Network analysis of migration pattern
To quantify the proportion and direction of recent migration patterns among populations in small and large patches, relative migration levels (Nm ) were estimated between populations calculated based on neutral SNPs using the GST and Nmparameters (Sundqvist et al., 2016) with the divMigrate function of the R package diveRsity (Keenan et al., 2013). This approach permits the determination of migration directional differences between genetically structured populations, where deterministic extinction and density-dependent processes govern source-sink population dynamics (i.e., metapopulation structure). Relative migration network analysis was estimated using only the complete dataset, as the probability of assigning the correct migration direction is directly related to the increase in the number of samples and loci in the dataset (Sundqvist et al., 2016). The direction and magnitude of recent migration patterns were summarized as a network from the significant estimated effective number of migrants (Nm ) based on 1000 bootstrap interactions and with a filter threshold of 0.35. Network graphics were generated using the qgraph R package (Epskamp et al., 2012).