Habitat identity influences species-area relationships in heterogeneous habitats

Loss of habitat is commonly identified as a reduction in area (or patch size) and its effects are often investigated using species-area relationships to evaluate loss of biodiversity. In many habitats, however, area alone is not sufficient to explain the decline of populations because natural habitats are rarely homogeneous. In fact, reductions in the diversity of habitats (e. g. destruction of microhabitats) might have even greater effects on the diversity of assemblages. Although these are 2 well-known attributes of habitats (i.e. area and diversity), there have been few empirical studies that have attempted to separate their effects by experimental manipulation. Experiments were carried out to separate the effects of number, relative proportion and identity of patches of habitat using additive designs. Predictions were tested using assemblages of marine molluscs colonizing artificial mimics of macroalgal habitats on rocky intertidal platforms. Our experiments to disentangle different attributes of habitat heterogeneity showed that these are complex and highly interactive. In particular, not all heterogeneous habitats had the numbers of species expected from a simple species-area relationship. Instead, the presence of more species in heterogeneous habitats could, to a great extent, be explained by the presence or absence of particular types of sub-habitats and not just by patch size. The complexity we found in responses of assemblages to composition of heterogeneous habitats was not predicted and could not have been observed without appropriate manipulations of habitat composition. A mechanistic understanding of the associations between species and particular types of habitats is essential for better predictions of species' responses to loss of natural habitats.