During spring migration, many birds travel thousands of kilometers across the North American continent. This massive undertaking poses key behavioural and physiological challenges. A major question is how these long-distance flights are impacted by human-induced changes to the landscape. We use citizen-science data to model the migration dynamics of 63 long-distance migratory bird species in response to human population density (HPOP). HPOP is a continent-wide measure of human influence that we expect to integrate many aspects of human activity, such as sensory and chemical pollution, buildings, roadways, and other structural changes to the environment. First, we use model selection to test whether the migration dynamics of each of our focal bird species are influenced by HPOP. Next, we use the fitted GAM models to derive species-specific metrics of site use and site selection behaviour during migration. We then conduct a series of comparative analyses to understand why HPOP has different effects on different migratory bird species.
Results/Conclusions
We first establish that HPOP influences migration dynamics in all 63 species in our study. Models containing HPOP had lower AIC values than the null models, and were better at predicting out of sample data. Our results also show how nearly all focal species in our analysis experience much greater HPOP during migration than they experience during breeding. We find that site selection during migration is influenced by HPOP in most species, although the direction of this effect varies: some species show attraction towards HPOP, whereas others avoid HPOP during migration. Further comparative analyses indicate that species variation in site use and site selection during migration are only weakly associated with breeding ecology and migration timing. Overall, these results establish that across bird species, migration dynamics are susceptible to the influence of human activity, and birds experience far greater levels of urbanization during migration than they do at other times in their annual cycle. These results fill an important gap by allowing us to determine species differences in these responses, and helps broaden our understanding of the migration ecology of North American birds.
