How sieve mesh size affects sample estimates of estuarine benthic macrofauna

The way mesh size influences the reliability of density and biomass estimates in macrofauna sampling was investigated for a benthic community in the shallow, upper reaches of a small, temperate estuary. Macrofauna was separated from the sediment with sieves of 0.25, 0.5 and 1.0 mm mesh aperture size and the performance of each gear type in terms of retention efficiency evaluated. The 0.25 mm sieve sampled all macrofauna adequately, but only 55% of all individuals were, on average, retained by the 0.5 mm mesh and a mere 8% by the 1.0 mm mesh. Mean retention efficiency for total biomass was 86% for the 0.5 mm and 49% for the 1.0 mm screen, respectively. Undersampling by bigger mesh sizes was most severe for juveniles and small species, resulting in biased interpretation of apparent community structure. Because population densities obtained by the two bigger screens were not only markedly less accurate (i.e., biased lower counts), but also less precise (i.e., higher standard errors), the use of 0.5 mm and 1.0 mm mesh sizes significantly increased replication levels. Consequently, for this specific habitat and community, sampling schemes which employ a 1.0 mm mesh were considerably more expensive than designs using finer screens. No single sampling design will pel-form optimally in each permutation of habitat and community, but our data emphasise the need for the stringent assessment of sampling errors and caution against the uncritical use of mesh sizes greater than 0.25 mm in estimating population densities of estuarine macrobenthos.