The use of streambed texture to interpret physical and biological conditions at watershed, reach, and subreach scales

Thesis (Ph. D.)--University of Washington, 1998Physical controls on bed-surface grain size and consequent implications for aquatic habitat are examined at a variety of spatial scales. Data compiled from laboratory studies show that reach-average median surface grain size (D50) varies inversely with sediment supply, while field studies of forest gravel-bed rivers conducted here demonstrate that D50 varies directly with bed shear stress as modified by hydraulic resistance due to channel walls, bars, and wood. Textural response is evaluated relative to a theoretical prediction of competent D50. I hypothesize that surface textures are altered through size-selective erosion or deposition caused by imbalances between sediment supply and transport capacity. Moreover, changes in surface texture that smooth or roughen the bed should alter both the boundary shear stress and the critical shear stress of grains moving over the bed, thereby creating a feedback process between bed-surface texture, sediment flux, and supply-transport state. At subreach scales, my field studies demonstrate that bed surfaces are composed of textural patches, the number and frequency of which varies directly with the magnitude and frequency of hydraulic roughness, due presumably to greater spatial divergence of transport capacity and sediment supply forced by the roughness elements. Bed-surface textures were classified using a new procedure that combines visual and quantitative measurements of patch grain-size distributions and offers reasonable statistical discrimination of differences in both mean grain size and variance between textural patches.I further examined the correspondence between bed shear stress and surface grain size in two forest pool-riffle channels using a theoretical shear-stress partitioning model. Reach-average velocities calculated from the model are within 7--8% of those measured in the field using a salt-tracer, providing support for the model. Observed values of D50 are in close agreement with those predicted from modeled bankfull bed stresses, indicating that surface grain size of the sites is in quasi-equilibrium with bankfull channel hydraulics.Greater textural variation at subreach scales creates a broader diversity of aquatic habitats for animals that rely on specific substrate sizes. Furthermore, textural fining caused by bar and wood roughness has the potential to create usable salmonid spawning-gravels in channels that otherwise would be too coarse. Coupling field measurements and theoretical predictions via digital elevation models, I propose a watershed-scale framework for assessing the potential influence of hydraulic roughness on salmonid spawning-habitat availability. Application of the model indicates that textural fining caused by bar and wood resistance can control more than 90% of the potential spawning habitat in mountain drainage basins studied here, with bar roughness most important for lower-mainstem reaches (41--73% of the potential habitat), and wood roughness important for upper-mainstem channels (25--53% of the potential habitat)