Cold tolerance,and not earlier arrival on breeding grounds,explains why males winter further north in an Arctic‐breeding songbird

Sex biases in distributions of migratory birds during the non‐breeding season are widespread; however, the proximate mechanisms contributing to broad‐scale sex‐ratio variation are not well understood. We analyzed a long‐term winter‐banding dataset in combination with spring migration data from individuals tracked by using geolocators to test three hypotheses for observed variation in sex‐ratios in wintering flocks of snow buntings Plectrophenax nivalis. We quantified relevant weather conditions in winter (temperature, snowfall and snow depth) at each banding site each year and measured body size and condition (fat scores) of individual birds (n > 5500). We also directly measured spring migration distance for 17 individuals by using light‐level geolocators. If the distribution pattern of birds in winter is related to interactions between individual body size and thermoregulation, then larger bodied birds (males) should be found in colder sites (body size hypothesis). Males may also winter closer to breeding grounds to reduce migration distance for early arrival at breeding sites (arrival timing hypothesis). Finally, males may be socially dominant over females, and thus exclude females from high‐quality wintering sites (social dominance hypothesis). We found support for the body size hypothesis, in that colder and snowier weather predicted both larger body size and higher proportions of males banded. Direct tracking revealed that males did not winter significantly closer to their breeding site, despite being slightly further north on average than females from the same breeding population. We found some evidence for social dominance, in that females tended to carry more fat than males, potentially indicating lower habitat quality for females. Global climatic warming may reduce temperature constraints on females and smaller‐bodied males, resulting in broad‐scale changes in distributional patterns. Whether this has repercussions for individual fitness, and therefore population demography, is an important area of future research.