Three subhypotheses were constructed following a review of the literature and development of a theoretical framework. First, it was hypothesized that the force-resistance index predicts channel cross-section area more appropriately than either tractive force or resistance offered by riparian vegetation. Second, total stream length upstream is important in predicting network volume, the prediction of which is a useful parameter for comparing flood-producing potentials for watersheds treated with flood detention dams. Third, weighted mean silt-clay percent in channel perimeter is a good estimator of channel width-depth ratio. Geology was included in the analysis as an indicator of the spatial distribution of channel form.

The t-test used in the analysis of the subhypotheses revealed significant differences between the slopes and intercepts of regression lines of sandstone streams and shale streams at a 0.005 significance level.

This research attempts to show how geology (lithology) and man control downstream channel form change. It was hypothesized that a change from sand to shale in the Middle Permiam beds of western Oklahoma produces profound differences in channel characteristics.

Combining the three successfully tested subhypotheses enabled estimation of the channel cross-section area fairly consistently. However, geology as an indicator factor was important in determining the spatial variation of downstream channel form change. Evidence of the effects of varying lithology on channel form sheds light on the environmental problems in fluvial research, and signals precautions necessary when results of research in one region are applied to another.