Aim: Mountains cover approximately 22% of the planet's terrestrial surface and have dramatic effects on climate and biodiversity. The rain shadow effect is a common feature on mountain ranges worldwide and its effects on ecology and evolution of species, particularly morphology, are incompletely known. Our aim is to identify the correlates that best describe morphological variation along a pronounced rain shadow gradient.Location: Patagonia and Tierra del Fuego (southern South America), including Valdivian and Magellanic forests and steppes.Taxa: Shaggy soft-haired mouse Abrothrix hirta (Order Rodentia, Family Cricetidae).
Methods:We measured 450 skulls of A. hirta from 67 localities between 35°S and 54°S, from the Pacific to the Atlantic coasts. We analysed geographic differences in skull morphology using 17 linear skull measurements. Discriminant function analysis revealed clear evidence for sexual dimorphism; thus, analyses were conducted for the entire dataset, and separately by sex. We implemented regression tree analysis to test the environmental correlates that best describe morphological variation along this gradient.Results: Skull size variation in shaggy soft-haired mice does not follow Bergmann's rule; however, latitude was the second node for PC1 of all samples. Regression tree analyses showed that the variables that best explained size for the various datasets were longitude, precipitation of the coldest quarter and temperature seasonality.Longitude appeared in eight of the nine regression tree analyses. We find that organisms attain greater sizes in the western Andes where there is no rain shadow effect, below 500 m of elevation.Main Conclusions: Morphological variation and size of A. hirta is highly structured along a longitudinal gradient, which mirrors patterns of primary productivity and the Andean rain shadow effect in this region, best described by the "resource rule".Mountains are a formidable land feature that clearly affects the morphology of species distributed at both sides of them.