Biases in pushnet and pullnet zooplankton samples were investigated by obtaining paired samples sequentially with the 2 types of net, alternating the order of net type to control for the effect of order. There were no significant differences between pushnet and pullnet samples in total numbers of zooplankton, dry weight, or ash-free dry weight. Daytime pullnet samples contained significantly greater numbers of relatively passive drifters (fish eggs, invertebrate eggs, and siphonophores). There were no such differences among samples taken on moonless nights. Copepods (Acartia spp.) were more abundant in pushnet than in pullnet samples in all 32 palrs of daylight samples, but not at night. Therefore, the bias for motde zooplankton probably resulted from avoidance influenced by light. The greater abundance of floating eggs in pullnet samples may have been a result of mixing the surface water w t h water 60cm below by turbulence created by the boat hull and the propeller There were generally no significant interactions in results between net type and date or location, therefore either pushnet or pullnet could be used in studies to test hypotheses because statistical analysis requires a standardized sampling technique withn the program and because w~thout interaction, any bias should be consistent across the sampling program. However, slnce all methods for sampling zooplankton are probably biased, a variety of sampling methods should be used in combination for estimating parameters.
Read full abstract