Meristic co‐evolution and genomic co‐localization of lateral line scales and vertebrae in Central American cichlid fishes

Abstract

Meristic traits are often treated as distinct phenotypes that can be used to differentiate and delineate recently diverged species. For instance, the number of lateral line scales and vertebrae, two traits that vary substantially among Neotropical Heroine cichlid species, have been previously suggested to co‐evolve. These meristic traits could co‐evolve due to shared adaptive, developmental, or genetic factors. If they were found to be genetically or developmentally non‐independent, this might require a more general re‐evaluation of their role in evolutionary or taxonomic studies. We expanded a previous analysis of correlated evolution of meristic traits (lateral line scales and vertebrae counts) in these fishes to include a range of phylogenetic reconstructions as well as the analyses of 13 Nicaraguan Midas cichlid species (Amphilophus spp.). Additionally, we performed qualitative traits locus (QTL) mapping in a F2 laboratory‐reared hybrid population from two ecologically divergent Midas cichlid fish species to discover and evaluate whether genomic co‐segregation might explain the observed patterns of meristic co‐evolution. Meristic values for these traits were found to morphologically differentiate some species of the Midas cichlid adaptive radiation. Our QTL analysis pinpointed several genomic regions associated with divergence in these traits and highlighted the potential for genomic co‐segregation of the lateral line and vertebrae numbers on two chromosomes. Further, our phylogenetic comparative analyses consistently recovered a significant positive evolutionary correlation between the counts of lateral line scale and vertebrae numbers in Neotropical cichlids. Hence, the findings of genomic co‐segregation could partially explain the co‐evolution of these two meristic traits in these species. Continuing to unravel the genetic architecture governing meristic divergence helps to better understand both trait correlations and the utility of meristic traits in taxonomic diagnoses and how traits in phenotypes might be expected to co‐evolve.