The spawning habitat associated with fall chinook salmon (Oncorhynchus tshawytscha) redd clusters was investigated in the Hanford Reach of the Columbia River. A conceptual spawning habitat model is proposed that describes how geomorphic features of river channels create hydraulic processes, including hyporheic flows, that influence where salmon spawn in unconstrained reaches of large mainstem alluvial rivers. Spatial point pattern analysis of redds showed that redd clusters averaged approximately 10 hectares in area and their locations were consistent from year to year. The tendency to spawn in clusters suggests fall chinook salmon's use of spawning habitat is highly selective. Hydraulic characteristics of the redd clusters were significantly different than the habitat surrounding them. Velocity and lateral slope of the river bottom were the most important habitat variables in predicting redd site selection. While these variables explained a large proportion of the variance in redd site selection (86 to 96%), some unmeasured factors still accounted for a small percentage of actual spawning site selection. Further investigation showed that the magnitude and chemical characteristics of hyporheic discharge were different between and within two spawning areas. Apparently, fall chinook salmon used chemical and physical cues from the discharge to locate spawning areas. Traditional spawning habitat models could be improved if they: used spawning area-specific, rather than river-specific; spawning characteristics; incorporated hyporheic discharge measurements; and gave further consideration to the geomorphic features that are present in the unconstrained segments of large alluvial rivers. Ultimately the recovery of endangered fall chinook salmon will depend on how well we are able to recreate the characteristics once common in alluvial floodplains of large rivers. The results from this research can be used to better define the relationship between these physical habitat characteristics and fall chinook salmon spawning site selection, and provide more efficient use of limited recovery resources. / Graduation date: 1999