Gear bias in field estimation of the amount of food consumed by fish

We made concurrent collections of small (85–100) yellow perch (Perca flavescens) at the bottom of Lake Erie with an active gear (otter trawl) and a passive gear (gillnet) within a 24-h sampling period. During daylight periods when yellow perch feeding activity was most prominent, gillnetted fish possessed significantly higher median food amounts in stomachs than did trawled fish. Estimated daily food consumption for small yellow perch was 5.86% of body weight by gillnet sampling and 3.98% by trawl sampling. Passive gears tend to sample the most active fish (more likely to be feeding) and may, therefore, yield upwardly biased estimates of food consumption for individuals within a population. Active gears are considered to give more accurate estimates because quiescent and low activity, as well as the more active members of a population, are sampled.