Integrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes

Addressing global water security challenges requires effective watershed management that seamlessly integrates ecosystem service (ES) assessments with governance and socio-economic aspects of water resource utilization. However, most existing modeling approaches focus narrowly on biophysical processes and lack mechanisms for incorporating local knowledge under uncertainty. Unlike previous ES-watershed models that focused on biophysical indicators, our approach integrates ecological modeling with local perceptions and governance probabilistically, representing uncertainty and socio-ecological interactions that are often omitted from ES tools. The rural water supply system (RWSS) supports ES for water access, biodiversity, and social cohesion but is absent from watershed planning. To address this gap, we developed a spatial Bayesian Belief Network (BBN) model using data from RWSS organizations to support participatory zoning. The spatial BBN framework was applied to the Coello River Basin, Colombia, integrating high-resolution land use, ecosystem service multifunctionality indices, and multidimensional poverty indicators with perception data from the community leaders. The model generated spatial probability maps of multifunctional ES zones, incorporating governance capacity and conservation willingness. Key variables affecting ES multifunctionality include participation, restrictions, and poverty. Areas of high multifunctionality are aligned with zones of strong governance and high ecological integrity. This BBN enhances planning by integrating community insights into ecological modeling. By quantifying both the probability of ES supply and the uncertainty of predictions, the framework provides actionable insights for integrated watershed management (IWM) and supports the localization of global Sustainability Development Goals (SDG) (SDG 6-clean water and sanitation and SDG 15-life on land) in diverse socio-ecological contexts.