Prediction of parasite infection dynamics in primate metapopulations based on attributes of forest fragmentation.

Although the effects of forest fragmentation on species and ecological processes have been the focus of considerable research in conservation biology, our capacity to predict how processes will be altered and which taxonomic or functional groups will be most affected by fragmentation is still poor. This problem is exacerbated by inherent temporal and spatial variability in fragment attributes. To improve our understanding of this interplay, we examined how various fragment attributes affect one potentially important ecological process, parasite infection dynamics, and considered how changes in this process affect host metapopulations. From August 1999 to July 2003 we surveyed red colobus (Piliocolobus tephrosceles) metapopulations inhabiting nine fragments (1.2 to 8.7 ha) in western Uganda to determine the prevalence and richness of strongyle and rhabditoid nematodes, a group of potentially pathogenic gastrointestinal parasites. We used noninvasive fecal flotation and sedimentation (n = 536) to detect parasite eggs, cysts, and larvae in colobus fecal samples. To obtain an index of infection risk, we determined environmental contamination with Oesophagostomum sp., a representative strongyle nematode, in canopy (n = 30) and ground vegetation plots (n = 30). Concurrently, physical (i.e., size, location, and topography) and biological (i.e., tree diversity, tree density, stump density, and colobine density) attributes were quantified for each fragment. Interfragment comparisons of nine potential factors demonstrated that an index of degradation and human presence (tree stump density) strongly influenced the prevalence of parasitic nematodes. Infection risk was also higher in the fragment with the highest stump density than in the fragment with the lowest stump density. These results demonstrate that host-parasite dynamics can be altered in complex ways by forest fragmentation and that intensity of extraction (e.g., stump density) best explains these changes.