| Headwater areas have received little attention from researchers, even though headwater catchments can comprise over 70% of the land area in most regions. Headwater streams, being tightly coupled to their adjacent hillslopes, are greatly affected by the upland landscape history in their watersheds. I employ this tight connection in developing a watershed classification system that explicitly links catchment characteristics and stream channel morphology. I test two approaches for classification. First, I assess whether a 'top-down' statistical clustering, based on landscape-scale attributes, forms catchment groups with significantly distinct types of stream channels. In the second 'bottom-up' approach, I test whether catchments grouped by their respective distinct types of stream channels show any significant relationships between channel morphology and landscape-scale attributes. With the top-down technique, I use a geographic information system (GIS) and a digital elevation model (DEM) to delineate 862 headwater catchments; I then use a two-step clustering procedure to create six groups based on catchment shape, area, land use, and geology. For the bottom-up approach, I use channel morphology data, from 51 sampled streams, in a principal components analysis (PCA) and a two-step cluster procedure to create five groups of catchments based on the similarities in stream channel morphology. I then use multinomial logistic regression to test the efficacy of predicting bottom-up group membership using landscape-scale attributes as independent variables. The top-down classification creates groups of catchments with different landscape-scale attributes, but these groups do not have significantly different types of stream channels; this is largely because the top-down approach cannot account for convergence, where different combinations of catchment attributes produce similar types of channel morphology. Being more process-driven, the bottom-up catchment groups represent transitional states in the expected fluvial response to hillslope disturbance; stream channels are either aggrading, degrading, or in dynamic equilibrium. Finally, bottom-up catchment group membership is predicted with better than 80% accuracy. This demonstrates the strong relationship between landscape-scale processes and stream channel morphology, which may allow researchers to anticipate possible changes in stream channel habitat as a function of proposed land use changes. |
