The evolutionary histories of alpine species are often directly associated with responses to glaciation. Deep divergence among populations and complex patterns of genetic variation have been inferred as consequences of persistence within glacier boundaries (i.e. on nunataks), while shallow divergence and limited genetic variation is assumed to result from expansion from large refugia at the edge of ice shields (i.e. massifs de refuge). However, for some species, dependence on specific microhabitats could profoundly influence their spatial and demographic response to glaciation, and such a simple dichotomy may obscure the localization of actual refugia. In this study, we use the Nebria ingens complex (Coleoptera: Carabidae), a water-affiliated ground beetle lineage, to test how drainage basins are linked to their observed population structure. By analyzing mitochondrial COI gene sequences and genome-wide single nucleotide polymorphisms, we find that the major drainage systems of the Sierra Nevada Mountains in California best explain the population structure of the N. ingens complex. In addition, we find that an intermediate morphotype within the N. ingens complex is the product of historical hybridization of N. riversi and N. ingens in the San Joaquin basin during glaciation. This study highlights the importance of considering ecological preferences in how species respond to climate fluctuations and provides an explanation for discordances that are often observed in comparative phylogeographic studies.

The genetic data, including DNA sequences, single nucleotide polymorphisms (SNPs), and site allele frequency spectrum (sfs), is collected from Nebria ingens complex samples. The DNA sequences were used to reconstruct the phylogeny of this species complex, and the the SNP and sfs data were used to conduct analyses regarding population structure and demographic history. The phylogeny reconstruction includes Bayesian tree with BEAST2 and haplotype network with popart. We use SNP data to analyze the population structure using vcftools, sNMF, and conStruct. We also use sfs data to reconstruct the demographic history of this species complex using stairway plot and ∂a∂i.

This dataset includes data and major scripts to repeat the analyses. The raw data such as DNA reads are stored in NCBI.