Metabolism of Pyrithiobac Sodium in Soils and Sediment, Addressing

Jul 5, 2016 - Ashok K. Sharma†, Lian Wen‡, Larry R. Hall‡, John G. Allan‡, and Brett J. Clark‡. † Stine Haskell Research Center, E.I. DuPo...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/JAFC

Metabolism of Pyrithiobac Sodium in Soils and Sediment, Addressing Bound Residues via Kinetics Modeling Ashok K. Sharma,*,† Lian Wen,‡ Larry R. Hall,‡ John G. Allan,‡ and Brett J. Clark‡ †

Stine Haskell Research Center, E.I. DuPont de Nemours and Co, 1090 Elkton Road,Newark, Delaware 19714, United States ABC Laboratories, Inc., 4780 Discovery Drive, Columbia, Missouri 65201, United States



ABSTRACT: Degradation of pyrithiobac sodium (PE350) was examined in a number of soils and sediments using 14C-PE350. It degrades primarily via microbial degradation which leads to the separation of the two rings of the molecule. Identification of several metabolites, many of which were minor products, helped to understand the formation of nonextractable residues (NER) and 14CO2. In all studies, unextractable residues accounted for a large portion (20−60%) of the residues. Traditional kinetics modeling treats NER and CO2 as a single compartment, stated as sink, and formation mechanism of such components individually is ignored. Since studies conducted with radiolabeled test substance provides an accurate measurement of NER and CO2, we have demonstrated that kinetics modeling with these compartments separately can be used to clarify degradation pathways, including the origin of NER and CO2. This work demonstrated that overall metabolism in soils and sediments proceeded via similar pathways, and kinetics modeling was useful in clarifying the degradation route and formation of NER in all studies. KEYWORDS: pyrithiobac sodium, soil metabolism, sediment degradation, aquatic degradation, nonextractable residues, bound residues, NER, kinetics modeling



INTRODUCTION Pyrithiobac sodium (E350, CAS Registry Number: 123343-168), sodium 2-chloro-6-[(4,6-dimethoxypryrimidin-2-yl)thio]benzoate is a herbicide commercialized for uses in cotton cultivation. It has a water solubility of 117−690 mg/L, which varies with pH of the solution. The product remains popular due to emerging resistance to many alternative herbicides. Even though the product was introduced many years ago, information on its degradation behavior in soil and sediments is limited.1,2 Reports on resistance mechanisms3,4 and product5−7 uses have appeared periodically. This study presents the degradation of PE350 in a number of soils and sediments, identification of its degradation products and kinetics modeling which help clarify the formation of bound residues. Extensive degradation of this compound in soils and sediments led to predominantly nonextractable residues (NER) and 14CO2. Characterization of NER is always challenging8 because structural information is difficult to ascertain. We have utilized kinetics modeling in combination with the nature of minor degradation products to shed light on the formation of NER and overall degradation in the environment.

Scheme 1. Radiolabel Locations in Test Substance

dissolved in 12.5 mL of acetonitrile and 12.5 mL of HPLC-grade water. The concentration was determined to be 1.08 mg/mL and used for PH-label treated samples. Incubation in Soils. Four soils were collected from various locations (Table 1), between March 11 and April 08, 2015, and the soils were stored at ca. 4 °C until used. Soils were passed through a 2 mm sieve. The moisture content of each soil was measured by analysis of aliquots of the sieved soil with a Mettler Toledo HR 73 halogen moisture analyzer. Various physico chemical characteristics and the microbial biomass of the soils were determined at Agvise Laboratories, Inc. (Northwood, ND, USA). Test samples were prepared by weighing portions of soil (50 g dry weight equivalent) adjusted to 100% of 0.1 bar (pF2) moisture into test flasks. Test samples were incorporated into a flow-through system, and incubated under the test conditions for 14 days prior to test item application. Test systems were maintained in darkness at a temperature of 20 ± 2 °C. All test vessels treated with the two radiolabels in separate vessels, and



MATERIALS AND METHODS DuPont prepared two versions of the test substance, PH-label labeled uniformly in the phenyl ring, and PYM-label was located at 2-position of the pyrimidine ring (Scheme 1). A 19.49 mg aliquot of [pyrimidine-2-14C]-PE350 (PYM-label) and 16.25 mg aliquot of nonlabeled PE350 were quantitatively transferred into an amber glass bottle with 15 mL of acetonitrile and 15 mL of HPLC-grade water. The concentration was determined to be 1.01 mg/mL (sodium salt equivalent), and used for dosing PYM-label treatments. Likewise, a 28.00 mg aliquot of [phenyl(U)-14C]-PE350 (PH-label) was quantitatively © 2016 American Chemical Society

Received: Revised: Accepted: Published: 5793

May 6, 2016 June 16, 2016 July 5, 2016 July 5, 2016 DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry Table 1. Characteristics of Soils Used soil name

Tift

Cheneyville

Raymondville

GreenvilleA

GreenvilleB

location

Tift County, GA

Rapides Parish, LA

Willacy County, TX

Washington County, MS

Greenville, MS

texture class

sandy loam

silty clay

sandy clay loam

silt loam

silt loam

sand (%) silt clay pH (in 0.01 M CaCl2) organic matter (%) CEC* (meq/100g) MWHC (% w/w at 1/3 bar)

78 7 15 5.7 1.6 5.3 8.9

6 47 47 7.5 1.3 20.6 27.3

50 19 31 7.5 2.0 16.3 22.9

41 52 7 5.8 0.8 6.8 14.3

24.4 64 11.6 6.5 1.1 10.5 19.4

leaving each water/sediment treatment group was passed sequantially through an air sampling tube (Supelco ORBO-32) to trap volatile compounds, followed by two traps containing 1N KOH to collect 14CO2. At each sampling interval, the dissolved oxygen content, redox potential, and pH of the water and the sediment were measured in the study samples. Samples were treated with the test material at 0.9 μg/mL based on volume of water for both radiolabels. The anaerobic water sediment system was set up just like the aerobic water sediment system, except nitrogen was passed through all test vessels during acclimation and the incubation periods in order to keep the system anaerobic. Sample Analysis. Soil samples were analyzed at zero time (immediately following application of test substance) and at 3, 7, 15, 30, 50, 70, 100, and 120 days of aerobic incubation. Sedimentwater samples (aerobic as well as anaerobic) were analyzed at days 0, 7, 15, 32, 50, 82, and 103. All soil and sediment samples were extracted sequentially, once with 9:1 (v/v) acetone-1% aqueous ammonium carbonate, and twice with 4:1 (v/v) acetone-1% aqueous ammonium carbonate. The second and third extractions included sonication and heating to 50 °C with the acetone/ammonium carbonate solutions prior to centrifugation. A final fourth extraction was carried out when the NER exceeded 10% of the applied radioactivity. The fourth extraction utilized Soxhlet extraction with 1% formic acid in acetonitrile for 1 h. Total radioactive content in the various solvent extracts was determined by LSC. Nonextractable 14C-residues were quantified by combustion analysis. Appropriate portions of the extracts were pooled and concentrated to a small volume. Concentrated extracts were filtered through a 0.2 μm filter before analysis by reverse phase HPLC. The surface water for each sediment sample was separated from the solids via centrifugation. The water supernatant was transferred to a graduated cylinder, and the solids removed from the water were combined with the remaining sediment for further extraction with organic solvents. The water volume was recorded, and triplicate 1 mL aliquots were assayed by liquid scintillation counting (LSC) on the day of sampling. A 5 mL aliquot of the centrifuged surface water was filtered (0.2 or 0.45 μm) prior to HPLC analysis. Water and the soil extracts were treated with 10% aqueous formic acid to achieve 0.1% formic acid in the samples for HPLC analysis. In order to confirm the identity of 14CO2 trapped in potassium hydroxide, the trap solutions were treated with saturated barium chloride. Quantitative precipitation of radioactivity as barium carbonate demonstrated that the radioactivity collected in these traps was 14CO2. HPLC analysis of samples was carried out on an Agilent system equipped with a UV detector set to monitor 254 nm wavelength (Agilent G1314B which employed Empower 3

connected to communal traps for collection of organic volatiles. Air was drawn through the test vessels and it passed through a cylinder containing water to humidify the air before entering the test system flasks. Air leaving the treated test system flasks was passed through a Supelco ORBO 32 adsorbent tube to collect any organic volatiles, an empty trap for overflow, and then through two traps containing 1N KOH (to collect organic volatiles). Vacuum pumps were used to maintain a steady flow of air through the systems. Each soil test vessel was treated with either PH or PYM-label test substance at a concentration of 0.9 μg/g (dry weight basis). Incubation of Aerobic Sediment Systems. Two sediments and associated water were collected from locations listed in Table 2. The sediment and water were stored at approximately Table 2. Characteristics of Sediments Used sediment name

Goose River

Chula

geographic location

Grand Forks, ND, USA

Chula, GA, USA

textural class (USDA)

clay LOAM

sand

sand % silt % clay % bulk density (g/cm3) CEC (meq/100 g) % moisture at 1/3 bar (%) % org.carbon (Walkley Black) pH [in 0.01m CaCl2] total nitrogen (%) total phosphorus (ppm) calcium (ppm) magnesium (ppm)

36 33 31 0.86 21.8 44.0 3.8 7.7 0.33 710 2550 710

98 0 2 1.37 2.6 2.8 0.52 5.8 0.04 97 310 32

4 °C in the dark until used. The sediments were wet-sieved through a 2 mm mesh prior to use in the study. Sediment and water characterization was carried out by Agvise Laboratories (Northwood, ND). The microbial biomass of the sediments was determined at the initiation and at study termination by a fumigation/extraction method. Portions of sieved sediment equivalent to 55.0 and 122.6 g dry weight, for the clay loam (Goose River) and sand (Chula) test systems, respectively, were dispensed into individual incubation vessels to obtain a depth of ca. 2.5 cm sediment. Corresponding water (265 mL) was added to achieve a depth of ca. 8 cm, which provided sediment to surface water ratios of ca. 1:3 (v/v). Each incubation group was incorporated into a separate flow-through system. The test systems were maintained in darkness at a temperature of 20 ± 2 °C. A gentle flow of moist air was passed through the test apparatus and introduced into the test system via a dip tube extending into the surface water of the vessels. The humidified air 5794

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry Software Build 3471), a radioactivity detector (IN/US Systems β-Ram 5C, with B-RAM software: Laura, ver. 4.1.7.70 from Lab Logic), and a Zorbax SB-C8 column (250 × 4.6 mm, 4 μm) maintained at 40 °C. Two mobile phases (A:10 mM ammonium formate in pH 2−3 water; and mobile phase B: acetonitrile) were used as a gradient which started at 2% B (0−2 min) increased linearly to 55% B at 42 min, followed by increase to 95% B at 50 min, and finally to 100% B at 52 min. Mass Spectral Confirmation of Metabolites. A Thermo Fisher Scientific, Q-Exactive Plus mass spectrometer equipped with an Agilent HPLC, using the same gradients and column were used for LC/MS analysis of various samples. The mass spectrometer was used in electrospray mode set to scan 100 to 800 m/z with a resolution of 140 000 in both positive and negative ion mode. Data was acquired using Thermo Xcalibur software (3.0.63). A comparison of reference standards of PE350, JW212, JW213, JW214, JW217, JY676, KG027, with metabolite retention times, accurate mass, and fragmentation patterns were used for identification. Selected soil extracts, surface water, and sediment extracts were used for structural elucidation. Samples of metabolites were partially purified via fraction collection or SPE cleanup before carrying out LC/MS work. As an example, PE350 showed a protonated molecular ion at m/z 327.02064 (accurate to ±5 ppm) in the positive ion mode and a deprotonated ion at m/z 325.0059 in negative ion mode. Peaks assigned to PE350 matched the HPLC retention time for the reference standard, as well as the protonated and deprotonated molecular ions. In addition, fragments corresponding to the loss of 44 mass units (43.9898; loss of CO2) and a fragment due to the pyrimidine ring portion (Scheme 2) helped clarify the presence of dimethoxy-pyrimidine.

Table 3. Mass Spectral Data for Various Components component

HPLC rettime

m/za

other ions/fragments

PE350 JW212 JW213 JW214 JW215 JW217 KG027

33.9 18.2 19.2 22.2 19.6 6.7 13.1

325.0059 (−) 310.9902 (−) 341.0004 (−) 372.9167 (−)b 200.9773 (−) 141.0294 (+) 296.9744 (−)

281.0158d, 155.0452c 267.0003d, 141.0294c 186.9618, 155.0451c, 137.0232

A1 A3 A2

11.2 9.6 9.8

216.9728 (−) 266.9198 (−) 234.9472 (−)

202.9743e, 156.9872d 143.0450 252.9843d, 165.0183, 127.0136c 218.9697e, 172.9825d 186.9620 [loss of SO3] 236.9442e

a c

(+) for positive ion; (−) for negative ion. bDisulfide dimer of JW214. Pyrimidine fragment. dLoss of CO2. eCl isotope.

Degradation led to the formation of two metabolites containing both rings, distinctly identified as JW212 and JW213, observed in both labels. While JW212 was observed in all soils, JW213 was detected in a few soils. These two metabolites together comprised 10% of total radioactivity at multiple intervals (Figure 6) while KG027 was observed as a minor metabolite (5% or less) in both systems. Further degradation of these two metabolites led to label specific smaller molecules. Parent compound as well as the minor metabolites degraded more rapidly in the Goose River system as opposed to the Chula sediment system, as noted by the larger proportion of radioactivity that was either incorporated into organic matter as NER or was mineralized to 14CO2. Minor metabolites in the PYM-label samples appeared as polar molecules, and their structural identity could only be inferred by comparison of retention times with the pyrimidine analogs shown in Scheme 3. As noted for the soil studies, such structures would likely undergo ring opening of the pyrimidine ring,12,13 and would be anticipated to generate 14CO2 in aerobic sediment as well. PH-labeled samples also showed additional polar degradates, which were very likely degraded benzoic acid type structures. The very low concentration of these metabolites in soil made identification difficult, however, a more significant occurrence in sediments enabled identification of two of these degradates as JW214 and JW215. JW215 was a significant metabolite and accounted for a maximum 19% of the applied radioactivity in Goose River sediment samples. Further oxidation of these molecules, especially at the sulfur atom, is wellknown,14,15 which possibly led to the formation of more polar molecules, degradation to 14CO2 and incorporation into NER. Residue composition in the total system, obtained after summation of the water and sediment phase component amounts, is illustrated in Figure 7. A notable difference between

dioxide. Transformation of small molecules into extensive amounts of 14CO2 and some assimilation of this radioactivity into the natural carbon pool is a logical rationale of the results. This is why test systems treated with the pyrimidine label (PYM label) always resulted in a higher proportion of 14CO2 as compared to the phenyl label (PH label) in the same test system. The phenyl label on the other hand, afforded a higher proportion of 14C incorporation into organic matter, and small benzene structures8−10 carrying polar functional groups (shown in Scheme 3, y) have been known to result in substantial bound residues. In order to support why PE350 itself does not lead to the bound residue formation, a kinetics modeling approach was undertaken to explain the formation of radioactive residues, NER and CO2. A kinetics model without any PE350 directly converting to NER was tested by fitting data to a model displayed in Figure 3. For comparison, a model which did allow a portion of PE350 to proceed directly to NER, in addition to the other transformation products, was also tested (Figure 4). The kinetics model would be acceptable if (i) the same model provided good data fit for both labels, (ii) the degradation rate for PE350 were identical when the two labels were modeled separately, within experimental variability (iii) different amounts of NER and CO2 formed for the two labels are consistent with the data observed. Degradation rates for PE350 in all systems, using both labels as replicates, as well as label specific modeling to explain NER and CO2, are listed in Table 4. Simple first-order (SFO) kinetics described the degradation of PE350 very well for all soils and sediments, therefore SFO kinetics was used for all kinetics models. Table 4. Degradation Rates of PE350 in Soils and Sediments DT50 (days) systema

P-onlyb

PHc

PYMc

Tift [Aer-Soil] Cheneyville [Aer-Soil]] Raymondville[Aer-Soil] GreenvilleA [Aer-Soil] GreenvilleB [Aer-Soil] Goose River [Aer-Sediment] Chula [Aer-sediment] Goose River [Anaer-Sediment] Chula [Anaer-Sediment]

44.9 80.6 30.5 81.8 59.3 25.2 118 28.1 51.4

49.9 84.3 34.3 86.9 61.2 24.6 115 30.5 53.7

46.5 80.1 36.8 77.2 61.9 26.9 117 31.8 55.4

a c

Aer = aerobic, Anaer = anaerobic. bCalculated parent decline only. Based on pathway fit for each label.

Raymondville soil, which showed the most extensive degradation was used for the initial kinetics evaluations. Figure 3 displays the data fit for the three largest components of the residue, viz., PE350, CO2, and NER from the two labels in this soil. Since the two labels generate dramatically different levels of NER and CO2, the two labels must be modeled separately. Observed data fit well with the model that PE350 itself is not incorporated into NER. Moreover, the fact that products degrade to CO2 at a faster rate in PYM-label, while product assimilation into NER is faster for PH-label, is consistent with the modeled transformations and observed data. An alternative model, where the parent compound was presumed to become a part of NER, also showed a decent fit for the data [Figure 5]. However, a closer scrutiny of the rate constants needed to achieve the data fit in the case of PYM-label requires NER to convert back into metabolites, otherwise a 5798

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry

Figure 5. Data fit in GreenvilleB soil.

Figure 6. PE350 and JW212 in two aerobic sediments (*PH-label, no* PYM-label).

Figure 7. Composition of residue in the aerobic sediment system, Goose River.

likely explanation for such an observation is that the pyrimidine ring was more vulnerable to ring-opened small molecules which could potentially be incorporated into natural products (as postulated for soil degradation) and also resulted in more mineralization to 14CO2.

the two radiolabels was in the quantity of component assigned as “others” and CO2. These “other” compounds, entirely composed of label specific metabolites, were found in much smaller proportion in the PYM-label, while significantly more NER and CO2 accounted for the residue in the pyrimidine label samples. A 5799

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry

Figure 8. Degradation of PE350 in anaerobic Goose River sediment system.

Figure 9. Optimized data fit for the Goose River aerobic sediment system (OtherM = all label specific metabolites).

Phenyl label samples showed very little 14CO2, but a lot more of the “others” components specific to the PH-label, as compared to PYM-label. Substantial portion of this residue was identified as JW214 and JW215, and sulfonic acid like structures were suspected in the remainder, even though they could not be identified with certainty. Significant formation of NER and slower degradation of the PH-specific degradates enabled benzoic acids to become incorporated into carbon pool, but very little mineralization was noted for PH-label. Assimilation of benzene-ring structures into organic carbon has been documented.9,10

Metabolism in Anaerobic Water Sediment Systems. As was the case in aerobic sediments, the majority of the radioactive residue remained in the surface water phase in the anaerobic sediment systems. The parent compound again metabolized at a faster rate in the Goose River system as compared to the Chula system, and the main route of degradation proceeded along the same pathway as was observed in the aerobic studies. However, anaerobic systems did display differences that are characteristic of reductive transformations, which are common under anaerobic conditions. Initial metabolism of PE350 did generate JW212 in amounts approaching 20% of applied radioactivity at multiple intervals, and the remaining residues appeared to be label specific 5800

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry

Figure 10. Alternate model for Goose River aerobic sediment.

Presumably ring opening of the pyrimidine ring was still operational even under anaerobic conditions. Kinetics Analysis of Sediments. Kinetics modeling of the dissipation of PE350 in soils employed “products” as a compartment to represent all metabolites collectively, because they all amounted to a maximum 20% at any interval. In aerobic sediments, however, JW212 consistently accounted for more than 10% of total residue at multiple sampling intervals, while the remaining metabolites, which were almost entirely label specific, represented the balance of degradation products. NER again accounted for a large portion of the total, and just like soil studies, showed significant differences between the two radiolabels. In sediment studies, 14CO2 was a significant portion of the residue from only the PYM-label. Therefore, JW212 was modeled as a separate compartment and designated as “OtherM”. Optimization of the data (Figure 9) displayed a good fit for all components, and again the model justified the observed degradation processes. Computed rate constants supported the metabolites being largely responsible for the generation of NER, little CO2 formation in PH-label, and accounted for significant differences in the amounts of “OtherM” in the two labels. Just as in soil studies, the optimized model supported PE350 not being sequestered into NER. All transformations in the model displayed in Figure 9 are quite consistent with logical chemical transformations

metabolites (i.e., ring separated metabolites). These included JW217 and B5363 in amounts sufficient to be identifiable via LC/ MS from the PYM-label samples, and JW214, JW215 and possibly A3 from PH-label samples. Unlike aerobic studies, very little 14CO2 was generated from either label, and most of the metabolites underwent incorporation into NER. Formation of compounds, such as JW215, under anaerobic conditions has been shown previously.15,16 Oxidative degradations were much slower in the anaerobic systems, and label specific degradates found ample opportunity to become assimilated into organic matter [NER]. Figure 8 attests to the composition of radioactive residues in the Goose River sediment system. An important point regarding JW215 is that it was only detected in the PH-label samples, which meant that the pyrimidine labeled carbon did not undergo reductive transformations to result in the formation of JW215. Metabolism in the anaerobic system presumably generated JW214, which was methylated, oxidized to the sulfonic acid, or reacted with sulfites and sulfates to produce A3, as evidenced by their detection as minor metabolites via LC/MS. Unlike the aerobic studies, incorporation of degradates into NER, as well as degradation into 14 CO2 for the PYM-label was more prevalent than for the PHlabel in anaerobic sediment. As a result, significantly lesser proportion of label specific degradates, and a higher level of NER and CO2, was found in the PYM-label, as illustrated in Figures 9. 5801

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802

Article

Journal of Agricultural and Food Chemistry

(10) Nowak, K.; Miltner, A.; Gehre, M.; Schaffer, A.; Kastner, M. Formation and Fate of Bound Residues from Microbial Biomass during 2,4-D Degradation in Soil. Environ. Sci. Technol. 2011, 45, 999−1006. (11) Bray, L.; Heard, N.; Overman, M.; Vargo, J.; King, D.; Lawrence, L.; Phelps, A. Hydrolysis of Prosulfuron at pH 5: Evidence for a Resonance-Stabilized Triazine Cleavage Product. Pestic. Sci. 1997, 51, 56−64. (12) Vogels, G.; Van der Drift, C. Degradation of Purines and Pyrimidines by Microorganisms. Bacteriol Rev 1976, 40 (2), 403−468. (13) Sarmah, A.; Sabadie, J. Hydrolysis of Sulfonylurea Herbicides in Soils and Aqueous Solutions: a Review. J. Agric. Food Chem. 2002, 50, 6253−6265. (14) Konopka, A.; Miller, R.; Sommers, L. Microbiology of the Sulfur Cycle. Agronomy Monograph, Tabatabai, M. A., Ed.; Soil Science Society of America, Madison, WI, 1986; Vol 27, pp 23−55. (15) Goldhaber, M.; Kaplan, I. Controls and Consequences of Sulfate Reduction Rates in Recent Marine Sediments. SSSA Special Publication 10, Kittrick, J. A., Fanning, D. S., Hossner, L. R., Eds.; Soil Science Society of America: Madison, WI, 1982; Chapter 2, pp 19−36. (16) Stets, E.; Hines, M.; Kiene, R. Thiol methylation potential in anoxic, low-pH wetland sediments and its relationship with dimethylsulfide production. FEMS Microbiol. Ecol. 2004, 47, 1−11.

The alternative model in which PE350 was allowed to incorporate into NER directly, was tested for the sediment system as well. The observed data could be optimized to achieve a good fit as shown in Figure 10. However, the data fitting required unacceptable rates, one of them being a portion of the NER must be allowed to degrade to JW212. Therefore, kinetics modeling supports the hypothesis that the NER is generated from assimilation of metabolites, and no significant portion of NER results from direct assimilation of PE350. Kinetics assessment for the anaerobic sediment system was also carried out using the model employed in Figure 10, and it gave good fit for anaerobic sediments. A summary of the degradation rates in all systems is summarized in Table 4. Small differences in the DT50s using data from individual labels are within the experimental variability observed with the two labels. In conclusion, metabolism of pyrithiobac sodium in aerobic soil, aerobic sediments, and anaerobic sediments displayed similar degradation pathways. In each system, the parent compound initially degraded to JW212 as the main metabolite, which further metabolized to many label specific compounds. Transformation of label specific products from the pyrimidine ring system preferentially underwent mineralization to 14CO2 while metabolites derived from the phenyl ring afforded a higher level of incorporation into NER. PE350 itself did not become part of nonextractable residues, and this view was supported by kinetics modeling. Kinetics assessment was invaluable in clarifying the conversion of label specific metabolites into NER and CO2, which were the major products of degradation from PE350.



AUTHOR INFORMATION

Corresponding Author

*[email protected]. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Singles, S.; Dietrich, R.; McFetridge, R. Degradation of Pyrithiobac Sodium in Soil in the Laboratory and Field. ACS Symp. Ser. 2002, 813, 207−221. (2) Gondim-Tomaz, M.; Franco, T.; Durrant, L. Biodegradation of Diuron and Pyrithiobac-sodium by White-Rot and Soil Fungi. Contaminated Soils, Sediments and Water 2005, 9, 21−32. (3) Liu, W.; Wu, C.; Guo, W.; Du, L.; Yuan, G.; Wang, J. Resistance Mechanisms to an Acetolactate Synthase (ALS) Inhibitor in Water Starwort. Weed Sci. 2015, 63 (4), 770−780. (4) Nandula, V.; Reddy, K.; Koger, C.; Poston, D.; Rimando, A.; Duke, S.; Bond, J.; Ribeiro, D. Multiple Resistance To Glyphosate And Pyrithiobac In Palmer Amaranth (Amaranthus Palmeri) From Mississippi And Response To Flumiclorac. Weed Sci. 2012, 60 (2), 179−188. (5) Smith, M.; Shaw, D.; Miller, D. In-Field Bioassay to Investigate the Persistence of Imazaquin and Pyrithiobac. Weed Sci. 2005, 53 (1), 121− 129. (6) Webster, E.; Shaw, D. Effect of Application Timing on Pyrithiobac Persistence. Weed Science 1997, 45 (1), 179−182. (7) Harrison, M.; Hayes, R.; Mueller, T. Environment affects cotton and velvetleaf response to pyrithiobac. Weed Science 1996, 44 (2), 241− 247. (8) Roberts, T. Non-Extractable Pesticide Residues in Soils And Plants. Pure Appl. Chem. 1984, 56, 945−956. (9) Barriuso, E.; Benoit, P.; Dubus, I. Formation of Pesticide Nonextractable (Bound) Residues in Soil: Magnitude, Controlling Factors and Reversibility. Environ. Sci. Technol. 2008, 42, 1845−1854. 5802

DOI: 10.1021/acs.jafc.6b02073 J. Agric. Food Chem. 2016, 64, 5793−5802