Physical and biological variables related to habitat preferences of rodents

This study determined habitat characteristics related to the presence of species of rodents and possible associations between pairs of these species that may affect their distribution. The study was conducted in the southern half of the Baja California Peninsula, Mexico. We used an environmentally constrained null model with 200 data sites that were sampled from 1999 to 2004 and measurements of different environmental factors obtained from electronic maps, meteorological stations, and measurements in the field. Although most of the isolated factors were not adequate for distinguishing between sites (rodents’ habitat), the combination of all the factors gave an acceptable explanation for the presence of rodent species. The low selectivity of the species for the variables used in the model could be related to the selection of variables that were not appropriate for all species and because an artifact of scale of several variables was employed in the analysis. The percentage of rock cover and vegetation types was most closely related to habitat preference for most species. The possible associations between pairs of these species shown by the null model, the negative interaction between two pairs of species (Peromyscus eva vs. P. fraterculus and Chaetodipus spinatus vs. C. fallax) have probably affected the distribution of these rodents.