Dolphin bycatch in New Zealand trawl fisheries, 1995–96 to 2006–07


Thompson, F. N., & Abraham, E. R. (2009). Dolphin bycatch in New Zealand trawl fisheries, 1995–96 to 2006–07. New Zealand Aquatic Environment and Biodiversity Report No. 36. 24 p. Retrieved from


In the 12 years from 1 October 1995 to 30 September 2007, a total of 107 dolphin captures were reported by Ministry of Fisheries observers in trawl and longline fisheries within New Zealand’s Exclusive Economic Zone. These captures included 80 common dolphins, 9 pilot whales, and 7 dusky dolphins that were reported killed in trawl fisheries. In longline fisheries over the same period, eight dolphins were captured and released alive, with one common dolphin and two pilot whales reported killed. There have been few observations in inshore trawl fisheries, and no data were available from setnet fisheries. Hector’s dolphin captures have previously been reported from both setnet and inshore trawl fisheries, but the observer database does not contain enough information to estimate the captures of either Hector’s or Maui’s dolphins.

The jack mackerel trawl fishery, operating off the west coast of the North Island, was responsible for 91% of observed dolphin mortalities in trawl fisheries. Most of these mortalities were of common dolphins. Only 15 vessels participated in this fishery, with over 95% of the effort being carried out by just 7 vessels. A Bayesian model was developed to estimate dolphin captures in the jack mackerel fishery. This was the only fishery with sufficient observed captures to allow a reliable estimate of total captures to be made. There were 0.8 capture events per 100 observed tows in the 2006—07 fishing year. Dolphins are often caught in groups; the mean number killed in a single capture event was 2.5 dolphins. The model had two stages: first the probability that a dolphin capture event occurred on a tow was predicted, then if a capture event occurred, the number of dolphins killed was estimated. The model was fitted using Markov chain Monte Carlo techniques.

Four covariates explained variation in the probability of a capture event occurring during a trawl: headline depth, the depth of the top of the trawl net; trawl duration, time spent trawling, from shooting the net to hauling; light condition at haul time, which depended on the time of day and the illumination of the moon; and whether the tow was north or south of Mount Taranaki. Headline depth had the most explanatory power, with more dolphins being caught when the headline depth was shallower. The model estimated that deepening the headline depth of a tow by 21 metres would have halved the probability of a dolphin capture event.

Annual trawl effort in the jack mackerel fishery has increased dramatically since the late nineties, from 405 tows by three vessels in the 1995—96 fishing year, to 2164 tows by eight vessels in the 2006—07 fishing year. Observed dolphin captures have also increased over the period, with 11 dolphins caught during observed tows in this fishery in 2006—07.

The model estimated that there were fewer than 5 dolphin captures a year between 1995—96 and 1998—99. As effort in the fishery increased there was initially a large increase in the estimated number of dolphin mortalities, which peaked at 174 (95% c.i.: 74 to 366) dolphins in 2002—03. Since then the number of dolphins caught each year has decreased. In 2006—07 the model estimated that 52 dolphins were killed in the large vessel jack mackerel fishery (95% c.i.: 22 to 106). This decrease has occurred despite the number of tows in the fishery remaining steady since 2002—03, at over 2000 tows per year.

The abundance of common dolphin on the west coast of the North Island is unknown, and so there is no basis for estimating the impact of the fisheries mortalities on the local population.