Recent Adaptation in a Threatened Salmonid Revealed by Museum Genomics

Abstract

Steelhead/rainbow trout (Oncorhynchus mykiss) is an imperilled salmonid with two main life history strategies: migrate to the ocean or remain in freshwater. Domesticated hatchery forms of this species have been stocked into almost all California waterways, possibly resulting in introgression into natural populations and altered population structure. We compared whole-genome sequence data from contemporary populations against a set of museum population samples of steelhead from the same locations that were collected prior to most hatchery stocking. We observed minimal introgression and few steelhead-hatchery trout hybrids despite a century of extensive stocking. Our historical data show signals of introgression with a sister species and indications of an early hatchery facility. Finally, we found that migration-associated haplotypes have become less frequent over time, a likely adaptation to decreased opportunities for migration. Since contemporary migration-associated haplotype frequencies have been used to guide species management, we consider this to be a rare example of shifting baseline syndrome that has been validated with historical data. We suggest cautious optimism that a century of hatchery stocking has had minimal impact on California steelhead population genetic structure, but we note that continued shifts in life history may lead to further declines in the ocean-going form of the species.

Keywords: adaptation; conservation genetics; fish; hybridization; natural selection and contemporary evolution; population genetics – empirical.

FIGURE 1

Sampling scheme and population genetic analysis. (A–C) Locations where fish were sampled from California rivers. Cities and main stems of sampled tributaries are written in grey. Each marker represents a sampling location, where X's are historical and O's are contemporary, coloured according to the legend in (D). For some rivers, the exact historical sampling location is unknown (see Note S3). Grey bars indicate impassable barriers. (D) PCA analysis of genomic data with distinct colours and symbols indicating separate geographic populations and time periods. Hatchery populations are in grey. (E) Unsupervised ADMIXTURE (K = 3 clusters) analysis. Each bar represents an individual, and each colour represents a different ancestry assignment.

FIGURE 2

Isolation by distance and PCA suggest persistent North–South population structure. Isolation by distance calculated by the outgroup f3 statistic between all pairs of populations (black dots) and all pairs of individual trout (green circles for non‐hybrid pairs, orange circles for pairs that include at least one hybrid) in (A) historical populations and (B) contemporary populations. The best fit line, Pearson's correlation and Mantel test p‐value are calculated using data from all pairs of populations (N = 10, black dots). Jitter was added to the x‐axis of the green circles to improve visualisation. (C) A PCA of contemporary and historical steelhead populations after removing trout with suspected hatchery or cutthroat trout ancestry. The dashed line separates historical and contemporary samples, with populations largely mirrored across the line.

FIGURE 3

Contemporary steelhead show minimal hatchery introgression. (A) Diagram of D‐statistic topology and interpretation, where a positive D‐statistic indicates excess allele sharing between hatchery and contemporary populations, and a negative D‐statistic indicates excess allele sharing between hatchery and historical populations. Z‐scores > 3 are considered moderately significant, with Z‐scores > 6 highly significant. (B) D‐statistics for all five rivers, quantifying significance of introgression between historical or contemporary steelhead populations and any of the five hatchery strains. Z‐scores coloured according to (A).

FIGURE 4

Contemporary steelhead genomes show signs of adaptation and lost genetic diversity. (A) Comparison of the proportion of the Omy05 migration‐associated haplotype in historical and contemporary populations. R indicates the resident/rearranged haplotype, and A indicates the anadromous/ancestral haplotype. (B) Comparison of genomic diversity present in historical (dotted lines), contemporary (solid lines), and hatchery samples (solid grey lines). Contemporary and historical hatchery hybrids were not included in the analysis.