Modelling ecological responses to flow: Making the most of available evidence and data

Historically, it has been common for environmental flow modelling to rely on expert judgment. However, the literature now contains increasing evidence and data on the ecological effects of flow alteration. Using examples from our work, we will outline our approach to making maximal use of this information in developing more objective and robust flow-ecology response relationships. First, conceptual models can be developed using the Causal Criteria approach to evidence assessment. This removes some of the subjectivity inherent in expert-based conceptual models. Second, the conceptual model must be translated into one or more numerical model statements. Evidence from the previous step should also inform these choices. Finally, the numerical model must be quantified with existing or new data. The Bayesian hierarchical approach to data analysis provides a powerful and flexible method for incorporating data from seemingly dissimilar sites to improve inferential strength and reduce uncertainty. In sum, these methods allow the development of robust, evidence-based flow-ecology relationships that employ all possible existing data and evidence. Given that a reliance on expert judgment has stemmed from the perception that insufficient data are available to model flow-ecology responses, such effective use of data should be an aim of any investigation.