All in the Family: Juvenile Sampling of Bull Trout for Genetic Monitoring

The presence of family spatial structure can bias results for some genetic estimators like the effective population size per generation (N_e) or the number of successful spawners per year (N_b). Ideally samples should contain a representative but random subset of individuals from all families while providing the minimum sample size required for precise estimates. In brook trout, work suggests a sample size of 75 age-0 fish collected from at least 3 different stream reaches is a successful strategy to estimate N_b. We used juvenile bull trout samples from three cohorts (age-0, age-1, age-2) collected from multiple stream reaches in each of three different streams to test for the effects of family structure on estimates of N_b. Similar to published results for brook trout, we found that samples from multiple stream reaches were required to account for family structure and accurately estimate N_b. Strong clustering of full sib pairs existed at distances up to approximately 600m, especially in age-0 fish, suggesting a minimum distance between sampling transects to reduce the potential of family bias.  Interestingly, cohort N_b estimates were smaller and showed less variation than observed in redd counts, varying from 9-26 successful spawners compared to 13-72 redds.