Whole genome resequencing reveals signatures of rapid selection in a virus-affected commercial fishery

Abstract

Infectious diseases are recognized as one of the greatest global threats to biodiversity and ecosystem functioning. Consequently, there is a growing urgency to understand the speed at which adaptive phenotypes can evolve and spread in natural populations to inform future management. Here we provide evidence of rapid genomic changes in wild Australian blacklip abalone (Haliotis rubra) following a major population crash associated with an infectious disease. Genome scans on H. rubra were performed using pooled whole genome resequencing data from commercial fishing stocks varying in historical exposure to haliotid herpesvirus-1 (HaHV-1). Approximately 25,000 single nucleotide polymorphism loci associated with virus exposure were identified, many of which mapped to genes known to contribute to HaHV-1 immunity in the New Zealand pāua (Haliotis iris) and herpesvirus response pathways in haliotids and other animal systems. These findings indicate genetic changes across a single generation in H. rubra fishing stocks decimated by HaHV-1, with stock recovery potentially determined by rapid evolutionary changes leading to virus resistance. This is a novel example of apparently rapid adaptation in natural populations of a nonmodel marine organism, highlighting the pace at which selection can potentially act to counter disease in wildlife communities.

Keywords: blacklip abalone; genetic adaptation; haliotid herpesvirus-1; infectious diseases; southeastern Australia; whole genome resequencing.

FIGURE 1

Sampling sites selected for population genomic analysis from southeastern Australia. Figure legend and colour coding of mapped sites indicate history of virus exposure. Refer to Table 1 for sample codes

FIGURE 2

Heatmap of pairwise estimates of genetic differentiation (F _ST) among sample locations based on (a) all 7,745,655 SNPs and (b) the 25,854 SNP loci associated with AVG exposure. *Virus‐unaffected sample locations

FIGURE 3

Plots of eigenvalues from the principal components analysis: (a) plot of axis 1 and 2 eigenvalues, and (b) density plot of axis 1 eigenvalues. Plots are based on candidate SNP genotypes from each of the 28 pooled whole genome resequencing libraries representing virus‐affected (red) and unaffected (black) fishing stocks

FIGURE 4

(a) Regression analysis indicating a positive linear relationship between number of candidate SNPs (BF > 50) and scaffold length. Outlier scaffolds with a greater frequency of candidate SNPs relative to scaffold length are plotted in red. (b) Linkage disequilibrium heatmaps of scaffolds with the greatest number of candidate SNPs (QXH01000030.1 and QXJH01000212.1) generated with the package ldblockshow (Dong et al., 2021). Heatmaps depict the pairwise linkage disequilibrium measure of D′ (refer to colour key) between each SNP with a BF ≥ 50, while green lines link the relative position of the candidate SNPs to the heatmap. In addition, black triangle sections represent detected haplotype blocks; these are genomic regions of low recombination