Use of a multiplicative model to estimate relative abundance from commercial cpue data

Abstract

A multiplicative model is developed for the CPUE of sole by UK beam trawlers fishing in the western English Channel between 1972 and 1982. The model includes a factor for vessel/vessel group as a means of adjusting for fishing power effects. Area, season, vessel and year are shown to have a significant effect on CPUE. Interactions between these main effects are investigated for a smaller model with grouped levels of each factor. Significant interactions are identified between the effects of year, season, vessel group and, to a lesser extent, area, but the variation in CPUE explained by each effect is small in comparison to the variation explained by main effects. Estimates of annual CPUE extracted from a main effects model (i.e. assuming no interactions) are shown to have a broadly similar trend to that observed in CPUE indices calculated using traditional rectangle and month averaging procedures. Some of the more obvious biases found in traditional indices are avoided. Approximately 60% of the total observed variation in CPUE was not explained by the model, and it is suggested that analysis of CPUE data disaggregated by age may be appropriate. © 1992 International Council for the Exploration of the Sea.