After quality editing, 39 102 autosomal SNPs for 15 723 birds remained for analysis.
M. Daley and R. Main provided birds for cineradiographic analysis.
We chose birds for this analysis due to their high trophic position, relatively high visibility and detectability during the breeding season, and high mobility, which we assume allow them to track environmental change rapidly [25], [26].
We chose to use birds for this analysis because they were the only taxonomic group for which a standardized, comprehensive database of conservation scores had been assembled that would allow both global risk and global responsibility to be evaluated for all species in all U.S. states.
The data for nine and ten day old birds were combined for analysis unless there was an effect of age on the growth parameter of interest (determined via regression analysis).
The sequences of 855 birds, including the 76 samples sequenced in the present study were aligned using ClustalX program [ 29], manually edited using GeneDoc [ 30] and the region conserved in all the birds was used for analysis.
In order to avoid pseudo-replication, we only used the first recorded night of each bird for this analysis.
In each sampling, choanal swabs (e-Swab, Copan Diagnostics, Canada) were collected from 5 birds for RT-PCR analysis (swabs were stored at −76°C until further analysis) and blood samples were collected from 15 birds.
Of these, 40 birds were randomly selected for analysis, 20 from each group.
We selected 10 individuals (20 chromosomes) with the highest sequence coverage as averaged over all 24 loci from each of the four populations from each species, respectively (40 birds in total per species), for analysis.
