Abstract: A systematic approach for prioritization of protected areas is the use of artificial intelligence. This approach employs computer algorithms based on an objective function to identify the best network of areas to be protected. Site selection algorithms are commonly used to identify areas of high conservation value. This study used three types of heuristic algorithms (simulated annealing, greedy, rarity) to prioritize areas for protection in Mazandaran Province of Iran using Marxan software. The goal was to select a conservation network with the smallest possible area in which maximum protection targets are achievable. The effects of spatial scale, algorithm, and zone compactness were also examined. We found that the existing network of protected areas is inadequate to achieve conservation targets. The algorithm results provided the best areas for supplementation of the current network. The simulated annealing algorithm provided the most plausible results for all scenarios. These results can be used to modify the existing boundaries of the protected areas network and introduce new sites for protection of plant and animal species.

Keywords: Optimization algorithms, Simulated annealing, Greedy, Rarity, Marxan, Systematic conservation planning