Patterns and Predictors of Small Mammal Phylogenetic and Functional Diversity in Contrasting Elevational Gradients in Kenya

Mountains of the Afrotropics are global biodiversity hotspots and centers of speciation and endemism; however, very few studies have focused on the phylogenetic and functional dimensions of Afromontane small mammals. We investigated the patterns and mechanisms of small mammal phylogenetic and functional diversity and assembly along elevational gradients in Mount Kenya, the second highest mountain in Africa, and a contrasting low mountain range, Chyulu Hills. We sampled 24 200-m interval transects in both sites; 18 in Mt. Kenya (9 each in the windward side, Chogoria, and the leeward side, Sirimon) and 6 in Chyulu. We extracted the mitochondrial Cytochrome b gene to reconstruct a time-calibrated species tree for estimating phylogenetic diversity indices [phylogenetic richness (PD), mean nearest taxon distance (PDMNTD), and nearest taxon index (PDNTI)]. A functional trait data set was compiled from the field-recorded measurements and published data sets for estimating functional diversity indices [functional richness (FD), mean nearest taxon distance (FDMNTD), and nearest taxon index (FDNTI)]. Several environmental variables representing water-energy availability, primary habitat productivity, and topographic heterogeneity were used to estimate the predictive power of abiotic conditions on diversity variances using generalized linear and generalized additive regression models. The PD and FD peaked around mid-elevations in Mt. Kenya, unimodally increased or decreased in Chogoria and Sirimon, and monotonically increased in Chyulu. The divergence and community structure indices-PDMNTD, FDMNTD, and PDNTI and FDNTI-were relatively weakly associated with elevation. Overall, the tendency of assemblages to be phylogenetically and functionally closely related than expected by chance decreased with elevation in Mt. Kenya but increased in Chyulu. Across the indices, the annual precipitation and topographic ruggedness were the strongest predictors in Mt. Kenya, evapotranspiration and temperature seasonality were the strongest predictors in Chyulu, while temperature seasonality and terrain ruggedness overlapped as the strongest predictors in Chogoria and Sirimon in addition to annual precipitation in the latter and normalized difference vegetation index in the former. The observed contrasting trends in diversity distribution and the strongest predictors between elevational gradients are integral to the sustainable management of the high faunal biodiversity in tropical Afromontane ecosystems.