Pesticide Transformation Products - American Chemical Society

---

Chapter 14

Phytotoxicity of Pesticide Degradation Products

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

Stephen O. Duke, Thomas B. Moorman, and Charles T. Bryson Southern Weed Science Laboratory, Agricultural Research Service, U.S. Department of Agriculture, P.O. Box 350, Stoneville, MS 38776

Pesticide metabolites and degradation residues can accumulate in soils and plants. Relatively little is known of the potential phytotoxicity of these compounds. High levels of certain pesticide metabolites, particularly herbicides, are phytotoxic. For instance, N-methyl-N'-[3-(trifluoromethyl)phenyl]urea, a metabolite of fluometuron, is a weak photosynthesis inhibitor. Crops are not likely to encounter sufficient concentrations of a single weakly phytotoxic pesticide degradation product to cause symptoms in most field situations. Some scientists have speculated that accumulated herbicide degradation products in soils have contributed to a less than anticipated growth of agricultural productivity in cotton and some other crops. However, there is little scientific evidence to validate this and some data indicate that this is not the case. For instance, levels of trifluralin metabolites in soil equivalent to many years of accumulation from high trifluralin use rates have been shown to have no effect on yield of cotton or soybeans. Similar results were obtained with the metabolites of diuron andfluometuronon cotton. In crops genetically engineered to be highly resistant to certain herbicides, metabolite accumulation at high herbicide application rates could be sufficient to cause phytotoxicity in some cases. In thefield,there are generally a large number of different degradation products of many different pesticides in the soil. Little is known of possible interactions of these combinations of metabolites on plant health. The topic of herbicide metabolite phytotoxicity has not been reviewed before and little literature exists on this topic. Much of the literature that exists is "hidden" in papers that emphasize other aspects of herbicides and their degradation products. A non-refereed compilation of the phytotoxicity of thousands of compounds, including some pesticide metabolites is available (2). The pesticide structure-activity and degradation data of many companies is arich,but usually unavailable, source of this type of information. In this short review we have attempted to provide sufficient coverage of the information that is available This chapter not subject to U.S. copyright Published 1991 American Chemical Society In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

14. DUKE ET AL

Phytotoxicity of Degradation Products

189

to allow the reader to appreciate our limited understanding and knowledge gaps of this topic. Some herbicides are applied to the plant in a form that is herbicidally inactive at the molecular site of action, but are metabolized to the active form by the target plant or soil microorganisms. Herbicides that must be activated by metabolic activity are termed proherbicides. For instance, the phenoxyalkanoate and aryloxyphenoxy alkanoate herbicides are generally applied as esters that must be hydrolyzed by the plant or microorganisms in order to become active (2). In the case of the aryloxyphenoxy alkanoates, only one of the two enantiomeric forms is herbicidally active. However, in preemergence treatments, soil microorganisms convert the inactive S enantiomer to the active R enantiomer, so that a mixture of the two forms is as herbicidally active as the exclusive R enantiomer (3)(Figure 1). Whether a proherbicide is phytotoxic through design or chance, its metabolically activated form should be viewed as a herbicide rather than as a degradation product. There may be cases of herbicides that have not been identified as proherbicides because the activity of the metabolites have not been examined in an appropriate in vitro bioassav. Proherbicides can have several potential advantages over their herbicidal metabolites, including, superior penetration and stability, selectivity based on metabolism to the active form, better translocation, and/or prolonged release of the active form. The active forms of proherbicides have been discussed elsewhere (4-6) and will not be considered in this review. The phytotoxicity of known metabolites is discussedfirst,followed by a review and discussion of the possible role of phytotoxic pesticide degradation products on crop yield. The role of pesticide degradation products in the loss of expected yields that have been observed in some crops in the recent past has been a matter of considerable debate, but of little experimentation. Finally, the potential importance of phytotoxicity of herbicide metabolites in crops genetically engineered to be herbicide resistant is discussed. Phytotoxicity of Specific Pesticide Metabolites Several factors should be considered in a survey of the phytotoxicity of pesticide metabolites. For instance, the selectivity of many herbicides, such as the sulfonylureas (7), is based on differential metabolism of the herbicide to herbicidally inactive compounds. In these cases, it is highly unlikely that there would be any significant accumulation of phytotoxic metabolitesfrommetabolism of the herbicide by the crop with which it is used. In target species, the herbicide is either not inactivated by metabolic activity or is metabolized to phytotoxic residues. Metabolites may vary between plant species. Just as herbicides can be selective between plant species, metabolites can differ in their phytotoxicity patterns. Microbial and plant metabolism of herbicides can be very different, so that a phytotoxic microbial degradation product might accumulate in soil that would not be produced in the plant. Few studies of the phytotoxicity of pesticide metabolites have considered these contingencies. Most of the literature describes experiments in which herbicide metabolites of the plant are tested for the same type of herbicidal activity as that possessed by the parent compound. Metabolites might have very different sites of action and/or different species selectivity than the parent compound. Distinct microbial metabolites have seldom been tested for phytotoxicity. Examples of phytotoxic herbicide metabolites are discussed below. Amitrole (lH-l,2,4-triazol-3-amine) is metabolized to 3-(3-amino-s-triazole-lyl)-2-arninoproionic acid (3-ATAL) by plants (8). This product is phytotoxic, but much less phytotoxic than the parent compound. Uncharacterized metabolites of

In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

190

PECTICIDE TRANSFORMATION PRODUCTS

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

amitrole have been reported to be more phytotoxic than the parent compounds (9, Bromoxynil (3,5-dibromo-4-hydroxybenzonitrile) can be converted to 3.5-dibromo-4-hydroxybenzoic acid by a microbial nitrolase. We found this metabolite to be completely ineffective as a photosynthetic inhibitor (unpublished data) and as a growth inhibitor in cotton (see section on herbicide-resistant crops). Interestingly, the 2,6-dibromophenol derivative is a potent growth regulator with auxin-like activity that is probably toxic at sufficiently high concentrations (J). Although this compound is not a known metabolite in lants, decarboxylation of the benzoate derivative of bromoxynil is considered a kely route of further degradation in soil (11). Decarboxylation of benzoates by soil microbes is very common. As mentioned above, the basis for selectivity of sufonylurea herbicides is differential metabolism. The hydroxylated metabolites of chlorsulfuron {2-chloroN-[[(4-methoxy-6-methyl-l,3,5^ formed by plants have reported to be nonphytotoxic (7). It is likely that any plant-derived metabolites of sulfonylurea herbicides that accumulate to significant levels are non-phytotoxic because selectivity could not be based on metabolism if phytotoxic metabolites accumulated. Phytotoxic intermediates might only accumulate to insignificant levels in non-target plants. Three metabolites of diclobenil (2,6-dichlorobenzonitrile) are phytotoxic (12) . One of the metabolites (2,6-chlorobenzoic acid) has strong auxin-like activity and the other two (3 and 4-hydro^-2,6-dichlorobenzonitrile) have stong contact (rapid foliar desiccation) activity. The selectivities of the parent compound, the benzoate, and the hydroxydiclobenils were all different. All three of these metabolites were determined to be as herbicidally active as the parent compound, but only the hydroxydichlobenils had the same mode of action as the parent compound. Chlorosis of the leaf margins of apple trees was attributed to 2.6-dichlorobenzamide, a product of degradation of dichlobenil by soil microbes (13) . This is an excellent example of the importance of examining metabolites for different types of activity and different selectivity than the parent compound. Fluometuron {^N-dimethyl-N'-p-itrifluoromethylJphenyljurea} inhibits photosynthesis by inhibiting photosystem II (PSII). This effect can be rapidly detected by measuring variable fluorescence increases in plant tissues in which PSII is inhibited. The metabolic degradation pathway of fluometuron is shown in Figure 2. We and others have found that N-methyl-N -[3-(trifluoromethyl)phenyl]urea (DMFM), thefirstmetabolite of fluometuron , is a weak photosynthesis inhibitor (Figure 3) (15, 16). In a leaf disc assay, Rubin and Eshel (15) found it to be about two- tofive-foldless active than fluometuron as a photosynthesis inhibitor in cotton and redroot pigweed (Amaranthus retroflexus L). However, it was almost as phytotoxic as the parent compound to whole plants (cotton and pigweed) when treated by soil incorporation of the compounds. Pigweed was much more sensitive to both compounds than was cotton. No other principle metabolites of fluometuron had a measureable effect on photosynthesis, however, two metabolites, TFMPU [3-(trifluoromethyl)phen5durea] and TFMA [3-(trifluoromethyl)aniline], were weakly phytotoxic in both soil and nutrient solution to pigweed and foxtail and to cotton only in nutrient solution. No data have been generated to indicate that DMFM or any other fluometuron metabolites cause phytotoxicity in the field to cotton, a species that is tolerant to fluometuron (17). However, no data are available on the potential for phytotoxicity of fluometuron metabolites to species that are not tolerant to the parent compound. DMFM is the principle metabolite of fluometuron in both cotton and cucumber, accumulating to higher levels than the parent compound within 12 and 96 h of application, respectively

E

In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

14. DUKE ET AL

Phytotoxicity of Degradation Products

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

Further ESTER S

ACID

o

r

191

degradation

conjugation

Hydrolysis

I n v e r s i o n of configuration R

Figure 1. Transformation of enantiomers of fluazifop in different environments. (1) Absent in soil and plant; (2) Rapid in soil and plant; (3) Rapid in soil; absent or very slow in plant; (4) Predominant in soil; significant in plant; (5) Limited in soil; significant in plant. (Adapted from ref. 3; reproduced with permission from ref. 2. Copyright 1988 Marcel Dekker)

Microbial biomass or Bound residue

II

III

IV

Figure 2. Degradation pathways of the substituted urea herbicides (I) fluometuron (X, = C F . , X, = H) or diuron (X, = Cl, X, = Cl). Metabolites offluometuronare DMFM (II), TFMPU JET), Sid TFMA (IV). Diuron metabolites are DCPMU (Π), DCPU (HI), DCA

In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

192

PESTICIDE TRANSFORMATION PRODUCTS

(18). Removal of the second methyl group of fluometuron completely eliminates any effect on photosynthesis (Figure 2) (16) or whole plant growth (79). Metabolites of other PS II inhibitors have been shown to have weak activity. For instance, Matsunaka (20) found 3,4-dichloroaniline (DCA), a plant derivative of propanil ^i-3,4-dichloropheiyl)propamide], to be three orders of magnitude less effective as a photosynthesis inhibitor than propanil. DCA is also a metabolite of substituted ureas such as diuron J^'-(3,4-dichlorophenyl)-IiN-dimethylurea]. Although there is no good evidence that the non-selective herbicide glyphosate [N-(phosphonomethyl)gtycine] is significantly metabolized by plants, it is metabolized to sarcosine, ammomet^phosphonic acid (AMPA), N-methylphosphinic acid, glycine, Ν,Ν-dimethylphosphinic acid, and hydroxymethylphosphonic acid by microorganisms (21). Only AMPA has been reported to significantly inhibit plant growth (22, 23). However, it was a much weaker growth inhibitor than glyphosate in ooth studies and there was no indication that its phytoxicity was evoked by inhibition of the shikimate pathway (the mechanism of action of glyphosate). In fact, AMPA stimulated synthesis of anthocyanin, a shikimate pathway product (22). Since glyphosate can be completely metabolized and used as a sole C or Ρ source by sou microbes (24), there is little possibility that AMPA causes any phytotoxicity infieldsituations. Methazole [2-(3,4-dicWorophenyl)-4-methyl-l,2,4-oxadiazolidine-3,5-dione] is a proherbicide that is metabolized to the potent PS II inhibitor l-(3,4-dichlorophenyl)-3-methylurea (DCPMU) (4,5,25). DCPMU is also a metabolite of diuron [N'-iS^dichlorophenylVNjN-dimethylurea] (Figure 1) and is 10 to 50 % as phytotoxic as diuron (26). Another metabolite, l-(3,4-dichlorophenyl)urea fDCPU), is about a twentieth as phytotoxic as DCPMU or methazole (27) and 0 to 12 % as phytotoxic as diuron (26). Hydroxylation of s-triazines to form 2-hydroxy derivatives results in metabolites with no herbicidal activity (28). Monodealkylation of atrazine r6-chloro-N-ethyl-N'-(l-methyle^ (formerly 2

ω

Fluometuron DMFM

ι—

ο ω

> ω oc

Control

Time (s)

Figure 3. Effects of 10 μΜ fluometuron and several of its metabolites on variable chlorophyll fluorescence in cotton leaf discs incubated for 2 h on the test solution. Assay procedures were as previously described (14). 240

Concentration (M)

Figure 4. Influence of a 24-h treatment with trifluralin (squares), TR-36M (triangles), or TR-40 (circles) on the mitotic index (cells in division per 1000 cells) of goosegrass root meristem cells (32). In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

194

PESTICIDE TRANSFORMATION PRODUCTS

(34). Methyl parathion degrades to 4-nitrophenol after being sprayed onto plant foliage of cotton and lettuce; however, it is not clear whether degradation occurs on or in the leaf (35). Oxidative phosphorylation is inhibited by 4-nitrophenol (36). Similarly, the phytotoxicity of fenitrothion [Ο,Ο-dimethyl cabbage was attributed to 4-mtrocresol (37). Whether other effects of methylparathion on plants, such as inhibition of lipid synthesis (38), are due to metabolites of the insecticide or to the parent compound has not been determined. Some insecticides or their metabolites inhibit degradation of herbicides, thereby increasing herbicide phytotoxicity (20,39). The above examples demonstrate that pesticide metabolites can be phytotoxic by several mechanisms. They can synergize herbicides by interference with herbicide degradation. They can have entirely different mechanisms of action and selectivity than their parent compounds. Soil metabolites can also have different modes of entry into the plant than the parent compound (e.g., soil metabolites of foliar-applied herbicides). Although there are many possible mechanisms by which pesticide metabolites might affect plants, whether they significantly affect crop yields in field situations is unknown or unclear with all pesticide-crop combinations. Effects of Pesticide Metabolites on Crop Yield Long-term pesticide use has several potential hazards, including the accumulation of parent herbicides and/or degradation products that are extremely persistent. Pesticide metabolites can be produced in soil by microbial or chemical processes or may be initially formed in plants and later enter the soil in plant residues. The effects of these metabolites, regardless of their source, may be direct or indirect. Metabolites can affect vital plant processes directly (e.g., photosynthesis) or indirectly (eg., by altering host plant resistance to a pathogen or insect or by affecting the composition of soil microflora). Long-term pesticide use has the potential effect of altering microbial communities in the soil that drive the cycling of plant nutrients, which could result in a crop response. Aside from quantitative effects of degradation products on yields, these compounds may also have effects on food quality. Assessing the effects of degradation products on yields requires quantitative knowledge of the crop response to a particular dose of toxicant and the levels of metabolites available to the crop underfieldconditions. Evidence of pesticide metabolite effects on crop yields have been obtainedfromexperiments comparing yield on metabolite-treated soil to that on nontreated soil. Other, more circumstantial evidence can be gatheredfromexperiments that compare the yields of pesticide-treated crops to that of crops grown without pesticides. These types of experiments are advantageous in determining the cumulative effects of both pesticides and metabolites under realistic farm management conditions. Comparisons of pesticide-treated and untreated crops can be confounded by other factors that affect crop yield, including pest pressure, climatic effects, or other agronomic factors that contribute to yield. The adverse effects of pesticides and their metabolites on crop yields can only be assessed in the absence of losses due to uncontrolled pests. Yield increases due to pest control can be greater than potential losses due to metabolite-induced damage. It is often difficult to separate the effects of degradation productsfromthe parent compound because of their potential similarity in mechanisms of action and their simultaneous presence in the soil. Also, many pesticides (particularly herbicides) have rate-dependent effects on crops that have only marginal tolerance, or rotational crops that may have no tolerance to a pesticide used on a previous crop.

In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by UNIV MASSACHUSETTS AMHERST on August 1, 2012 | http://pubs.acs.org Publication Date: March 21, 1991 | doi: 10.1021/bk-1991-0459.ch014

14· DUKE ET AL.

Phototoxicity of Degradation Products

195

Pesticide degradation products are highly variable in their concentration and persistence in soils. Pesticides that have considerable structural similarity may have similar or common metabolites produced in soils. For example, the weak photosynthesis inhibitor 3,4-dichloroaniline is common metabolite of the photosynthesis-inhibiting herbicides diuron, linuron [N'-(3,4-dichlorophenyl)-Nmetho^r-N-methylurea], propanil ^-(3,4-dicMorophenyl)propanamidej, and swep (methyl 3,4-dichlorocarbanilate). Many reports indicate that, as pesticides degrade in soil, only small amounts of metabolites appear. If it is assumed that a degradation product has a half-life of twice that of the parent pesticide, the maximum metabolite concentration in soil from a single application would be 50% of the applied amount (40). The available evidence for many different classes of herbicides supports the view that that metabolites generally do not accumulate. For instance, after 20 years of continuous atrazine use, metabolite concentrations ranged from 14 to 296 Mg/kg in soil (41\ Concentrations in the roots and shoots of oats grown in this soil rangedfrom73 to 116 j*g/kg, but oat growth was not different from growth in soil without atrazine use. In addition to the formation of extractahle metabolites, studies with C-labeled pesticides indicate that substantial amounts of the pesticide and/or metabolites are converted to bound residues. Typically, between 20 and 70% of the applied pesticide may become bound residue (42). Bound residues include pesticides and degradation products that have become covalently bonded to soil organic matter or compounds that have diffused into the organic or mineral matrix sufficiently to become unextractable by conventional solvent extraction methods. This process occurs fairly quickly and concurrently with the pesticide degradation process. The unextractable nature of bound residues suggests low bioavailability. However, studies with triazine herbicides showed that strong extraction procedures recovered 10 to 15% of parent pesticidefromthe bound residue (43, 44). In addition to prometryn [^N'-bisil-methylethyn-ô-imethylthioJ-l^jS-triazine2,4-diamine] (20% of added), the metabolites nydroxypropazine [2-hydroxy4,6-b^isopror^lamino)-s-triazineJ (7%) and deisoproiylpropetryn [2-methylthio)-amino-6-(isopropylamino)-s-triazme (11%) were extracted with strong base from the bound residue pool (43). Smaller amounts of pesticide and metabolites are releasedfrombound forms by natural processes. Bioavailability, defined as residue uptake by plants, is shown in Table I. These residues are bioavailable in quantities generally belqjy 2% of the applied pesticide. The residues are generally measured as bound C or C m plants and generally represent both uncharacterized metabolites and parent pesticides. Lee et al. (45) reported that the uncharacterized bentazon residues releasedfrombound forms (see Table I) had no effect on maize growth and that maize took up approximately 40% of freshly added [ C]bentazon [3-(l-methylethyl)-(lH)-2,l,3-benzothiadiazin-4(?H)-one 2,2,-dioxide]. All available evidence suggests that while bound residues in soil represent a large pool of parent pesticides and metabolites, only a small fraction of the residues are available to plants. Plant uptake of pesticide metabolites has been examined in experiments using fjeshly added metabolites instead of bound residues. Barley uptake of [ C]4-chloroaniline after 20 weeks was only 0.3% of the 1.25 mg/kg added to soil (50). Ryegrass treated with a glucoside conjugate of hydroxymonolinuron only tooJc up between 0.19 and 0.23% of the added conjugated metabolite [(phenyl^ C)-3-(4-chlorophenyl)-l-(hydroxymethyl)-l-methoxyurea-)3-Dglucoside] (51). The limited plant uptake of these pesticide metabolites may be related to their relatively strong sorption to soils, their rapid transformation in soil, or poor efficiency of root uptake. Nevertheless, these results tend to 14

4

In Pesticide Transformation Products; Somasundaram, L., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

196

PESTICIDE TRANSFORMATION PRODUCTS

confirm the results obtained with the less characterized bound residues discussed previously. Chlorinated anilines taken up by plants appear to become bound principally to the ligninfraction(52). Table L Plant Uptake ofPesticide Residues ReleasedfromBound Forms by Natural Processes a

Uptake Bound* -(%ofapplied C)-

Plant

26 44

2.3 (34) 1.8(42)

Maize Maize

Parathion

16