Preventing extinctions requires understanding macroecological patterns of vulnerability or persistence. However, correlates of risk can be nonlinear, within-species risk varies geographically, and current-day threats cannot reveal drivers of past losses. We investigated factors that regulated survival or extinction in Caribbean mammals, which have experienced the globally highest level of human-caused postglacial mammalian extinctions, and included all extinct and extant Holocene island populations of non-volant species (219 survivals or extinctions across 118 islands). Extinction selectivity shows a statistically detectable and complex body mass effect, with survival probability decreasing for both mass extremes, indicating that intermediate-sized species have been more resilient. A strong interaction between mass and age of first human arrival provides quantitative evidence of larger mammals going extinct on the earliest islands colonized, revealing an extinction filter caused by past human activities. Survival probability increases on islands with lower mean elevation (mostly small cays acting as offshore refugia) and decreases with more frequent hurricanes, highlighting the risk of extreme weather events and rising sea levels to surviving species on low-lying cays. These findings demonstrate the interplay between intrinsic biology, regional ecology and specific local threats, providing insights for understanding drivers of biodiversity loss across island systems and fragmented habitats worldwide.

Populations along steep environmental gradients are subject to differentiating selection that can result in local adaptation, despite countervailing gene flow and genetic drift. In montane systems, where species are often restricted to narrow ranges of elevation, it is unclear whether selection is strong enough to influence functional differentiation of subpopulations differing by a few hundred meters in elevation. We used targeted capture of 12,501 exons from across the genome, including 271 genes previously implicated in altitude adaptation, to test for adaptation to local elevations for two highland hummingbird species, Coeligena violifer (n=62) and Colibri coruscans (n=101). For each species, we described population genetic structure across the complex geography of the Peruvian Andes and, while accounting for this structure, we tested whether elevational allele-frequency clines in single nucleotide polymorphisms (SNPs) showed evidence for local adaptation to elevation. Although the two species exhibited contrasting population genetic structures, we found signatures of clinal genetic variation with shifts in elevation in both. The genes with SNP-elevation associations included candidate genes previously discovered for high-elevation adaptation as well as others not previously identified, with cellular functions related to hypoxia response, energy metabolism, and immune function, among others. Despite the homogenizing effects of gene flow and genetic drift, natural selection on parts of the genome evidently optimizes elevation-specific cellular function even within elevation range-restricted montane populations. Consequently, our results suggest local adaptation occurring in narrow elevation bands in tropical mountains, such as the Andes, may effectively make them “taller” biogeographic barriers.

Dietary adaptation is a major feature of phenotypic and ecological diversification, yet the genetic basis of dietary shifts is poorly understood. Among mammals, Neotropical leaf-nosed bats (family Phyllostomidae) show unmatched diversity in diet; from a putative insectivorous ancestor, phyllostomids have radiated to specialize on diverse food sources, including blood, nectar, and fruit. To assess whether dietary diversification in this group was accompanied by molecular adaptations for changing metabolic demands, we sequenced 89 transcriptomes across 58 species, and combined these with published data to compare ∼13,000 protein coding genes across 66 species. We tested for positive selection on focal lineages, including those inferred to have undergone dietary shifts. Unexpectedly, we found a broad signature of positive selection in the ancestral phyllostomid branch, spanning genes implicated in the metabolism of all major macronutrients, yet few positively selected genes at the inferred switch to plantivory. Branches corresponding to blood- and nectar-based diets showed selection in loci underpinning nitrogenous waste excretion and glycolysis, respectively. Intriguingly, patterns of selection in metabolism genes were mirrored by those in loci implicated in craniofacial remodelling, a trait previously linked to phyllostomid dietary specialisation. Finally, using simulations, we show that the widely-used branch-site model is likely to be misspecified, with the implication that it is too conservative and probably under-reports true cases of positive selection. Our findings point to a complex picture of adaptive radiation, in which the evolution of new dietary specialisations has been facilitated by early adaptations combined with the generation of new genetic variation.