Iterative evolution of large-bodied hypercarnivory in canids benefits species but not clades

Ecological specialization has costs and benefits at various scales: traits benefitting an individual may disadvantage its population, species or clade. In particular, large body size and hypercarnivory (diet over 70% meat) have evolved repeatedly in mammals; yet large hypercarnivores are thought to be trapped in a macroevolutionary "ratchet", marching unilaterally toward decline. Here, we weigh the impact of this specialization on extinction risk using the rich fossil record of North American canids (dogs). In two of three canid subfamilies over the past 40 million years, diversification of large-bodied hypercarnivores appears constrained at the clade level, biasing specialized lineages to extinction. However, despite shorter species durations, extinction rates of large hypercarnivores have been mostly similar to those of all other canids. Extinction was size- and carnivory-selective only at the end of the Pleistocene epoch 11,000 years ago, suggesting that large hypercarnivores were not disadvantaged at the species level before anthropogenic influence. Balisi and Van Valkenburgh analyse the effect of body size and dietary specialization on extinction regimes in North American Canidae. They find that hypercarnivory, which evolved independently multiple times, does not increase species-level extinction but is associated with extinctions of clades.