A Bayesian Belief Network - Based approach to link ecosystem functions with rice provisioning ecosystem services

The complex interactions between environmental and anthropogenic components have significantly influenced rice cultivation. The clear understanding of these interactions is important to (i) optimize rice provisioning ecosystem service (ES) supply, (ii) minimize negative impacts on other ES and (iii) choose suitable strategies for sustainable agriculture. Impacts of environmental and anthropogenic components on rice provisioning ES supply largely depend on site selection and farming practices. The demand for rice can be determined by the size of the population and imports/exports of rice products. Rice provisioning ES supply and demand need to be balanced if the goal is an import-independent and sustainable agriculture. As a decision support tool, Bayesian Belief Networks (BBN) are used for quantifying various ES supply types, demands as well as their budgets. The BBN network presented in this study was developed through interviews, expert knowledge, geographical information systems and statistical models. The results show that the capacity of rice provision can be optimized through site selection and farming practice. The results can help to reduce crop failures and to choose suitable areas for the use of new practices and technologies. Moreover, the presented BBN has been used to forecast future patterns of rice provision through effective or ineffective options of the environmental and human-derived components in eight scenarios. Thereby, the BBN turns out to be a promising decision support tool for agricultural managers in predicting probabilities of success in scenarios of agricultural planning.