Patterns on a parr: Drivers of long‐term salmon parr length in U.K. and French rivers depend on geographical scale

• Published in: Freshwater Biology
• Main Authors: Gregory, Stephen D., Nevoux, Marie, Riley, William D., Beaumont, William R. C., Jeannot, Nicholas, Lauridsen, Rasmus B., Marchand, Frederic, Scott, Luke J., Roussel, Jean‐Marc
• Other Authors: Department for Environment, Food and Rural Affairs
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2017
• Subjects: Atlantic salmon; Salmo salar
• Online Access: http://dx.doi.org/10.1111/fwb.12929; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1111%2Ffwb.12929; https://onlinelibrary.wiley.com/doi/pdf/10.1111/fwb.12929

Abstract Understanding the geographical scales at which environmental variables affect an individual's body size, and thus their mortality risk, can inform management strategies to help conserve wild populations under climate change. Yet, our current understanding of these relationships is based on studies done at different scales that report inconsistent findings. We predicted that temperature‐related variables (e.g. winter temperature) influence body size at a “regional” scale, that is, affecting individuals in geographically independent catchments similarly, whereas non‐temperature‐related variables (e.g. conspecific competitor density) exert a “local” influence, that is, affecting individuals in geographically independent catchments differently. We developed statistical models to test our predictions using body length measures of a large and long‐term sample of juvenile Atlantic salmon ( Salmo salar ) from three rivers in the U.K. and France. We developed mixture models to predict the individual juvenile salmon ages objectively from their body length. We then developed linear mixed models to describe inter‐annual changes in mean length of the youngest (age 0) cohort of juvenile salmon from river‐specific seasonal variables, and tested whether they exerted their influence at a “local” or “regional” scale. All models accounted for spatio‐temporal differences in sampling protocols and individual reproductive strategy. We estimated and interpreted coefficients using Bayesian theory. Our findings supported our predictions. Juvenile salmon were longer in years of higher overwinter water temperature and this effect was best parameterised as a single “regional” coefficient applicable to all three rivers. Similarly, spring mean temperature was best parameterised with a single “regional” nonlinear coefficient. In contrast, juvenile salmon were shorter in years of high densities of competing conspecifics and their interaction with total mean discharge and these effects were represented by “local” river‐specific ...