Building hydrology-food web model to assess environmental flows for indigenous fish in the upper Yangtze mainstem

The construction of cascade dams in the upper Yangtze River has severely impacted indigenous fish resources. Current hydro-ecological operations primarily focus on fish spawning but ignore other critical life stages, such as juvenile rearing, resulting in limited recovery of adult fish. To solve this problem, we develop an innovative Hydrology-Food Web Model (HFM) for the mainstem reach downstream of Xiangjiaba Dam (XJB), the lowermost mega-dam in the upper Yangtze mainstem, to simulate the relationships between the annual dynamics of XJB discharge and the production of basal ecological groups (prey) and fish in the downstream reach. HFM exhibits satisfactory predictive performance owing to three key features: 1) its framework incorporates the requirements of all life history stages of two major fish groups, i.e., indigenous fish producing pelagic and adhesive eggs; 2) ecological group models are built based on biologically perceptive spatiotemporal scales and hydro-ecological mechanisms, ensuring high predictive accuracy; 3) a structurally simplified yet mechanistically robust food web model effectively integrates basal group models to predict fish production. Using HFM, we can assess environmental flows for the reaches downstream of XJB, satisfying hydrological needs of both fish spawning/hatching and basal groups (primarily floodplain plants). Our model provides a more comprehensive and ecologically realistic tool for river management. It is applicable to ecological operation and biodiversity restoration in regulated rivers worldwide. The main innovations include: 1) advocating for environmental flows assessments targeting entire life cycles of most species within protected groups; and 2) establishing two principles for hydrology-biology model development—matching model’s spatiotemporal scales to biological perception and employing mechanism-driven quantitative modeling.