Defining minimum environmental flows at regional scale: application of mesoscale habitat models and catchments classification

In the context of water resources planning, this work defined a possible approach to quantify minimum environmental flows (e-flows) at a regional scale. Focusing on catchments smaller than 50km2, the problem was addressed through mesoscale habitat models and a catchment classification technique (regression tree algorithm). Within the Piedmont region in NW Italy, 25 reference streams were chosen on the basis of the natural condition of the flow regime and fish community. Mesohabitats were sampled for hydromorphic and fish parameters following the mesoscale habitat models approach. Logistic regression models, along with 55 habitat descriptors, were then used to build multivariate habitat suitability criteria, identifying the habitat characteristics mostly used by the target fish species. These models were applied to each stream reach and used to classify each mesohabitat into suitability categories. The reference minimum discharge for each stream was derived from habitat–flow rating curves. Finally, to define the regional criteria, the study domain was split according to the regression tree classification, defining homogenous sub-regions distinct on both e-flows and catchment/stream characteristics. This bottom–up approach used a catchment classification technique based on the environmental requirements of the fish communities and demonstrated potentials for further applications to defining e-flows at regional scales.