Modeling the Geographical Distribution and Fundamental Niches of Cacajao spp. and Chiropotes israelita in Northwestern Amazonia via a Maximum Entropy Algorithm

Abstract

We modeled the geographical distribution of 4 pithecine primate species: brown-backed bearded sakis (Chiropotes israelita) and 3 black uakaris (Cacajao melanocephalus, C. hosomi, and C. ayresi) that inhabit remote regions of western Amazonas, Brazil. We applied a maximum entropy algorithm modeling program (MAXENT) to field data Boubli collected from 1991 to 2007. We used 23 environmental coverage variables to model the distribution of the primates. The layers were related to precipitation, temperature, topography, and ecological bioregions or Ecoregions. The predicted distribution for Cacajao hosomi was strongly associated with the Negro-Branco Moist Forest and Guianan Highlands Moist Forests Ecoregions, and the Worldclim variables Bio3 (isothermality), Bio4 (temperature seasonality) and Bio17 (precipitation of the driest quarter). Cacajao melanocephalus was strongly associated with Japura/ Solimoes-Negro Moist Forests, Caqueta Moist Forests, Purus Varzea Flooded Forests, Rio Negro Campinaranas, and Cordillera Oriental Montane Forests, Ecoregions. Cacajao ayresi was strongly associated with Negro-Branco Moist Forest and Rio Negro Campinarana Ecoregions as well as Worldclim Bio3 (isothermality). Chiropotes israelita was also strongly associated with Worldclim Bio3 (isothermality) followed by the Negro Branco Moist Forests and Guianan Piedmont and Lowland Moist Forests Ecoregion, and to the Guianan Highland moist forests. These results show a great overlap between the bearded saki and 2 black uakaris, Cacajao hosomi and C. ayresi. Given that one cannot attribute the separation between the species in the Rio Negro-Rio Branco interfluvium to the existence of geographical barriers such as rivers, we suggest that the present geographical boundaries and thus coexistence of the 3 pithecines north of the Rio Negro is maintained by competitive exclusion or stochastic events. Until more surveys are conducted, the present geographical distributions of the pithecines and the mechanism maintaining their boundaries in the Rio Negro-Rio Branco interfluvium will remain uncertain. One important contribution of our model is to identify areas of higher probability of occurrence that might be helpful in guiding future survey expeditions and choices of areas for future conservation of pithecines.