Fig. 5: Sensitivity of the multi-generation implicit dispersal model predictions to dispersal trait and sampling design.

Correlation (R² = 0.31*** using a linear regression model fit) between the PLDs categories (x-axis) and the proxy of sampling extent D_btw divided by the optimal number of modelled generations (y-axis). n = number of studies whose model predictions of genetic observations are significant per PLD categories. On each box, the central mark, the bottom and top edges indicate the indicates the median, the 25th and the 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using a cross symbol. Note that Susini et al.⁹¹ has been removed from this figure as it displays the minimal number of localities of our compilation (i.e., three localities while two sampled populations are indeed compassed in the same locality), which does not permit a robust evaluation of D_btw/Opt gen ratio. When considering this study, R² = 0.23***. Source data are provided as a Source Data file.