Genetic Exchangeability - Clustering Analyses
Three different genetic clustering analyses were used to evaluate the potential for gene flow (see Table 2 for summary). The STRUCTURE analysis with All_M20 resulted in an optimal K value of eight (Figure 3a), with the two monophyletic lineages A. microunicolor andA. unicolor A clustering separately; however, A. unicolor B formed six clusters that did not correspond to monophyletic groups. STRUCTURE analyses with UniB-only matrices produce two to three clusters with all but one data set recovering non-monophyletic clusters (Figure 4). UniB_M30 does recover the three monophyletic lineages withinA. unicolor B, but some individuals have mixed cluster assignments (Figure 4). In contrast, both PCA and VAE clustering analyses (Figure 3b, 3c, & 3d) for All-M20 indicate three distinct clusters consistent with monophyletic lineages: A. microunicolor , A. unicolor A, and A. unicolor B. The VAE analysis with UniB_M30 shows some separation between A. unicolor B lineages; however, there is still some overlap between them when considering standard deviation around samples (Figure 4). The lineages within A. unicolor B were not recovered as discrete clusters in any of the three clustering analyses.