Assessing spatiotemporal predator-prey patterns in heterogeneous habitats

Disentangling the contribution of biotic interactions (density-dependent) and environmental heterogeneity (density-independent) to the formation of spatial patterns between predators and prey is crucial for a better understanding of food-web interactions Most techniques for the analysis of spatial patterns assume that abiotic processes influence the distribution of individuals with similar intensity at all locations of a study area (stationarity) This simplification may result in a spurious description of predator prey associations in environmentally heterogeneous habitats In a spatially explicit way we sampled ground-active linyphiid and lycosid spiders and their Collembola prey along a forest-meadow gradient and analysed the change in spatial relationships with time We used techniques of point pattern analysis and pair-correlation functions to summarize spatial patterns To disentangle the contribution of biotic interactions and environmental heterogeneity on pattern formation we compared observed functions with those arising from null models either assuming environmental homogeneity or accounting for habitat heterogeneity All taxa were aggregated at the three sampling periods if habitat homogeneity was assumed, but only linyphiid spiders were still clustered after accounting for environmental heterogeneity A similarly contrasting result was present for the spatial relationship between predators and their prey, with association under the assumption of homogeneity, but strong repulsion that intensified with time if accounting for environmental heterogeneity Results from additional bivariate null models under which either predator or prey locations were fixed, suggest that Collembola showed lower activity density in more suitable, but predator-rich habitats Biotic interactions were important drivers of the spatial distribution of ground-active predators and their decomposer prey in the analysed forest floor food-web However, these structuring forces remain hidden when using simple spatial models that ignore environmental heterogeneity Therefore, for understanding predator prey interactions in spatially complex habitats, such as grasslands and forests, spatial models considering habitat heterogeneity are indispensible