Microhabitat Segregation of Three Species of Pocket Mice (Genus Chaetodipus) in Coastal Baja California, Mexico

—I compared vegetation and substrate characteristics at capture locations of three syntopic species of pocket mice (Chaetodipus: Heteromyidae) to determine differences in microhabitat affinities of these ecologically similar rodents in coastal Baja California, México. Principal components analysis revealed that capture locations of C. spinatus had disproportionately higher cover of medium and large rocks, and tended to have higher plant species diversity than capture locations of the other two species. C. spinatus and C. arenarius, the two smallest species, differed most in their microhabitat affinities, suggesting spatial segregation that may minimize competition for similar food resources. C. arenarius was captured most often at sites with the finest-textured soils, whereas C. rudinoris, the largest species, was widespread and frequently associated with gravel and small rocks. No species showed any apparent preference for plant species or significant amounts of overhead cover, a conclusion supported by the high densities of C. rudinoris on small, barren islands in the Gulf nearby. My results represent the first quantitative descriptions of the habitat associations of these species, which are wholly or predominantly restricted in their distributions to Baja California. Studies of desert rodents have provided important insights into the mechanisms that permit species coexistence (Kotler and Brown 1988; Brown and Harney 1993). Desert rodent communities are often incredibly diverse, containing as many as 14 species, and in the case of North American deserts, are dominated by granivorous heteromyids (Brown and Harney 1993). The co-occurrence of multiple, ecologically similar heteromyid species has been attributed to subtle differences in diet (e.g., Brown and Leiberman 1973; M’Closkey 1978), foraging efficiency (Price and Heinz 1984; Price and Waser 1985; Brown 1989), predator escape abilities (Kotler 1984; Longland and Price 1991) and body size (Bowers and Brown 1982), all of which reflect the long-term consequences of competition, predation risk and biogeographical history (Brown and Harney 1993). These differences usually are manifest in patterns of microhabitat use (e.g., Price 1978; Wondolleck 1978; Thompson 1982), with species separated locally in space by the amount and species composition of plant cover and soil and substrate characteristics (Reichman and Price 1993). As part of a study of the ecology of coastal and insular populations of rodents in northern Baja California, I captured three species of pocket mice (Chaetodipus arenarius, C. rudinoris and C. spinatus) in very close proximity to one another, at locations with seemingly similar vegetation and substrate features. I recorded microhabitat characteristics at capture locations of each species to look for evi1 Stapp: Microhabitat Segregation of Three Species of Pocket Mice (Genus Chaetodipus) Published by OxyScholar, 2005 38 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES dence of microhabitat partitioning that might explain the co-occurrence of these otherwise ecologically similar species, which comprise approximately 75% of captures of rodents in the study area (Stapp and Polis 2003). These species have their distributions either wholly (C. arenarius) or predominantly (C. spinatus, C. rudinoris) in Baja California (Lackey 1991a, b; Paulson 1988), yet little is known of their ecology outside of the general habitat descriptions provided in taxonomic references (e.g., Burt 1932; Huey 1964). Riddle et al. (2000) recently proposed that populations of C. baileyi west of the Colorado River are a distinct species, C. rudinoris. To my knowledge, this study provides the first quantitative descriptions of the habitat affinities of these species.