Presenter/Author Information

Jenifer Ticehurst
C. A. Pollino

Keywords

bayesian networks, natural resource management, model development

Start Date

1-7-2010 12:00 AM

Description

In 2007 we undertook ‘capacity building’ with six Natural Resource Management (NRM) bodies within Australia, where the aim was to train staff in how to develop Bayesian networks (Bns). Previously, the NRM staff had expressed interest in developing Bns themselves to assist with their target setting, planning and reporting needs, so that investments in on-ground activities can be better targeted to outcomes in resource condition. Concurrently, we were developing generic Bns ‘collaboratively’ with the same groups. Six months after completing the initial training, none of the NRM regions had made any significant progress in the development of their own Bns. Follow-up surveys, two and a half years later, found that the development of Bns by the NRM regions themselves had largely been limited to conceptual diagrams and influence diagrams. The NRM regions who had made the most progress were those that had staff complete other external training, and those who had committed additional funding to external projects, rather than just internal management and target setting. The time commitment required to develop the Bns and lack of data resources remained the major limitations. Since the 2007 Bn training exercise, fewer participants believe that it is valuable and feasible for the NRM regions to develop their own Bns, but a greater proportion can see the usefulness of the Bn approach to their work. If completed Bns are the measure of success, then it is best to build collaborative models, but detailed capacity building in the initial stages of this project aided the depth of the collaborative feedback, the building of a working relationship between the researchers and stakeholders, and provided a systems approach to environmental management for the stakeholders. Consequently, we would recommend building both capacity and collaborative models to improve NRM decision making processes and to increase adoption of decision making tools.

Share

COinS
 
Jul 1st, 12:00 AM

Build collaborative models or capacity? Reflections from two years on.

In 2007 we undertook ‘capacity building’ with six Natural Resource Management (NRM) bodies within Australia, where the aim was to train staff in how to develop Bayesian networks (Bns). Previously, the NRM staff had expressed interest in developing Bns themselves to assist with their target setting, planning and reporting needs, so that investments in on-ground activities can be better targeted to outcomes in resource condition. Concurrently, we were developing generic Bns ‘collaboratively’ with the same groups. Six months after completing the initial training, none of the NRM regions had made any significant progress in the development of their own Bns. Follow-up surveys, two and a half years later, found that the development of Bns by the NRM regions themselves had largely been limited to conceptual diagrams and influence diagrams. The NRM regions who had made the most progress were those that had staff complete other external training, and those who had committed additional funding to external projects, rather than just internal management and target setting. The time commitment required to develop the Bns and lack of data resources remained the major limitations. Since the 2007 Bn training exercise, fewer participants believe that it is valuable and feasible for the NRM regions to develop their own Bns, but a greater proportion can see the usefulness of the Bn approach to their work. If completed Bns are the measure of success, then it is best to build collaborative models, but detailed capacity building in the initial stages of this project aided the depth of the collaborative feedback, the building of a working relationship between the researchers and stakeholders, and provided a systems approach to environmental management for the stakeholders. Consequently, we would recommend building both capacity and collaborative models to improve NRM decision making processes and to increase adoption of decision making tools.