Conclusion
Low-coverage WGS is a powerful and potentially cost-effective approach
for population assignment studies. We demonstrate that high assignment
accuracy can be obtained for weakly differentiated populations, even for
individuals with very low sequencing coverage (< 0.1X). We
further demonstrate the importance of balancing the effective sample
sizes of source populations to avoid assignment bias due to variation in
the precision of allele frequency estimation. By applying these methods
to the American Redstart, we reveal broad-scale parallel migration and
highlight unique population-specific patterns of connectivity. In
combination with our demographic analysis, we demonstrate the importance
of the Northern Temperate-Greater Antilles migratory population to the
total abundance of the species. Furthermore, our identification of
nonbreeding regions for the genetically distinct breeding populations
provides a foundation for a full annual cycle approach towards
preserving genetic diversity. Together, our results provide a valuable
framework for studies that aim to use lcWGS to understand the ecology
and evolution of migratory species.