Integrative measures of consumption rates in salmon: expansion and application of a trace element approach

1. Establishing reliable estimates of consumption is necessary for understanding the physiology, bioenergetics and trophic relationships of organisms. For fish, the inability to measure consumption directly prevents a mechanistic understanding of habitat– foraging relationships. Building upon established models for 137 caesium (Cs) mass balance in fish, we used natural abundances of a stable geologically derived isotope of Cs to estimate consumption rates over the first growing season for Atlantic salmon Salmo salar and to derive a general model that provides integrative estimates of consumption rates for individuals of all sizes. 2. To test the reliability of the trace metal approach we (i) performed a sensitivity analysis of model parameters and (ii) parameterized the model with site-specific data, including gut contents, Cs concentrations of invertebrate prey and assimilation rates. 3. We applied the method in two sites to make the first in situ determinations of consumption rates of individual age-0 salmon at post-larval and fry stages, for fish as small as 0·1 g. Consumption estimates were most responsive to changes in three parameters: Cs body burdens, Cs concentration in prey items and assimilation efficiency, all of which could be measured with high precision using inductively coupled plasma mass spectrometry. 4. The assimilation efficiency of Cs measured on field-caught age-0 salmon was approximately 60%. Consumption rates at 2 weeks post-stocking were highly variable in both sites, ranging from no detectable consumption to 8·5% fresh weight (fw) day − 1 . By the end of the growing season, consumption rates were less variable (2–4% fw day − 1 ). 5. Synthesis and applications. This study is the first to demonstrate that background levels of geologically derived Cs can be used to estimate consumption rates of fish. Our results show that extremely low consumption rates during the first 6 weeks of life correspond closely with the critical survival period in other fish populations, and suggest a mechanism for a hypothesized survival bottleneck at this time. The results implicate the importance of early season habitat availability when considering management priorities of fish. Additionally, our use of stable Cs at natural abundance concentrations permits the global application of this trace element approach for estimating consumption rates of fish as well as other organisms. This general approach can be adopted for conservation and management settings in which it is necessary to identify suitable foraging habitat of a species or to quantify the relationship between consumption and growth.