Fragmented nature: consequences for biodiversity

We discuss how fragmentation of resources and habitat operate differently on species diversity across spatial scales, ranging from positive effects on local species coexistence to negative effect on intermediate spatial scales, to again positive effects on large spatial and temporal scales. Species with different size and mobility can be regulated by different processes at the same spatial scale, a principle that may contribute to diversity. Differences in species richness between local communities may be differentially regulated at larger spatial scales. This causal connection between local and regional processes has several practical conservation implications. We furthermore show that fractal geometry can be a valuable tool in the separation of the effects of habitat loss (percentage cover), habitat fragmentation (contiguity) and habitat (texture). Especially important may be the notion that the same effective degree of fragmentation can exist at in a very aggregated habitat (one big patch) and a very sparse, random landscape (many small, nearby patches). The geometric behaviour and covariance of these three basic parameters of spatial configuration needs further investigation. The fractal approach is tested using data on heathland habitat configuration and biodiversity in 36 Dutch landscapes of each 9 km 9 km. Fractal geometry was indeed successful in separating the patterns of habitat loss from habitat fragmentation in a subset of the landscapes, despite covariance across all landscapes. Habitat loss and habitat fragmentation both had a negative effect on diversity of heathland breeding birds, while only habitat loss seemed to affect butterfly species richness. We conclude that fractal geometry seems to be a promising approach for linking population and community processes to landscape spatial structure. # 2002 Elsevier Science B.V. All rights reserved.