Estimating hybridization in the wild using citizen science data: A path forward

Genomic evidence of introgression in natural populations has reinvigorated the study of hybridization in recent years. Still, it is largely unknown how frequently individual organisms mate across species lines. Recently, Justyn et al. suggested that eBird, one of the world's largest citizen science databases, may supply adequate data for estimating hybridization rates. Here, we compare Justyn et al.'s estimates-and their conclusions that hybridization is rare-with estimates from museum and molecular data. We also estimate hybridization using eBird observations from areas and times when hybridization is possible, namely, in contact zones during the breeding season. These estimates are all considerably higher than those reported in Justyn et al., emphasizing that inferences from multiple datasets can differ radically. Finally, we demonstrate an approach for predicting the location of hybrid zones using eBird data, which can be done with high confidence and with unprecedented resolution. We show that citizen science data, far from settling the question of how frequently bird species hybridize, instead offer a promising step toward more focused study of hybrid zones.