Population viability analysis in urban wildlife management: modelling management options for Sydney's quarantined bandicoots

Managers of urban wildlife must make transparent, quantitative decisions about environmental impacts but are challenged by the complexity of these impacts, which can interact with environmental variability to cause long-term changes. Here I use population viability analysis (PVA) to exemplify its potential in urban wildlife management, modelling the endangered population of long-nosed bandicoots at North Head. An “off the shelf” software package, VORTEX, was used to model data on population size, animal life span, reproduction rates, litter size, and mortality rates to simulate the population's dynamics over time, taking into account fluctuations in these parameters due to environmental conditions. Various management options for the long-term recovery and potential impacts of urban development were then modelled by varying adult mortality and carrying capacity.

The bandicoot population was highly sensitive to increases in adult mortality. Under the basic model, with a 30% chance of fox arrival each year and the carrying capacity of the headland at 120 bandicoots, the population had a 10% chance of going extinct within 20 years due to chance events. When adult mortality was increased to 11%, 14% and 16% the probability of extinction by 2020 increased to 15%, 24% and 32% respectively. If the carrying capacity of the headland is increased (to 200 animals), elevated adult mortality had a much lower impact on the chances of extinction (6%, 9% and 15% for the 3 levels of mortality). However, if development reduces the carrying capacity (to 75 individuals on the headland), elevated adult mortality leads to increased chances of extinction in 20 years to 31%, 42% and 46% for the 3 levels. Hence the model showed that for bandicoots, managing the byproducts of urbanisation, such as road kills or predation by pets is more immediately important than small changes in habitat availability, but the effects are additive. More generally, the exercise showed how PVA can be used to avoid subjective and short sighted assessment of management decisions to provide a quantitative comparison of a complex network of management options.