MATERIALS AND METHODS
Target population and samples
Tachysurus ichikawai is nocturnal species, resides in pools and flats with slow currents. It does not exhibit significant migratory patterns over its lifespan, and it is not uncommon for individuals to reside in a single pool for multiple years (Watanabe, 1995). The average lifespan for this species ranges from 3 to 6 y, with sexual maturity typically attained at approximately 2 to 3 y (Watanabe, 1994a, 2008). The sex ratio within the breeding population is skewed towards females (Watanabe, 1994a), likely due to the higher mortality rate of adult males. Breeding is concentrated during June and July, during which males establish breeding territories and females visit them, thus demonstrating continuous polygamy (Watanabe, 1994a, 1994b, 2008).
The area 70 km upstream from the mouth of the Toyokawa River, which flows into Mikawa Bay, is an important habitat for a population of this species. In this location, Shitara Dam is under construction (impounded water surface area: approximately 3 km²), and when the dam is completed, the habitat of the population will disappear. Therefore, surveys and consideration of specific conservation measures have been ongoing for about 20 years since 2003. The Shitara Dam Construction Office has conducted an environmental impact assessment (EIA) of the dam project based on the EIA Law; the assessment report was published in 2007. This report recommends conservation measures for T. ichikawai because of the impact of the project, and translocation experiments are being conducted (Fig. 1).
Samples were collected from three successive generations in a single breeding population (Fig. 1): the F0 generation, collected near the dam construction site and used for breeding at the facility; the F1 generation, produced by breeding between F0 males and females at the facility; and the F2 generation, produced by breeding between F1 individuals under natural conditions at the translocated site. F0 rearing and breeding at the facility were conducted in multiple tanks containing multiple F0 males and females or one F0 male and one F0 female each, with each tank used for a single breeding season. For this study, all F0 individuals used for breeding in each tank will be referred to as a “family”. Some of the F1 individuals born in each family were released to the transfer sites in the spring and fall, resulting in F1 individuals from 12 families having been released by spring 2019 (families A1, A2, A3, A4, A5, A6A7A9, A10, A11, A12, A15, A16 and A17). Table S1 contains a summary of information about the F0 individuals in each family.
The reproductive contribution of individual families was estimated by determining the number of F2 individuals derived from each family. The simplest way to achieve this is to check the parent–offspring relationships between F1 and F2 generations. However, as previously stated, the F1 individuals that were released were not sampled for genetic analysis. Therefore, it was imperative to investigate GG relationships between F0 and F2generations, as well as UANN relationships between siblings of the released F1 individuals and wild-bred F2individuals (Fig. 1).
The selection criteria for individuals used in determining relationships were as follows. First, if samples from all F0individuals from a particular family were available, all F0 individuals from that family were included for analysis. This applied to families A2, A3, A6A7A9, and A17. Second, if a sample from at least one F0 individual in a family was not obtainable, the F1 individuals (the siblings of the released F1 individuals) from that family were analyzed instead. This approach was applied to families A1, A4, A5, A12, A15, and A16. For families where a single male and female pair were used for breeding, a single F1 individual was selected, given that all F1 individuals within the family were FS. For families where breeding involved multiple F0individuals, two F1 individuals were analyzed. As for the F2 generation, individuals born in the NN6 pool, one of the release sites and the pool that showed the highest reproductive success after release, were selected for analysis. Samples from the F2 generation consisted of two year-groups: the 2018 year-group (F2 individuals born during the breeding season around June–July 2018, sampled in June 2019 at approximately 1 year old) and the 2019 year-group (F2 individuals born during the breeding season around June–July 2019, sampled in June 2020). A total of 70 individuals were chosen for analysis based on these criteria: 15 F0, 8 F1, and 47 F2 individuals (35 from the 2018 year-group and 12 from the 2019 year-group).