Prediction of the consumption rates for juvenile Atlantic salmon (Salmo salar L.) on the scale of habitat patches: Development of empirical models based on stomach contents

Rivers and streams are unstable environments in which estimation of energetic costs and benefits of habitat utilization are the daunting exercise. Empirical models of food consumption may be used to estimate energetic benefits based on abiotic and biotic conditions in patches of habitat. We performed thirty daily surveys of fish stomach contents to estimate the consumption rates for juvenile Atlantic salmon (Salmo salar) in a river. The data were used to assess whether variations of daily consumption rates existed within the river, and to develop empirical models that could predict fish consumption rates using abiotic and biotic conditions as independent variables. Daily consumption rates based on stomach content surveys in the field (range: 0.15-1.49 g dry/(100 g wet day)) varied significantly depending on habitat patch (500-1000 m2), summer period, and sampling year. Variables such as water temperature, numerical density of salmon, water depth and moon phase explained 83-93% of the variations in daily food consumption rates. Daily consumption rates tended to increase with water temperature and depth, and were also higher near a full moon. However, they tended to decrease with the numerical density of salmon. Our work suggests that empirical models based on independent variables that are relatively simple to estimate in the field may be developed to predict fish consumption rates in different habitat patches in a river.