Invited Lecture - Abstract

Since the conservation of biodiversity became an issue of international significance, governments and managers of land have been interested in how to plan landscapes that will allow biodiversity to prosper. Problems of landscape planning for conservation range from methods of designing reserve systems to rules for reconstructing ecosystems.

I will discuss two general sorts of problem in landscape planning. First I introduce the problem of optimal reserve system design and I show how this raises a range of computational and communication challenges. This work is in collaboration with the Great Barrier Reef Marine Park Authority. Simulated Annealing is the primary optimisation tool we have used. Second I discuss a problem of optimal landscape reconstruction in time and space for an endangered Australian bird. We have used Stochastic Dynamic Programming to solve this problem. In both cases questions arise as to whether the computational difficulties are insignificant compared to the data uncertainty and the communication of results to decision-makers. The work has been done in collaboration with Michael Westphal (UC Berkeley) and Ian Ball (Australian Antarctic Division)