Fisheries, Subsistence, and Habitat
Publications Searchable Database
Search Again| Division: | Commercial Fish |
| Title: | An evaluation of the Bendix smolt counter used to estimate outmigrating sockeye salmon smolt in the Kvichak River, Alaska, and the development of a replacement sonar, 2000–2001 |
| Author: | Maxwell, S., A. Mueller, D. Degan, D. Crawford, L. McKinley, and N. Hughes. Maxwell, S., A. Mueller, D. Degan, D. Crawford, L. McKinley, and N. Hughes |
| Year: | 2009 |
| Report ID: | Alaska Department of Fish and Game, Fishery Manuscript No. 09-02, Anchorage |
| Abstract: | Smolt abundance estimates are useful in predicting adult salmon returns, setting escapement goals, and partitioning freshwater and ocean survival. Within the Bristol Bay area, several riverine sonar systems, Bendix smolt counters, have provided reasonable abundance estimates of sockeye salmon smolt for many years, but recently, the estimates appeared inflated. The Bendix system was initially validated, but not retested after substantive changes were made. We evaluated potential reasons for the failure, tested a new acoustic-video system, and compared estimates from the new and old systems. We combined a side-looking sonar to estimate smolt passage and up-looking video cameras to study behavior. The video data showed that smolt migrated primarily in the top 0.3 m of the river, their velocity was similar to current velocity, their body aspect varied considerably, and the majority migrated between late evening and early morning. The side-looking sonar defied a basic echo integration assumption, which assumes a uniform distribution, so it was necessary to modify the equivalent beam angle formula used in echo integration to account for the actual smolt distribution. We expected the inconsistencies in body aspect to reduce the smolts’ target strength; however, the acoustic scaler (-46.4± 2.3 dB) was similar to a scaler derived from video methods (-47.3 dB). Paired video and acoustic counts were strongly correlated, but linear regression results were not statistically equal to one (95% confidence interval 0.68–0.80). The acoustic-video abundance estimate was 15.3 M, considerably lower than the Bendix estimate of 325.9 M. The greatest differences occurred during daytime passage. Comparison and behavioral information from the video data gave us confidence in the side-looking acoustic estimates. However, processing the video data was time-intensive. A single up-looking sonar might efficiently provide similar information. We were unable to correct the historical Bendix estimates, but suggest studies that may provide a means. |
| Keywords: | Bendix, hydroacoustic, Kvichak River, Oncorhynchus nerka, salmon, smolt, sockeye salmon, sonar, split beam, underwater acoustics, underwater video |