Search

Island rule describes the graded trend of gigantism in small‐bodied species to dwarfism in large‐bodied species inhabiting islands, but causal explanations remain unresolved. We used geometric morphometrics to quantify cranial morphology of 544 meadow vole Microtus pennsylvanicus samples across 11 island and 3 mainland populations from the Outer Lands of New England (Atlantic) and the Alexander Archipelago of Alaska (Pacific). We compared the thermoregulation and endurance (TRE) and ecological release (ER) hypotheses using all‐subsets linear models employing residual randomization permutation procedures (rrpp), and Akaike information criterion (AIC) for model selection. We decoupled direct and indirect effects of island variables on size using path analysis. We evaluated shape with principal components analysis (PCA) and Procrustes ANOVA on Procrustes shape coordinates, then assessed the impact of static allometry and TRE and ER variables on shape. Six Atlantic island populations exhibit significant signals of gigantism with the largest voles occurring on the smallest islands lacking predators. ER explains 63% of cranial size differences. Island area has a significant total effect on size by influencing the number of mammalian predators, resulting in a 0.011 increase in unit centroid size for a 100 km2 decrease in island area. This corresponds to a predicted 0.9% change in size for every 100 km2. Given static allometry, cranial shape does not respond to insularity independent of size. These results suggest that island rule is a latent evolutionary process whose manifestation depends on nuanced biogeographic and ecological contexts that have important conservation and taxonomic implications. [ABSTRACT FROM AUTHOR]