Using digital video monitoring systems in fisheries: application for monitoring compliance of seabird avoidance devices and seabird mortality in Pacific halibut longline fisheries
Ames, RT.; Williams, GH.; Fitzgerald, SM.
NOAA Technical Memorandum NMFS-AFSC 152 i-xii, 1-93
2005
Accession: 024092158
The incidental take of seabirds, including rare takes of the endangered short-tailed albatross (Phoebastria albatrus), is known to occur in the Alaskan longline fishing fleet. Current fishery regulations do not require observer coverage in the Pacific halibut fishery unless a vessel is greater than 60 feet length overall (LOA) and participating in other federally managed fisheries. The lack of at-sea observations has resulted in little information on the amount of seabird bycatch and on the compliance with seabird avoidance measures within the halibut fishery. The National Marine Fisheries Service (NMFS) contracted with the International Pacific Halibut Commission (IPHC) on a project to examine the feasibility of electronic monitoring systems (EMS) in the Pacific halibut longline fleet operating off the state of Alaska. The project was conducted on two of the IPHC's chartered stock assessment survey vessels fishing in Alaska during 2002. The objectives of the project were to 1) examine the ability of an electronic monitoring system to provide images that would allow an analyst to monitor seabird avoidance devices for regulatory compliance; 2) determine the feasibility of using video images for detecting and identifying incidentally caught seabirds; and 3) discuss options for the future use of electronic monitoring as a fishery management tool. To determine if an electronic monitoring system (EMS) could be used to assist in compliance determination, a system was placed on the vessels to record images of halibut gear being set and the performance of seabird avoidance devices, or streamer lines, during the setting. Vessel and video observations were compared on 106 setting events. The EMS video observations proved to be successful in detecting streamer line deployment and relative position on 100% of the daytime sets when two cameras were used. The results of the streamer line performance evaluations suggest that accurate performance recognition was positively related to the increase in image recording speed and the video analysts' ability to distinguish measured interval markings that were attached to the streamer lines. The ability of a video analyst to recognize and identify retrieved seabirds was examined by intentionally setting previously caught frozen seabirds on the fishing gear. No birds were caught incidentally during this study. Using 63 seabird specimens, the results showed a positive relationship between correct seabird species identification and EMS recording frame rates. At a fast recording speed 91% of the deliberately set seabirds were correctly identified as seabirds and 64% were identified accurately according to species. Nine of 12 albatross (Diomedea spp.) specimens were correctly identified to species; 1 was determined to be an unidentified albatross, and 2 were incorrectly identified. During a second examination 14 members of the NMFS's North Pacific Groundfish Observer Program staff examined video images of six retrieved seabirds. The results indicated that correct seabird identification is related both to the analyst's knowledge of distinguishing species characteristics, and to the size of the seabird. A third independent evaluation, where an analyst had no advance knowledge of birds being retrieved on the gear, showed that an analyst was capable of detecting 96% of the seabirds deliberately set with the gear, and that 79% of the specimens were correctly identified to species. estimated at $2.7 million. Coverage levels of 100% for vessels greater than 125 feet LOA, 30% vessels for 60-124 feet LOA, and no coverage of vessels less than 60 feet LOA was estimated at $0.41 million for an on-board observer program, whereas the cost of an EMS was estimated at $0.22 million. Electronic monitoring costs could be reduced if fewer hauls were sampled on each vessel than currently realized. In conclusion, this study suggests that an EMS monitoring program would produce accurate data and enable compliance evaluations for seabird avoidance devices. In addition, an EMS program would be able to detect a high proportion of incidentally caught seabirds. However, additional work is needed on seabird image identification and verification methods and testing the effects of soak time on the physical characteristics of seabirds. The potential costs of two monitoring programs were estimated at two levels of coverage for the halibut fishery off Alaska. The cost of complete monitoring of all setting and haul backs was estimated at $8.5 million for an on-board observer program; the cost of an EMS was estimated at $2.7 million. Coverage levels of 100% for vessels greater than 125 feet LOA, 30% vessels for 60-124 feet LOA, and no coverage of vessels less than 60 feet LOA was estimated at $0.41 million for an on-board observer program, whereas the cost of an EMS was estimated at $0.22 million. Electronic monitoring costs could be reduced if fewer hauls were sampled on each vessel than currently r ealized. In conclusion, this study suggests that an EMS monitoring program would produce accurate data and enable compliance evaluations for seabird avoidance devices. In addition, an EMS program would be able to detect a high proportion of incidentally caught seabirds. However, additional work is needed on seabird image identification and verification methods and testing the effects of soak time on the physical characteristics of seabirds.