Energy Density and Dry Matter Content in Fish: New Observations and an Evaluation of Some Empirical Models. Issue 6 (2nd August 2017)

• Record Type: Journal Article
• Authors: Johnson, Brett M.; Pate, William M.; Hansen, Adam G.
• Is Part Of: Transactions of the American Fisheries Society. Volume 146:Issue 6(2017)
• Page Start: 1262
• Page End: 1278
• Publication Date: 2017-08-02
• Subjects: Fish culture -- Periodicals; Fishery management -- Periodicals; Pisciculture -- Périodiques; 639.2
• DOI: 10.1080/00028487.2017.1360392
• Languages: English
• ISSNs: 0002-8487

Abstract

Abstract: Energy density (ED) is an indicator of fish nutritional status, physiological status, and fitness. Estimates of ED of predators and prey are also needed for bioenergetics modeling, but direct measurements of ED are difficult to obtain. Hence, investigators often borrow published values from the same species or related species. Alternatively, models have been developed that predict ED from dry matter content (DM). The scarcity of published ED values makes data borrowing and the generality of predictive modeling difficult to evaluate. We report new ED measurements derived from bomb calorimetry for six freshwater fish species (Arctic Char Salvelinus alpinus, Brown Trout Salmo trutta, kokanee Oncorhynchus nerka, Lake Trout Salvelinus namaycush, Rainbow Trout O. mykiss, and White Sucker Catostomus commersonii ) and compared them to previously published data. We used our data to validate existing ED:DM models and to fit new ones. We also quantified bias that could result from borrowing inappropriate ED values for bioenergetics modeling. We collected a range of fish sizes from two reservoirs of differing productivity and measured whole‐body ED (wet‐mass basis) and DM. Our data substantially increase the range of ED data for Arctic Char, kokanee, and White Suckers. Two multispecies ED:DM models predicted the ED of our samples accurately (mean root mean square error [RMSE] < 500 J/g), even at the extremes for the range of prediction where data used for model development were limited. Taxon‐specific models performed less well (mean RMSE = 775 J/g), and some appeared highly biased (RMSE 956−1, 900 J/g). Bioenergetics model simulations showed that Lake Trout prey consumption could be overestimated by as much as 22% when using borrowed EDs for Lake Trout and their prey, but prey consumption estimates fell within ±2% of observed when ED was predicted from DM. When direct measures of ED are unattainable, measurement of DM and prediction of ED from one of the published multispecies models offer a practical and accurate method for bioenergetics modeling and other studies requiring information on fish energy content.