Using river temperature to optimize fish incubation metabolism and survival: a case for mechanistic models

Allocating reservoir flows to societal and ecosystem needs under increasing demands for water and increasing variability in climate presents challenges to resource managers. In the past, managers have operated reservoirs to achieve flow and temperature compliance points based on qualitative predictions of competing needs. Because it is difficult, if not impossible, to assess whether meeting such compliance points is efficient or equitable, new strategies for regulation are being advanced. Critical to these strategies is the need for new models with sufficient biological details to identify the effects of reservoir operations on organism growth and survival in real time. This paper evaluates the adequacy of three models of differing complexity for managing the Sacramento River temperature during the incubation of winter-run Chinook salmon. The models similarly characterize temperature-and density-dependent mortality from egg through fry survival, but use different spatial and temporal resolutions. The models all fit survival data reasonably well, but predict different reservoir operations to protect fish. Importantly, the models with the finer spatial/temporal resolution predict reservoir operations that require less flow and better protect fish when water resources are limited. The paper illustrates that shifting the focus of management from meeting compliance points to meeting the metabolic needs of the organisms yields efficiencies and identifies when water is needed and when it can be saved.