Equivalence factors for standardizing catch data across multiple beach seine nets to account for differences in relative bias

Abstract

We describe a method for modelling the relative effects of seine net biases and for deriving equivalence factors to standardize fish abundance data sets collected using multiple sampling gears. Nearshore fish communities were sampled from 10 sites in each of the basin and riverine portions of the Swan–Canning Estuary, Western Australia, using beach seine nets of three different lengths (21.5, 41.5 and 133 m). The resulting data were subjected to generalized linear modelling to derive equivalence factors relating catches from the two larger net types to those from the 21.5 m net. Equivalence factors were derived on the basis of functional habitat guilds of fish (small benthic, small pelagic, demersal, pelagic). Prior to standardization, catches from the 41.5 and 133 m nets consistently underestimated fish densities relative to those from the 21.5 m net. Following standardization, the degree to which fish densities were underestimated by the two larger nets was reduced and/or eliminated for most guilds, and particularly in the case of the 133 m net. For both of the larger nets, standardized estimates of total fish density across all species were far closer to those recorded using the 21.5 m seine, thus indicating that standardization of the fish abundance data had greatly reduced the overall effects of the biases introduced by the different net types. This approach could be applied to other systems and sampling methods, to facilitate more robust comparisons of fish abundances between studies with divergent sampling methodologies.