Building models for water management are particularly critical and challenging in regions with limited ecohydrological information. Frameworks for water management depend on the availability of data and parameterized hydrological models simulating the functioning of the hydrosystems to be managed. The availability of relatively long term empirical data may provide a solid ground in the understanding of hydrosystems to adapt frameworks to data-deficient conditions. We applied and adapted the Eco-engineering decision scaling EEDS framework from Poff et al. (2016) to establish e-flows in several mountain streams providing with potable water to Quito, the 2.7-million inhabitants capital of Ecuador. As in the EEDS, we developed a collaborative initiative with the main stakeholders involved in water management: the metropolitan drinking water company of Quito (EPMAPS) and the water fund for Quito (FONAG) in charge of catchment conservation. In a common agreement, we sampled over two years, flow and benthic fauna in 12 head water stream sites of two of the main catchments providing potable water to Quito. Our preliminary hypothesis was that empirical data from streams subjected to withdrawals and natural streams offer the possibility to quantify thresholds for managing e-flows from water infrastructures. As an application of the EEDS, our objectives were threefold: 1) to evaluate benthic fauna response to flow in streams monitored for water withdrawal, 2) to analyze thresholds in flow–benthic fauna relationships, and 3) to identify trade-offs for acceptable eco-engineering management levels. We propose a five-step roadmap to define ecological flows in data defficient hydrosystems were a priori thresholds cannot be defined. We propose a System' database that is built upon empirical information and instead a Vulnerability analysis a Thresholds analysis to identify change points on flow-ecology relationships.
Results/Conclusions
Our findings suggest that flows of 20% of the mode of the period might be the least e-flow condition that these hydrosystems can support to recover from impacts and reach to some extent the observed natural conditions. In contrast, with remaining flows of 40% in low-flows due to withdrawals, ecological metrics seemed to behave in a similar way as the conditions found in natural low-flows. Although our findings are constrained to the empirical data generated from monitoring streams, the flow–benthic fauna relationships found in the threshold analysis provide in the light of scarce-data hydrosystems, one of the first attempts to implement the eco-engineering decision scaling framework adapted to define environmental flows in hydrosystems threatened by flow alteration for water allocation.