In New Zealand, planning for freshwater quality and quantity management is being delegated increasingly to collaborative stakeholder groups (CSGs), as encouraged by the Government’s recent freshwater reforms. Comprising major water users and communities of interest, CSGs typically identify a wide variety of ecological, economic, social and cultural values related to a river watershed, values that interact in highly complex ways.
Given this complexity, making water policy and management decisions that achieve acceptable outcomes across many values is an enormous challenge for a stakeholder group, especially because of the stakeholders’ different levels and areas of knowledge regarding freshwater. Robust decision-support tools are needed. Such tools need to be highly interactive and transparent, due to the collaborative nature of the planning process and stakeholder perceptions of scientific models as ‘black boxes’.
In the greater Heretaunga and Ahuriri area (Hawke’s Bay), a CSG was convened by the regional council to revise the regional water plan and improve the balance between ecological and increasing human demands for water. A Structured Decision Making (SDM) approach was adopted, within which stakeholders identified a range of objectives and the policy/management options that could be used to achieve them. SDM requires the stakeholder group to estimate the consequences of different policy options on the range of objectives they have identified, so they can choose the best policies for the revised water plan. These estimates must be seen as reasonable and unbiased, thus requiring a transparent decision-making process.
To develop a transparent process, the CSG constructed ‘influence diagrams’ that represented their understanding of the cause-effect relationships linking policy/management choices to outcomes for the different objectives. The influence diagrams were refined with the help of council science staff and other experts, including individuals from the stakeholder group, and agreed on with the CSG.
The same expert group then quantified the relationships between nodes of the influence diagrams, thus converting them into Bayesian Networks (BNs). BNs quantify cause-effect relationships between variables using probabilities to reflect the strength of influence and the certainty with which the relationships are known. Scientific theory, empirical studies and expert opinion are all used to set the probabilities, which can be updated as new information becomes available. Once quantified, the BNs will help the CSG evaluate and compare different policy options so that an acceptable outcome is likely to be achieved for all relevant objectives.
Some members of the CSG were initially resistant to using BNs, mainly due to a perception of models as obscuring the decision-making process and taking control away from stakeholders. Effort was given to presenting the BNs as a decision-support tool that organises and integrates existing knowledge. The value of the Bayesian Networks lies both in the final product, which is a transparent tool to support decision-making, and in the development process, which engages both stakeholders and technical experts in identifying the key values, issues and processes in river systems.
External collaborators:
Jim Sinner (Cawthron Institute), Tim Sharp (Hawkes Bay Regional Council), and Suzie Greenhalgh (Landcare Research)