Conceptual and Bayesian networks: two new methodologies to analyze the provision of ecosystem services in agroecosystems

ROSITANO, F.; FERRARO, DIEGO OMAR

Laval University, Québec City, Canada

Congreso; ISEM 2009 Conference: Ecological Modelling for Enhanced Sustainability in Management; 2009

Agroecosystems, as well as undisturbed ecosystems, provide services to satisfy humanity. Ecosystem services from agriculture include provision of food, regulation of water and climate, cultural services, as well as enhanced supporting services (such as soil fertility). However, the ability and efficiency of agroecosystems to offer ecological services depend on how these ecosystems are managed: stress and degradation might reduce services. Various methods have been used to estimate the market value of ecosystem services. However; the ecosystem economic value is not always correlated with the environmental value or the long-term benefits. For this reason, it is necessary to develop tools for valuing (not only economically) and understanding the service provision changes due to agricultural management. We propose two new methodologies for the evaluation of potential impacts of agricultural activities on the ecosystem services: conceptual and bayesian networks. Both methods (one qualitative and another quantitative) have the same main graphical structure as they involve a group of nodes and connections between them. Conceptual networks are representations of the knowledge acquired; whereas Bayesian networks are presentations of conditional probability between nodes. This probabilistic method is useful for dealing with uncertainty, subjectivity and the actual limited knowledge we have about the consequences of agricultural management on ecosystems. In this model, we firstly obtained qualitative networks and then, we quantified these relationships and made the probabilistic model. The qualitative and quantitative information came from a literature survey and interviews with experts. These networks represent the logical links between actions (e.g. nitrogen fertilization) and consequences (e.g. nitrogen leaching) in agroecosystems. The model developed in this work represents a novel approach for assessing and valuing the supply of ecosystem services as affected, in a positive or negative way, by the agricultural activities. Furthermore, this model can give an inference of the degree of sustainability of agroecosystems.