9484 - Laserfiche WebLink

Home Browse Search

December 2002 FLOW-SEDIMENT EFFECTS ON RIVERINE FISH 1969, Muth and Snyder 1995), and abundant nonnative minnows (Cyprinella, Notropis, and Pimephales) pro- vide ample forage for the next several years. As Col- orado pikeminnow mature (6-9 yr), the need for larger forage fish is not met in lower reaches of the Colorado River mainstem where native prey fish are scarce. Low body condition prompts many Colorado pikeminnow to disperse to upper reaches and tributaries where larger prey (suckers and chubs) are more abundant (Os- mundson et al. 1998). This progressive dispersal pat- tern results in relatively segregated life stages and adult densities are surprisingly clumped near the upstream margins of their range. The amount (kilometers) of suitable adult habitat in the Colorado River is substantially less than that avail- able in the Green River system and this accounts in part for the difference in size of the respective popu- lations. Additionally, the spring hydrograph of the Green River mainstem downstream of the Yampa River confluence more closely approximates natural condi- tions than does the hydrograph of the mainstem upper Colorado River. Although a population estimate is not yet available for the Green River system, Osmundson and Burnham (1998) estimated only 300-400 adult pikeminnow (?500 mm long) in the mainstem Colo- rado River during 1991-1994 and considered the pop- ulation vulnerable to extirpation. The Colorado River in the upper basin Setting.-The 373-km-long study area, situated in western Colorado and eastern Utah, encompasses the past and present range of the Colorado pikeminnow in the mainstem Colorado River upstream of the Green River confluence (Fig. 1). River locations are herein described in river kilometers (rk) upstream of this confluence (rk 0.0). The upstream-most 76 rk is a transition zone between warm- and cold-water fish communities and historic usage of this area by Col- orado pikeminnow is unknown. Today, upstream range of pikeminnow is truncated by a diversion structure at rk 303 that has blocked upstream fish movements since 1911. Excluded from study were the Gunnison and Dolores rivers, two primary trib- utaries entering the Colorado River within the study area. Headwaters of the three tributaries are located in the Rocky Mountains and the Colorado Plateau. The annual hydrograph is dominated by spring snow- melt that typically begins in late April, peaks in late May or early June, and recedes in July. Most runoff is derived from high-elevation basins underlain by erosion-resistant rocks, whereas most of the sedi- ment is derived from surface erosion of sedimentary rocks, primarily shale, in low-elevation basins (Iorns et al. 1965, Liebermann et al. 1989). Common, lo- calized, summer thunderstorms dramatically in- crease river turbidity but generally have little effect on mainstem discharges (Van Steeter and Pitlick 1998). Drainage area upstream of the Gunnison River 1721 confluence is -22700 km'; -46200 km2 upstream of the Dolores River confluence, and -70 000 kmz upstream of the Green River confluence. Within the study area, the river flows southwesterly, bisecting the Roan Mesa and Paradox Basin physiographic provinces of the Colorado Plateau (Liebermann et al. 1989). The gradient varies considerably (Fig. 1) and open valleys with floodplains (strata 4, 6, 8, 9, and 11) alternate with canyons (strata 1, 2, 3, 5, 7, and 10). Habitat richness is generally highest in flood- plains where multichannel segments are common; in more confined sections, single-channel segments predominate. Gravel-cobble substrates comprise the channel bed in most areas, although silt and sand dominate in some low-gradient areas. Regulation history.-Three mainstem, low-head, diversion dams built-just upstream of Palisade,- Col- orado between 1883 and 1916 collectively divert -43 m3/s from the river during the irrigation season (April through October) and -80% of this water is returned through groundwater and numerous drains entering the river at various sites for 70 km down- stream. Sediment plumes are normally observed at the mouths of return canals during the irrigation sea- son. The middle and upper dams block all upstream fish migration. In the headwaters of the Colorado mainstem, 15 dams with individual capacities >6.2 X 106 m3 were constructed, most between 1943 and 1968 (Liebermann et al. 1989). Collectively, the wa- ter volume stored in these reservoirs is equal to about one half of the mean annual streamflow of the upper Colorado River. Many of these were designed to serve transbasin diversions to the east side of the Rocky Mountains, and today, a mean of 14% (7-30% range) of the annual streamflow of the upper Colo- rado River is lost from the basin. In addition, res- ervoirs store runoff in the spring and release it slowly over the rest of the year to generate power and satisfy irrigation demands. This results in a flatter hydro- graph, with lower spring runoff flows and higher summer and winter base flows (Van Steeter and Pi- dick 1998). Earlier in the century (1902-1942), the median peak discharge (highest annual day) of the Colorado River (rk 298) upstream of the Gunnison River confluence was 838 m3/S; in more recent years (1969-1999), it was only 411 m3/s, a decline of 51%. Similarly, major dams in the Gunnison River basin were constructed primarily during 1937-1966 and the median peak flow upstream of the confluence with the Colorado River declined from 489 m3/s during 1897- 1936 to 216 m3/s during 1967-1999, a 56% decline (synthesis of U.S. Geological Survey [USGS] gauge data). Median peak discharge of the Dolores River near its mouth (historically about 195 m3/s), also declined following the construction of McPhee Reservoir in 1984, but only by -6%. While spring flows in the Colorado River have declined, sediment inputs have probably not. Thus, suspended sediment that was once