Verification of Hydroacoustic Estimates of Fish Abundance in Ohio River Lock Chambers

Abstract

Hydroacoustic assessments of fish populations have become commonplace in the last 15 years. However, as use of hydroacoustics for stock assessment has become simpler, need for enumerative validation has increased, particularly for inland, multispecies systems. This study compared fish abundance estimated from 1-d rotenone surveys with estimates obtained from a 120-kHz Simrad split-beam acoustic system within lock chambers of the Ohio River. A total of five lock chambers were surveyed by both methods in fall 1997 and 1998. Abundance estimates ranged from 10,340–17,887 fish in the rotenone survey and from 11,543–14,962 fish in the hydroacoustic surveys. The two survey methods were highly and positively correlated (r = 0.938), and corrected lock rotenone estimates were within −12% to +20% of those derived from hydroacoustic surveys. Comparison of the size distributions showed significant differences in the size distributions of fish between methods, but peak modes for both methods occurred at the same range (75–99 mm). Differences in abundance estimates between the two methods were small and can be accounted for by differences in variability in 1-d capture efficiency of the rotenone method. Differences in the size distributions may be attributable to fish behavior (schooling) that may cause underrepresentation of smaller targets. Processing acoustic data on small spatial scales, as is possible in the Digital Echo Visualization and Information System (DEVIS) and the Simrad EP500 software (using layers), should result in better precision in hydroacoustic estimates in large rivers such as the Ohio River. The results of this study should provide further confidence to managers interested in acoustic stock assessment in large rivers.