Distinguishing species boundaries from geographic variation

Abstract

In an era of unprecedented biodiversity loss, the need for standardized practices to describe biological variation is becoming increasingly important. As with all scientific endeavors, species delimitation needs to be explicit, testable, and refutable. A fundamental task in species delimitation is distinguishing within-species variation from among-species variation. Many species that are distributed across large geographic areas exhibit levels of genetic variation that are as great or greater than those that exist between well-defined sympatric species. Here, we provide a workflow to distinguish between intra- and interspecific genetic variation and apply the workflow to a taxonomically problematic group of frogs (the Rana pipiens complex, or leopard frogs) that are widely distributed across Mexico and Central America. Our workflow makes use of recent advancements that pair genome-scale datasets with model-based species delimitation methods, while emphasizing the need for positive evidence of reproductive isolation to confirm the validity of geographically contiguous species boundaries. We find that intraspecific geographic variation in widespread leopard frog species has resulted in considerable taxonomic inflation of species. Ten currently recognized species are not supported in our analyses, and we here synonymize them with previously named taxa. Furthermore, we find positive evidence for the presence of three undescribed species. In addition to proposing these taxonomic changes, we provide descriptions of the data or analyses that would be needed to refute and overturn our recommendations. We recommend that all species delimitation studies (especially of geographically variable groups) clarify what new evidence would be sufficient to change the taxonomic recommendations.