Abstract
The genus Amatitlania contains two closely related cichlid species in eastern Costa Rica, namely A. septemfasciatus in northern streams, and A. myrnae in southeastern watersheds. A tectonically uplifted ridge separates the geographic ranges of the two species, and although this barrier is predicted to restrict movement, occasional flooding events may enable fish to cross the divide between watersheds and hybridize with one another. This study used Double Digest Restriction Site Associated Sequencing (ddRADseq) to test whether gene flow is occurring between species, quantify genetic diversity across their ranges, and identify locations where hybridization is most likely occurring. Ninety-six individuals were sampled from twelve river sites spanning pure species and putative hybrid zones. DNA was extracted from fin clips, and DNA libraries were prepared using a modified protocol for Double-Digest Restriction Site-Associated DNA Sequencing (ddRADseq). Single Nucleotide Polymorphism (SNP) discovery was performed in the software program, STACKS, followed by filtering in the programs VCFtools and PLINK. Population structure was investigated using Principal Component Analysis (PCA) and Bayesian clustering using STRUCTURE. Nucleotide diversity (π) and pairwise genetic differentiation (FST) were calculated using VCFtools.
PCA and STRUCTURE analysis identified two main groups, corresponding to the parental species, with four sites (Rio Bananito I at Railway Bridge, Rio Bananito II Spider Bridge, Rio Carbon-Cano Negro, and Quebrada Christine) containing individuals with mixed ancestry, confirming active or ancestral hybridization. Pure A. myrnae populations exhibited ~10× higher nucleotide diversity (π ≈ 1.8×10-5) than all other groups (π ≈ 2.0×10-6–3.0×10-6), suggesting that this population has potentially retained a higher percentage of ancestral variation or multiple sublineages of the population exist within the watershed. FST values ranged from low (0.04) between hybrid groups to high (0.38) between pure parental species, reflecting strong but incomplete genomic separation between the two species.
These results demonstrate that the ridge does not entirely prevent gene flow and that hybridization is geographically structured. This has implications for conservation planning, as genetic introgression may either threaten species integrity or represent natural evolutionary processes that were not previously known. Future research should incorporate hydrologic modeling and increased sampling south of the ridge to test whether hybridization is unidirectional and better track which streams fish are moving from and to in this area.