Climate-driven fitness decline in Japanese Chum Salmon reshapes North Pacific salmon biogeography, with reproductive consequences: A Bayesian network analysis

Abstract

Japanese Chum Salmon (JCS) supported by one of the world largest hatchery programs have experienced severe declines in abundance, return rate, and egg size. To investigate the underlying mechanisms, we analyzed a 21-year time series of reproductive traits of age-4 hatchery fish from 13 rivers together with climate and salmon abundance data using a bootstrap-supported Bayesian network. Here, we assumed that environmental variables can affect the salmon populations, but not vice versa, and that there could be maternal effect on reproductive traits, but not the other way around. The final network identified a major pathway from NPGO3 through JCS abundance to Russian Chum Salmon (RCS) abundance, followed by pathways from RCS to female fork length (FL) and from FL to egg size. A separate maternal pathway linked FL to egg size through fecundity. All corresponding edges had edge strength >0.97, and standardized path coefficients were statistically significant. These findings suggest that climate-driven fitness decline in JCS has reshaped salmon biogeography across the North Pacific, with consequences for reproductive traits, including a decline in egg size.