Abstract
The proper target strengths to be used in establishing target-strength-to-fish-length regressions for use in interpreting acoustic measurements of fish abundance are shown to be the averages of target-strength functions with respect to the orientation distribution of the fish. Averaged target-strength regressions are computed for cod, Gadus morhua, at 38 kHz, for which measurements of dorsal aspect target-strength functions are available, for three kinds of behaviour as defined by observed and postulated tilt-angle distributions. Errors incurred in abundance estimation due to the misuse of target-strength regressions are computed and found to be significant. Empirical evidence for the behaviour dependence of echo energy is adduced through an analysis of observations of a non-monotonic dependence of echo energy on fish-number density for encaged aggregations of fish. The need for measurements of orientation distributions of fish whose abundance is to be determined acoustically is noted.

This publication has 0 references indexed in Scilit: