Table 1.

Success rate of Mantel‐based methods to identify the true resistance model when compared to all the competing resistance models in a landscape cluster (including the discriminant surface). Success rate is the proportion of all 50 MC replicates in which the true resistance model outperformed all other resistance models in a cluster. Results are pooled across truth 10U and truth 100SQ. Log refers to whether the cost distances were log‐transformed to better linearize their relationship with genetic distance. Eq indicates whether the genetic data used were from a generation that had reached genetic equilibrium or whether a prior generation was used. Numbers in table are means with standard deviations in parentheses

	Log = N Eq = N	Log = N Eq = Y	Log = Y Eq = N	Log = Y Eq = Y	Mean across approaches
Causal modeling (α = 0.05)	0.009 (0.044)	0.313 (0.426)	0.153 (0.310)	0.388 (0.450)	0.216 (0.374)
Causal modeling (α = 0.005)	0.013 (0.052)	0.326 (0.427)	0.173 (0.323)	0.414 (0.448)	0.231 (0.379)
Relative support (RS)	0.430 (0.425)	0.397 (0.424)	0.511 (0.405)	0.547 (0.423)	0.471 (0.420)
Simple Mantel r	0.462 (0.414)	0.468 (0.455)	0.522 (0.419)	0.554 (0.437)	0.501 (0.429)
Mean across methods	0.228 (0.367)	0.376 (0.434)	0.339 (0.405)	0.475 (0.442)	0.354 (0.421)