Influence of Sampling Location, Timing, and Hatching Origin on the Prediction of Energy Density in Juvenile Pink Salmon

摘要

Accurate estimation of energy density of fish is important for biogenetic models. Our objectives for this study were to determine which variables could be used to predict energy density instead of estimating energy density directly with bomb calorimetry. Secondly, we examined the variability in energy density relative to the sampling location within the Gulf of Alaska, the stock of origin, and the year the fish was sampled. Juvenile pink salmon Oncorhynchus gorbuscha were collected from the Gulf of Alaska during July 2001 and 2002. Energy density (J/g of wet weight) was estimated using bomb calorimetry. Hatchery stocks were identified from otolith thermal marks, and non-thermally marked fish were assumed to be wild. Energy density differed significantly by transect (P < 0.000), year (P < 0.000) hatchery stock (P = 0.001), and the interaction of origin and transect (P = 0.018). Body size was not related to energy density. However, % dry weight (dry weight/wet weight) was related to energy density (R2 = 0.93) and thus can be used in regressions to estimate energy density. We used energy densities predicted from a regression with % dry weight in bioenergetic modeling simulations. Error associated with energy density predictions affected bioenergetic models of body growth by up to 7-8% over a 30-day period. This error increased as the water content of fish increased and as the energy density decreased. Biological factors should be considered when predicting energy densities so that errors are minimized.