Comparing echosounder efficiency using field observations

Abstract

Time series, using different echosounders or large-scale multiship acoustic surveys, can be criticized because equipment changes might affect the final results. This criticism was addressed previously by comparing the results from different vessels using echo integration on the target species. The acoustically estimated standing stock of Nile perch (Lates niloticus) in Lake Victoria, East Africa, declined to 50% between successive surveys six months apart in 2007, prompting the criticism that a change in echosounder was responsible for this observation. This concern has been addressed, using data from the same four small localities around the lake, Emin Pasha Gulf, Nyanza Gulf, Speke Gulf and the vicinity of the Sesse Islands, from six surveys, spanning the time when the change in echosounder occurred. For three of the locations, echo integration and single target detections within the first bottom echo indicated no significant differences in echosounder performance. Results from the fourth location, Sesse Islands, showed very low backscatter, possibly due to a layer of detritus on the lake bed. It is concluded that all data are equally comparable, providing echosounders are correctly calibrated with the vessel being stationary, although there may still be differences under operational conditions. Characteristics of intercalibration sites are discussed in this study. The results also show changes in substrate, likely attributable to local environmental changes between surveys.