Spawning salmon depend on fluvial processes to maintain the bed texture of gravel bedded rivers as suitable spawning and incubation habitat. The coarse texture of a gravel bed is maintained by flow strengths capable of moving the gravel and removing fine sediments, thereby providing loose substrate that enhances its ability to conduct flow (i.e., its hydraulic conductivity). The flow strength that corresponds to the beginning of movement (i.e., entrainment) of a grain on a gravel bed is variable and, therefore, it needs to be measured to predict flow levels capable of bed texture maintenance. Lower flows can deliver fine sediment, which may accumulate in the gravels overlying salmon nests (i.e., redds) reducing the hydraulic conductivity and impairing salmon embryo development. I examine these processes to explain the variability in (i) the rate of change in the proportion of a gravel size fraction entrained as a function of flow strength and (ii) the rate of decrease in hydraulic conductivity as a function of the cumulative transport of fine sediment that depends on flow level.

I used tracer gravel and cobble grains and a two dimensional flow model to determine the flow strengths capable of gravel bed entrainment using binary plots of the occurrence or absence of tracer movement to approximate the fraction of the bed sediment entrained as a logistic function of increasing flow strength. The method provides an approximation of the flow strength capable of entraining the least resistant through the most resistant grains, thereby providing a new method to approximate the flow strength that is capable of fully entraining the bed. The results are confirmed by comparing the measurements of two study sites.

To measure the resistance of individual gravel and cobble grains to downstream movement, I used force gages and a theoretical force balance model that incorporates the frictional resistance of a grain and the lift and drag forces applied by stream flow. I measured the frictional resistance of grains at six sites with different morphologies to evaluate influences on grain resistance. Grain resistance varied depending on the grain size and sorting, amount of infiltrated sand, streamwise position along a gravel bar, and degree of fluctuation in the flow strength. The force balance model predicts that the dimensionless instantaneous flow strength capable of entraining a given proportion of a relative grain size varies along a bar. I tested the force balance model predictions by comparison with the tracer measurements from the tracer study. The differences between the force balance predictions and the tracer entrainment measurements are due to the definition of the flow strength as the time-averaged value from the 2D flow model and the instantaneous value from the force balance model. Adjusting the force balance model predictions by a simple factor that quantifies their difference and that corresponds with an index describing the fluctuations of the instantaneous flow strength about its time-averaged value produces agreement with the tracer measurements. This adjustment to force gage measurements from two sites on opposite limbs of a bar indicates that the time-averaged dimensionless flow strength capable of entraining a given proportion of a relative grain size is constant along a bar.

I constructed artificial redds at five sites to monitor changes in hydraulic conductivity and sand accumulation with cumulative bed load transport. I applied two backfill treatments to examine the effects of the presence and absence of grains smaller than 6.4 mm. Hydraulic conductivity varies temporally due to sand deposition and spatially due to the initial sand content of the backfill and the local sediment supply rate. The rate of change in hydraulic conductivity as a function of increasing cumulative transport is statistically significant but does not differ between backfill treatments or with location. I predicted the effect of flow on egg survival using an empirical relationship between hydraulic conductivity and egg survival with results that indicate hydraulic conductivity is dependent on sediment transport rate as a function of flow rate.

This work provides additional understandings of the relationship between flow and bed texture maintenance processes and useful tools for managing gravel bed rivers.