9484 - Laserfiche WebLink

Home Browse Search

1736 D. B. OSMUNDSON ET AL. Ecological Applications Vol. 12, No. 6 framework is set in motion, the bed becomes a source of suspended sediment as fines are winnowed out. How much fine sediment accumulates in the bed is a function of sediment load, local velocity, and the period of time separating bed-mobilizing events. Flows of sufficient magnitude to fully mobilize the bed now occur at a lower frequency than during historic times, largely due to river regulation. The result has been a lengthening of time between such events (Fig. 7). For strata 6-11, estimates of historic recurrence intervals averaged 1.2-2.7 yr, depending on the stra- tum. Our monitoring results of DFS in strata 8 and 9 indicate that sediment begins to accumulate in the bed within the first year following a flushing event and DFS continues to decline through at least the first four years if flows do not exceed the threshold for widespread bed mobilization. Flows reaching or exceeding the thresh- old for initial'motion were ineffective at reversing de- clines in DFS. In the 34 yr since most water devel- opment projects were completed, mean duration sep- arating flushing events doubled in strata 9-11, tripled in stratum 8 and increased fivefold in strata 6 and 7. From 1957 to 1983, strata 6 and 7 underwent a 26-yr period without a widespread mobilizing event. Simi- larly, stratum 8 lacked such flows during 1965-1978 (14 yr). Historically, the longest interval for strata 6- 11 was 3 yr. Given the link between benthic biomass and fine sediment indicated here, we infer that in recent times these strata- underwent artificially extended pe- riods of lowered productivity with a corresponding negative effect on resident fish populations. This, cou- pled with range reduction and increases in nonnative fish of limited forage value, have almost certainly re- duced the capacity of the upper Colorado River to sup- port adult Colorado pikeminnow. Management implications Spatial distribution of the community's primary pis- civore was related to the relative abundance of potential prey fish. Our results also indicated that total surface area of the river and the availability of non-riffle/run mesohabitats (pools, eddies, backwaters, etc.) posi- tively influenced adult pikeminnow densities in com- bination with food availability. Relationships with food availability were stronger than with habitat area alone. Hence, traditional instream flow methodologies that strive to increase standing stocks of target species by emphasizing the importance of preferred physical hab- itat and the manipulation of base flows may be met with limited success in the upper Colorado River and perhaps in other similarly regulated rivers. Such efforts may prove beneficial, however, if coupled with man- agement actions designed to enhance food availability. Our work provides a first step towards understanding how changes in flow regime in the upper Colorado River might affect the foodweb supporting the main- channel fish community. To increase standing crops of primary producers and consumers, strategies aimed at restoring the frequency of flushing events should be explored. Runs were the most pervasive habitat type in the river and the most in need of more frequent flushing of fine sediment. Riffles were relatively clean during our longitudinal study regardless of location; however, the 1994-1995 sampling may have represented a particularly "clean" period: discharges adequate to fully mobilize the bed occurred during 1993, a year preceding the start of our study, and again during 1995, just prior to our last sampling effort. Results of the 6-yr DFS monitoring in strata 8 and 9 showed that riffles, though generally more free of fine sediment than runs, can nonetheless accrue fines over time. In all strata, riffles made up a small percentage of the overall total area. Riffles were especially rare in downstream strata 1-5; there they apparently act as localized centers of bioproduction amid a channel dominated by unproductive runs. It is difficult to surmise the degree to which the lon- gitudinal distributions of fine sediment documented here deviate from the historical norm and the degree to which management efforts might be successful in enhancing conditions in some strata. Indeed, the period of our longitudinal study may have coincided with a relatively clean period, representing substrate condi- tions difficult to improve upon. Thus, benthic inver- tebrates, with their quick recovery time, might have been at near-optimum levels; however, fish populations, due to lag effects, may have required more time to respond to improved conditions. Hence, our fish abun- dance results may have reflected antecedent or long- term average conditions. Clearly, the main objective of providing flows that flush fines from the bed is to make runs more biologically productive more of the time. Bankfull flushing flows serve other purposes bene- ficial to Colorado pikeminnow and to the native fish community in general. Such flows are often channel- forming flows that scour encroaching vegetation and maintain side channels (Van Steeter and Pitlick 1998), improve substrate conditions for spawning and repro- ductive success, entrain organic debris and perhaps negatively affect small non-native fish abundance (Stanford 1994, McAda and Ryel 1999). Complex river segments created by lateral channel movement in un- confined reaches produce mesofeatures of the channel such as riffles, and midchannel or shoreline bars, there- by providing a mosaic of productive littoral zones (Thorp 1992) juxtaposed with other mesohabitats (pools, eddies, and backwaters) preferred by Colorado pikeminnow and other native fish (Osmundson et al. 1995). Although the base-flow stage can be manipu- lated to maximize preferred mesohabitats and habitat heterogeneity, periodic flood flows are required to cre- ate or maintain the channel features that give rise to these base-flow habitats. These effects are important functions of the natural flow regime in the upper Col- orado River. Restoring the frequency of bankfull events to near historical levels may, yet be possible through