Evolutionary opportunity and the assembly of faunas and communities
Community abundance patterns suggest at least broadly comparable resources for anoles in tropical montane forests on Hispaniola and Jamaica. So why are Jamaica’s montane communities less diverse? Ecological explanations such as colonization limitation are unlikely—all Jamaican anole species not found in highland communities are otherwise distributed islandwide, and no obvious barriers prevent colonization. Instead, diversity differences are best explained by the relative paucity of high-elevation specialists in the species pool of Jamaica – indeed, computationally “deleting” the endemic highland subfaunas of each island reconciles their divergent relationships between local species richness and elevation (Fig. 4). Thus, to understand Jamaica and Hispaniola’s highland community differences, we must turn to the evolutionary buildup of their species pools (Jetz & Fine 2012; Cornell 2013; Mittelbach & Schemske 2015).
Speciation-driven contributions to highland diversity could arise either by evolutionary colonization, in which lowland-derived lineages evolve to tolerate or specialize on cool upland conditions, or by in situ radiation from a highland ancestor (Ricklefs 2006; Rosindell & Phillimore 2011). The first mechanism has occurred on both islands, but appears difficult, with only one instance on Jamaica and a just a handful in northern Hispaniola. Indeed, evolutionary transitions of this sort appear to be rare, with one study estimating that speciation across biomes is ~25 times less likely than within them (Crispet al. 2009). The second mechanism, in situ highland radiation, occurred exclusively on Hispaniola, where a clade of highland-endemic anoles diversified into a variety of ecological types (Figure 3e), raising the taxonomic, functional, and phylogenetic uniqueness of the Hispaniolan highlands. The smaller cumulative area of Jamaica’s highlands likely precluded enrichment of its montane fauna via this mechanism. Allopatric separation of populations is likely a pre-requisite for speciation in anoles (Glor et al. 2004, 2005), and anoles likely need at least 3000km2 to undergoin situ speciation (Losos & Schluter 2000). The highland “sky island” in northern Hispaniola above 700m clears this threshold with 4000km2 (Fig. 1), yet Jamaica’s equivalent zone occupies less than 400km2. Thus the rarity of cross-biome speciation events, plus insufficient space for subsequentin situ radiation, appear to have deprived highland Jamaica from the 2-3 additional species supported locally at similar elevations on Hispaniola.
In contrast to the highlands, the lowlands on both islands are large, well over the 3000km2 threshold. Even though the lowlands of northern Hispaniola are nearly 5X larger than in Jamaica (18,500 versus 3,900km2), alpha-diversity in the lowlands is identical, suggesting that local diversity caps predominate, and have been reached. Northern Hispaniola’s greater total diversity thus appears to come about not because it is larger per se , but because its high-elevation biome is large enough to generate an additional set of distinct species, boosting diversity from 6 species (lowland only) to 11 (total).
A growing body of recent work has revealed the importance of geographical and environmental determinants of evolutionary radiation in setting regional diversity patterns (Fine & Ree 2006; Jetz & Fine 2012; Quintero & Jetz 2018). Our work shows that such “evolutionary opportunities” can also structure communities by producing species capable of maintaining local diversity equilibria. Such equilibria are consistent with evidence from temporal studies of local-scale diversity, which provide evidence for stable community-level richness limits over timescales ranging from decades to tens of millions of years (Bambach 1977; Brown et al. 2001; Gotelli et al. 2017; Closeet al. 2019). However, they contrast with local diversity patterns in Hawaiian trees, one of very few other groups to be investigated for signatures of macroevolution on local community structure (Craven et al. 2019). In that system, greater island-level evolutionary diversity was linked to increased local-scale diversity, suggesting that local diversity caps in that system either have not yet been reached, or are more easily overridden. Understanding this difference will require further research, and we suggest that the likelihood that evolution will strengthen local controls will depend on potentially system-specific mechanisms of dispersal, competition, and coexistence. For example, compared to tropical trees, anoles are poorer dispersers (Williams 1969), more likely to engage in interference competition (Schoener 1977; Culbertson & Herrmann 2019), and, lacking dormancy, less likely to benefit from temporal storage effects (Chesson & Warner 1981), all of which could favor the primacy of local controls in anole communities.