The stock size of sandfish in the northern Sea of Japan was estimated by a virtual population analysis (VPA) and sensitivity analyses were attempted on the VPA estimate. The stock size estimates were approximately 600-900 million until 1975, but since 1976 they have rapidly decreased. In the sensitivity analyses, the estimates of absolute stock size were not sensitive against the changes in the fishing mortality coefficient for terminal age and the measurement error in catch-at-age. This suggested that the relative stock size remains almost unaffected by the error in the data used in the VPA, if the degree of catch-at-age error and the natural mortality coefficient is correct. The relationships between the biomass estimated by the VPA and the density index from Danish seine fisheries, and between the biomass and the catch per unit effort (CPUE) from the experimental survey using Danish seine nets, were also examined. The density index and the CPUE indicated significant relations with the biomass. Consequently, the CPUE is useful to monitor the relative stock size in a timely manner, and the VPA estimate and the CPUE should be utilized for adjusting the total allowable catch in the multiseasons.

No abstract

No abstract

Three different techniques for stock identification, namely scale patterns, otolith thermal marks, and single-nucleotide polymorphism (SNP) markers, were applied to the same mixture samples of chum salmon collected in a setnet fishery in the Sea of Okhotsk, off Hokkaido, Japan, to evaluate consistency and accuracy among the techniques. The scale pattern and SNP analyses provided similar estimates for two regional origins (Hokkaido and Honshu) with data for 2011-2013; for each year, the Hokkaido-origin stock was dominant in September and early October, but the Honshu-origin stock was dominant by late October. The SNP analysis specified five geographical origins of chum salmon (the Okhotsk coast, Hokkaido and Honshu coasts of the Sea of Japan, and Hokkaido and Honshu coasts of the Pacific), whereas the other techniques did not identify multi-regional origins. Recaptures of otolith-marked chum salmon were used for estimating the proportions of fish originating from the Hokkaido Sea of Japan region; these estimates differed from those based on the SNP analysis. This study demonstrated that scale pattern and SNP analyses have the potential to provide nearly consistent estimates for the two regional origins, and can possibly help us to understand the interception of migrating chum salmon in Japan.

No abstract