Effects of Land-Cover Changes on Mammals in a Neotropical Region: a Modeling Approach.

Land-cover changes translate into shifts in habitat available to wildlife species. I analyzed the effects of land-cover changes on habitat availability for 54 mammal species in a 2.7 million-ha area in southern Mexico and northern Guatemala. I considered the regional variability of these changes and the effect of variation in management and development trends. Using cluster analysis, I grouped mammal species into assemblages with similar associations of land-cover type. Based on data from a remote-sensing, land-cover change analysis (1974-1986), I created simple linear Markov models for a heuristic tool to simulate land-cover changes over a 60-year period and to explore temporal trends of change in habitat availability for the mammal species. I used elasticity analyses to identify land-cover transition probabilities critical for these trajectories. Of the 12 land-cover classes considered, four dominated the area: grasslands, tropical secondary vegetation, tropical moist forests, and wetlands. Transition probabilities to and from these four land-cover types were key in determining the availability of habitat for mammals. Thus, the relatedness of mammal species to these land-cover types seems critical for their long-term persistence in the region. According to general simulated trends in fluctuation of habitat availability, I classified species into seven categories: (1) opportunistic and highly adaptable species, occupying most of the area; (2) opportunistic, human-commensal species, with restricted range in the study area; (3) species with stable and moderate habitat availability; (4) species with moderate but increasing habitat availability; (5) species with originally moderate but declining habitat availability; (6) species with already reduced and sharply declining habitat availability; and ( 7) species with declining habitat availability and a restricted range. Markovian trajectories suggest a declining trend of habitat availability for 32 species (59%). Land-cover change trends, however, were highly variable between the nine subregions. As a result, habitat availability for many species varied regionally. Likewise, changes in management and development policies and trends in the study area will lead to contrasting habitat availability for declining and increasing species, but not for the rest of the species. The approach I used is useful for (1) assessment of land-cover changes resulting from different development trends and management practices and (2) exploration of how changes may affect species habitat availability and survival perspectives. This examination can be accomplished for a substantial part of a biota and for entire regions, even in the context of limited information.