Abstract:
1. An explicit spatial understanding of population dynamics is often critical for effective management of wild populations.
Sophisticated approaches are available to simulate these dynamics, but are largely either spatially homogeneous or agentbased,
and thus best suited to small spatial or temporal scales. These approaches also often ignore financial decisions crucial to
choosing management approaches on the basis of cost-effectiveness.
2. We created a user-friendly and flexible modelling framework for simulating these population issues at large spatial scales –
the Spatial Population Abundance Dynamics Engine (SPADE). SPADE is based on the STAR model (McMahon et al. 2010) and
uses a reaction-diffusion approach to model population trajectories and a cost-benefit analysis technique to calculate optimal
management strategies over long periods and across broad spatial scales. It expands on STAR by incorporating species
interactions and multiple concurrent management strategies, and by allowing full user control of functional forms and
parameters.
3. We used SPADE to simulate the eradication of feral domestic cats Felis catus on sub-Antarctic Marion Island (Bester et al.
2002) and compared modelled outputs to observed data. The parameters of the best-fitting model reflected the conditions of
the management programme, and the model successfully simulated the observed movement of the cat population to the
southern and eastern portion of the island under hunting pressure. We further demonstrated that none of the management
strategies would likely have been successful within a reasonable timeframe if performed in isolation. 4. SPADE is applicable to a wide range of population management problems, and allows easy generation, modification and
analysis of management scenarios. It is a useful tool for the planning, evaluation and optimisation of the management of wild
populations, and can be used without specialised training.
Description:
Appendix S1. SPADE manual.
Appendix S2. Details of algorithms used in SPADE.
Appendix S3. Details of statistical models.
Appendix S4. Source code for SPADE package.
Appendix S5. Description of potential issues in using STAR.