Sampling characteristics and biases of enclosure traps for sampling fishes in estuaries
We evaluated the sampling characteristics of enclosure traps in estuaries in southern California, USA. Using enclosure traps that sampled 0.25, 0.5, and 1-m2 footprints, we found that enclosure trap size significantly affected estimates of fish density and the precision of these estimates. The highest estimates were produced by the 0.5-m2 trap and the lowest by the 0.25-m2 trap. Precision of the density estimates improved with increasing trap size, while the proportion of zero values in the data sets decreased and estimates of species richness increased. The largest trap was difficult to use in the field and often did not function properly; thus we concluded that intermediate enclosure trap sizes offered the best compromise between statistical and logistical considerations. By examination of burrows in sediment cores taken in fished out enclosure traps, we found no evidence to support the widely held view that burrow-dwelling fishes evaded capture by hiding in burrows. We also used mark-recapture techniques to estimate recovery efficiency in 0.43-m2 enclosure traps. Recovery efficiency averaged 91% and did not differ significantly among estuaries or sampling stations within estuaries. Based on extensive netting within enclosure traps, we determined that in areas with dense fish populations (>90 fish 0.43-m?2), netting could be ceased after the first sweep that captured no fish with only a trivial effect on the estimate of density. In more sparsely populated areas, netting had to continue until 2-3 sweeps had captured no fish in order to obtain estimates of density that were within 90% of the actual values present. Overall, we found enclosure traps to be effective tools for sampling small, abundant fishes in shallow estuaries in southern California, but we recommend that care be taken when choosing trap size and sampling (netting) effort within traps in order to optimize their sampling characteristics.