The application of Bayesian networks to evaluate risks from multiple stressors to water quality of freshwater ecosystems

It is difficult to predict and manage the ecological consequences of multiple water quality stressors on our freshwater systems. This due dynamism source-stressor-response relationships factors including lack data, complex impact pathways risks, uncertainties that are parameterise. We present a risk-based probability modelling approach using Bayesian network (BN), at spatial scales. illustrate use this approach, by evaluating probable effects altered associated with sources in three case study rivers South Africa. Water land activity were used describe conceptual risk pathways, parameterise BNs model stressors. The BN demonstrated endpoints selected for reflected risks levels input variables. further electrical conductivity was just as sensitive more salt model. able accurately represent all systems irrespective data base size. can contribute towards sustainable resource management.