Habitat suitability and centrality—not peripherality—predict demographic performance in a Neotropical lizard

Abstract

AbstractAimThe centre–periphery hypothesis (CPH) sustains that peripheral populations are subjected to harsher environmental conditions, which reduces demographic performance manifested, among others, by lower abundance, higher inbreeding and higher developmental instability. However, range centrality/peripherality is not necessarily a good proxy for environmental suitability since geographical distributions are often irregular in shape, barriers may limit dispersal towards suitable areas, and ancestral range shifts may have occurred unevenly in each direction. We evaluated predictions of the CPH for populations of the Neotropical lizard Notomabuya frenata by modelling the effects of environmental suitability and geographical centrality/peripherality upon demographic performance.LocationBrazilian Cerrado.TaxonNotomabuya frenata (Squamata, Scincidae).MethodsWe used populational values of body size, body mass, body condition, tail autotomy and fluctuating asymmetry (FA; a proxy of developmental instability) to estimate demographic performance. Then, we built correlative models to estimate environmental suitability for each sampled population. Finally, we used hierarchical Bayesian models based on a combination of stochastic partial differential equations and integrated nested Laplace approximations to model the relationships between demographic performance and predictor variables (geographical centrality/peripherality and environmental suitability) and make predictions across the geographical distribution of N. frenata.ResultsGeneralized boosting produced the best environmental suitability models. The total suitability area spanned more than 4 million km2 in cis‐Andean South America. All scalation counts presented some degree of asymmetry. Only three response variables—all related to FA—responded significantly to predictor changes. Contrary to the predictions of the CPH, one scalation count presented higher asymmetry closer to the range center and another in more suitable environments. Peripherality did not significantly influence any of the response variables.Main ConclusionsPopulations may respond differently to geographical centrality/peripherality and environmental suitability, as these characteristics do not necessarily coincide. Studies testing the CPH should ideally incorporate these different predictors of demographic performance.