Reproductive bias in the reintroduced population
By combining the kinship estimates from COLONY 2.0 and Sequoia, we were able to identify at least one family for all F2individuals. This is a significant accomplishment of our study, as there have been few instances where detailed kinships, such as GG and UANN relationships, have been identified in a highly inbred population.
We next investigated whether there was any bias in reproductive contribution among the identified families. In the 2018 year-group, there was no statistically significant difference between the contributions from the three families that were likely involved in reproduction. This is a desirable outcome, as outlined in the introduction, because bias among families can result in a loss of genetic diversity (Willoughby et al., 2017). However, in the 2019 year-group, we observed a significant bias in breeding contribution among families. This was largely influenced by the fact that no F2 individuals were found from the A4, A5, or A6A7A9 families. We anticipated that F1 individuals from these families would likely have had few offspring in 2019. It is probable that a reduction in the water level of the NN6 pool, caused by the collapse of a downstream weir due to a flood in fall 2018, had a negative impact on the growth and survival rates of F1individuals in families A4, A5, and A6A7A9 (Shitara Dam Construction Office, unpublished information). Although T. ichikawai typically reaches sexual maturity at age 2–3 y (Watanabe, 1994a, 2008), it is possible that only a few F1 individuals from the A4, A5, and A6A7A9 families had reached maturity during the 2019 breeding season because of delayed growth. Even if some F1 individuals from these families had reached maturity, they would have been smaller than the F1 individuals from the A1, A2, and A3 families, which were one year older, and therefore at a disadvantage in breeding competition. Moreover, the low water level may have decreased the overall reproductive rate in 2019. The number of F1individuals able to produce offspring in 2019 was small, as only 12 F2 samples from year-group 2019 were obtained, despite a vigorous sampling effort, of which six had exactly the same estimated grandparents (m14A and f7A, m18A and f10A), and six had exactly the same grandparents (m18A and f10A), indicating that the number of F1 individuals that were able to produce offspring in 2019 was small.