USGS - science for a changing world

Scientific Investigations Report 2008–5027

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008–5027

Back to Table of Contents

Discussion

Pesticide Occurrence in the Lower Clackamas River Basin

Pesticide occurrence in the lower Clackamas River basin was widespread, particularly in the tributaries, but also in the mainstem Clackamas River. Analyses of samples from four storm events identified some of the tributaries (Rock and North Fork Deep Creeks, for example) that contributed relatively high quantities (or loads) of pesticides to the Clackamas River upstream of drinking-water intakes. In some streams, pesticide concentrations exceeded aquatic-life benchmarks, and these findings can be used to focus and prioritize current and future efforts related to pesticide and land management, stream restoration, and salmon recovery.

The occurrence of pesticides in the Clackamas River basin is not unexpected given the large amount of urban and agricultural land in the drainage basin, where pesticides are frequently applied, and these results are similar to those from other studies. The most frequently detected pesticides in the Clackamas River basin—atrazine, simazine, metolachlor, diuron, and the organophosphate insecticides diazinon and chlorpyrifos—also were the most frequently detected pesticides in the Willamette River basin in Oregon (Rinella and Janet, 1998) and in other rivers across the United States (Gilliom and others, 2006). Several of the pesticides detected in the Clackamas River basin also were detected downstream in the Willamette River at Portland (21 pesticides during 2004–2005), and 5 have been detected downstream in the Columbia River (Jennifer Morace, U.S. Geological Survey, oral commun., 2006), but it is unclear how much the Clackamas River contributes compared with other major rivers in the Willamette River basin, such as the Molalla and Tualatin Rivers.

Pesticide occurrence in the Clackamas River is influenced by runoff from the tributaries and antecedent streamflow conditions in the mainstem prior to rainfall events. Streamflow in the lower Clackamas River is dynamic during the rainy season (fall to spring), responding to water releases from upstream dams, patterns and intensity of rainfall, snowmelt, and rain-on-snow events. Winter or spring storms can deliver precipitation to the lower basin during cold periods when moisture in the upper basin remains as snow. During such times, freezing levels may be low enough to reduce streamflow from the upper basin, which can result in less dilution water for the lower mainstem. At such times, and after heavy rainfall, pesticide concentrations in the lower Clackamas River can be elevated from tributary inputs in the lower basin.

The significance of these mostly trace-level concentrations of pesticides, however, is not yet known, but future studies could examine potential effects on aquatic life and human health. Identifying which compounds are present, when, and at what concentrations is a first step towards understanding the contamination potential posed by pesticides, and this information can be used to guide future pesticide reduction strategies to improve water quality in affected areas.

Potential Effects of Pesticides on Aquatic Life

Figure - refer to figure caption for alternative text description The Clackamas River supports the last remaining wild coho salmon stock in the Columbia River basin. (Photograph by Tim Shibahara, Portland General Electric.)

Many pesticides have the potential to harm nontarget organisms, especially benthic invertebrates, fish, amphibians, and various stream microbes (Nowell and others, 1999). Biota in the lower Clackamas River and the lower-basin tributaries are exposed to pesticides, sometimes at concentrations high enough to exceed aquatic-life benchmarks. Aquatic life in the Clackamas River and some of its tributaries include various anadromous and resident fish species, amphibians, plants, and other organisms. Declines in some fish populations, including winter steelhead, spring chinook, and coho salmon have resulted in their being included on the Endangered Species List (National Marine Fisheries Service, 2006). Potential explanations for such declines have included overharvesting of fish, hydroelectric dams, poor-quality stream habitat, and degraded water quality from pesticides and other contaminants. Understanding the potential cumulative effects of the combined influences on aquatic life is challenging, and the understanding of the effects of pesticides alone, for example, is not complete because most toxicity research focuses on single compounds, not mixtures. The chemical and (or) physical conditions in streams may affect aquatic life through mechanisms related to stress (and sometimes-resulting disease), feeding, and reproduction, but such cumulative effects are not yet well understood.

There also exists potential for sediment-bound pesticides to affect benthic organisms. This study examined the occurrence of pesticides dissolved in water, not those associated with streambed sediments. Some pesticides such as pyrethroid insecticides, for example, may adhere to sediments and cause toxicity to benthic invertebrates (Amweg and others, 2006). In some places, pyrethroids are being used as an alternative to organophosphate insecticides, which are more toxic to humans than pyrethroids. Because the pyrethroids insecticides accumulate in sediments, benthic organisms may be exposed to elevated concentrations in low gradient pools and riffles affected by sedimentation. Sediment and turbidity levels were high in many of the Clackamas River basin tributaries during the storm sampling in 2005 (appendix C, table C4) due to erosion of stream banks, resuspension of sediment from the streambed, and nonpoint source runoff from the drainage basin. It is not known whether such sedimentation causes sorption of pesticides that tend to adhere to sediment particles, but such a hypothesis could be examined with further study.

Previous studies found invertebrate assemblages in upper Noyer and North Fork Deep Creeks to be severely impaired (Cole and Hennings, 2004). The invertebrate assemblage quality was poorest in the headwaters, and improved somewhat at the downstream sites in the lower forested canyon reaches of these and other streams, including Rock and Richardson Creeks. Specific conductance also was lower at the downstream sites, indicating fewer dissolved ions in water compared with the upstream sites and, potentially, improved water quality (Cole and Hennings, 2004). The water quality at sites in these lower reaches may be affected by low-ion content ground water, which might help decrease contaminant concentrations (and lower water temperatures), but improved physical habitat quality probably also benefited benthic invertebrate assemblages. Headwater streams in the Noyer Creek and North Fork Deep Creek basins have less intact and narrower riparian zones, with some concentrated agricultural and rural residential areas (Cole and Hennings, 2004). Downstream reaches in forested canyons have greater amounts of intact riparian vegetation and contain cobble-substrate riffles that are more suitable for benthic invertebrates (for example, see cover photograph of Noyer Creek). Dewberry and others (1999) found the diversity of aquatic insect assemblages in Rock and Sieben Creeks to be suppressed by factors including habitat impairment, and in light of the current study findings, pesticides also may affect benthic invertebrates and other aquatic life in these streams.

Pesticides occasionally exceeding their respective USEPA aquatic-life benchmarks in this study included the insecticides diazinon, chlorpyrifos, endosulfan, and azinphos-methyl. The diazinon concentrations in storm samples collected from Carli, Sieben, and Rock Creeks, for example, exceeded the USEPA aquatic-life criterion for benthic invertebrates of 0.1 μg/L by a factor of as much as 2.5. Diazinon and other organophosphate insecticides are designed to impair nervous system function through inhibition of the enzyme acetylcholinesterase (Nowell and others, 1999). Exposure to these compounds may inhibit the activity of this enzyme in organisms such as benthic invertebrates, amphibians, and fish. Diazinon impairs predator avoidance behavior and homing ability in Chinook salmon at concentrations of 1 and 10 μg/L, respectively (Scholz and others, 2000). Although these concentrations are much higher than those detected in the Clackamas River basin, the effects of sustained or multiple exposures to diazinon are not well understood. Diazinon was detected in the lower Clackamas River on six occasions (fig. 15), and peak concentrations may not have been measured by this study, given the small number of samples collected during storms. Although diazinon sales for residential usage ended on December 31, 2004, diazinon in storage likely has been used since then; diazinon continued to be detected into 2005.

Another compound that exceeded its criterion was the organophosphate insecticide chlorpyrifos. Chlorpyrifos concentrations in Noyer and North Fork Deep Creeks were 0.14 and 0.17 μg/L, respectively, and exceeded the USEPA chronic and acute aquatic-life benchmarks of 0.041 and 0.083 μg/L (table 6). The chlorpyrifos concentration in extract from an SPMD deployed in North Fork Deep Creek was the highest among all 28 sites sampled for the EUSE study (Ian Waite, U.S. Geological Survey, written commun., 2007) and highest among all sites nationally (Bryant and others, 2007). Because the SPMDs were deployed for an extended period (30 days), the high value suggests a relatively high average concentration over time in North Fork Deep Creek compared with the other sites. This is consistent with the results from the 2005 pesticide storm event samplings, when chlorpyrifos was detected in two-thirds of the sites in the Deep Creek basin (appendix C, table C1). In 2005, the chlorpyrifos concentrations in North Fork and Noyer Creeks (up to 0.17 and 0.14 μg/L, respectively) were substantially greater than the highest value in 2000 (0.056 μg/L in Rock Creek; Carpenter, 2004). In 2005, chlorpyrifos was detected in only one other tributary, Trillium Creek at a concentration of 0.005 μg/L. Chlorpyrifos also exceeded non-USEPA benchmarks in the mainstem Clackamas River on two occasions (table 6).

The prevalence of pesticide mixtures in Clackamas River basin streams presents challenges for understanding how aquatic life in these streams might be affected. Some stocks of salmon, winter steelhead, and sea-run cutthroat trout continue to use tributaries, including Eagle, Clear, Deep, and Rock Creeks to spawn and rear, and are sometimes exposed to multiple pesticides. Some of the tributaries, such as Rock and Tickle Creek, still support coho salmon populations despite threats from a variety of potential contaminants, including pesticides.

One of the complicating factors in determining safe exposure levels for aquatic life for pesticides is that laboratory studies typically involve only a single compound and do not consider additive or possibly synergistic effects of multiple pesticide exposure. Although it might be logical to assume that two pesticides with the same mode of action (such as the orthophosphate insecticides chlorpyrifos and diazinon, which inhibit the same acetylcholinesterase enzyme) would act in an additive fashion, certain pesticides may affect the toxicity of others through various physiological mechanisms that are just beginning to be understood. For example, the toxicity of orthophosphate insecticides was shown to increase markedly by simultaneous exposure to the herbicide atrazine (Pape-Lindstrom and Lydy, 1997). Other studies (Kao and others, 1995) have found a potential mechanism for this: atrazine exposure stimulates Cytochrome P450 and general esterase activity in insects that increases production of oxon degradates such as diazinon-oxon and malathion-oxon from the parent compounds. Ironically, the degradation to oxon compounds produces a more toxic degradation compound. A recent study of frog tadpoles found that oxon derivatives such as diazinon-oxon, chloroxon, and maloxon (degradates of diazinon, chlorpyrifos, and malathion, respectively) were between 10 and 100 times more toxic than the parent insecticide compounds (Sparling and Fellers, 2007).

The PTI values suggest that benthic invertebrates were more at risk than fish at most sites, and it is unclear how other aquatic life may be affected. Benthic invertebrate assemblages were highly degraded in lower Tickle Creek during the EUSE study (Ian Waite, U.S. Geological Survey, written commun., 2007). Good habitat quality was found in the lower mainstem of Tickle Creek during the EUSE biological and habitat survey, and included cobble riffles in a mostly forested canyon with abundant riparian vegetation, so some other factor, possibly exposure to pesticides, wastewater-treatment plant effluents, or other contaminants may be affecting benthic invertebrates in lower Tickle Creek.

According to the PTI (fig. 13), fish assemblages in Tickle Creek were most at risk from the organochlorine insecticide endosulfan, which was detected at a total concentration of 0.11 μg/L, which included 0.067 μg/L endosulfan I (alpha endosulfan) and 0.039 μg/L endosulfan II (beta endosulfan). Note that these two compounds were not distinguished in toxicity tests used for the PTI (appendix D, Table D1). Although the total endosulfan (I + II) concentration in Tickle Creek (0.11 μg/L) was less than the Oregon DEQ benchmark for acute exposure (0.22 μg/L), it is greater than the Oregon DEQ chronic benchmark of 0.056 μg/L and benchmarks suggested by NAS/NAE and Canada (table 6). Although the storm-runoff samples collected for this study probably reflect short duration exposure (making acute benchmarks more appropriate than chronic benchmarks), it is not clear whether peak concentrations were captured during sampling or how long such high concentration pulses persist. The endosulfan concentration in Tickle Creek might be more indicative of chronic exposure levels during periods of active runoff. If this were true, the lower chronic benchmark would be more appropriate. Repeated or prolonged exposures to elevated concentrations of endosulfan (or other contaminants) might have contributed to the lackluster condition of juvenile coho salmon and quality of benthic invertebrates in Tickle Creek found during the 2004 EUSE study (Ian Waite, USGS, oral commun., 2004), but more study would be needed before specific conclusions regarding connections to pesticides can be made.

Doane Creek, a tributary of North Fork Deep Creek, had the second highest PTI for fish (fig. 13), largely due to the occurrence of the potent orthophosphate insecticide azinphos-methyl (AZM), which was detected at a concentration of 0.21 μg/L. AZM is highly toxic to freshwater fish and invertebrates—the risk assessment for AZM (U.S. Environmental Protection Agency, 2001) states that

“…if [AZM] enters a water body in sufficient quantities, it can result in death and reproductive effects in aquatic organisms, and there is also potential exposure and risk to birds, mammals, and bees from direct spray, drift, and surface AZM residues.”

Although the AZM QA spike results in the current Clackamas River basin study suggested a positive bias for AZM of about 17 percent (from four QA blank water spikes), the AZM concentration in Doane Creek, when corrected for this bias, would be reduced 17 percent to 0.179 μg/L. This value—the only AZM detection during the study—exceeded the USEPA aquatic-life acute benchmark for benthic invertebrates (0.06 μg/L) and approximates the USEPA aquatic-life acute benchmark for fish (0.18 μg/L) (table 6).

Although not examined during this study, exposure to pesticides or other contaminants may cause sublethal effects on aquatic life, such as deformities during early developmental stages or diminished reproductive success from disruption of endocrine system function. Developmental deformities in frog gonads, for example, have been documented in laboratory experiments by Hayes and others (2002a) from exposure to the herbicide atrazine. Twenty percent of male frogs studied developed deformities at atrazine concentrations as low as 0.10 μg/L. This concentration is lower than the USEPA MCL allowable in drinking water (3 μg/L) by a factor of 30. Atrazine is the most widely used pesticide in the world and is the most frequently detected pesticide in streams nationwide (U.S. Geological Survey, 1999); it was the second most frequently detected pesticide in the Clackamas River basin—detected in nearly one-half of the samples collected (table 3). Sublethal effects on aquatic life, such as impaired reproduction and development from exposure to pesticides have been documented by numerous laboratory and field studies. Cope and others (1970) reported delayed fish spawning (in bluegill) from exposure to 2,4-D; Von Westernhagen and others (1989) showed reduced fish fertilization from exposure to dieldrin; Choudhury and others (1993) and Baatrup and Junge (2001) demonstrated reproductive system disruption in fish from exposure to carbaryl and p,p’-DDE, respectively; Hayes and others (2002b) found developmental irregularities in frogs from low-level exposure to atrazine. Hayes and others (2006) also found that a nine-pesticide mixture had profound effects on the development of frog larvae by delaying metamorphosis. Because frogs took longer to reach maturity, they were smaller as adults, presumably because they used more of their energy reserves before reaching a feeding age than the control group. Colborn and others (1993) reported endocrine system disruption in wildlife and humans from exposure to pesticides. Pesticides also may affect fish behaviors, including predator avoidance and homing (Scholz and others, 2000), swimming (Matton and Laham, 1969), and feeding (Bull, 1974). One of the challenges in understanding toxicity is that in the past, most studies were designed to detect effects on growth or survival (LC50 tests, for example)—not on sublethal effects such as those described above. Determining such effects on aquatic life is complicated by the multiple-compound exposures, by variations in concentrations (including high-level pulses that occasionally occur), and by interactions with streambed sediment where pesticide residues may accumulate over time in areas affected by erosion.

Pesticides in Source and Finished Drinking Water

Of the 57 pesticides or degradates detected in tributary streams draining into the Clackamas River upstream of the treatment plant intakes, 26 were detected in the mainstem Clackamas River or in samples of source water from the study water treatment plant on the lower Clackamas River. Of these, 15 pesticides and degradates—11 herbicides, 3 insecticides, and 1 fungicide—were detected in samples of finished drinking water from the study drinking water-treatment plant (table 3).

Only one of the four water-treatment plants on the lower Clackamas River was examined during the SWQA drinking water study. Consequently, these results characterize just a portion of the water supply derived from the Clackamas River, from a water treatment plant that uses direct filtration, one of four treatment technologies used by the municipal water providers, along with sand filtration, membrane filtration, and conventional water treatment.

The finding of pesticides in finished drinking water derived from the Clackamas River is consistent with other studies of medium to large sized integrator-type rivers and reservoirs conducted in other parts of the United States. A recent pilot study by the USGS and USEPA examined raw and finished drinking water from 12 water-supply reservoirs across the country found that conventional water treatment did not completely remove pesticides and degradates, and that 9–30 compounds were detected in finished water in each area (median number of pesticide compounds detected was 23) (Blomquist and others, 2001; Coupe and Blomquist, 2004).

Although concentrations of pesticides in finished drinking water derived from the lower Clackamas River were all well below USEPA standards and other human-health benchmarks, current benchmarks do not account for multiple compound exposures. In addition, some of the compounds are likely or possible carcinogens, endocrine disruptors, and (or) acetylcholinesterase inhibitors (table 9), and may warrant further study and monitoring.

The pesticides having the highest human health Benchmark Quotients (BQs) in this study were diuron, ethoprop, simazine, and pronamide (table 8). Maximum concentrations of pesticides in finished drinking water were less than their respective human-health benchmarks by a factor of between 11 and 350,000 (fig. 14). Three of the compounds detected in finished water—diazinon-oxon (a degradate of diazinon), CIAT (deethylatrazine, a degradate of atrazine), and the insect repellent DEET—have no established MCL or HBSL benchmark for which to compare (table 4) because the toxicity data needed to calculate HBSL values are lacking. CIAT was frequently detected in tributary and mainstem samples, occurring in 31 percent of samples overall (table 3). CIAT is formed in the environment from the degradation of atrazine, a commonly used herbicide. In the USEPA OPP Human-Health Risk Assessment for atrazine, the toxicity of CIAT was considered as equivalent to that of the parent compound atrazine (U.S. Environmental Protection Agency, 2003b). The consensus-based protocol for HBSL development (Toccalino and others, 2003), however, does not currently permit the use of toxicity data from a parent compound to calculate a HBSL for a degradate. The MCL for atrazine is 3 μg/L, which is 600 times higher than the CIAT concentration (0.005 μg/L) detected in the one sample of finished water.

Detections of pesticides in finished water samples collected in 2004 and 2005 differ from previous results from 2000–2001 (Carpenter, 2004) and from the routine compliance monitoring by the water providers over the past several years that has detected no pesticides in finished water. One of the possible explanations for this difference is that prior to 2004, all of the USGS samples were processed without the use of a dechlorinating agent to stop the chlorine activity and subsequent degradation of pesticides (see “Methods”). This quenching procedure was added in 2004 for the second phase of the SWQA study (Carter and others, 2007). Because previously collected finished-water samples did not receive a dechlorinating agent, pesticides that may have been present in the samples could have been oxidized by residual chlorine prior to being analyzed at the laboratory. In addition, laboratory methods used during the USGS studies had considerably lower detection limits for pesticides compared with the routine compliance monitoring.

A comparison of pesticide concentrations in a limited number of samples with and without the dechlorinating agent provides some indication of the potential effects of chlorine on many pesticides, with fewer compounds being detected and at lower concentrations in the unquenched samples than in the quenched samples (appendix A, tables A2 and A3). Many of the percent recoveries for quality control spiked samples were zero for unquenched drinking-water samples, indicating oxidation of these pesticide compounds. Based on these data, chlorination may be effective at decreasing concentrations of certain pesticides in finished water, although more analyses are needed to verify these results. Many pesticides, however, transform into degradates through oxidation by chlorine in public distribution lines and in chlorinated drinking-water samples prior to analysis. The degradation of pesticides into degradates forms new compounds that are generally less toxic, but in some cases, such as for diazinon-oxon (degradate of the orthophosphate insecticide diazinon) and 3,4-dichloroaniline (degradate of the herbicide diuron), the degradates have greater toxicity than the parent compounds. Although some pesticide degradates were examined during this study, the full suite of pesticide degradates that could form from the 63 pesticides detected were not characterized.

The occurrence of simazine and diuron in finished drinking water is consistent with their high rates of detection in the lower-basin tributaries and in the lower Clackamas River mainstem. These two herbicides occurred in 50–70 percent of tributary samples, sometimes at elevated concentrations (1–2 μg/L). Simazine is a selective herbicide used to control broadleaf weeds and annual grasses on nursery and field crops, including Christmas trees, hazelnuts, and cane berries. Simazine also may be used to control aquatic plant growth in farm ponds, swimming pools, and fish hatchery ponds (Extension Toxicological Network, 1996).

Diuron was detected in finished drinking water on four occasions at a maximum concentration of 0.18 μg/L, which was 11 times less than the low HBSL value (table 8). Diuron also was frequently detected during the study, occurring in 44 percent of samples. Diuron was first registered for use in 1967. It is applied as a pre- and post-emergent herbicide, with approximately two-thirds of its use on agricultural crops and the remaining third on noncrop areas such as along roads and other right-of-ways (table 10). It is also used to control mildew, as a preservative in paints and stains, and to control algae in commercial fish production, residential ponds and aquariums (U.S. Environmental Protection Agency, 2003d).

One of the primary degradates of diuron, 3,4-dichloroaniline (DCA), may warrant further study given the frequent occurrence of diuron and the relative lack of data for DCA because it was analyzed for in a small number of samples (18 samples analyzed compared with 93 for diuron) (table 3). DCA was frequently detected during the EUSE study, occurring in two-thirds of samples collected from Tickle and North Fork Deep Creeks (appendix C, table C2). A recent review of the environmental toxicity and degradation of diuron by Giacomazzi and Cochet (2004) indicated a greater toxicity from DCA compared with diuron. The USEPA has completed an “Effects Determination” for diuron to evaluate exposure of endangered and threatened salmon and steelhead species to diuron and the potential for indirect effects on these fish from damage to their aquatic plant cover in water bodies in California and the Pacific Northwest. The USEPA concluded that agricultural crop uses of diuron will not have effects on Pacific salmon and steelhead, except at certain high-use rates on walnuts, filberts, and peaches, and that noncrop uses may affect 25 salmon and steelhead evolutionarily significant units (ESUs). For those ESUs that may be affected by diuron use, the USEPA will consult with the National Marine Fisheries Service to determine what protective measures are needed (U.S. Environmental Protection Agency, 2003d).

Another pesticide detected in finished drinking water during the September 2005 storm event was the insecticide/nematocide ethoprop, at a concentration of 0.006 μg/L, which was 175 times less than the low HBSL value of 1 μg/L (table 8). Ethoprop is classified by USEPA as a likely human carcinogen (table 9). This low HBSL corresponds to a 1-in-1 million cancer risk for ethoprop, which at higher concentrations is a likely human carcinogen (table 9). This insecticide was detected in 18 or one-third of storm samples, with nearly all of the detections in the lower-basin tributaries (table 3). Although ethoprop was detected in the Clackamas River and in finished drinking water (once), its occurrence in the mainstem Clackamas River was not fully characterized by this study as it was analyzed only during the four storms sampled for the USGS/CWMG studies, not during the routine sampling for the SWQA study.

The low-level detections of pesticides and degradates inform watershed managers about their presence and gives some idea of their respective levels in source and finished drinking water from one of the four treatment plants on the lower Clackamas River. The presence of pesticides and degradates raises questions regarding the potential for effects on aquatic life in the lower-basin tributaries and the lower Clackamas River, and on human health from exposure to low-level concentrations of pesticides. It is uncertain, for example, what the cumulative effects might be on human health from simultaneous exposure to multiple pesticide compounds, and current regulations do not yet consider the spectrum of interactions that may occur among pesticides and other contaminants that may be present.

Potential Pesticide Sources

The pesticides detected in the Clackamas River basin come from a wide variety of sources. The diverse land use in the study area and unpredictable water management (pumping, irrigation, collection, and release) make it challenging to identify sources. Pesticide applications are made along roads and on agricultural fields, harvested forests and urban landscaping, especially in the lower Clackamas River basin where agricultural and urban land is concentrated. One survey estimated that at least 116 have been used in the Clackamas River basin (Hassanein and Peters, 1998), but the actual numbers may be much higher given that there are approximately 11,000 pesticide products registered for use in Oregon.

A more recent report on pesticide occurrence in the Clackamas River basin estimated that as much as one-half of the agricultural pesticide use could be on nursery and greenhouse crops, with lesser amounts applied to pastureland, Christmas trees, alfalfa and hay fields, hazelnut orchards, and grass seed fields (Carpenter, 2004). Findings from the current study also suggest that nursery, floriculture, and greenhouse operations continue to be a significant source of pesticides in the Clackamas River basin.

A Source Water Assessment was conducted in 2003 by the Oregon Department of Environmental Quality and the Oregon Department of Human Services (2003) with guidance from the Clackamas River Water Providers identified 961 individual or area-wide potential contaminant sources upstream of the drinking-water intakes in the lower Clackamas River, with 445 sources posing a moderate-to-high risk. Specific sources include 55 high density housing areas, 33 high maintenance lawn areas, 6 golf courses, 3 wastewater-treatment plants, 173 irrigated and 200 nonirrigated agricultural operations, 22 pesticide/fertilizer storage areas, and 35 ponds, some of which collect irrigation tail water from agricultural land.

The collective influence of land use, topography, drainage network, and patchy nature of storms contributes to producing variable runoff of water, sediment, and pollutants from these basins during storms. Drainage basins affected by urbanization—Carli, Cow, and Sieben Creek basins, and other streams around Estacada, Boring, and Sandy—collect and transfer stormwater to streams through drains, culverts, and other engineering conveyance systems. Large amounts of overland runoff with high levels of suspended sediment also may transport dissolved and sediment-bound contaminants to the lower Clackamas River upstream of drinking-water intakes. Basins with relatively steep topography and large amounts of impervious areas—Sieben, Rock, and Richardson Creek basins, for example—respond quickly to rainfall and are often highly turbid after storms (appendix C, table C4).

Figure - refer to figure caption for alternative text description Tractor and boom sprayers are used to apply pesticides on a variety of agricultural crops in the Clackamas River basin. (Photograph taken May 2, 2003.)

The sources of pesticides detected in the Clackamas River basin are difficult to identify because most have multiple uses (table 10). Furthermore, data collected for Oregon’s Pesticide Use and Reporting System (PURS) will be at a relatively coarse scale, and not specific enough to locate sources. Pesticide applications in the Clackamas River basin will likely be incorporated into a larger report for the entire Willamette River basin. The PURS data will be useful, however, for identifying potentially important chemicals not currently being analyzed. Only a small fraction of the 11,000 pesticide products registered for use in Oregon were tested during this study, which makes pesticide use data especially helpful for designing pesticide monitoring plans.

The transport of pesticides from their target areas to waterways occurs from several sources, including: (1) surface runoff from urban and rural areas, agricultural fields, roadside ditches and culverts (which are sprayed directly for vegetation control), greenhouses and nurseries, and other source areas, (2) erosion of soils treated with chemicals, especially pesticides with high Koc values (appendix E) that tend to adhere to sediment, (3) atmospheric drift, and (4) ground water, whereby pesticides travel into aquifers or move through shallow flow paths to streams.

Pesticides used on the landscape may be transported into streams, exposing aquatic life to pulses of toxic runoff and also may travel to drinking-water intakes as was demonstrated in this study. Although highly soluble compounds—those with high water solubility or low Koc values in appendix E—tend to move from the land at relatively high rates, additional factors also may explain the fate of pesticides in the environment. These include the chemical half life (rate of breakdown in water or soil), pattern and extent of chemical use, and physical or hydrologic characteristics of the drainage basin.

Many studies have shown that while streams and rivers are most vulnerable to pesticide contamination and tend to have higher pesticide concentrations, ground water also may contain pesticides. This source of pesticides merits careful attention because ground water contamination is difficult to reverse. The importance of ground water as a pesticide source for surface waters in the Clackamas River basin is not, however, known but may be important in certain areas where surface runoff containing pesticides recharges ground water. Future studies examining the surficial geology and ground-water quality beneath nurseries, golf courses, and urban areas could begin to characterize pesticide concentrations in some of the high-risk areas identified during the Source Water Assessment (Oregon Department of Environmental Quality and the Oregon Department of Human Services, 2003).

Figure - refer to figure caption for alternative text description Landscaping ornamentals are grown in a Rock Creek basin nursery. (Photograph taken January 2, 2003.)

Concentrations of some compounds, including CIAT (deethylatrazine), metalaxyl, and simazine were somewhat elevated in samples collected during low-flow conditions in the Deep Creek basin during the EUSE study, particularly in North Fork Deep and Tickle Creeks. The pesticide detections in these streams during nonstorm conditions indicate a continuous non-storm-derived source such as ground-water inflows, irrigation return flows, or wastewater-treatment plant effluent. Both North Fork Deep and Tickle Creeks receive effluent from the community of Boring and the city of Sandy, respectively. Currently, Sandy’s wastewater is used to irrigate nursery stock during the dry months, and although wastewater inputs to surface water are reduced, some amount may enter ground water.

One of the most commonly detected pesticides in the Clackamas River basin was the herbicide glyphosate, the active ingredient in many household, agricultural, and forestry herbicide products such as RoundUP™, Rodeo™, and Accord™ (table 10). The average glyphosate concentration in tributary samples was 3.5 μg/L, and the highest concentration was 45.6 μg/L in middle Rock Creek at 172nd Avenue during the September 2005 storm. Glyphosate has a relatively high water solubility (900,000 mg/L) and moderate half life in soil (47 days) (appendix E). Most glyphosate products contain surfactants that are designed to make the chemical spread and stick to surfaces, and therefore, have a low tendency to runoff or enter ground water despite its high water solubility. Although surfactants may retard movement of glyphosate, it may be transported to streams on sediment particles. Although sediment-associated transport of glyphosate to streams may explain its frequent occurrence during storms—71 percent of samples contained glyphosate (table 3), it is also one of the most widely used herbicides.

Figure - refer to figure caption for alternative text description Irrigated container nursery in the Sieben Creek basin. (Photograph taken July 10, 2003.)

Agriculture—About 100,000 acres of land are used for agriculture in Clackamas County. In the Clackamas River basin, agricultural land is concentrated on the high plateau between the Clackamas and Sandy Rivers (pl. 1). Some agricultural land also is located adjacent to or within the floodplain of the Clackamas River. Although diverse crops are grown, pastureland, hay fields (mostly alfalfa), nurseries, and greenhouses make up about 65 percent of the agricultural land in the basin. Clackamas County also is one of the top Christmas tree producing counties in the country. According to the National Agricultural Statistics Service (2002), 18 herbicides, 12 insecticides, and 4 fungicides are used on Christmas trees in Oregon. Several of these pesticides (including atrazine, hexazinone, simazine, triclopyr, and chlorpyrifos) were detected in the Clackamas River basin during this study, but the individual contribution from each of the types of agriculture was not part of the study design.

The greatest amount of agricultural land is located in the Deep Creek basin, which is drained by tributaries including Noyer, North Fork Deep, and Tickle Creeks. These streams drain basins containing the highest percentage of agricultural land—approximately 33 to 47 percent of the total basin area was agricultural land (table 1). Rock and Richardson Creek basins also contain substantial amounts of agricultural land (about 31% each), along with some rural residential and urban land. Deep, Noyer, Richardson, and Rock Creeks have cut into a large plateau in the northern part of the lower Clackamas River basin, forming drainages that are relatively flat in the headwaters and descend through steep forested canyons before joining the mainstem Clackamas River to the south. Streams draining this part of the basin contribute sediment, nutrients, and pesticides to the Clackamas River, particularly during storms.

Nursery, floriculture, and greenhouse crops—In 2003, there were 12,700 acres of nursery land in Clackamas County (Oregon Department of Agriculture, 2005), with much of the acreage located within the Clackamas River basin (pl. 1). In 2000, the top agricultural commodity in Oregon was production from nurseries and greenhouses and sales in Oregon have increased from $347 million in 1993 to $844 million in 2004 (Oregon Department of Agriculture, 2005). The results of this study show that Deep Creek and its tributaries, including Noyer, North Fork Deep, Tickle, and Rock Creeks are among the largest contributors of pesticides to the Clackamas River during storms. All these streams have potential to be impacted from various sources, including large nursery operations. The contributions from nursery operations relative to other types of agriculture, or from rural residential use in these areas were not, however, specifically examined during this study.

Figure - refer to figure caption for alternative text description Herbicides are used for weed control in Christmas tree plantations. (Photograph taken July 10, 2003.)

A 2003 survey of nursery and floriculture operations (National Agricultural Statistics Service, 2004) reported about 275 herbicides, insecticides, and fungicides applied to nursery and floriculture crops in program States during 2003. The survey reported aggregated data on the types and amount of chemicals used in Oregon and four other states—California, Michigan, Pennsylvania, and Florida. Some data in the National Agricultural Statistics Service survey were reported by State, including the number of operations, total number of pesticides used, and qualitative information on patterns in pesticide use. Applications of pesticides occurred in open areas and inside greenhouses to control various pests. Most pesticides were applied manually in 2003 (80 percent) using backpack or power hydraulic sprayers, and about 20 percent of applications for all States were made using a tractor and boom sprayer. There were more than 25,000 reports of chemical usage from about 900 nursery and floriculture operations in Oregon during 2003, some of which are in the Clackamas River basin. Nearly 600,000 lbs of active ingredient were applied in Oregon for agricultural purposes, with 40 percent of nursery and floriculture operations applying pesticides based mostly on a preventative schedule (National Agricultural Statistics Service [NASS], 2004), and not in response to an active threat. The 2003 NASS survey determined that 51 percent of operators in the program States actively surveyed for pests, and that in Oregon, 11 percent of operators use pheromone traps as part of an integrated pest management program (IPM) to monitor for pests. IPM potentially allows for early detection and treatment of pest infestations, which can prevent loss of crops. Early detection also may reduce the amount of chemical required to treat the spread of a particular infestation, which over the long term, may reduce the need for preventative applications in the future.

Figure - refer to figure caption for alternative text description Cane berries are an important agricultural commodity in Clackamas County. (Photograph taken July 10, 2003.)

Ninety-two percent of current-use pesticides detected in the Clackamas River basin (table 10) were on the National Agricultural Statistics Service list of pesticides applied to nursery, greenhouse, and floriculture crops. Given the potential for extensive pesticide use on nursery and greenhouse crops in the Clackamas River basin, the results presented in this report may underestimate the relative contribution from nursery operations because many of the compounds used at nurseries were not analyzed. Acephate, for example, was the most commonly used insecticide for nursery operations (National Agricultural Statistics Service, 2004), yet it was not tested during this study.

Glyphosate, oxyflurfen, and oryzalin were the most commonly applied herbicides, and the fungicide of choice was chlorthalonil. Chlorthalonil was detected just once, at the mouth of Noyer Creek in May 2005, at a concentration of 0.26 µg/L. The insecticide imidacloprid was detected in Noyer Creek downstream from Highway 212 (September 2005) at a concentration of 4.5 μg/L (appendix C, table C1). According to the NASS survey, this compound was used by 20 percent of nursery operations in all program States (National Agricultural Statistics Service, 2004). Imidacloprid is used as a less toxic alternative to orthophosphate insecticides to control sucking insects such as aphids in various nursery and floriculture crops (National Agricultural Statistics Service, 2004). Scholz and Spiteller (1992) reported that imidacloprid breaks down faster in soils with plant ground cover compared to fallow soils, although there is potential for imidacloprid to move through sensitive or porous soil types with large amounts of gravel or cobble. Because of its moderate water solubility (10 mg/L) and relatively long half life in soil (48–190 days; appendix E), imidacloprid can be transported in irrigation runoff, especially from steep slopes.

Figure - refer to figure caption for alternative text description Herbicides are applied for vegetation control along roads in the Clackamas River basin. (Photograph taken July 10, 2003.)

Many nurseries collect irrigation tail-water in ponds, and re-use of water during the summer irrigation season is a common practice. In many cases, nursery and farm ponds are formed from small impoundments on small streams, and unlined ponds may lose some water to the surrounding area, recharging the shallow ground-water system. This water may enter streams through springs or in upwelling areas downstream. Ponds typically are drained in the autumn, before the onset of heavy rains. Releases from these ponds could be an important source of pesticides to Clackamas River basin tributaries and the mainstem Clackamas River, but more information is needed to quantify their contributions. Determining the factors influencing breakdown rates for certain compounds, including high-use compounds or those with relatively high toxicity—including endosulfan, diazinon, and chlorpyrifos, for example, could improve management of this potential pesticide source.

Forestry—The use of pesticides on forestland includes selective control of insect pests and invasive plant species in problem areas, and broad scale herbicide applications during site preparation following harvest on private timberland to control under story vegetation during the early stages of regeneration. Applications of herbicides also are used to control noxious nonnative plant species such as Himalayan Blackberry, Scotch Broom, English Ivy, Purple Loosestrife, and Japanese Knotweed. These invasive plants have the potential to displace native vegetation, reduce replanted tree growth, alter habitat, reduce forage for grazing animals, and cause economic damage and other effects.

Efforts to remove Japanese Knotweed from riparian areas in the Deep Creek and other Clackamas River basin streams have included plant stem injection with Rodeo™, a formulation of glyphosate that does not contain surfactants present in RoundUP™ that are more toxic to some aquatic life than glyphosate itself (Pesticide Action Network, 1996). In addition, there are an estimated 420,000 acres of National Forest System lands in the Pacific Northwest Region Six that are currently infested with invasive plants. The official Record of Decision, which includes Federal lands within the Clackamas River basin, includes provisions for using herbicides to control invasive plants (U.S. Forest Service, 2005).

The amount of pesticides applied in the Clackamas River basin on private, State, and Federal forestland is not readily available, but pesticide use on the Mount Hood National Forest, which comprises most of the Federal land in the upper Clackamas River basin, is relatively insignificant. The herbicide glyphosate is used sparingly to control invasive plants—in 2006, 2 acres were treated with 0.25 gal of glyphosate (Rodeo™) to control spotted knapweed along USFS Road 46 in the upper Clackamas River basin (Mark Kreiter, U.S. Forest Service, written commun., 2007). Pesticides also may be used to control insect pests on the forest, and in 1989, an outbreak of western spruce budworms (Choristoneura occidentalis) in the upper Clackamas River basin was treated with the biological insecticide Bacillus thuringiensis kurstaki (BT), which was aerially applied to more than 7,595 acres (Sheehan, 1996). This insecticide was not among those analyzed during the USGS study. No other pesticides are applied in the Clackamas River basin on the Mount Hood National Forest at this time (Jennie O’Connor, U.S. Forest Service, written commun., 2007).

Private forestland in the Clackamas River basin occurs primarily in the lower basin, especially in the Eagle and Clear Creek basins, but also in other basins (for example, upper Deep Creek and in other localized areas). Pesticide use on private forestland in the Clackamas River basin is unknown, but data from a nearby drainage basin, the McKenzie River basin (south of the Clackamas River basin), indicates that pesticide uses on private forestland may be significant. For example, about 97,650 acres of forestland in the McKenzie River basin was projected to be treated with the herbicides 2,4-D, glyphosate, hexazinone, metsulfuron, triclopyr, and imazapyr in 2006 (Morgenstern, 2006).

In the Clackamas River basin study, sampling did not focus specifically on forestland, but forestland was a large component of the basin land cover for a few of the streams sampled, which can provide some insights into pesticide concentrations and loadings from these largely forested basins. Overall, fewer pesticide compounds were detected in storm-runoff samples collected from Eagle and Clear Creeks in May and October 2000 (2 and 5 pesticides each, respectively) compared with streams draining agricultural or urban land. Because of their higher streamflows, however, Clear and Eagle Creeks contributed 19 and 12 percent of the total measured atrazine load in the lower Clackamas River in May 2000 (Carpenter, 2004). Although atrazine can be used on conifer trees on forestland or Christmas tree plantations—plentiful in Clear and Eagle Creek basins—atrazine also is used for agricultural purposes. The nonstorm-runoff samples collected from the mostly forested upper Deep Creek basin during the 2003–04 EUSE study contained six pesticide compounds, with the forestry and Christmas tree herbicide hexazinone being detected in all six samples (appendix C, table C2). Detection of hexazinone is consistent with the high amount of forestland in the upper Deep Creek basin (53 percent). Despite the large amount of forestland, pesticide use on rural residential areas, pasturelands, or along right-of-ways also may contribute to detections in these streams, so specific studies focused on forestland are needed to fully evaluate this potential source.

Urban uses—Pesticides are used in urban areas to control weeds and insect pests on lawns, gardens, and ornamental trees and plants, and in homes to control pests such as ants and fleas. During the past 20 years, about one-half of homes in the United States were treated with pesticides for nonstructural pests (Templeton and others, 1998). About 55 percent of the pesticides detected in the Clackamas River basin have urban uses, and several herbicides are applied along fences, utility lines, roads and other right-of-ways in urban areas (table 10). Many urban-use pesticides were detected in the Clackamas River basin, including atrazine, metolachlor, simazine, prometon, diuron, and 2,4-D. These were the most common herbicides detected in urban streams nationwide (U.S. Geological Survey, 1999; Gilliom and others, 2006).

The two most highly urbanized streams in the Clackamas River basin—Cow and Carli Creeks—have about 90 percent urban land, and drain large amounts of impervious area such as buildings, roads, and parking lots that convey rainfall runoff to the Clackamas River. These streams had between 7 and 12 pesticides detected during the 2 storms, with some occurring at relatively high concentrations (appendix C, table C1). The diazinon concentration in Carli Creek, for example, was 0.25 μg/L, which exceeded the USEPA aquatic-life criterion of 0.1 μg/L (fig. 8D). Streamflow in each of these urban creeks was relatively high for their drainage area during the May and September 2005 storms, resulting in higher water yields compared with other less developed basins (fig. 5A). Pesticide yields (mass per unit area) in these basins also were the highest of all basins sampled during the May and September 2005 storms (fig. 5C).

Wastewater effluents—Although the quality of wastewater in the Clackamas River basin was not examined during this study, treated effluent from the city of Estacada is discharged to the Clackamas River upstream of River Mill Dam, and North Fork Deep and Tickle Creeks receive treated effluent from the community of Boring and the city of Sandy, respectively. Effluent from Sandy is routed to a nearby nursery to irrigate ornamental nursery stock from about May through October. Leakage from failed or failing septic systems, which can be a source of many different kinds of contaminants, including pesticides, also may introduce wastewater into the surrounding soils and aquifers connected to the Clackamas River and some of its tributaries.

In addition, the Source Water Assessment study identified 194 areas with septic systems and 27 large capacity septic systems in the basin that have potential to release wastewater to ground water flowing into the Clackamas River upstream of drinking-water intakes (Oregon Department of Environmental Quality and the Oregon Department of Human Services, 2003). Basinwide, however, there may be thousands of individual septic systems.

Golf courses—The extent of pesticide use on golf courses in the Clackamas River basin is unknown. There are six golf courses located within the drainage basin, and considering that many golf courses in Oregon treat turf for various fungal, insect, and weed pests, golf courses are another potential pesticide source. About 50 percent of the pesticides detected in the Clackamas River basin have reported use on golf courses (Barbash, 1998) (table 10). More specific information on golf course applications in the basin could help quantify this potential source, and may become available through existing or future Pesticide Use Reporting surveys.

Atmospheric deposition—Pesticides and other chemicals also may be transported through the air and later deposited on land and into waterways. For example, orthophosphate insecticides in two Oregon streams, Hood River and Mill Creek (tributaries of the Columbia River), were detected following periods of chemical applications on orchard crops, and may be related to atmospheric drift, mixing operations, or other aspects of their use (Gene Foster, Oregon Department of Environmental Quality, oral commun., 2006). In another study, chlorpyrifos, diazinon, trifluralin, and other pesticides were detected in air samples collected in Sacramento, California (Majewski and Baston, 2002). Pesticides were detected in wet deposition (rain) (Capel and Wotzka, 1998), and in snow samples from Mount Rainier National Park, Washington (Hageman and others, 2006). Three of the four most frequently detected pesticides in the Mount Rainier snow (dacthal, chlorpyrifos, and endosulfan) also were detected in the Clackamas River basin during 2000–2005.

Potential Future Studies

Additional monitoring could track contaminants that may pose a future threat, for pesticides identified during this study, or from other compounds that may be identified through the PURS, which began in 2007. Candidate streams for follow-up studies include Tickle, North Fork Deep, and Noyer Creeks (all Deep Creek tributaries), and Rock Creek, where some of the highest loads and concentrations of pesticides were measured during this study. Future studies might also focus on Cow and Carli Creeks, which had the highest pesticide yields during the May and September 2005 storms.

The seasonal contributions from select streams also could be evaluated with monthly sampling, for example, to better understand the relations between the timing of pesticide applications and detections in streams. Such monitoring could better quantify contaminant contributions from potential sources identified in this study, such as urban developments or certain types of agriculture, including, for example, nursery operations or Christmas tree plantations.

Future studies may utilize autosampling devices that could collect water during periodic storm-runoff events, for example, to provide more detailed information on the temporal occurrence and transport of contaminants including pesticides. Passive sampling equipment such as semipermeable membrane devices (SPMDs) and polar organic chemical integrative samplers (POCIS) could provide time-weighted concentrations for certain hydrophilic compounds present in streams or the mainstem Clackamas River. Some SPMD data were collected for three Deep Creek basin streams during the EUSE study in 2004 (Ian Waite, U.S. Geological Survey, written commun., 2007), and results from future studies could be compared to results of the 2004 study. Alvarez and others (2004) used POCIS samplers to identify the presence of select pesticides, including diuron, in surface water. Diuron was among the most frequently detected pesticides in the Clackamas River basin during the present study—occurring in the tributaries, mainstem Clackamas River, and in samples of source and finished drinking water—making POCIS a viable option for future monitoring of this herbicide.

Detailed time-series data collected over the course of a storm hydrograph could provide insights into the dynamic nature of pesticide transport within these basins, and could better quantify their overall contributions during storms. Such data would provide much needed information about the duration of pesticide occurrence in the Clackamas River and at the downstream drinking-water supply intakes during storms. Time-series data also could determine the concentrations and duration of exposure for aquatic life in the Clackamas River and its tributaries.

Future studies could examine the cumulative effects of nursery and farm pond drawdown on the Clackamas River in autumn, when the combination of released pond water and storm runoff may produce spikes in pesticide concentrations during this susceptible period, when dilution water is in shorter supply. If warranted, future studies could analyze pesticides in fish tissue and conduct physiological studies to determine potential impairment to biological functions.

Reductions in the offsite transport of pesticides to streams may be achieved by developing and implementing best management practices (BMPs) to reduce erosion or reducing chemical application rates. The Oregon Department of Environmental Quality (DEQ) is working on a pesticide Stewardship Partnership in the Clackamas River basin, for example, with the objective of identifying streams with elevated levels of pesticides (orthophosphate insecticides and triazine herbicides—atrazine and simazine) and helping to implement BMPs. DEQ and other agencies, including Oregon Departments of Environmental Quality, Agriculture, Human Services, and Forestry, and the Clackamas County Soil and Water Conservation District (SWCD) have formed a water quality working group and are working collaboratively on this issue. Targeted monitoring before and after implementation of specific projects initiated by the working group might identify BMPs that can identify mechanisms involved in offsite transport from sources such as pond discharges or runoff of irrigation water. In addition, educational programs aimed at reducing pesticide contamination are currently being developed by the Clackamas River Basin Council in cooperation with Clackamas Watershed Management Group, USGS, Portland METRO, and Clackamas County SWCD.

Additional monitoring of source and finished drinking water could verify the results presented here, and examine treatment options for the various types of water treatment plants that utilize the river. Additional monitoring of the source water could provide information on the seasonal patterns in pesticide occurrence in the basin, and identify trends in concentrations over time that may occur. Continued monitoring for pesticides is especially important for the lower Clackamas River and its tributaries because of the encroaching development from Portland, which has expanded its urban growth boundary into parts of the lower basin near Damascus, including parts of Rock and Richardson Creeks (fig. 1). Population growth in this area is expected to be considerable in the coming years, which poses additional threats to water quality.

Pesticide concentrations in finished drinking water reported herein may be higher than concentrations farther along in the distribution system (for example, at the customers’ taps) because finished water samples were preserved with a dechlorinating agent to stop chlorine activity (see Water Sample Processing and Laboratory Analysis section), whereas pesticides in the distribution system continue to be exposed to residual chlorine. Continued oxidation of pesticide compounds by chlorine would be expected to occur in the distribution system, resulting in lower concentrations at customers’ taps. Even with relatively short contact time (about 90 min), chlorination did appear to oxidize many of the organic compounds in this study, in one case transforming the insecticide diazinon in source water to its degradate diazinon-oxon in finished drinking water. A small number of split samples with and without the dechlorinating agent showed significant differences for some compounds (appendix A, table A3; appendix B, table B2). Although this study was not designed to fully characterize water treatment, comparisons between pesticide concentrations in source and finished water can provide some indications about the removal of pesticides by the process of direct filtration.

Future studies could evaluate treatment options for the different types of compounds, if concentrations should increase to levels approaching human-health benchmarks. Such studies would benefit from more precise estimates of travel time through the water-treatment plant (for example, time from source to finished water) to ensure comparability between the two samples. Tracer studies, for example, could ensure that accurate comparisons are made. Accurate travel times are especially important during storms because pesticide concentrations may change rapidly as runoff from different areas of the basin reaches source-water intakes.

In the current study, PAC (powdered activated carbon) appeared to be somewhat effective at decreasing concentrations of some pesticides. In most cases, however, concentrations in the source water were so low (often close to the detection level) that measured decreases in finished water may not be statistically significant. Although PAC has been shown to be effective at decreasing concentrations of pesticides and other organic contaminants elsewhere (Westerhoff and others, 2005), additional studies could determine the potential effectiveness of PAC in these waters. PAC appeared to be less effective at decreasing or removing pesticides during storms, possibly because of interference by high concentrations of suspended sediment in source water. The September 2005 sample of PAC-treated finished water, for example, contained several pesticides, including diazinon-oxon, simazine, ethoprop, metolachlor, 2,4-D and propiconazole, among others (table 5). Higher doses of PAC may be required to remove the pesticides from highly turbid water, in this case about 100 NTRU (appendix C, table C4). There was no apparent association between the physical properties of the pesticides, such as the organic carbon partitioning coefficient (Koc) or water solubility (appendix E), that determined the likelihood of a pesticide being removed through treatment with PAC or chlorine, although future studies could evaluate removal efficiencies at varying levels of PAC or evaluate other treatment options using controlled laboratory experiments.

Back to Table of Contents

AccessibilityFOIAPrivacyPolicies and Notices

Take Pride in America logoUSA.gov logoU.S. Department of the Interior | U.S. Geological Survey
URL: http://pubs.usgs.gov/sir/2008/5027
Page Contact Information: Publications Team
Page Last Modified: Thursday, 10-Jan-2013 18:43:07 EST