Comparing results of black petrel capture interactions with bottom longlines using different data collection methods

The effect of additional observations of non-target captures of black petrels (Procellaria parkinsoni) by electronic monitoring (EM) and observers was evaluated for commercial small-vessel bottom longline (BLL) fisheries. Moreover, the influence of observer and/or camera presence on fisher’s compliance to report protected species captures (specifically black petrels) was assessed. The results showed that estimated black petrel captures were lower when fitting models to observer data and EM data combined (i.e., increasing the proportion of monitored fishing events in the data) compared with model fits against observer data alone. Simulating data with different proportions of assessed video footage revealed strong biases of estimated black petrel captures (against the ‘true’ number of simulated captures) for scenarios that are comparable with the proportion of assessed video footage in the actual data that are used to estimate captures. Hence, current bycatch models seem to overpredict black petrel captures in small-vessel BLL fisheries in FMA 1. However, comparing fisher-reported black petrel captures (i.e., those being independently reported from observers or camera footage if present) with model estimates suggested reasonable alignment between both, even when only cameras were present and video footage was not assessed. For fishing events that had neither an observer on-board nor cameras present, the fisher-reported captures were below capture estimates, but as aforementioned these also seem to be overpredicted. Therefore, the results highlight two main benefits from EM, which are the ability to increase the proportion of monitored fishing events (ideally close to 100%), thus reducing the need for estimating captures or at least reduce bias, and the need for more accurate reporting of captured birds when cameras are present.