Abstract
We evaluated the use of total body electrical conductivity (TOBEC) for determination of whole-body water content of yellow perch, Perca flavescens, and alewife, Alosa pseudoharengus. We used multiple linear regression with backwards stepwise elimination to test the capability of TOBEC values, wet weight, and total length in predicting whole-body water content of yellow perch and alewife. We found that wet weight was the best predictor of whole-body water content. The inclusion of TOBEC values in multiple linear regressions did not improve the predictive capability of wet weight over simple linear regressions that used wet weight alone (r(2) increased by only 0.00002 to 0.0005). We reanalyzed the data from three previous studies that used TOBEC to evaluate the tissue composition of fish. Again we found that the inclusion of TOBEC values in regression functions with wet weight as the other independent variable did not substantially improve the predictive capability of functions that used wet weight alone (r(2) increased by only 0.00003 to 0.0007).