Modeling the influence of flow on invertebrate drift across spatial scales using a 2D hydraulic model and a 1D population model

Methods for creating explicit links in environmental flow assessments between changes in physical habitat and the availability and delivery rate of macroinvertebrates that comprise fish diets are generally lacking. Here, we present a hybrid modelling approach to simulate the spatial dynamics of macroinvertebrates in a section of the Merced River in central California, re-engineered to improve the viability of Chinook salmon. Our efforts focused on quantifying the influence of the hydrodynamic environment on invertebrate drift dispersal, which is a key input to salmon bioenergetics models. We developed a two-dimensional hydrodynamic model that represented flow dynamics well at baseflow and 75% bankfull discharges. Hydraulic predictions from the 2D model were coupled with a particle tracking algorithm to compute drift dispersal, where the settling rates of simulated macroinvertebrates were parameterized from the literature. Using the cross-sectional averaged velocities from the 2D model, we then developed a simpler 1D representation of how dispersal distributions respond to flow variability. These distributions were included in 1D invertebrate population models that represent variability in drift densities over reach scales. Dispersal distributions in the 2D simulation and 1D representation responded strongly to spatial changes in flow. When included in the 1D population model, dispersal responses to flow 'scaled-up' to yield distributions of drifting macroinvertebrates that showed a strong inverse relationship with flow velocity. The strength of the inverse relationship was influenced by model parameters, including the rate at which dispersers settle to the benthos. Finally, we explore how the scale of riffle/pool variability relative to characteristic length scales calculated from the 1D population model can be used to understand drift responses for different settling rates and at different discharges. We show that, under the range of parameter values explored, changes in velocity associated with transitions between riffles and pools produce local changes in drift density of proportional magnitude. This simple result suggests a means for confronting model predictions against field data.