No evidence for a link between forest herbicides and offspring sex ratio in a migratory songbird using high-throughput molecular sexing.

Many species that use or require early-successional forest are of conservation concern, including a number of songbirds that have experienced long-term population declines. In this study, our initial goal was to test whether herbicide application intensity was linked to offspring sex ratio in the White-crowned Sparrow (), a species that requires early-successional forest within forested landscapes. However, a rapid and accurate method using direct PCR to sex a large sample of birds ( > 1000 individuals) was unavailable, so our secondary goal was to develop a new approach for rapidly determine offspring sex. We obtained blood samples from sparrow young during the 2013-2014 breeding seasons in regenerating conifer plantations that were treated with one of four treatments (i.e. light, moderate, and intensive herbicide application, or no-spray control). We then optimized a protocol that used a commercially available, direct PCR kit to amplify sex-specific fragments of the CHD (chromo-helicase-DNA-binding) genes directly from whole blood stored in lysis buffer. Using this approach, we found no evidence that offspring sex ratio was linked to herbicide application intensity or to food availability across herbicide treatments. Our molecular sexing technique was 100% accurate when validated on known-sex adults, and 99.9% of our blood samples amplified successfully after being stored in lysis buffer stored for up to 3 years. The application of direct PCR for sexing birds eliminated the need for DNA extraction and substantially reduced sample processing time, cost, and the opportunity for errors during the extraction step. We conclude that forest herbicide application intensity does not influence sparrow offspring sex ratio in our study system, and that our approach provides a rapid, accurate, and tractable method for sexing birds that can facilitate studies that require processing of a large number of samples.