Designing effective protection for New Zealand dolphins using an agent-based approach
Elisabeth Slooten, Joseph Cahill-Lane and Jacob Nabe-Nielsen
Hectorâ€™s and MÄui dolphins have been heavily impacted by mortality in fishing nets. Bycatch over the last four or five decades has resulted in dolphins declining to around 30% of original population size and MÄui dolphins to below 5% of their former numbers. Our analysis of the population effects of bycatch is based on knowledge of the spatial and temporal distribution of dolphins, gillnets and trawling vessels, and is validated using bycatch estimates from government observer programmes. Our main goal is to assess the effectiveness of protection measures introduced in 2020 that aim to reduce dolphin deaths in gillnet and trawl fisheries. These new regulations ban one or both of these fishing methods in part of the dolphinsâ€™ habitat. We model a time period of 30 years to evaluate how population size changes over this period, in response to current dolphin protection as compared to protection measures proposed by the IWC and IUCN. We developed an agent-based simulation model to study how population growth rates for the New Zealand dolphin were influenced by different levels of protection from bycatch in gillnets and trawls. The model uses a spatial structure that matches data on the distribution of dolphins and fisheries, and builds on the processes that influence population dynamics in nature. The model produces several patterns that emerge from the dolphinsâ€™ movement behaviour: The dolphinsâ€™ movements with regard to water depth and distance from shore drive their spatial distribution, replicating dolphin distribution data from surveys; their tendency to move towards trawlers influences the number of encounters they have with trawl nets, and therefore the risk of dying by drowning in a net. The probability of a dolphin dying when it is in the same location as a trawler or gillnet was estimated in the Banks Peninsula area (Fisheries areas 20 and 22). This is the only area with sufficient observer data to result in scientifically robust estimates of dolphin deaths in these two fishing gears. We estimate the capture probability for gillnets and trawlers within the model. That results in a total of 37 dolphins caught per year in this area, which is the official government estimate. We use this information to estimate the probability of capture when a dolphin or group of dolphins encounters a gillnet or trawler. We carry out sensitivity testing to estimate the effectiveness of current protection given government estimates of bycatch, as well as bycatch levels more consistent with recent reports from fishermen, independent observers and on-board video monitoring.