USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5164

Back to Table of Contents

Influence of Drawdown on Sediment and DDT Transport and Deposition

Owing to a lack of sufficient high-turbidity events and suspended-sediment samples, only limited interpretations can be inferred from the turbidity–SSC correlations. First, with the exception of CGRO and to a lesser extent SFCO, the slope of the regression lines may be biased by a few individual samples such that additional samples might have produced substantially different estimates of concentrations, and therefore average and annual load estimates, on the basis of monitored turbidity. Second, estimation of concentration during several events from most sites is based on extrapolation of regression equations at turbidities beyond the maximums sampled, and so may not adequately account for differential sources or character of suspended sediment at those high concentrations and flows. Third, the turbidity and SSC sampling network in place during spring and summer 2002, the period of highest turbidity release from Cougar Reservoir, was not adequate to evaluate the potential effects on the mainstem McKenzie River.

Nevertheless, releases from Cougar Reservoir apparently were the primary source of suspended sediment to the mainstem during 2002–03, although full quantification of that contribution and its effect on deposition in the mainstem is not possible with the available data. Loads entering the reservoir were less than one-half of those exiting the reservoir during 2002 and 2003, and about 75 percent in 2004. Blue River Reservoir, although releasing a relatively high suspended-sediment concentration (annual discharge weighted average greater than 8 mg/L), has relatively low flow, so the loads were reduced in comparison to those from Cougar, and the water from upstream on the McKenzie River was relatively clear, so loads were kept small despite discharge being more than three times higher than those released from Cougar Reservoir. Data from 2003 and 2004 indicate that the reach between the South Fork and Vida was actually a source rather than a sink for suspended sediment. Deposition downstream of Vida (for example, in Leaburg Reservoir) could not be estimated on the basis of available data. Overall downstream sediment transport decreased during the course of the study as export from Cougar Reservoir declined with the equilibration of the exposed stream channels. Thus, even though large storm events in water year 2004 resulted in elevated loads at upstream, unregulated sites, the loads at CGRO and VIDA were reduced compared to previous years.

Furthermore, the flood-control reservoirs in the McKenzie River basin apparently act as traps for large particles and preferentially allow smaller, fine material to pass through to downstream reaches. In the South Fork, upstream erosion of the exposed deltaic sediments in Cougar Reservoir during the drawdown period caused mobilization of several decades’ worth of deposited fine sediment and clays. However, this also apparently occurs annually, albeit to a lesser extent, in Blue River Reservoir, where the annual winter reservoir drawdown results in an elevated baseline turbidity. Evidence for this sediment fractionation process is threefold: (1) SSC samples at regulated sites (CGRO and BLUE, below Cougar and Blue River Reservoirs, respectively) were composed of a substantially high percentage of fine material than samples from the unregulated sites (on the basis of a relatively small number of samples), (2) the slopes of the turbidity-SSC regressions are lower for the regulated sites than unregulated sites (SFCO and MRBO), indicating that a lower concentration of suspended sediment is being transported per unit of turbidity at those sites, and (3) infiltration bag results indicated that deposition of fine material was greater at regulated than at unregulated sites during the 1-year deployment period.

One factor that may partially explain the high accumulation of fine sediment in infiltration bags at regulated compared to unregulated sites is the relative frequency of bed-moving events. The storm of December 13, 2003, and possibly one or two others, caused mobilization of large rocks, trees, and some reconfiguration of gravel beds at the most unregulated sites (SFCO and MRBO). During such events, fine sediments that have been lodged in the interstices of gravel beds would likely be scoured from the streambed, remobilized, and transported downstream, thus reducing overall accumulation of fine sediment. At the regulated sites (BLUE and CGRO), scouring flows were not experienced and typically are not during even the largest midwinter storms, because the reservoirs’ intended function is to decrease stormwater peaks, metering the peak flow to downstream reaches more slowly. Therefore, deposited fine sediments in the reaches downstream of the flood-control reservoirs are not scoured annually and the fine material can better accumulate within the pore spaces. If reservoir outflows can be manipulated to produce periodic pulses of high flow during periods when downstream flooding is not a concern and when the reservoir outflows are relatively sediment free (that is, SSLs entering the reservoirs are small), it might be possible to mobilize the streambed enough to decrease accumulation of fine sediment. Blue River Reservoir typically released higher turbidity water than Cougar Reservoir during winter baseflow (nonstorm) periods (U.S. Geological Survey, 2003, 2004), which is one reason that deposition at BLUE may have been so much higher than at any of the other sites. Therefore, mobilization of the bed downstream of Blue River Reservoir might be particularly effective.

Deposition of fine material as determined from infiltration bags is best considered a relative measurement rather than an absolute deposition rate, because the streambed above and immediately surrounding the bag (an area with radius approximately three to four times the radius of the infiltration bag) is thoroughly disturbed during the burial process. The use of experimental rocks and gravel allows for more direct comparison between individual bags and between sites than would the use of native gravel. Given that the median diameter of the experimental gravel was mostly similar to that of native gravel, the porosity of the freshly installed experimental gravel is assumed to be substantially higher than in undisturbed native gravel. Thus, the pore spaces would be more susceptible to additional accumulation of fine sediment during the subsequent deployment period; therefore, deposition rates of fine material derived from infiltration bags probably are overestimates of actual, recent deposition rates in native gravels in the McKenzie River. By comparison, the freeze cores extracted by Gregory Stewart and others (Oregon State University, unpub. data, 2002) may provide an upper bound of long-term (and recent) deposition of fine material, but without resolution of recent deposition rates.

Findings from the infiltration bags indicated increased deposition of silts and clays (< 0.063-mm fraction) and decreased deposition of sand at regulated sites as compared to the unregulated sites in 2003–04. For the most part, this was similar to the findings from freeze cores collected in 2002 shortly after the reservoir drawdown (table 7, fig. 12), although the amount of sand in freeze cores was somewhat higher at all sites except SFCO (Gregory Stewart and others, unpub. data, 2002). In the freeze cores, the less-than-2-mm fraction was as high as 18 percent at the unregulated site MRBO as compared to about 14 percent at the regulated CGRO during 2002. However, they found only 3.8 percent of the less-than-2-mm fraction at SFCO. Freeze cores were not collected in Blue River.

The pattern of enrichment in silt and clay (<0.063 mm) in infiltration bags at regulated sites also was similar to the findings in freeze cores (table 7, fig. 12), although fine fractions generally were lower in infiltration bags relative to the overall sample than the fine fractions in the freeze cores. However, expressed as a percentage of the <2 mm fraction, the silt and clay proportions were quite similar in the two studies. The <0.063 mm fraction constituted the smallest fraction of the <2 mm material upstream of Cougar Reservoir and the highest downstream of Cougar Reservoir, and also was high downstream of Blue River Reservoir. In the Oregon State University study, variabilities in errors estimated from replicate freeze cores were 10 percent in clay and 50 percent in silt, and the fine fraction likely was underestimated by 8–50 percent, in part because of the dry-sieving method used (Sarah Lewis, Oregon State University, written commun., 2006).

Deposition of sand generally was decreased in infiltration bags at regulated sites during 2003-04 as compared to the freeze cores because sand was removed during the installation of the bags and was never replenished due to the trapping of sand and larger-sized particles in the reservoirs. In contrast, the freeze cores sampled native bed material that had been minimally disturbed in previous years, and high sand concentrations were likely to be the result of many years of accumulation. Although both the freeze cores and infiltration bag techniques attempt to evaluate sediment deposition in the bed, the two methods are fundamentally different. Freeze cores sample native sediments but with no ability to extrapolate to the time period during which deposition occurred, whereas infiltration bags used nonnative (experimental) sediments and had a well-constrained time of exposure. For the two methods to show similar patterns and generally similar fractions of sand and silt plus clay lends credence to the results of either method individually.

The deposition in the South Fork during 2003–04 probably was less than that during the drawdown period and subsequent winter, from spring 2002 to spring 2003, because average turbidities were not as high for a sustained period as immediately following the initial reservoir drawdown, during 2002 (fig. 6C). A reportedly thick, visible surface layer of settled, fine duff over the bed sediment at CGRO that was easily resuspended when disturbed and was observed during spring – summer 2002 (Doug Cushman, U.S. Geological Survey, oral commun., 2002) was not as evident during 2003–04. The same may be true in the mainstem McKenzie River downstream of the South Fork, which transported a prolonged plume of increased turbidity from the South Fork during 2002 (U.S. Army Corps of Engineers, 2003) but had only episodic periods of high turbidity following the largest events during 2003–04.

DDT and its metabolites were detected in deposited fine sediment in the streambed at CGRO and at VIDA, and concentrations were low. ΣDDx was likely mobilized only during large storms, and, on the basis of turbidity measurements, may have been detectable only a few times in the water column during the course of the study, if at all. The greatest downstream transport period of ΣDDx would likely have occurred during April–June 2002, when turbidities were greater than 50 FNU for most of May, with sporadic spikes of more than 100 FNU between April and June (appendix A); individual storms in January, February, and December 2003 were probably the only other times that detectable amounts of ΣDDx could have been resuspended and transported downstream, also likely at low concentrations less than 0.001 parts per billion. However, ΣDDx was not detected in samples collected during those events, so the transport probably was minimal.

From the available evidence following initial applications of DDT to portions of the upper McKenzie River basin to fight budworm activity during the 1950s (Dolph, 1980; Moore and Loper, 1980), Cougar Reservoir apparently acted as a trap for sediment and residual DDT and metabolites until the time of the construction project in 2002. Meanwhile, other portions of the upper McKenzie River basin, where DDT also likely had been sprayed in the 1950s and 1960s, acted as a source of ΣDDx to the lower basin between the 1950s and 2002 without a significant depositional basin to retain it upstream of Leaburg Reservoir. The detection of DDE in infiltration bags at VIDA, downstream of the South Fork, is consistent with this hypothesis. That no ΣDDx was detected in the mainstem upstream of Cougar Reservoir indicates that this upstream source probably has been mostly degraded or exported in the intervening decades.

Decreases in ΣDDx concentrations apparently have been greatest in streambanks and sediments that are routinely wetted— residual concentrations in organic horizons in samples from forested soils upslope from Cougar Reservoir were significantly greater than those in the bed sediments (U.S. Army Corps of Engineers, 2003). Concentrations in soil samples from the forest floor in nearby regions collected by the U.S. Forest Service were equal to or slightly higher than those in streambanks collected by the Corps of Engineers, but were only about one-tenth of the highest concentration measured by the Corps of Engineers, which also was from the forest floor (Dave Kretzing, U.S. Forest Service, McKenzie Bridge, Oregon, unpub. data, 2003). Once construction began in 2002 at Cougar Reservoir, with the accompanying erosion and downstream transport of deposited sediments, the reservoir apparently began acting as a source for ΣDDx to the South Fork and mainstem McKenzie River downstream. However, as a result of chemical and biological degradation and dilution, concentrations were relatively low and probably were not detectable even at sub-part-per-trillion concentrations, except when turbidities were highest during storm runoff and other construction-related events. This study did not examine ΣDDx in tissues of fish or other aquatic organisms, or top-level aquatic predators such as otters or raptors (see, for example, Henny and others, 1980), so the extent to which ΣDDx has bioaccumulated in the McKenzie River basin is unknown.

Back to Table of Contents