A finding we didn't anticipate is that these small, fragmented populations share many large/very large structural variants - presumably inversions, presumably influenced by balancing selection. We don't know what their phenotypic consequences are, but they have substantial effects on gene expression