A Genetic Algorithm Solver for Pest Management Control in Island Systems

摘要

Island conservation management is a truly multidisciplinary problem that requires considerable knowledge of the characteristics of the ecosystem, species and their interactions. Nevertheless, this can be translated into an optimisation problem. Essentially, within a limited budget, a manager needs to select the conservation actions according to expected payoffs (in terms of protecting or restoring desired species) versus cost (the amount of resources/money) required for the actions. This paper presents the problem in terms of a knapsack formulation and develops optimisation techniques to solve it. From this, decision-support software is being developed, tailored to meet the needs of pest control on islands for conservation managers. The solver uses a Genetic Algorithm and incorporates a simplified model of the problem. The solver derives strategies that reduce the number of threats, allowing the preservation of desired species. However, the problem model needs further refinement to derive truly realistic options for conservation managers.