Aim: Freshwater fish radiations are often characterized by multiple closely-related species in close proximity, which can lead to introgression and associated discordance of mitochondrial and nuclear characterizations of species diversity. As a case in point, single locus nuclear versus mitochondrial analyses of New Zealand’s stream-resident Galaxias vulgaris complex have yielded conflicting phylogenies. Our goal is to use genome-wide divergence patterns among these fishes to evaluate the potential role of mitochondrial capture in obscuring species diversity and to understand how ancient and anthropogenic drainage modification explains this diversity.

Location: Freshwater ecosystems of New Zealand. Methods: We generate and analyze a genome-wide data set comprising 52,352 SNPs across 187 Galaxias specimens to resolve the phylogeny of this recent fish radiation. We conduct phylogenetic, PCA, STRUCTURE, and ABBA-BABA analyses to evaluate the evolutionary relationships of lineages in the context of natural and anthropogenic river drainage alterations.

Results: In addition to the 11 previously recognized stream-resident lineages, genome-wide data reveal a twelfth candidate species (G. ‘Pomahaka’), apparently obscured by introgressive mitochondrial capture. We identify additional examples of mito-nuclear discordance and putative mitochondrial capture, likely mediated by geological and anthropogenic modification of drainage boundaries.

Main conclusions: Our study highlights the need for genome-wide approaches for delimiting freshwater biodiversity. Genetic data also reveal the influence of drainage history on freshwater biodiversity, including the rapid divergence of recently fragmented fish populations, and the conservation genetic risks of anthropogenic translocations events.

This is a dataset of 187 Galaxias sp. from across their range in Aotearoa New Zealand. They have been sequenced using Genotyping by-Sequencing, a variant of RAD-Seq. For detailed methods, please refer to the README and the script. Do not hesitate to contact the authors.

This folder includes the main script used to go from raw data to SNPs, as well as the code used to produce the phylogenetic tree.

The demultiplexed data can be obtained from the Short-Read-Archive as specified in the article.

For any questions, please do not hesitate to contact Ciaran Campbell (ciarancampbell10@gmail.com) or Ludovic Dutoit (ludovic.dutoit@otago.ac.nz).