Determining sampling date interval for precise in situ estimates of cumulative food consumption by fishes

Abstract

We tested the influence of sampling date interval (SDI) on precision of in situ estimates of cumulative food consumption by fishes. Daily rations of stream-dwelling green sunfish (Lepomis cyanellus) and impoundment-dwelling bluegill (Lepomis macrochirus) were estimated for 30 consecutive days using a low-effort procedure. Cumulative consumption by each species over the 30-day period (and 95% CIs) was estimated using Monte Carlo simulations. The effect of SDI on cumulative consumption estimates was examined by calculating cumulative consumption for SDIs of 1, 2, 3, 4, 5, 6, 7, 10, 14, and 30 days; the 1-day SDI served as a standard for evaluation of other SDIs. Cumulative consumption estimates began to fall outside the 95% CI for the 1-day SDI at SDIs of 3-4 days and did so with with increasing frequency as SDI increased. Error in estimating cumulative consumption was almost always [Formula: see text]15% relative to the 1-day SDI standard at SDIs of 5 days or less but was as high as 26 and 39% at SDIs of 6 and 7 days, respectively. Our results suggest that sampling at least every 5 days may be required to obtain precise estimates of cumulative consumption by fishes in lotic systems and small impoundments.