Understanding of how risk factors affect populations across their annual cycle is a major challenge for conserving migratory birds. For example, disease outbreaks may happen on the breeding grounds, the wintering grounds, or during migration, and are expected to accelerate under climate change. The ability to identify the geographic origins of individuals impacted, especially outside of breeding areas, might make it possible to predict demographic trends and inform conservation decision making. However, such an effort is made more challenging by the degraded state of carcasses and resulting low quality of DNA available. Here we describe a rapid and low-cost approach for identifying the origins of birds sampled across their annual cycle that is robust even when DNA quality is poor. We illustrate the approach in the common loon (Gavia immer), an iconic migratory aquatic bird that is under increasing threat on both its breeding and wintering areas. Using 300 samples collected from across the breeding range we develop a panel of 158 SNP loci that diverge across six genetic subpopulations. Using this SNP panel we identify the breeding grounds for 167 live nonbreeding individuals and carcasses. For example, genetic assignment of loons sampled during botulism outbreaks in parts of the Great Lakes provide evidence for the significant role the lakes play as migratory stopover areas for loons that breed across wide swaths of Canada, and highlights the vulnerability of a large segment of the breeding population to botulism outbreaks that are occurring in the Great Lakes with increasing frequency. Our results illustrate that the use of SNP panels to identify breeding origins of carcasses collected during the non-breeding season can improve our understanding of the population-specific impacts of mortality from disease and anthropogenic stressors, ultimately allowing more effective management.

Through a large collaborative effort, we collected samples from breeding, wintering and migrating common loons from across North America during the years 1979 and 1992-2016. Samples for genetic analysis included blood samples collected from birds captured in known breeding areas across North America during the breeding season, and samples collected from wintering or migratory birds, including: a) feathers from bird carcasses resulting from botulism outbreaks in the Great Lakes or collisions at solar facilities in southern California, b) blood or feathers collected from birds that were dead or moribund from mostly unknown causes from Florida and the northeastern seaboard, and c) blood from living, healthy birds from Morro Bay, CA, and coastal Louisiana. Tissue samples and previously extracted DNA were sent to UCLA for SNP genotyping and/or library preparation. Samples were transferred to UCLA under a protocol approved by the Research Safety and Animal Welfare Administration, University of California Los Angeles: ARC# 2017-073-03, approved 10/28/2017. The datasets provided include 1) a RAD-PE dataset (COLO-cleaned-150-100K.unrel...vcf) containing genotypes of 129 breeding loon samples and an accompanying file containing sample metadata, 2) a fluidigm genotype dataset(ped and map file) that includes the genotypes of 467 breeding, wintering and migrating loons at 158 loci, and an accompanying file containing sample metadata, 3) an input dataset for conducting PCA in base R based on ddRAD sequencing of 106 breeding loons (A1634.SNPs). Please see the paper Larison et al. Leveraging Genomics to Understand Threats to Migratory Birds [details TBP] for details.