides, Ligustrum vulgare and Rosa spp. This belongs to the Hippophaeo-Ligustretum. It is in this community that a more erect form of A. prostratus may have evolved through local selection of genotypes which can compete with taller scrub vegetation. VEGETATION OF ASPARAGUS PROSTRATUS 235 TABLE 1. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 1 (N = 2) Mean altitude, m 30 Mean angle of slope, degrees 30 Mean vegetation height, cm 18 Mean vegetation cover, % 100 Asparagus prostratus V Brachypodium pinnatum V Carpobrotus edulis V Crithmum maritimum m Daucus carota hi Festuca rubra hi Holcus lanatus m Species with constancy II occur in 20T-40% of quadrats of the group. III in 401-60%, IV in 60T-80% and V in 801-100%. TABLE 2. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 2 (N = 37). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 25 25 23 82 Asparagus prostratus V Daucus carota V Festuca rubra V Agrostis stolonifera III Anneria maritima III Dactylis glomerata III Lotus comiculatus III Scilla verna III Anthyllis vulneraria II Holcus lanatus II Plantago lanceolata n Plant ago maritima ii Silene uniflora ii Sonchus oleraceus h TABLE 3. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 3 (N = 9). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 11 25 22 41 Asparagus prostratus V Crithmum maritimum IV Festuca rubra IV Armeria maritima m Daucus carota hi Galium mollugo hi Inula crithmoides ii Plantago maritima ii Sonchus oleraceus n 236 ATLANTIC ARC TABLE 4. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 4 (N = 9). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 11 24 6 68 Asparagus prostratus V Plantago coronopus V Armeria maritima IV Anthyllis vulneraria III Bromus hordeaceus III Catapodium marinum III Festuca rubra subsp. III Beta vulgaris maritima II Cerastium diffusum II Cochlearia danica II Crithmum maritimum II Daucus carota II Erodium cicutarium agg. II Herniaria ciliolata II Hypochaeris radicata II Leontodon saxatilis II Limonium binervosum agg. II Sagina maritima II Spergularia rupicola II Trifolium occidental II Trifolium striatum II TABLE 5. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 5 (N = 20). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 15 11 29 86 Asparagus prostratus V Festuca rubra V Ammophila arenaria IV Anthyllis vulneraria m Car ex arenaria hi Galium verum hi Hypochaeris radicata hi Poa humilis hi Agrostis stolonifera ii Armeria maritima ii Calystegia soldanella ii Dactylis glomerata ii Euphorbia portlandica ii Holcus lanatus ii Leontodon saxatilis ii Lotus comiculatus ii Plantago lanceolata ii Sonchus oleraceus n VEGETATION OF ASPARAGUS PROSTRATUS 237 TABLE 6. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 6 (N = 8). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 7 16 8 95 Asparagus prostratus V Lagurus ovatus V Eryngium maritimum IV Carex arenaria IV Aetheorhiza bulbosa III Catapodium marinum III Phleum arenarium III Allium sphaerocephalon II Anisantha diandra II Bromus hordeaceus II Cerastium diffusion II Crepis vesicaria II Elytrigia juncea II Ephedra distachya II Euphorbia portlandica II Geranium mode II Geranium purpureum II Helichrysum stoechas II Herniaria ciliolata II Mibora minima II Plantago coronopus II Polycarpon tetraphyllum II Raphanus maritimus II Senecio vulgaris II Torilis nodosa II Vulpia fasciculata II Several of the distinctive vegetation types noted in the field are not separated in the TWLNSPAN analysis, possibly due to their relative rarity. In Spain and England it occurs with Schoenus nigricans on dry sea cliff-slopes. In Belgium it occurs in species-rich mesotrophic grassland inland, admittedly on old dunes, but with scarcely any maritime associates, and in the Netherlands it grew for a while on an inland railway where it was probably introduced with sand ballast (J. van Alphen, pers comm. 2000). In Spain, England and Wales it occurs on the edge of, or rarely in, Ulex scrub, and sometimes in the associated Ericaceous heathlands too. At Portland, England it occurs on the top edge of Atriplex portulacoides saltmarsh grading into Triplenrospermum maritimnm/Carex arenaria/Elytrigia juncea shingle vegetation on the ballast edge of an old railway line - it is not clear if the Asparagus has grown into the Atriplex or the Atriplex has spread above the line of the highest tides. At this site, and also on sheltered cliffs in Spain, its roots may exceptionally be submerged by high or storm tides, but in general it prefers free-draining soils above the tidal limit and appears highly drought-tolerant. The 14 vegetation types in which A. prostratus occurs in Britain are listed in Table 9 following British Plant Communities (Rodwell 1991-2000). In Cornwall, it occurs in a range of maritime cliff communities, rarely extending into Prunus spinosa scrub. In Wales the sites are mixed maritime cliff and sand dune communities (the latter are essentially developed over wind-blown sand on cliffs, rather than on dunes), persisting rarely within invasive Ulex europaeus . scrub. At Portland (Dorset) it occurs in Atriplex portulacoides saltmarsh. It is tempting to relate the European quadrats to Rodwell’ s vegetation types but, as pointed out by Dirkse (1998), the validity of communities is usually limited to the areas from which the samples were derived. 238 ATLANTIC ARC TABLE 7. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 7 (N = 6). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 16 4 8 95 Asparagus prostrates V Car ex arenaria V Erodium cicutarium agg. V Geranium molle V Senecio jacobaea V Festuca rubra IV Plantago lanceolata IV Sedum acre IV Arenaria serpyllifolia III Bromus hordeaceus III Crepis capillaris III Koeleria macrantha III Leontodon saxatilis III Poa humilis III Cerastium ar\’ense II Conyza canadensis II Diplotaxis tenuifolia II Phleum arenarium II Rubus caesius II Taraxacum sp. II Soil pH measurements from 19 sites are given in Table 10. The average soil pH is 7-3, suggesting it is predominately a calcicole, but pH values as low as 5-0 have also been recorded. Salt spray often raises the pH of soils above that of the natural soil components. DISCUSSION Asparagus prostratus grows in a surprisingly broad range of sea-cliff, sand dune and other coastal vegetation types, and many of these are widespread around the coast of Europe. This leads us to consider that its overall rarity cannot be due to lack of habitat or restriction to a specific vegetation type, but is more likely to relate to other biological factors such as the chances of seeds being dispersed to the right niche by birds or mammals, or the availability of specific niches within the vegetation for regeneration. It is difficult to pin-point such niches, but it usually grows in open vegetation, often where there is some bare sand or rock crevices which may be essential for establishment of seedlings. We have observed seedlings in these kinds of situations in many of the larger populations (e.g. Bennallick 2001; Evans & Evans 2003). The creeping rhizomes it may produce have huge root balls 70 cm or more across. Once established it can persist for many years, perhaps even after the original vegetation has changed (e.g. calcareous dunes become leached, or cliff vegetation becomes scrubby). It is uncommon in rank, closed grasslands, and sites with such vegetation tend to have a few large old plants and no regeneration. At some sites it is clearly associated with bird perches. It is surprising how uniform and recognisable most of the European coastal vegetation types are across the range from Spain to the Netherlands. Perhaps the most distinct of these were the species-rich dunes in Brittany which formed their own group. The main TWINSPAN divisions are largely related to habitat and not geographical location. Most of the vegetation types in which it occurs are ungrazed or lightly grazed. Grazing may, however, be an important determinant of its local frequency, either though direct consumption of the fronds, through breakage of the brittle stems or by creation of open patches for regeneration. Many of the Spanish sites are extensively VEGETATION OF ASPARAGUS PROSTRATUS 239 TABLE 8. SPECIES CONSTANCY DATA FOR ASPARAGUS PROSTRATUS VEGETATION TYPE GROUP 8 (N = 3). FOR CONSTANCIES, SEE TABLE 1. Mean altitude, m Mean angle of slope, degrees Mean vegetation height, cm Mean vegetation cover, % 23 9 70 100 Asparagus prostratus V Rubus cae sius V Arrhenatherum elatius IV Carex arenaria IV Anisantha sterilis II Bromus hordeaceus II Calamagrostis epigejos II Diplotaxis tenuifolia II Elytrigia juncea II Elytrigia repens II Festuca rubra II Galium aparine II Galium mollugo II Helictotrichon pubescens II Hippophae rhamnoides II Hordeum murinum II Ligustrum vulgare II Poa humilis n Quercus robur h Rosa canina h Rosa pimpinellifolia ii Rosa rubiginosa ii Senecio sylvaticus h Urtica dioica n Viola tricolor subsp. curtisii ii TABLE 9. BRITISH PLANT COMMUNITIES (RODWELL 1991-2000) VEGETATION TYPES RECORDED FOR ASPARAGUS PROSTRATUS IN BRITAIN H7b Calluna vulgaris-Scilla verna heath, Viola riviniana sub-community MClb Crithmum maritimum-Spergularia rupicola maritime rock crevice community, Inula critlvnoides sub-community MC5b Armeria maritima-Cerastium diffusum maritime therophyte community, Anthyllis vulneraria sub-community MC8a Festuca rubra-Armeria maritima maritime grassland, typical sub-community MC9b Festuca rubra-Holcus lanatus maritime grassland, Dactylis glomerata sub-community MCI lb Festuca rubra-Daucus carota ssp. gummifer maritime grassland, Ononis repens sub-community SD6e Ammophila arenaria mobile dune community, Festuca rubra sub-community SD6f Ammophila arenaria mobile dune community, Poa pratensis sub-community, SD7a Ammophila arenaria -Festuca rubra dune community, typical sub-community SD7c Ammophila arenaria -Festuca rubra dune community, Ononis repens sub-community SD8a Festuca rubra-Galium verum fixed dune community, typical sub-community SD9 Ammophila arenaria-Arrhenatherum elatius dunes SMI 4a Halimione portulacoides salt-marsh, Halimione portulacoides sub-community W22 Prunus spinosa-Rubus fmticosus under-scrub, indeterminate sub-community W23 Ulex europaeus-Rubus fruticosus under-scrub, indeterminate sub-community 240 ATLANTIC ARC TABLE 10. SOIL PH MEASUREMENTS FROM ASPARAGUS PROSTRATUS SITES Locality Soil pH Cadgwith, England 6-0 Kynance, England 60 Anse de Dinan, France 8-4 Goulven, France 8-5 lie de Noirmoutier, France 8-1 lie D’Oleron, France 7-6 La Faute-sur-Mare, France 8-7 Pointe de la Torche, France 7-8 Port de Bestree, France 5-0 Ajo, Spain 8-3 Barayo, Spain 5-6 Bermeo, Spain 7-3 Faro San Juan de Nieva, Spain 7-1 Isla de La Virgen del Mar, Spain 7-9 Moniello, Spain 7-3 Praia de Morouzos, Spain 8-0 Punta da Estaca de Beres, Spain 6-5 San Juan de La Canal, Spain 6-2 Santona, Spain 7-9 grazed by cattle or horses, sometimes with winter burning to control scrub. In France and Belgium a few sites are grazed by stock, but rabbits can cause close cropping of the grasslands though they do not seem to eat A. prostratus itself. In Ireland, the very large population on the dunes of Ballyteige Burrows is extensively grazed by cattle. The Dutch, English and Welsh sites are ungrazed except by rabbits, but part of one Welsh site is grazed by sheep and cattle. As grazing tends to result in the sward being opened up, some grazing may be beneficial for regeneration, but it is usually absent from heavily grazed sites. In the Netherlands, and to a lesser extent Belgium and France, it was clearly behaving as a nitrophile in disturbed sand dune areas. At Coepelduynen, Netherlands its prime habitat was the disturbed edges of dune scrub composed of bird-sown, woody species such as Hippophae, Ligustrum, Rosa and Euonymus. Other Dutch sites are associated with places where fishermen clean their nets (J. van Alphen, pers. comm. 2000). This contrasts with the typical pristine, nutrient-poor sea-cliff and dune habitats in Spain, Britain and Ireland, but may indicate that disturbance and nutrient-enrichment associated with human activities is not necessarily prejudicial to its survival. Asparagus prostratus usually occurs within 100-200 m of the sea, but can persist inland. Although clearly highly tolerant of salt spray, it is presumably either sensitive to climatic factors such as frost (though it can survive inland in gardens through cold winters), or to edaphic factors, perhaps through competition with other species in the absence of significant soil salinity. It is probably most abundant on calcareous soils and rocks such as limestone or shell-sand (Table 10), but it is not confined to such soils, and is known to also occur directly on serpentine, sandstones, slates, shales and occasionally granite, or in the soils derived from these rock types. Although it can persist in patches of the alien Carpobrotus edulis, it may be difficult for it to regenerate due to the dense cover of stems and leaves. It may also be threatened by the spread of alien plants on coastal habitats in general; these should be controlled, not only for the Asparagus but also for the many other rare and special western European species. AC KNOWLEDGMENTS We would like to thank Jacques van Alphen, Ian Bennallick, Ma Femandez-Carvajal, Elizabeth Davies, Andy Karran, Quentin Kay, Wouter Van Landuyt, Andrew Malloch, Alan Orange and Guido Rappe. VEGETATION OF ASPARAGUS PROSTRATUS 241 REFERENCES BENNALLICK, I. J. (2001). Survey of Asparagus prostratus Dumort. in Cornwall 2001. Unpublished contract survey report to National Trust/BSBI. CURTIS, T. G. F. & Mcgough, H. N. (1988) The Irish Red Data Book. Vol. 1. Vascular Plants. Dublin: Stationery Office. Davies, E. W. (1961). The ecology and distribution of Asparagus officinalis subsp. prostratus. Proceedings of the Botanical Society of the British Isles 4: 211. DlRSKE, G. M. (1998). The validity of general purpose flora-based classification of vegetation. IBN Scientific Contributions 14. DLO Institute for Forestry and Nature Research (IBN-DLO), Wageningen. EVANS, S. B. & A. E. (2003). Lusitanian Tour of Europe May 2002. Asparagus prostratus on the coasts of northern Spain and the Atlantic coast of France as far north as Normandy. Unpublished report to National Museums & Galleries of Wales. Kay, Q. O. N., Davies, E. W. & T. C. G. Rich (2001). Taxonomy of the western European endemic Asparagus prostratus (A. officinalis subsp. prostratus ) (Asparagaceae). Botanical Journal of the Linnean Society' 137: 127-137. MALLOCH, A. J. C. (1999). VESPAN III. A computer package to handle and analyse multivariate species data and handle and display species distribution data. 3rd edition. University of Lancaster, Lancaster. Mennema, J., Quene-Boterenbrood. A. J. & Plate, C. L. (1980). Atlas of the Netherlands flora. 1. Extinct and very rare species. W. Junk, The Hague. Rich, L. K. & Rich, T. C. G. (1998a). Distribution and ecology of Asparagus prostratus. 7. Cantabria, Northern Spain, May 1997. Unpublished report. RICH, L. K. & Rich, T. C. G. (1998b). Distribution and ecology o/Asparagus prostratus. III. Brittany, France, May 1998. Unpublished report. RICH, L. K. & Rich, T. C. G. (1998c). Distribution and ecology of Asparagus prostratus. TV. South-east Ireland, Autumn 1998. Unpublished report. RICH, L. K. & Rich, T. C. G. (1999a). Distribution and ecology of Asparagus prostratus. V. Ashirias, Spain, May 1999. Unpublished report. RICH, L. K. & Rich, T. C. G. (1999b). Distribution and ecology of Asparagus prostratus. VI. Wales, 1999. Unpublished report. RICH, L. K. & Rich. T. C. G. (2001). Distribution and ecology of Asparagus prostratus. IX. Brittany, France, 2001. Unpublished report. Rich, L. K. & Rich, T. C. G. (2003). Distribution and ecology of Asparagus prostratus. VIII. England, 1999- 2002. RICH, T. C. G. (1998). Distribution and ecology of Asparagus prostratus. II. Northern France, June 1997. Unpublished report. RICH, T. C. G. (2000). Distribution and ecology of Asparagus prostratus. VII. Tlie Netherlands 2000. Unpublished report. National Museums & Galleries of Wales. Rich, T. C. G. (2002). Occurrence of the Western European endemic Asparagus prostratus (A. officinalis subsp. prostratus ) (Asparagaceae) in Germany. Eloristische Rundbriefe 32: 65-67. Rich, T. C. G., Bennallick, I. J.. Cordrey, L., Kay, Q. O. N., Lockton, A. & Rich, L. K. (2002). Distribution and population sizes of Asparagus prostratus Dumort., Wild Asparagus, in Britain. Watsonia 24: 183-192. RICH, T. C. G., Rappe, G. & Landuyt, W. Van (2002). Distribution and ecology of Asparagus prostratus. X. Belgium 2002. Unpublished report. National Museums & Galleries of Wales. Rodwell, J. S. ed. (1991-2000). British Plant Communities. Vols. 1-5. Cambridge University Press, Cambridge. Van Der Meijden, R.. Ode, B., Groen, K. L. G., Witte, F. M. & Bal, D. (2000). Endangered and vulnerable vascular plants in the Netherlands. Basic report with proposal for the Red List. Gorteria 26(4): 85-208. UK BIODIVERSITY Group (1998). Tranche 2 Action plans. Volume 1. Vertebrates and vascular plants. English Nature, Peterborough. WILKINSON, L. K. (1999). Asparagus prostratus Dumort. (Liliaceae), in WlGGINTON, M. J. ed. British Red Data Books 1. Vascular plants. 3rd ed. Joint Nature Conservation Committee, Peterborough. 243 A strategy for the conservation of Genista pilosa L. (Hairy Greenweed) in Picardie, France P. SALIOU Centre Regional de Phytosociologie/Conser\’atoire Botanique National de Bailleul, Hameau de Haendries, 59270 Bailleul, France*1 and J-C. HAUGUEL Conservatoire des Sites Naturels de Picardie, 1, place Ginkgo, Village OASIS Diary, 80044 Amiens Cedex 1, France* ABSTRACT Genista pilosa L. (Hairy Greenweed) is a medio-European species, with the north-western edge of its distribution in the extreme south of Britain and in Picardie, one of three regions in France where the plant is legally protected. This note explains how the Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul and the Conservatoire des Sites Naturels de Picardie are developing a strategy for the safeguard of G. pilosa in Picardie. Ke WORDS: distribution (Europe. France. Britain), conservation. DISTRIBUTION Genista pilosa L. (Hairy Greenweed) (Plate 29) is a medio-European species, with the western/ north-western boundary of its European distribution in the south and west of Britain and western France (Fig. 1, reproduced from Meusel et al. (1965)). The species is absent from Brittany and from much of the north of France (Lambinon et al. 1999), and is rare in Britain where it is restricted to a handful of coastal sites in Cornwall and Pembrokeshire, and two inland areas in Wales (Merioneth and Breconshire) (Wigginton 1999). In France, G. pilosa is frequent away from the west and north (Fig. 2, reproduced from Dupont (1990)). It occurs in a variety of oligotrophic (infertile) habitats, including dry calcicolous grassland, siliceous rock outcrops, dry heaths and forest margins on acidic substrates. G. pilosa is legally protected in three French regions: Picardie. Pays-de-Loire and Poitou-Charentes (Fig. 3). In Picardie, there are three populations of G. pilosa, all located in the same area: Versigny, Montaigu and Royaucourt-et-Chailvet. CONSERVATION ACTION IN PICARDIE The Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul (CRP/ CBNBF) and the Conservatoire des Sites Naturels de Picardie (CSNP) work in partnership to safeguard habitats and species in Picardie (Hendoux & Blondel 1999), including remaining populations of G. pilosa. ^email : crp.cbnbl@wanadoo.fr *'e-mail : csn.picardie@wanadoo.fr 244 ATLANTIC ARC FIGURE 1. Map showing distribution of Genista pilosa in Europe (from Meusel et al. 1965). FIGURE 2. Map showing the distribution of Genista pilosa in France (from Dupont 1990). GENISTA PILOSA IN FRANCE 245 Figure 3. Map of French regions where Genista pilosa is protected by law (Picardie, Pays-de-Loire and Poitou-Charentes). The conservation strategy for G. pilosa has three stages: • CSNP and CRP/CBNBL carry out surveys of the species and its habitats, and assess threats to G. pilosa populations. • CSNP negotiates with the owners of sites to secure proper management, or else tries to gain direct control of the land and the use to which it is being put. At the same time, CRP/CBNBL collects seeds for ex situ conservation and defines experimental management protocols for conservation of the species in situ. • CSNP instigates programmes of site management in line with that being recommended by CRP/CBNBL, and keeps detailed management records. CRP/CBNBL carries out scientific support and studies, and assesses whether any strengthening of the populations needs to be carried out, using material from the ex situ collection. ACKNOWLEDGMENTS The authors wish to thank Direction Regionale de FEnvironnement de Picardie, Conseil Regional de Picardie, Reserve Naturelle de France, and the town of Versigny for their support. REFERENCES DUPONT, P. (1990). Atlas partiel de la Flore de France. Secretariat de la Faune et de la Flore. Collection Patrimoines Naturels. Volume No. 3. Serie Patrimoine genetique. Museum National d'Histoire Naturelle, Paris. HENDOUX, F. & BLONDEL, C. (1999). Mission d’ assistance scientifique aupres du Consen’atoire des Sites Naturels de Picardie pour la Resen’e Naturelle de Versigny. Pars I, 93 p. Lambinon, J., De Langhe, J.-E., Delvosalle, L. & Duvigneaud, J. (1999). Nouvelle Flore de la Belgique, du Grand-Duche de Luxembourg, du Nord de la France et des Regions voisines (Pteridophytes et Spermatophytes). 2eme tirage, edition du Jardin Botanique National de Belgique, Meise. MEUSEL, H., JAGER, E. & WEINERT, E. (1965). Vergleichende Chorologie der Zentraleuropaischen Flora. Volume 1. 2 vols. Jena: Gustav Fischer, Jena. WlGGINTON, M. J. ed. (1999). British Red Data Books. 1. Vascular plants. 3rd edn. Joint Nature Conservation Committee, Peterborough. 247 Romulea columnae Sebast. & Mauri (Sand Crocus) refound in Cornwall after 121 years and Juncus capitatus Weigel (Dwarf Rush) new to E. Cornwall (v.c. 2) I. J. BENNALLICK Lower Polmorla, St.Wenn, Bodmin, Cornwall, PL30 5 PE, England* ABSTRACT This paper gives details of two major botanical discoveries in E. Cornwall in 2002. First, the rediscovery of Romulea columnae at Polruan, in open, draughted cliff-top grassland quite unlike the dune grassland in which it occurs at Dawlish Warren, S. Devon (its only other British locality), but similar to its cliff-top habitats in the Channel Islands. Second, the discovery of Juncus capitatus at The Blouth, the first record of this species in E. Cornwall, and a species otherwise restricted in the British Isles to the Channel Islands, the Lizard peninsula (W. Cornwall) and Anglesey. Keywords: vegetation survey, National Vegetation Classification, maritime cliff-top grassland, botanical discovery, distribution. ROMULEA COLUMNAE Whilst undertaking a vegetation survey of the Polruan to Polperro S.S.S.I. on the south coast of E. Cornwall (v.c. 2) for English Nature, during the summer of 2001, the author came across a plant community around numerous cliff-top rocky outcrops that appeared to correspond to the Armeria maritima-Cerastium diffusion ssp. diffusum maritime therophyte community, Aira praecox sub¬ community (MC5c) of the National Vegetation Classification (Rodwell 2000). This is a widespread community of maritime cliff-tops and other open, droughted/summer-parched coastal habitats in south-west Britain and, interestingly, RodwelTs account of MC5c includes Romulea columnae amongst the suite of ‘rare species’ associated with it, presumably on account of its occurrence in similar cliff-top situations in the Channel Islands, and in short dune grassland at Dawlish Warren, S. Devon (v.c. 3). Romulea columnae was last recorded in E. Cornwall in 1879 and 1881 (Davey 1902, 1909). The description of the site in Davey (1902) reads as follows: “this little rarity was found on the cliffs near Polruan, first in May 1879, and again in May 1881, by Miss Kemp. A specimen collected in 1879, and which has been submitted to me for examination, is in Mrs Graham’s Herbarium”. In Davey (1909) the site description is “top of the cliffs near the Coastguard Battery at Polruan, May 1879 & 1881, by Miss Kemp”. Mrs Graham’s Herbarium now resides at the Royal Cornwall Museum, Truro, but the specimen of R. columnae went missing before the collection was placed in the care of the Museum. Since the original discovery of Romulea near Polruan, several people have attempted to refind it, without success. Often, the areas searched were those close to the Coastguard Lookout (SX1250), with special attention being given to any patches of ground matching Davey’ s (1909) description of a “grassy slope by the sea”. Many of the coastal slopes in this general area have become much overgrown with scrub in recent decades, and would now seem to be unsuitable for the species, and because of this Romulea had long been presumed extinct. Even so, several patches of suitable habitat were located during the 2001 survey, but the fieldwork was carried out after the time that R. columnae, if present, would have flowered. Thus, it was decided that a second visit the following year would be worthwhile, to see if Romulea could be found within the patches of vegetation that were referable to Rodwell’ s (2000) MC5c. On a sunny day, 4 May 2002, chosen as a good time to look for open flowers of the Romulea, a search for plants was carried out in MC5c on and around rock outcrops in SX1350. On an outcrop *e-mail: ian@bennallick.fsnet.co.uk 248 ATLANTIC ARC to the east of the area first considered for searching, many plants of Moenchia erecta were located. A rather uncommon species in Cornwall, this prompted a more intensive search of the species- rich, closely grazed turf in this area, and it wasn’t long before the small fruits of an unknown plant were found. Despite the similarity of the leaves, the fruits were wrong for Scilla verna and S. autumnalis, both of which are found in such situations along the south coast of Cornwall, but only as far east as Nare Head (SW9237). Having ruled out these two species, it was quickly realised that the small fruits were those of Romulea columnae (Plates 32 & 33). Ironically it was the very same place that the author had stood with Steve Payne, the National Trust warden, when discussing the logistics of the vegetation survey in 2001 - little was it realised at the time that Romulea was lurking underfoot, just waiting to be discovered the following spring! In the species account for Romulea in the Red Data Book (Wigginton 1999), N. F. Stewart states that “...it grows on well drained, leached sandy or gravelly soils in established sand dune grassland” at Dawlish Warren in S. Devon, at that time its sole mainland British locality. The habitat at Polruan is in shallow turf on and around cliff-top rocky outcrops. McClintock (1975) states that the species “is a delightful speciality of the Channel Islands that bespangles the cliffs of the south and the sandy turf of the north on sunny spring days”. McClintock also notes that “Knowlton saw [Romulea columnae ] on Monday 6 June 1726 ...always upon the highest part of the rocks”, and evidently in much the same habitat as at Polruan. Interestingly, just like at Polruan in 2002, when Knowlton first discovered Romulea in the Channel Islands his plants were in fruit rather than in flower (McClintock 1975). While Romulea occurs at its Cornish site in similar cliff habitat to that in the Channel Islands, the site at Polruan nevertheless shares several associates with the Dawlish site. These include Moenchia erecta, Myosotis ramossissima, Ornithopus perpusillus, Plantago coronopus, Rumex acetosella and Veronica ar\>ensis. At Polruan the community in which Romulea occurs is closest to MC5c; Aira praecox and Festuca ovina are also found, along with MC5 constants Armeria maritima, Festuca rubra, Cerastium diffusum and Sedum anglicum. The author revisited the site on 6 May 2002 in the company of R. J. Murphy, Dr C. N. French, R. Parslow, M. Atkinson, T. Atkinson, P. Green and M. J. Stribley. On this visit plants of Romulea were confirmed around not only the outcrop first noted but also on one near by. The vegetation on and around these outcrops is kept free of scrub by grazing cattle and sheep (Plate 31), and rabbits are also in evidence. The population was estimated at around 1,500 plants in total for the site. On small ledges below the second outcrop some of the Romulea plants were quite large and straggly, growing around and through wind-pruned Ulex europaeus. Although some authorities give the numbers of flowers per plant as between one and three, Stace (1997) gives the number as ‘one to several’, and those at Polruan varied between two and seven. A specimen was collected for verification and this is now with R. J. Murphy, ready to be placed in a National Herbarium (BM). Digital photographs of the specimen were taken by P.G. Bennallick, and these were examined by Dr T.C.G. Rich, who compared them with herbarium material from Dawlish, Guernsey, Jersey and France and found them to be a perfect match. Digital photographs of plants on site were also taken by M. J. Stribley (Plates 32 and 33). The owners of the site (National Trust) and English Nature are fully aware of the discovery and its importance, and the future management of this part of the S.S.S.I. is currently being discussed. JUNCUS CAPITATUS In the same week as the discovery of Romulea at Polruan, the author was contacted by W. Eyre, the National Trust warden for the length of coast between Black Head and Pendower, E. Cornwall (v.c. 2). He was wondering whether he might have found Minuartia verna on The Blouth, a headland to the north of Nare Head (SW9238). The geology here is very interesting, as the rocks that outcrop are of the same type as those occurring on the Lizard peninsula, both Nare Head and The Blouth being basalt/dolerite outcrops interspersed with gabbro and serpentine. On the Lizard, Minuartia verna grows on and around serpentine outcrops, these being ultrabasic, and as two other Lizard specialities, Orobanclie alba and Allium schoenoprasum, had also been recorded near Nare Head, it seemed likely that Minuartia verna might also occur there. The site was visited on 9 May 2002 by the author and W. Eyre. The plants thought to be Minuartia verna proved to be Sagina subulata. However, with the find of Romulea columnae still ROMULEA COLUMNAE AND J UNCUS CAPITATUS NEW TO E. CORNWALL 249 fresh in the mind, it was decided to investigate the communities similar to MC5c in this area, just in case Romulea was present here too. None was found, but while searching in some shallow erosion pans, similar to those found on The Lizard, several plants of Radiola linoides were seen, as well as a small reddish Juncus species that proved on closer inspection to be Juncus capitatus (Plate 30). Around 100 plants of J. capitatus were recorded, and on a second visit on 14 May 2002 with Dr P. Gainey, M. Atkinson and T. Atkinson, a second colony of around 20 plants was found near by. /. capitatus occurs at The Blouth in open, short vegetation, with plenty of bare ground. Associated species include Anthoxanthum odoratum, Aira caryophyllea, A. praecox, Festuca ovina, Erica cinerea, Plantago coronopus, Radiola linoides, Ornithopus perpusillus , with occasional Sedum anglicum, Hypochaeris radicata, Vulpia bromoides, Lotus comiculatus, Lotus subbiflorus, Danthonia decumbens and, more rarely, Hypericum humifusum, Hyacinthoides non-scripta, Luzula campestris and Teucrium scorodonia. The site was heavily sheep-grazed in 2001, when restrictions over livestock movements due to the Foot and Mouth crisis had necessitated the use of this coastal area for grazing once adjacent fields had become ‘exhausted’. It is possible that the increase of grazing in 2001 had created the open habitat needed for Juncus capitatus to appear in 2002 (from a buried seed-bank?). Juncus capitatus is found in some years in quantity in suitable habitats on the Lizard peninsula, W. Cornwall (v.c. 1), and in the Channel Islands and Anglesey (Wigginton 1999; Preston et al. 2002). In W. Cornwall there are also old records from West Penwith, but the site at The Blouth is new: it is the first record of this species from E. Cornwall (v.c. 2), and represents a significant extension to its British range. CONCLUSION The rediscovery of Romulea columnae at Polruan and the discovery of Juncus capitatus new to E. Cornwall (v.c. 2) should be a major encouragement to those botanists engaged in the search for further records of rare and scarce plants in Cornwall. The cliff-top turf at Polruan is similar to that in which Romulea is found in the Channel Islands, and this is a widespread plant community in Cornwall (and elsewhere in south-west England) wherever suitable coastal habitats exist. This suggests that it would be worth searching for Romulea in similar areas elsewhere, particularly along the south coast of Cornwall and in S. Devon. Juncus capitatus may also be worth searching for at other Cornish sites, although opportunities for this species may be limited by the restricted occurrence of the rock types which it appears to favour. ACKOWLEDGMENTS Many thanks to all those who were involved with the field visits, to N. F. Stewart for information on the habitat and associated species of Romulea at Dawlish Warren, to Dr T. C. G. Rich who confirmed the identification of Romulea, and to Rosaline J. Murphy for her help and advice throughout. Thanks also to the National Trust for permission to visit both sites, and to W. Eyre for his help with recording at The Blouth. REFERENCES Davey, F. H. (1902). A Tentative List of the Flowering Plants and Ferns etc. known to occur in the County of Cornwall. P. Chegwidden, Penryn. Davey, F. H. (1909). Flora of Cornwall. P. Chegwidden, Penryn. McCLINTOCK, D. (1975). Wild Flowers of Guernsey. Collins, London. PRESTON, C. D., Pearman, D. A. & Dines, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. RODWELL, J. S. ed. (2000). British Plant Communities. Vol. 5. Maritime Communities and Vegetation of Open Habitats. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of the British Isles. 2nd ed. Cambridge University Press, Cambridge. WIGGINTON, M. J. ed. (1999). British Red Data Books 1 Vascular Plants. 3rd ed. Joint Nature Conservation Committee, Peterborough. 251 Hypericum linariifolium x humifusum : a hybrid adapted to the Atlantic coast R. A. JONES Countryside Council for Wales, Plas Gogerddan, Penrhyncoch, Aberystwyth, Ceredigion SY23 IBP, Wales * and M. D. SUTTON Countryside Council for Wales, Llanion House, Llanion Park, Pembroke Dock, Pembrokeshire SA72 6DY, Wales ABSTRACT Hybrids between the nationally - and internationally - rare Hypericum linariifolium Vahl. (Toadflax-leaved St John’s-wort) and the more widespread H. humifusum L. (Trailing St John’s-wort) are reported for coastal heath in west Wales. Their combined drought-tolerance and prostrate habit may have a selective advantage in this markedly windswept situation. Keywords: Trailing St John’s-wort, Toadflax-leaved St John’s-wort, hybridisation, evolutionary adaptation, coastal heaths, Pembrokeshire, Ll>n peninsula. HYPERICUM LINARIIFOLIUM IN WALES - OLD RECORDS Hypericum linariifolium Vahl. (Clusiaceae) (Plate 35) is a rare and threatened plant in Britain (McDonnell & Muddeman 1999) and probably elsewhere in its range, across Atlantic Europe and Madeira (Fig. 1). In Wales, at its northerly limit, the species can now only be found on a few rocky outcrops in south Llyn (Caernarfonshire) but there are intriguing historic records from Trearddur. Anglesey (1915), Abermaw/Barmouth in Meirionydd (1907) and Radnorshire (perhaps Llandegly 1945)1. None of these outliers ever seem to have been refound but, then again, wild fluctuations are also typical of H. linariifolium in Britain, with whole populations vanishing in some years and flourishing in others. The early records from Llyn paint a vivid picture of its unexpected disappearance and recovery (Babington 1889; Griffiths 1895; Carr 1897; Druce 1920) and there is a distinct possibility that some colonies still remain in their old haunts - and maybe in other suitable habitat. They should be looked for on south-facing rocky slopes with a thin acidic soil, where there is a low risk of frost. THE PEMBROKESHIRE PLANTS Given the presence of Hypericum linariifolium in both Devon and Cornwall, it is perhaps surprising that the species has never been found in Pembrokeshire (inset, Fig. 1). In fact, most of the suitable habitat there seems to be occupied by the more common H. humifusum L. although, on closer inspection, some populations in Pembroke actually seem to be neither species - or perhaps a Another record from near Buildwas, Shropshire, in 1919, seems to have been of the same isolated and ephemeral kind. *e-mail: a.jones@ccw.gov.uk 252 ATLANTIC ARC FIGURE 1. World distribution of Hypericum linariifolium. HYPERICUM LINARIIFOLIUM x H. HUMIFUSUM 253 TABLE 1. HYPERICUM LIN ARIIF OPIUM x H. HUMIFUSUM IN WALES Location Habitat Grid Ref. Date Recorder Determiner Near Solfach, St David’s Coastal and inland heaths Head & Airfield SM72 2001- -2002 Matt Sutton N. K. B Robson & Matt Sutton Dinas Fawr - Newgale Coastal heath SM82 2002 Matt Sutton Matt Sutton Lydstep Head Coastal heath (Old Red Sandstone) SS09 2002 Stephen Evans Stephen Evans Trwyn Cilan Coastal grassland SH22 2002 Iwan Edgar N. K. B. Robson bit of both. In 2001 one of us (M.S.) found several Hypericum plants in coastal heath near St. Davids/Tyddewi with moderately large flowers and half-upright stems, more or less pointed, ciliate sepals and narrow, fairly glandular leaves. These are all features belonging to one or other species but not in this combination or to this extent. These plants are neither Toadflax-leaved nor Trailing St. John's-wort. Intermediates between H. linariifolium and H. humijusum have been described as both forms and varieties of the two species or even as evidence of two imperfectly-separated H. humijusum subspecies (Pugsley 1915). The alternative interpretation, however - that they represent hybrids or introgression between two closely-related species (e.g. McClintock 1975; Robson 1990) - fits more closely with their distribution and ecology and is supported by recent molecular genetic evidence (Kay & John 1995). For these reasons the St Davids’ material has been identified as H. linariifolium x humijusum (N. K. B. Robson pers. comm.) and further hybrid populations have since been identified from coastal sites and windswept inland heaths elsewhere in Wales (Table 1). Probable hybrids are also found at the western tip of Llyn, in a markedly more exposed habitat than that favoured by H. linariifolium , and they may well be present elsewhere in west Wales - and perhaps in southern England too2. EVOLUTIONARY IMPLICATIONS It is often thought - and perhaps especially in the case of rare plants - that hybridisation represents a breakdown of natural barriers and the loss of biological distinctiveness. Hypericum linariifolium and H. humijusum seem to have come together (at least partly) through climate change, however, with the more warmth-demanding species contracting at the edge of its range. And, in one sense, this process has also probably conserved something of H. linariifolium. A number of plants along the Atlantic coast have adapted to its extreme windiness through the evolution of prostrate subspecies and ecotypes (e.g. Cytisus scoparius subsp. maritimus, Genista tinctoria subsp. littoralis, Juniperus communis subsp. hemisphaerica). The descendents of Toadflax-leaved St. John's-wort in some sites seem to have acquired their prostrate growth habit through hybridisation, carrying with them the drought-tolerance of their rare southerly parent with the prostrate, wind- resistance of the trailing species. In this way they have successfully adapted to a distinct niche on very windswept coastal heath. ACKNOWLEDGMENTS Thanks to Dr N. K. B. Robson, Stephen Evans and Iwan Edgar. :The first British record of H. linariifolium, from Cape Cornwall (1839), was of the intermediate ‘var. approximatum ’ and this variety is also reported from The Lizard and elsewhere. Interestingly, the equivalent H. humijusum ‘var. ambiguum ’ is also found near Land's End (Pugsley 1915). 254 ATLANTIC ARC REFERENCES BABINGTON, C. C. (1889). Hypericum linariifolium in Carnarvonshire. Journal of Botany 27: 187. CARR, J. W. (1897). Hypericum linarifolium [sic] in Carnarvonshire. Journal of Botany 35: 58-59. DRUCE, G. C. (1920). H. linariifolium. New County Records: Botanical Society & Exchange Club of the British Isles 5: 645. Kay, Q. O. N. & John, R. (1995). The conservation of scarce and declining plant species in lowland Wales: population genetics, demographic ecology and recommendations for future conservation in 32 species of lowland grassland and related habitats. Science Report 1 10. Countryside Council for Wales, Bangor. GRIFFITH, J. W. (1895). Flora of Anglesey and Carnarvonshire. Bangor. MCCLINTOCK, D. (1975). The Wild Flowers of Guernsey. Collins, London. MCDONNELL, E. J. & MUDDEMAN, J. L. (1999). Hypericum linariifolium L., in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants, pp. 196-197. Joint Nature Conservation Committee, Peterborough. PUGSLEY, H. W. (1915). British forms of Hypericum humifusum and H. linariifolium. Journal of Botany 53: 162-170. ROBSON, N. K. B. (1990). British and Irish St. John’s-worts, in A Guide to Some Difficult Plants; Illustrated Articles from the Wild Flower Society Magazine 1973-1988. Wild Flower Society. 255 Lotus angustissimus L. (Slender Bird Vfoot- trefoil) in S.W. England S. J. LEACH English Nature, Roughmoor, Bishop's Hull, Taunton, Somerset, TA1 5AA, England * ABSTRACT In the early 1990s, while new data were being gathered for the third edition of the British Red Data Book, surveys were undertaken to establish the distribution and conservation status of Lotus angustissimus L. in S. W. England. This note gives a brief summary of the results of these surveys and of more recent work on the species. Keywords: distribution, conservation, management, cliff-top grassland, scrub. Red Data Book. INTRODUCTION In 1993-94, as part of the fieldwork for the revision of the British Red Data Book (Wigginton 1999), R. D. Porley, J. H. S. Cox and I carried out an extensive survey of Lotus angustissimus L. (Slender Bird’s-foot-trefoil, Plate 34) in Devon and Cornwall (Leach 1995). The aims of the project were threefold: first, to produce an up-to-date assessment of the distribution and conservation status of L. angustissimus in S.W. England; second, to establish whether the species was declining, and if so for what reasons; and third, to consider the plant’s habitat requirements and assess whether instigation of appropriate habitat management on its remaining sites would improve its long term chances of survival. Our results are summarised below. DISTRIBUTION Since 1987 Lotus angustissimus has been reported from 22 10-km squares in Britain and 11 10-km squares in the Channel Islands (Pearman 2002). As such, it is no longer regarded as Nationally Rare (i.e. <15 10-km squares in Britain), and in the third edition of the Red Data Book (Wigginton 1999) it was listed as Nationally Scarce. However, it has a restricted distribution in Britain, and our surveys in 1993-94 indicated that in S.W. England the main populations are concentrated within just three areas: • The north Cornish coast, at Pentire Point (SW98); 200-300 plants in 1994 • The south-east coast of Cornwall, between Polruan (SX15) and Penlee Point (SX44); 600+ plants in 1994 • The coast of the South Hams district of Devon, between the Salcombe Estuary (SX73) and Start Point (SX83); 6,500+ plants (and at least a dozen localities) in 1993-94. More recent surveys of the South Hams coastline have turned up several new localities, and we now have records of around 25 sub-populations from this area, although the species may not be present at every one of these in a single year. Thus, the coastline between the Salcombe Estuary and Start Point appears to be the main stronghold for L. angustissimus in S.W. England. Since the mid 1990s, however, N. F. Stewart, R. Smith and others have been finding it in S. Devon at several inland sites in the Teign valley, and near Bishopsteignton and Dawlish. Several of these populations are very large, containing thousands of plants in a ‘good’ year, and appear to rival even the largest of the coastal colonies. *email: simon.leach@english-nature.org.uk 256 ATLANTIC ARC HABITAT Lotus angustissimus is usually found in submaritime or paramaritime cliff-top grassland and in open patches amongst cliff-top scrub, around rock outcrops and especially on path-side banks. It seems to have a preference for thin, drought-prone soils, and sunny, relatively sheltered locations with a S.E. to S.W. aspect. Our releves indicated that L. angustissimus tends to occur in short, well-grazed or trampled grassland conforming to the Hypochaeris radicata sub-community of the Festuca ovina-Agrostis capillaris-Rumex acetosella grassland (N.V.C. community Ulf (Rodwell 1992)), or to transitions between Ulf and the Rumex acetosella sub-community of the Ulex europaeus-Rubus fruticosus agg. scrub (W23b (Rodwell 1991)). Agrostis capillaris, Crepis capillaris, Dactylis glomerata, Hypochaeris radicata and Vulpia bromoides were constant (i.e. in >60% of quadrats) in stands containing L. angustissimus , while Anagallis arvensis, Cerastium fontanum, Holcus lanatus, Plantago lanceolata, Rumex acetosella, Senecio jacobea, Sonchus asper, Trifolium dubium and T. repens were all frequent (41-60%). The Nationally Scarce Lotus subbiflorus was also a frequent associate, and may help to ‘hide’ L. angustissimus on sites where the latter is only present in small quantity. CONSERVATION ISSUES The balance between Ulf and W23b is crucial. Too much scrub can rapidly lead to the demise of L. angustissimus , along with an array of small annuals with which it is frequently associated, e.g. Lotus subbiflorus, Medicago polymorpha, Ornithopus perpusillus, Trifolium glomeratum, T. scabrum, T. striatum. Many extant colonies are small, and even the largest populations vary considerably in size from year to year. It can be notoriously erratic in its appearance: one year there may be thousands of plants, whilst in the next it can be almost impossible to find. Nevertheless, our surveys indicate that some populations - particularly in W. Cornwall - have been lost altogether, and it is likely that L. angustissimus has declined in many areas (including S.S.S.I.s) following the cessation of traditional management practices such as grazing and burning. Annual cutting back of path-side scrub has been important in enabling the plant to persist at many of its cliff-top localities. More extensive scrub control - along with the re-introduction of grazing - is now being attempted at some sites (e.g. Prawle Point, S. Devon), with the conservation of this species very much in mind. However, it may be that on more exposed sites the presence of at least some scrub is actually beneficial, helping to provide the sheltered conditions that L. angustissimus seems to favour. REFERENCES LEACH, S. J. (1995). Lotus angustissimus - a sur\>ey of selected sites in Devon and Cornwall. Unpubl. report. Joint Nature Conservation Committee, Peterborough. PEARMAN, D. A. (2002). Lotus angustissimus, in PRESTON, C. D., PEARMAN, D. A. & DINES, T. D. eds. New Atlas of the British & Irish Flora, p. 658. Oxford University Press, Oxford. RODWELL, J. S. ed. (1991). British Plant Communities. Vol. 1. Woodlands and scrub. Cambridge University Press, Cambridge. RODWELL, J. S. ed. (1992). British Plant Communities. Vol. 3. Grasslands and montane communities. Cambridge University Press, Cambridge. WlGGINTON, M. J. ed. (1999). British Red Data Books, 1 Vascular Plants, 3rd ed. Joint Nature Conservation Committee, Peterborough. 257 Puccinellia foucaudi (Coste) Hackel on the French mid- Atlantic coast: taxonomic status, morphology and distribution C. LAHONDERE Vice-President, Societe Botanique du Centre-Ouest, 94 avenue du Parc, 17200 Royan, France ABSTRACT Puccinellia foucaudi (Coste) Hackel (Poaceae) is a taxon recorded along the estuaries of the French mid- Atlantic coast, from the Vilaine River down to the Gironde River. The plant is distinguished from other closely related Puccinellia species by its tufted growth, up to at least one metre in height, and by its ability to produce stolons after anthesis. Keywords: Puccinellia, taxonomy, estuary, stolon, rhizome. HISTORICAL BACKGROUND The oldest mention of this taxon that we have been able to find is in Foucaud (1893). Rouy (1913), in the last volume of his Flore de France, described it briefly under the name of ‘Atropis foucaudi Hackel’, and considered it to be a race of Atropis maritima Griseb (= Puccinellia maritima (Huds.) Pari.). In the third volume of his Flore descriptive de la France , Coste (1937) described it under the name of ‘ Glyceria foucaudi (Hackel) Coste’, and raised it to the rank of a distinct species. This taxonomic rank was subsequently confirmed by Fournier (1961). Guinochet & Vilmorin (1978), in the third volume of their Flore de France, also considered it to be a distinct species and named it Puccinellia foucaudi (Coste) Hackel. The apparent restriction of this species to the coast of the French mid-Atlantic (Charente-Maritime) may explain why there is no mention of it in foreign/ European Floras: Volume 5 (1980) of Flora Europaea, for example, does not cite it. However, we suspect that once British and Irish botanists have been made aware of this taxon - and how it differs from Puccinellia maritima - it may also be found on the Atlantic shores of the British Isles. DISTINGUISHING BETWEEN PUCCINELLIA FOUCAUDI AND P. MARITIMA I have studied Puccinellia foucaudi in detail at two Charentais stations: the bank of the Charente estuary in Saint-Laurent-de-la-Pree, upstream from Rochefort; and the Bonne Anse Bay (Commune of Les Mathes) near the lighthouse at La Coubre. At both sites, P. foucaudi forms spaced out clusters seeming to be independent one from the other; plants are tall, with height ranging from 04 to 1-2 m, most frequently between 0-7 and 0-9 m. We took a few clumps from both sites in June, after anthesis, then hosed down the mud and sand so that we could observe the underground organs of the plant. Several clumps had one or two stolons clearly visible amidst the network of roots (Fig. 1). Those nude, unrooted stolons (except at the level of the first intemode visible in Fig. 1) were about 40 cm long, and displayed at their tips a bud not yet rooted in the substratum. At Bonne Anse, the P. foucaudi station is next to vast expanses of P. maritima. The two species are easily distinguished from each other not only by their size - P. maritima hardly reaching more than 0-5 m - but particularly by their growth form: ‘turfy’ for P. maritima, in clusters for P. foucaudi. Guinochet & Vilmorin (1978) distinguished two groups in the Puccinellia genus: the first with “sterile, radicate, stoloniform innovations”, including both P. maritima and P. foucaudi ; the second without such “stoloniform innovations”, including all the other French Puccinellia species. In Flora Europaea, there is a distinction between stoloniferous Puccinellia rooting at the nodes (“stolons present, rooting at nodes”) and other Puccinellia species lacking stolons; all the European Puccinellia species belong to the second group, apart from P. maritima (with extravaginal stolons) and P. phryganodes (with intravaginal stolons) - a taxon from Arctic Europe. Thus, in Flora Europaea it is clear that P. foucaudi could be subsumed within P. maritima - which would agree with the opinion of some French botanists. We differ from this view, however. 258 ATLANTIC ARC Figure 1. Base of the plant with one stolon. on the grounds that the stolons of P. foucaudi do not root, except at their extremities. From our own observations we think that P. maritima is a rhizomatous species, its turfy or cespitose character resulting from the development of numerous “short rhizomes squeezed against one another” (Marouf 2000). P. foucaudi, on the other hand, is a stoloniferous species, that is to say it produces stolons, where a stolon is defined as a superficial, “creeping stem with long intemodes... with limbs of a reduced size and different from those growing on the main stem, ending in a bud which, where it is in contact with the ground, produces adventitious roots, which makes it able to have new plantlets” (Marouf 2000). In our opinion, these definitions correspond well, from a morphological point of view, with what we observed about P. foucaudi and P. maritima during our investigations on the Charentais coast. The mobility of superficial deposits (mud and sand) cause the stolons to be covered quickly, and as they decay the various individuals become independent of one another. The presence or absence of stolons is essential for us to be able to distinguish P. foucaudi from P. maritima, and we suspect that confusion between them has been partly due to the vocabulary chosen by some writers. Thus, Coste (1937), writing about P. foucaudi, notes that it has “a rootstock sprouting numerous, long stoloniform tillers” while P. maritima is “a plant... sending out horizontal, radicate stoloniform tillers”. The choice of the phrase “stoloniform tiller”, used also by other authors, is unfortunate, for the outcome is that many botanists (though not Coste!) have been overlooking the rare P. foucaudi, assuming it to be just a particularly vigorous form of the common P. maritima. It should be noted that Coste (1937) writes that the “stoloniform tillers” are “numerous” in P. foucaudi, whereas our observations have shown us that this is not the case. PUC CINELLIA FOUCAUDI - MORPHOLOGY AND PHENOLOGY MORPHOLOGY 1. The upright stems are distinctly grooved. The limbs of the leaves are soon twisted and the ligule is triangular with a blunt tip (Fig. 2). 2. The panicle is pyramidal with upright, slanting branches; very rarely, after anthesis, the branches of the panicle may get closer to the main axis. It is 15-27 cm long, most frequently 20-23 cm. 3. The spikelet (Fig. 3) is 7-14 mm long, with an average of 8-12 mm; it has between 6 to 10 flowers. 4. The lower glume has one rib, more rarely 3, and is 1 -5-3-2 mm long. 5. The upper glume has three ribs and is 2-4-3 -8 mm long. 6. The lemma has five ribs and is 2-8^4-2 mm long. 7. The palea has two keels with cilia and is the same size as the lemma. The panicle of P. foucaudi appears to be much longer than that of P. maritima [3-15 (-25) cm] according to most Floras, but the morphology and size of all parts of the spikelet of P. foucaudi is almost identical with those of P. maritima. PUC CINELLIA FOUCAUDI 259 Figure 2. Ligule of Puccinellia foucaudi showing FIGURE 3. Spikelet, ruler scale in mm. triangular shape and blunt tip. Note distinct grooves on stem. PHENOLOGY Our observations indicate that Puccinellia foucaudi is an earlier-flowering species than P. maritima. In 1994, we noted that the latter started to flower in early June on the Charentais coast, by which time the flowering of P. foucaudi was well over in the Charente estuary. CHOROLOGY1 The following notes are relevant in any assessment of the habitat and geographical extent (distribution) of Puccinellia foucaudi in France. It was recorded by Foucaud (1893) and Rouy (1913) from the “banks of the Charente River from Rochefort to Port-des-Barques”, and in the “sandy marshes of Fouras and Yves”. Coste (1937) noted it from “marshes and ditches, on the salty banks of rivers in Charente-Inferieure”2, while Fournier (1961) also had it from the same area, on “salty muds and in ditches”. Drawing, presumably, from previous accounts, Guinochet & Vilmorin (1978) described the species as occurring on the “salty muds of Charente-Maritime”. More precise details of localities than the above include those of Dupont (1954, 1974), who noted that P. foucaudi was to be found on the banks of the maritime Vilaine River (near Fegreac); it has since disappeared, following construction of the Arzal dam (Dupont 2001; and pers. comm.), though is still present in the Bourgneuf Bay in Loire-Atlantique (Dupont 2001). Corillion (1953) ]A word originating in the 19th century meaning “the scientific study of the geographical extent of things” [ed.] 2Charente-Inferieure being the former name of Charente-Maritime, the latter thought to be less derogatory! 260 ATLANTIC ARC recognised it in the Breton Marsh and in the Isle of Noimoutier (Vendee). Lahondere (1993) mentioned it from the banks of the Charente River upstream from Rochefort, from Saint-Laurent- de-la-Pree to Port-Neuf; more recently, I have also recorded it from the Bonne Anse Bay near the Coubre lighthouse, north of Royan (Charente-Maritime). Lastly, we have learnt of it being found in 1997 in Ars-en-Re (Charente-Maritime) (J.-M. Tison pers. comm.). CONCLUSION Puccinellia foucaudi is a taxon that can be easily distinguished from the common P. maritima by its tall stature and long stolons. We hope that its identity as a true species will be recognised more widely - and its distribution and ecology made clearer - before measures of protection can be contemplated should its status as a French/European endemic species be confirmed. We are uncertain, at present, whether this taxon is truly endemic to the mid-Atlantic coast of France, or whether it may be present, but overlooked, in other Atlantic coastal regions of Europe. ACKNOWLEDGMENT I wish to thank Yves Peytoureau, Secretary, S.B.C.O., for translating this paper into English. REFERENCES Corillion, R. (1953). Phanerogames interessantes pour la Bretagne. Bull. Soc. Sc. Bretagne 27: 55-64. COSTE, H. (1937). Flore descriptive de la France. Vol. 3. Lib. des Sciences et des Arts. Paris. Dupont, P. (1954). La vegetation des marais de la Vilaine maritime. Bull. Soc. Sc. Bretagne 29: 65-104. Dupont, P. (1974) - Additions a la flore de Loire-Atlantique, de Vendee et du Morbihan. Bull. Soc. Sc. Nat. Ouest France 72: 33-38. DUPONT, P. (2001). Atlas floristique de la Loire-Atlantique et de la Vendee. Vol. 1: Etat et Avenir d’un Patrimoine; Vol. 2: Cartes et Commentaires. Siloe. edit. Nantes. Tutin, T. G., Heywood, V. H., Burges, N. A., Moore, D. M., Valentine, D. H., Walters, S. M. & RICHARDSON, I. B. K. eds. (1980). Flora Europaea. Vol. 5. Cambridge University Press, Cambridge. FOUCAUD, J. (1893). Atropis foucaudi Hackel. Bull. Soc. Bot. Rochelaise 15: 43-44. FOURNIER, P. (1961). Les Quatre Flores de la France. P. Lechevalier, Paris. Guinochet, M. & VlLMORIN, R. de (1978). Flore de France. Fasc. III. C.N.R.S. Paris. Lahondere, C. (1993). Contribution a l’etude de deux especes littorales: Oenanthe foucaudi Tesseron. Puccinellia foucaudi Holmberg. Bull. Soc. Bot. Centre-Ouest., N.S. 24: 41-60. MAROUF, A. (2000). Dictionnaire de botanique: les Phanerogames. Dunod, Paris. ROUY, G. (1913). Flore de France. E. Deyrolle, Paris. 261 Status of Teucrium scordium L. (Water Germander) in north-west France B. TOUSSAJNT, B. DESTINE & F. HENDOUX Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul, Hameau de Haendries, 59270 Bailleul, France* ABSTRACT Teucrium scordium L. (Water Germander) is a polymorphic species: two subspecies are known in France, subsp. scordium and subsp. scordioides. In this paper we briefly discuss the results of a biometric study of populations in N. W. France and Brittany; and we suggest that one population in dune slacks on the Merlimont Nature Reserve (Pas-de-Calais) may be referable to the Mediterraneo- Atlantic subsp. scordioides, while others on that site appear to be intermediate between subsp. scordioides and subsp. scordium. We give a summary of the biology, autecology and distribution of T. scordium in the Nord/Pas-de-Calais region. A survey of all known populations in the region has allowed us to highlight priorities for the long-term conservation of this threatened species. Keywords: autecology, distribution, conservation. INTRODUCTION Teucrium scordium (Water Germander) is a rare and declining species in northern France and southern Britain. We were fortunate that a ‘Trans-Channel’ Interreg programme allowed us to make a thorough survey of this species in the Nord/Pas-de-Calais region, including detailed assessments of its taxonomy, autecology, distribution and conservation status (Blondel et al. 2001). In considering the long-term safeguard of T. scordium we were particularly concerned to identify threats to its remaining sites, and to compare the vulnerability and magnitude of decline of coastal and inland populations in the region. TAXONOMY Two subspecies of Teucrium scordium are recognised in France: subsp. scordium, the main subspecies, a declining taxon, but still widely distributed and found at both inland and coastal sites; and subsp. scordioides (Schreber) Arcangeli, apparently restricted to Mediterraneo-Atlantic coasts where it grows in dune-slacks or in ponds associated with coastal marshes. The two taxa differ most obviously in their pubescence (Plate 37), which is much more developed in subsp. scordioides (it can look quite ‘woolly’), but there are other morphological characters mentioned in the Floras that may be used to separate the subspecies (Table 1). Subsp. scordioides is known to occur as far north as Basse-Normandie (western coast of Cotentin), although the true identity of some of these northern coastal populations still requires verification. This is certainly the case with those populations occurring in the Pas-de-Calais dunes - in the Merlimont Nature Reserve - where much of the T. scordium displays characters suggesting subsp. scordioides ; one of the aims of the present work was to determine whether or not these populations are indeed referable to this Mediterraneo-Atlantic taxon or to the more widespread subsp. scordium. To this end, a biometric study was carried out on five ex situ populations and ten in situ populations (Legrand 1996). Material of subsp. scordioides from *e-mail: crp.cbnbl@wanadoo.fr 262 ATLANTIC ARC TABLE 1. MORPHOLOGICAL DIFFERENCES BETWEEN TEUCRIUM SCORDIUM SUBSP. SCORDIUM AND SUBSP. SCORDIOIDES subsp. scordium subsp. scordioides Plant hairiness Can be quite hairy, but not strongly pubescent Strongly pubescent - can be very hairy (‘woolly’ -looking) Size of the plant 10-20 cm 20-60 cm Habit very branched stem almost simple Stolons with small leaves with scale-like leaves Leaf shape sessile but not amplexicaul, toothed mainly near the tip, length/width ratio of 2-3: 1. amplexicaul, toothed all around, length/ width ratio of 1-2: 1 . Brittany (Erdeven) was used as a morphological ‘reference’ for that subspecies. 16 characters were measured on each population; of these, pubescence (density of the hairs), leaf-shape (width/length ratio) and number of teeth per leaf were found to be the most reliable characters for distinguishing between populations and assigning them to either of the two subspecies. A graph plotting ‘hairiness’ against leaf width/length ratio (Fig. 1) shows that all the sampled inland populations cluster together, having a leaf morphology consistent with subsp. scordium; but that, in contrast, the Merlimont dune-slack populations are morphologically rather variable, with some plants/sub¬ populations appearing to be intermediate between the two subspecies, and one sub-population, at least, close to the Breton samples and therefore seemingly referable to subsp. scordioides (Plate 38). It could be argued, of course, that plants showing ‘intermediate’ morphology may simply reflect phenotypic plasticity due to edaphic or climatic stress. Such a view is untenable, however, as we found that even after seven years of ex situ cultivation the morphological differences between the various populations are still evident. Thus, it seems likely that the observed morphological differences do indeed have a genetic basis: is it possible that ‘intermediate’ populations have arisen by hybridisation between the two subspecies, or are they simply variants within a continuous spectrum of variation in leaf shape/leaf hairiness that is, in effect, represented at its extremes by the two named subspecies? FIGURE 1. Variation in leaf hairiness and leaf width/length ratio of Teucrium scordium populations in N.W. France (The first four sites in the key are inland, the rest are coastal). STATUS OF TEUCRIUM SCORDIUM IN NORTH-WEST FRANCE 263 Further studies on a larger geographic scale would be useful. For example, we are uncertain whether the Breton populations of Teucrium scordium from which we collected our reference material are truly the same taxon as that named as subsp. scordioides from the Mediterranean region. Detailed molecular work would be especially useful in helping to make sense of the morphological variation displayed by this species across its European range. However, even without this work - and despite the taxonomic uncertainties - we remain convinced that the coastal populations in the Nord/Pas-de-Calais region are morphologically (and probably also genetically) separable from the relatively widespread subsp. scordium , and that, as such, they deserve to be specially protected. BIOLOGY Teucrium scordium is highly allogamous (insect-pollinated, e.g. by bees and hoverflies (Plate 36). Seeds are generally dispersed by the wind, but dispersal by water can also occur in particular situations (e.g. in winter-flooded dune slacks). Achenes are rough, with a thick, woody tegument (Plate 39), which allows the seed to remain viable for several years in sediments (buried seed- bank). In cultivation, good germination responses were obtained with seeds refrigerated at -20° C for 7 years. Vegetative reproduction is also very effective, as T. scordium produces not only underground rhizomes but also creeping stolons that root easily. The species usually forms large patches and counting individuals is therefore often impossible. For our population surveys (see below), we took a pragmatic approach, assuming that each erect stem was equivalent to one plant. HABITAT Inland populations of Teucrium scordium in northern France grow in damp, light or half-shaded and quite open habitats, usually on base-rich soils. Water level varies between seasons. The plant used to occur quite widely in alluvial meadows ( Bromion racemosi ) but we now find it usually in damper biotopes, e.g. in wet meadows ( Oenanthion fistulosae ) and on river or ditch banks ( Oenathion aquaticae ) (Fig. 2). Meadow systems Class Atrhenatheretea elatioris Order Agrostietea stoloniferae Phragmiti australis - Magnocaricetea elatae Agrostietalia Eleocharitetalia stolonifera palustris Phragmitetalia australis Alliance Bromion racemosi Oenanthion fistulosae Oenanthion aquaticae - - - - - - ► Teucrium scordium Figure 2. Diagrammatic representation of the range of vegetation types within which Teucrium scordium may be found in N.W. France. Teucrium scordium L. Repartition U.T.M. {1 KmJ) 264 ATLANTIC ARC Figure 3. Distribution of Teucrium scordium in the Nord/Pas-de-Calais region. Symbols denote date-class of last known record at each site. STATUS OF TEUCRIUM SCORDIUM IN NORTH-WEST FRANCE 265 iut’liire U 3 KC-hlaUf, "if- A ff*y- ^ w 3*|*I «1# j^i Teucrium scordium L • Localisation ce Is population * rrth*' r f,B F"enait +•*?'.' BAILLfcl L FIGURE 4. Example of site map, showing the location of Teucrium scordium sub-populations within the Merlimont dunes. 266 ATLANTIC ARC In dune slacks, it occurs at several different topographic levels: it may be found in low-level amphibian communities ( Samolo-Litorelletam ), Carex trinervis- dominated vegetation (. Drepanoclado-Caricetum trinervis), high-level amphibian communities ( Carici scandinavicae- Agrostietum maritimae), and hygrophilous marshes ( Calamagrostio-Juncetum subnodulosi or Ophioglosso-Calamagrostietum epigei). DISTRIBUTION IN N.W. FRANCE Teucrium scordium is now mainly restricted to sandy coasts of Pas-de-Calais, Somme and to a few valleys in the north of the region (Plaine de la Lys et Plaine de la Scarpe) and in Picardie (the Oise valley and a few localities of the Aisne department) (Fig. 3). In Haute-Normandie, it is now considered extinct (last record 1983). The species decreased dramatically in the Nord/Pas-de- Calais region in the 20th century, with 33% of its coastal sites and 64% of its inland sites apparently being lost during that period. Overall, it seems that just over half (53%) of the populations of our region have now disappeared (Fig. 3). A survey of all the known populations in the Nord/Pas-de-Calais region was carried out in recent years: seven coastal and twelve inland sites were surveyed, and detailed maps were drawn showing the exact positions of all populations of Teucrium scordium. An example of one of our site maps is shown in Figure 4. At each site we estimated the population size of T. scordium and carried out a threat assessment (Figs 5 & 6); and, for those populations occurring in a habitat that was sub- optimal (poorly managed), we looked at the restoration potential, and made suggestions to site managers about how their management could be modified to protect this species better. Just three of the 19 sites were judged to be ‘not threatened’, all of them coastal - at Merlimont, Mont-St- Frieux and Baie de Canche - and supporting by far the largest populations; the remaining 16 sites 5000 2500 - o o o LO A mm o o o LO A o 111 o i o _ c\j _ A Not threatened — I - 1— o o o LO o Q. CO c 0 ^ _0 c 0 0 c o O O =3 Q 0 CO FIGURE 5. Population size and threats to coastal populations of Teucrium scordium in Nord/Pas-de-Calais. STATUS OF TEUCRIUM SCORDIUM IN NORTH-WEST FRANCE 267 endangered critically restricted presumed extinct FIGURE 6. Population size and threats to inland populations of Teucrium scordium in Nord/Pas-de-Calais. (including all the inland sites) were thought to be under threat at the time of survey - apart from one of the largest inland populations, at Nivelle, which was not assessed. Essentially, and not surprisingly, the largest populations are also considered to be the most secure/least threatened, while small populations (all those having less than a thousand plants) are thought to be endangered or ‘critically restricted’ (Figs 5 & 6). It is clear that the remaining inland populations, in particular, are under severe threat: only three sites support over 100 plants, and not even these can be considered safe; furthermore, three sites thought to be still extant are now presumed extinct as no plants of T. scordium were found during the survey. CONCLUSION Our investigations have shown the very critical situation of inland populations of Teucrium scordium in Nord/Pas-de-Calais. Most of them are already endangered or presumed extinct, and only one of them currently benefits from site protection and suitable management. As for many wetland plants, the main threats to T. scordium are water pollution, drainage, agricultural intensification and the planting of poplars. On the coast, we have shown that dune slack populations have intermediate morphologies with the Mediterraneo-Atlantic subsp. scordioides , while one population could even be referred to this taxon. Three protected sites are harbouring important populations, and we consider these to be secure, at least in the short to medium term. We know of four other coastal sites supporting T. scordium , but they support only very small populations and their long-term safeguard is uncertain. The protection and proper management of the remaining inland populations and the small coastal populations of T. scordium is a priority. Unfortunately, ex situ conservation often seems, at 268 ATLANTIC ARC first sight, to be the only immediate way to preserve those tiny (often isolated) colonies occurring in pockets of suitable habitat on private land; far more effective, however, would be to work with site managers and owners to get them to manage their land sympathetically - only then will we be able to secure a long-term future for T. scordium in northern France. ACKNOWLEDGMENTS We thank Yves Peytoureau, Secretary, Societe Botanique du Centre-Ouest, for reading the English text of this paper; thanks too to the referees and one of the editors of the present volume (S.J.L.) for their comments on an early draft. REFERENCES BLONDEL, C., VALENTIN, B., Destine, B. & Jorant, J.-A. (2001). Plan de conservation de la Germandree des marais (Teucrium scordium L.) pour la region Nord/Pas-de-Calais. Unpublished report (111 pp. + annexes). Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul pour l’Union Europeenne, le Conseil Regional Nord/Pas-de-Calais et la D.I.R.E.N. Nord/Pas-de-Calais. LEGRAND, L. (1996). Etude taxonomique et biometrique des populations de Teucrium scordium L. sur le territoire du C.R.P./C.B.N.B.L. Unpublished report (74 pp. + annexes). Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul. 269 Restoration and conservation management for the habitat of Mibora minima (L.) Desv. (Early Sand-grass) in northern France F. BASSO & F. HENDOUX Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul, Hameau de Haendries, 59270 Bailleul, France * ABSTRACT Mibora minima (L.) Desv. (Poaceae) is a widely distributed species in Europe, but is very rare and threatened in the north of France. In this paper we describe the distribution and ecology of M. minima on sand dunes in the Nord/Pas-de-Calais region of France and neighbouring dune systems in Belgium and the Netherlands. The species occurs in grasslands on poor sandy soils, in short vegetation having a high cover of bryophytes and low cover of associated vascular plants. The management regime of dunes in the Nord/Pas-de-Calais region is different on each site, but measures are now in place to restore and manage dune grasslands in a way that will maintain areas of suitable habitat for M. minima. The great potentialities of these sites, and the dispersal capacity of Mibora, suggest that we do not have to develop specific management for this species. On the other hand, the removal of shrub vegetation has to be encouraged through the cutting back of scrub and the grazing of dune grassland by sheep, cattle, or even horses to keep those areas free of scrub. Cutting/mowing for hay may also be beneficial. However, rabbit grazing would seem to be the best way to maintain the right habitat conditions for M. minima. Keywords: ecology, habitat conservation and management, sand dunes, Nord/Pas-de-Calais. DISTRIBUTION AND FREQUENCY OF MIBORA MINIMA Mibora minima (L.) Desv. (Early Sand-grass) is a tiny, autumn-germinating, early-flowering annual grass which is widely distributed in Europe (Dupont 1962), having a predominantly Atlantic-Mediterranean distribution. On the Atlantic coast of Europe it is recorded from Belgium, the Netherlands, France, the Iberian peninsula, Germany and Britain; elsewhere, it is recorded from Austria, Italy, the Balkan peninsula and N. Africa (Meusel et al. 1965). Records from the floristic data bank of the Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul (C.R.P./C.B.N.B.L.) show the distribution of M. minima in the north of France (Fig. 1), in the regions of the C.R.P./C.B.N.B.L. territory, that is to say Nord/Pas- de-Calais (NPdC), Picardie (Pic) and Haute-Normandie (HN). There are records from several periods, from its first mention in the 1920s right up to the present day. The species is observed in both coastal and inland localities, the latter especially in Picardie and Haute-Normandie. In Picardie, where M. minima is very rare and vulnerable (C.R.P.-C.B.N.B.L., unpublished), it occurs on coenozoic acidic sands near Laon (in the south-west of the region) and in the south of the Oise department. In Haute-Normandie, it is also considered to be rare and vulnerable (Collectif de Botanique 2000), with inland localities along the Seine River on old muddy-sandy alluvial terraces. In Nord/Pas-de-Calais (Boullet et al. 1999) M. minima is considered to be a native species, and as in the other regions appears to be under some threat due to changes to its coastal dune habitats. Habitat management of these dune sites is now being introduced, in an attempt to restore suitable areas of open dune grassland. *e-mail: crp.cbnbl@wanadoo.fr 270 ATLANTIC ARC 25 24 23 22 21 0 1 2 3 4 FIGURE 1. The distribution of Mibora minima in the north of France. C.R.P./C.B.N.B.L.: DIGIT ALE Nord/ Pas-de-Calais, Systeme d’lnformations Floristiques et Phytosociologiques. Symbols denote date-classes of records. STUDY SITES The Mibora minima populations we studied more particularly, and from which our releves were recorded, are found in the following coastal sites of the Nord/Pas-de-Calais region (Fig. 2). THE DECALCIFIED OLD DUNES AT GHYVELDE (NORD DEPARTMENT) This dune system is about 5 km long and 300-600 m wide, extending to about 200 ha on both sides of the Franco-Belgian border, among the polders of maritime Flanders. The sea is almost 3 km away. These dunes are a vestige of a sandy sea-cordon, submerged opposite a former shore. Sand was deposited offshore, parallel to the coast, during the Flandrian transgression, from the end of the last glaciation (8000 B.C.) The cordon emerged progressively and isolated the hinterland from the sea, and the sandy substrate is nutrient-poor and leached (lixiviated), with very low MIBORA MINIMA CONSERVATION IN NORTHERN FRANCE 271 22 21 0 1 2 3 A Figure 2. Sand dune sites of Mibora minima in the Nord/Pas-de-Calais region. C.R.P./C.B.N.B.L.: DIGITALE Nord/Pas-de-Calais, Systeme dTnformations Floristiques et Phytosociologiques. Symbols denote date-classes of records. amounts of calcium carbonate. The vegetation and flora of the Ghyvelde Dunes are very unusual in the context of the Flemish maritime plain, and the site has a great value from the point of view of natural heritage. RECENT CALCIFEROUS LITTORAL DUNES OF FLANDERS (NORD DEPARTMENT) This group of sites, comprising about 500 ha of sand dunes, is situated between Dunkirk and the Franco-Belgian border, but geomorphologically it continues into Belgium (Westhoek dunes). The Flemish dunes are composed of three sites: from west to east, they are the Dewulf Dunes, the Marchand Dunes (Natural Reserve) and the Perroquet Dunes. Those dunes were formed by sand silt, parallel to the shore, as was the case with the Ghyvelde Dunes, from the 4th and 7th centuries to the Middle Ages, during the different phases of the Dunkerquian transgression. The prevailing west to south-west winds modelled their relief and built up high dunes (parabolic dunes for example). Soils are generally calciferous, and ground-water, reaching the surface in many places, supplies several ponds and dune slacks. CALCIFEROUS DUNES BETWEEN BERCK AND LE TOUQUET (PAS-DE-CALAIS DEPARTMENT) A very large (1300 ha) expanse of sand dunes covering three areas separated by towns and tourist complexes: the Mayville-Le Touquet Dunes, the Stella-Plage Dunes and the Merlimont Dunes. The latter includes the Cote d’Opale State Biological Reserve (450 ha) which is actively managed by the Forestry Board. The other dune areas are not currently being managed. This group belongs to the geomorphological complex of the maritime plain of Picardie, characterised by a border dune cordon, subject to strong erosion, and a large inter-dunal flooded plain, located in front of an older, inland cordon. 272 ATLANTIC ARC mm r 3 - 1 a cm 0 b FIGURE 3. Some differences between Mibora minima subsp. litorea (left) and subsp. minima (right) (reproduced from Ortez et al. 1999). (a): spikelets, (b): inflorescences. Within these three major dune complexes, the main populations of Mibora minima are to be found in the Ghyvelde Dunes. In the Nord/Pas-de-Calais region the species is now restricted to these coastal sites, having been lost from its former inland localities (Fig. 2). Intensive agriculture is probably the main reason for its disappearance. The dunes at Ghyvelde and Perroquet, and those between Berck and Le Touquet, are interesting European sites and are proposed for the European natural sites network. Natura 2000. TAXONOMY Two infraspecific taxa of Mibora minima are mentioned in the literature (Ortiz et al. 1999): M. minima subsp. litorea and M. minima subsp. minima. The main differences between the two subspecies are in the length and arrangement of the spikelets (Fig. 3). Subsp. litorea is described from sand dunes in the north-west Iberian peninsula. From the samples examined, we could not establish for certain which taxon is present on the dune systems of the Nord/Pas-de-Calaisa region; however, the individuals we examined (seven samples from different sites, spring 2003) showed relatively strong, thick spikelets, which is said to be characteristic of subsp. litorea. Clearly, more taxonomic investigations are needed; it would be interesting to extend these studies to include all populations of M. minima along the European Atlantic coast, including those in England and Wales. ECOLOGY Concerning ecological conditions (Zwaenepoel & Vanhecke 1995, and our own observations) M. minima tends to occur in short, open dune grassland vegetation. These grasslands can differ in the characteristics of the underlying soil, having a wide amplitude especially with regards to pH. From a phytosociological point of view, dune-grassland containing M. minima belongs to the Koelerion albescentis Alliance, being found in both the hemicryptophytic acidiphilous grassland of the Festuco filiformis-Galietum maritimi and the bryo-lichenic calciferous grassland of the Phleo arenarii-Tortuletum ruraliformis (Plate 40). Such communities are rare across the Nord/Pas- de-Calais region as a whole, though they may be locally quite common in the coastal sand dune systems (especially the Phleo-Tortuletum). MIBORA MINIMA CONSERVATION IN NORTHERN FRANCE 273 Phytosociological releves are presented in Table 1, to illustrate the botanical composition of the plant-communities in which Mibora grows in the Nord/Pas-de-Calais region of northern France (data collected from the Ghyvelde, Perroquet and Stella-Plage Dunes). Releves are arranged according to the pH of the substratum, from acidophilous-low acidophilous grassland (rel. 1 and 2 at Ghyvelde) to calciferous grasslands (rel. 3 and 4 at Perroquet and Stella-Plage). In rel. 1, which is close to Festuco filifonnis-Galietum maritimi , the bryophytic layer is very dense ( Polytrichum juniperinum, Dicranum scoparium). The herbaceous layer is less abundant ( Luzula campestris, Festuca filiformis, Rumex acetosella and Galium verum var. maritimum), but there may be a considerable cover of annual species in spring (Air a praecox, Teesdalia nudicaulis). Mibora can become locally abundant in this type of vegetation, but also occurs more widely in a grassland TABLE 1. BOTANICAL COMPOSITION OF DUNE GRASSLAND CONTAINING MIBORA MINIMA (COVER-ABUNDANCE SCORES USING BRAUN-BLANQUET SYSTEM) Sites Ghy Ghy Per SPI Releve No. 1 2 3 4 Area (m2) 4 2 2 2 Ht of herbaceous vegetation (cm) 1-15 1-10 1-15 1-10 Cover of herbaceous vegetation (%) 30 40 40 50 Cover of Bryophytes (%) 100 90 100 10 Cover of Lichens (%) 40 10 40 Mibora minima 12 22 22 22 Festuca filiformis + Galium verum var. maritimum 12 Rumex acetosella 22 + Aira praecox 11 r Teesdalia nudicaulis 2 r Luzula campestris 22 r 22 Carex arenaria 22 11 11 11 Senecio jacobaea r 2 2 r Festuca rubra subsp. arenaria 22 11 22 Phleum arenarium + 22 11 Cerastium semidecandrum 11 22 11 Erophila vema + + i Myosotis ramosissima + + r Arenaria serpyllifolia r 22 r Myosotis arvensis + + Viola curtisii r 11 Saxifraga tridactylites r Hypochaeris radicata + Polytrichum juniperinum 22 Dicranum scoparium 33 Polytrichum piliferum 12 Hypnum cupressiforme var. lacunosum 22 12 Racomitrium elongation + 44 Tortula ruraliformis 44 23 Brachythecium albicans 23 Homalothecium lutescens 23 * Ceratodon purpureus var. purpureus 11 Tortella flavovirens 12 Bryum argenteum r Lichens (all spp.) 33 11 33 274 ATLANTIC ARC community characterised by Racomitrium elongatum 1 (rel. 2). The latter community has only been observed in the Ghyvelde Dunes, where it could be a substitute for Phleo-Tortuletum on soils that are either decalcified or in the process of decalcification. More phytosociological investigations are required in order to describe this vegetation properly. In the Perroquet Dunes (rel. 3) and the Stella-Plage Dunes (rel. 4) the bryo-lichenic grassland of Phleo arenarii-Tortuletum ruraliformis is widely developed. Tortula ruraliformis constitutes the main component of the bryophytic layer, with Brachythecium albicans and Homalothecium lutescens. The herbaceous layer is characterised by Phleum arenarium and Viola curtisii (very rare in Nord/Pas-de-Calais and a protected species in France). This grassland can present fragmentary (Stella-Plage) or more structured forms (Perroquet). Mibora may be found frequently in these grasslands, its cover reaching even 25-30 % in small patches. At Ghyvelde, horses (Haflinger race) have been grazing grassland all year round since 1999, together with rabbits, and the site is closed to public access. On other sites, there is no horse grazing, but, as in the Perroquet Dunes, grassland is mown and/or grazed by rabbits, or, as in Stella-Plage Dunes, only grazed by rabbits. In the Perroquet Dunes, relict grassland with Mibora also remains on the sides of infrequently used footpaths. In all cases, the species occurs in more or less open grassland (with or without mosses). When herbaceous vegetation becomes dense, Mibora becomes rare and may quickly disappear. Clearing of shrub vegetation Ligustro vulgaris - Hippophaetum rhamnoidis grasslands Calciferous sands Lixiviated sands Phleo arenarii- Tortuletum ruraliformis Mibora minima SOIL SEED-BANK? Figure 4. Restoration of grasslands after clearance of dune scrub. Festuco filiformis- Galietum maritimi CONSERVATION AND MANAGEMENT OF THE MIBORA MINIMA HABITATS According to the particular ecology of Mibora minima in the dunes of northern France (oligotrophic, calciferous or decalcified sands), the main threats to the species appear to be eutrophication, competition from taller/denser vegetation and/or growth of scrub vegetation, and outright destruction of vegetation by natural erosion or due to too much trampling by humans or animals. lRacomitrium elongatum differs from R. canescens by longer and unforked nerve, ending in a white hair point. R. elongatum is calcifuge (Pierrot 1994). M1B0RA MINIMA CONSERVATION IN NORTHERN FRANCE 275 Currently, there is no programme of management or conservation aimed specifically at ensuring the survival of M. minima , but we have found that the routine habitat management at these sites is generally successful in maintaining and restoring areas of dune grassland suited to this species. For example, dune scrub ( Ligustro vulgaris-Hippophaetum rhamnoidis) has developed and covered many areas, due to a variety of factors (including the decline of rabbit populations owing to the myxomatosis epidemic), but these areas are now being cleared for grassland restoration (Plate 41). Grasslands regenerate from a substratum without any vegetation (as is the case in areas formerly occupied by dense Hippophae scrub) and from disturbed soil, after clearing. These open grasslands evolve towards Phleo-Tortuletum or Festuco-Galietum communities, depending on the soil type (Basso 2001-2002; Basso & Baliga 2000-2001) (Fig. 4). The grassland species may recolonise from wind-dispersed seeds or from the soil seed-bank. M. minima might also be a soil seed-bank strategy species (Zwaenepoel & Vanhecke 1995). CONCLUSIONS Thanks to the current state of the Mibora minima populations in the different sites, and according to the great potentialities for its habitats and its colonisation capacity, we do not think it necessary to develop specific conservation measures for this species. On the other hand, restoration efforts in clearing scrub and favouring dry grasslands are clearly beneficial to this species, and should continue. The importance of rabbit grazing should not be underestimated either. Indeed, we suspect that rabbit grazing alone could be sufficient to maintain suitable vegetation communities for M. minima , if the size of their populations could be increased - although repeated outbreaks of myxomatosis, combined with hunting pressure, are important factors currently limiting rabbit populations. Other forms of management intervention, including grazing (e.g. horse grazing on the Ghyvelde Dunes) and mowing for hay, may also be useful management measures, though eutrophication effects of livestock grazing have been little studied and are poorly understood. M. minima tolerates light trampling, and footpaths may contribute to its conservation by providing patches of open habitat and potential corridors for its spread. ACKNOWLEDGMENTS We wish to thank Yves Peytoureau and Simon Leach for their help in checking the English text for this paper. REFERENCES BASSO, F. (2001-2002). Bilan et evaluation de la gestion de la xerosere de la Reserve Naturelle de la Dune Marchand. Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul pour le Conseil General du Nord, Bailleul. BASSO, F. & Baliga, M.-F. (2000-2001). Bilan et evaluation de la gestion de Vhygrosere de la Dune Marchand. Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul pour le Conseil General du Nord, Bailleul. Boullet, V., Desse, A. & HENDOUX, F. (1999). Inventaire de la flore vasculaire du Nord/Pas-de-Calais (Pteridophytes et Spermatophytes): raretes, protections, menaces et statuts. Version no. 2/25.10.1998. Bulletin de la Societe du Nord France 52(1): 1-67. COLLECTIF De BOTANIQUE (2000). Inventaire de la flore vasculaire de Haute -Normandie ( Pteridophytes et Spermatophytes): raretes, protections, menaces et statuts. Versionl/1 1.05.2000, DIREN Haute- Normandie, Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul. Bailleul. DUPONT, P. (1962). La flore atlantique europeenne. Introduction a V etude du secteur ibero-atlantique - Documents pour les cartes des productions vegetales, ser. Europe-atl., Vol. 1. Faculte des Sciences, Toulouse. MEUSEL, H., JAGER, E. & WEINERT, E. (1965). Vergleichende Chorologie der zentraleuropaischen Flora. Gustav Fischer, Jena. Ortiz, S., Rodriguez-Oubina, J. & Guitian, P. (1999). Taxonomic characterization of littoral sabuline populations of Mibora minima (Poaceae) in the northwestern Iberian Peninsula. Nordic Journal of Botany 19(5): 581-586. 276 ATLANTIC ARC Pierrot. R. B. et al. (1994). Contribution a l’inventaire de la bryoflore fran9aise (annee 1993). Bulletin de la Societe Botanique du Centre-Ouest, Nouvelle serie, 25: 369-371. Zw AENEPOEL, A. & VANHECKE, L. (1995). The distribution, ecology and phytosociology of Mibora minima (L.) Desv. in Belgium. Biologisch jaarboek dodonaea, 1994 62: 134-156. 277 Viola kitaibeliana Schult. (Dwarf Pansy) in the Isles of Scilly R. E. RANDALL Girton College, Cambridge, CB3 OJG, England* ABSTRACT This note summarises historic and recent records of Viola kitaibeliana Schult. (Dwarf Pansy) in the Isles of Scilly. It occurs here, as elsewhere on the Atlantic coasts of W. Europe, in short, open turf on drought-prone sandy soils by the sea. Rabbit grazing, human trampling and other forms of ground disturbance are important in maintaining the tightly-grazed, open-textured or ‘patchy’ swards in which this species grows. Footpath management, and the cutting of vegetation on sand dunes, has resulted in the appearance of several new V. kitaibeliana colonies. KEYWORDS: sand dunes, Festuca rubra-Galium verum fixed dune grassland (SD8), rabbit grazing, trampling. INTRODUCTION Viola kitaibeliana Schult. (Dwarf Pansy) is a tiny spring-flowering annual restricted within the British Isles to the Channel Islands (Guernsey, Herm and Jersey) and the Isles of Scilly (Plate 42) (Tresco, Tean, Bryher and possibly St Martin’s). It is listed as Vulnerable in the British Red Data Book (Parslow & Wigginton 1999). Its wider European distribution is summarised by Randall (2004). Along the Atlantic seaboard of western Europe V. kitaibeliana occurs (as in the British Isles) as a native of sand dunes and short turf on sandy soils, from the Iberian peninsula northwards to north-east France; its northernmost locality in Europe is said to be near Ambleteuse, Pas-de-Calais (Lousley 1956). Elsewhere, it is widespread in southern and central Europe, extending eastwards to Syria, Turkey and the eastern Ukraine. It is not clear if V. kitaibeliana is native to N. America: in Indiana, for example, it is regarded as an alien of European origin, while in Kansas it is thought to be probably native. As a garden plant in the U.S.A. it goes by the name of ‘Johnnyjump Violet’. V. kitaibeliana includes V. nana (DC.) Godr. Other commonly used synonyms outside Britain are V. rafinesquii Greene, V. minima Presl. and V. valesiaca Thomas. In north-west Spain and Portugal the smaller V. kitaibeliana var. henriquesii (Willk) Beck and a more robust var. machadiana (Coutinho) Coutinho have been recognised (Franco 1971). In the British Isles, the first known record of V. kitaibeliana was from Trimen’s visit to the Channel Islands in 1871 (McClintock 1975). The earliest published record for the Isles of Scilly is Cumow’s (1876) [as V. curtisii]; the species was discovered on Tresco by J. Ralfs in July 1873 and named by him as V. curtisii var. mackaii in a letter to F. Townsend in 1876 (Ralfs 1876). It was recognised as V. nana Corb. by Baker (1901) but Babington (1881) correctly described it as a “small form from Scilly... very like V. parvula Tin.” HABITATS AND PLANT COMMUNITIES MAINLAND EUROPE Throughout its European range, Viola kitaibeliana typically occurs on well-drained siliceous, sandy soils. In Hungary, its type location, V. kitaibeliana is an early coloniser, with other therophytes, of sand steppe grassland. In Cyprus, Meikle (1977) described it as occurring in vineyards, gardens and fields or in bare, sandy places on grassy hillsides between 625-1250 m. *e-mail: rerlO@cam.ac.uk 278 ATLANTIC ARC Strid (1980) found it widely on Mount Olympus, Greece, in similar locations to V. arvensis, while elsewhere in Greece it is a plant of dry, open habitats (Strid 1986). In Turkey it is recorded on stony slopes, screes and ‘macchie’ from sea level to 1800 m (Davis 1965). Along the Atlantic coasts of Portugal, Spain and France, it typically occurs in sand dunes and open sandy grassland near the sea. ISLES OF SCILLY In the Isles of Scilly, V. kitaibeliana is a plant of sandy soils, where the wind-blown sand is derived from granite and the dominant constituents are grains of quartz and feldspar (Randall 2004). The climate of the Isles of Scilly is quite different to that found in the central and eastern European parts of the range of V. kitaibeliana , but is similar to that of the Channel Islands and coasts of the adjoining European mainland (e.g. Brittany, Biscay): the climate is markedly ‘oceanic’ in nature, with relatively equable temperatures (February mean 7-3°C; August mean, 16-5°C), a regular supply of precipitation (mean 810 mm/year), near freedom from snow and frost (mean of 350 days/year with min. air temperature 5°C or above) and high incidence of sunshine (mean of 1752 hours/year). In those rare years when frosts have been recorded, V. kitaibeliana may be killed in exposed locations, and populations then have to re-establish from buried seed. Severe drought may also be a problem in some years, with low precipitation in spring leading to poor flowering in at least one population (Rushy Bay, Bryher) studied in the 1990s. A common feature of all its sites on the Isles of Scilly is that they are open and sunny, with plenty of bare ground for seeds to germinate, and with surrounding vegetation low enough for competition for light to be limited. V. kitaibeliana is found here in open, sandy grassland or grass- heath, or in derelict arable fields near the coast. It may be quickly shaded out where encroachment by Pteridium aquilinum occurs, and where the dune grassland is ‘closed’ V. kitaibeliana plants tend to be depauperate but usually survive to flowering and seed-set. Rabbit grazing (and burrowing) is important in maintaining the short, open turf, and the preferential grazing of taller (often perennial) species helps to reduce competition. It is notable that several of the largest, most persistent, colonies are in areas intensively used by rabbits. Many Viola kitaibeliana populations occur on ground adjacent to footpaths or track-ways and are consequently trampled by people: this disturbance produces a relatively open sward and helps to keep potential competitors at bay. Trampling may also aid seed dispersal. For example, in 1986 plants were recorded at a new site on Bryher, away from the main population but along a frequently used footpath, and it is speculated that seed may have reached the new site with human assistance; equally, on Tean, several colonies have recently appeared along footpaths and cleared areas created by the Isles of Scilly Wildlife Trust. On Bryher, where paths have been mown through taller dune vegetation, it seems to grow particularly well along the edge of the mown areas. Viola kitaibeliana is associated with a wide range of species. Randall (2004, Table 2) lists species occurring with V. kitaibeliana on the Isles of Scilly, concluding that the vegetation stands here are referable to the Luzula campestris sub-community of the Festuca rubra-Galium verum fixed dune grassland (SD8) (Rodwell 2000). The commonest associates are Stellaria pallida , Cerastium fontaum , C. diffusion, Sagina procumbens, Aphanes australis, Lotus comiculatus. Trifolium occidental. Euphorbia portlandica. Geranium molle, Myosotis ramosissima, Veronica arvensis, Galium verum, Hypochaeris radicata, Senecio vulgaris, Festuca rubra and the moss Syntrichia ruraliformis. Records of Viola kitaibeliana on the Isles of Scilly suggest that it is usually occurs only temporarily, or sporadically, in sites opened up by human disturbance, but that ‘naturally’ open locations - such as in areas grazed and disturbed by rabbits, or in patches of semi-fixed dune grassland where there is still some sand movement - may hold more persistent populations. Numbers of plants can vary greatly from one year to the next: at a few sites it may be very abundant in some years, with up to several thousand plants in just a few square metres (Parslow & Wigginton 1999). Currently, the largest population in Scilly extends over c. 200 m2, at Rushy Bay, Bryher. VIOLA KITAIBELIANA IN THE ISLES OF SCILLY 279 SUMMARY OF RECORDS FROM SCILLY Details are given below of the first records of Viola kitaibeliana at each site on the Isles of Scilly, together with the results of recent surveys to check on the current status of all known populations. BRYHER Occurring in semi-stable dune grassland, heavily rabbit-grazed. Rushy Bay (SV876142): 1st record 1951 (Tresco Abbey Herbarium); still present in 2003. Rushy Bay (SV875142): 1st record 1986; still present 2003. Very close to the last site. Great Porth (SV8714): 1st record 1986; not present 2003. ST MARY’S Pellistry Bay (SV928119), sandy field near shore: 1953 (Lousley 1971), but not seen at this site since 1957. ST MARTIN'S Sandy fields close to shore. Field below Chapel Down (SV937153): 1st record 1990 (P. Sargeant, pers. comm.); not present 2003. Field below school (SV9215): 1st record 1887 (Ralfs’ herbarium, Kew); seen at intervals since then (Lousley 1971), but not present in 2003. Presumed extinct on the island, but two plants reported in 2003 after an absence of several years (G. Fenton pers. comm.) - exact location uncertain. TEAN Dune grassland and track-ways. West Porth (SV906165): 1st record 1953 (Lousley 1971), said to be ‘plentiful’; seen again in 1982 (R. Parslow pers. comm.) and 1992 (R. E. R.); not present 2002, but refound in 2003. Black Porth (SV909167): 1st record 1988; not present in 2003. Black Porth (SV909166): 1st record 1990; not present in 2003. East Porth (SV909164): 1st record 2002; still present in 2003. Tean Sound Dune (SV910164): 1st record 2002 (R. E. R.); not present in 2003. Tean Sound Bay - tracks (SV911164 and SV911165): two colonies, 1st record 2002; both still present in 2003. TRESCO Sand dunes. New Grimsby (SV 889150): 1st record 1873 (Ralfs 1876); not recorded from this locality since 1914. Borough (SV899149): 1st record uncertain, but still extant here in 2002. Pentle Bay (SV899144): 1st record 2000; not present in 2003. Appletree Banks (SV899139): 1st record 1936 (Raven, pers. comm. 1980); present in 2003. PHENOLOGY Viola kitaibeliana is a winter-annual therophyte, with seeds germinating in the autumn and young plants continuing to grow through the mild Scillonian winter. Knight (1996) shows it as ‘possibly in flower’ as early as February on Scilly; Lousley (1971), on the other hand, recorded it starting to flower at the end of March and in full flower in April and May, while Parslow & Wigginton (1999) have it flowering in late March and April. Individual plants can flower for 4-5 weeks (Randall 2004). In draughted habitats on sand dunes and in coastal turf, plants can shrivel up and disappear quite early in the season (before the end of May); in arable fields or on damper sites, however, late-flowering plants may persist into June or even early July. McClintock (1975) noted that it was in flower in the Channel Islands as late as 5th July in 1902, but there (as on Scilly) it has usually finished flowering by early June. 280 ATLANTIC ARC AC KNO WLEDGMENTS Thanks to Will Wagstaff and Rosemary Parslow for unpublished data, and to Chris Rolfe and Steve Boreham for their assistance with the soil analyses. I would also like to thank one of the editors of this volume (S.J.L.) for helpful improvements to an early draft of the manuscript. REFERENCES BABINGTON, C. C. (1881). Manual of British Botany , 8th ed. London. BAKER, E. G. (1901). Some British violets. Journal of Botany, 39: 9-12. CURNOW, W. (1876). A botanical trip to the Scilly Isles. Hardwicke’s Science Gossip 1876: 162. DAVIS, P. H. (1965). Flora of Turkey and the East Aegean Islands. Vol. 1. University Press, Edinburgh. FRANCO, J. do A. (1971). Novo Flora de Portugal. Sociedade Astoria Lda, Lisboa, Portugal. Knight, L. (1996). Isles of Scilly Flora Checklist. Isles of Scilly Environmental Trust, St Mary’s. LOUSLEY, J. E. (1956). Le Monde des Plantes, 319: 19. LOUSLEY, J. E. (1971). The Flora of the Isles of Scilly. David & Charles, Newton Abbot. MCCLINTOCK, D. (1975). The Wild Flowers of Guernsey. Collins, London. Meikle, R. D. (1977). Flora of Cyprus. Bentham-Moxon Trust, Royal Botanic Gardens, Kew. Parslow, R. & WiGGINTON, M. J. (1999). Viola kitaibeliana Schultes, in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants, p. 388. Joint Nature Conservation Committee, Peterborough. Ralfs, J. (1876). Letter to F. Townsend dated 2 October, 1876 in library of South London Botanical Institute. RANDALL, R. E. (2004). Biological Flora of the British Isles. Viola kitaibeliana Schult(es). Journal of Ecology 92: 361-369. RODWELL, J. S. ed. (2000). British Plant Communities. Volume 5. Maritime Communities and Vegetation of Open Habitats. Cambridge University Press, Cambridge. STRID, A. (1980). Wild Flowers of Mount Olympus. Goulandris Natural History Museum, Athens, Greece. STRID, A. (1986). Mountain Flora of Greece. Cambridge University Press, Cambridge. 281 Polygonum maritimum L. (Sea Knotgrass) in Cornwall R. J. MURPHY Shangri-la, Reskadinnick, Camborne, Cornwall TR14 OBH, England ABSTRACT This note summarises historic and recent records of Polygonum maritimum L. in Cornwall. The species is erratic in its appearance, but recent Cornish records at Gunwalloe Church Cove and Par Beach, along with its discovery at several new sites elsewhere along the south coast of England, suggest that more extensive searches of beaches in Cornwall may prove fruitful. Keywords: conservation, distribution, population counts. INTRODUCTION One might expect that a plant at the northern limit of its European range - and with a habitat as tourist-ridden in summer and storm-swept in winter as a Cornish beach - would be, at very best, somewhat erratic in its appearance. This is certainly the case with Polygonum maritimum (Plate 43). Belonging to the Mediterranean-Atlantic element of the British flora (Preston & Hill 1997), this species occurs “...throughout Europe on Atlantic, Mediterranean and Black Sea coasts, extending northwards to the Channel Islands, Britain and Belgium” (Walls & Wigginton 1999). In Britain, P. maritimum receives special protection by being listed on Schedule 8 of the Wildlife & Countryside Act (1981), but its distribution (and persistence) may be largely beyond human control since it favours coarse sand or shingle on exposed storm beaches where there is much shifting of beach material. It has made appearances at one time or other in various counties along the southern coasts of England, from the Isles of Scilly to E. Sussex. It also occurs at a single site in Ireland (Curtis & McGough 1988). In recent years P. maritimum has turned up at several new localities (and reappeared at several old ones): since 1987 it has been reported from sites in Cornwall, Dorset, Hampshire, the Isle of Wight, W. Sussex and E. Sussex (Harmes & Spiers 1993; Walls & Wigginton 1999; Akeroyd 2002). There is a suggestion that this trickle of new sightings could represent a real increase, perhaps correlated with the recent run of mild winters and hot summers (Akeroyd 2002). THE CORNISH RECORDS Styles (1962) reported that the most recent British record of P. maritimum had been a Cornish one, from “near Mullion”, in about 1939, but he considered that the species had since become “...extinct on the mainland of Great Britain”. Prior to 1939 there had been records from several Cornish localities (Fig. 2): Swanpool (SW83) [1897], Loe Bar (SW62) [1899], Gunwalloe Church Cove (SW62) [1925-1935], Poldhu Cove (SW61) [1936], and Par Beach (SX05) [1935]. Despite its loss from Cornwall and other southern coastal counties of England, Styles (1962) struck a more optimistic note when he stated that he saw no reason why the plant “could not become established again... given the right conditions”. This optimism seemed to be borne out by the discovery in 1972 of a few plants at a new locality at Lantic Bay (SX15). Subsequent records from this site are summarised in French et al. (1999). At first the population grew in size - the species was recorded there again in 1978, and then in 1982 when P. maritimum was said to be plentiful. It then began to decline, becoming much more intermittent in occurrence, with just a single plant being seen there in 1990. It was seen again in 1991, but not since. 282 ATLANTIC ARC k 1 0 9 8 7 6 5 4 3 2 FIGURE 2. Polygonum maritimum in Cornwall. O Previous records, no longer extant, • 2000+: still at Gunwalloe Church Cove (SW62) and Par (SX05). Severe winter storms in 1989-90 had a marked effect on many Cornish beaches, shifting sand and shingle to an enormous extent, and it is thought that this disturbance, together with beach clearing work carried out by the National Trust, may have been responsible for the reappearance of P. maritimum in 1990 at Gunwalloe Church Cove (Murphy 1991). This colony has flourished, with 250 plants recorded there in 1993 (Wigginton 1999), “more than 500" in 1998 (French et al. 1999), and 92 plants in 2000. In 1998 several plants appeared on Par Beach (SX05), while the following year two plants were found at Poldhu Cove (SW61), both old localities where P. maritimum had not been seen for more than 60 years. At present there are good populations at Gunwalloe Church Cove and Par Beach1. The latter site was visited on 12th May 2003 by a group of B.S.B.I. members who had been attending the ‘Atlantic Arc’ symposium, and they counted 15 plants. Given the sporadic nature of this plant’s occurrence and its apparent increase in recent years, a search of more of Cornwall’s storm beaches could prove rewarding. There is now a third Cornish colony: 26 plants were discovered in September 2004 on a stony revetment by the foreshore at Marazion (SW5130). The photo in Plate 43 was taken at this site. POLYGONUM MARITIMUM IN CORNWALL 283 REFERENCES AKEROYD, J. R. (2002). Polygonum maritimum, in PRESTON, C. D., PEARMAN, D. A. & DINES, T. D. eds. New Atlas of the British and Irish Flora , p. 187. Oxford University Press, Oxford. CURTIS, T. G. F. & McGough, H. N. (1988). The Irish Red Data Book. 1. Vascular Plants. Stationery Office, Dublin. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. HARMES, P. A. & SPIERS, A. (1993). Polygonum maritimum L. in East Sussex (v.c. 14). Watsonia 19: 271- 273. MURPHY, R. J. (1991). Polygonum maritimum L. in Cornwall. Botanical Cornwall 5: 27-29. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. STYLES, B. T. (1962). The taxonomy of Polygonum aviculare and its allies in Britain. Watsonia 5: 177-214. WALLS, R. M. & WlGGINTON, M. J. (1999). Polygonum maritimum L., in WlGGINTON, M. J. ed. British Red Data Books 1. Vascular Plants, 3rd ed., p. 296. Joint Nature Conservation Committee, Peterborough. 285 Mentha pulegium L. (Pennyroyal) in Britain and Ireland R. D. PORLEY English Nature, Crookham Common, Thatcham, Berkshire RG19 BEL, England * ABSTRACT This note discusses the current status of Mentha pulegium (Pennyroyal) in Britain and Ireland. Its decline as a native, mainly due to habitat loss, has been matched in recent years by its increase as an introduction as a contaminant of grass-seed mixtures. Establishing whether particular populations are native or introduced is an important prerequisite for effective conservation action under the U.K. Biodiversity Action Plan. Suspected or known alien populations seem to comprise plants that are taller and more robust than their native counterparts. Preliminary results of molecular studies suggest that introduced populations can be separated genetically. Keywords: conservation, native and introduced populations, genetic fingerprinting. INTRODUCTION Mentha pulegium is a European Southern-temperate species (Preston & Hill 1997) that is Red- listed as ‘Vulnerable’ in Britain (Chatters 1999) and included as a Priority Species in the U.K. Biodiversity Action Plan (U.K. Biodiversity Group 1998). M. pulegium is also included on Schedule 8 of the Wildlife & Countryside Act 1981, giving it full legal protection in England, Wales and Scotland. It has declined steeply in both Britain and Ireland, having been recorded as a native in a total of 293 10-km squares (regardless of date), but in only 24 squares during the period 1987-1999 (Walker 2002). A couple of additional presumed-native sites have recently been found in Devon, including a large population in dry cliff-top grassland on the South Hams coast, near Hallsands (SX8139), but overall this species continues to decline - a trend that had begun before 1930. HABITATS AND REASONS FOR RECENT DECLINES Mentha pulegium is a short-lived perennial that typically occurs on seasonally inundated or flushed ground overlying water-retentive clay soils. In southern England its stronghold in such habitats is in the New Forest (Hampshire), where it is a characteristic plant of grazed village greens and commons; but even here it has declined, and many populations have been lost. In Cornwall and elsewhere it grows on disturbed tracks and paths, while in Devon it may also be found in quite dry, rather species-poor sheep- or cattle-grazed coastal grassland. In Northern Ireland it occurs at a single site, in wet grassland along the margins of Lough Beg. Outright habitat loss and the demise of traditional management (especially grazing of commons and heaths) appear to be the main reasons for the decline of M. pulegium. When occurring in marginal, rather scruffy habitats, it is frequently at risk from development pressures, or from a human tendency to find such places aesthetically displeasing and to want to see them ‘tidied up’. RECENT MOLECULAR WORK TO DISTINGUISH BETWEEN ‘NATIVE’ & INTRODUCED' POPULATIONS In recent years there has been a marked increase in the number of Mentha pulegium records thought to be introductions (records from seven 10-km squares 1970-1986, and a further 46 *e-mail: ron.porley@english-nature.org.uk 286 ATLANTIC ARC Cluster 2 FIGURE 1: UPGMA analysis of Mentha pulegium , from Cowan, Devey & Fay (2002). Site details: 2, East End triangle (New Forest) SU369976; 3, Norley Marl Pits (New Forest) SU364976; 4, Cadnam Green (New Forest) SU294153; 6-7, Bovey Heathfield (Devon) SX832761; 8-11, Ardingly Reservoir (Sussex) TQ3329; 12-16, Lough Beg (N. Ireland) H970930; 17-18, Penhale (Lizard) SW688182; 19-20, Lowland Point (Lizard) SW793194; 21-24, Bray Pennyroyal field (Berkshire) SU915783; 26, 28, Eastham Ferry (Cheshire) SJ365817; 29-31, Anderton Nature Park, Northwich (Cheshire) SJ652751; 33-34, 36, Pawlett Hams (Somerset) ST2643 & ST2743; 37-38, Brechfa Pool (Brecknock) SOI 18377; 40-41, Hallsands (Devon) SX8 18397. Material collected in 2000, except for Hallsands which was collected in 2001. squares 1987-1999 (Walker 2002)). These alien populations probably originated from contaminated grass-seed mixtures, as they often appeared in newly sown grassland on roadsides, behind sea-walls and on reservoir margins (Kay 1996; Leach 1996; Briggs 1997). Several observers have noticed that probable introductions comprise plants that appear to be more robust and erect than their native counterparts, although there is some overlap with putative native populations. MENTHA PULEGIUM IN BRITAIN AND IRELAND 287 Our commitment to protecting native populations of M. pulegium meant that we needed, if possible, to distinguish between ‘native’ and ‘introduced’ populations. English Nature therefore asked Kew to carry out genetic fingerprinting studies on a range of material collected from Britain and Northern Ireland. The high level of genetic variability among the samples is unusual in the British and Irish flora. Four distinct clusters emerged (Fig. 1): clusters 1 and 2 consisted entirely of material gathered from presumed native populations, while cluster 3, and to a lesser extent cluster 4, contained samples from supposedly introduced populations. It is interesting to note that samples from dry cliff-top grassland at Hallsands, S. Devon, lie very clearly within one of the ‘native’ clusters, suggesting that our view of M. pulegium as a species of “seasonally inundated grassland... within and around ephemeral pools and runnels” (Chatters 1999) may need to be revised. In view of the rate of decline of native populations in recent years, and the wide genetic variability of populations (both ‘native’ and ‘introduced’), Kew’s interim recommendation was that every attempt should be made to protect all populations of M. pulegium, while giving priority to those populations thought to be native or probably so (Cowan et al. 2002). The genetic fingerprinting work is continuing, however, and since 2002 we have collected leaf samples from a number of other M. pulegium populations - including several on sites being considered for notification as S.S.S.I.s - in anticipation that molecular analysis will help us to determine whether such populations are likely to be recent introductions or truly native. ACKNOWLEDGMENTS My thanks to Robyn Cowan and colleagues at Kew for undertaking the molecular analysis, and to all those English Nature staff, B.S.B.I. members and others who assisted with the collection of leaf samples for analysis. REFERENCES BRIGGS, M. (1997). Non-native Mentha pulegium (Pennyroyal). BSBI News 74: 50. CHATTERS, C. (1999). Mentha pulegium L., in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular plants, pp. 244-245. Joint Nature Conservation Committee, Peterborough. COWAN, R. S., Devey, D. & Fay, M. F. (2002). Genetic fingerprinting studies of Mentha pulegium. Royal Botanic Gardens, Kew: unpublished report to English Nature. Kay, G. M. (1996). Mentha pulegium in grass seed. BSBI News 72: 46. Leach, S. J. (1996). Contaminants in grass seed. BSBI News 73: 23-25. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. U.K. BIODIVERSITY Group (1998). Tranche 2 Action plans. Volume 1. Vertebrates and vascular plants. English Nature, Peterborough. Walker, K. (2002). Mentha pulegium, in PRESTON, C. D. PEARMAN, D. A. & DINES, T. D. eds. New Atlas of the British and Irish Flora, p. 213. Oxford University Press, Oxford. 289 Hypericum undulatum Schousboe ex Willd. (Wavy St John’s-wort) in S.W. England and Wales R. D. PORLEY English Nature, Crookham Common, Thatcham, Berkshire RG19 8EL, England * ABSTRACT This note discusses the current status of Hypericum undulatum (Wavy St John’s-wort) in England and Wales. The species is thought to have declined in recent decades, mainly due to agricultural intensification or neglect, but there are now various conservation initiatives in place designed to protect this species and the western fen- meadow/rush pasture vegetation in which it is usually found. Keywords: distribution, conservation. Culm grassland, Rhos pasture. DISTRIBUTION Hypericum undulatum Schousboe ex Willd. (Clusiaceae) is an Oceanic Southern-temperate species (Preston & Hill 1997), restricted in Europe to the Atlantic fringe, from W. Spain and Portugal northwards to England and Wales. It also occurs on Madeira and the Azores. In Britain it is listed as ‘Lower Risk - Nationally Scarce’, having been recorded as occurring in 49 10-km squares during the period 1970-1992 (Leach & Wolton 1994), and in 55 10-km squares during the period 1987-1999 (Robson 2002). Surprisingly, it is unknown from Ireland, apart from a single (possibly mis-labelled) herbarium specimen from Glengarriff, W. Cork (Robson 1958, 2002). HABITAT AND PLANT COMMUNITIES H. undulatum occurs predominantly within the Cirsio-Molinietum, and in Britain most of its core populations are within the Juncus acutiflorus-Erica tetralix sub-community of the Molinia caerulea-Cirsium dissectum fen-meadow (N.V.C. community M24c (Rodwell 1991)). It is a characteristic species of damp, moderately species-rich Molinia and Juncus pastures in S.W. England (known locally as Culm grassland) and in W. Wales (rhos pasture). H. undulatum prefers areas where there is some lateral water movement and base enrichment, and requires bare ground for seedling establishment (Leach & Wolton 1994). Typical associates include Angelica sylvestris, Cirsium palustre, Filipendula ulmaria, Lotus pedunculatus, Mentha aquatica and Senecio aquaticus. Atlantic species such as Carum verticillatum, Pinguicula lusitanica and Scutellaria minor may also be found on sites supporting H. undulatum. Hypericum tetrapterum is a frequent associate, and can be difficult to separate vegetatively from H. undulatum , having forms with wavy leaves and variably winged tetraquetrous stems. However, the leaf width/length ratio is a reliable way to distinguish between the two species: samples collected by the author averaged 1:21 (range 1:1-3 — 1*8) in H. undulatum (n = 17, upper 2 rows of leaves in Pig. 1) and 1:1-5 (range 1:1-3— 1*8) in H. tetrapterum (n = 16, lower 2 rows of leaves). RECENT DECLINES AND CONSERVATION INITIATIVES H. undulatum is now better recorded than in the past, and this disguises to some extent the significant losses that have occurred, mainly due to habitat destruction or neglect. Lor example, in N. Devon a survey of land outside Sites of Special Scientific Interest (S.S.S.I.s) within a sample area of 40 km2 showed a 65% loss in the area of Culm grassland over the 5-year period 1984 to *e-mail: ron.porley@english-nature.org.uk 290 ATLANTIC ARC ^ k A ^ undulatum mmnm tetrapterum ••«##*** FIGURE 1 . Hypericum undulatum and H. tetrapterum - a comparison of leaf shape. 1989. Agricultural intensification - including drainage, ploughing, reseeding and the use of fertilisers - has clearly been a major problem in many areas; but ‘neglect’ (i.e. lack of management) has been a problem too, with some heathy pastures becoming too rank and overgrown for this species. Today, many remaining populations of H. undulatum in England and Wales are protected as S.S.S.I.s, and Molinia and Juncus pastures are listed as a Priority Habitat under the U.K. Biodiversity Action Plan. Importantly, there are now agri-environment schemes in place to provide farmers with an incentive to put grazing animals on the land. Optimal management for H. undulatum involves cattle grazing in summer and occasional burning in winter to remove litter accumulation. Seedling recruitment on disturbed ground is probably from a persistent seed-bank, and numbers of plants vary greatly from year to year. Getting neglected land back into good shape is proving beneficial for H. undulatum on some sites, with populations increasing rapidly once grazing and burning management has been reinstated. REFERENCES LEACH, S. J. & Wolton, R. J. (1994). Hypericum undulatum Schousboe ex Willd., in STEWART, A., PEARMAN, D. A. & PRESTON, C. D. eds. Scarce Plants in Britain, pp. 215-216. Joint Nature Conservation Committee, Peterborough. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. ROBSON, N. K. B. (1958). Hypericum undulatum Schousb. ex Willd. in Ireland? Irish Naturalists' Journal 12: 269. ROBSON, N. K. B. (2002). Hypericum undulatum, in PRESTON, C. D. PEARMAN, D. A. & DINES, T. eds. New Atlas of the British and Irish Flora, p. 213. Oxford University Press, Oxford. RODWELL, J. S. ed. (1991). British Plant Communities Volume 2: Mires and heaths. Cambridge University Press, Cambridge. 291 Conservation of Spiranthes romanzoffiana Cham. (Irish Lady’s-tresses) in Scotland - the role of twin lateral bud production R. GULLIVER, M. GULLIVER Carraig Mhor, Imeravale, Port Ellen, Isle of Islay, Argyll PA42 7AL, Scotland E. GRANT Drimcruy, Uig, Isle of Coll, Argyll, PA78 6TB, Scotland J. ROBARTS 233, Bruernish, Isle of Barra, Western Isles, Scotland M. KEIRNEN Grianan, Drumclach, Isle of Colonsay, Argyll, PA61 7YR, Scotland S. J0NCH M0LLER, A. BEARE Teglvaerksgade 6, 2th, Copenhagen 2000 0 Denmark and C. SYDES Scottish Natural Heritage, 2 Anderson Place, Edinburgh, EH6 5NP, Scotland ABSTRACT Mature capsule development in Spiranthes romanzoffiana Cham. (Orchidaceae) has never been observed in Scotland, suggesting that seed production in Scottish populations is now rare, extremely rare or non-existent, and may only involve small quantities of seed. Therefore, vegetative reproduction is likely to be of considerable importance. One way in which S. romanzoffiana plants reproduce vegetatively is through the production of twin lateral buds. On Colonsay in 2000 and 2001 rates of twin bud production in study populations were 2-4% and 1-8% respectively, whilst the incidence of twin plants was 4-3% and 3-0% (based on single visit data). Losses of one of the twin buds, and of the buds produced by established twin plants, occurred over time. On Coll, rates of twin bud production in 2000 and 2001 were much the same as on Colonsay (2T%, 3-7% and 0%; in samples), but the incidence of twin plants in samples in 2000 was 26-3% and 22-0% (single visit data). These results indicate that the relative importance of this form of vegetative reproduction may vary from island to island. Our results contrast with genetic studies which indicate very high genetic diversity in a population on Coll. Understanding the various ways in which this species reproduces in Scotland is essential, so that comprehensive conservation strategies and detailed management plans can be developed for its long term survival. Keywords: orchid, conservation, vegetative reproduction, grazing, buds. 292 ATLANTIC ARC INTRODUCTION Spiranthes romanzoffiana Cham. (Irish Lady’s-tresses) (Plate 44) is a U.K. Biodiversity Action Plan species (U.K. Biodiversity Action Group 1999). It was included in the second edition of the British Red Data Book (Perring & Farrell 1983), but following a minor spate of new records in the 1980s it is now listed as Nationally Scarce (Horsman 1994; Cheffings 2004). In Britain, between 1987 and 1999 there were records of the species from 17 10-km squares, all but one of these in western Scotland; whilst in Northern Ireland (where it is legally protected) it was recorded from eight 10-km squares over the same time period (Preston et al. 2002). The sole English locality is in Devon, on the western fringes of Dartmoor. The species is specially protected in the Republic of Ireland and included in the Irish Red Data Book (Curtis & McGough 1988). S. romanzoffiana has an unusual distribution. It is not known from any other locations in Europe, and at the European scale has a clear Atlantic distribution. It is, however, widespread in northern N. America (distribution map in Arft & Ranker 1998), and Preston & Hill (1997) assign it to their Oceanic Boreal-montane element. Historically there have been several explanations for this distribution. It may have survived in western Europe through the last Ice Age (concept of Hulten (1937); discussed in Heslop-Harrison (1953); and in Henderson (2001)). Alternatively, it may have colonised by seed or plant fragments carried by geese from N. America to Greenland, and then from Greenland to western Europe (concept of Heslop-Harrison (1953); advocated by Horsman (1998); discussed in Henderson (2001)). However, S. romanzoffiana has not been recorded from Greenland, and so may have colonised Europe directly by seed blown from N. America (Ettlinger 1997, p. 125), or carried on the feet or feathers of American vagrant waders blown off course on migration (F. Horsman pers. comm.). Observations of plants in autumn in Scotland have been made on Barra since 1995 (J.R.) (not reported here), on Colonsay since 1999 and on Coll since 2000 (E.G.) (Plates 45 & 46). Most of these observations have been of ‘capsules’, although in some instances these have been examined for seeds. Mature, fully-developed capsules such as occur in Spiranthes spiralis (Gulliver et al. 2004a in press) have not been observed in Scottish S. romanzoffiana; and the capsules that have been examined, including material from Coll, have never contained seed (e.g. C. Wilcock pers. comm.). However, in Ireland there is one indisputable recent case of capsules containing seed: in 2003 D. Lupton (pers. comm.) collected 46 seeds from a single plant growing on the southern shore of Lough Conn in Co. Mayo - the seeds were found in dried, dehisced material, but appeared to have originated from just two capsules. Pollen has been shown to be viable in Scottish plants. In experimental manipulations with recipient plants at sites GF and WC on Coll pollen tubes failed to fertilise the ovules up to 10 days after pollination (Wilcock 2002). Forrest (2001) and Forrest et al. (2004) have studied the molecular genetics of the species. Three different lines of evidence led to the recognition of two groups: a ‘northern’ group (chloroplast microsatellite haplotype A), comprising populations from Barra, Coll and Vatersay; and a ‘southern’ group (haplotype B), comprising the Colonsay populations together with all the sampled Irish populations, from Counties Antrim, Fermanagh, Londonderry, Tyrone, Galway, and Mayo. The results showed high genetic diversity in the northern populations, suggesting that sexual reproduction may be an important process in this group. On the other hand there was a low genetic diversity in the Colonsay and Irish samples, suggesting a highly uniform set of populations which could be explained in at least three possible ways: (1) a high incidence of vegetative reproduction; (2) sexual reproduction but with low diversity due to a previous bottleneck period; or (3) agamospermy. Based on our current state of knowledge, the observational (and some experimental) data imply a radically different level of importance of sexual reproduction for the northern populations in comparison with conclusions of the genetic analysis. Mature plants of S. romanzoffiana may be in one of three states: flowering, vegetative or underground. The underground phase can last up to six years (J.R.). The term ‘dormancy’ is not used in this paper because mycotrophy is likely to be taking place throughout the underground period. A lateral bud may arise alongside the stem of a flowering plant, alongside the ‘rosette’ of leaves in a vegetative plant or above the centre of the group of roots in an underground plant. Lateral buds normally appear above ground from July to October. Those which survive develop into either a flowering or vegetative plant during the following year. One means of vegetative CONSERVATION OF SPIRANTHES ROMANZOFFIANA IN SCOTLAND 293 reproduction is the production of two lateral buds on the same plant (an occasional occurrence); rarely there may be three such buds and very occasionally four (J.R. pers. obs.). Three American Spiranthes species can produce new plants from root tips (Catling 1990): two of these are stated to produce root buds (Correll 1950), i.e. to have both a) root buds which arise part way along the root, or at its tip, and b) lateral buds. Using phraseology as it is conventionally applied to the genus, root buds are different from lateral buds. Some authors have incorrectly used the term root buds when referring to the lateral buds in S. romanzoffiana. Additional material on the biology of the species can be found in Forrest (2001), Forrest et al. (2004), Gulliver et al. (2004, in press a and b), Gulliver & Gulliver (2004), Gulliver et al. (2003), Gulliver (2002), Henderson (2001) and Robarts (2000). These accounts contain further references and background information. METHODS The incidence of established (mature) twin plants (ETP) was recorded in two large populations at sites KA and KB on Colonsay in 2000 and 2001, and in one on Coll (site CA) in 2000. Data were also gathered from two smaller populations on Colonsay (sites KC and LF) but the interpretation of percentage values from these populations should be undertaken with caution. For concise site details see Gulliver et al. (2003). Mature twin plants had a gross morphology that was similar for each of the pair, though certain detailed features (e.g. number of flowers per inflorescence) sometimes differed. Twin plants may both be flowering (F/F), or one flowering and the other vegetative (F/V), or both vegetative (V/V). Three categories of twins were recognised. Category ‘a’ twin plants were broadly similar in the F/ F state in the current year and had a separation of less than 30 mm. Category ‘b' twin plants were F/V and V/V pairs with a separation of less than 30 mm, but where morphological similarities/ differences could not be determined with certainty. Data on plants with a separation of 30-100 mm are also presented (Category ‘c’). Further details are given in the relevant tables and in Gulliver (2002). The categories have been re-defined in comparison with those given in Gulliver (2002), but the totals (a+b+c) remain the same. The incidence of single lateral buds and new twin lateral buds (NTLBs) was recorded from Colonsay and Coll in 2000 and 2001 at the same time as the mature twin plants. The core data were gathered on a single visit in August/September, with selected data gathered on visits at other times of year. RESULTS ESTABLISHED TWIN PLANTS (ETPs) Detailed results of our surveys on Colonsay and Coll are given in Table 1. On Colonsay in 2000 at sites KA, KB, KC and LF (pooled) the incidence of category ‘a’, ‘b’ and ‘c’ plants (combined) was 4-3% in 2000 and 3-0% in 2001. On Coll in 2000, in 15-4 nr quadrats at site CA (sample 1) the maximum estimate of percentage pairs was 26-3% (one category ‘a' plant, 22 category 'b' and two category ‘c’). Again on Coll, for a sample of 47 flowering and 12 vegetative plants at site CA (sample 2), the maximum estimate of percentage pairs in 2000 was 22-0% (no category ‘a’ plants, ten category ‘b’ and three category ‘c’). No previous knowledge existed about the plants in sample 2. The area occupied by the population at site CA (c. 1-4 ha) was divided into 6 m strips and all inflorescences counted (n = 184) between 19 and 20 August 2000. Four pairs of flowering stems (considered to be twins) were located, including one pair also in the quadrats. The percentage incidence of F/F twins in this set of 184 plants was 2-2%. NEW TWIN LATERAL BUDS (NTLBs) Our records of the incidence of new twin lateral buds are summarised in Table 2. On Colonsay at sites KA, KB. KC and LF (pooled) in 2000 and 2001 the overall incidence of detected NTLBs was 24% and 1-8% respectively. On Coll the incidence of NTLBs in 15 quadrats at site CA (sample 1) 294 ATLANTIC ARC TABLE 1. DETAILS OF ESTABLISHED TWIN PLANTS (ETP), FLOWERING (F) AND VEGETATIVE (V) IN 2000 AND 2001 GROWING ON COLONSAY AND IN 2000 ON COLL IN THREE CATEGORIES", a) SEPARATION OF FLOWERING STEMS 30 MM OR LESS, RECORDED IN THE FIELD AS TWINS (SEE TEXT) b) SEPARATION OF STEMS AND/OR BUDS IN V/F AND V/V PAIRS 30 MM OR LESS c) SEPARATION OF STEMS AND/OR BUDS IN F/F, F/V AND V/V PAIRS BETWEEN 31 MM AND 100 MM: RECORDED ON A SINGLE VISIT IN AUGUST/SEPTEMBER (8/9): FURTHER DETAILS IN GULLIVER (2002) PLANTS DESCRIBED IN TABLE 2 ARE NOT INCLUDED5 Site codes Nature of pairs and separation of stems in Appendix 1 Number of pairs of plants Percentage pairs of plants (ETPs). (ETPs), In categories a), b), and c) i.e. as a percentage of all plants examined in 8/9 Year 2000 2001 Colonsay Site KA: 02-04/09/00 (n = 107): 30/08/-06/09/01 (n = 166) 2000 2001 a) 0 1 0 0-6 b) 1 1 0-9 0-6 c) 5 2 4-7 1-2 Colonsay Site KB: 31/08-05/09/00 (n = 73): 28/08/01 (n = 124) a) 0 0 0 0 b) 1 2 1-4 1-6 c) 0 1 0 0-8 Colonsay Site KC: 31/08/00 (n = 4): 27/08/01 (n = 15) a) 0 1 0 6-7 b) 0 0 0 0 c) 0 0 0 0 Colonsay Site LF: 30/08-02/09/00 (n = 26): 04/09/01 (n = 25) a) 1 1 3-8 4-0 b) 1 1 3-8 4-0 c) 0 0 0 0 Colonsay 4 sites pooled category ; a) only 1 3 0-5 0-9 category a), b) and c) 9 10 4-3 30 Coll Site CA Sample 1) 14/08/2000, (EG), 15 Quadrats (4 m") , (n = 95), a) 1 - 10 - b) 22 -■ 23-2 - c) 2 - 2-1 - category a), b) and c) 25 - 26-3 - Coll Site CA Sample 2) 21-22/8/2000 (n = 59: 47F & 12V)C a) 0 - 0 - b) 10 - 16-9d - c) 3 - 5-1 - category a), b) and c) 13 - 22-0 - Notes a Categories redefined cf. Gulliver 2002, totals (a+b+c) remain the same. b Also not included in this compilation are cases where only one member of a previously known ‘pair’ was present. c For all other entries the sample zone or site had been visited previously. d The sward immediately adjacent to the flowering stems was searched, but the rest of the area occupied by the two groups of flowering plants (R1 and R2, i.e. sample 2) was not searched. CONSERVATION OF SPIRANTHES ROMANZOFFIANA IN SCOTLAND 295 TABLE 2. NEWLY GENERATED TWIN LATERAL BUDS (NTLBS) PRODUCED BY FORMERLY SINGLE, MATURE PLANTS, RECORDED IN AUGUST/SEPTEMBER (8/9) NR = NOT RECORDED (DATES IN TABLE 1) Site name Number of new twin lateral buds (NTLB) detected - bold. Separations of buds (centre - centre) - plain text Total number of Percentage new twin lateral mature plants detected buds (NTLBs) with bud (8/9 visit only) separation less than 12 mm, generated during summer; i.e. as a percentage of number of mature plants examined in 8/9 Year Colonsay 2000 2001 2000 2001 2000 2001 SiteKA 3 3, 2,8 2 l,5a 107 166 28 12 Site KB 2 5,5 3 5,5,5 73 124 2-7 24 Site KC 0 0 4 15 0 0 Site LF 0 1 1 6C 26 25 0 40 Colonsay 4 sites pooled Coll Site CA: (EG) 5b 210 330 24 18 1 ) 15 4m2 Quadrats 2 5,5 5 NR 95 134d 2 1 3-7 2) Sample of 59 (47F & 12V) 0 NR 59 NR 0 NR Notes a Also three triplets. b For all sites pooled a total of 12 twin plants were generated anew in growing year 2000/2001 but in only five cases were the twin buds present in August/September 2000, see Table 3. c A further three twin buds (over all sites pooled) were detected in October 2001, see Table 3. d Includes 1 1 cases where basal leaves were present on 28-29 July; but only rotten leaves, stems, or buds were present on 20-3 1 August. in 2000 and 2001 was 2T% and 3-7% respectively, whilst there was a 0% incidence in the sample of 47 flowering and 12 vegetative plants (sample 2) in 2000. The distance separating these newly generated buds ranged for 1 to 8 mm (Table 2). CHANGES OVER TIME The time taken for twin buds to become separate individuals is not known. However, the progress of new twin buds and new twin plants was monitored over a 15 month period on Colonsay. Five pairs of buds were present in August/September 2000 (Table 3, part 1). One bud from the set of five twins which was present in October 2000 had disappeared (died or eaten) by February 2001. No further new twin buds were detected in October 2000 or February 2001. In May 2001 five pairs of small twin plants were located which had developed between February and May 2001 (in each case a single mature plant was present in this position in August/September 2000). These five were in addition to the four surviving sets of twins from the original sample. Two further pairs of new late developing twin plants were recorded in August/September 2001. Allowing for May to August losses, the net number of new twin plants in the growing year August 2000 to August 2001 was four. During the August/September 2001 survey six newly generated twin buds (12 buds in all) were detected (Table 3, part 2). By October, four of these (33-3%) were missing, either died or eaten. However, three new pairs of twin buds were detected at that time. One set of triplets (i.e. the standard one bud plus two others) was present in August/September and October. The survival/ loss of single buds at the base of twin plants in 2000/2001 is shown in Table 3 (part 3). 296 ATLANTIC ARC The increase in cumulative numbers of recorded pairs of twin buds from August/September 2001 to October (Table 3, part 2) shows the importance of multiple late summer and autumn site visits. At least two, and preferably three or more site visits are needed to reveal the scale of losses between late summer and the following spring. The separations of the twin buds recorded in August/September 2001 varied between 1 and 5 mm (Table 2). Given the close proximity of twin buds it seems highly unlikely that slug grazing is responsible for the loss of individual buds. If slugs were involved, we suspect that, having consumed one bud of the pair, they would then consume the other bud; yet in the plants examined only single buds were lost. Hence it is provisionally concluded that the loss of four buds in the ‘twins’ between August/September and October 2001 was probably not due to slug predation, but rather due to death from some other cause. TABLE 3. NUMBERS OF NEWLY GENERATED SECOND MEMBERS OF TWIN BUDS (TB), TWIN SMALL PLANTS (TSP), OR LATE DEVELOPING TWIN SMALL PLANTS (LDTSP) AT FOUR SITES, POOLED, ON COLONSAY (KA, KB, KC, LF), PLUS DETAILS OF A TRIPLET BUD (TRB), RECORDED THROUGH THE YEAR, SHOWING GAINS AND LOSSES (PARTS 1 & 2); AND DATA ON BUD SURVIVAL OF ESTABLISHED TWIN PLANTS (PART 3) THE TRIPLETS IN AUGUST 2001 CONSISTED OF TWO NEW BUDS IN ADDITION TO THE STANDARD LATERAL BUD Records from Status of Gains from Net numbers Cumulative Losses from newly previous of ‘new’ buds numbers of previous produced recording or small plants ‘new’ buds or recording plants period i.e. second small plants i.e. period members of second the pair members of the pair Part 1: 2000/2001 growing year Newly generated twin (second) buds or small plants 30 August-5 September 2000 TB 5 5 21-25 October 2000 TB 5 5 12-14 February 2001 TB 4 5 1 5-6 May 2001 TSP 5 9 10 27 August-6 September 2001 LDTSP3 2 4 12 7b Part 2: 2001/2002 growing year Newly generated twin (second) buds or triplets (in the second row, shown in square brackets) 27 August-6 September 2001 TB 6 6 27 August-6 September 2001 TRB [2] [2] 3-7 October 2001 TB 3 5 9 4 3-7 October 200 1 TRB [2] [2] Part 3: Survival/Loss of buds of twin plants present or newly generated in the 2000/2001 growing year with both mature plants present in May and/or August/September 2001 Each member of the pair of new twin plants (n = 12; May and August/September records) is theoretically capable of producing a bud. In seven cases both plants were recorded with a bud, their subsequent survival is indicated below Both buds survived from August/ September 2001 to October 2001 -4 One bud survived from August/ September 2001 to October 2001 -2 Neither bud survived from August/ September 2001 to October 2001 -1 Notes a Each of these late developing plants produced a bud. h Some of the 27 August-6 September losses may have been due to early death of above ground parts of the plant. CONSERVATION OF SPIRANTHES ROMANZOFFIANA IN SCOTLAND 297 RATES OF BUD PRODUCTION, BUD LOSSES AND INCIDENCE OF TWIN PLANTS On Colonsay there was an incidence of 2-4% new twin buds and 4-3% twins (categories a-c) in 2000 and 1-8% and 3-0% respectively in 2001. If all the newly generated twin buds had survived, the incidence of twin plants would have been many times that of the annual rate of production of new twin buds, and this was not the case. Important losses occurred throughout the growing year (Table 3), due either to a) death of the bud or b) grazing and the non-replacement of lost tissue. On Coll there was an incidence of 2-1% new twin buds and 26-3% twins (categories a-c) in the quadrats (sample 1) and 0% and 22-0% respectively in sample 2 in 2000. These results suggest a high survival of twins at this site. However the incidence of twins with both members flowering was still low (2-2%). There was an incidence of 3-7% new twin buds in the quadrats (sample 1) in 2001. DISCUSSION Comparisons can be validly made of the incidence of twin buds, either as a percentage of all buds at a single survey date or from a set of fairly close dates. However, it should be recognised that bud counts will vary in time as late-developing buds break the soil surface and existing above¬ ground buds are ‘lost’ as a result of death or being grazed to below ground level (Table 3). The recorded losses of one of a pair of twin buds and the linked finding of a low incidence of twin plants indicates that twin bud production is not a major means of vegetative reproduction in the populations on Colonsay. In theory the meristem of a bud which had been grazed to ground level may give rise to replacement tissue later in the growing season, but this phenomenon has not, as yet, been detected by us. Similarly, it is possible that a replacement twin bud might be generated in the following year. In addition to the losses that occur in the first year (Table 3), it is likely that there will be losses in subsequent years due to the death of one of the twin plants. Furthermore the observed incidence of twin plants may be affected by an alternation in appearance above ground of members of a twin pair. Such a phenomenon would be very hard to detect. On Coll at site CA there was a low rate of production of twin buds (2T% in sample 1 in 2000, 3-7% in 2001: 0% in sample 2 in 2000) but a high percentage of twin plants, suggesting a high rate of bud (and twin plant) plant survival. Thus, for this population, twin bud production might be an important means of vegetative reproduction. For Coll our results from site CA reveal a possible significant means of vegetative reproduction together with an absence of observed mature capsule development at sites CA, GF (e.g. Fig. 2) and WC. By contrast, Forrest (2001) and Forrest et al. (2004) in their molecular studies of material from site CA showed a very high genetic diversity in the population. (All the Coll material used by Forrest (2001) and Forrest et al. (2004) came from site CA.) These differing results emphasise the need for a deeper understanding of the reproductive system(s) used by the species. It is stated in Preston et al. (2002) that Spiranthes romanzoffiana has rhizomes. All orchids have rhizomes in the absolute sense. In S. romanzoffiana, as in the entire genus Spiranthes, the rhizome is vertical (Rasmussen 1995) and short (e.g. only a few millimetres long). All morphological investigations so far published have reported that the visible underground organs are roots (e.g. Godfery 1922, 1924, 1930; Mousley 1924 a and b; Butcher 1961; Roles 1965; Ross-Craig 1973). These are mycorrhizal. Rasmussen (1995, p. 1 29) has suggested that root fragments in the Orchidaceae may be capable of a period of mycotrophic existence and before developing into new plants. Such a mechanism would be an effective means of vegetative reproduction and its possible existence would repay further investigation. S. romanzoffiana does occur in wet habitats grazed by cattle and sheep (and slugs), so there is the strong possibility of periodic root severance by the hooves of stock, or by slugs grazing underground. The American species S. odorata ( =S . cernua var. odorata ), S. cernua and S. laciniata can produce new plants from the root tips (Catling 1990). Similarly, new plants arising from the tips of roots of S. odorata growing in cultivation have been reported (R. Manuel pers. comm.). These observations show the possibility of conversion of attached root tissue to stem tissue, a process which might also occur in detached root fragments. If such a mechanism occurred in S. romanzoffiana , new - but genetically identical - plants would become established across a range of distances from the mother plant; and some of these might be expected to have mother plant to daughter plant separation distances similar to those of twins. 298 ATLANTIC ARC A major part of the efforts to resolve the issue of the balance between vegetative and sexual reproduction in S', romanzoffiana should be directed at the use of underground observation techniques to monitor the production and survival of lateral buds and roots, ideally coupled with some experimental root fragmentation. Such observations and manipulations should proceed with the approval of all the relevant conservation agencies. ACKNOWLEDGMENTS Thanks to the following groups and organisations for their assistance with the studies of Scottish S. romanzoffiana , including Scottish Natural Heritage, the Scottish Executive Rural Affairs Department, the Botanical Research Fund and the Professor Blodwen Lloyd Binns Bequest Fund of the Glasgow Natural History Society. Our grateful thanks to all the owners, farmers and crofters for permission to visit the sites on numerous occasions, and to the following for help in various ways: David Benham, Dr Neil Cowie, Alan Forrest, Dr Frank Horsman, David Lang, Roger Latour, Simon Leach, Dr Marilyn Light, Alex Lockton, Dr Deborah Long, Darach Lupton, Richard Manuel, Rae McKenzie, Rose Murphy, Wendy Nelson, Philip Oswald, Professor Clive Stace and Dr Chris Wilcock. The work was undertaken as part of a Scottish Natural Heritage research project with additional funding provided by the Botanical Research Fund and the Institute of British Geographers. REFERENCES Arft, A. M. & Ranker, T. A. (1998). Allopolyploid origin and population genetics of the rare orchid Spiranthes diluvialis. American Journal of Botany 85 (1): 1 10-122. Butcher, R. W. (1961). A new illustrated British flora. Leonard Hill, London. Catling, P. M. (1990). Biology of the North American representatives of the Subfamily Spiranthoideae, in SAWYERS, C. E. North American native terrestrial orchid propagation and production, pp. 46-67. Brandywine Conservancy, Chadds Ford, Pennsylvania, U.S.A. CHEFFINGS, C. (2004). New plant status lists for Great Britain. BSBI News 95: 36-43. CORRELL, D. S. (1950). Native orchids of North America. Chronica Botanica, Waltham Massachusetts, U.S.A. CURTIS, T. J. F. & Mcgough, H. N. (1988). The Irish Red Data Book. 1 Vascular Plants. The Stationery Office, Dublin. Ettlinger, D. M. Turner (1997). Notes on British and Irish Orchids. Privately published by the author, Dorking, Surrey. FORREST, A. (2001). Genetic structure within and among populations of Spiranthes romanzoffiana Cham, in Scotland. M.Sc. Thesis. Royal Botanic Garden Edinburgh and University of Edinburgh. Forrest, A. D., Hollingsworth, M. L., Hollingsworth, P. M, Sydes, C. & Bateman, R. M. (2004). Population genetic structure in European populations of Spiranthes romanzoffiana set in the context of other genetic studies on orchids. Heredity 92: 218-227. GODFERY, M. J. (1922). Spiranthes romanzoffiana. The Orchid Review 30: 261-264. GODFERY, M. J. (1924). Spiranthes romanzoffiana. The Orchid Review 32: 357-358. GODFERY, M. J. (1930) Spiranthes romanzoffiana. The Orchid Review 38: 291-295. GULLIVER, R. L. (2002). Dr Richard Gulliver action plan research into the ecology of Spiranthes romanzoffiana in Scotland. Unpublished report to Scottish Natural Heritage. Gulliver, R. & Gulliver, M. (2004). Looking for Irish Ladies on Scottish Islands. Aspects of the conservation biology of Irish Lady’s-tresses orchid Spiranthes romanzoffiana. The Orchid Review 112 (1256): 89-92, 106, 110, 118. Gulliver, R., Gulliver, M., Keirnen, M., & Sydes, C. (2004a; in press). Studies on the conservation biology of Irish lady’s-tresses orchid, Spiranthes romanzoffiana. 1. Population sizes, grazing, vegetation height and capsule status at reference sites. The Glasgow Naturalist 24 (2). Gulliver, R., Gulliver, M., Keirnen, M. & Sydes, C. (2004b; in press). Studies on the conservation biology of Irish lady’s-tresses orchid, Spiranthes romanzoffiana. 2. The establishment of 10 exclosures, dung counts and further studies on associated Juncus taxa (species and hybrid rushes). The Glasgow Naturalist 24 (2). Gulliver, R. L., Gulliver, M. & Sydes, C. (2003). The relationship of a Biodiversity Action Plan (BAP) orchid, Spiranthes romanzoffiana , to grazing in the West of Scotland. Aspects of Applied Biology 70: 143-150. CONSERVATION OF SPIRANTHES ROMANZOFF1ANA IN SCOTLAND 299 HENDERSON, S. A. (2001). The vegetation associated with Spiranthes romanzoffiana Cham. (Orchidaceae), Irish Lady’s-tresses, on the Isle of Coll, Inner Hebrides. Watsonia 23: 493-503. Heslop Harrison, J. (1953). The North American and Lusitanian elements in the flora of the British Isles, in LOUSLEY, J. The Changing Flora of the British Isles. T. B uncle and Co, Arbroath. HORSMAN, F. (1994). Spiranthes romanzoffiana Cham., in STEWART, A., PEARMAN, D. A. & PRESTON, C. D . eds. Scarce Plants in Britain, p. 398. Joint Nature Conservation Committee, Peterborough. HORSMAN, F. (1998). Spiranthes romanzoffiana and the Greenland White-Fronted Goose. BSBI News 79: 49- 50. HULTEN, E. (1937). Outline of the history of the arctic and boreal biota during the quaternary period. Stockholm. MOUSLEY, H. (1924a). Spiranthes romanzoffiana. The Orchid Review 32: 71-77. MOUSLEY, H. (1924b). Further notes on the underground development of Spiranthes romanzoffiana and Spiranthes cemua. The Orchid Review 32: 296-300. PRESTON, C. D., PEARMAN, D. A. & Dines, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. PERRING, F. H. & FARRELL L. (1983). British Red Data Books 1 Vascular Plants. The Royal Society for Nature Conservation, Nettleham, Lincoln. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. RASMUSSEN, H. N. (1995). Terrestrial Orchids from Seed to Mycotrophic Plant. Cambridge University Press, Cambridge. ROBARTS, J. (2000). A study of the orchid Spiranthes romanzoffiana on the island of Barra. Hebridean Naturalist 13: 30-35. ROLES, S. J. (1965). Flora of the British Isles - Illustrations. Vol. 4. Cambridge University Press, Cambridge. Ross-Craig, S. (1973). Drawings of British Plants, Vol. 8. G. Bell & Sons, London. U.K. BIODIVERSITY Action Group (1999). Tranche 2 Action Plans Volume 3. Plants and fungi. English Nature, Peterborough. WlLCOCK, C. (2002). Pollination failure in Scottish populations of Spiranthes romanzoffiana Cham., a rare European orchid. Unpubl. report to Scottish Natural Heritage. APPENDIX 1. DETAILS OF TWIN PLANTS IN 2000 & 2001 OVERALL COUNTS PRESENTED IN TABLE 1 Nature of twin followed by separation of stems (stem/bud in F/V pairs, bud/bud for V/V pairs) in mm COLONSAY 2000 Site KA: 02-04/09/00: a) 0; b) V/V, 23; c) V/V, F/V, F/V, F/V, V/V, 35, 48, 80, 58, 50 Site KB: 31/08-05/09/00: a) 0; b) V/V, 27; c) 0 Site KC: 31/08/00: a) 0; b) 0; c) 0 Site LF: 30/08-02/09/00: a) F/F, 10; b) V/V, 8; c) 0 COLL 2000 Site CA Sample 1)15 4 m2 Quadrats, 14/08/2000, (EG) a) F/F, 17; b) fifteen V/V, 7 F/V mean separation 10-0; c) V/V, V/F*, 89, 57 Site CA Sample 2) Sample of 59 (47F & 12V) 21-22/8/2000 a) 0; b) ten F/V mean separation 13-2; c) 3 F/V, 45, 52, 70 COLONSAY 2001 Site KA: 30/08/-06/09/01: a) F/F, 10; b) V/V, 20; c) V/V, F/F, 35, 48 Site KB: 28/08/01: a) 0; b) V/V, F/V, 10, 5; c) F/V, 65 Site KC: 27/08/01: a) F/F, 10; b) 0; c) 0 Site LF: 04/09/01: a) F/F, 15; b) V/V, 8; c) 0 * V/F to retain sequence of plant code numbers 301 The turlough form of Ranunculus repens L. (Creeping Buttercup) S. WALDREN Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6, Ireland* and D. LYNN & S. MURPHY Depart?nent of Environment and Local Government, National Parks & Wildlife Service, 7 Ely Place, Dublin 2, Ireland ABSTRACT Turloughs are temporary water bodies in areas of karst limestone. They are almost entirely restricted to Ireland, and though the flooding regime is largely seasonal, flooding can occur at any time of the year following heavy rainfall. This unique hydrological regime has resulted in an impoverished but highly interesting flora, including a morphologically distinct form of Ranunculus repens with highly dissected leaves. This genetically determined leaf morphology is accompanied by other physiological traits, including different patterns of seasonal carbohydrate accumulation, greater underwater photosynthesis and respiration, and altered patterns of stomatal distribution, as compared with non-turlough populations. In addition, while non-turlough populations show petiole elongation typical of the ‘depth-accommodation’ response when submerged, no such petiole elongation occurs in turlough populations, which survive several metres of submergence annually as fully green, metabolically active plants. AFLP analysis has revealed that turlough populations of R. repens may show closer genetic affinities to each other than to non-turlough populations, suggesting a common evolutionary ancestry. Our detailed studies of selected turlough basins have shown that despite the capacity of R. repens for profuse clonal growth a relatively large number of distinct genotypes occur in these habitats. We suggest that this genetic diversity is a result of microhabitat heterogeneity, in terms of flood depth and duration, with different genotypes occupying different micro-environments within the turlough basin. It therefore seems probable that strong selective pressures have maintained various morphological and physiological traits in turlough populations of Ranunculus repens to enable its survival in such a challenging environment. Keywords: Ireland, morphology, physiological adaptation to submergence, population biology. AFLP. AN INTRODUCTION TO TURLOUGHS - ENVIRONMENT AND ECOLOGY Turloughs are temporary water bodies occurring over Carboniferous limestone in Ireland. The word ‘turlough’ comes from the Irish ‘Tuar Loch\ meaning dry lake (Praeger 1932), which accurately summarises these features: depressions in the landscape which might appear as a lake in winter, yet often with no sign of standing water during summer. Turloughs have also been referred to as ‘vanishing fields’ (MacGowran 1985), again alluding to the temporary nature of the water body. In general, turloughs occupy depressions in karst limestone, often where there is only a thin drift layer. Though some may have inflow streams, outflow from turloughs is entirely subterranean via swallow holes or estevelles (Coxon 1996). This direct link with subterranean aquifers means that water levels in turloughs rise during periods of high water table or high charging, and subside with the water table. In practice, this means that turloughs are generally flooded during the winter *e-mail: swaldren@tcd.ie 302 ATLANTIC ARC (Plate 49a), but dry during summer (Plate 49b), though the direct connection with ground water means that changes can occur rapidly and, given Ireland’s relatively high rainfall, temporary flooding may occur at any time of the year. In 2002 water levels in turloughs that we have been studying remained consistently high until July, whereas in 2003 water levels had receded to summer levels as early as April, only to rise again during the very wet late spring and early summer. It is clear that plants and animals occurring in turloughs must be able to cope with fluctuating, and to some degree unpredictable changes in water level. Many of the plant species occurring in turlough basins survive the winter period under several metres of water, and then face unpredictable flooding during spring, summer and autumn. The depth, duration and timing of flooding are therefore major determinants of vegetation composition and species distribution, though different soil types - peats, marl, gleys and drift - also play some role. Typically there is a pronounced zonation of vegetation with depth and duration of submergence, with shrubby communities occurring in the upper, rarely flooded zones, grading through herb and sedge-rich communities to marsh and fully aquatic vegetation in the permanent pools which may occupy the lowest part of the basin. Turlough vegetation communities have been described by Praeger (1932), Ivimey-Cook and Proctor (1966), Louman (1984), MacGowran (1985), Goodwillie (1992, 2003) and Lynn & Waldren (2001a, 2003c). Turloughs are often managed for summer grazing, mainly by cattle or sheep, and grazing intensity strongly affects sward development. One of the most intriguing aspects of turlough vegetation is the occurrence of species that would not normally be considered wetland plants (e.g. Lotus corniculatus ) and many wetland species that would not normally be subjected to complete submergence for long periods (e.g. Ranunculus repens ). Turloughs are largely restricted to the Carboniferous limestone of Ireland, particularly in the mid-west through Co. Clare, Co. Galway, Co. Roscommon and Co. Mayo. Outside Ireland, a turlough exists in south Wales at Pant Y Lynn. The Breckland meres are also temporary water bodies lacking an overground outflow, but as they overlie chalk their flooding regime is very different to that of turloughs. The ‘poljes’ of Yugoslavia are also glacial hollows on Carboniferous limestone; however the presence of thick sediments on the floors of these hollows again leads to a very different hydrology to that occurring in Irish turloughs. As a result of their restricted distribution and unique geomorphology, and also partly because of the unique assemblages of plant and animal communities they contain, turloughs have been designated as a Priority Habitat in the 1992 E.U. Habitats Directive. Accordingly, many turloughs have been designated as candidate Special Areas of Conservation by the Irish Government, and many others nominated as National Heritage Areas. A few, including part of the Coole/Garryland complex in Co. Galway, are National Nature Reserves, biogenetic reserves and designated Ramsar sites. THE TURLOUGH FORM OF RANUNCULUS REPENS - MORPHOLOGY A wide range of mechanisms to cope with survival of submergence occurs in turlough plants, including avoidance through the annual habit (e.g. Persicaria mitis ) or dormancy (e.g. Potentilla anserina ), development of an aquatic habit (e.g. Galium palustre), depth accommodation by petiole elongation (e.g. Rumex crispus ) and submergence tolerance (e.g. Ranunculus repens). Ranunculus repens occurs widely in turlough basins and in surrounding pasture land, however the plants growing in the zones of turloughs prone to flooding have characteristically highly dissected, more or less glabrous leaves, very different from the leaves of typical R. repens growing in pasture (Plate 48; Lynn & Waldren 2001a, 2003a). The turlough plants are also typically smaller and, when growing in short swards, the leaves are usually held close to the substrate surface. We calculated the ratio of leaf perimeter to the square root of leaf area as a measure of dissection (see Kincaid & Schneider 1983; Lynn & Waldren 2001a) in seed-derived plants raised as families from known wild maternal parents, and demonstrated that turlough populations have a significantly higher index of dissection (Fig. 1). The narrow-sense heritability estimate for variation in these families was 72% (Lynn & Waldren 2001a), indicating a high degree of additive genetic variation in leaf dissection transmitted from parent to offspring. TURLOUGH FORM OF RANUNCULUS REPENS 303 Area (cm) Figure 1 . Plot of square root of leaf area versus leaf perimeter for turlough (open symbols) and non-turlough (filled symbols) populations of Ranunculus repens. The area (a) and perimeter (p) of each leaf was measured from digitally scanned tracings of leaf outlines. The linear regressions equations for the two habitat types are turlough: p = 22-6a-23-3 (P<0-001); non-turlough: p = 13-7a-16-3 (P<0-001). The slopes for the two habitats are significantly different (ANOVA, P<0-001). See Appendix 1 for site locations. In turloughs, R. repens remains green during submergence and usually shows a small amount of leaf development. In the more oligotrophic turloughs we have studied during submergence using scuba apparatus, R. repens stands out as fresh green growth among a matrix of greyish sedges that show very little growth during submergence. Ranunculus flammula, also a common component of turlough communities, also remains winter-green, but develops long, more or less linear submerged leaves, with long petioles and small lamina. Where the depth of submergence is relatively shallow, these long leaves of R. flammula reach the surface and re-establish contact between the plant and the atmosphere. This so-called depth accommodation response (Ridge 1987), or ‘snorkeling’, greatly facilitates gaseous exchange, and hence metabolic activity. Many other wetland plants show this response to submergence, thought to be mediated by the plant growth regulator ethylene (Ridge 1987; van der Smam et al. 1993). The turlough form of R. repens shows no such depth accommodation response. Submerged plants in the field retain their small rosettes with short petioles, although the leaves are generally held more erect when submerged, possibly a buoyant response but perhaps also due to ethylene- mediated flexure of the petiole base (Voeseneck et al. 1997). Exactly the same response is seen in pot-grown plants in controlled environments and submerged in aquaria. However, when the typical pasture or ruderal form of R. repens is subjected to experimental submergence, there is a considerable extension of petioles and stimulation of new leaf growth (Lynn & Waldren 2003b). Provided that these leaves can reach the surface and regain gaseous exchange with the atmosphere, photosynthesis and metabolism can occur and the plant may well survive. However, we have dived over some of our experimental plots and recorded 6 m of water over R. repens in turloughs, and 304 ATLANTIC ARC suspect that in some narrow basins, flood water may well regularly reach a depth of 10 m. It is highly unlikely that the depth accommodation response of the pasture form of R. repens could facilitate survival in even 1 m of flood water; in addition, the increased growth must be highly demanding on stored reserves of carbohydrate, given that photosynthesis will be reduced under submerged conditions (Lynn & Waldren 2002). Thus, by not displaying a depth accommodation response, turlough populations of R. repens avoid futile attempts to reach the surface in deep water and the consequent demands on stored nutrients that such a response would entail. PHYSIOLOGICAL ADAPTATIONS TO SUBMERGENCE IN RANUNCULUS REPENS The growth of petioles, leaves or other plant parts during submergence is dependent on a supply of stored carbohydrates. As photosynthesis is significantly reduced under submerged conditions in both turlough and non-turlough forms of R. repens (Lynn & Waldren 2002), it is unlikely that de novo synthesis of carbohydrates could match demands placed by depth accommodation. Stored carbohydrate will be required to rapidly re-establish growth once flood waters recede, and the turlough form of R. repens stores large amounts of starch in its roots. Monitoring of carbohydrate levels in roots of turlough and non-turlough populations throughout the year showed a large peak of starch in spring, shortly before the usual period of emergence and spring growth (Lynn & Waldren 2003b); these were far higher concentrations than recorded in the non-turlough population studied. In the laboratory we were able to demonstrate higher levels of both photosynthesis and respiration in submerged leaves of turlough plants compared to non-turlough plants (Lynn & Waldren 2002), suggesting the turlough form may be able to acquire additional photosynthate during spring submergence; as very few additional leaves are produced while submerged, presumably much of this photosynthate can be directed to storage products. We also found much higher levels of ethanol-soluble carbohydrate (including sucrose etc.) and ethanol- insoluble carbohydrate (probably fructans - absent from non-turlough populations) in turlough plants, again suggesting a very different pattern of carbohydrate metabolism in turlough and non- turlough R. repens. Both turlough and non-turlough populations of R. repens readily develop aerenchyma, gas spaces or lacunae in cortical cells of the root which enable gaseous exchange between leaves and roots buried in waterlogged substrate (Lynn & Waldren 2003b). However, the pattern of stomata, the pores in leaf surfaces which regulate gas exchange, varied considerably between populations. Non-turlough plants had a much higher stomatal index on the lower surface compared to the upper leaf surface, while this difference between leaf surfaces was much less apparent in the turlough plants. Coupled with the highly dissected and glabrous leaves, this may be responsible for rates of aerial photosynthesis being almost two times higher in turlough plants than in non-turlough plants. Turlough plants also had a slight but consistently higher rate of respiration (Lynn & Waldren 2002). The dissected leaf form may also facilitate gaseous exchange in submerged conditions, particularly by decreasing the boundary layer of relatively static water (or air) around a leaf. The net result of these physiological and morphological characteristics is that the turlough form is much better able to survive submergence than the non-turlough forms we have studied. When grown experimentally, drained or flooded (not submerged) treatments had little effect on either ruderal or turlough populations, whereas submergence greatly reduced the biomass of the ruderal population (Fig. 2a). Submergence also greatly increased the amount of dead material on each plant in the ruderal population (Fig. 2b), and when further experiments were conducted in artificially hardened water, as might be experienced in a typical limestone turlough, the onset of senescence and death was even more rapid in the ruderal plants (Lynn & Waldren 2003b). POPULATION BIOLOGY AND SYSTEMATICS It is clear that the large morphological and physiological differences between turlough and non- turlough populations of R. repens are a result of the severe selection pressures imposed by life in an amphibious environment. Despite the great variability of clones of R. repens in Ireland (Lynn & Waldren 2001b), the dissected leaf form consistently occurs in turloughs. Has this form repeatedly TURLOUGH FORM OF RANUNCULUS REPENS 305 8 i Drained Flooded Submerged T reatment □ Ruderal □ Turlough "O CO CD -o c o tr o Q. O ol 0.5 i 0.4 - 0.3 - 0.2 - 0.1 - 0 Drained Flooded Submerged Treatment (B) FIGURE 2. Mean total dry weight (± SE) of living plant tissues (a) and proportion of dead material (b) from a ruderal (River Dodder) and turlough (Lough Gealain) population of Ranunculus repens grown experimentally in drained, flooded (but not submerged) or fully submerged conditions for three weeks (n = 4). The interactions between population and flooding treatment are significant in both cases (ANOVA- total dry wt: P1 mm wide), with an acute apex and are entirely - or almost entirely - smooth (Fig. 2). The dense, hyaline papillae in ‘ repens' are very distinctive but also microscopic (needing at least x 20 magnification), and so easily overlooked (Fig. 3). Any fruits need to be fully ripe and dried for proper determination. (Note that the fruits of ‘typical’ Baldellia can also have occasional small protrusions or papillae, especially along the dorsal surfaces, but these are quite unlike the dense covering in ‘repens\ ) FIGURE 2. Baldellia fruits: ‘typical’ Baldellia , from Ceredigion (above) and stoloniferous Baldellia, from Pembrokeshire (below). BALDELLIA RAN UN C ULOIDES SUBSP. REPENS IN WALES 317 Figure 3. Fruit of stoloniferous Baldellia from Pembrokeshire, showing hyaline papillae. TABLE 4. FURTHER INFRASPECIFIC CHARACTERS IN BALDELLIA RANUNCULOIDES (SUPPLEMENTARY TO THE CHARACTERS LISTED IN TABLE 1, FROM KERN & REICHGELT (1950)) Subsp. rcinunculoides Subsp. repens Achenes Inflorescence Habitat* Flowering period Population structure Deeply keeled (>1 mm wide) and smooth, with an acute beak and without numerous hyaline papillae when ripe (x 20 magnification). upright umbel or tier of 2(— 3?) whorls without leaves, growing from c. April-August. temporary gaps within calcareous or mildly brackish pools, dune slacks, ditches and mesotrophic water-bodies. June to July (-August) scattered, detached individuals. Narrow (<09 mm wide) with a persistent hooked beak and numerous hyaline papillae when ripe (x 20 magnification). decumbent, leafy, indeterminate shoot, usually rooting at nodes and growing throughout the year shorelines and long-standing gaps in weakly acidic, Littorelletean heathland pools and oligotrophic lakes. June to October (-November) locally-abundant, networks of vegetatively spreading clones * There is some intersection between these two habitats, e.g. in coastal heathland pools or disturbed peaty habitats overlying calcareous clays. The difficulty with this character, however, is that fruits are not always present in creeping Baldellia : the Anglesey plants, for example, seem to be entirely self-incompatible (both in the field and in cultivation), whereas the plants from Pembrokeshire populations are partially self-fertile, at least towards the end of the year. However, the absence of fruits - or at least the presence of aborted fruits on the receptacle - is practically diagnostic because ‘typical' upright Baldellia is so highly self-fertile. 318 ATLANTIC ARC STOLON/INFLORESCENCE CHARACTERS A more constant feature in the two variants of Baldellia is their stem character - and, indeed, they are both most readily described in these terms. The inflorescence in ‘creeping’ plants is a decumbent, leafy and largely indeterminate shoot, whereas in the ‘upright’ variant it is, essentially, an erect umbel or tier of 2 (-3?) whorls without leaves. Flowers in the stoloniferous variant are also usually solitary, mixed with vegetative buds or (more rarely) held in short, ascending racemes. Very often (e.g. in submerged plants) the stoloniferous inflorescence does not actually have any flowers and grows as an entirely vegetative axis. This is a very distinctive structure, termed a ‘pseudostolon’ (Charlton 1971), where vegetative buds arise in the axes of bracts or bracteoles; and it is a structure found elsewhere within the Alismataceae (cf. Luroninm). Very occasionally, ‘typical’ Baldellia can develop secondary vegetative buds in the inflorescence, if they fall over (Kern & Reichgelt 1950; Triest & Vuille 1991) but these can be seen in the field as an exception, unlike the network of leafy, creeping stems in ‘ repens'. Highly impoverished rosettes of typically upright plants can also arise with only solitary flowers (‘forma pumilis Gluck’), bearing a resemblance to inflorescence nodes in the creeping plant but, again, these are isolated plants, not aggregated or connected by horizontal stems. Very notably, the flowering stems in ‘typical’ Baldellia seem to always die down after fruiting whilst in stoloniferous plants - especially in the Anglesey population - they continue growth, even during the winter. In nature, these two developmental attributes could each have an adaptive significance: the more short-lived and fruiting stem type being suited to temporarily open, productive habitats, and ramifying vegetative plants occupying a perennial niche in low-nutrient waters. These resemble the competitive and stress-tolerant strategies, respectively, of Grime et al. (1988). Upright Baldellia is also usually found as scattered individuals, most visible in summer, but the creeping form is very often locally common and evident through the year - if only as submerged vegetative rosettes. These ecological and phenological differences add to the existing list of characters and help to identify populations in the field, even when the two types are growing in close proximity to one another. DISCUSSION Baldellia ranunculoides (. sensu lato ) is a highly variable taxon with considerable morphological plasticity. Much of this variation seems to have very little taxonomic significance and British and Irish botanists have, understandably, tended to see creeping or stoloniferous plants simply as a habitat-form or ecotype of the ‘typical’, upright Baldellia. However, there is good evidence from cultivation experiments, and from the clear correlation of a number of morphological characters, that the two variants are genetically distinct and consistent with Baldellia ranunculoides subsp. repens and B. ranunculoides subsp. ranunculoides (perhaps even B. repens and B. ranunculoides ) described in the mainland European literature. Recent work, summarised above, has highlighted several characters, in addition to those listed in Table 1, that appear to be useful in distinguishing these two taxa (Table 4). FURTHER WORK The rediscovery - and re-description - of Creeping Water-plantain in Wales inevitably raises a number of questions. There are, reportedly, similar populations in Jersey and in S. Kerry (J. Banks & E.C. Mhic Daeid pers. comm.), but also some instances of confusion with Luroninm natans. There is a need for more detailed surveys and for a review of herbarium and literature sources to help define the distribution, trends and current conservation status in this taxon. At the same time, a wider review of Baldellia in the British Isles should aim to characterise the range of variation and the extent of intermediates. It is unclear, for instance, how far the differences between Anglesey and Pembroke populations (e.g. average stem width and degrees of self-incompatibility) amount to transitional forms and whether the distinctive fruit and stem characters always occur together. Vuille (1988) notes the existence of intermediates in Italy and perhaps Belgium and there is a need for reciprocal crossings between ‘typical’ and creeping variants and, also, a re-examination of the more microscopic attributes (e.g. pollen characteristics) in relation to large morphological differences. BALDELLIA RANUNCULOIDES SUBSP. REPENS IN WALES 319 All of these questions bear upon other taxonomic uncertainties. The question of repens/ ranunculoides as species or subspecies, for instance, relies on an understanding of intermediates and genetic barriers. Molecular genetic data may help resolve another problem, too: whether Baldellia is in fact congenerous with Echinodorus. Whatever the details of its taxonomy, Creeping Water-plantain - Dyfr lyriad ymlusgawl - has been largely overlooked in Britain and Ireland for almost 200 years. The reasons for this go back to the earliest descriptions and interpretation of some very variable material. Smith’s English Botany is still a remarkably good account of the British and Irish flora but, in this case perhaps, Davies’ Welsh Botanology is the better guide. ACKNOWLEDGMENTS Particular thanks to Arthur Chater for valuable comments on an early draft of this paper, and to Tim Rich for genuine inspiration in the field. Anna Williams translated Lindblad and Stahl (1990) and Alan Hale took the photographs of Baldellia fruits. REFERENCES CAVANILLES, A. I. (1791). leones et descriptions plantarum quae aut sponte in Hispania crescunt aut in hortis hospitanhir. Type Regia, Madrid. Charlton W. A. (1973). Studies in Alismataceae II. Inflorescences of Alismataceae. Canadian Journal of Botany 51: 775-789. Dandy, J. E. (1980). Baldellia , in Tutin T. G. et al. eds. Flora Europaea Vol. 5, p. 2. Cambridge University Press, Cambridge. Davies, H. (1813). Welsh Botanology. W. Marchant, London. Day, P. (1983). Rare Plant Sur\>ey of Anglesey. Unpublished report. Nature Conservancy Council, Bangor. GLUCK, H. (1905). Biologische und morphologische Untersuchen iiber Wasser- und S u m pfg e wach s e . Gustav Fischer, Jena. GRIME J. P., HODGSON, J. & HUNT R. (1988). Comparative Plant Ecology: a functional approach to common British plants. Unwin Hyman, London. HORWOOD, A. R. & NOEL, C. W. F. (1933). The Flora of Leicestershire and Rutland. Oxford University Press, Oxford. JONES, R. A. & Rich, T. C. G. (1998). Luronium natans! Baldellia ranunculoides! Alisma, in RICH, T. C. G. & Jermy, A. C. Plant Crib 1998, pp. 314-317. B.S.B.I., London. Kern, J. & Reichgelt, Th. (1950). Over enige kritische planten van onze flora. Nederlandsch Knudkundig Archief LVII: 244-261. LAWALREE, A. (1959). Baldellia repens (Lamk.) van Ooststroom en Belgique. Bulletin du Jardin Botanique de I’Etat, Bruxelles 29 (1): 7-14. LINDBLAD, R. & Stahl, B. (1990). Revsvalting, Baldellia repens, i Norden. Svensk Botanisk Tidskrift 84: 253-258. PRESTON, C. D. & Croft, J. M. (1997). Aquatic Plants in Britain and Ireland. Harley Books, Colchester. PRESTON, C. D., Pearman, D. A. & Dines, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. PRIMAVESI, A. L. & Evans, P. A. (1988). Flora of Leicestershire. Leicestershire Museums, Leicester. Punt, W. & Reumer, J. W. (1981). The northwest European pollen flora 22. Alismataceae. Review of Paleobotany and Palynology 33: 27-44. Roelofs, J. G. M. (1983). Impact of acidification and eutrophication on macrophyte communities in soft waters in the Netherlands. 1. Field observations. Aquatic Botany 17: 139-155. SCULLY, R. W. (1916). Flora of County Kerry. Hodges, Figgis & Co., Dublin. SELL, P. D. & Murrell, G. (1996). Flora of Great Britain and Ireland. Volume 5. Cambridge University Press, Cambridge. Smith, J. E. (1828). English Botany. Longman, Rees, Orme Brown & Green, London. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. Triest, L. & VuiLLE, F-L. (1991). Isozyme variation in several seed collections and hybrids of Baldellia (Alismataceae), in Triest, L., ed. Isozymes in water plants. Opera botanica Belgica 4: 37-48. VuiLLE, F-L. (1988). The reproductive biology of the genus Baldellia (Alismataceae) Plant Systematics and Evolution 159: 173-183. WALTERS, M. & Gilmour, J. (1955). Wild Flowers. Collins, London White, J. W. (1912). The Bristol Flora. John Wright & Sons. Bristol. 321 Isoetes histrix L. (Land Quillwort) on the Lizard peninsula R. J. MURPHY Shangri-la, Reskadinnick, Camborne, Cornwall TR14 OBH, England and C. N. PAGE Gillywood Cottage, Trebost Lane, Stithians, Truro, Cornwall TR3 7DW, England ABSTRACT This note summarises what is known about the distribution and ecology of Isoetes histrix L. (Land Quillwort) at the northern extremity of its world range on the Lizard peninsula, W. Cornwall. Keywords: Land Quillwort, ecology, conservation management, ephemeral pools, erosion pans. DISTRIBUTION AND STATUS Isoetes histrix L. (Isoetaceae) was first found in Britain in 1919, when a plant from the Caerthillian Slopes on the Lizard peninsula, W. Cornwall (v.c. 1), was accidentally uprooted by F. Robinson. However, its presence here was not confirmed until 1937 when it was rediscovered by Dr. R. Melville of Kew. I. histrix is a plant of the Mediterranean and Atlantic coasts of Europe, and not surprisingly assigned to the Mediterranean-Atlantic element of the British flora by Preston & Hill (1997). In mainland Britain it is found only on The Lizard, where it is largely confined to outcrops of serpentine on the western side of the peninsula (Fig. 1). It also occurs in the Channel Islands. I. histrix is well adapted to the weather conditions of this far south-westerly comer of Britain, making its active growth during the mild, wet winters and passing the hot, dry summers as a resting, underground corm. A Red Data Book species (Byfield & Pearman 1999), it has continued to attract much interest following the work of John Hopkins in 1979 and that during the University of Bristol Lizard Project from 1982 to 1984. HABITATS AND ASSOCIATED SPECIES As it is a poor competitor it is well adapted to growing on the shallowest, skeletal soils that tend to support very little in the way of vegetation. If these soils are also moist (or flooded) in winter and subject to severe drought in summer, then they are even more suitable. There are three habitats on the Lizard peninsula that meet the needs of Isoetes histrix : • Footpaths, especially those on or near the coast • Erosion pans, especially those with fragments of exposed serpentine • Shallow soils around serpentine outcrops on slopes with a southerly aspect - these often support the highest populations (Frost et al. 1982) Recently it has been realised that this species can also survive in short herb-rich turf on shallow soils, as long as overall vegetation cover does not exceed 80% (Byfield & Pearman 1999). Though Isoetes histrix cannot tolerate excessive competition there is still quite a list of associated species and some of these can be confused with it. Associated species include several of the Lizard specialities including Herniaria ciliolata, Juncus capitatus, Trifolium bocconei and T. strictum. Amongst the associates that can look remarkably similar to Isoetes are Allium 322 ATLANTIC ARC 6 7 8 FIGURE 1. The distribution of Isoetes histrix on the Lizard peninsula, W. Cornwall (1-km square records) O Pre 1987, • 1987. FIGURE 2. Isoetes histrix - 20 March 2002, near Holestrow, Lizard (M. J. Stribley). ISOETES HISTRIX ON THE LIZARD 323 schoenoprasum, Scilla autumnalis and S. vema; all these have linear, rather curved leaves like Isoetes, but the ‘cart-wheel’ appearance of the latter (Fig. 2), with its usually narrower leaves, often less than 1 mm in width, and with a double band of stomata on their upper surface, can be very distinctive. (The stomata on the upper leaves of S. autumnalis are more scattered.) Population numbers can change markedly from year to year according to weather conditions and the availability of suitable habitat patches. Where tourist pressures are not excessive they can help in maintaining the open conditions this species requires. Populations on coastal paths, for example, depend for their survival on the passage of human feet. In addition, the introduction of grazing by Shetland ponies and Dexter and Highland Cattle (‘hairy lawn-mowers’!) is going a long way towards keeping the cliff-top grasslands and rocky slopes in the right condition for I. histrix, along with so many of the Lizard’s other rare species. One further note: as can be seen in Fig. 1, there are few inland sites on The Lizard for this species, and most colonies occur on exposed south- or west-facing slopes on the western side of the peninsula. In March 2002, a botanical recording group met on The Lizard on a day when strong winds were blowing spume and sea-spray a good distance inland, and it was almost impossible to get across the inlet at Caerthillian Cove. It was a marked reminder of its coastal habitats in Guernsey, where Isoetes grows “...close enough to the sea to get wind-blown spray” (Ryan 1990). REFERENCES Byfield, A. J. & Pearman, D. A. (1999). Isoetes histrix Bory, in WlGGINTON, M. J. ed. British Red Data Books. 1. Vascular Plants. 3rd. ed., p. 200. Joint Nature Conservation Committee, Peterborough. Frost, L. C., Hughes, M. G. B., Nichols, C. & Lawman, J. M. (1982). A total population estimate of the land quillwort (Isoetes histrix) at The Lizard District and recommendations for its conservation. Bristol: University of Bristol. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. Ryan, P. (1990). The land quillwort in the Channel Islands. Pteridologist 2: 28-30. 325 Pilularia globulifera L. (Pillwort) in Cornwall I. J. BENNALLICK Lower Polmorla, St.Wenn, Bodmin, Cornwall, PL30 5 PE, England* ABSTRACT This note summarises historic and recent records of Pilularia globulifera L. (Pillwort) in Cornwall. In Britain this is a declining species. It has certainly been lost from a number of its outlying sites in Cornwall, but i is still extant at many localities within its stronghold areas on the Lizard peninsula and the mid-Cornwall moors. A greater appreciation of its habitat requirements, and an improved ability amongst botanists to recognise it in the field, have led to several new populations being discovered and old sites - thought to have been lost - being found again in recent years. Keywords : conservation, distribution, ephemeral w'ater bodies, lakes, ponds. Lizard peninsula, mid-Cornwall moors. INTRODUCTION In Britain, Pilularia globulifera L. (Marsileaceae) is Nationally Scarce (Jenny 1994), having been recorded in 98 10-km squares since 1987 (Preston et al. 2002), while in Ireland it is a Red Data Book species (Curtis & McGough 1988), with records from just seven 10-km squares since 1987 (Preston et al. 2002). It is a Priority Species in the U.K. Biodiversity Action Plan (U.K. Biodiversity Group 1998). In Cornwall, as in many other parts of Britain and Ireland, it appears to be a declining species: there are records of it from a total of 14 Cornish 10-km squares regardless of date, but from only nine (64%) of these squares between 1987 and 1999 (Preston et al. 2002). This decline is further highlighted by examining the records at tetrad (2-km square) scale: Pilularia has been recorded from a total of 34 tetrads in Cornwall, but since 1980 from only 17 (50%) of these (French et al. 1999). Many of these losses have been of relatively isolated populations and, although some losses may have occurred within its core areas (especially in W. Cornwall), large extant populations can still be found in its main strongholds on the Lizard peninsula (SW61, SW71 and SW72) and the mid-Comw'all moors area north of St Austell (SW95. SW96, SX05 and SX06). RECENT SURVEYS In recent years field botanists in Cornwall have been actively searching for Pilularia , and in many cases repeated checking of old sites within the core areas on The Lizard and mid-Cornwall moors has led to its rediscovery in places where it was previously considered to have been lost. At several of these localities we have noted that water levels fluctuate markedly from season to season, and the species not only occurs in shallow water at the edge of lakes or pools, but also in ephemeral water bodies (e.g. summer-dry ponds and puddles on rutted track-ways). At many sites we suspect that intermittent, low-level disturbance, for example by livestock and farm vehicles, is important in helping to maintain the open conditions that Pilularia seems to require. The mid-Cornwall moors have a history of tin-streaming, and Pilularia can be found in the mosaic of small pools and hollows that have been created by this form of mineral extraction. Breney Common (SX06) and Goss Moor (SW95 & SW96) are good examples of w7here this has happened, and Pilularia occurs at both these sites in scattered and sometimes large populations. Interestingly, more recent disturbance can also lead to colonisation. For example, Pilularia has *e-mail: ian@bennallick.fsnet.co.uk 326 ATLANTIC ARC been found colonising soft mud in new drainage culverts constructed to drain the A30 trunk road that traverses Goss Moor. The species also occurs in the disturbed China-clay extraction areas, as at Treskilling Pit (SX05) where a huge population occurs around the edge of the disused flooded China-clay pit. The other core area for Pilularia in Cornwall is the Lizard peninsula. This area has a fascinating history of land-use and this - together with the unusual geology and the oceanic climate - has led to the development of internationally important wetland and heathland habitats. Within these habitats, many ephemeral puddles or pools (and some permanently wet pools) are found, often closely associated with the many track-ways that criss-cross the peninsula - some of which are traditional routes that have been in use for hundreds of years. Pilularia is found on The Lizard in all these (usually relatively open) habitats, with important populations at several sites, including Hayle Kimbro Pool (SW61), Croft Pascoe Pool (SW71) and Traboe Cross (SW72) - at the last of these it was rediscovered after an absence of 15 years (French et al. 1999). To untrained eyes Pilularia can be infuriatingly difficult to find amongst other aquatic vegetation, especially when growing with Eleogiton fluitans or the elongated aquatic form of Juncus bulbosus (var. fluitans). Difficulties of identification can mean that in some well-vegetated sites it may be still present but - despite searches - undetected. For example, in 1998 the species was refound on a site at Trelow Downs (SW96) following a gap of 20 years, after cattle had helped to open up the vegetation and enrichment from their dung had produced some very lush (and easily spotted) Pilularia\ It is hoped that with better understanding of this plant’s habitat requirements and continued training of botanists to be able to identify it in the field, future recording in Cornwall may result in further populations being discovered (and old ones rediscovered) not only within its strongholds on The Lizard and mid-Cornwall moors, but also in outlying areas where it may not have been seen for several decades (see Plate 50). REFERENCES CURTIS, T. G. F. & MCGOUGH, H. N. (1988) The Irish Red Data Book. Vol. 1. Vascular Plants. Dublin: Stationery Office. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. JERMY, A. C. (1994). Pilularia globulifera L., in STEWART, A., PEARMAN, D. A. & PRESTON, C. D. eds. Scarce Plants in Britain , pp. 311-312. Joint Nature Conservation Committee, Peterborough. PRESTON, C. D., Pearman, D. A. & Denes, T. D. eds. (2002). New Atlas of the British and Irish Flora. Oxford University Press, Oxford. U.K. BIODIVERSITY Group (1998). Tranche 2 Action plans. Volume 1. Vertebrates and vascular plants. English Nature, Peterborough. 327 Ranunculus ophioglossifolius Vill. (Adder’s-tongue Spearwort) in the Nord/Pas-de-Calais region of northern France C. BLONDEL Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul, Hameau de Haendries, 59270 Bailleul, France* ABSTRACT This note summarises recent work on the distribution, habitat preferences and current status of Ranunculus ophioglossifolius Vill. (Adder’s-tongue Spearwort) in the Nord/Pas-de-Calais region. The species is restricted to just three sites (all coastal), and it is now the subject of a detailed conservation plan. The largest extant populations lie in the Low Slack valley, near Boulogne-sur-Mer. Keywords: population counts, conservation plan, agri-environment schemes. DISTRIBUTION AND HABITATS Ranunculus ophioglossifolius Vill. (Ranunculaceae) has a predominantly Mediterraneo-Atlantic distribution in Europe (Hulten & Fries 1986), being widespread in southern Europe from Spain to Greece and the Crimea, but restricted to just a few, highly disjunct (outlying) stations in the north of its range (northern France, southern England and Sweden (Gotland)). There are three localities in the Nord/Pas-de-Calais region of northern France (Blondel et al. 2001), at Tardinghen Marsh, Low Slack valley and Visemarest meadows; unlike those in Britain, all three sites are on or near the coast (Fig. 1). The vegetation in which the species occurs in our area may be described as ‘subhalophilous’, and is characterised by such species as Eleocharis palustris and Oenanthe fistulosa (the Eleocharo palustris-Oenanthetum fistulosae de Foucault 1984). Figure 1 . The three populations of Ranunculus ophioglossifolius in Nord/Pas-de-Calais *e-mail : crp.cbnbl@wanadoo.fr 328 ATLANTIC ARC THE LOW SLACK VALLEY POPULATION The Low Slack valley is a large area of alluvial floodplain situated near Boulogne-sur-Mer. It is an exceptional natural area in Nord/Pas-de-Calais, supporting a remarkable diversity of fauna, flora and habitats. There are extensive areas of hay-meadows, with hunting ponds here and there. The populations of Ranunculus ophioglossifolius at this site are the largest in northern France: about 20 stations/discrete sub-populations have so far been registered, containing between them more than 10,000 plants. The species appears to occur within a range of seasonally flooded vegetation types in this area (Fig. 2). Securing the conservation of this exceptional area remains a high priority: many patches of land are (or were, until recently) being managed in traditional fashion as hay-meadows, and the grassland they support is very largely agriculturally unimproved. Substantial areas of grassland have been successfully protected through a local agri-environmental campaign led by Parc Nature l Regional des Caps et Marais d’Opale. It is important, however, that these management practices be continued in the future; it is feared that the withdrawal of active measures to protect the environment of the valley might constitute a serious threat to its biodiversity - and would soon lead to the demise of several key populations of R. ophioglossifolius. REFERENCES Blondel, C., Valentin, B., Destine, B., Hendoux, F. et al. (2001). Plan de Conservation de la Renoncule a feuilles d’ophioglosse (Ranunculus ophioglossifolius Vill.) pour la region Nord/Pas-de-Calais. Programme Interreg II “Biodiversite - Transmanche”. Restauration, Protection et Gestion conservatoire d’especes menacees dans la region Transmanche. Centre Regional de Phytosociologie/Conservatoire Botanique National de Bailleul, pour PUnion Europeenne, le Conseil Regional Nord/Pas-de-Calais, la D.I.R.E.N. Nord/Pas-de-Calais. Bailleul. FOUCAULT, B. (de) (1984). S ystematique, structuralisme et synsystematique des prairies hygrophiles des plaines atlantiques frangaises. These, Rouen. HULTEN, E. & FRIES, M. (1986). Atlas of North European Vascular Plants. Koeltz Scientific Books, Konigstein. RANUNCULUS OPHIOGLOSSIFOLIUS IN NORTHERN FRANCE 329 Ti G C 73 ba ro o o rt o' P p 3 Vi ft o ft -! ft 3 3“ P 3^ P <— ► on 3' =r o' 3“ >3 a a a 3 O a “§ a- ^ • a a" Cn C-3 Vi 72 o O Vi O 3 ft r o c/t p o 7T < EL o T Berge abrupte Berge progressive 'X Rigole de drainage Berge 4 pente douce PI’ 2V. ^ ioia Glycerio-Sparganion Frange d’Helophytes Eleocharo-Oenanthetum lineaire Bromion racamosii Eleocharo-Oenanthetum spatial Roseliere a Baldingere Eleocharo-Oenanthetum spatial Friche hygrophile Eleocharo-Oenanthetum lineaire Ceinture d’Helophytes Eau libre 331 Echiurn plantagineum L. (Purple Viper’s-bugloss) at Boscregan, St Just - its conservation and management L. BUTTERFIELD The National Trust, The Count House, Botallack, St Just, Penzance, Cornwall TR19 7QQ, England * ABSTRACT This note summarises successful attempts by the National Trust since 1995 to conserve Echium plantagineum (Purple Viper’s-bugloss) at Boscregan, St Just (W. Cornwall). The reinstatement of a low input farming regime has led to rapid increases in the population of this and other arable species, and its re-introduction onto a neighbouring National Trust farm is now being proposed. Keywords: archaeophyte, arable weeds, conservation, National Trust. INTRODUCTION Echium plantagineum L. (Boraginaceae) is a Mediterranean-Atlantic species (Preston & Hill 1997), occurring in Britain at the extreme northern edge of its range, and able to survive in the far west of Cornwall because of the mild climate of the area. It has been considered by many authors to be a native species in S.W. England (e.g. Margetts & David 1981; Cuddy 1991; Stace 1991; Ashton 1999; Butterfield 1999), although in the New Atlas of the British and Irish Flora it is described as an archaeophyte in the British Isles (Welch 2002). The species was first recorded in the Channel Islands in 1690 and in Cornwall in about 1856. ECHIUM PLANTAGINEUM AT BOSCREGAN There are a number of Cornish records of Echium plantagineum , most of these referring to casual plants turning up on waste ground or horticultural fields (French et al. 1999) - apart from at Boscregan, St Just, where it has been known since at least the 1870s. It was formerly widespread in St Just parish, but is now very largely restricted to Boscregan Farm, where it occurs in several arable fields owned by the National Trust. Its distribution appears to be limited by modern agricultural practices; in the past, it would have been spread between fields through seeds from one field being harvested with the crop, introduced into animal bedding with straw or hay and then spread on other fields with manure. Given the fact that this is a Red Data Book species, classified as ‘Endangered’ in Britain (Ashton 1999), it seemed entirely appropriate that the National Trust should be attempting to halt the decline of E. plantagineum and, hopefully, ensure its long term survival. RECENT MANAGEMENT HISTORY AND SURVEYS OF ECHIUM. PLANTAGINEUM Boscregan Farm has been under National Trust ownership since 1983, but the farm was purchased with an established long-term tenancy attached to it, and as such the Trust initially had little influence over how the land was being managed. In 1997 the farm came within the extended West Penwith Environmentally Sensitive Area (E.S.A.), which went some way to introducing a better *e-mail: lindsey.butterfield@nationaltrust.org.uk 332 ATLANTIC ARC 150 Metres Figure 1. Typical summer distribution of Echium plantagineum; dots show individual plants, numbers indicate groups of plants on 14 August 1997.. 0 150 Metres FIGURE 2. Typical autumn distribution of Echium plantagineum seedlings. ECHIUM PLANTAGINEUM IN WEST CORNWALL 333 management regime, while in 2001 the National Trust decided to take the farm in hand, assuming direct responsibility for its management and with nature conservation - and especially the conservation management of E. plantagineum - as a priority. The National Trust began surveying E. plantagineum and other arable plants at Boscregan in 1995, using nine fixed 2 m x 2 m quadrats (recorded monthly) combined with whole-field population counts and distribution mapping undertaken at least once a year. Farm management details, including sowing and harvesting dates, were also recorded. Results of these surveys indicated that in summer most E. plantagineum plants were found in the field margins (Fig. 1 & Plate 51), and that whilst some fields held very large numbers of plants (Plate 52), in others either very small numbers or none at all were found. E. plantagineum is an autumn-germinating annual, and our recording in autumn showed that large numbers of seedlings and young plants occurred over a much wider area at that time compared with the distribution in summer, including in the middle of some fields (Fig. 2). This suggested that it might be beneficial to leave one or more of these fields fallow every so often, to give these seedlings a chance to flower and increase the seed-bank. CURRENT MANAGEMENT AND FUTURE PLANS Our surveys in recent years established that E. plantagineum requires annual disturbance of the soil to allow its seeds to germinate, and tends to do best in barley fields with wide, thinly-sown field margins. As an autumn-germinating annual, it clearly benefits from fields being left as stubble over the winter. It is vulnerable to competition from vigorous arable weeds and ruderal species like Arctium minus , and does best under a ‘low input' agricultural regime. The increased knowledge gained from these surveys has helped us to devise an appropriate management scheme for the plant, and the current Management Plan for the farm is based on the view that the conservation of E. plantagineum far outweighs any economic return from the barley crop. Field margins are broad and thinly-sown with barley, whilst inputs are kept to the minimum required to keep the soil in good heart and suppress problematic vigorous perennial weeds. The costs of running the farm in this way are assisted by E.S.A. payments, and the E.S.A. winter stubble agreement also benefits ground-feeding birds. Most importantly, one field a year is cultivated but left ‘fallow’ the following year to allow E. plantagineum and other arable species to build up their seed-banks. The results of taking this approach to the management of the farm have been rapid and spectacular, with abundant E. plantagineum and Chrysanthemum segetum now recorded every year, along with an increasing population of Misopates orontium and other scarce arable plants. We are now planning to re-introduce E. plantagineum to Letcha - the neighbouring National Trust farm - where a wide diversity of arable plants were known to occur just 15 years ago, but where the fields have been in pasture for at least a decade. REFERENCES ASHTON, P. A. (1999). Echium plantagineum L., in WlGGINTON, M. J. ed. British Red Data Books, 1. Vascular plants , p. 131. Joint Nature Conservation Committee, Peterborough. Butterfield, L. (1999). Boscregan - last refuge of the Purple Viper’s Bugloss? British Wildlife 10: 166-191. Cuddy, M. (1991). Flora of St Just in Penwith. Cornwall Institute of Cornish Studies. Cornish Biological Records Unit, Redruth. French, C. N., Murphy, R. J. & Atkinson, M. G. C. (1999). Flora of Cornwall. Wheal Seton Press, Camborne. Margetts, L. J. & David, R. W. (1981). A Review of the Cornish Flora, 1980. Institute of Cornish Studies, Redruth. Preston, C. D. & Hill, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical Journal of the Linnean Society 124: 1-120. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. WELCH, D. (2002). Echium plantagineum, in PRESTON, C. D., PEARMAN, C. D. & DINES, T. D. eds. New Atlas of the British and Irish Flora, p. 496. Oxford University Press, Oxford. 334 ATLANTIC ARC List of participants (including all authors) ABBOTT, P. & C. ARNOLD, H. R. ATKINS, J. & S. ATKINSON, M. BALDOCK, N. BARTER, G. M. BASSO, F. BEARE, A. BENNALLICK, I. J. BRERET, C. M. BIORET, F. BLONDEL, C. BLOOMFIELD, D. L. BONNER, I. R. BRAVEN, J. & C. BRISTOW, C. BRUCE, J. BURCH, J. BURNS, A. BUTTERFIELD, L. BYFIELD, A. J. CAMERON, A. CARY, J. CHATER, A. O. CHEFFINGS, C. M. CLARKSON, S. COPPING, A. CORDREY, L COTTINGHAM, K. COX, J. H. S. COXON, P. CRAIG, J. S. CROFT, J. & M. CROW, E. DANIELS, R. E. DESTINE, B. DINES, T. D. DINGLE, T. J. DUCKETT, J. G. ELLIS, R. G. EVANS, S. & A. FRENCH, C. N. GARBETT, G. G. GEE, D. GOATER, B. & J. GODFREY, M. GRANT, E. GREENWOOD, E. F. Cedar Croft, 73 Ridgeway, Leeds LS8 4DD CEH Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE28 2LS Summerfield House, Brough, Kirkby Stephen, Cumbria CA17 4BX Clarence House, Higher Downgate, Callington, Cornwall PL 17 8HL Dartmoor National Park, Parke, Bovey Tracey, Devon TQ13 9JQ 1 Glan-yr-Afon, Gwaelod y Garth, Cardiff CF15 9HP C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul Teglvaerksgade 6, 2th, Copenhagen 2000 0 Denmark Lower Polmorla, St Wenn, Bodmin, Cornwall PL30 5PE 8 rue P. Cezanne, 17138 Saint-Xandre Universite de Bretagne Occidentale, CS 93837, 29238 Brest C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul Hortons, Mascall Lane, South Weald, Brentwood, Essex CM 14 5LJ Cae Trefor, Tyn y Gongl, Anglesey LL74 8SD Penally, The Crescent, Crapstone, Yelverton, Devon PL207PS 46 Chatsworth Way, Carlyon Bay, St Austell, Cornwall PL25 3SN The National Trust, 33 Sheep Street, Cirencester, Gloucestershire GL7 IRQ Royal Botanic Gardens, Kew, Richmond, Surrey 3 Rosliston Road, Stapenhill, Burton-upon-Trent, Staffordshire DE15 9RJ The National Trust, Old Farmhouse, Pentireglaze, Wadebridge, Cornwall PL27 6QY 21 Fishers Road, Totton, Southampton, Hampshire SO40 4HW The National Trust, Lower Lanner Farm, Penrose, Helston, Cornwall TR13 OP A 34 Tweedale Close, Mursley, Milton Keynes, Buckinghamshire MK17 0SB Windover, Penyrangor, Aberystwyth, Ceredigion SY23 1BJ 20 Cromwell Road, Letchworth, Hertfordshire SG6 IDT Maitlands West, Distillery lane, Colchester, Essex C02 8EZ The Nook, Swamp Lane, Roydon, Diss, Norfolk IP22 5FY The National Trust, 33 Sheep Street, Cirencester, Gloucestershire GL7 IRQ Trevethin, School Road, Pwll, Llanelli, Carmarthenshire SA15 4AL Wall Cottage, East Holme, Wareham, Dorset BH20 6AG Department of Geography, Trinity College, Dublin 2, Ireland 29 Lode Hill, Downton, Wiltshire SP5 3PW 12 Spaldwick Road, Stow Longa, Huntingdon, Cambridgeshire PE28 0TL 62 Bannawell Street, Tavistock, Devon PL 19 0DP c/o C.E.H. Dorset, Winfrith Technology Centre, Dorchester, Dorset DT2 8ZD C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul Rhyd y Fuwch, Bethel, Nr Caernarfon, Gwynnedd LL55 3PS The Bam House, Poundstock, Bude, Cornwall EX23 0DG School of Biological Sciences, Queen Mary, University of London, London El 4NS 41 Marlborough Road, Roath, Cardiff CF23 5BU Glan-y-mor, Dinas ross, Newport, Pembrokeshire SA42 0UQ 12 Seton Gardens, Weeth Road, Camborne, Cornwall TR14 7JS Hillside Cottage, Cammarth, Catharrack, Redruth, Cornwall TR16 5SA 28 Hungerford Road, Bournemouth BH8 0EH The Ridge, 27 Hiltingbury Road, Chandler’s Ford, Eastleigh, Hampshire S053 5SR 3 Castleton Avenue, Bamehurst, Bexley Heath, Kent DA7 6QT Drim cuy, Uig, Isle of Coll, Argyll PA78 6TB 10 Gayton Parkway, Gayton, Wirral, Merseyside CH60 3SS PARTICIPANTS 335 GULLIVER, R. & M. HALLIDAY, G. HARVEY, H. W. HARVEY, H. J. HAUGUEL, J-C. HAYWARD, K. HENDOUX, F. HODGSON, R. M. H. HOLLINGS, M. & O. HOPKINS, F. HOPKINS, J. J. HOUSTON, L. HOWELLS, M. ILIFF, J. & M. J0NCH M0LLER, S. JOHNSON, N. & G. JONES, R. A. KEARSLEY, D. KEIRNEN, M. KETTLE, I. J. KING, M. P. KINGSTON, N. KITCHEN, M. & C. LAHONDERE, C. LANSDOWN, R. V. H. LEACH, S. J. LEWIS, T. & N. LIGRONE, R. LINDOP, M. LISTER, F. M. LISTER, J. & K. LORIOT, S. LOVATT, A. LUPTON, D. LYNN, D. MAGNANON, S. McABENDROTH, L. McDonnell, e. j. McVEIGH, A. M. MEREDITH, H. M. MHIC DAEID, E. C. MURPHY, R. J. MURPHY, S. MYLES, S. NEWTON, A. O’CONNOR, R. OUNSTED, J. PAGE, C. N. Carraig Mhor, Imeravale, Port Ellen, Isle of Islay, Argyll PA42 7AL 26 Mowbray Drive, Burton-in-Kendal, Camforth, Lancashire LA6 INF 29 Lode Hill, Downton, Wiltshire SP5 3PW The National Trust, 33 Sheep Street, Cirencester, Gloucestershire GL7 1QW Conservatoire des Sites Naturels du Picardie, 1 place Ginkgo, 80044 Amiens 6 Crouch Drive, Witham, Essex CVM8 1TD C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul Gorselands, Axtown Lane, Yelverton, Devon PL20 6BU St Cyprien, Ham Manor Close, Angmering, Littlehampton, West Sussex BN16 4JD Conservatoire Botanique National de Brest, 52 allee du Bot, 29200 Brest English Nature, Northminster House, Peterborough, Cambridgeshire PEI 1UA 2 West Grove, Montpelier, Bristol BS6 5LS 63 Ashleigh Road, Derwen Fawr, Swansea SA2 8EE Eithin Tewion, Cilycwm, Llandovery, Dyfed SA20 0TF Teglvaerksgade 6, 2th, Copenhagen 2000 0 Denmark Cassandene, Station Road, Soberton, Hampshire S031 3QU Yr Uchelgaer, Io Lon Penparcau, Aberystwyth, Ceredigion SY23 1PB c/o C.C.W., Winchway House, Winch Lane, Haverfordwest SA61 1RP Grianan, Drumclach, Isle of Colonsay, Argyll PA61 7YR 75 Dupont Road, Raynes Park, London SW20 8EH 91 South Court Avenue, Dorchester, Dorset DTI 2DA Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6 The Cottage, Bevingtcn, Berkely, Gloucestershire GL13 9RB Societe Botanique du Centre-Ouest, 94 avenue du Parc, 17200 Royau Floral Cottage, Upper Springfield Road, Stroud, Gloucestershire GL5 1TF English Nature, Roughmoor, Bishop's Hull, Taunton, Somerset TA1 5AA 12 Cannisland Park, Parkmill, Swansea SA3 2ED Departimento di Scienze Ambientali, Seconda Universita di Napoli, Via. Vivaldi 43, 81100 Caserta, Italy 36 Woodland Hill, Whitkirk, Leeds LSI 5 7DG Broadmoor Farm, Stoketon, Saltash, Cornwall PL 12 4SA The National Trust, Killerton House, Broadclyst, Exeter, Devon EX5 3LE Conservatoire Botanique National de Brest, 52 allee du Bot, 29200 Brest The Eden Project, Bodelva, St Austell, Cornwall PL24 2SG Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6 Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6 Conservatoire Botanique National de Brest, 52 allee du Bot, 29200 Brest Department of Biology, University of Plymouth, Plymouth, Devon PL4 8AA The Old Gasworks Cottage, The Lerbume, Wedmore, Somerset BS28 4ED 15 Willow Road, Great Horwood, Milton Keynes, Buckinghamshire MK17 0QH 17 Esplanade Road, Newquay, Cornwall TR7 1QR Avondale, Moynalty, Kells, Co. Meath, Ireland Shangrila, Reskadinnick, Camborne, Cornwall TR14 0BH National Parks & Wildlife Service, 7 Ely Place, Dublin 2 E.R.C.C.I.S., Five Acres, Allett, Truro, Cornwall TR4 9DJ 6 Stanley Walk, Exmouth, Devon EX8 5QD Springthwaite, Aisgill Moor Cottages, Kirkby Stephen, Cumbria CA7 4JY Apple Tree Cottage, Woodgreen Common, Fordingbridge, Hampshire SP6 2BD Gillywood Cottage, Trebost Lane, Stithians, Truro, Cornwall TR3 7DW 336 ATLANTIC ARC PANKHURST, T. J. PARKER, J. PARKER, S. J. PARSLOW, R. E. PEARMAN, D & A PERRING, F. & M. PEYTOUREAU, Y. PORLEY, R. D. PORTCH, M. F. PRATT, E. PRESSEL, S. PRESTON, C. D. PRIEST, S. PRYCE, R. D. RANDALL, R. E. REDGRAVE, L. RICH, T. & L. ROBARTS, J. ROBSON, N. & E. RUMSEY, F. R. SALIOU, P. SAUSMAREZ, N. de SELBY, P. SHEAHAN, M. C. SMITH, R. SOUTHEY, J. F. SPROULL, J. STENBERG, L. STEWART, E. STEWART, N. F. STRIBLEY, M. J. SUTCLIFFE, J. SUTTON, M. SYDES, C. TAYLOR, S. THOMSON, P. TOUSSAINT, B. TURNER, C. WALDREN, S. WALKER, A. WALLS, R. & J. WATSON, G. K. WHEELER, B. WHITEWAY, B. J. WOODHEAD, F. WOODWARD, S. F. 44 The Avenue, Leighton Bromswold, Huntingdon, Cambridgeshire PE 18 OSH The Botanic Gardens, Cory Lodge, Bateman Street, Cambridge CB2 1JF 26 Laburnum Road, Wellington, Somerset TA21 8BL 17 St Michael’s Road, Ponsanooth, Truro, Cornwall TR3 7ED Algiers, Feock, Truro, Cornwall TR3 6RA Green Acre, Wood Lane, Oundle, Peterborough PE28 5TP Societe Botanique du Centre-Ouest, 230 rue de la Soloire, 16200 Nercillac English Nature, Foxhold House, Crookham Common, Thatcham RG19 8EL 4 Belgrave Close, Abergavenny, Gwent NP77 7AP 7 Bay Close, Swanage, Dorset BH19 IRE School of Biological Sciences, Queen Mary, University of London, London El 4NS C.E.H. Monks Wood, Abbots Ripton, Huntingdon, Cambridgeshire PE28 2LS Yonder Cottage, Ashford Hill, Thatcham, Berkshire RG19 8 AX Trevethin, School Road, Pwll, Llanelli, Carmarthenshire SA15 4AL Girton College, Cambridge CB3 OJG 1 Kingston Drive, Nailsea, Bristol BS48 4RB Department of Botany, National Museum of Wales, Cathays Park, Cardiff CF10 3NP 233 Bruemish, Isle of Barra, Western Isles, Scotland Bum Edge, 48 Granville Road, Limpsfield, Oxted, Surrey RH8 ODA Department of Botany, Natural History Museum, Cromwell Road, London SW7 5BD C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul Point House, Point, Devoran, Truro, Cornwall TR3 6NJ 12 Sedgewick Road, Bishopstoke, Eastleigh, Hampshire SO50 6FH 81 Westmoreland Road, Names, London SW13 9RZ Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6 2 Old Butterleigh Road, Silverton, Exeter, Devon EX5 4JE 3 Mill Cottages, Lower Sticker, St Austell, Cornwall Fastlagsvagen 13, 126 48 Hagerstan, Sweden 33 Woodlands Street, Milnagavie, Glasgow G62 8NS Cholwell Cottage, Posbury, Crediton, Devon EX 17 3QE 20 Green Close, Truro, Cornwall TR1 2DD English Nature, Northminster House, Peterborough, Cambridgeshire PEI 1UA c/o C.C.W., Winchway House, Winch Lane, Haverfordwest SA61 1RP Scottish Natural Heritage, 2 Anderson Place, Edinburgh EH6 5NP 46 Common Road, Bressingham, Diss, Norfolk IP22 2AZ Hall Pool, Marden, Hereford HR1 3EN C.R.P./C.B.N.B.L., Hameau de Haendries, 59270 Bailleul 5 Mill Road, Great Gransden, Sandy, Bedfordshire SGI 9 3 AG Trinity College Botanic Garden, Palmerston Park, Dartry, Dublin 6 31 Hughenden Gardens, Glasgow G12 9YH 16 Leigham Vale Road, Bournemouth, Dorset BH6 3LR 50 Sunningdale Drive, Bromborough, Wirral CH63 0JE 64 Station Road, Horrabridge, Yelverton, Devon PL20 7RD 8 Bosvean Gardens, Illogen, Redruth, Cornwall TR16 4DH 28 Hungerford Road, Bournemouth, Dorset BH8 0EH 19 Highfield Road, Groby, Leicester LE6 0GU PLATE 1. Franklyn Hugh Perring (1927-2003) botanist, nature conservationist and author PLATE 2. Kynance Cove in the Lizard. The Lizard is a battered fragment of oceanic crust which is, relative to most other rocks in SW England, rich in magnesium and iron and deficient in silica. This gives the vegetation on the Lizard a distinctly different character to the rest of Cornwall (p. 22). Plate 3. Cliff section through Quaternary deposits exposed in a cliff west of Spit Beach, near Par. Cornwall. Resting on the Devonian slate bedrock is a raised beach, with a considerable thickness of Head overlying it. Just under the soil is a layer of loessic material c. 0-5 m thick (pp. 26 & 27). Plate 4. Perran Sands near Perranporth. The dunes are formed of carbonate-containing sand blown up from the beach by strong westerly winds. They cover about 4 km2. The carbonate is derived from the shells of marine organisms. ‘Plumes’ of alkaline soil are derived from this sand being blown inland (p. 28). PLATE 5. Estuary at Looe, south Cornwall. During the coldest periods of the Quaternary, when sea level fell by about 120 m. this valley would have contained a river flowing seawards about 30 m below present sea level. This river would have had huge erosive capacity when spring meltwater floods rushed down. Sea level rise since the end of the last cold period of the Quaternary has flooded the valley to produce a ‘ria' (p. 26). PLATE 6. Newbridge ball clay pit in the Bovey Basin, South Devon. Beds of sand, clay and lignite alternate, see earthmoving plant for scale. Studies of the palynology and the fossil wood associated with the lignites have enabled a good picture of the flora to be built up. Two floras are represented (Selwood 1984): a swamp flora in the basin, and an upland flora dominated by Sequoia couttsiae. Climate was sub-tropical or warm temperate (p. 24). PLATE 7. An 'ice-wedge' exposed in a low cliff at the back of Spit Beach, near Par, Cornwall (p. 27). Plate 9. Bulb field on St Mary's with Eastern Gladiolus, Gladiolus communis subsp. byzantinus. Photo: R. Parslow (p. 71). PLATE 10. Whorled Caraway, Carum verticillatum showing close-up of flowers (a) and leaves (b). Photo: R. Pryce (p. 88). PLATE 11. A rushy M23 hay meadow with abundant Whorled Caraway, Carum verticillatum near Llandybie on the Carmarthenshire coalfield. Photo: R. Pryce (p. 87 & 88). PLATE 12. Ground-level view of fern-rich Celtic Hedges along a bridle-track in a Cornish coastal landscape. Lizard peninsula, c. 20 m alt.. May. Photo: C. Page (p. 93). PLATE 13. Ground-level view of fern-rich Celtic Hedges along a quiet lane in a Cornish inland landscape included within the Parish survey quoted. Stithians Parish, south of Truro, c. 100 m alt, July. Photo: C. Page (p. 97). PLATE 15. Ophrys vasconica (O. et E. Danesch) P. Delforge: Le Chateau d'Oleron. 05.1998. Photo: J.-M. Mathe (p. 1 15). PLATE 17. Liparis loeselii (L.) L. C. M. Richard: Saint-Trojan. Oleron. 7.07.1998. Photo: J.-M. Mathe (p. 117). Plate 14. Ophrys passionis Sennen ex J. et P. Devillers-Terschuren: Fief Melun. Le Chateau d'Oleron. Oleron. 15.04.1991. Photo: J.-M. Mathe (p. 115). PLATE 16. Epipactis phyllanthes G. E. Smith. Saint- Trojan. Oleron. . 01.07.1998. Photo: J.-M. Mathe (p. 116). PLATE 18. Long-headed Clover, Trifolium incarnation subsp. Molinerii. Photo: R. Parslow (pp. 121 & 125). Plate 19. Low Bar, Penrose looking west across shingle vegetation. Photo: NT/K. N. Alexander (p. 135). PLATE 20. Irish Heath, Erica erigena R. Ross. PLATE 21. Mackay’s Heath, Erica mackaiana Bab. Photo: S. Waldren (p. 147). Photo: S. Waldren (p. 147). Plate 22. A stand of Irish Heath, Erica erigena in the Charite marsh, in the Medoc region of S.W. France 5 January 2003. Photo: Y. Peytoureau (p. 178). Plate 23. Campylopus brevipilus (arrowed) growing with C. introflexus on the Lizard Downs (p. 167 & 172). Plate 24. Tuft of Campylopus brevipilus (arrowed) transplanted into C. introflexus, Thursley Bog. Surrey (p. 167 & 172). PLATE 25. Weissia controversa var. densifolia forming extensive yellow-green mats on lead mine spoil, Ventongimps, East Cornwall (p. 171). PLATE 26. Cultures of protonemata of Weissia controversa var. densifolia (a) and W. controversa var. controversa (b). Their highly distinctive appearance suggests that var. densifolia merits specific status (p. 173). PLATE 27. Vigurs’ Eyebright, Euphrasia vigursii Davey. Photo: R. Hodgson (p. 181). PLATE 28. Wild Asparagus, Asparagus prostratus Dumort. Photo: R. Parslow (p. 231). PLATE 29. Hairy Greenweed, Genista pilosa L. PLATE 30. Dwarf Rush, Juncus capitatus Weigel at Photo: J. C. Haugel (p. 243). Mullion, Lizard peninsula, 3 May 2004. Photo: I. J. Bennallick (p. 249). Plate 31. The Polruan Sand Crocus, Romulea colwnnae Sebast & Mauri site 6 May 2002. Photo: I. J. Bennallick (p. 248). Plate 32. Sand Crocus. Romulea columnae Sebast & Mauri at Polruan, close-up showing fruits, 6 May 2002. Photo: M. J. Stribley (p. 248). PLATE 33. Sand Crocus, Romulea columnae Sebast & Mauri at Polruan, view of whole plant 6 May 2002. Photo: M. J. Stribley (p. 248). PLATE 34. Slender Bird’s-foot-trefoil, Lotus angustissimus L. Photo: R. D. Porley (p. 255). PLATE 35. Toadflax-leaved St John's-wort. Hypericum linariifolium Vahl. Photo: T. C. G. Rich (p. 251). PLATE 37. (a) Teucrium scordium subsp. scordium from the Oise valley (Picardie) (b) T. scordium subsp scordioides from Brittany. (Both plants in ex situ cultivation.) (p. 261 ). PLATE 39. Seeds of Water Germander, Teucrium scordium L. (p. 263). PLATE 40. Phleo arenarii-Tortuletum ruraliformis grassland with Early Sand-grass, Mibora minima (L.) Desv. in the Perroquet Dunes. Photo: D. Mercier (p. 273). PLATE 41 . Clearing scrub in the Dunes of Merlimont to restore habitat for Early Sand-grass, Mibora minima. (p. 275). PLATE 42. Dwarf Pansy, Viola kitaibeliana Schult. - a tiny spring-flowering annual restricted within the British Isles to the Channel Islands and the Isles of Scilly. Photo: R. Parslow (pp. 67 & 277). Plate 43. Sea Knotgrass. Polygonum maritimum L. Photo: M. J. Stribley (p. 281). Plate 44. Irish Lady’s-tresses, Spiranthes romanzoffiana Cham. Photo: S. Waldren (p. 292). Plate 45. Irish Lady’s-tresses, Spiranthes romanzoffiana Cham, at site GF on Coll, 8 August 2002. Photo: E. Grant (p. 292). PLATE 46. Irish Lady’s-tresses, Spiranthes romanzoffiana Cham, at site GF on Coll, 1 October 2002, showing withered, unexpanded capsules. Photo: E. Grant (p. 292). PLATE 47. North Cornish coast looking north from Godrevy near the lighthouse. Photo: T. Opie (p. 53). Plate 48. Typical leaves collected from Creeping Buttercup, Ranunculus repens L. collected from a turlough (left) and a pasture adjacent to a turlough (right). Photo: S. Waldren (p. 302). a PLATE 49 Coolorta turlough. County Clare, Ireland, photographed from approximately the same position during winter (a) and summer (b). Photos: S. Waldren. (p. 302). PLATE 50. Pillwort, Pilularia globulifera L., Cargenwen Reservoir, Crowan (near Redruth) SW64, abundant in reservoir margin, refound 12 October 2004 after a gap of 40 years. Photo: I. J. Bennallick (p. 326). PLATE 51. Purple Viper’ s-bugloss, Echium plcintagineum L. thriving in a broad field margin at Boscregan. Photo: L. Butterfield (p. 333). PLATE 52. Masses of Echium plcintagineum and Chrysanthemum segetum in a field corner at Boscregan. Photo: L. Butterfield (p. 333). Whiffs, THE NATURA1 HISTORY MUSEuf f.t 2 2 SEP m PURCHASED BOX ANY LIBRARY •v«^4 This volume comprises a series of 45 papers and posters presented at the symposium ‘Botanical Links in the Atlantic Arc’, organised by the Botanical Society of the British Isles (B.S.B.I.) and held in May 2003, in Camborne, Cornwall. French abstracts have been provided by Yves Peytoureau. This Anglo-Hiberno-French meeting, the brainchild of the late Dr Franklyn Perring, gave botanists an opportunity to come together to celebrate the wild plants of the Atlantic coastal regions of Europe, and to consider a whole range of topics concerning their distribution, autecology, conservation and management. Without his vision and dogged persistence, the symposium would never have happened and this volume is respectfully dedicated to his memory. English Nature assisted financially when the symposium was being planned, and have provided a substantial grant towards the cost of publishing these proceedings. ISBN 0 901158 33 X The cover photograph by Stuart Croft (R.S.P.B.) shows the west coast of The Lizard, Cornwall.