Do diel changes in patchiness exhibited by schooling fishes influence the precision of estimates of fish abundance obtained from hydroacoustic transect surveys?

Abstract

Rainbow smelt (Osmerus mordax) were studied using hydroacoustic methods (echo integration and dual-beam) to determine whether fish patchiness resulting from diel vertical migrations and schooling behavior cause lowered precision in estimates of fish abundance when assessed with hydroacoustic technique. A parallel transect design with randomly selected transect locations was applied. Cluster sampling analysis was used to assess sample error. Smelt were dispersed at night and patchily distributed during the day. The observed changes in fish distribution had no influence on the precision of abundance estimates. However, when compared seasonally, sample errors were found to decrease in association with an increase in fish abundance and in relation to a decrease in the large-scale heterogeneity of fish.