Temporal and spatial heterogeneity of trophic interactions within below-ground food webs

Abstract

Real food webs are dynamic multi-dimensional systems, whereas the descriptions of real food webs often do not capture this complexity in that they have been confined to a single habitat (the community web sensu Cohen, 1978) and do not represent changes in time (Paine, 1988). We present an analytical approach that uses univariate and multivariate statistics and simulation modeling to study pattern within food webs. As an illustration of the approach, we compared below-ground food webs from natural and agricultural ecosystems in terms of their architecture, temporal dynamics of the biomass of functional groups, and the temporal and spatial dynamics of energy channels. The complexity and diversity of below-ground food webs are similar to the detritus-based food webs of other terrestrial and aquatic habitats. The pattern of the flow of nitrogen through the below-ground food webs of the Shortgrass Steppe of North America is similar to that of the food web of agricultural soils of reclaimed marine sediments in The Netherlands. The webs are compartmented along dominant flows of energy (energy channels) originating from primary production and detritus. Comparisons of the connectedness descriptions, and implementations of cluster analysis, canonical discriminant analysis and analysis of variance of temporal biomass of functional groups within food webs of soils from North America and The Netherlands, indicate that the detritus energy channel can be further compartmented into a fungal and bacterial channel. For winter wheat soils in The Netherlands, the degree of compartmentalization appears to depend on management practice. Consumers of fungi were separated in time from consumers of bacteria in the integrated management practice, while little separation was observed in conventional practice. Our study indicates (1) that analyses of food webs should aim to project the web onto the principal niche dimensions food, habitat and time, and (2) that quantitative measures of community structure — identification of functional groups, the biomass and productivity of functional groups, and the flow of nutrients within energy channels — are useful measures of food web structure.