Appendix A: Population groupings
For comparisons, we formed three focal groups: SSI generalist-only, SSI radiating, and the Caribbean. The PCA (Figure A1) indicates that generalists, snail-eaters, and scale-eaters on SSI form distinct clusters based on their genetic variation. The PCA visualization also suggests that: 1) Oyster Pond scale-eaters and generalists fall well within the 95% confidence intervals for their species designations, justifying their inclusion in the SSI radiating population analyses and 2) Mermaid Pond “scale-eaters” (designated as such based on their morphological traits) do not fall within the 95% confidence interval for genetically clustering scale-eaters, and are instead closer to the centroid of generalists. Although previous studies suggest that Mermaid Pond contains scale-eating individuals (Martin and Wainwright. 2013), more recent work suggests that these individuals are distinct from scale-eaters (Richards et al. 2021. Richards and Martin. 2022). Furthermore, there are no published sightings of snail-eating pupfish in Mermaid Pond. For these reasons, we included Mermaid Pond in the generalist-only population calculations as opposed to the SSI radiating populations. Finally, snail-eaters have previously been documented in Moon Rock and Wild Dilly Pond, but scale-eating specialists have never been found in these locations (Martin 2014). Considering that we did not have access to any of the snail-eater specimens from these locations, and that these populations have never contained the full radiation of pupfish,
we included Moon Rock and Wild Dilly Pond in the generalist-only population calculations as opposed to the SSI radiating populations.