Open Collections

Abstract

Communal roosting is a form of social behaviour exhibited by hundreds of species of birds where unrelated individuals spend the resting time together. Different hypotheses have been proposed to explain the origins of this behaviour, including predator avoidance, thermal regulation, and increased foraging efficiency through the Information Center Hypothesis (ICH) which suggests information transfer at roost sites. To test for potential evolutionary drivers of communal roosting, I first compiled ecological data from all landbirds as this behaviour has been more consistently studied on landbirds. I collected communal roosting information from online databases and published literature, and classified communal roosting behaviour as present (1) or absent (0). I then used body mass (kg), trophic guild, Hand-wing Index (HWI) and brain to body size ratio as potential predictor variables and evolutionary relatedness using phylogenetic trees from VertLife, a website of resolved phylogenies for all extant species. I fit generalized linear models using biological variables as predictors and communal roosting behaviour as a response after accounting for phylogeny. I then performed model comparison between the null model with phylogeny only and the full model to test whether CR can be explained with the aforementioned variables. My results show that this behaviour is positively correlated to body mass and HWI independent of phylogeny, and is more persistent in scavengers than any other trophic guild. My findings expand on Beauchamp’s study, and support that communal roosting cannot be explained through phylogeny only. I use these results to highlight the importance of communal roosting sites for birds, with particular emphasis on obligate scavengers, and discuss the implications of foraging efficiency from the lens of the Information Center Hypothesis.