The prediction model of tourism impact assessment in vegetation canopy cover of Qhamishloo National park and Wildlife Refuge

The protected areas are managed by ecological targets including ecological and biological protection of nature and at the same time the goal of tourism use. The management of natural ecosystems in protected areas and where tourism creates many environmental impacts is a global challenging issue in the field of protection. The aim of this study is to model the vegetation canopy cover reduction in Qhamishloo national park and wildlife refuge in order to evaluate the impact of tourism using artificial neural network to determine the influence of the ecological variables and severity of tourism on protected areas. This study has been performed in Qhamishloo national park and wildlife refuge with an area of 10 hectares intensive tourism zone and 100 hectares extensive tourism zone. In this study to evaluate the impact of tourism on vegetation canopy cover reduction, 100 inventory sample plots were used to measure ecological and tourism variables and vegetation canopy cover density changes along one year (spring 2017-2018). Artificial neural network modeling method has been used to predict the reduction of vegetation canopy cover density using 11 environmental variables. According to the results, the model with the structure of 1-8-11 (11 input variables, 8 neurons in hidden layer and 1 output variable) with regard to the maximum coefficient of determination in the three categories of training, validation and test data set which equal 0.95, 0.87, and 0.93 respectively, declare the best function of structural optimization. On this basis, the intensity of tourism, land slope, soil salinity, soil depth and percentage of organic matter of soil with the coefficient of determination of 8.59, 2.02, 1.88, 1.81 and 1.65 respectively show the highest effect on the vegetation canopy cover reduction in the region. The proposed model in this study provides a decision support system in tourism impacts assessment in the protected areas and enables prediction of the vegetation canopy cover reduction in recreational zones of national parks.