|
INTRODUCTION
<br />The native fish community of the Colorado River is adapted to a dynamic riverine environment
<br />characterized by pronounced seasonal fluctuations in water discharge, temperature and sediment
<br />load, and significant longitudinal variations in channel morphology, gradient and substrate sediment
<br />size. Many aspects of this environment have now been altered either directly or indirectly by
<br />human activities. Reservoir operations and flow diversions in the Colorado and Gunnison River
<br />basins have reduced peak discharges by 29-38% (Van Steeter and Pitlick, 1998). As a result, both
<br />rivers have lost some of their capacity to transport sediment, causing an overall narrowing of the
<br />main channel and a loss of associated side channels and backwaters. These changes, along with
<br />the introduction of non-native fish species, appear to have adversely affected several native fish
<br />species, including the Colorado pikeminnow (Ptychocheilus lucius), formerly known as the
<br />Colorado squawfish. The Colorado pikeminnow is one of four federally listed endangered species
<br />in the upper Colorado River basin and its population remains quite small (Osmundson and
<br />Burnham, 1998).
<br />Recent studies of native fish abundance by Osmundson et al. (1998) indicate that juvenile and adult
<br />pikeminnow are located in distinctly different reaches of the Colorado River (defined here as that
<br />portion of the Colorado River upstream of the Green River confluence). Most of the juvenile and
<br />subadult pikeminnow (fish less than about 7 years in age and 550 mm in length) are found in
<br />canyon-bound reaches below Westwater Canyon, UT, whereas most of the adults (fish greater than
<br />550 mm in length) are found in alluvial reaches near Grand Junction, CO. The difference in
<br />location of adults and juveniles presumably reflects a transition in the need for certain resources and
<br />habitats, which are governed in turn by differences in physical conditions within the upper and
<br />lower reaches. The upper reach near Grand Junction is characterized by coarse bed materials, and
<br />simple to complex channel patterns that provide a wide variety of habitats, including pools, riffles,
<br />runs and backwaters (Osmundson et al., 1995; Lamarra, 1999). In contrast, the reach below Moab,
<br />UT, is characterized by lower habitat heterogeneity (Lamarra, 1999), and sandy-silty bed materials
<br />that are mobile over a wide range of flows. The overall channel pattern in this reach is less complex
<br />than it is in gravel-bed reaches upstream, but the mobility of the substrate results in the formation of
<br />numerous sand bars and shallow embayments (backwaters) that provide important nursery habitat
<br />for young-of-the-year fish.
<br />The physical characteristics of river channels typically change downstream, and the shape of the
<br />channel (hydraulic geometry) continually evolves to carry the water and sediment supplied from the
<br />drainage basin (Gilbert, 1877; Mackin, 1948; Rubey, 1952). In one of the first attempts to quantify
<br />and explain the longitudinal trends in river channel characteristics, Leopold and Maddock (1953)
<br />presented a set of empirical equations showing that the width, depth, velocity, slope, and sediment
<br />load of rivers varied downstream as power functions of discharge; these equations were referred to
<br />collectively as "hydraulic geometry" relations. Data presented in subsequent studies showed that
<br />rivers in many different environments are characterized by broadly similar hydraulic geometry
<br />relations, meaning the bankfull width, depth and velocity vary downstream in roughly the same way.
<br />Specifically, it has been noted in many studies that the bankfull width increases downstream to
<br />about the 0.5 power of discharge, while the bankfull depth increases to about the 0.4 power of
<br />discharge, and the velocity changes little if at all (Simons and Albertson, 1960; Leopold et al., 1964;
<br />Church, 1992; Knighton, 1998). Recent attempts to predict the hydraulic geometry of self-formed
<br />alluvial channels focus more on physical processes of sediment transport through a cross section
<br />(e.g. Parker, 1979; Ikeda et al., 1988; Pizzuto, 1990). Parker's theory for the formation of a stable
<br />gravel-bed channel predicts that the bankfull width and depth are set by a discharge that produces
|