Fuzzy modelling of habitat suitability using 2D and 3D hydrodynamic models: Biological challenges

摘要

The last 10 years has seen a considerable improvement in our ability to measure and to model the complex flow fields in a range of river environments, even to the point that we have numerical schemes and data acquisition strategies that allow us to model the interaction between the detailed geometry of individual clasts and clast clusters and the associated three-dimensional flow and sediment transfer fields. At the same time, habitat modeling has developed significantly, as we have developed methods for coupling hydrodynamic models to measured habitat preferences. Here, we show how we have developed this linkage through coupling of a depth-averaged hydrodynamic model to habitat preferences, using a fuzzy rule-based approach. We demonstrate how this approach can be used to determine habitat suitability for two fish species (Atlantic salmon and Brown trout), in relation to spawning, nursery and rearing habitat, and related to proposed changes in compensation flow releases downstream from a dam, and possible future climate changes to the 2050s. The results demonstrate the crucial need to explore flow-biology interactions at the within- reach scale, especially in relation to low flows. In reflecting on these findings, we show that whilst three-dimensional modeling may reduce some of the uncertainties that follow from a depth-averaged approach, its applicability is limited by both the small scale of predictions generated and the lack of ecological knowledge of how to interpret those scales of prediction. The scales of habitat that matter in ecological terms are larger and more continuous, questioning approaches based on modeling 'representative reaches' and necessitating exploration of new ways of coupling biological and hydrodynamic knowledge.