Genetic assignment with isotopes and habitat suitability (gaiah), a migratory bird case study

Abstract

Summary Identifying migratory connections across the annual cycle is important for studies of migrant ecology, evolution and conservation. While recent studies have demonstrated the utility of high‐resolution SNP‐based genetic markers for identifying population‐specific migratory patterns, the accuracy of this approach relative to other intrinsic tagging techniques has not yet been assessed. Here, using a straightforward application of Bayes’ Rule, we develop a method for combining inferences from high‐resolution genetic markers, stable isotopes and habitat suitability models, to spatially infer the breeding origin of migrants captured anywhere along their migratory pathway. Using leave‐one‐out cross validation, we compare the accuracy of this combined approach with the accuracy attained using each source of data independently. Our results indicate that when each method is considered in isolation, the accuracy of genetic assignments far exceeded that of assignments based on stable isotopes or habitat suitability models. However, our joint assignment method consistently resulted in small, but informative increases in accuracy and did help to correct misassignments based on genetic data alone. We demonstrate the utility of the combined method by identifying previously undetectable patterns in the timing of migration in a North American migratory songbird, the Wilson's warbler. Overall, our results support the idea that while genetic data provides the most accurate method for tracking animals using intrinsic markers when each method is considered independently, there is value in combining all three methods. The resulting methods are provided as part of a new computationally efficient R‐package, gaiah, allowing broad application of our statistical framework to other migratory animal systems.

An evaluation of isotopic (δ2H) methods to provide estimates of avian breeding and natal dispersal

“…, Ruegg et al. ). Therefore, there is a wide range of possibilities to refine these methodologies that are not yet broadly implemented.…”

“…However, future studies that are able to incorporate tracers of longitude (e.g., genetic markers, Ruegg et al. ) in a multi‐variate assignment within a Bayesian framework (Hobson et al. ) could be used in conjunction with the methods employed in this study.…”

“…Under climate change predictions, the northward movement of species may be favored (Virkkala et al 2008, Coristine andKerr 2015). However, future studies that are able to incorporate tracers of longitude (e.g., genetic markers, Ruegg et al 2017) in a multi-variate assignment within a Bayesian framework ) could be used in conjunction with the methods employed in this study.…”

“…), genetic markers (Ruegg et al. ), and trace elements (Poesel et al. ) obtained from individuals sampled across numerous breeding populations can provide a coarse but pragmatic alternative.…”

“…This lack of progress is understandable because of the immense challenge of marking and following enough individuals over space and time to provide meaningful inference into the frequency, scale, and biological significance of such movements. Given the challenges of studying dispersal using traditional marking techniques such as radio telemetry, data from the use of intrinsic markers such as stable isotopes (Hobson et al 2004, Studds et al 2012, genetic markers (Ruegg et al 2017), and trace elements (Poesel et al 2008) obtained from individuals sampled across numerous breeding populations can provide a coarse but pragmatic alternative. Moreover, even if precise natal or previous breeding origins are not discernable, it may still be possible within acceptable margins of error, to identify those individuals that derive from the local population vs. immigrants.…”