Optimizing the prediction of discard survival of bottom-trawled plaice based on vitality indicators

Description

Routine, at-sea observations of an organism’s vitality (responsiveness, vigor, and visible condition) are needed to address management policies related to post-release fate (i.e., discard survival), as well as growing welfare concerns. An animal’s involuntary motor responses to external stimuli (i.e., reflexes), external physical condition (i.e., injury) and vigor of movements (i.e., activity) upon capture can be rapidly collected after immediate capture impact. This has been a common approach to integrate the effects of multiple stressors a fish experiences during fishing. Using vitality scores or aggregated indices (e.g., a simple proportion: mean score of impaired reflexes and present injury or sum of impaired reflex and present injury scores divided by the total number of tested reflexes and injury types, R&I index) does not always allow for accurate predictions of post-release survival. One reason could be that due to collapsing individual reflex and injury scores into a single index, any differential contributions (or loadings) of individual reflexes/injuries to observed mortality events are not accounted for. To test whether some re-flexes and injuries may be more relevant over others in contributing to survival, in this study, the performance of suitable optimization functions was evaluated to optimize the loadings of individual reflex and injury attributes (either at individual or aggregated at trip level), and com-pared with conventional vitality proxy indicators (i.e., the original R&I index; number of absent reflexes; number of present injuries; number of absent reflexes and present injuries; categorical vitality index; reflexes and injuries as individual covariates ; and only the environmental variables). The proposed method was tested by using two datasets of bottom-trawl-caught plaice collected following a harmonized protocol in Belgium and Denmark. From 14 and 3 commercial fishing trips, and 51 and 18 gear deployments of Belgian beam- and Danish otter trawlers, respectively, in total, 736 and 386 undersized plaice (~ 23 cm TL) were assessed for vitality and delayed survival from the monitored trips in Belgium and Denmark, respectively. Alive individuals were sampled and monitored for their post-capture fate by being held in captivity until fishing-capture related mortality events were not observed for at least three days. It was difficult to predict either the average (per trip) survival probability of a group of fish (i.e., those discarded per trip) or that of individual fish without ancillary information about capture (environmental) conditions. Knowing what kind of gear was used and the temperature at the fishing depth was needed to improve predictions based on vitality information, and much more so than optimizing the loadings of individual reflex/injury attributes. Hemorrhages around the head region constituted the single most relevant predictor, whereas point bleeding injury assessments were redundant to score. A categorical vitality index provided a viable alternative to the more labour-intensive, scoring method of reflex responsiveness.

Period	Feb 2023
Event title	ICES-FAO Working group on fishing technology and fish behaviour (WGFTFB)
Event type	Conference
Location	Kochi, IndiaShow on map
Degree of Recognition	International