A Bayesian Belief Network Decision Support Tool for Watering Wetlands to Maximise Native Fish Outcomes
- First Online:
- 384 Downloads
Wetlands are productive and diverse habitats for native fish but can be highly degraded, particularly in the Murray-Darling Basin (MDB), south-eastern Australia. Wetland management requires tools and processes that facilitate the synthesis and application of knowledge for decisions concerning the allocation of environmental water to wetlands to improve environmental outcomes. This paper describes the development of a Decision Support Tool (DST), based on a Bayesian Network designed to provide the best available science and support adaptive management of environmental flows into wetlands. The DST predicts the probability of improvements in fish population health as defined by abundance, population structure and fish condition for introduced common carp and three native species of fish: carp gudgeon, Australian smelt, and golden perch. Model sensitivity and validation showed that fish response varied depending on model inputs, but that responses from the DST were an accurate reflection of fish responses in wetlands based on field data. Ultimately, the success of this DST is dependent on its adoption by wetland managers. Throughout the entire development process, adoption of the DST has been promoted through engagement with managers and subsequently, through initiatives to integrate it into current management initiatives.