Disentangling evolutionary, geometric and ecological components of the elevational gradient of diversity.

Despite the high importance and risk of mountain ecosystems in global biodiversity conservation, the mechanisms giving rise to and maintaining elevational biodiversity gradients are poorly understood, limiting predictions of future responses. Species richness peaks at lowlands for many taxa, which might be a consequence of mountain shape, reducing available area in highlands. For other taxa, diversity can be highest at mid elevations, suggesting the presence of mechanisms that counteract the influence of geometry. Here, we mechanistically investigate the role of mountain geometry (smaller at the peak) interaction with ecological niche width, diversification, and altitudinal dispersal to investigate the relative roles of these processes in shaping elevational biodiversity gradients. We simulated landscapes and lineages until species richness stop increasing and showed that the disproportionately large area of lowlands provides opportunity for higher species accumulation than any other elevation, even when available niche width and per-capita diversification rate are uniform across altitudes. Regardless of the underlying Elevational Diversity Gradient, altitudinal dispersal always plays a stronger role in maintaining highland than lowland diversity, due to unequal areas involved. To empirically test these predictions resulting from our model, we fit dynamic models of diversification and altitudinal dispersal to three mountainous endemic radiations whose species richness peaks in mid and high-elevation. We find that highland diversity is explained by increased diversification rates with elevation in Fijian bees, whereas niche availability is more likely to explain high altitude diversity in frailejon bushes and earless frogs, suggesting these clades are still growing. Our model and findings provide a new framework for distinguishing drivers of diversity dynamics on mountainsides and allow to detect the presence of clade-specific mechanisms underlying the geometry-diversity relationship. Understanding of these ecological and evolutionary forces can allow increased predictability of how ongoing land use and climate changes will impact future highland biodiversity.