The Impact of Shifting Sampling Frames On Relative Abundance Indices: Lessons Learned for the Japanese Longline Fishery

Bias can be introduced into relative abundance time series calculated from catch and effort data when shifts in the spatial and temporal sampling frame are not accounted for. This bias can be severe when multiple species with non-overlapping distributions are intercepted and the catch and effort information is used to develop individual or aggregated abundance trends. The Japanese longline fleet expanded spatially and temporally across the world’s oceans over the first 20 years after the final MacArthur line was removed following World War II. A distorted picture of the impact these industrial fisheries had is the result of failing to account for a shifting sampling frame.  When developing relative abundance trends from populations sampled in such a manner it is essential to impute catch rate values in those areas not sampled. When such calculations are performed for tuna and billfish species captured in the Japanese longline a different picture on stock status is apparent.