UNBOUNDED DIVERSITY DYNAMICS IN CENOZOIC CARNIVOROUS MAMMALS

Classical niche theory predicts that species richness is limited by the availability of ecological niche space within local assemblages and that interspecific competition should cause species with identical ecological niches to occupy separate geographical spaces. Theoretically, γ diversity is therefore not only limited by local niche packing but also the geographic area available for the dispersal of controphic species. If γ diversity is bounded, the expectation is that, over time, rates of diversification slow, rates of extinction accelerate, rates of range expansion slow, and/or niches become saturated as the upper bound is reached. Because rates of diversification, extinction, and dispersal can vary for multiple ecological reasons (e.g., neutral biodiversity dynamics), we propose testing hypotheses of bounded diversity dynamics using morphological and phylogenetic metrics for the ecological niche overlap. As a test for an upper bound on γ diversity, we calculated phylogenetic and functional diversity, geographic range size (95% minimum convex polygons), and diversification rates (mark recapture methods) for carnivorous mammals during the Cenozoic (66 Ma – present). We show that niche occupation, diversification rates, and geographic range size wax and wane in a way that is inconsistent with bounded diversity dynamics. For comparison we programmed a null model using a phylogenetically explicit birth-death simulation in which secular decline in diversification rates is unrelated to γ diversity, range size, range overlap, or niche packing. Simulation results produce clades whose ranges gradually increase with time due to Brownian motion, even as clade diversity is declining. Our findings suggest that, on the timescales of the terrestrial fossil record, unsaturated communities may be a common phenomenon because they exist in a balance amongst the processes of invasion, dispersal, extirpation, and disturbance, which cumulatively prevent them from reaching any theoretical upper bound on regional diversity.