Figure 1.
Poecilia thermalis and its natural habitat.
A.-F. Representative specimens of Poecilia thermalis (males on top, females below). A. & B. represent artistic depictions of the species from the original description by Steindachner [35]. C.-F. are photos of freshly collected specimens of the P. thermalis from the large La Esperanza spring in 2012. G. The type locality of P. thermalis at La Esperanza (large spring), Chiapas, Mexico, a typical sulfidic spring habitat. H. A second, smaller sulfide spring (La Esperanza small spring) in close proximity to the type locality, which is also inhabited by P. thermalis.
Table 1.
Summary of the taxonomic history of Poecilia thermalis and P. sulphuraria in chronological order.
Figure 2.
Overview of the study area and sampling localities of three Poecilia species in southern Mexico.
Sampling of P. thermalis, Poecilia mexicana mexicana, and P. sulphuraria in southern Mexico. The colors represent sulfidic (yellow) and non-sulfidic (blue) sites and the numbers represent localities as described in Table 2 of the main manuscript. For orientation purposes, we included black lines representing major roads and gray areas representing major towns in the region.
Table 2.
Overview of samples used in this study for morphometric and population genetic analyses.
Figure 3.
Body shape variation of Poecilia species in the lateral projection.
Depicted are mean principal component scores along the first two principal component axes for each site for P. thermalis (yellow circles), P. sulphuraria (yellow triangles), as well as sulfidic and non-sulfidic populations of P. mexicana across the 23 study sites in southern Mexico. The thin-plate spline transformation grids represent shape variation along each principal component axis.
Table 3.
Results of a multivariate analysis of covariance on lateral body shape of Poecilia from sulfidic and non-sulfidic habitats.
Figure 4.
Convergent changes in body shape of Poecilia species from sulfidic and non-sulfidic habitats in the lateral projection.
Depicted are the mean divergence scores (± SEM; derived from the H2S term in the MANCOVA) for each site for the three formal species (P. thermalis, P. sulphuraria, and P. mexicana) across the 23 sites in southern Mexico from sulfidic (yellow) and non-sulfidic (blue) populations including. The numbers correspond to sites as described in Table 2.
Table 4.
Results of a univariate analysis of covariance on the lateral body shape divergence vector scores between Poecilia from sulfidic and non-sulfidic habitats.
Figure 5.
Body shape variation of Poecilia species in the dorsal projection.
Depicted are mean principal component scores along the first two principal component axes for each site for P. thermalis (yellow circles), P. sulphuraria (yellow triangles), as well as sulfidic and non-sulfidic populations of P. mexicana across the 23 study sites in southern Mexico. The thin-plate spline transformation grids represent shape variation along each principal component axis.
Table 5.
Results of a multivariate analysis of covariance on dorsal body shape of Poecilia from sulfidic and non-sulfidic habitats.
Figure 6.
Convergent changes in body shape of Poecilia species from sulfidic and non-sulfidic habitats in the dorsal projection.
Depicted are mean divergence scores (± SEM; derived from the H2S term in the MANCOVA) for each site for the three formal species (P. thermalis, P. sulphuraria, and P. mexicana) across the 23 sites in southern Mexico from sulfidic (yellow) and non-sulfidic (blue) populations. Numbers correspond to sites as described in Table 2.
Table 6.
Results of a univariate analysss of covariance on the dorsal body shape divergence vector scores between Poecilia from sulfidic and non-sulfidic habitats.
Figure 7.
Hierarchical cluster analysis of Poecilia populations from sulfidic and non-sulfidic environments based on body shape variation in the lateral and dorsal projections.
Colors denote sulfide-adapted (yellow) and non-adapted (blue) populations of three species, Poecilia thermalis (yellow circles), P. sulphuraria (yellow triangles), and P. mexicana. The shapes represent the drainages (diamonds- Tacotalpa, squares- Puyacatengo, circles- Ixtapangajoya, and triangles- Pichucalco) and the numbers correspond to sites as described in Table 2.
Figure 8.
Bayesian tree from phylogenetic analysis of Poecilia species for five markers rooted with poeciliid outgroups.
Phylogenetic analyses of two mitochondrial and three nuclear genes (5337 base pairs) yielded nodal support values (in percent) represent (from top to bottom) Bayesian Posterior Probabilities, as well as RAxML, and GARLI bootstrap support values. Asterisks denote nodal support of ≥95% for all three methods. Nodes with no values present either had low values or were of little interest for this study.
Figure 9.
Genetic structure among populations based on microsatellites analysis in N = 272 individuals from 10 populations.
The top panel is a bar plot showing the assignment scores of individuals by STRUCTURE with K = 2 with yellow representing P. thermalis and P. sulphuraria as a cluster from the Puyacatengo and Pichucalco sulfidic drainages and blue representing P. m. mexicana from non-sulfidic sites in the same drainage. The bottom panel is a bar plot showing the assignment scores clustering at K = 7, the second most likely number of distinct groups.