Mammal functional diversity and habitat heterogeneity: Implications for hominin habitat reconstruction.

Hominin habitats are frequently described as 'mosaic' based on interpretations of fossil assemblages comprising taxa with divergent functional adaptations (e.g., both grazers and browsers). This interpretation rests on an assumption that mammal functional diversity is positively associated with habitat heterogeneity. We test this assumption using modern mammal data for 141 sites in Africa. Species average body mass and locomotor and dietary information was compiled for all species >500 g. The functional diversity of each species assemblage was measured using five metrics: locomotor richness, trophic richness, functional richness, functional divergence, and functional evenness. We used a high-resolution woody cover estimate for sub-Saharan Africa to compute the coefficient of variation of percentage of woody cover for each site. We used a published land cover classification to compute the number of habitat patches and the number of distinct habitat types at each site. Multiple regressions were conducted at 9 different spatial resolutions (pixel size ranging from 100 m to 30 km) to explore the relationship between functional diversity and habitat heterogeneity metrics. The overall number of species found at a site is strongly positively associated with functional richness, locomotor richness, and trophic richness at all spatial resolutions. The number of habitat types at a site and the density of habitat patches show a modest positive relationship with most functional diversity metrics at most spatial resolutions. The coefficient of variation in woody cover is a very poor predictor of functional diversity. The locomotor and trophic richness of mammal communities are positively associated with habitat heterogeneity and may be useful for reconstructing aspects of heterogeneity at hominin sites. However, the overall number of species is an important confounding variable, which in turn depends on the sample size of the overall fossil sample. Researchers should carefully consider impacts of sample size on faunal reconstructions of habitat heterogeneity.