Abstract:
Variation in Catch Per Unit Effort (CPUE) on a temporal scale have thus far been largely attributed to variation in recruitment and annual abundance in the New Zealand scampi fishery. We looked to explain some of this variability using a historical fishing pressure variable derived from the function nn2 in the RANN library. It was found that in all of the main fishing grounds some of the yearly variation in catch rate could be explained by the historical fishing pressure variable. Some of the yearly variation was still evident in the fishery and this was looked into by examining the residual plots from each ground. Smaller sub-sets from each ground were looked at to see if fishing that exhibited a different distribution in yearly occurrences produced similar output. This was found to be true, and in some cases were found to be modelled better than the ground as a whole. In the second part of this research we looked to examine the differences in distance between consecutive tows. Four variables were used to explain these differences, and these variables fitted to a classification tree. Three years were examined from each fishing ground and similar trends were seen throughout the fishery as a whole. Decisions to move were largely based upon current catch weights and these were proportional to the CPUE for that year. It was also found that had a tow occurred at night a vessel was more likely to stay in place and fish throughout the day, presumably because day time fishing on average produced a better catch rate. One ground was not modelled well by the classification tree approach which we surmised to be due to the spatial complexity of the ground which became evident from the fishing pressure analysis.
