Biases of STRUCTURE software when exploring introduction routes of invasive species.

摘要: Population genetic methods are widely used to retrace the introduction routes of invasive species. The unsupervised Bayesian clustering algorithm implemented in STRUCTURE is amongst most frequently these methods, but its ability provide reliable information about has never been assessed. We simulated microsatellite datasets evaluate extent which results provided by were misleading for inference routes. focused on an invasion scenario involving one native and two independently introduced populations, because it sole that can be rejected when obtaining a particular with analysis at K = 2 (two clusters). Results classified as “misleading” or “non-misleading”. investigated influence effective size, bottleneck severity number loci type frequency results. showed obtained 10% all datasets. Our highlighted categories output. first occurs population low level diversity. In this case, populations may very similar, despite their independent histories. second category from convergence issues K = 2, strong and/or large numbers resulting high levels differentiation between three populations. Overall, risk being misled context inferences moderate, important remain cautious diversity genuine multimodality runs involved.