A study of the flora and invertebrate fauna of magnesian grassland (of Thrislington plantation SSSI), with particular reference to plant architecture and invertebrates

Abstract

1/. Relationships between plant architecture and invertebrates have been widely described, (Lawton 1983 and 1986, Southwood, Brown and Reader, 1979). This study aimed to investigate some part of the invertebrate fauna with particular reference to plant architecture and the affects of disturbance. 2/.Thrislington was chosen as the study site as it consisted of a number of transplantated sites (Fig. 1). These represent a series of 8 Magnesian limestone grassland sites of known age and site history. Previous studies of die flora (Shimwell 1968 and Parks 1982 and 1989) and invertebrates (Sheppard, 1987 & 1989 and Woodford and Bruce, 1983) have produced a detailed history of this Grade 1 SSSI prior to it's transplant. 3/.A series of 25 fixed quadrats (2 square metres) were positioned throughout 5 of the 8 transplanted sites i.e. 1,3,4,6,8. Five quadrats were placed randomly within each of the five chosen sites. These quadrats formed the basis of die vegetation, architecture and invertebrate surveys. 4/.The survey yielded 82 plant species occuning within the fixed quadrats and 112 species throughout the 5 sites surveyed. The sites were found to become more homogeneous with increasing age. The diversity of plant species was greatest on site 6 (removed 3 years ago) with 84 species occurring on that site and lowest diversity on site 1 i.e. 59 species and moved 8 years ago. 5/.The architecture survey data and vegeatation data were used to run DCA analyses. The analyses were carried out on the individual quadrat data and both showed a certain degree of separating with respect to the sites they were taken from in diat vegetation (in terms of species) and architecture become increasing homogeneous with age, at Thrislington. The CCA analyses showed that % bare earth and % moisture content were most influencing the biplot along die first and second axes. 6/.A DCA and CCA analyses were carried out on weevils, spiders, carabids and "all invertebrates". The "invertebrate" CCA showed that site age (axis 1) and the height classes 0-2 & 2-5cm (2nd axis) explained the greatest amount of variance. The weevil CCA showed that pitfall position and 10-20cm were the two most important (measured) variables. The spider CCA showed the greatest association with average moss depth and 2-5cm. While the carabids showed the greatest association with 2-5cm and % moss cover. 7/.In addition to the quadrats and pitfalls further invertebrate groups were selected for study by different techniques - walking transect for butterflies, hand searches for molluscs and counts for Cercopidae nymphs within each of the 25 fixed quadrats. The butterfly transects revealed that the small heath was the commonest of the 13 species recorded and Site 3 had constantly high counts throughout the study period. The mollusc survey yielded 8 species (5 snails and 3 slugs). The commonest large snail was Cepaea nemoralis. No significant correlation was found between Cepaea numbers and environmental variables measured, using correlation coefflcents. The Cercopidae nymphs counts produced a total of 44 host plant species, five of which accounted for 80% of the total counts. An estimate of density was calculated for each site and an average density per square metre on. each of the four sampling occasions. Multivariate techniques using the DCA and CCA analyses run on the CANOCO programme were found to be a robust and reliable techniques for investigating the interrelationships between species and environmental variables.