374 North American Journal of Fisheries Management 25:374–390, 2005 [Article]  Copyright by the American Fisheries Society 2005 DOI: 10.1577/M04-018.1 Indicators of Energetic Status in Juvenile Coho Salmon and Chinook Salmon M. TRUDEL* Fisheries and Oceans Canada, Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, British Columbia V9T 6N7, Canada S. TUCKER Department of Biology, Dalhousie University, Halifax, Nova Scotia B3H 4J1, Canada J. F. T. MORRIS Fisheries and Oceans Canada, Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, British Columbia V9T 6N7, Canada D. A. HIGGS Fisheries and Oceans Canada, West Vancouver Laboratory, 4160 Marine Drive, West Vancouver, British Columbia V7V 1N6, Canada D. W. WELCH Fisheries and Oceans Canada, Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, British Columbia V9T 6N7, Canada Abstract.—Bioenergetic models frequently rely on published values or models for estimating the energy density of fish, principally because of the cost and effort of obtaining direct measurements. In this study, we developed empirical models of energy density for free-ranging juvenile coho salmon Oncorhynchus kisutch and Chinook salmon O. tshawytscha sampled at sea from the west coast of Oregon to Kodiak Island, Alaska, and we evaluated the accuracy of published energy density models commonly used for these species. Our analyses showed that the energy density of juvenile coho and Chinook salmon was strongly correlated to percent dry weight and proximate constituents (especially lipid and, to a lesser extent, protein concentrations) but poorly correlated to body size and condition factor. Percent dry weight of whole fish was the single best predictor of energy density for both species, accounting for more than 90% of the variance in energy density. We also found that percent dry weight in the muscle tissue accounted for 65% of the variance in energy density. Changes in energy density mainly reflected changes in lipid composition. These results indicate that accurate estimates of energy density could be obtained at low effort and cost for juvenile coho and Chinook salmon simply by determining the water contents in whole-fish or muscle samples. Published models overestimate the energy density of juvenile coho and Chinook salmon collected from the Pacific Ocean. This may result from the extrapolation of the models to different size-classes, life stages, or habitats. More caution is needed when models are extrapolated to conditions beyond those that were used for their development. Bioenergetic models have been used to study numerous processes, including fish growth (Hew- ett and Kraft 1993), the accumulation of contam- inants (Trudel and Rasmussen 2001), nutrient cy- cling (Kraft 1992), food web dynamics (Rand and Stewart 1998), fish migration (Stockwell and John- son 1999), and habitat selection (Brandt and Kirsch 1993), as well as for contrasting manage- ment strategies (Negus 1995). Because energy is * Corresponding author: trudelm@pac.dfo-mpo.gc.ca Received February 7, 2004; accepted July 19, 2004 Published online March 7, 2005 the currency being quantified in bioenergetic mod- els, accurate estimates of energy densities (i.e., energy contents per unit weight) are required in these models (Hartman and Brandt 1995). Instead of using direct measurements obtained using bomb calorimetry, bioenergetic models fre- quently rely on (1) published values, (2) proximate constituents, or (3) models for estimating the en- ergy density, possibly because of the cost and time required to process a large number of samples and the difficulty of obtaining homogenates of large fish for whole-body measurements. The first ap- proach usually assumes that the average value de-