TY - JOUR

T1 - A Bayesian Belief Network

T2 - Based approach to link ecosystem functions with rice provisioning ecosystem services

AU - Dang, Kinh Bac

AU - Windhorst, Wilhelm

AU - Burkhard, Benjamin Felix

AU - Müller, Felix

N1 - Funding Information: The study was embedded in the LEGATO (Land-use intensity and Ecological EnGineering – Assessment Tools for risks and Opportunities in annual crop based production systems) project, funded by the German Ministry of Research and Education within their funding measure Sustainable Land Management; Funding No. 01LL0917 . This study was co-financed by the Vietnamese Government Scholarship (911). Furthermore, we would like to thank all our LEGATO colleagues and especially Associate Professor Isoda Yuzuru from the Graduate School of Science at Tohoku University in Japan and all our colleagues in the Institute for Natural Resource Conservation, Department of Ecosystem Management at Kiel University , Germany. The authors would like to thank Mrs. Angie Faust for language corrections of the manuscript.

PY - 2019/5

Y1 - 2019/5

N2 - 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.

AB - 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.

KW - Agriculture

KW - Ecosystem service budget

KW - Ecosystem service demand

KW - Ecosystem service supply

KW - Scenario

KW - Socio-ecological system

UR - http://www.scopus.com/inward/record.url?scp=85046167933&partnerID=8YFLogxK

U2 - 10.1016/j.ecolind.2018.04.055

DO - 10.1016/j.ecolind.2018.04.055

M3 - Article

AN - SCOPUS:85046167933

VL - 100

SP - 30

EP - 44

JO - Ecological indicators

JF - Ecological indicators

SN - 1470-160X

ER -