A comprehensive spatial decision model and its application of ecological compensation for agricultural land in reserve areas

Reserve areas are the key coverage areas of ecological protection compensation in China. In this paper, the generality and particularity of eco-compensation decision in reserve areas were analyzed, and a decision framework for eco-compensation was built considering the systematicness of decision and multi-dimensional effects. Based on the emphasis on expressing uncertainty in decision information, a decision model with "GIS (Geographic Information System)-BN (Bayesian Network)" as the core was constructed. The Maoli Lake Watershed, where the Hunan Maoli Lake National Wetland Park is located, was selected as a case area to design four agricultural land-use adjustment measures aimed at improving the water environment of the Maoli Lake, such as the reduction of chemical fertiliser application, and to form a number of eco-compensatory schemes to be selected. The decision tools were developed by self-programming to realize the fine decision of priority scheme and implementation area of agricultural land eco-compensation in the case area. The results showed that: (1) Ecological compensation for agricultural land in nature reserve areas had a multidimensional effect, and its decision-making should seek synergies between the conservation needs of the reserve and regional economic and social stability and development. (2) Ecological compensation for agricultural land comprised multiple key components, such as land use adjustment measures, target land types, compensation standards, and implementation areas. These elements could be combined in various ways to form comprehensive decision-making plans. This multi-component approach offerd a more systematic and holistic framework compared to single-factor decision-making models. (3) The priority programme for ecological compensation of agricultural land in the case area was the implementation of crop rotation and fallow for paddy fields and drylands, with compensation rates at the second level; regional preferences at the village level of the administrative districts could be implemented more efficiently. (4) The decision model in this paper could entirely transfer uncertainty information from the microscale of plot to the mesoscale of watershed, and the assessment results could simultaneously suggest the potential benefits and possible risks of the decision.