Investigation of the herbicide glyphosate 
and the plant growth regulators chlormequat and mepiquat 
in cereals produced in Denmark. 

Food Additives and Contaminants, 2001, v.18, n.10 Oct01

Kit Granby* and Martin Vahl

Institute of Food Research and Nutrition, Danish Veterinary and Food Administration, M6rkh6j Bygade 19, DK-2860 S6borg Denmark

(Received 11 September 2000; revised 5 December 2000; accepted 5 January 2001)

Keywords: glyphosate, AMPA, chlormequat, mepiquat, LC-MS, cereals, grain

* To whom correspondence should be addressed. e-mail: kgr@fdir.dk

An LC-MS/MS method for analysing glyphosate and aminomethylphosphonic acid (AMPA) in cereals was developed. The method is based on extraction with water and detection of the ions from the fragmentation m/z 170--->88 (glyphosate) and m/z 112--->30 (AMPA), using electrospray interface in the positive mode. Investigation from the harvests of 1998 and 1999 showed residues of glyphosate and/or its degradation product AMPA in more than half of the cereal samples produced in Denmark. The average concentration of glyphosate in 46 samples from the 1999 harvest was 0.11 mg/kg compared with 0.08 mg/kg for the 1998 harvest (n = 49). Thus, the figures were well below the maximum residue limit (MRL) and no violations were observed. The plant growth regulators chlormequat and/or mepiquat were investigated in cereals from the Danish harvest of 1999 where 83% of the samples contained chlormequat (n = 46) compared with 87% of the samples from the 1997 harvest (n = 52) . The average concentration of chlormequat in 1999 was 0.32 mg/kg compared with 0.23 mg/kg in 1997. At 2.9 mg/kg, one sample of wheat bran was exceeding the MRL of 2 mg/kg for wheat. The intakes of the pesticides through the diet of cereals were estimated to comprise 0.04% of the acceptable daily intake (ADI) for glyphosate and 1 % of the ADI for chlormequat for an adult Dane.

Glyphosate (N-phosphonomethyl glycine) is used world wide as a herbicide, e.g. towards couch grass, and for withering of commodities before harvest. The pre-harvest interval is 10 days. It is a systemic pesticide that, when translocated throughout the plant, inhibits the production of some aromatic amino acids essential for plant growth (Franz et al. 1997). The major natural metabolite in plant and soil is aminomethylphosphonic acid (AMPA). Glyphosate is the most frequently used pesticide in Denmark. In 1999 761 tons were used in Denmark and in 1998 881 tons were used, comprising ca 20% of the total amount of pesticide used in Denmark or 36% of the total herbicide used (Danish EPA 2000).

The plant growth regulator chlormequat (2-chloro-N,N,N-trimethylethylammonium) and mepiquat (1,1dimethylpiperidinium) are also frequently used in Denmark. The major usage is as stalk stabilizers. Chlormequat inhibits cell elongation; hence it shortens and strengthens the stem, thereby preventing the cereals from lodging during the wintertime. In 1999 241 tons of chlormequat chloride and 1.4 tons of mepiquat chloride were used. The chlormequat used comprised 94% of the total amount of plant growth regulators used (Danish EPA 2000).

In 1997 the Danish Veterinary and Food Administration developed an LC-MS/MS method for analysing chlormequat and mepiquat in cereals (Vahl et al. 1998, Juhler and Vahl 1999).

The methods developed for analyses of glyphosate in cereals have often been based on HPLC with fluorescence detection (Moye et al. 1983, Cowell et al. 1986, Tuinstra and Kienhuis 1987, Wigfield et al. 1994, Sen and Baddoo 1996, Hogendoorn et al. 1999). These methods require a clean-up, e.g. on resin columns. However LC-MS/MS methods may be preferable as they are more specific and the need for sample cleanup before detection is reduced. Vreeken et al. (1998) developed an LC-MS/MS method for glyphosate in water using derivatization with 9-fluorenyl methoxycarbonyl chloride (FMOC) before LC-MS/MS. In the present paper, an LC-MS/MS method based on extraction with water and online clean-up before ion chromatography separation and MS/MS detection without any derivatization is presented.

The analytical methods have been used to investigate the contents of glyphosate, AMPA, chlormequat and mepiquat in cereals produced in Denmark. Results using cereals from the 1997, 1998 and 1999 harvests are presented.

Sampling

The investigation includes cereals for human consumption. Until the quality of the grain is known it is not certain whether the grain is to be used for bread production or, for example, for feedingstuff. Therefore the samples were mostly collected at wholesalers and mills, where the quality is known, and only a few samples were collected at the major farms and bakeries. This implies that samples often are a mixture of grain from several farms. Samples of 1 kg or more were collected randomly from the grain harvests of 1997, 1998 and 1999 and investigated as part of the Danish National Pesticide Survey. The samplings were performed by the Municipal Food Control units in Denmark.

Materials

Pesticide standards of glyphosate (98%), aminomethylphosphonic acid (AMPA) (99%), chlormequat chloride (99%) and mepiquat chloride (99%) were purchased from Dr Ehrenstorfer (Augsburg, Germany). Labelled glyphosate, 2-¹³C¹⁵N-glyphosate (99% ¹³C, 98% ¹⁵ N) was purchased from Cambridge

Isotope Laboratories (Massachusetts, US). Labelled ¹³C-chlormequat was synthesized as described by Vahl et al. (1998). All solvents were of HPLC grade and all other chemicals were of analytical grade. Extraction tubes, vials, etc. were made of plastic.

The samples were analysed on a triple quadrupole LC-MS/MS instrument. The HPLC system was configured with a Jasco PU980 gradient pump, an HP 1050 isocratic pump, an HP1050 autosampler and a Rheodyne column-switching valve. Electrospray ionisation mass spectrometry was performed using a Micromass Quattro 11 instrument with Masslynx software.

The method used for analysing chlormequat and mepiquat in cereals was developed in 1997 (Vahl et al. 1998, Juhler and Vahl 1999). The method is based on extraction with methanol-water-acetic acid; cleanup on a C18 cartridge, separation by liquid chromatography on a Spherisorp S5 ODS1 column and detection by MS/MS using 13C-chlormequat as internal standard for quantification. The fragmentation pathways were for chlormequat m/z 122--->58, for ¹³C-chlormequat m/z 125--->61 and for mepiquat m/z 114--->98. The product ions were detected using electrospray MS/MS in the positive mode. The limit of detection (LOD) based on three times the signal to noise level was 0.004 mg/kg for chlormequat and 0.001 mg/kg for mepiquat.

Extraction and clean-up. Analytical portions of 3 g were extracted twice with 25.0 ml deionized water for 10 min by ultrasonication. The extracts were centrifuged at 500 x g for 10 min and 10 ml of the combined supernatant were filtered through a 0.20 µm filter (Sartorius Minisart SPR15). The first 1-3 ml of the filtrate was discarded. A mixture of 0.5 ml filtrate, 25 µl 5% nitric acid and 25 µl 2-¹³C¹⁵N-glyphosate (50 µg/ml in deionized water) was transferred to autosampler vials for analysis by LC-MS/MS. Reference solutions of glyphosate and AMPA were mixed with blank extracts to make five matrix-matched standards at 0.0015 µg/ml-0.15 µg/ MI.

Electrospray ionization mass spectrometry in the positive mode was performed. The electrospray capillary was at 3.75 kV, the cone at 30 V and the extractor at 2 V. The ion source temperature was 90°C, the desolvation temperature 200°C and the flow rates for nitrogen desolvation gas and spray gas 500 l/h and 20 l/h respectively. The pressure of the argon used for CID (collision induced dissociation) was 7.5 x 10-4 mbar and the collision energy 10 eV. Data were acquired in MRM (multiple reaction monitoring) mode. The ions monitored were m/z 88, product ion of the glyphosate protonated molecule (m/ z 170), m/z 90, product ion of the ¹³C¹⁵N-glyphosate protonated molecule (m/z 172) and m/z 30, product ion of the AMPA protonated molecule (m/z 112). Quantification of glyphosate and AMPA was done by the internal standards method using the ratio of the areas of the m/z 88 (glyphosate) or the m/z 30 (AMPA) and the m/z 90 (¹³C¹⁵N-glyphosate) product ions in the mass chromatograms.

Recovery studies were done as true replicates at three levels. A reagent blank, a sample of blank reference and a sample of positive reference were included in each series of analysis.

Results and discussion

Analytical method for glyphosate and AMPA

The analytical method for glyphosate and AMPA worked well, especially for glyphosate. The internal standard ¹³C¹⁵N-glyphosate was used to compensate for any matrix-dependent response on the MS/MS-detector. When the glyphosate standards were compensated for the response of the internal standard ¹³C¹⁵N-glyphosate, the relative responses of matrix-matched standards and solvent standards showed no significant difference. However, the matrix standards of AMPA showed significantly lower responses compared with solvent standards. The mass chromatograms of the product ions of the internal standard (13C15Nglyphosate), glyphosate and AMPA appear in figure 1. The limit of detection (LOD) based on three times the signal to noise level was 0.01 mg/kg for glyphosate and 0.04 mg/kg for AMPA. Recoveries from spiked samples, within series coefficient of variation (repeatability) and between series coefficient of variation (reproducibility) are shown in table 1. The recoveries are acceptable for both glyphosate and AMPA at the three spiking levels 0.1  0.2 and 0.5 mg/kg. The repeatabilities and reproducibilities for glyphosate are fine, while those for AMPA are poor at low concentration levels.

Results of glyphosate, AMPA, chlormequat and mepiquat in cereals from the harvests of 1997, 1998 and 1999

The results of glyphosate and AMPA in 49 cereal samples from 1998 and 46 cereal samples from 1999 are shown in figures 2 and 3. For both the 1998 and 1999 harvests glyphosate was found in more than half of the samples (LOD 0.01 mg/kg), while AMPA was present in less than 15% of the samples (LOD 0.04 mg/kg). On average glyphosate was present in ca 60% of the samples. The highest content of glyphosate in 1999 was found in a sample of wheat bran (1.62mg/kg) and in 1998 in a sample of barley grain (1.25 mg/kg). The highest contents of AMPA were one order of magnitude lower. The European Communities (EC) have established maximum residue limits (MRLs) for glyphosate in cereals at 5 mg/ kg for wheat, rye; triticale; 20 mg/kg for barley, oat sorghum and 0.1 mg/kg for other cereals. The level 0.1 mg/kg is defined as the analytical detection limit. Hence the contents of glyphosate in the cereals investigated were well below the MRLs. In 1999, in addition to the 46 samples investigated four samples of organically grown cereals were analysed. A sample of organically grown rye contained 0.09 mg/kg glyphosate. The content has been reported to the local authorities. 

A comparison of the contents of glyphosate in 1999 and 1998 (figure 3) shows about the same concentration levels. The average concentration of glyphosate in 1999 was 0.11 mg/kg and in 1998 0.08 mg/kg. There has been a debate in the Danish press about the usage of glyphosate and the plant growth regulators, and in order to meet the consumer requirements, today some bread is labelled `grown without usage of `Roundup' or plant growth regulators'. Also in 1999, the Danish agriculture organizations and bakeries announced in the Danish press that they would not use glyphosate on grain for bread production (human consumption). This did not seem to reduce the content of glyphosate from the 1999 harvest as it did not decline compared with the investigation from 1998, but a future investigation of glyphosate in cereals from the 2000 harvest will show if the agreement has affected the concentration levels. 

In 1999 the plant growth regulator chlormequat was found in 83% of the cereal samples compared with 87% in 1997. The high frequency of findings may not directly reflect the usage of the pesticide in the fields, as many samples may be mixtures from several farmers (the same is the case for the glyphosate findings). However, chlormequat is present in most of the cereal products used in Denmark. The average concentration of chlormequat was 0.32 mg/kg in 1999 compared with 0.23 mg/kg in 1997 (figure 3). Mepiquat is mostly found in rye samples and at concentration levels ca one order of magnitude lower than chlormequat. 

The European Communities (EC) have established MRLs for chlormequat in cereals at 2 mg/kg for wheat, rye, barley, 5 mg/kg for oat and 0.05 mg/kg for other cereals. In addition the Danish authorities have established a provisional MRL of 1 mg/kg for mepiquat in barley and rye. The highest figure of chlormequat (2.9 mg/kg) was found in a sample of wheat bran, which exceeded the MRL of 2 mg/kg for wheat. This content has been reported to the local authorities. Although the number of wheat bran and wheat flour samples are relatively few, it may be concluded from figure 3 that the relative contents of glyphosate and chlormequat are highest in the bran, lower in the grain and lowest in the flour. The concentrated content in bran is expected and in accordance with previous surveys. 

Table 1. Performance data of the analytical method for glyphosate and AMPA.

                           Coefficient of variation     
                     Within series       Between series 
                     Degree of           Degree of               Recovery 
                      freedom   CV(%)     freedom   CV(%)      n   Mean(%)
Glyphosate (mg/kg)
   Wheat spiked at 0.1   4      12          4       14         10    84
   Wheat spiked at 0.2   5       8          5       17         12   100
   Wheat spiked at 0.5   4      11          4       15         10    87
   House reference 1.3                     13        7
AMPA (mg/kg)
   Wheat spiked at 0.1   4      59          4      105         10   113
   Wheat spiked at 0.2   5      14          5       39         12   104
   Wheat spiked at 0.5   4       9          4       20         10    97
   House reference 0.06                    13       86

Figure 2. Contents (mg/kg) of glyphosate, AMPA, chlormequat and mepiquat in cereals produced in Denmark (n - 1997: 52; 1998: 49 and 1999: 46) .

Figure 3. Comparison of' the average contents of glyphosate (1998/1999) and chlormequat (1997/1999) for different cereal products and for all samples. The figures above the columns refer to the number of samples investigated for each product.

Table 2. Comparison of chlormequat with other surveys: sampling year, country code, no. (n) of .sample analysed/with residues, commodity and concentration range (mg/kg).   

                       Glyphosate                                            Chlormequat                      
1990,            C,    n = 55/1  Barley 4.7          1995,            S,      n = 48/3   Rye 0.3-0.5
Wigfield et al.  1994                                Andersson et al. 1996

1998,            DK,   n = 4/3   Barley 0.22-1.25    1996,            S,      n = 52/0   Rye 0.00
This study                                           Andersson and Pålsheden

1999,            N,    n = 8/2   Rye > 0.01-0.03     1997,            S,      n = 25/2   Rye > 0.5-0.52
Varran et al.    2000                                Andersson and Pålsheden

1999,            DK,   n = 12/2  Rye 0.013-0.015     1997,            DK,     n = 10/9   Rye 0.03-1.08
This study                                           This study

1990,            C,    n = 76/1  Wheat 4.3           1998,            S,      n = 28/2   Rye > 0.4-0.42
Wigfield et al.  1994                                Andersson et al. 1999

1998,            DK,   n = 27/17 Wheat 0.01-0.44     1999,            DK,     n = 12/9   Rye 0.02-1.02
This study                                           This study

1999,            DK,   n = 20/14 Wheat 0.01-0.87     1999,            N,      n = 5/5    Rye > 0.05-0.13
This study                                           Varran et al. 2000

1999,            S,    n = 47 /3 Wheat > 0.1-0.17    1997,            DK,     n = 30/27  Wheat 0.004-0.62
Andersson et al. 2000                                This study

1999,            N,    n = 79/30 Wheat > 0.01-0.19   1999,            DK,     n = 20/15  Wheat 0.004-0.62
Varran et al.    2000                                This study

                                                     1999,            N,      n = 39/16  Wheat > 0.05-0.33
                                                     Varran et al. 2000

Comparison with other surveys

A comparison of the chlormequat contents in rye and wheat with other surveys showed higher concentrations in the present study compared with the other Scandinavian surveys (table 2). The glyphosate contents in rye were the same but the contents in wheat were higher in the present study compared with other Scandinavian surveys. The Danish results were lower compared with two studies of glyphosate residues in Canada (table 2).

Intake assessment

FAO/WHO has established an Acceptable Daily Intake (ADI) for glyphosate at 0.3 mg/kg body weight (bw) and for chlormequat at 0.05 mg/kg bw. The average daily diet of cereals based on 100 g wheat, 59 g rye, 7.7 g oat and 0.4 g barley are consumption data from a national survey of dietary habits carried out by the Danish Veterinary and Food Administration (Lyhne et al. 1995). The method of this survey was based on a combination of personal interviews and self-administered diet records from 3098 persons. Based on the average concentrations of the pesticides, the intake of glyphosate from cereals for an adult at 60 kg is 7 µg/day and the intake of chlormequat is ca 40 µg/day. These intake calculations do not include processing factors from milling and baking. The pesticide intake from cereals corresponds to 0.04% of ADI for glyphosate and 1 % of ADI for chlormequat for an average adult Dane. In addition to the contribution from cereals, humans are exposed to the pesticides through other commodities, e.g. chlormequat may be present in pears and glyphosate in peas. However, so far the Danish National Pesticide Survey has concentrated the efforts on residue measurements of glyphosate, AMPA, chlormequat and mepiquat in cereals.   

Acknowledgements

We acknowledge the skilled technical assistance of Inge Schröder, Lene Bai Jensen and Mette Engel.  

References  

ANDERSSON, A., and Pålsheden H., 1997, Pesticide residue monitoring of fruit, vegetables and cereals in Sweden 1996. Report concerning directives 90/642/EEC, 86/362/EEC and Commission Recommendations 96/199/EC, 96/738/EC. National Food Administration, Sweden, August, 1997.

ANDERSSON, A., and PÅLSHEDEN, H., 1998, Pesticide residue monitoring of fruit, vegetables and cereals in Sweden 1997. Report concerning directives 90/642/EEC, 86/362/EEC and Commission Recommendation 96/738/EC. National Food Administration, Sweden, August, 1998.

ANDERSSON, A., PÅLSHEDEN, H., BERGH, T., and JANSSON, A., 1996, Pesticide residues in food of plant origin 1995. National Food Administration Report 10/96, Sweden.

ANDERSSON, A., JANSSON, A., and JAHRL, S., 1999, Pesticide residue monitoring of fruit, vegetables and cereals in Sweden 1998. Report concerning directives 90/642/EEC, 86/362/EEC and Commission Recommendation 97/822/EC. National Food Administration, Sweden, October, 1999.

ANDERSSON, A., JANSSON, A., and STRANDLER, S., 2000, The Swedish monitoring of pesticide residues in food of plant origin 1999. Report concerning directives 90/642/EEC, 86/362/EEC and Commission Recommendation 1999/333/EC. National Food Administration, Sweden, October, 2000.

COWELL, J. E., KUNSTMAN, L. J., NORD, P. J., STEINMETZ, J. R., and WILSON, G. R., 1986, Validation of an analytical residue method for analysis of glyphosate and metabolite: An interlaboratory study. Journal of Agricultural and Food Chemistry, 34, 955-960.

DANISH ENVIRONMENTAL PROTECTION AGENCY (DANISH EPA), 2000, Pesticide statistics 1999. Report 11/2000 (Orientering fra Milj6styrelsen) (in Danish).

FRANZ, J. E., MAO, M. K., and SIKORSKI, J. A., 1997, Glyphosate: A Unique Global herbicide, ACS Monograph 189 (Washington, DC: American Chemical Society).

HOGENDOORN, E. A., OSSENDRIJVER, F. M., DIJKMAN, E., and BAUMANN, R. A., 1999, Rapid determination of glyphosate in cereal samples by means of pre-column derivatisation with 9-fluorenylmethyl chloroformate and coupled-column liquid chromatography with fluorescence detection. Journal of Chromatography A, 833, 6773.

LYHNE, N., FAGT, S., GROTH, M. V., HARTKOPP, H. B., MØLLER, A., OVESEN, L., and WARMING, D. L., 1995, Danish Food Habits 1995, Main results (In Danish), The Danish Veterinary and Food Administration, Report 235, June, 1995.

JUHLER, R. K., and VAHL, M., 1999, Residues of chlormequat and mepiquat in grain. Results from the Danish national pesticide survey. Journal of AOAC International, 82, 331-336.

MOYE, H. A., MILES, C. J., and SCHERER, S. J., 1983, A simplified high-performance Liquid chromatographic residue procedure for the determination of glyphosate herbicide and (aminomethyl) phosphonic acid in fruit and vegetables employing postcolumn fluorogenic labelling. Journal of Agricultural and Food Chemistry, 31, 69-72.

SEN, N. P., and BADDOO, P. A., 1996, Determination of glyphosate as N-nitroso derivative by high performance liquid chromatography with chemiluminescence detection, International Journal
Environmental and Analtyical Chemistry, 63, 107-117.

TUINSTRA, L. G. M. T., and KIENHUIS, P. G. M., 1987, Automated two-dimensional HPLC residue procedure for glyphosate on cereals and vegetables with post-column fluorogenic labelling, Chromatographia, 24696-24700.

VAHL, M., GRAVEN, A., and JUHLER, R. K., 1998, Analysis of chlormequat residues in grain using liquid chromatography-mass spectrometry (LC-MS/MS). Fresenius Journal of Analytical Chemistry, 361, 817-820.

VARRAN, G. T., HOLEN, B., and OTLO, A., 2000, Norwegian monitoring of pesticide residues in fruit, vegetables and cereals, The results of 1999. Norwegian Food Control Authority, 2000.

VREEKEN, R. J., SPEKSNIJDER, I., BOBELDIJK-PASTOROVA, I., and NOIJ, TH. H. M., 1998, Selective analysis of the herbicide glyphosate and aminomethyl phosphoric acid in water by on-line solid-phase extraction-high-performance liquid chromatography-electrospray ionization mass spectrometry. Journal Chromatography A, 794, 187-199.

WIGFIELD, Y. Y., DENEAULT, F., and FILLION, J., 1994, Residues of glyphosate and its principle metabolite in certain cereals, oilseeds, and pulses grown in Canada, 19901992. Bulletin of Environmental Contamination and Toxicology, 53, 543-547.