Genotyping-in-Thousands by sequencing (GT-seq) is a method that uses next-generation sequencing of multiplexed PCR products to generate genotypes from relatively small panels (50-500) of targeted single-nucleotide polymorphisms (SNPs) for thousands of individuals in a single Illumina HiSeq lane. This method uses only unlabelled oligos and PCR master mix in two thermal cycling steps for amplification of targeted SNP loci. During this process, sequencing adapters and dual barcode sequence tags are incorporated into the amplicons enabling thousands of individuals to be pooled into a single sequencing library. Post sequencing, reads from individual samples are split into individual files using their unique combination of barcode sequences. Genotyping is performed with a simple perl script which counts amplicon-specific sequences for each allele, and allele ratios are used to determine the genotypes. We demonstrate this technique by genotyping 2068 individual steelhead trout (Oncorhynchus mykiss) samples with a set of 192 SNP markers in a single library sequenced in a single Illumina HiSeq lane. Genotype data were 99.9% concordant to previously collected TaqMan TM genotypes at the same 192 loci, but call rates were slightly lower with GT-seq (96.4%) relative to Taqman (99.0%). Of the 192 SNPs, 187 were genotyped in ≥90% of the individual samples and only 3 SNPs were genotyped in <70% of samples. This study demonstrates amplicon sequencing with GTseq greatly reduces the cost of genotyping hundreds of targeted SNPs relative to existing methods by utilizing a simple library preparation method and massive efficiency of scale.

Unlike most anadromous fishes that have evolved strict homing behaviour, Pacific lamprey (Entosphenus tridentatus) seem to lack philopatry as evidenced by minimal population structure across the species range. Yet unexplained findings of within-region population genetic heterogeneity coupled with the morphological and behavioural diversity described for the species suggest that adaptive genetic variation underlying fitness traits may be responsible. We employed restriction site-associated DNA sequencing to genotype 4439 quality filtered single nucleotide polymorphism (SNP) loci for 518 individuals collected across a broad geographical area including British Columbia, Washington, Oregon and California. A subset of putatively neutral markers (N = 4068) identified a significant amount of variation among three broad populations: northern British Columbia, Columbia River/southern coast and 'dwarf' adults (F(CT) = 0.02, P ≪ 0.001). Additionally, 162 SNPs were identified as adaptive through outlier tests, and inclusion of these markers revealed a signal of adaptive variation related to geography and life history. The majority of the 162 adaptive SNPs were not independent and formed four groups of linked loci. Analyses with matsam software found that 42 of these outlier SNPs were significantly associated with geography, run timing and dwarf life history, and 27 of these 42 SNPs aligned with known genes or highly conserved genomic regions using the genome browser available for sea lamprey. This study provides both neutral and adaptive context for observed genetic divergence among collections and thus reconciles previous findings of population genetic heterogeneity within a species that displays extensive gene flow.

To elucidate the mechanisms of thermal adaptation and acclimation in ectothermic aquatic organisms from differing climates, we used a common-garden experiment for thermal stress to investigate the heat shock response of redband trout (Oncorhynchus mykiss gairdneri) from desert and montane populations. Evidence for adaptation was observed as expression of heat shock genes in fish from the desert population was more similar to control (unstressed) fish and significantly different (P ≤ 0.05) from those from the montane population, while F1 crosses were intermediate. High induction of heat shock proteins (Hsps) in the montane strain appeared to improve short-term survival during first exposure to high water temperatures, but high physiological costs of Hsp production may have led to lower long-term survival. In contrast, the desert strain had significantly lower heat shock response than the montane fish and F1 crosses, suggesting that these desert fish have evolved alternative mechanisms to deal with thermal stress that provide better balance of physiological costs. Genomewide tests of greater than 10 000 SNPs found multiple SNPs that were significantly associated with survival under thermal stress, including Hsp47 which consistently appeared as a strong candidate gene for adaption to desert climates. Candidate SNPs identified in this study are prime targets to screen more broadly across this species' range to predict the potential for adaptation under scenarios of climate change. These results demonstrate that aquatic species can evolve adaptive responses to thermal stress and provide insight for understanding how climate change may impact ectotherms.

Little is known of the genetic basis of migration despite the ecological benefits migratory species provide to their communities and their rapid global decline due to anthropogenic disturbances in recent years. Using next generation sequencing of restriction-site associated DNA (RAD) tags we genotyped thousands of SNPs in two wild populations of migratory steelhead and resident rainbow trout (Oncorhynchus mykiss) from the Pacific Northwest of the United States. One population maintains a connection to the sea, whereas the other population has been sequestered from its access to the ocean for more than fifty years by a hydropower dam. Here we performed a genome-wide association study (GWAS) to identify 504 RAD SNP markers from several genetic regions that were associated with the propensity to migrate both within and between the populations. Our results corroborate those in previous quantitative trait loci (QTL) studies and provide evidence for additional loci associated with this complex migratory life history. Our results suggest a complex multi-genic basis with several loci of small effect distributed throughout the genome contributing to migration in this species. We also determined that despite being sequestered for decades, the landlocked population continues to harbor genetic variation associated with a migratory life history and ATPase activity. Furthermore we demonstrate the utility of genotyping-by-sequencing and how RAD-tag SNP data can be readily compared between studies to investigate migration within this species.

Parentage-based tagging (PBT) is a promising alternative to traditional coded-wire tag (CWT) methodologies for monitoring and evaluating hatchery stocks. This approach involves the genotyping of hatchery broodstock and uses parentage assignments to identify the origin and brood year of their progeny. In this study we empirically confirmed that fewer than 100 single nucleotide polymorphisms (SNPs) were needed to accurately conduct PBT, we demonstrated that our selected panel of SNPs was comparable in accuracy to a panel of microsatellites, and we verified that stock assignments made with this panel matched those made using CWTs. We also demonstrated that when sampling of spawners was incomplete, an estimated PBT rate for the offspring could also be predicted with fewer than 100 SNPs. This study in the Snake River basin is one of the first large-scale implementations of PBT in salmonids and lays the foundation for adopting this technology more broadly in the region, thereby allowing the unprecedented ability to mark millions of smolts and an opportunity to address a variety of parentage-based research and management questions.