Relative brain size explains migratory/resident tendency in birds: Partial altitudinal migration in Asian house martins.

Migration is widespread among animals but varies in its manifestation with differences in direction, distance and obligatory nature. Understanding the evolution of migration requires insight into not only the development of this behaviour but also the loss of it. Partial migration, where some individuals within a population migrate while others stay, provides a unique opportunity to identify the proximate factors determining migratory/resident behaviours. In this study, we tested four hypotheses-the body size, arrival time, dominance and behavioural flexibility hypotheses-regarding phenotypic contributions to the loss or gain of migration in the Taiwan population of Asian house martins Delichon dasypus. This population exhibits partial altitudinal migration, with some martins remaining at mountain breeding grounds year-round and some migrating to lower elevations during winter. Our results most supported the behavioural flexibility hypothesis, which predicts that resident individuals tend to have larger brains than migratory ones, potentially associated with higher levels of foraging innovation. We argue that surviving in the harsh winter conditions in mountainous areas requires large brains, an energetically expensive trait that may further inhibit migration in resident Asian house martins. This creates a potential positive feedback loop where the demands of residency select for increased brain size while larger brains simultaneously facilitate residency. We also found that residents tended to have relatively smaller beaks, which likely help reduce heat loss in mountainous regions during winter. Our findings suggest that partial migration in Taiwan's Asian house martins resulted from the emergence of residency in large-brained individuals in a previously migratory population. We extend the behavioural flexibility hypothesis, traditionally applied to interspecific comparison, to demonstrate its explanatory power for intraspecific variations. Additionally, we integrate this hypothesis with selection imposed by high-elevation hypoxia to elucidate the evolutionary link between brain size and partial altitudinal migration.