Continuous low-level aquatic monitoring (CLAM) samplers for pesticide contaminant screening in urban runoff: Analytical approach and a field test case.

Monitoring of surface waters for organic contaminants is costly. Grab water sampling often results in non-detects for organic contaminants due to missing a pulse event or analytical instrumentation limitations with a small sample size. Continuous Low-Level Aquatic Monitoring (CLAM) samplers (C.I.Agent® Solutions) continually extract and concentrate organic contaminants in surface water onto a solid phase extraction disk. Utilizing CLAM samplers, we developed a broad spectrum analytical screen for monitoring organic contaminants in urban runoff. An intermediate polarity solid phase, hydrophobic/lipophilic balance (HLB), was chosen as the sorbent for the CLAM to target a broad range of compounds. Eighteen urban-use pesticides and pesticide degradates were targeted for analysis by LC/MS/MS, with recoveries between 59 and 135% in laboratory studies. In field studies, CLAM samplers were deployed at discrete time points from February 2015 to March 2016. Half of the targeted chemicals were detected with reporting limits up to 90 times lower than routine 1-L grab samples with good precision between field replicates. In a final deployment, CLAM samplers were compared to 1-L water samples. In this side-by-side comparison, imidacloprid, fipronil, and three fipronil degradates were detected by the CLAM sampler but only imidacloprid and fipronil sulfone were detected in the water samples. However, concentrations of fipronil sulfone and imidacloprid were significantly lower with the CLAM and a transient spike of diuron was not detected. Although the CLAM sampler has limitations, it can be a powerful tool for development of more focused and informed monitoring efforts based on pre-identified targets in the field.

Relative toxicity of bifenthrin to Hyalella azteca in 10 day versus 28 day exposures.

Many watersheds in the Central Valley region of California are listed as impaired due to pyrethroid-associated sediment toxicity. The Central Valley Regional Water Quality Control Board is developing numeric sediment quality criteria for pyrethroids, beginning with bifenthrin. Criteria are being developed using existing data, along with data from 10 d and 28 d toxicity tests with Hyalella azteca conducted as part of the current study. A single range-finder and 2 definitive tests were conducted for each test duration. Median lethal concentrations (LC50s), as well as LC20s and inhibition concentrations (IC20s) were calculated based on measured whole sediment bifenthrin concentrations and interstitial water concentrations. Sediment LC50s were also corrected for organic C content. Average LC50s were not significantly different in 10 d versus 28 d tests with H. azteca: 9.1 and 9.6 ng/g bifenthrin for 10 d and 28 d tests, respectively. Average LC20 values were also similar with concentrations at 7.1 and 7.0 for 10 d and 28 d tests, respectively. Bifenthrin inhibition concentrations (IC20s) based on amphipod growth were variable, particularly in the 28 d tests, where a clear dose-response relationship was observed in only 1 of the definitive experiments. Average amphipod growth IC20s were 3.9 and 9.0 ng/g for 10 d and 28 d tests, respectively. Amphipod growth calculated as biomass resulted in IC20s of 4.1 and 6.3 ng/g for the 10 d and 28 d tests, respectively. Lack of a clear growth effect in the longer term test may be related to the lack of food adjustment to account for amphipod mortality in whole sediment exposures. The average C-corrected LC50s were 1.03 and 1.09 µg/g OC for the 10 d and 28 d tests, respectively. Interstitial water LC50s were determined as the measured dissolved concentration of bifenthrin relative to interstitial water dissolved organic carbon. The average LC50s for dissolved interstitial water bifenthrin were 4.23 and 4.28 ng/L for the 10 d and 28 d tests, respectively. In addition, a set of 10 d and 28 d tests were conducted at 15 °C to assess the relative toxicity of bifenthrin at a lower temperature than the standard 23 °C test temperature. These results showed that bifenthrin was more toxic at the lower temperature, with LC50s of 5.1 and 3.4 ng/g bifenthrin in 10 d and 28 d tests, respectively. Amphipod growth at 15 °C after a 28 d exposure resulted in the lowest effect concentration of all experiments conducted (IC20 = 0.61 ng/g). This article discusses how bifenthrin dose-response data from 10 d and 28 d exposures inform development of sediment quality criteria for this pesticide for California Central Valley watersheds.

Methods for deriving pesticide aquatic life criteria for sediments.

In this review, we evaluated three main current approaches for deriving sediment quality guidelines: empirical, mechanistic (equilibrium partitioning), and spiked sediment toxicity testing approaches. Empirical approaches determine ranges of sediment concentrations that are likely or unlikely to cause toxicity, based on large datasets of matching sediment chemistry, field, and laboratory toxicity data. The empirical approaches are not suitable for determining SQC for specific pesticides because (I) direct cause-effect relationships between single sediment contaminants and toxicity cannot be discerned; (2) chemistry measurements have not accounted for bioavailability, which leads to numeric values with high uncertainty and low reliability; and (3) for many pesticides, little or no matching chemistry and toxicity data are available. In the EqP approach, SQC are derived by entering existing aquatic toxicity data into the equilibrium-partitioning model. This approach is practical for pesticides with water quality criteria in place, but the assumption of equilibrium in aquatic ecosystems is questionable, and the EqP approach neglects available sediment toxicity data. The SSTT approaches utilize sediment toxicity data, creating a scientifically defensible foundation for SQC, but experimental uncertainties regarding spiking technique and equilibration times are yet to be eliminated. The species sensitivity distribution approach generates criteria with confidence intervals, providing a measure of uncertainty, but requires relatively large datasets, whereas the assessment factor method lacks quantification of uncertainty but only requires few data to calculate conservative criteria. Several existing methodologies incorporate a combination of approaches that is dependent on data availability and the physicochemical properties of the compound of interest.A summary of the differences and similarities between key elements of the seven methodologies emphasized in this review is displayed in Table 6. One important element regarding sediment contamination is the incorporation of bioavailability and multiple exposure routes, which must be addressed to achieve a technically defensible methodology. It is crucial that bioavailability be incorporated in both criteria derivation and compliance determination (sampling) to ensure that data are comparable. Recent research on bioavailability of sediment contaminants has indicated that the freely dissolved pore water fraction corresponds well with uptake and toxicity. For species having significant exposure via ingestion of contaminated food and/or sediments and/or direct sediment contact, exposure may be underpredicted if these additional exposure routes are overlooked. Future SQC methodologies will be greatly improved by accounting for factors relevant for bioavailability and exposure pathways. To develop a completely new methodology, existing methodologies offer valuable building blocks that are well suited for adaptation. A new method will be more reliable and robust if it utilizes more refined risk assessments than currently are available in existing methodologies. To date, the most comprehensive methodologies for deriving single numeric SQC are those of the Netherlands and the EU,which include both SSTT and EqP approaches.

Rotenone formulation fate in Lake Davis following the 2007 treatment.

In September 2007, Lake Davis (near Portola, California) was treated by the California Department of Fish and Game with CFT Legumine, a rotenone formulation, to eradicate the invasive northern pike (Esox lucius). The objective of this report is to describe the fate of the five major formulation constituents-rotenone, rotenolone, methyl pyrrolidone (MP), diethylene glycol monethyl ether (DEGEE), and Fennedefo 99-in water, sediment, and brown bullhead catfish (Ameiurus nebulosus; a rotenone-resistant species) by determination of their half-lives (t(1/2)) and pseudo first-order dissipation rate constants (k). The respective t(1/2) values in water for rotenone, rotenolone, MP, DEGEE, and Fennedefo 99 were 5.6, 11.1, 4.6, 7.7, and 13.5 d; in sediments they were 31.1, 31.8, 10.0, not able to calculate, and 48.5 d; and in tissues were 6.1, 12.7, 3.7, 3.2, and 10.4 d, respectively. Components possessing low water solubility values (rotenone and rotenolone) persisted longer in sediments (not detectable after 157 d) and tissues (<212 d) compared with water, whereas the water-miscible components (MP and DEGEE) dissipated more quickly from all matrices, except for Fennedefo 99, which was the most persistent in water (83 d). None of the constituents was found to bioaccumulate in tissues as a result of treatment. In essence, the physicochemical properties of the chemical constituents effectively dictated their fate in the lake following treatment.

Soil and glass surface photodegradation of etofenprox under simulated california rice growing conditions.

Photolysis is an important degradation process to consider when evaluating a pesticide's persistence in a rice field environment. To simulate both nonflooded and flooded California rice field conditions, the photolytic degradation of etofenprox, an ether pyrethroid, was characterized on an air-dried rice soil and a flooded rice soil surface by determination of its half-life (t(1/2)), dissipation rate constant (k) and identification and quantitation of degradation products using LC/MS/MS. Photodegradation was also characterized on a glass surface alone to rule out confounding soil factors. Measured photolytic dissipation rates were used as input parameters into a multimedia environmental fate model to predict etofenprox persistence in a rice field environment. Photolytic degradation proceeded at a faster rate (0.23/day, t(1/2) = 3.0 days) on the flooded soil surface compared to the air-dried surface (0.039/day, t(1/2) = 18 days). Etofenprox degradation occurred relatively quickly on the glass surface (3.1/day, t(1/2) = 0.23 days or 5.5 h) compared to both flooded and air-dried soil layers. Oxidation of the ether moiety to the ester was the major product on all surfaces (max % yield range = 0.2 ± 0.1% to 9.3 ± 2.3%). The hydroxylation product at the 4' position of the phenoxy phenyl ring was detected on all surfaces (max % yield range = 0.2 ± 0.1% to 4.1 ± 1.0%). The air-dried soil surface did not contain detectable residues of the ester cleavage product, whereas it was quantitated on the flooded soil (max % yield = 0.6 ± 0.3%) and glass surface (max % yield = 3.6 ± 0.6%). Dissipation of the insecticide in dark controls was significantly different (p < 0.05) compared to the light-exposed surfaces indicating that degradation was by photolysis. Laboratory studies and fate model predictions suggest photolysis will be an important process in the overall degradation of etofenprox in a rice field environment.

Aerobic versus Anaerobic Microbial Degradation of Etofenprox in a California rice field soil.

The microbial degradation of etofenprox, an ether pyrethroid, was characterized under anaerobic (flooded) and aerobic (nonflooded) California rice field soil conditions by determination of its half-life (t1/2) and dissipation rate constant (k) and identification and quantification of degradation products at both 22 and 40 °C using LC-MS/MS. The overall anaerobic t1/2 at 22 °C ranged from 49.1 to 100 days (k=-0.0141 to -0.0069 days(-1)) compared to 27.0 days (k=-0.0257 days(-1)) at 40 °C, whereas under aerobic conditions the overall t1/2 was 27.5 days (k=-0.0252 days(-1)) at 22 °C compared to 10.1-26.5 days (k=-0.0686 to -0.0262 days(-1)) at 40 °C. The biphasic dissipation profiles were also fit to a first-order model to determine the t1/2 and k for both the fast and slow kinetic regions of the dissipation curves. Hydroxylation at the 4'-position of the phenoxy phenyl ring was the major metabolic process under anaerobic conditions for both 22 °C (maximum% yield of applied etofenprox mass=1.3±0.7%) and 40 °C (max % yield=1.2±0.8%). Oxidation of the ether moiety to the ester was the major metabolite under aerobic conditions at 22 °C (max% yield=0.5±0.1%), but at 40 °C increased amounts of the hydroxylated form were produced (max% yield=0.7±0.2%, compared to 0.3±0.1% for the ester). The hydrolytic product of the ester, 3-phenoxybenzoic acid (3-PBA), was not detected in any samples. Sterilized control soils showed little etofenprox degradation over the 56-day incubation period. Thus, the microbial population in a flooded soil was able to transform and contribute to the overall dissipation of etofenprox. The simulated summer temperature extreme (40 °C) increased the overall degradation.

Partitioning of etofenprox under simulated California rice-growing conditions.

BACKGROUND: The pyrethroid insecticide etofenprox is of current interest to rice farmers in the Sacramento Valley owing to its effectiveness against the rice water weevil, Lissorhoptrus oryzophilus Kuschel. This study aimed to describe the partitioning of etofenprox under simulated rice field conditions by determining its Henry's law constant (H) (an estimate of volatilization) and organic carbon-normalized soil-water distribution coefficient (K(oc)) at representative field temperatures. A comparison of etofenprox and lambda-cyhalothrin is presented using a level-1 fugacity model. RESULTS: Experimental determination of H revealed that etofenprox partitioned onto the apparatus walls and did not significantly volatilize; the maximum value of H was estimated to be 6.81 x 10(-1) Pa m(3) mol(-1) at 25 degrees C, based on its air and water method detection limits. Calculated values for H ranged from 5.6 x 10(-3) Pa m(3) mol(-1) at 5 degrees C to 2.9 x 10(-1) Pa m(3) mol(-1) at 40 degrees C, based on estimated solubility and vapor pressure values at various temperatures. Log K(oc) values (at 25 degrees C) were experimentally determined to be 6.0 and 6.4 for Princeton and Richvale rice field soils, respectively, and were very similar to the values for other pyrethroids. Finally, temperature appears to have little influence on etofenprox sorption, as the log K(oc) for the Princeton soil at 35 degrees C was 6.1. CONCLUSION: High sorption coefficients and relatively insignificant desorption and volatilization of etofenprox suggest that its insolubility drives it to partition from water by sorbing to soils with high affinity. Offsite movement is unlikely unless transported in a bound state on suspended sediments.