Comparison of fish density estimates from repeated hydroacoustic surveys on two Wyoming waters

The ability to actively sample fish populations is a major advantage of hydroacoustic assessment. This technique does not affect fish behavior, and it typically produces more precise abundance estimates than do other gears. Thus, hydroacoustic surveys repeated on a closed population should produce similar fish density estimates. We sought to demonstrate this on inland waters using multiplexed side- and down-looking transducers and target-tracking enumeration methods. To test the repeatability of hydroacoustically derived fish density estimates, we conducted complete hydroacoustic surveys daily over three consecutive days on two Wyoming reservoirs, with a parallel transect design at one reservoir and a zigzag design at the other; this mimicked two designs commonly used during routine hydroacoustic surveys. These repeated surveys were performed on each reservoir monthly from April to September 2001. We used a Kruskal-Wallis test to detect significant differences in fish density estimates among repeated surveys for each lake during each month. There were no significant differences among daily surveys when sampling conditions were constant. During periods of increased wind we increased the aiming-angle of the side-looking transducer to minimize wind noise interference at the lake surface. This produced significantly lower density estimates than on calm days. We concluded that sampling under marginal conditions affected the results of the survey and inhibited the comparability of samples over time.