Variability in fish size/otolith radius relationships among populations of Chinook salmon

Abstract

Back-calculation of growth trajectories from otolith microstructure is a valuable tool for understanding mechanisms underlying variability in growth among fish populations. We analyzed fish length/otolith radius relationships for Snake River spring/summer Chinook and Snake River fall Chinook salmon, listed as separate “Evolutionarily Significant Units” (ESUs) under the US Endangered Species Act, to determine whether these ESUs shared relationships. In addition, we analyzed otoliths from seven separate populations within the Snake River spring/summer Chinook ESU to assess the variability in relationships among populations, which are much more closely related than ESUs. We also examined several potential functional forms for the equations. We found that the separate ESUs had significantly different fish length/otolith radius relationships, but that variability in otolith growth rate could not explain the difference. Relationships among populations within the spring/summer Chinook ESU did not vary nearly as much as those between ESUs. The quadratic model and the power model fit the data equally well, and constraining these models to pass through a biological intercept (estimated fish length and otolith radius at hatching) resulted in only a slight decrease in model fit. To test the ability of the models to back-calculate fish lengths, we predicted the length at tagging for 17 PIT-tagged fall Chinook that were measured at release and at recapture. The back-calculation demonstrated little bias (<1 mm FL, on average) and relatively small standard deviation (∼3.5 mm) for the best model. When we repeated the back-calculation with data from both ESUs combined, bias increased substantially (to 15 mm FL), demonstrating the importance of determining the proper taxonomic level at which to combine data within a species.