Download the publication
Demographic assessment
Summary
A demographic assessment was conducted of female New Zealand sea lions, aimed at identifying the proximate demographic causes of population change of New Zealand sea lion populations at the Auckland Islands, focussing on the two largest breeding colonies at Sandy Bay (Enderby Island) and Dundas.
State space demographic models fitted to mark-recapture, pup census and age distribution observations were developed using NIWA’s demographic modelling software SeaBird to estimate year-varying survival, probability of pupping and age-at-first-pupping.
For the Sandy Bay population, variation was observed in all demographic rate estimates when using the model configuration with lowest AIC (model run 8). Generally low pupping rates (including occasional years with very low estimates), a declining trend in cohort survival to age 2 since the early 1990s and relatively low adult survival (age 6-14) since 1999 may explain declining pup counts at Sandy Bay since the late 1990s. However, the lack of resighting effort prior to 1998 limits the assessment of demographic rates during the period of increasing pup production prior to 2000.
Similar time-trends in survival at age were obtained with respect to year for the Dundas population when adopting a similar model configuration to that used for Sandy Bay (model run 9).
The demographic rate estimates obtained for Sandy Bay were then used in a correlative analysis aimed at identifying the ultimate causes of population change, which accompanies this report (Roberts et al., 2014).
Correlative assessment
Summary
A correlative assessment was conducted with the aim of identifying the potential causes (eg, fishery-related mortality, climate, disease and others) of demographic variation and population change in New Zealand sea lions at the Auckland Islands.
Year-varying demographic rate estimates for females at Sandy Bay (Roberts et al., 2014) were related to a collated dataset of climatic, dietary, biological and fishery-related observations. Hypothetical biological and demographic responses to candidate drivers of population change were identified prior to the correlative assessment.
In most cases, the time series of available data were short and were mostly available for the period of population decline and this compromised the power of correlative assessments. Also, a large number of tests were performed and we would expect some tests to have been ‘significantly’ correlated by random chance.
A correlation with cohort survival to age 2 years was consistent with disease-related mortality affecting a decline in survival after 2005. Prior to 2005, pup mass at 3-weeks appeared to have been a good predictor of cohort survival to age 2.
Poor correlations were obtained when relating survival at ages 2-5 (juveniles) or age 6-14 (adults) to estimated captures and interactions in the Southern arrow squid trawl fishery at the Auckland Islands (SQU6T). However, a strong negative correlation was observed between survival at ages 6-14 (1999-2004) and cohort survival to age 2 in the previous year (1998-2003), which would be consistent with the high energetic costs of lactation affecting maternal survival during this time period.
Climate indices including Inter-decadal Pacific Oscillation (IPO) and sea surface height (SSH) were well-correlated with the occurrence of an array of key prey species in the diet, from an analysis of scats (Stewart-Sinclair, 2013). However, a longer time series of climate and diet data, with cyclic fluctuations, would be needed to establish a causative correlation with diet.
Variable diet composition, a decline in maternal condition, changes in milk quality and in pup mass, and depressed pupping rates, are all consistent with changes in nutritional status, though some of these responses could also occur in response to pup mortality that was not driven by nutritional stress.
The relationships and mechanisms identified should be considered as indicative only, though highlighted some new areas for thorough assessment. This should include a model-based assessment (as opposed to correlative) so that it will be possible estimate effect sizes, which will be necessary for explaining the causes of population decline.
Publication information
Roberts, J., Fu, D., Doonan, I., & Francis, C. 2014. New Zealand sea lion: demographic assessment of the causes of decline at the Auckland Islands. Demographic model options: demographic assessment. Report prepared by NIWA for the Department of Conservation, Wellington. 142p.
Roberts, J. & Doonan, I. 2014. NZ sea lion: demographic assessment of the causes of decline at the Auckland Islands. Demographic model options: correlative assessment. Report prepared by NIWA for the Department of Conservation, Wellington. 58p.