Modeling the Affects of Climate on Individual Growth Variability of Marine Fish: A Bayesian Hierarchical Approach

Annual growth-increments in hard structures of marine organisms provide an integrated measure of an animal’s growth rate over its life span that when related to environmental variability reveal evidence for biophysical coupling.  The functional response between climate variability and animal growth is strengthened when such a response is seen across diverse species under the influence of physical processes in a given ecosystem.  In this paper, we developed an integrated approach to analyzing the relationship between growth increment data and climate using Hierarchical Bayesian methods.  By taking advantage of nested data structure, we fit a hierarchical Bayesian model to growth increment data from numerous individuals over different species in the North Pacific Ocean and test for different covariance structures among random individual effects, long- and short-term environmental effects and residual error.  Sea surface temperature (SST) and the pacific decadal oscillation (PDO) entered the model as covariates explaining significant variability in growth across both bivalves and fish.  Once the effects of age on growth-increment data were removed, growth variability correlated positively with SST on an inter-annual basis while PDO accounted for longer term growth rate trends that were particularly evident between the 1940-1950 warm-cold and 1970-1985 cold-warm regime transitions.  Accounting for PDO and SST greatly improved model skill in reconstructing climate conditions prior to the instrumental record.  Our results indicate that biophysical coupling likely occurs at several temporal scales.