Comparing Electronic Monitoring and human observer collected fishery data in the tropical tuna purse seine operating in the Western and Central Pacific Ocean

Electronic Monitory (EM) systems have been proven a valid tool for collecting fishery dependent data. They are being widely used in many fisheries as a complement or alternative to human observers to increase the monitoring coverage of fisheries. However, considering its wide application, following agreed minimum standard, it is important to compare the congruence between the information collected by EM and observers. We compared EM and two sets of different observer data collected on 6 trips of tuna purse seiners in the Eastern and Western and Central Pacific Ocean to analyze the similarity of fishing set type identification, estimation of tuna and bycatch catches between both monitoring systems. Overall EM was a valid tool to estimate the type of fishing set. Retained total catch of tunas by set was estimated by EM as reliable as that by both observer programs and logbook. When comparing the information by set, EM estimation of the main species, such as skipjack and bigeye and the combination of bigeye/yellowfin, was proven to be less accurate but statistically similar to the estimates made by both observers’ programs. EM tended to underestimate the retained catch of skipjack in comparison to both observers estimates and slightly overestimate bigeye and yellowfin, the overestimation being less pronounced for bigeye than for yellowfin. For bycatch species, EM is able to identify main bycatch species as observers do. However, the capability of EM to estimate the same number of bycatch items in comparison to IATTC and WCPFC observers varies greatly by species group. For sharks, which are the main bycatch issue in the FAD purse seine fishery, the overall congruence between EM and observers was high. EM and IATTC observer identified a similar overall number of individual sharks, however, WCPFC observers estimated lower number of shark individuals than the other two monitoring systems when considering all trips together.