Seasonal growth patterns of wild juvenile fish:: partitioning variation among explanatory variables, based on individual growth trajectories of Atlantic salmon (Salmo salar) parr

1. We present an empirical, analytical model that estimates both temperature and seasonal response functions for the growth of wild juvenile fish without the need for costly tank experiments in less realistic conditions. 2. Analysis of monthly recapture data on the lengths and weights of individual wild juvenile fish demonstrates that simple temperature-driven models of growth can be less informative than more realistic, empirical, models. 3. A case study of wild Atlantic salmon parr (Salmo salar) showed that: most growth took place in a 10-week period in spring, at temperatures below those that previous published models report as necessary for rapid growth and at faster rates than the maximum that previous models predicted. 4. Temperature and fish size allometry explained 45% of growth variation, but the effects of temperature were significantly and markedly different at different seasons. 5. Seasonal effects explained an additional 18% of the variation and were strongly associated with the abundance of potential 'drift' food. 6.The seasonal patterns for growth in length and weight were different, implying differential allocation of resources to structural and reserve tissues. 7.The growth patterns of sexually maturing male parr and emigrants also differed in comparison to other parr. 8.Condition factor, length at first capture and seasonal interactions with both early maturity and smolting explained another 7% of the variation. 9. However, individual fish did not grow consistently better, or worse, than the 'average' fish. 10. This study emphasizes the necessity to test the adequacy of laboratory-based physiological models with suitably detailed field data and to focus model refinement by identifying processes which otherwise confound interpretation.