Arachnologische Mitteilungen 35

52 S. Buchholz & V. Hartmann half of the 19 th century and is under British control today. It is periodically used for training (tanks, track vehicles) with different intensities from site to site (S TARKMANN et al. 1993, G ROSSE 1995). Site descriptions Seven sites (D1-D7) were studied at the military training area (Tab. 1). Five sites (D1-D5) were located in open habitats whereas D6 and D7 were situated on woodland edges. Methods Pitfall traps were used to monitor the active den- sities of the ground-dwelling spiders for one year (08.04.2002 to 18.04.2003). In five sites (D1-D5) five traps (glass jars, diameter 9 cm) were installed with a distance of 5–10 m, arranged as a cross. In two sites (D6, D7) three traps were placed in line, about 5 m apart. A 3% formalin solution with deter- gent was used as the killing agent and preservative. The traps were emptied fortnightly during the warm seasons and once a month during the winter. In addition to pitfall trapping, one hand sampling session was carried out in June 2002. Statistical methods Dominances [D %] were calculated according to E NGELMANN (1978). Measurements of diversity were reported for each site. For alpha-diversity (diversity in individual sample units) the Brillouin index was calculated using H B = (ln N! – ∑ ln n x !)/N (N = total number of individuals; n x = specimens of species x). For an approximation of factorial Stirling’s formula (ln n! ≈ n * ln n) was used. Fur- thermore the Brillouin-Evenness E B was calculated as follows: E B = H B / H Bmax (H B = Brillouin index; H Bmax = ln S where S is the total number of species) (K RATOCHWIL & S CHWABE 2001). Comparison of the epigeic spider communities was made by Principal Component Analysis (PCA) using Canoco 4.5. To compare the sites, data were standardized (individual sums x 100/number of sampling days/number of pitfall traps). For PCA the abundances of each species were log transformed. Only species with a dominance of >1% per site were included in the analyses. Results Altogether, 11194 mature spiders from 141 species and 20 families were caught by pitfall trapping and hand sampling (Tab. 2). The most abundant families were Linyphiidae, Lycosidae and Gnapho- sidae. Dominant species were Pardosa monticola and Pardosa palustris . Alopecosa cuneata , A . pulverulenta , Drassyllus pusillus , Hahnia nava , Pardosa hortensis , P . lugubris s. str. and Trochosa terricola were also nume- rous. Rare or previously rarely recorded species were Micaria subopaca , Panamomops mengei , Parapelecopsis nemoralis and Pelecopsis mengei . A status of endan- germent for North Rhine-Westphalia is given for 18 species: Five species are endangered (category 3) and one, Euophrys herbigrada , is highly endangered (category 2). For 11 species an endangerment can be assumed (category V). Meioneta simplicitarsis is considered extinct (category 0). Tab. 1 : Site characteristics (hum = humidity: hum1 = very dry, hum2 = dry, hum3 = fresh, hum4 = humid; sha = shading: sha1 = open, sha3 = partially shaded, sha4 = shaded; dist = disturbance). site habitat vegetation structure humidity shade disturbance D1 heathland mosaic of bare sand and Calluna shrubs hum2 sha1 habitat management (once a year) D2 rough grazing dense herbal-layer with a low height of vegetation (about 10 cm) hum2 sha1 grazing, training (periodical) D3 dry grassland sparse herbal-layer in places bare sand and mosses hum1 sha1 animals (e. g. rabbits) (periodical) D4 rough meadow dense herbal-layer (about 20-30 cm) overgrown with shrubs hum3 sha1 - D5 rough meadow dense herbal-layer (about 30-40 cm) overgrown with shrubs hum3-4 sha1 - D6 edge of a pine-forest ( Pinus silvestris ) mosaic tussocks and shrubs in places bare sand hum2 sha3 training (periodical) D7 edge of a oak - forest ( Quercus robur ) dense and high-growing herbal-layer (about 50 cm) hum4 sha4 -