Abiotic Factors and Competitive Exclusion Drive Assembly Patterns in Two Insular Gecko Adaptive Radiations Displaying Ecomorphological Convergence
Skipwith, P. L.; Castillo-Rodriguez, N.; Zenil-Ferguson, R.
Show abstract
Adaptive radiation theory posits that speciation in such lineages is largely driven by ecological opportunity in concurrent morphological expansion in response to niche availability. Here, we use a phylogenomic estimate of Australasian diplodactyloid geckos in combination with meristic and ecological data to infer patterns of ecological diversification, quantify signatures of stabilizing selection, and the factors driving speciation processes. Specifically, we focus on two relatively young but speciose and ecomorphologically diverse assemblages from the ancient islands of New Caledonia and New Zealand. Models accounting for stabilizing selection recover shifts in morphospace along many branches that also experienced shifts in ecological guild as inferred from ancestral state reconstructions. We find convergent evolution to be present between the two insular lineages as they independently transitioned to similar guilds from different ancestral ecologies. Community assembly is integral to understanding the dynamics of adaptive radiations and various studies focused on identifying if biotic or abiotic factors drive character suits and sympatry in diverse lineages. Bayesian and multiple regression analyses suggest that abiotic factors rather than interspecific competition dictates phenotypic divergence in both insular lineages. Rather, species seem to diverge phenotypically in allopatry and environmental factors, such as climate, in combination with competitive exclusion drive phenotypic overlap in sympatry. This study provides the first modern assessment of convergence for diplodactyloid geckos and provides robust evidence indicating that similar selective pressures have shaped morphological diversity in these disparate as well the factors affecting sympatry.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.