Population estimates and projections of Maui dolphin (Cephalorhyncus hectori maui) based on genotype capture-recapture, with implications for management of mortality risk
Justin G. Cooke , Debbie Steel, Rebecca Hamner, Rochelle Constantine and C. Scott Baker
International Whaling Commission
The Māui dolphin (Cephalorhyncus hectori maui) is endemic to the west coast of the North Island of New Zealand and has been subject to incidental catches in fishing operations. An individual-based, stage-structured population model was fit to individual genetic identification data collected during 2001-16. A number of variants of the population model were fitted which all yielded essentially the same estimates of recent population size and trend over the period 2001-16, but the variant with sex-specific survival rates and no individual heterogeneity in capture probability provided the best fit. The population is estimated to have declined over this period with approximately 95% probability. The best fitting model showed no evidence of a change in mortality rate over the period 2001-16, but the fit to the data was only slightly poorer (ΔAIC ~ +1.5) when the mortality rate was assumed to have been reduced by 50% after 2008. The population size has become so low (median estimate less than 60 animals aged 1+ in 2016) that the predicted population trajectories are noticeably influenced by random individual birth and death events.
This model was projected forward to 2050 for a range of different assumed values for the intrinsic rate of increase (r0) and the carrying capacity (K). The projections show that in order to place the population on a recovery trajectory with 95% confidence, the anthropogenic mortality hazard (annual risk of a human caused mortality per individual) would need to be reduced to 20% or less of the average 2001-16 level. If the anthropogenic mortality hazard has already been reduced relative to the average 2001-16 level by fishing restrictions, then the further reduction in hazard that is required is proportionally less.