Warp strike in New Zealand trawl fisheries, 2004–05 to 2006–07

Edward R Abraham & Finlay N Thompson


Abraham, E. R., & Thompson, F. N. (2009). Warp strike in New Zealand trawl fisheries, 2004–05 to 2006–07. New Zealand Aquatic Environment and Biodiversity Report No. 33.


Since the 2004–05 fishing year, Ministry of Fisheries observers have been carrying out warp strike observations, recording the numbers of birds that are struck by trawl warps during 15 minute observation periods. In 2004–05, the observations were aimed at exploring the relation between offal discharge and warp strikes. In 2005–06, monitoring of mitigation devices was introduced and the observations include those made as part of an experiment exploring the efficacy of warp-strike mitigation devices. In 2006– 07, regulations requiring the use of warp mitigation devices were introduced, and the observations are from ongoing monitoring. In this report, a summary is presented of all the strike data collected from 2004–05 to 2006–07. The analysis provides an overview of the strike data and an indication of the main trends. During this period a total of 12 097 strike observations were made. Most observations (58.8%) were made in the squid trawl fishery, with some also being carried out in hoki (16.9%) and other trawl fisheries.

The average warp strike rate has decreased since 2004–05, when warp strike observations were first made in New Zealand fisheries. For small birds (all birds other than albatrosses and giant petrels), the average warp strike rate has decreased from 3.03 to 0.55 strikes per hour. For large birds (albatrosses and giant petrels), the average warp strike rate has decreased from 2.35 to 0.39 strikes per hour. The decrease is associated with an increase in the use of mitigation devices, which were made mandatory in January 2006 for all trawlers over 28 m in length fishing in New Zealand waters. In 2004–05 mitigation was not compulsory and was used during 48% of the observations. In 2006–07, mitigation was used during 92% of observations. The most frequently used mitigation devices in 2006–07 were tori lines, which were used during 56% of observations. Statistical modelling confirms that the strike rate of both large and small birds is reduced when tori lines are used as a mitigation device. Other devices such as warp scarers and bird bafflers are not found to significantly reduce the warp strike rate.

The warp strike dataset provides strong evidence that the increased use of tori lines has led to a reduction in the number of albatross mortalities. From 2004–05 to 2006–07, there was a decrease in large bird warp capture rate from 6.7 to 0.4 birds per 100 tows. Capture rates of small birds on the warps are too low to infer a trend in the rate. Statistical modelling of large bird warp captures shows that they decrease when tori lines are used, and increase when there is continuous discharge. While the warp capture rate of large birds has decreased, there has been no consistent change in the net captures.

The analysis also shows the importance of offal management, with few warp strikes or captures being recorded in the absence of offal discharge. Across all the data the average large bird warp strike rate is 0.02 strikes per hour when there is no discharge, compared with 3.22 strikes per hour when discharge is continuous. The large bird warp capture rate when no discharge was recorded during the warp observations was 0.2 birds per 100 tows, compared with an average capture rate of over 7 birds per 100 tows when offal or discards were discharged.

For every large bird captured on the warps, there are 208 (95% c.i.: 150 to 290) warp strikes, and for every small bird captured on the warps there are 7610 (95% c.i. 3800 to 36 000) warp strikes. It is likely that birds are killed by warp strike but not brought on board, and so not counted as captures. Currently, estimates of seabird mortality in New Zealand fisheries are based solely on landed captures and the true mortality from trawl fishing is likely to be underestimated. An estimate of the ratio of warp strikes to warp mortality is needed to allow total mortality to be assessed.