Extrapolation of spatiotemporal phenotypic distributions
We characterized candidate SNP genotypic distributions across time and space to better understand the ecological niche of life history traits. These spatiotemporal distributions were characterized using the candidate SNPs with the most robust associations with body size (chromosome 02) and sexual maturity (chromosome 01). Distributions of representative SNPs of the other adaptive chromosomes (Chromosomes 04 and 22) were also characterized (Fig. S1, S2).
We used two independent datasets to characterize spatial and temporal distributions of genetic variation (Table S1). These datasets were independent of each other and separate from the association testing samples, and they were optimally suited for these characterizations. For the spatial dataset, we primarily used collections of larvae and juveniles (95% of dataset of N=3,435) but included some adult collections that were distributed widely across the species’ range. Larvae and juveniles were the ideal life stage to represent genotypic distributions of individuals that successfully spawned at discrete locations throughout the range. Adult collections were used to fill in portions of the range where larval samples were not available. Genotyping was partially conducted with a TaqMan assay panel (Hesset al. 2015), which overlapped the GT-seq panel by 85 SNPs they had in common. COLONY v. 2.0.6.5 (Jones and Wang 2010) was used to reconstruct full-sibling families (Wang 2004) using the 85 shared SNPs on each of the 70 collections. We analyzed all collections together as one using the following parameter settings: polygamous mating for males and females without inbreeding, full-likelihood, medium length of run, no allele updating, and no sibship priors. Only 1 collection out of the 7 adult collections had full siblings (N=13, Stamp River, B.C.) which were maintained to accurately represent this small spawning segment. We excluded duplicate genotypes, 797 full siblings, and collections with fewer than 5 individuals, resulting in a final set of 57 collections consisting of a total of 2,581 individuals each representing a unique family (Table S2). This dataset was then used to calculate allele frequencies across collections for the representative candidate SNPs Etr_464 and Etr_5713 within the adaptive regions on chromosomes 01 and 02, respectively.
For the temporal dataset, we used individuals collected from two successive spawning runs at Willamette Falls (2014 – 2015; N of 868 and 581, respectively) over which it was possible to randomly sample the majority of the annual adult migration of Pacific lamprey (typically Feb – August) in weekly strata. A daily abundance estimate (Whitlocket al. 2019) was used to expand candidate SNP allelic proportions in the weekly strata. One biological complexity was that a portion of the adults encountered before May probably overwintered and experienced shrinkage in body size due to advanced maturation (Beamish 1980). Therefore, in addition to characterizing allele frequencies of candidate SNPs Etr_464 and Etr_5713, we categorized fish by body length to provide insight into the transition between overwintered fish and newly-arrived migrants.