Department of Biology
Species incidence and diversity in response to habitat loss and fragmentation: insights from fractal geometry
Monday, October 24, 2011
AA-G008, 5:00 PM
Urban environments often yield aggregated patterns of land use associated with positive feedbacks in human economic activity (shops tend to be near other shops) and zoning restrictions. Such aggregation, together with constraints imposed by natural landscape features such as rivers or topography, can often yield fractal distributions of urban habitats. Fractal distributions imply that the spatial pattern evident at any one extent is similar at all extents, but mobile animals or dispersal phases of plants likely have characteristic spatial scales of response and thus different encounter rates with preferred habitats. Fractal habitats can vary in their fragmentation, represented by fractal dimension H, for a given proportional cover of habitat h across a landscape. Here I present a landscape coexistence model of species incidence and biodiversity derived from differential utilization by trophically-similar species of habitat clusters that differ in size and “quality” (resource density). The joint fractal distribution yields unique or “exclusive” sets of habitat clusters of a particular size and quality for species that differ in their scale of response to the environment, which is approximated well by species body size. The model predicts incidence and diversity of species that differ in body size as a function of habitat amount, cluster size, and fragmentation. Minimum habitat cluster size for a species is predicted to scale directly with body size with an exponent of 3.2 in rare habitats to 4 in abundant habitats. Species-area relations depend on both habitat distribution and quality. Aggregation of habitat can mitigate habitat loss, especially in poorer quality or less productive habitats. In contrast, aggregation may lead to lower species richness in more productive habitats. The exponent, z, for power law species-area relationships depends as strongly on habitat quality as area, but may be relatively insensitive to habitat fragmentation except in the poorest and richest quality habitats. The exponent is predicted to lie between 0.1 and 0.15 for most habitats, considerably lower than the exponent derived from a lognormal distribution of abundances (May 1975). A review of a limited number of empirical studies of bird and insect species incidence and richness across landscapes with different proportions and aggregation of habitat generally support the qualitative predictions of the landscape coexistence model. It therefore may provide considerable insights and hypotheses about the effects of urban landscape planning on species extinction and persistence and biodiversity.
- Ritchie, Mark. 2002. “Competition and Coexistence of Mobile Animals.” Chapter 5, Ecological Studies, Vol. 161. U. Sommer, B. Worm, Eds. [PDF]
The work presented in this seminar is documented in more detail in Ritchie’s 2009 monograph, Scale, Heterogeneity, and the Structure and Diversity of Ecological Communities. This book has been placed on 2-hour reserve in the Newcomb Reading Room for the benefit of anyone interested in reading more. Chapter 7, “Biological Conservation in Fractal Landscapes”, is available in PDF form here.