Hybridization between coastal cutthroat trout (Oncorhynchus clarkii clarkii) and steelhead (O. mykiss) was assessed in the Smith River, California. Individuals were categorized as pure or as one of 10 hybrid classes using 30 ‘diagnostic’ single-nucleotide polymorphisms positioned on 26 separate chromosomes. Most of the individuals examined (n = 876), were pure coastal cutthroat trout (n = 634) or pure steelhead (n = 213), and 29 individuals were identified as having hybrid ancestry. Among hybrids, first generation hybrids (n = 15) and coastal cutthroat trout backcrosses (n = 12) were the most common. No individuals were identified as backcrosses to SH, suggesting the presence of genetic or behavioral mechanisms constraining such backcrosses, or the growth and survival of their progeny. Mitochondrial DNA of 14 of 15 F1 hybrids was of steelhead origin, suggesting that hybridization was driven primarily by sneak-mating of male coastal cutthroat trout with female steelhead. Evaluation of classical phenotypic characters for coastal cutthroat trout and steelhead (i.e., jaw slash, maxillary length, and hyoid teeth) were not reliable by themselves for identification of either pure parental fish or hybrids. In contrast, analysis with geometric morphometrics revealed distinctive body shapes for pure coastal cutthroat trout and steelhead, and the combination of classical traits and geometric morphology was mostly accurate in distinguishing them. However, first generation hybrids and backcrosses overlapped completely with parental types, highlighting challenges in hybrid identification using phenotypic traits.

This data was collected by sampling coastal cutthroat trout and steelhead in the Smith River. Fish were captured using night netting, hook and line, electrofishing, and weirs from mid-May through August 2013. A tissue sample, photograph, and fork length measurement were collected from 874 fish. Phenotypic characteristics of jaw slash color intensity, maxillary length compared to posterior margin of eye, and presence of hyoid teeth were also recorded for each fish in the field. We generated genotypes for 96 single-nucleotide polymorphisms (SNPs) using TaqMan^TM 5’ nuclease assays (Applied Biosystems Inc.) and 96.96 Dynamic Genotyping Arrays (Fluidigm Corporation), with imaging on an EP1 instrument l (Fluidigm). Genotypes were manually called using the SNP Genotyping Analysis Software (3.0.2, Fluidigm). NewHybrids was configured to assign fish to ten hybrid classes characterized by the expected frequencies of loci having 0, 1, or 2 alleles originating from each parental type. 

Body morphologies of CCT, SH, and their hybrids were compared using landmark-based geometric morphometrics. Landmark locations were as follows: 1) tip of snout; 2) anterior margin of eye; 3) posterior margin of eye; 4) posterior end of operculum; 5) posterior end of maxillary; 6) origin of pectoral fin; 7) origin of pelvic fin; 8) origin of anal fin; 9) anterior attachment of ventral membrane of caudal fin; 10) base of middle caudal rays; 11) anterior attachment of dorsal membrane of caudal fin; 12) origin of adipose fin; 13) origin of dorsal fin; and 14) posterior end of the neurocranium. To ensure landmark placement did not migrate, and to adjust for imperfect digitizing, each fish was landmarked twice and results of the digitizing events were averaged. All morphometric analyses were performed in Morpho J.