The previous discussions of river variability have applications to river management. The important concept is that river reaches vary in both location and through time. For example, an 11-mile (c. 17 km) reach of the Arkansas River (Figure 1.3) near Leadville, Colorado can be divided into eight reaches of different morphology. Gradient in this gravel-bed stream varies from 0.0067 to 0.011, sinuosity from 1.12 to 1.34, width from 60 to 110 feet (c. 18–33 m), calculated bankfull discharge from 330 to 1,060 cfs (c. 9–30 m3/s), and one reach is anastomosing. Any overall river modification or rehabilitation scheme would seem to fail because of the reach to reach variability of the river.

In order to prevent such problems, there are three concerns when undertaking practical work or, in fact, during any river investigation. These concerns are:

1. An investigation should always consider not only the site of interest, but upstream and downstream river reaches to determine if the reach of concern is representative of the river. That is, an investigator should back away from the specific problem site and view it in a broader context.

2. Rivers may range in sensitivity from very to not at all. An attempt should be made to evaluate river and reach sensitivity to determine if change is likely (Figure 11.2).

3. The multiple hypothesis approach should always be considered in an attempt to explain or anticipate river behavior. That is, the most obvious conclusion may be incorrect.

Concern 1: a broader perspective

A good example of Concern 1 was the problems associated with a bridge over the Cimarron River (Figure 1.3) near Perkins, Oklahoma (Keeley, 1971).