Our results remain largely unchanged. The number of migrants per generation remain much higher than one between Ramphocelus melanogaster and R. carbo , as expected if these populations are merging genomically. Previously, our incorrectly calculated number of migrants had suggested much higher movement of individuals into R. melanogaster from R. carbo than in the reverse. Instead, we now find that the number of migrants moving into R. carbo fromR. melanogaster is slightly higher than in the reverse. Here we show an updated Figure 4 with the correct number of migrants per generation, and we update Table S4 to have the correct parameter values for this model. Coded model files used for demographic analyses deposited in Figshare (dx.doi.org/10.6084/m9.figshare.14981556) have also been updated.

The authors apologize for any inconvenience this may have caused.