Using Bayesian networks (BNs) for mapping stakeholders behaviors in integrated water resource management with a focus on irrigated agriculture in Al Batinah region of Oman.

The problems of water resource management and its interaction with society have to be tackled from an integrated perspective taking into account the interdependence of environmental, political, social and economic factors. The conventional approaches being currently used, to study water resources management (WRM), lack to reflect the mutual relationship between water resources and societies. However, varieties for approaches to developing models of complex systems are available. Bayesian networks (BNs), is an approach, which can integrate data and knowledge of different types and from different sources in which, causal links join the variables. A BN is a type of decision support system based on a probability theory that implements Bayes’ rule. This work is focusing on the case of Al-Batinah coastal plain in Oman where lots of small-scaled farms practice agriculture. There is no control over the water amounts abstracted from the coastal aquifer. Therefore, the coastal aquifer is at risk due to seawater intrusion. Since groundwater replenishment is limited, the region is facing a problem of water deficit, which also influences the sustainability of agricultural production. Based on a social survey conducted regarding the Al Batinah case, the existing situation generates conflicts between different stakeholders (SHs) which have different interests regarding water availability, sustainable aquifer management, and profitable agricultural production. Therefore, the development of appropriate management strategies for a transition towards a stable and sustainable future hydrosystem states is required. This work aims to evaluate the implementation potential of several management interventions and their combinations by analysing behaviors and opinions of relevant SHs (farmers and decision makers) in the region. This should support decision makers (DMs) in taking more informed decisions. Data were collected through a social survey. Differences were examined statistically between opinions of farmers and decision makers regarding potential interventions. Additionally, the approach of BN was used for mapping stakeholders’ behaviors and to show the strength of a relationship between dependent and predictor variables. The findings suggest that BNs provided an enhanced understanding of the presence and strength of causal relationships. The hypothesis for most of the variables, in the structure of the network, worked logically. Moreover, it is possible to determine the implementation potential of management interventions regarding their acceptance, and additionally, triggers can be identified to increase this potential. Nevertheless, management interventions should also be evaluated regarding economic and environmental criteria. [ABSTRACT FROM AUTHOR]