Assessment of the risk of commercial fisheries to New Zealand seabirds, 2006–07 to 2014–15

Seabirds are incidentally captured during commercial fishing operations, but assessing the impact of fishing mortalities on seabird populations is hampered by a lack of information. Seabird bycatch is not fully known, and understanding of seabird populations is limited. To manage the effects of fisheries on seabird populations, risk-based approaches are used to assess the population impacts of commercial fisheries on seabirds. In New Zealand, assessments of the risk to seabird populations from fisheries bycatch have been based on a comparison of the ratio between estimates of incidental captures and estimates of seabird population productivity, following the Spatially Explicit Fisheries Risk Assessment framework. This report presents an update of the previous assessment of the risk of commercial fisheries in New Zealand, for 71 seabird taxa that breed in the New Zealand region.

In this risk assessment, estimates of incidental captures were derived from the captures of seabirds recorded by government observers onboard fishing vessels and from data on fishing effort between 2006–07 and 2014–15, and the risk was calculated using annual average fishing effort for the period from 2012–13 to 2014–15. A Population Sustainability Threshold (PST) was used for seabird population productivity, a generalisation of the Potential Biological Removal (PBR) index, based on the total number of breeding pairs, and including the uncertainty in all demographic parameters explicitly.

A total of 16 species were estimated to have a risk that was non-negligible. Black petrel (Procellaria parkinsoni) was identified to be the species most at risk from commercial fisheries, and the only species in the “very high risk” category. There were seven species in the next highest estimated risk ranking, “high risk”, including (in decreasing order of median risk ratio) Salvin’s albatross (Thalassarche salvini), flesh-footed shearwater (Puffinus carneipes), southern Buller’s albatross (Thalassarche bulleri bulleri), Westland petrel (Procellaria westlandica), Gibson’s albatross (Diomedea antipodensis gibsoni), New Zealand white-capped albatross (Thalassarche cauta steadi), and Chatham Island albatross (Thalassarche eremita). A further four taxa had an estimated risk ranking of “medium risk”, and four taxa were categorised to be at “low risk”. The risk category of the remaining 55 seabird populations was “negligible risk”.

The mean number of potential fatalities of the 71 taxa in New Zealand commercial trawl, bottomlongline, surface-longline, and set-net fisheries was estimated at 14 400 (95% c.i.: 11 900–17 500) birds annually. Most fatalities were estimated to occur in trawl fisheries, especially in the inshore fleet (although the estimate is highly uncertain). The estimate of annual potential fatalities includes cryptic mortalities: birds that are killed by the fishing activity but not brought on-board the fishing vessel or not included in captures reported by fisheries observers. The term “potential fatalities” is used to indicate the inherent uncertainty associated with estimating these cryptic fatalities.

Significant changes were made to the methodology to address limitations identified in previous risk assessments. Updates to the methodology included the use of a Population Sustainability Threshold (PST) for seabird population productivity, based on the total number of breeding pairs (rather than the lower quartile used previously in the PBR). This update included changes in the correction factors to meet the long-term goal of populations remaining above half their carrying capacity, in the presence of environmental variability. Other changes from the preceding risk assessment included the use of allometric modelling to reduce variability in the estimates of age at first reproduction and of adult survival. Both parameters were used in calculating the population growth rate under optimal conditions (r_max).

Updates from the preceding risk assessment also included the use of an integrated model for estimating fisheries mortalities, to prevent them from exceeding the total annual mortality of the adult population, and to ensure that estimated mortalities, seabird population size, and adult survival were mutually consistent. In addition, the proportion of captures released alive was estimated from the data, and half of number of fatalities from the number of observable captures, was disaggregated between fishery groups in trawl fisheries; vulnerability to capture was estimated in a single model across all fishing methods; for selected fisheries, the vulnerability was allowed to vary between the period before and after 2010; the recent split of Stewart Island shag (Leucocarbo chalconotus) into two separate species (Otago shag L. chalconotus and Foveaux shag L. stewarti) was incorporated in the current analysis; furthermore, for black petrel, an updated at-sea distribution was derived.

Applying these changes led to an overall decrease in the estimated risk for all taxa from the previous risk assessment, with most decreases resulting from the removal of the correction factor for the population growth rate rmax, and the use of the total population size in the PST calculation.

The current seabird risk assessment supports the “National Plan of Action – 2013 to reduce the incidental catch of seabirds in New Zealand fisheries”, which sets out the New Zealand government’s framework for reducing the impact of fishing on New Zealand seabird populations.