Geographic origins and genetics of eastern and Great Lakes mallards

Eastern and Great Lakes populations of mallards ( Anas platyrhynchos ) have experienced a significant decline in recent years. These subpopulations are increasingly wild × game‐farm mallard hybrids because of widespread releases of game‐farm individuals. Concurrent with an increasing prevalence of releases, a near 50% decline in mallard populations in the United States occurred, while abundances remained stable in Canada. We aimed to refine our understanding of the metapopulation dynamics of eastern North American and Great Lakes mallards to provide information useful in population and harvest management. We used stable isotope and genetic techniques during pre‐hunting season (July–September) banding to determine if banding location was representative of hatch or molt origin of mallards and if wild mallards captured and banded had more northern origins than wild × game‐farm hybrids. Mallards are expected to be largely of local origin during the pre‐hunting season, but nearly 50% of our sample had an origin north of their banding site, suggesting substantial movements during the banding period. We detected a similar percentage of wild × game‐farm hybrid prevalence for the eastern mallard population (~89%), but a substantial increase in the Great Lakes region (~75%) compared to prior studies. However, we did not detect strong evidence for geographic or temporal variation in isotopic values (i.e., origins) of wild and hybrid mallards, which suggests that genotypes of mallards occurred together throughout the sampling period. Our results suggest that banding location of mallards in eastern North America does not equate to breeding ground origin or genotype (wild or hybrid), and we recommend investigation of other methods to understand if vital metrics differ among regions and genotypes. The movement we inferred during the banding season could potentially violate important assumptions that birds do not move among banding units and confound population vital rates estimated using banding returns. Thus, we recommend that current integrated population models consider eastern mallards as a single population because their movement throughout the banding period makes assessment at smaller geographic units invalid.