Watsonia 26: 347–358 (2007) ECOLOGY OF EUPHRASIA VIGURSII 347 A fine scale study of selected environmental and floristic parameters in three populations of Euphrasia vigursii (Davey), a rare annual endemic to Devon and Cornwall L. GRANADOS1 and S. D. LANE2 School of Biological Science, University of Plymouth, Plymouth PL4 8AA ABSTRACT surveys have revealed only 25 active sites. This represents approximately 50% loss of localities Euphrasia vigursii is an endangered hemiparasitic in Cornwall mainly due to habitat destruction annual, endemic in Devon and Cornwall. This (French et al. 1999). In Devon there are only 2 investigation has evaluated fine scale patterns of sites, rediscovered in 1995 and 1998 and floristic composition and nutrient distribution in situated on the north-west perimeter of selected populations of E. vigursii in Devon and Cornwall to increase understanding of their ecology Dartmoor National Park. Ecologists at the for use in future management strategies in Devon. National Park aim to conserve and increase the Two particular questions were formulated: population of E. vigursii as part of the 1. Is E. vigursii immediately surrounded by other Dartmoor Biodiversity Action Plan, as it has particular plant species? been categorised in the priority list of the UK 2. If the soil was analysed in the direct location of steering group report on biodiversity as being the site of E. vigursii would the pH, Nitrogen, of ‘global conservation concern’ (HMSO Phosphorus and Potassium content be different from 1995). In order to establish a management the areas where E. vigursii is not present? strategy that would address the objectives of Phytosociological analysis indicated that E. vigursii is associated with heath and acidic grassland com- Dartmoor National Park authority, a greater munities typically containing Ulex gallii, Agrostis understanding of the ecology of E. vigursii and curtisii, Calluna vulgaris, together with Molinia specifically knowledge of the local populations caerulea and Potentilla erecta mire. Within the is essential. individual sites spatial scale comparison (macro vs. The processes and mechanisms that result in micro) indicated significant differences in species being rare and endemic are complex Phosphorus levels and pH levels but these were not traits involving interactions of abiotic, biotic consistent across the sites. Germination of E. vigursii and historical factors. Investigating the factors at microenvironment level was closely correlated which influence the distribution of E. vigursii with the presence of A. curtisii rather than its other known host species Ulex gallii. This is contrary to must therefore involve an amalgamation of published research which suggests that E. vigursii demographic, genetic, resource allocation avoids the more open, species-poor areas dominated (Davy & Jefferies 1981; Palmer 1987; by A. curtisii. Correlations between the environ- Schemske et al., 1974) alongside historical mental parameters assessed may reflect present man- systematic and stochastic perturbations (Shaffer agement regimes within the Dartmoor populations. 1981). E. vigursii is a hemiparasitic summer annual, KEYWORDS: spatial scale, species distribution, an erect to 20 cm branched therophyte (Grime hemiparasitic, soil nutrients, conservation, Vigurs’ 1996) located in heathland containing Ulex Eyebright. gallii and Agrostis curtisii in both Devon and Cornwall. The leaves are dull greyish green INTRODUCTION often suffused with violet or black (Stace 1999) due to anthocyanin pigment (Pankhurst 1977). E. vigursii (Vigurs Eyebright) is confined to The corolla is usually lilac/purple/reddish and Cornwall and Devon where it is considered on the upper leaves and bracts are long endangered and is classified as vulnerable glandular hairs. E. vigursii is only one of four (Cheffings & Farrell 2005). Until 1979 it had diploids in the genus and considered a stable been located in 53 sites in Cornwall; later hybrid between Euphrasia micrantha and 1e-mail:[email protected] 2e-mail: [email protected] 348 L. GRANADOS & S. D. LANE FIGURE 1. Location of study sites in Devon and Cornwall. Euphrasia anglica (Pankhurst 1977). An open target species E. vigursii to determine at grass sward provided by regular grazing and individual plant level the composition of controlled burning (swaling) is considered neighbouring plant associations together with characteristic of the habitat in which E. vigursii recordings of soil nutrients (Potassium, Phos- is found (Norman Baldock, Ecologist Dartmoor phorus and Nitrogen) and pH status. The National Park; pers. comm.. 2000). Very little second survey involved macrosampling to give specific research on E. vigursii has been con- a general indication of the vegetative com- ducted: therefore, the aims of this study have munity composition in which E. vigursii was been to evaluate some of the ecological and found together with the soil nutrient and pH environmental parameters influencing three status within areas where E. vigursii was not separate populations of E. vigursii to gain present. better understanding of its ecology and habitat requirements and to answer two specific SITE 1 questions: Lydford, Dartmoor National Park, Devon. Altitude: 320 m. Sample site size: 10 m × 50 m 1. Is E. vigursii immediately surrounded by In the surveyed area Ulex spp. (gorse) and other particular plant species? heather formed a mosaic of islands surrounded 2. If the soil was analysed in the direct by pathways containing a grass and herbaceous location of E. vigursii would the pH, sward. E. vigursii was located in most Nitrogen, Phosphorus and Potassium con- instances at the parameter of mature and new tent differ from areas where E. vigursii is vegetative growth of U. gallii and where A. not present? curtisii in the ‘pathways’ was of low sward height and also where periodic swaling had taken place. MATERIALS AND METHODS SITE 2 Two surveys were carried out at each of three St. Neot, Bodmin Moor Cornwall. locations in Devon and Cornwall (Fig. 1). The Altitude: 170 m. Sample site size: 7 m × 50 m. first survey entailed micro sampling of the The surveyed area contained a linear grass ECOLOGY OF EUPHRASIA VIGURSII 349 and herbaceous sward, 15 m wide, which cut COMMUNITY CLASSIFICATION AND through a patchy landscape of gorse, heather PHYTOSOCIOLOGY and scrubland at the perimeters. E. vigursii was situated within the short grass sward from the In order to define the floristic community in edge of tall Ulex spp. at the perimeter of the which E. vigursii occurs the floristic data survey area to within 5 m of the linear grass (using percentage cover data) were classified sward. using two-way indicator species analysis and the computer program TWINSPAN. The SITE 3 TABLEFIT programme was then used to link St. Agnes, nr Redruth, Cornwall. the resulting communities in to the categories Altitude: 85 m. Sample site size: 3·5 m × 50 m of the National Vegetation Classification In the surveyed area there was a blanket of (Dinsdale 1997). dwarf gorse and heather species with pathways The total micro floristic data (frequency formed by walkers and erosion by the weather. data) were also run through TWINSPAN and E. vigursii was situated approximately 20 m Detrended Correspondence Analysis computer from the cliff edge on one of the many worn program applied using PC ORD in order to paths that connect the coastal path that runs examine which were the associative and most along the cliff edge with a parallel inland frequent floristic species with E. vigursii. This pathway. It was present amongst shorter veg- was carried out for the micro floristic data etation at each side of the pathway although a between and in each individual site. greater percentage of the population was situated north east facing and was therefore RESULTS sheltered from south-westerly winds. Floristic data and soil cores were obtained from SURVEY METHODS a total of 180 quadrats collected from the three sites. The 90 micro quadrats yielded 21 plant Each survey site was marked out into a 1 m2 species in total while the macro quadrats grid framework and random co-ordinates used contained 28 plant species (Table 1). Bare for macro sampling (1 m2 quadrats). Micro ground and dead material were also recorded . sampling used 40 mm circular quadrats that All the species occurred at both survey scales, were positioned non-randomly; the quadrat was apart from Poa spp. which was only found only placed over the target species (E. vigursii in the in the micro quadrats. centre) and species frequency within the microsampling area recorded. Soil samples CLASSIFICATION AND PHYTOSOCIOLOGY were collected 10 mm from the stem of Euphrasia vigursii. For the macro survey a 1 Interpretation of two-way indicator species m2 collapsible quadrat was randomly posi- analysis and definition of plant communities tioned in the survey area and the percentage The classification groups resulting from cover for each species was recorded. TWINSPAN analysis and synoptic frequency Identification of the floristic species was of occurrence within each group are shown in carried out using keys by Stace (1999), Table 2. Five TWINSPAN groups were Hubbard (1985) and Fitter et al. (1995). defined and their associated NVC plant Soil samples were taken from the centre of communities categorised (Table 3). Groups 1 the quadrat using a 15 mm × 150 mm metal and 2 were heath communities of Site 3, St corer, to a depth of 25 mm to 75 mm de- Agnes. In group 1 Agrostis curtisii , Erica pending on soil texture. In each survey site 60 cinerea, Potentilla erecta, and Ulex gallii, soil samples were taken; 30 in the macro Thymus spp., Serratula tinctoria and Ulex quadrats and 30 in the circular micro quadrats. europaeus were frequent; however, there was a low frequency of Erica tetralix. In group 2 ANALYSIS OF SOIL SAMPLES Ulex gallii, Thymus species, Erica tetralix, Carex flacca, Serratula tinctoria, Viola spp., In total the surveys yielded 180 soil samples. Agrostis curtisii and Potentilla erecta were Each was crushed and put through a 2 mm frequent.