Integrating hydrology and ecology models into flexible and adaptive decision support tools: The Ibis DSS

Terminal wetlands in the semi-arid regions of northern NSW are important ecological refuges for native fauna, especially during dry times. Many of these systems have become increasingly stressed by human induced changes in the hydrologic regime of the rivers flowing into the lakes. For example, the Narran Lakes is a Ramsar wetland recognised for its geomorphological significance and the importance of the system as habitat and drought refuges for waterbirds and other species. The high level of regulation in the headwaters of the lake poses considerable obstacles for managers responsible for managing the health of the Narran Lakes. Decision Support Systems (DSS) can support decision making processes by providing users with a tool that shows the relationships between drivers of a system and outcomes. Drivers of wetland systems can be different management actions (e.g. environmental flow releases) or 'uncontrollable' drivers like climate variability. The complex biophysical, ecological and social context in which wetlands are managed mean that a DSS that supports decision-making should address uncertainty in data, knowledge and predictions, and allow users to explore the sensitivity of outcomes to management actions, uncontrollable drivers and uncertainty. It will be most useful in the long term if used as part of an adaptive process where objectives, strategies and knowledge are continually updated to test hypotheses and reduce key uncertainties, using a monitoring, research and evaluation program which the DSS could be augmented to inform. The IBIS DSS is being developed to predict the ecological outcomes of environmental flows in wetland systems in NSW. Two applications are in development: one each for the Gwydir Wetlands and Narran Lakes. It is intended to be used by wetland managers to inform and support on-ground management decisions, and model developers who can input new data and rules into the DSS and test hypotheses. The DSS links hydrologic-hydraulic models to Bayesian Network ecology models. Relationships between hydrology and ecological response are not fully established for either of the Gwydir Wetlands or Narran Lakes systems. For example, what constitutes the start and end of an event and what are the important characteristics of events? Triggers or water requirements for breeding and/or growth, which may or may not be known, can vary between species or over a species life cycle. Also, multiple factors may influence ecological response. In the Narran Lakes application of the IBIS DSS, the likelihood of waterbirds abandoning their nests is determined by the length of a hydrologic event, and the minimum depth under nests, maximum day-to-day decrease in water surface elevation, and number of 'cold' days during the event. The individual impact of these factors on abandonment is known with varying degrees of certainty. However, the combined impact of these factors is not so clear. The IBIS DSS allows users to systematically specify rules that define event characteristics and aid development of the probabilistic data in the Bayesian Network. This allows testing of alternate model structures and adds flexibility to the DSS that lends its use as part of an adaptive monitoring, research and evaluation program.