Recent U.S. experience with electronic monitoring, seabird monitoring, and incorporation into standard management protocols
Electronic monitoring (EM) projects underway in Alaskan fisheries include development of protocols or systems which improve species identification, development of additional system automation, and maintaining or improving compliance with seabird bycatch mitigation measure regulations. One method that is being trialed in a longline fisheries to identify the species of incidentally caught birds involves vessel crew voluntarily holding the bycaught bird up to a camera for several seconds. To date, there is only one instance of this request having been carried out, which suggests that other procedures should be developed in consultation with vessel crew and enforcement personnel. Alternatively, a technological approach could be used for species identification. Progress has been made on automated image processing algorithms for species identification and length measurement, through the use of training datasets of fish images collected from the fishery. Currently, these machine learning algorithms can identify 43 groundfish species with 94% accuracy and collect length measurement that is within 1 cm of actual length. This functionality is being extended to seabird species. The use of EM to monitor compliance with deployment of streamer lines is also being studied, on large and small vessels. The initial results indicate a lower success rate (92%) when monitoring smaller vessels.