Simple haul mitigation devices were trialled on two pelagic longline vessels and one demersal longliner. During at-sea observations it was apparent that birds consistently followed the vessel using different circular flight patterns depending on the wind direction relative to the vessel. This influenced how easily they could access the area beside the hauling station and what proportion of their time was available for searching for baits. During one pelagic longline trip birds were observed selectively taking sanma baits, in preference to squid, from branchlines in front of the vessel.
Due to low capture and direct interaction rates, it was necessary to use bird attendance in the area around the longline as a proxy for risk. Model results showed that mitigation devices reduced the number of birds moving into the area immediately around the hauling station. On the demersal longliner, retrieving surface floats also reduced bird attendance beside the hauling station.
Data collected in real time allowed for investigation of the influence of additional variables on the numbers of birds moving into the area beside the hauling station. For the models fitted to pelagic longline data, and both pelagic and demersal longline data, higher proportions of squid bait reduced the number of birds entering the area beside the hauling station. The model fitted to demersal longline data showed that higher wind speeds increased the number of birds entering the area beside the hauling station.
Although not selected in the final models, observations of bird behaviour during the haul indicated that wind strength and direction relative to the vessel influenced the ease with which birds could access baited hooks. Exploring these relationships statistically would require larger real time data sets. However, we note that plausible effects of wind direction were apparent in some of the models fitted during the variable-selection process.
The use of EM data allowed for generation of a longer-term data set, and it is recommended that the mitigation employed by the vessel should be routinely recorded when collecting data from video footage, to allow for analysis across larger data sets.
This work shows that simple and cheap hauling mitigation devices can reduce risk to birds during longline hauling with minimal impact on fishing operations. It is recommended that all longline vessels are encouraged to and supported to develop hauling mitigation.