Skip to main content
. 2016 May 21;6(12):4115–4128. doi: 10.1002/ece3.2154

Figure 1.

Figure 1

Summary of simulation methods used in this investigation. Each landscape was replicated five times and simulated using the same seed. A set of discriminant landscapes (using a different seed) was generated to represent a different underlying environmental variable. Gene flow was simulated across the true resistance model for 50 Monte Carlo replications, and Mantel‐based tests were evaluated using causal modeling (α = 0.05, 0.005), relative support, and simple Mantel r.