Modeling pulse dynamics of juvenile fish enables the short-term forecasting of population dynamics in Japanese pufferfish: a latent variable approach and hindcasting

Short-term forecasting of population dynamics fish stocks is required for scientific management recommendations such as catch quotas but difficult to perform accurately because the inevitable time lag between data collection and implementation. Here, we developed a method using small-scale survey juvenile Japanese pufferfish ( Takifugu rubripes) shorten this achieve accurate short-term forecasting. A at local sandy beach provides strength year classes before fisheries recruitment; however, use raw owing small sample size large observation errors. We found that random-effects model overcame these problems more predicted pulse patterns rates derive standardized recruitment index than fixed-effects model. stock assessment outperformed models without with respect hindcasting bias prediction skill. This study highlights applicability latent variable approach standardizing thereby unbiased dynamics.