• Birds;
  • eigenvector analysis;
  • evolution of migration;
  • migratory species;
  • niche conservatism;
  • oscine passerines;
  • phylogenetic eigenvector regression;
  • range size variation;
  • Rapoport's rule



To quantify the contributions of environment, phylogeny and geography to variation in the breeding and non-breeding geographical range sizes of oscine passerines.


Western Hemisphere.


Breeding range sizes were estimated for 420 species, and non-breeding ranges were estimated for 122 migratory species. Phylogenetic, environmental and geographical (spatial) eigenvectors were used to partition cross-species variation in range size. The strengths of environmental and phylogenetic signals were quantified and compared among all species, and between migratory and sedentary oscines.


Phylogenetic, environmental and geographical structure explained most of the variation in range size, accounting for 95% of the variation in breeding range sizes of migratory birds. The three components overlapped extensively, with most variation explained by differences in environmental niches. Models for breeding ranges of migratory species contained the strongest phylogenetic, environmental and geographical signals at the species level. In contrast, models for non-breeding ranges of migratory species contained the weakest phylogenetic and environmental signals (5.7% and 65.2% of variance explained, respectively). The phylogenetic signal was consistently stronger for migratory breeding ranges than for the other groups.

Main conclusions

Oscine range sizes contain a low to moderate phylogenetic signal that overlaps with environmental and geographical associations. The significance of phylogenetic signal suggests that the evolution of range size is not entirely labile, which is probably a result of the non-labile evolution of associated traits. Environmental, geographical and phylogenetic variables can account for most of the variance in species-level range size, with qualitatively similar patterns for migratory and sedentary species. Nonetheless, the stronger environmental and phylogenetic signals in the breeding ranges of migratory species may reflect both that migration is a phylogenetically conserved trait and that the subset of species able to breed in ‘recently’ deglaciated regions is more severely constrained by macroclimatic filtering.