The role of electronic monitoring in assessing post-release mortality of protected species in pelagic longline fisheries

Protected species bycatch occurs in longline fisheries worldwide with accurate documentation limited to trips with a fishery observer onboard. Electronic monitoring (EM) provides an additional data stream to detect these protected species interactions and to assess the likelihood of an animal’s survival after their incidental capture and release from fishing gear. In this study, protected species interactions were reviewed from EM video collected from the Hawaii longline fisheries to evaluate if determinations of mortality or injury severity (serious or non-serious) could be made for cetaceans, and if the percent likelihood of post-interaction mortality could be assigned for sea turtles. Cetacean injury determinations were made based on criteria developed by the National Marine Fisheries Service (NMFS 2012; NMFS 2023), with our study focusing on the location and amount of attached fishing gear at capture and at release and on the health condition of the animal. The percent likelihood of mortality for sea turtles was assigned based on criteria such as: 1) whether the sea turtle was a leatherback (Dermochelys coriacea) or hardshell, 2) its injury category derived from its hooking/entanglement location, and 3) the amount of fishing gear remaining at release, per Ryder et al. (2006, Table 1). When uncertainty existed in protected species data, then in some cases an injury determination could not be made for cetacean interactions or a potentially higher percent likelihood of mortality was assigned for sea turtles. During this study, we reviewed video footage of eight small cetaceans and 37 sea turtle interactions incidentally captured on Hawaii-based longline fishing trips. Our study demonstrated that injury determinations could be made for most cetacean interactions, with the reviewer able to see the general location of the attached fishing gear, if the animal was alive or dead at release, and if the amount of fishing line at release was enough to result in a serious injury. However, in two cases where cameras were dirty or had poor views (e.g., limited view of the water out from the vessel), no determination could be made as the amount of trailing line at release was unknown. For sea turtles, EM video could be used to estimate the percent likelihood of mortality (per Ryder et al. 2006) with high certainty for most sea turtles caught. Improvements during the study to camera resolution allowed reviewers to generally discern the hooking/entanglement location and to determine the release condition, which was based on the amount of trailing gear at release (per Ryder et al. 2006). However, some uncertainty occurred in cases where the sea turtle was hooked around the mouth or released with attached fishing gear. We provide recommendations and specifications for EM cameras and handling requirements that will enable reviewers to assess the likely post release mortality. EM cameras should have a resolution of at least 4 megapixels and, at a minimum one camera per vessel should have a view that extends out from the vessel over the water while a separate camera view should cover the entire deck. In addition, recording at a higher frame rate of 30 frames per second (fps) may increase the probability of seeing a hook to discern the specific attachment location and to distinguish dark-colored protected species, such as leatherback sea turtles and false killer whales, (Pseudorca crassidens) from the dark-colored ocean at night. Fisher handling guidelines are needed to remove fishing gear and to coil remaining fishing line after removal from the animal within the view of the EM cameras to improve estimation of the amount of fishing line that remains attached after protected species are released. Implementation of EM technology can provide a better understanding of the impacts of longline fisheries on protected species stocks worldwide and can lead to better management of fisheries.