Density-dependent population growth in a reintroduced population of North Island saddlebacks

1. Reintroductions provide a good opportunity to study density-dependent population growth, as populations can be studied at a range of densities and the change in density is not confounded with environmental conditions. An understanding of density dependence is also necessary to predict dynamics of reintroduced populations under different management regimens, and assess the extent to which they can be harvested for further reintroductions. 2. We monitored a North Island saddleback (Philesturnus rufusater) population for 6 years after reintroduction to Mokoia, a 135 ha island in New Zealand that was made suitable for saddlebacks by eradicating introduced Norway rats (Rattus norvegicus). We modelled adult and juvenile survival using Program MARK, and modelled numbers of young fledged per pair using Proc Mixed in SAS with individual female as a random factor. 3. Juvenile survival clearly declined as the population increased, and the decline was closely correlated with the number of breeding pairs. Reproduction also showed a clear decline that was explained by two factors: a difference in quality between territories occupied immediately after reintroduction and those occupied later, and an overall decline as the number of pairs increased. Reproduction was also strongly affected by age, and this needed to be accounted for when modelling density dependence. 4. A stochastic simulation model incorporating these dynamics closely predicted the observed population growth. The equilibrium population size was insensitive to density dependence in reproduction, but highly sensitive to density dependence in juvenile survival. 5. The model is being used to plan management strategies for potential reintroductions of saddlebacks to mainland areas with predator control. The species is currently confined to predator-free islands and one fenced mainland sanctuary.