Authors: Andrew N. Black, Erangi J. Heenkenda, Samarth Mathur, Janna R. Willoughby, Brian L. Pierce, Sarah J. Turner, David Rizzuto, J. Andrew DeWoody
Speciation is often driven by selective processes like those associated with viability, mate choice, or local adaptation, and “speciation genes” have been identified in many eukaryotic lineages. In contrast, neutral processes are rarely considered as the primary drivers of speciation, especially over short evolutionary timeframes. Here, we describe a rapid vertebrate speciation event driven primarily by genetic drift. The White Sands pupfish (Cyprinodon tularosa) is endemic to New Mexico’s Tularosa Basin where the species is currently managed as two Evolutionarily significant units (ESUs) and is of international conservation concern (Endangered). Whole-genome resequencing data from each ESU showed remarkably high and uniform levels of differentiation across the entire genome (global FST ≈ 0.40). Despite inhabiting ecologically dissimilar springs and streams, our whole-genome analysis revealed no discrete islands of divergence indicative of strong selection, even when we focused on an array of candidate genes. Demographic modeling of the joint allele frequency spectrum indicates the two ESUs split only ~4 to 5 kya and that both ESUs have undergone major bottlenecks within the last 2.5 millennia. Our results indicate the genome-wide disparities between the two ESUs are not driven by divergent selection but by neutral drift due to small population sizes, geographic isolation, and repeated bottlenecks. While rapid speciation is often driven by natural or sexual selection, here we show that isolation and drift have led to speciation within a few thousand generations. We discuss these evolutionary insights in light of the conservation management challenges they pose.
Evolutionarily significant units (ESUs) encompass lineages of genetically and ecologically distinct organisms that are effectively managed as intraspecific taxonomic units which merit increased conservation focus (
1–
4). Defining such conservation units is fraught with both political and biological challenges, but genomic data can refine these boundaries in natural populations (
5–
7). Population genomics can clarify the phylogeography and the timing of boundary emergence. Both are important as the cessation of gene flow ultimately puts ESUs or incipient species on different evolutionary trajectories, each subject to the vagaries of drift and selection (
8). Rapid sympatric speciation is well-known (e.g., due to polyploidy or chromosomal rearrangements) (
9), but allopatric speciation usually “arises gradually and incidentally as a result of mutation, genetic drift, and the indirect effects of natural selection driving local adaptation” (
10). This paper is concerned with rapid evolution in allopatry, the relative roles of drift and selection in driving speciation, and the conservation management implications of our population genomic insights.
The endangered (
11) White Sands pupfish (
C. tularosa) is a small desert killifish that is endemic to the Tularosa Basin in New Mexico where it inhabits four sites. Two of the four sites, Salt Creek (SC) and Malpais Spring (MS), are occupied by natural populations (
Fig. 1). The other two sites, Lost River (LR) and Mound Spring, are anthropogenic in origin. The LR population was founded in 1970, likely by fish translocated from SC (
12). The source of
C. tularosa in Mound Spring is less certain, but allozyme and microsatellite marker data suggest that it was also founded from the SC population (
13) between 1967 and 1973 (
12). Two ESUs are currently recognized due in part to pronounced microsatellite differentiation (mean F
ST = 0.39 across two loci) as well as private alleles (
13).
Attempts to identify reciprocal monophyly of gene genealogies, a condition often required for ESU designation (
2), have proven futile as single gene surveys have revealed little to no variation in mitochondrial control region DNA sequences (
13). However, there is additional genetic support for the two ESUs in the form of significant heritable differences in body shape between fish at these sites (
14,
15). There are substantial differences in water chemistry, hydrologic characteristics, and water flow between the MS (ESU1) and SC+LR (ESU2) sites (
12,
16). This is important because adaptive variation is a critical component in shaping population discreteness and, as a result, ESU classification (
4,
17).
Here, we used population genomic data from MS, SC, and LR pupfish to identify the primary evolutionary mechanism driving allopatric divergence between the two ESUs. We did so while simultaneously considering the effects of putatively neutral genomic regions and adaptive regions (19). We reconstructed the demographic history of each ESU, examined patterns of divergence and lineage sorting between ESUs, tested for signatures of selection (including at candidate genes that we expected to underlie local adaptation), and performed formal tests for species delimitation. Our genomic data indicate that isolation and drift, rather than selection for locally adapted variants, have been the primary drivers for substantial and rapid population divergence, leading to the evolution of incipient vertebrate species in only a few thousand generations.
Suggested Citation
Black, A.N., E.J. Heenkenda, S. Mathur, J.R. Willoughby, B.L. Pierce, S.J. Turner, D. Rizzuto, J.A. DeWoody. 2024. Rapid vertebrate speciation via isolation, bottlenecks, and drift. Proceedings of the National Academy of Sciences 121:e2320040121.
