Patterns and Mechanisms of Heterogeneous Breeding Distribution Shifts of North American Migratory Birds

There is widespread evidence that species distributions are shifting in response to climate change. Warming temperatures and climate niche constraints are hypothesized drivers of northward shifts in temperate migratory bird breeding distributions, but heterogeneity in the direction of distribution shifts suggests that the climate niche hypothesis does not explain all changes in distributions. We propose that: 1) changes in migration costs and benefits related to dampened seasonal differences between breeding and winter areas, 2) sensitivity to supplemental cues that affect duration of migration and onset of reproduction, 3) a latitudinal mismatch-driven fitness gradient, or a combination of these drivers may explain southward distribution shifts. We examined latitudinal shifts in breeding distribution centroids for 73 species of migratory birds from 1994 to 2017 across eastern, central and western regions of North America using Breeding Bird Survey data and tested if life history characteristics related to the above hypotheses and population status were associated with shift patterns. We found that 44% of regional centroid shifts were southward, 55% were northward, and several species shifted in different directions in different regions. Migratory strategy and protandry predicted breeding distribution centroid shifts, although they tended to be more predictive of northward shifts than southward shifts. There was evidence that supplemental cues explained some southward shifts because herbivorous birds tended to shift southward compared to insectivores, or raptors that shifted northward. Shifts in centroids were not explained by trends in abundance, suggesting that centroid shifts were not attributable to population declines or increases at distribution margins. Our results show the prevalence of heterogeneous breeding distribution shifts, including often overlooked southward shifts, and suggest that more work is needed to develop alternative hypotheses that would explain southward shifts in distributions.