Evaluation of Passive Samplers for Long-Term Monitoring of Organic Compounds

A period of moderate rainfall occurred during the passive sampler deployment period (fig. 2). Although this was a relatively small storm, it was preceded by an extended period of dry weather and was the first measurable runoff event of the autumn rainy season. The event produced a discernible increase in flow in the McKenzie River while the samplers were deployed (fig. 2).

Polar Organic Chemical Integrative Samplers

Chemical analyses.—The herbicide metolachlor, used for grass and broadleaf weed control in crops, was the only pesticide detected in extracts from the deployed POCIS (table 2). Metolachlor was detected in one of the samples from the treatment-plant inlet, but the detection was less than the MQL and metolachlor was not detected in the replicate sample (table 2; column B).

Three wastewater-indicator chemicals were detected in deployed POCIS (table 3). Cholesterol, a sterol that is synthesized by animals and to a lesser degree by plants and fungi and is used as an emulsifying agent in cosmetics and pharmaceuticals, was quantified in one of the treatment-plant inlet replicates and was detected but not quantified in one of the Camp Creek replicates. Diethyl phthalate, a widely used plasticizer and common background contaminant, was present at quantifiable levels in both treatment-plant inlet replicates and one of the Cedar Creek replicates. Another common plasticizer, diethylhexylphthalate (DEHP), was identified at a concentration less than the MQL in one of the treatment-plant inlet replicates.

Yeast Estrogen Screen.—Estrogenicity was not measurable in POCIS extracts from the treatment-plant inlet. Estrogenicity was measurable in samples from Cedar Creek and Camp Creek but at levels near the quantitation limit for this method (table 4). No estrogenic response was observed from any of the blanks, indicating that the sampler matrix and sample processing steps did not contribute to the total measured estrogenicity. The maximum estimated EEQ value of 0.06 ng/L was 130 times less estrogenic than the EC₁₀ concentration of 7.8 ng/L for the natural hormone 17β-estradiol.

Most studies to date that have used the YES on POCIS extracts have been focused on effluents from wastewater-treatment plants or industrial processes, and data from relatively pristine streams are scarce. During a study to assess runoff from livestock operations, however, Matthiessen and others (2006) used YES on extracts from POCIS deployed at sites upstream and downstream of the suspected sources of contamination. With the exception of one high outlier, YES results from their upstream (control) sites ranged from less than detection to 26.5 ng/L with a median EEQ of 0.4 ng/L. This median is approximately an order of magnitude greater than EEQ levels measured in Cedar and Camp Creeks.

Quality control.—The apparent lack of agreement between some replicate POCIS pairs is not an indication of poor data quality, but rather a manifestation of normal variability near method detection limits. With the exception of one unquantified detection of cholesterol in one of the treatment-plant inlet samplers, no pesticides, wastewater-indicator chemicals, or estrogenicity were detected in extracts from field-blank POCIS (appendix A, tables A1–A3).

Semipermeable Membrane Devices

Chemical analyses.—Several of the commonly identified PAHs were detected in the SPMDs deployed at each site (table 5). Fluorene, phenanthrene, fluoranthene, pyrene, chrysene, 1-methylfluorene, and 2-methylphenanthrene were detected at concentrations at or less than the MQL. These PAHs are common byproducts of combustion from sources such as wood smoke and automobiles.

The persistent chlorinated pesticides dacthal, chlorpyrifos, trans-chlordane, and DDT, as well as pentachloroanisole (PCA), which likely results from the degradation of the pesticides pentachlorophenol and(or) pentachloronitrobenzene, were detected in SPMD extracts from each of the study sites (table 6). In addition, endosulfan was detected in both replicates from the Cedar Creek and Camp Creek sites. Hexachlorobenzene, trans-nonachlor, DDE, DDD, and dieldrin were detected in both replicates from the Cedar Creek site, and heptachlor epoxide was detected in both replicates from the treatment-plant inlet. With the exception of PCA, all detections were near or less than the MQL. PCA was detected at quantifiable concentrations in all samples. The concentrations of PCA were similar to or slightly higher than concentrations detected during multiple sampling periods in the Columbia Slough system in Portland, Oregon (McCarthy, 2006, 2008).

Quality control.—As with the POCIS data, the apparent lack of agreement between some SPMD replicates is not an indication of poor data quality and is common when compounds are detected at concentrations near method detection limits. Other than trace concentrations of 2-methylnaphthalene, fabrication and field-blank SPMDs were free from target PAHs and organochlorine compounds (appendix A, tables A4–A5).

Estimation of water concentrations.—Water concentrations estimated from SPMD data that were quantified (table 7) indicate that the compounds detected were present in water at picogram-per-liter (parts per quadrillion) concentrations. For the compounds listed in table 7 for which health based screening levels (HBSLs) are available (Toccalino and others, 2008), the HBSLs were 4–6 orders of magnitude greater than the water concentrations estimated during this study.