Sensitivity of the seabird risk assessment to three years without captures

The ‘’National Plan of Action — 2013 to reduce the incidental catch of seabirds in New Zealand fisheries’’ (NPOA) sets out a five-year plan for reducing seabird mortalities in commercial fisheries, to be achieved by 30 June 2018. Over this period, the NPOA requires that ‘’species currently categorised as at ’very high risk’ or ‘high risk’ from fishing move to a lower category of risk’’. The risk ranking of seabird taxa within New Zealand’s Exclusive Economic Zone was estimated in a recent assessment of the risk of commercial fisheries to seabirds. Here, we used simulated data to test the sensitivity of the risk assessment to a reduction in captures. In particular, we simulated an additional three years of fishing effort and observer data, identical to the most recent three years of fishing, but with no observed seabird captures in any fishery.

We successfully fitted a model to these data, allowing for a step change in vulnerability. The estimated risk ranking over the three simulated years was predicted to be ‘’negligible risk’’ for most species. Nevertheless, the risk category of four species was still estimated to be non-negligible, including black petrel, Salvin’s albatross, flesh-footed shearwater, and southern Buller’s albatross. Black petrel was still estimated to be at ‘’very high risk’’, despite the lack of captures. This outcome was caused by a lack of observer coverage in small-vessel bottom-longline and inshore trawl fisheries. The model indicated that a large number of seabird fatalities was still statistically possible, given the overlap between these fisheries and the four species, even though no captures were reported.

This result demonstrates that with current levels of observer coverage, it will not be possible to meet the goals of the NPOA, and reduce the risk category of the seabird species estimated to be most at risk. Even if there are no captures over three years, a reduction in the estimated risk will not be achieved without increasing observer coverage. We recommend that a more detailed simulation study is carried out to demonstrate the observer coverage required to satisfy the NPOA goals, for a range of reductions in capture rates.