Pohnpei, Federated States of Micronesia Analysis of North Pacific Shark Data from Japanese Commercial Longline and Research/Training Vessel Records WCPFC-SC7-2011/EB-WP-02

This paper analyses North Pacific longline operational data from research and training vessel surveys (1992-2008) and commercial longline logbook records (1993-2008) provided by Japan for onsite analysis in Shimizu during January-March 2011. Both data sets required filtering to remove records believed to under-report actual shark catches. The analysis was based on 7,974 sets representing 10 vessels in the research and training vessel surveys and 88,129 sets representing 112 vessels in the commercial longline fleet. Application of different filtering methods could result in larger sample sizes, but this benefit would need to be weighed against the probability of increasing the presence of under-reported catches in the filtered database. When considering the selection and application of data filters it is important to recall that if vessels began releasing/discarding (and not reporting) sharks in recent years, filtering may not fully correct for this effect, and declining catch rate trends would thus potentially be exaggerated. On the other hand, if reporting practices do not change but shark stock abundance actually does diminish over time, declining catch rates would be expected. The challenge is to apply a filter which removes those catch records which are under-reported, but retains those which are low but accurate