Molecular Analysis of Seaside Sparrow (Ammospiza maritima) Diet: Methodological Insights and Ecological Applications

Throughout time, several techniques have been employed to study avian diet, each having its own strengths and weaknesses. In the past two decades, DNA-based techniques have advanced from novel “proof of concept” studies to standard methodology accepted across several fields of study. In ornithology, these DNA-based techniques have evolved from simply detecting a single prey species of interest in the diet to in-depth description of diet—with often many prey species—from hundreds of individual birds at a time. Despite the wide application of DNA metabarcoding, there remain technical uncertainties and methodological challenges that have yet to be resolved. As these methods continue to be improved, descriptions of diet will only continue to improve, especially when thoughtfully paired with other methods. I provide context for the use of DNA metabarcoding amongst other methods in the field of avian diet ecology, highlighting the unique challenges and advantages of each approach, and finally looking forward to how these methods may be employed together to provide the most accurate data, and thus the most accurate descriptions of resource use by birds to provide ecological insight. Next, I address one methodological uncertainty of metabarcoding in an effort to improve sampling methods for future avian diet studies and improve the basic knowledge of the application of this method. By comparing the descriptions of diet produced from both fecal samples and stomach contents, I show that fecal samples are the ideal sample type to use in avian diet metabarcoding. Finally, I apply DNA metabarcoding to an ecological question, using this method to evaluate how Seaside Sparrows respond to two wide scale disturbances that affected southeast Louisiana: the Deepwater Horizon oil spill (2010) and Hurricane Isaac (2012), showing that in resource poor environments, Seaside Sparrows individuals appear to become less specialized, with resource use most stable on unoiled sites