Toward adaptive management of parera (Anas superciliosa) and mallard (A. platyrhynchos) duck in New Zealand : a thesis presented in fulfilment of the requirements for the degree of Masters of Science in Conservation Biology at Massey University, Manawatu, New Zealand
Wildlife exploitation is encumbered with uncertainty. To ensure sustainability of
wildlife populations managers must understand the consequences of, and account for,
uncertainty in their decisions. This is most pertinent if the goal is to optimise or
maximise the harvest or take.
Uncertainty can be separated into four main categories: environmental variation,
partial management control, structural uncertainty (e.g., density dependence) and
partial observability. This thesis examines the first three categories in the context of
mallard (Anas platyrhynchos) and parera (grey duck, A. superciliosa) harvest in New
Zealand, and specifically addresses sustainable and maximum annual mallard harvest.
A simple heuristic harvest model is proposed to represent a population subject to a
seasonal annual harvest. The heuristic model is then converted into a series of
quantitative models that can be used to predict the effect of regulations on hunter
behaviour (partial management control). Specifically, how regulations may affect
hunter effort (hours hunted) and the consequences of hunter effort on, harvest rates,
survival, and productivity. Survival and productivity were further evaluated as a
function of post-harvest population size (structural uncertainty). Harvest rate,
survival, and productivity data were derived from 22,500 (1,024 recaptures; 3100
recoveries) mallard and parera banded from 1997 to 2009 in the Eastern and Hawke’s
Bay Fish and Game Regions and a telemetry study of 46 mallard in the Eastern
Region. Harvest data and reporting rate estimates were derived from a randomised
hunter survey over the study period.
In the Eastern Region hunter effort explained changes in survival better than any of
the other candidate models ( = 0.851 i w ). In the Hawke’s Bay changes in survival was
explained by changes in season length ( = 0.334 i w ), hunter effort ( D = 0.739 c QAIC ;
= 0.231 i w ), and spring temperature in the year of banding (SpcT) ( D = 0.1.53 c QAIC ;
= 0.155 i w ). Correlation of harvest rates and effort approached significance
(P=0.053) in the Eastern Region for adults only while in the Hawke’s Bay data there
was no relationship. This was assumed a consequence of reporting rate confounding
harvest rate estimates as correlation between hunter effort and harvest was good in
both Eastern (R=0.85, t(10)=5.3193, P<0.001) and Hawke’s Bay (R=0.76, t(8) = 3.3878,
P = 0.0095).
A deterministic model was developed (from the quantitative models), to maximise
annual harvest subject to the criteria that harvest should not compromise the ability to
maximise the following season’s harvest. The performance of the quantitative models
was validated using a partially stochastic model to simulate harvest. Harvest
simulations were used to predict 2010 (outside of the study period) harvest (41,549
mallard and parera; SE=3,552) in the Eastern Fish and Game Region. Simulations
predicted harvest accurately (42,045; SE=1,992). Simulations indicated that mallard
harvest was not sustainable over a 10 year period when juvenile female: adult female
ratios £ 0.8 when constrained by Eastern Regions regulation set (season length 30 to
71 days). When productivity increased ( ³ 0.95 juvenile female: adult female) long
term harvest was viable under the most relaxed season constraint (71 days). This has
important implications when managing breeding habitat.
It was proposed that managing populations within similar climate zones would reduce
environmental uncertainty. Survival of mallard and parera were analyzed using a set
of linear climate covariate models fitted to data from 91,500 mallard and parera
banded throughout New Zealand (1969–2009). Climate explained changes in survival
better than or was comparable to the alternate candidate models in 11 of 17 data sets.
The quantitative models in this thesis provide a platform for Fish and Game managers
to initiate an adaptive management approach to mallard and parera harvest
management in New Zealand.
Should Fish and Game wish to review current mallard and parera management areas,
establishing management units on homogenous climate zones would contribute to
creating a good management system.