Developing population models for guiding reintroductions of extirpated bird species back to the New Zealand mainland
- Wildlife Ecology Group, Institute of Natural Resources, Massey University, Private Bag 11 222, Palmerston North, New Zealand
- Present address: Department of Biology, Duke University, Box 90338, Durham, North Carolina 27708-0338, U.S.A.
Population models are useful tools to guide management as they allow us to project growth and persistence of wildlife populations under different scenarios. Nevertheless, good data are needed to produce reliable models, and this requirement is problematic in some situations. North Island saddlebacks (Philesturnus rufusater) were reintroduced to Boundary Stream Mainland Island in September 2004, and this was the first time this species had occurred in an unfenced mainland area since their extirpation in the 19th century. This situation creates a challenging scenario for population modelling, as this species has never been studied in the presence of mainland predators, and management of these predators will be the key factor determining whether the population survives. In this paper we present an approach for developing a “prior model” before a reintroduction takes place. We use data from the reintroduced saddleback population on Mokoia Island to develop a model of how saddleback populations are regulated in the absence of mammalian predators. We use this model to project growth of a reintroduced population when vital rates are reduced by predation and also to project responses of source populations to harvesting of birds for translocation. We then incorporate data from the reintroduced North Island robin (Petroica longipes) population at Paengaroa Mainland Island to model the relationship between population parameters and predator tracking rates. The combined model can be used to predict the level of predator control needed to ensure growth of the saddleback population, but the prediction is contingent on guessing the relative vulnerability of robins and saddlebacks to predation. We envision using a Bayesian approach to update such prior models as site-specific data become available after reintroduction.