Landscape and site effects on Collembola diversity and abundance in winter oilseed rape fields in eastern Austria

Soil animals are rarely considered in landscape ecology although recent findings show that landscape composition and habitat fragmentation may exert a strong influence on their communities. We assessed the relationships between landscape parameters and the species richness of Collembola (springtails) in agricultural eastern Austria, hypothesizing that the richness of surface active (epigeic) species are more significantly correlated with landscape-scale parameters than soil living (endogeic) species. We used pitfall traps and soil samples to measure species richness and abundance of the two life forms in 29 oilseed rape fields in differently structured landscapes. Epigeic species richness was positively correlated to landscape parameters like landscape diversity, habitat richness, isolation of open habitats and area of oilseed rape fields at small to medium (250–1000 m) and at larger (1000–1750 m) landscape scales around the investigated sites. Endogeic species richness was also significantly correlated with landscape parameters like landscape diversity and road-side strip length. Total species richness was significantly correlated with landscape diversity at a larger scale, isolation of open habitats and proportion of woody area at a smaller spatial scale. Landscape diversity measures explained a higher percentage of the diversity of epigeic (42%) than of endogeic diversity (32%). Epigeic abundance was not significantly correlated to any of the site parameters, but with landscape parameters like oilseed rape area, woody area and minimum distance to dry grassland. Endogeic abundance was only correlated to oilseed rape cover. Our results suggest that parameters associated with landscape diversity can be good predictors of Collembola diversity at two spatial scales: At small scales probably due to active migration from bordering hedges, forests or grasslands; and at larger scales possibly due to passive wind dispersal. Soil ecologists rarely take landscape-scale variables into account, and our findings may contribute to broader recognition of the importance of landscape parameters in investigations of soil biota.