Potential Biases when Using Otolith-Fish Size as an Inference of Relative Growth Rates among Lake Michigan Lake Trout Populations

Quantitative measurements of fish growth require length at age data which is labor intensive to obtain and is subject to age measurement error (reader subjectivity when ageing scales, otoliths, etc.).  Otolith-fish size relationships may provide a comparison of relative growth rates among individuals or populations and may be useful in detecting spatial, temporal, or environmental changes in growth.  Nonproportional growth between fish otoliths and body size is a common phenomenon arising from: 1) age effects due to continual, albeit not necessarily linear, daily otolith deposition over a fish’s life and 2) growth-rate effects whereby faster growing fish have larger otoliths than slower growing fish of the same age.  In the absence of environmental, temporal, spatial, and phylogenetic differences it has been suggested that otolith-fish size measurements enable relative growth comparisons at an individual or population scale.  We analyzed the otolith-fish size relationships from samples of known-age lake trout recaptured in Lake Michigan fall surveys, 2006-2008.  We used generalized additive models with terms for age and growth-rate and test whether otolith mass varies by recapture year, sex, strain, and location.  Results demonstrate potential biases when using lakewide otolith-fish size relationships to compare relative growth rates between strains and locations.  We present the magnitude of these environmental and phylogenetic differences in otolith size at age and discuss sampling design considerations for otolith size based metrics in ageing applications including age-otolith mass regressions and length mediated mixture analysis.