Determining the ecological risk levels of Shadegan wetland based on probability method

Wetlands are very valuable and sensitive ecosystems that in recent years due to destructive activities have been faced a lot of risks. Determining the the levels of ecological risk for these dynamic ecosystems could provide a clear understanding from their current conditions. To determining the levels of ecological risk is needed a flexible tool about the kinds of data that can be corrected with updated information and change variables. On this basis, Bayesian belief network are suitable tool to determining risk levels and in this study was used to determining ecological risk levels of Shadegan Wetland. Initially determining of the variables and relations between them through expert judgment was done and influence diagram was formed then by determining the states and filling the probability distribution tables, Bayesian belief network was completed. In order to survey the sensitivity of model, "diagnostic model" and "predictive model" were used and sensitivity analysis was done in Netica software. According to the results in output node of Bayesian belief network, ecological risk levels for Sadegan Wetland are (low: 10, moderate: 17 and high: 73) so it is in danger of disappearing. Ultimately, strategies for risk control was proposed.