Making biodiversity discovery more efficient: An exploratory test using Mexican birds

TLDR: A test is developed in which well-known and broadly distributed species are used to predict the distributional areas of poorly-known congener species, indicating that all of the ‘unknown’ species chosen for testing were predicted significantly better than random expectations.

Abstract: Recent studies have explored the possible utility of ecological niche modeling as a tool for species discovery, with promising initial results, but no detailed test has as-yet evaluated the success or failure rates of the method. We use a comprehensive distributional summary for the birds of Mexico to develop a test, in which well-known and broadlydistributed species are used to predict the distributional areas of poorly-known congener species. The results indicated that all of the ‘unknown’ species chosen for testing were predicted significantly better than random expectations by the ecological characteristics of at least one well-known congener species. These results bode well for broader applications of using biodiversity informatics tools to guide species discovery, but further testing and development of protocols is still needed.