Serial sampling bridges a gap between ecological and genetical definitions of immigrant: an empirical test in a grey-sided vole population

Abstract

The assignment test is believed to have the potential to estimate the contemporary rate of immigration and is considered as an alternative to ecological methods (e.g., mark-recapture and radio/GPS tracking). However, the conventional assignment test provides different results from those of the mark-recapture method, when it is applied to populations with low genetic differentiation from one another. The difference derives from a difference in the definition of “immigrant”; immigrants in the mark-recapture method are individuals coming from outside a predefined population, whereas immigrants in the assignment test are individuals coming from other genetically differentiated populations. We propose a sampling design of the assignment test bridging a gap between the ecological and genetical definition of immigrant. The assignment test application based on this design identifies immigrants referring to the allele frequency of a focal population before dispersal occurs. We tested the efficiency of the proposed sampling design by comparing immigration rates estimated by the assignment test with those from mark-recapture records of a grey-sided vole (Myodes rufocanus) population, which was not significantly genetically differentiated from the surrounding populations. The proposed design provided 7–30% closer results to those from the mark-recapture in accuracy, in comparison with results from the conventional assignment test, and in the detection power 23–54% improvement was observed. The serial sampling for the assignment test could bridge the differences between ecological methods and the assignment test.