Risk assessment for pirimiphos-methyl residues resulting from crosscontamination 1

Transcription

1 EFSA Journal 2011;9(11):2436 REASONED OPINION Risk assessment for pirimiphos-methyl residues resulting from crosscontamination 1 European Food Safety Authority 2 European Food Safety Authority (EFSA), Parma, Italy SUMMARY In accordance with Article 43 of Regulation (EC) No 396/2005 3, the European Commission submitted to EFSA a request to provide a scientific opinion on the setting of temporary MRLs for pirimiphosmethyl based on information on cross contamination. The request was supported by studies/data provided by stakeholders (Intercéréales, Euromaisiers, Fediol and Association des Amidonniers et Féculiers) demonstrating that for maize and oilseeds a MRL of 0.5 mg/kg would be appropriate. EFSA derives the following conclusions based on the information submitted in support of the request, the Good Agricultural Practices (GAPs) notified by Member States, the Draft Assessment Report (DAR) prepared under Council Directive 91/414/EEC (United Kingdom, 2003), the Commission Review Report on primiphos-methyl (EC, 2005), the conclusion on the peer review of the pesticide risk assessment of the active substance pirimiphos-methyl (EFSA, 2005), the JMPR Evaluation report (FAO, 2003, 2006) as well as the conclusions from the previous EFSA opinions on pirimiphos-methyl (EFSA, 2007, 2008, 2009). The toxicological profile of pirimiphos-methyl was assessed in the framework of the peer review under Directive 91/414/EEC and the data were sufficient to derive an ADI value of mg/kg bw/d and an ARfD of 0.15 mg/kg bw. Studies investigating the direct post-harvest treatment in cereals were assessed in a previous EFSA reasoned opinion and it was concluded that for wheat, barley, millet, oats and sorghum an MRL of 5 mg/kg would be appropriate. Plant protection practices which need to be considered for the potential cross-contaminations were notified by Member States and basically can be grouped in two categories: direct treatment of stored cereals (leading to residues in silos after unloading or on handling circuits) and structural treatment of empty storage rooms or facilities which get into contact with food. The studies on cross-contamination of maize stored in silos which previously hosted treated cereals demonstrate that residues above the LOQ are likely to occur. The studies also revealed that the residues are not uniformly distributed in a batch when sub samples are taken during the process of 1 On request from the European Commission, Question No EFSA-Q , issued on 14 November Correspondence: pesticides.mrl@efsa.europa.eu 3 Regulation (EC) No 396/2005 of the Parliament and of the Council of 23 February OJ L 70, , p Council Directive 91/414/EEC of 15 July OJ L 230, , p Suggested citation: European Food Safety Authority; Risk assessment for pirimiphos-methyl residues resulting from cross-contamination. EFSA Journal 2011;9(11):2436. [48 pp.] doi: /j.efsa Available online: European Food Safety Authority, 2011

2 unloading a silo. The highest residue (0.533 mg/kg) accounted ca. three times the mean residue calculated for the whole batch. Handling of untreated crops on the handling circuit was also identified as a potential source of crosscontamination. Residues in maize moved on a handling circuit that was previously used to transport a treated batch of wheat were decreasing over time, but the study also showed that a low contamination rate cannot be avoided under practical conditions. The concentration in the highest contaminated sub sample was ca. 3 times higher compared with the mean residue concentration measured in the trial. Residue transfer from surface treated silos was observed in studies using dynamic sampling devices. The mean residue concentration in a trial using an application rate which is not representing the most critical situation compared with the authorised GAPs accounted for mg/kg. The highest residue concentration in a sub sample was ca. 2.7 times higher. Lacking more information it is not clear whether the expected residue concentration is proportionate to the dose rate applied. In studies which used static sampling methods no cross-contamination of stored crops was observed. In monitoring samples on maize residues above the current MRL of 0.05 mg/kg were observed in 16.61% of the samples. In 95% of the samples the residues were below 0.3 mg/kg. Cross-contamination was also observed in oilseeds (rapeseed and sunflower seeds). In a study based on monitoring samples taken in storage premises which also handled other cereals, the mean residues were in the range of mg/kg to 0.13 mg/kg. These results were supported by monitoring data provided by the European oilseed industry. The median residue concentrations ranged from 0.07 mg/kg to 0.09 mg/kg with maximum residues up to 1.9 mg/kg. Processing studies are available to estimate the transfer from unprocessed cereals to cereal based food. In addition to the studies that were already reported in previous EFSA opinions, in the framework of this request data were presented for maize which showed that in flaking grits a significant reduction of the residue concentration is expected. EFSA assessed whether the presence of residues in maize and oilseeds caused by cross-contamination would require a modification of the MRL proposals for food of animal origin. Taking into account the information regarding residue levels in these crops it is concluded that the overall dietary burden for livestock will not change and that MRLs as proposed previously (0.01 mg/kg) would be appropriate. The studies provide evidence that measurable pirimiphos-methyl residues are likely to be found in maize and oilseed stored or handled in premises that are using pirimiphos-methyl on cereals in accordance with the Good Agricultural Practices. An MRL of 0.5 mg/kg as proposed by the stakeholders is expected to cover most of the situations investigated in the studies provided. Since the studies investigated worst case scenarios, it is expected that in reality in most cases the residue levels in maize and oilseeds would be significantly lower than the proposed MRL. If cross-contamination is also likely to occur in other cereals (rye, buckwheat, rice) the MRL proposal for maize could be extrapolated to these cereals as well. Analytical methods for MRL enforcement are available. EFSA performed an acute consumer risk assessment for the crops under consideration. It is concluded that for the consumption of maize containing residues at the highest level observed in a study on cross-contamination (0.533 mg/kg), taking into account the reduction during processing, no consumer health risk is identified (0.1% of the ARfD). For the consumption of oilseeds containing residues at the level of the proposed MRL of 0.5 mg/kg, the exposure is also well below the toxicological reference value (0.1 to 1.9% of the ARfD). In the long-term exposure assessment, EFSA updated the calculations performed in the previous EFSA opinion on pirimiphos-methyl (EFSA Scientific Report (2009)294), assuming that maize, buckwheat, rice and rye contain residues at the mean residue level observed in the most appropriate EFSA Journal 2011;9(11):2436 2

3 study on cross-contamination (0.198 mg/kg). In addition, consumers are supposed to be exposed to residues in wheat, barley, millet, oats and sorghum which were treated directly according to the agricultural practices. EFSA also took into account the concentration/reduction of the residue during processing. The total chronic exposure to pirimiphos-methyl is calculated to range from 0 to 98.5% of the ADI. EFSA notes that the calculations are affected by uncertainties, but in total are based on conservative assumptions. Therefore it is concluded that no long-term consumer health risk is expected with regard to pirimiphos-methyl residues in food of plant and animal origin. EFSA however recommends to monitor the residue situation in cereals and oilseeds to verify the conservatism of the assumptions made in this risk assessment. In conclusion, the following MRLs are recommended: Code number (a) Commodity Existing EU MRL Proposed EU MRL Justification for the proposal Oilseeds 0.05* 0.5 The information provided demonstrated that the MRL proposal would be exceeded in less than 5%. No consumer intake concerns were identified (acute exposure <2% of the ARfD, max. chronic exposure 3.25% of the ADI) Barley 5 5 See EFSA Scientific Report (2009) Buckwheat Results for maize were extrapolated to buckwheat. If it is confirmed that for buckwheat the same storage facilities are used as for wheat, barley, millet, oats and sorghum and cross-contamination is likely, the setting of a MRL at the level of 0.5 mg/kg should be considered. The risk assessment did not indicate a consumer health risk (acute exposure <0.05% of ARfD, max. chronic exposure 1.4% of the ADI) Maize The information provided demonstrated that the MRL proposal would be exceeded only in exceptional cases. The data are not sufficient to derive a MRL proposal with statistical methods. No consumer intake concerns were identified (acute exposure 0.1% of the ARfD, max. chronic exposure 0.49% of the ADI) Millet 5 5 See EFSA Scientific Report (2009) Oats 5 5 See EFSA Scientific Report (2009) Rice Results for maize were extrapolated to EFSA Journal 2011;9(11):2436 3

4 Code number (a) Commodity Existing EU MRL Proposed EU MRL Justification for the proposal Rye rice and rye. If it is confirmed that for rice and rye the same storage facilities are used as for wheat, barley, millet, oats and sorghum and cross-contamination is likely, the setting of a MRL at the level of 0.5 mg/kg should be considered. The risk assessment did not indicate a consumer health risk (rye: acute exposure: 0.1% of the ARfD, max. chronic exposure 16.6% of the ADI); (rice: acute exposure: 0.2% of the ARfD, max. chronic exposure 3.9% of the ADI) Sorghum 5 5 See EFSA Scientific Report (2009) Wheat (including triticale) 5 5 The long-term exposure was recalculated taking into account consumption data for processed cereals. The maximum long-term intake accounted for 89.9% of the ADI Spices (seeds) 5 5 See EFSA Scientific Report (2009) Spices (fruit and berries) Other plant commodities Products of animal origin See Appendix B * (a): According to Annex I of Regulation (EC) No 396/2005. (*): Indicates that the MRL is set at the limit of analytical quantification. 0.05* See EFSA scientific report (2008) 164 EFSA proposes to set these MRLs on temporary basis and to review the MRLs within the next years, taking into account the results of monitoring data. These monitoring data should then also serve as basis to derive MRLs by means of statistical methods. It was noted that the application rates for the direct treatment of cereals notified to EFSA are higher than the application rate of residue trials used to derive the MRL of 5 mg/kg. EFSA therefore recommends modifying the authorisations at national level to be in line with the GAP supported by the manufacturer and to ensure that the MRL of 5 mg/kg is not exceeded. EFSA also recommends implementing Good Storage Practices which aim to reduce residues in food and feed. If possible, separate silos should be used for storing crops which were treated and crops for which a primiphos-methyl treatment is not allowed. In addition, cleaning processes should be introduced to avoid as far as possible infestation by insects. Particular attention should be paid to cleaning of transport circuits and silos that have been in contact with pirimiphos-methyl. Before loading a silo adhering dust containing pirimiphos-methyl residues resulting from surface treatment or residues from stored crops being treated with the pesticide should be removed. Considering the phenomenon of segregation, the last batch of cereals remaining in the silo at the end of the emptying process should be discarded and should not be used for food or feed purposes. EFSA Journal 2011;9(11):2436 4

5 KEY WORDS Pirimiphos-methyl, maize, oilseeds, MRL application, Regulation (EC) No 396/2005, consumer risk assessment EFSA Journal 2011;9(11):2436 5

6 TABLE OF CONTENTS Summary... 1 Table of contents... 6 Background... 7 Terms of reference... 7 The active substance and its use pattern... 8 Assessment Methods of analysis Methods for enforcement of residues in food of plant origin Methods for enforcement of residues in food of animal origin Mammalian toxicology Residues Nature and magnitude of residues in plant Primary crops Rotational crops Nature and magnitude of residues in livestock Consumer risk assessment Acute risk assessment Chronic risk assessment Long-term exposure assessment Conclusions and recommendations References Appendix A. Good Agricultural Practices (GAPs) Appendix B. Pesticide Residues Intake Model (PRIMo) Existing EU maximum residue limits (MRLs) Abbreviations EFSA Journal 2011;9(11):2436 6

7 BACKGROUND On 1 August 2011, in accordance with Article 43 of Regulation (EC) No 396/2005 4, the European Commission submitted to EFSA a request to provide a scientific opinion on the setting of temporary MRLs for pirimiphos-methyl based on information on cross contamination. The request was supported by studies/data provided by stakeholders (Intercéréales, Euromaisiers, Fediol and Association des Amidonniers et Féculiers) demonstrating that for maize and oilseeds a MRL of 0.5 mg/kg would be appropriate. EFSA accepted the mandate for this request on 13 September 2011 and included it in the EFSA Register of Questions with the reference number EFSA-Q and the following subject: Pirimiphos-methyl - Risk assessment for pirimiphos-methyl residues resulting from crosscontamination In the letter accepting the mandate EFSA highlighted that the supporting documentation was not sufficient to perform the requested assessment and the need to provide additional information/clarifications was identified. EFSA accepted the proposed deadline of 24 October 2011 for completing the assessment, if the required information was provided before 20 September Euromaisiers provided additional information by on 16 September On 23 September 2011, the European Commission forwarded information collected from Member States to EFSA; on 30 September 2011 a second batch of information was submitted by the European Commission to EFSA. EFSA then proceeded with the assessment of the application as required by Article 43 of the Regulation. TERMS OF REFERENCE In accordance with Article 43 of Regulation (EC) No 396/2005, EFSA shall, based on the information provided by the European Commission and the stakeholders (Intercéréales, Euromaisiers, Fediol and Association des Amidonniers et Féculiers), provide a reasoned opinion on the risks for European consumers associated with primiphos-methyl residues in cereals and oilseeds resulting from crosscontamination during storage. EFSA shall provide detailed information on the contribution of the individual crops to the overall dietary burden to identify scenarios which would be acceptable from a consumer safety point of view. 4 Regulation (EC) No 396/2005 of the Parliament and of the Council of 23 February OJ L 70, , p EFSA Journal 2011;9(11):2436 7

8 THE ACTIVE SUBSTANCE AND ITS USE PATTERN Pirimiphos-methyl is the ISO common name for O-2-diethylamino-6-methylpyrimidin-4-yl O,Odimethyl phosphorothioate (IUPAC). CH 3 N (CH 3 CH 2 ) 2 N N S P(OCH 3 ) 2 O Pirimiphos-methyl belongs to the class of organothiophosphates. It is a broad spectrum insecticide and acaricide with contact and respiratory action. The biochemical mode of action is via inhibition of the acetylcholinesterase. Pirimiphos-methyl was assessed in the second stage of the peer review process under Directive 91/414/EEC 5 with the United Kingdom as designated Rapporteur Member State (RMS). EFSA published its conclusion on the peer review in 2005 (EFSA, 2005). The evaluated representative uses as insecticide comprise spraying to control a broad spectrum of insects in wheat, barley and oats grains during storage and treatment of structural surfaces in empty grain stores at application rates up to 4 g pirimiphos-methyl per ton and 50 g per 100 m², respectively. The peer review process was finalised with the decision on inclusion of pirimiphos-methyl in Annex I of Directive 91/414/EEC which was published in Directive 2007/52 6 and entered into force on 1 October The Annex I inclusion was restricted to the use as insecticide for a post-harvest storage treatment only. Authorisations for other uses had to be revoked by 31 March According to the provisions of Regulation (EC) No 1107/2009 7, pirimiphos-methyl is considered as approved under the mentioned Regulation. EU MRLs for pirimiphos-methyl in products of plant and animal origin have been set for the first time in These MRLs have been amended in 2000 for fruit and vegetables and have been transferred to Annex II of Regulation (EC) No 396/2005 without further amendments. Additional MRLs for commodities that were not covered by the former European MRL legislation were established in Annex III B by Regulation (EC) No 839/ The existing EU MRLs for pirimiphos-methyl are summarised in Appendix B to this document. It is noted that Codex Alimentarius has established CXLs for cereal grains (7 mg/kg, resulting from a post-harvest use of 4 to 8 g a.s./t), wheat bran (15 mg/kg), and certain products of animal origin (all 0.01 mg/kg equivalent to the limit of quantification). In addition, MRLs for spices have been established by Codex (3 mg/kg for spices (seeds) and 0.5 mg/kg for spices (fruits)). Similar to the EU residue definition, also Codex Alimentarius considers the parent compound as the relevant residue definition. 5 Council Directive 91/414/EEC of 15 July OJ L 230, , p Commission Directive 2007/52/EC of 16 August 2007 amending Council Directive 91/414/EEC to include ethoprophos, pirmiphos-methyl and fipronil as active substances. OJ L 214, In Directive 2011/31/EU a restriction for the use of the active substance with regard to the operator exposure was added. 7 Regulation (EC) No 1107/2009 of the European Parliament and of the Council of 21 October OJ L 309, , p Commission Regulation (EC) No 839/2008 of 31 July OJ L 234, , p EFSA Journal 2011;9(11):2436 8

9 In 2008 EFSA was asked by the European Commission to assess the safety of the MRLs for pirimiphos-methyl. Taking into account the available information, EFSA concluded that for fruits and vegetables the MRLs could be lowered to the limit of quantification (0.05 mg/kg) because at EU level all uses for these crops had to be revoked and it was considered as not justified to maintain the MRLs, unless they are needed as import tolerances. EFSA also concluded that the MRLs for cereals at the level of 5 mg/kg could potentially pose a consumer health risk. However, since detailed information on the consumption of processed cereal based food was not available at that time, the risk assessment was performed with conservative assumptions overestimating the real exposure. EFSA concluded that for demonstrating that the existing MRLs are safe, information on the consumption of processed cereal based food would be required in order to refine the exposure assessment (EFSA, 2008). In 2009, EFSA was requested by the European Commission to perform a refined risk assessment for the existing MRLs of pirimiphos-methyl for cereals, taking into account the additional information provided by Member States regarding consumption of cereals in processed form (e.g. breakfast cereals, bread, beer etc). In addition, EFSA was asked to take into account that the manufacturer of pirimiphos-methyl will no longer support the post-harvest use of the product on buckwheat, maize, rye and rice. EFSA concluded that, taking into account the provided information on processed cereal based food, the expected residues on stored cereal grain (barley, millet, oats, sorghum and wheat) resulting from the use of pirimiphos-methyl at a dose rate of 4 g/t, the reduction/concentration of residues in processed cereals and the proposed revised MRLs for products of animal original, the long-term exposure might still exceed the ADI. However, EFSA had some doubts regarding the consumption of wheat bran by UK infants and UK toddlers, the most critical sub-populations in the chronic exposure assessment. EFSA concluded that if it is confirmed that bran is consumed in the form of breakfast cereals, the MRL of 5 mg/kg on wheat would be acceptable (EFSA, 2009). The responsible UK authorities confirmed by an sent on 13 September 2010 that bran consumption by UK infants and toddlers is almost exclusively in the form of processed cereals (98% and 93% of bran consumed as processed cereals by infants and toddlers respectively). Thus, the MRL of 5 mg/kg for wheat would not be expected to pose a consumer health risk. Before the decision was taken to amend the existing MRLs as proposed in the reasoned opinion of EFSA (EFSA, 2009), the European Commission was contacted by stakeholders, who were concerned about possible MRL exceedances caused by cross-contamination during storage and handling of maize and oilseeds if all the MRLs for crops not supported by the manufacturer are set at the LOQ level. The representatives of Intercéréales, Euromaisiers, Fediol and Association des Amidonniers et Féculiers therefore requested to set a MRL of 0.5 mg/kg for maize and oilseeds to avoid trade problems. GAPs which need to be considered for the potential cross-contaminations were notified by Czech Republic, Germany, Spain, Netherlands, Hungary, Ireland, Italy, Latvia, Lithuania, Portugal, Slovakia, Estonia, Greece, France and the United Kingdom and are listed in Appendix A. The GAP can be grouped in two categories: Structural treatment of empty storage rooms or facilities which get into contact with food at dose rates of 25 to 750 g/1000 m³ or 4 to 140 g/100 m² Direct treatment of stored cereals with application rates between 2 and 8.5 g/t (causing crosscontamination in crops stored in the silos which previously were used to store treated crops) It is noted that the application rates for the direct treatment of cereals notified to EFSA are higher than the application rate of residue trials used to derive the MRL of 5 mg/kg. EFSA therefore recommends modifying the authorisations at national level to be in line with the GAP supported by the manufacturer. EFSA Journal 2011;9(11):2436 9

10 ASSESSMENT EFSA bases its assessment on the submitted information provided by Intercéréales (Arvalis, 2011), Euromaisiers (2011a, b), Fediol (2011) and Association des Amidonniers et Féculiers (AAF) (2011), the GAPs notified by EU Member States, the Draft Assessment Report (DAR) prepared under Council Directive 91/414/EEC (United Kingdom, 2003), the Commission Review Report on pirimiphos-methyl (EC, 2005), the conclusion on the peer review of the pesticide risk assessment of the active substance pirimiphos-methyl (EFSA, 2005), the JMPR Evaluation report (FAO, 2003, 2006) as well as the conclusions from the previous EFSA opinions on pirimiphos-methyl (EFSA, 2007, 2008, 2009). The assessment is performed in accordance with the legal provisions of the Uniform Principles for the Evaluation of the Authorization of Plant Protection Products set out in Regulation (EU) No 546/ and the currently applicable guidance documents relevant for the consumer risk assessment of pesticide residues (EC, 1996, 1997a, 1997b, 1997c, 1997d, 1997e, 1997f, 1997g, 2000, 2004, 2010, 2011). 1. Methods of analysis 1.1. Methods for enforcement of residues in food of plant origin Analytical methods for the determination of pirimiphos-methyl residues in plant commodities were assessed in the DAR and in the conclusion on the peer review under Directive 91/414/EEC (United Kingdom, 2003; EFSA, 2005). It was demonstrated that the DFG S 19 multi-method can be used for enforcement practice: pirimiphos-methyl is extracted with acetone/hexane and is analysed by GC with either thermoionic or mass selective detector. The LOQ for monitoring purposes is 0.05 mg/kg. The method was validated on high water content, dry, acidic and oily matrices. Thus the peer review concluded that an adequate analytical method is available Methods for enforcement of residues in food of animal origin The analytical methods for the determination of pirimiphos-methyl residues in commodities of animal origin were evaluated in the DAR and in the conclusion on the peer review under Directive 91/414/EEC (United Kingdom, 2003; EFSA, 2005). It was concluded that the DFG S19 method based on GC-MS is applicable. The LOQ that can be achieved in routine monitoring is 0.01 mg/kg. 9 Commission Regulation (EU) No 546/2011 of 10 June 2011, OJ L 155, , p EFSA Journal 2011;9(11):

11 2. Mammalian toxicology The toxicological properties of pirimiphos-methyl were assessed in the peer review under Directive 91/414/EEC. The toxicological reference values derived are summarised in Table 2-1. Table 2-1. Overview of the toxicological reference values Pirimiphos-methyl Source Year Value (mg/kg bw/d) Study relied upon ADI EFSA year rat, 2 year dog, supported by human data Safety factor ARfD EFSA Rat, acute neurotoxicity It should be noted that JMPR reviewed the toxicological data of pirimiphos-methyl in 2006 (FAO, 2006) and derived higher ADI and ARfD values (0.03 mg/kg and 0.2 mg/kg). The ADI established by JMPR is based on a NOAEL of 0.25 mg/kg bw per day in a 28-day and a 58-day study in human volunteers, and a safety factor of 10. The ARfD is based on a NOAEL of 0.15 mg/kg bw (inhibition of acetylcholineesterase activity in rats from a single-dose study of neurotoxicity). The ARfD of JMPR and EFSA are based on the same study; the difference is due to different practice of the rounding of values. EFSA Journal 2011;9(11):

12 3. Residues 3.1. Nature and magnitude of residues in plant Primary crops Nature of residues The metabolism of pirimiphos-methyl in primary crops was evaluated by the RMS United Kingdom (UK, 2003) and reviewed by EFSA (EFSA, 2005) in the framework of the peer review under Directive 91/414/EEC. It was concluded that the metabolism of pirimiphos-methyl is sufficiently addressed and that for stored grain the relevant residue for monitoring and risk assessment is the parent compound only Magnitude of residues In support of the request to assess cross-contamination of maize and oilseeds stored in places where crops are stored that were treated in accordance with the authorised uses, EFSA received supporting information and studies from Euromaisiers (2011a, b), Fediol (2011), Intercéréales (Arvalis, 2011) and the European Starch Industry Association (AAF) (2011). In addition, studies assessed in the framework of the peer review were considered. The studies are discussed below in detail and at the end of the section, general conclusions are reported. a. Cereals - post-harvest treatment In the DAR results of four trials representing the representative GAP (post-harvest treatment by spraying of cereals with an EC formulation at a dose rate of 4 g/ton) were presented (United Kingdom, 2003). The trials in barley, wheat (two trials) and oats were designed as decline studies, measuring the pirimiphos-methyl residues immediately after the treatment (day 0) and in intervals of 7 days, 4 and 8 and ca. 25 weeks. During storage, no significant decline of the residue concentration was observed. In the trial with oats and barley residues were also measured in the rub-off and in other fractions and/or processed products. The detailed results are presented in Table 3-1. The highest residues for cereal grain within the decline studies (i.e. 1.1 mg/kg, 2.5 mg/kg, 2.8 mg/kg and 3.2 mg/kg) were selected to derive the MRL proposal and to calculate the input values for the consumer risk assessment (STMR 2.65 mg/kg, HR 3.2 mg/kg). The results were extrapolated to the whole group of cereals. It was concluded that for post-harvest treatment of cereals a MRL of 5 mg/kg would be required (EFSA, 2005, 2008, 2009). Because of potential consumer health risks, the manufacturer decided to support only the use on barley, millet, oats, sorghum and wheat. Thus, for the other cereals (buckwheat, maize, rye and rice), the MRL should be set at the limit of quantification (0.05 mg/kg). EFSA Journal 2011;9(11):

13 Table 3-1: Overview of the available residues trials data (source: DAR, United Kingdom, 2003) Crop/ Variety Barley Wheat Wheat Oats Country year Germany 1989 Germany 1989 Germany 1989 Germany 1989 Application rate per treatment kg as/ton No. of treatments Portion analysed Grain Grain Grain Grain Grain Rub-off Pearl Withholding period Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Cleaning residue Husk rub off Kernel pre dry Kernel post steam Rolled oats Residues Parent Residues Desethylpirimiphosmethyl (R036341) <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 < < <0.05 <0.05 <0.05 Comments Reference 10 EC formulation used. EC formulation used. EC formulation used. EC formulation used. Hayward G J, Harradine K J Hayward G J, Harradine K J Hayward G J, Harradine K J Hayward G J, Harradine K J (1) 10 The reference to the studies can be found in the DAR, Annex B.7 (UK, 2003). EFSA Journal 2011;9(11):

14 b. Cross-contamination of maize via treated wheat Study 1: Euromaisiers (2011a) submitted a study investigating the residue concentration in maize which was stored in silos that previously hosted wheat treated with a commercial pirimiphos-methyl containing formulation (EC formulation, 5% a.s.). The silo was of cylindrical shape and concrete walls with a capacity of 160 tons. The silo was filled with 120 tons of wheat, on which the pirimiphos-methyl formulation was applied by means of a spray equipment directly on the grain moving through the loading line of the silo to ensure a homogeneous distribution. The application rate on wheat was 7.5 g/t. The total amount of pirimiphos-methyl applied was therefore calculated to be 900 g/silo (120 t *7.5 g (application rate per ton)). After storing the treated batch in the silo for 7 days the wheat was removed and transferred into another silo. Samples of the wheat were analysed before and after storage. After unloading the wheat, the bottom of the silo was emptied and the conveyor elements were run for 15 minutes empty to avoid cross-contamination of the batch to be stored in the silo by direct contact of treated wheat. According to the sponsor of the study, this is the usual cleaning practice in storage premises. As a succeeding crop, a batch of 90 tons of untreated maize was stored in the silo. The pirimiphosmethyl residues were measured in the maize before loading and after 8 days of storage when the maize was unloaded from the silo. The samples of wheat and maize were taken in accordance with Regulation (EC) No 401/ describing the sampling methods for official control of mycotoxins in food. At regular time intervals primary samples of 100 to 200 g were taken (dynamic sampling). After combining of primary samples the analytical samples derived from the bulked primary samples were analysed with a gaschromatographic method. The results of the residue analysis are summarised in Table 3-2. Table 3-2: Description of the study on cross-contamination Description of handling Analytical results (pirimiphosmethyl in mg/kg) Wheat- residue analysis before loading Loading silo with 120 t of wheat Wheat- pesticide treatment with pirimiphos-methyl at a dose rate of 7.5 g/t, 7 d storage Wheat- residue analysis when unloading silos after 7 days Comment Bulked sub samples 1 to 5 bulked sub samples 6 to 10 Total amount of pirimiphos-methyl applied per 120 t: 900g Bulked sub samples 1 to 25 Bulked sub samples 26 to 50 Bulked sub samples 51 to 75 Mean: 3.39 mg/kg Median: 3.64 mg/kg Total amount of pirimiphos-methyl (recovered) per 120t: 407 g 11 Regulation (EC) No 401/2006 of 23 February 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs, OJ L70, , p EFSA Journal 2011;9(11):

15 Description of handling Cleaning of the silo Analytical results (pirimiphosmethyl in mg/kg) Maize - residue analysis before loading Maize- loading of the silo with 90 t of maize Maize- residue analysis after storage, unloading after 8 days Comment Bulked sub samples 1 to 5 bulked sub samples 6 to 10 Bulked sub samples 1 to 25 Bulked sub samples 26 to 50 Bulked sub samples 51 to 75 Mean: mg/kg Median: mg/kg Total amount of pirimiphos-methyl (recovered): g Before loading, the untreated wheat contained low residue concentrations (mean residue mg/kg), probably caused by contaminations during transport or the conveyor systems in the storage premises. The mean residue concentration of wheat treated with 7.5 g a.s./t after storage was 3.39 mg/kg, thus higher than the residues observed in the residue trials reported in Table 3-1 (application rate 4 g a.s./t). The mean residue concentration in untreated maize before storage in the silos was mg/kg. When unloading the silo, the highest residues were observed in the last bulked sub samples containing residues at a level of mg/kg whereas the first and the second sub samples contained significantly lower residue concentrations (0.035 and mg/kg, respectively). The sponsor of the study explained this observation with the dynamic of unloading of vertical silos where the grains in contact with the walls of the silo which would be most likely in contact with pesticide residues remaining on the surface would be unloaded at the end of the unloading process, whereas the grain stored in the central zone of the silo would be the first to be removed. EFSA agrees that this is a plausible interpretation of the results. Thus, if the last sub-lot of the untreated stored grain which was contaminated during storage with pirimiphos-methyl is placed on the market without mixing or homogenisation with other lots, the residue concentration would exceed the LOQ of 0.05 mg/kg. Assuming that the three sub samples of wheat and maize are representative for the aliquot of the treated stored products, EFSA calculated the recovery of the applied active substance in wheat and the carry over in maize. The 120 t of wheat treated with 900 g pirimiphos-methyl at the loading contained 407 g pirimiphos-methyl at the unloading. The difference between the applied pirimiphos-methyl (900 g) and the pirimiphos-methyl recovered in the stored wheat (407 g) was therefore 493 g primiphosmethyl (55% of the applied dose). The non-recovered active substance may have been lost by degradation/metabolism, evaporation, or part of it may have remained in the silo adhering on the silo walls or on the silo bottom as rub-off. It is assumed that by cleaning the empty silo part of the primiphos-methyl residues was removed from the bottom of the silo. In total 3.62 % of the nonrecovered active substance was recovered in the succeeding maize (0.21% in the first third, 0.17% in the second third and 3.24% in the last third of the total maize load). Thus, 3.24% of the residues are accumulating in the last third of the stored maize. These considerations would support the proposal of Euromaisiers that the MRL for the succeeding crops stored in silos which hosted treated crops should be set at the level corresponding to 10% of the MRL for the treated crop. EFSA Journal 2011;9(11):

16 EFSA also notes that the samples were analysed by an accredited laboratory using gas chromatography with nitrogen/phosphorus detector system (NPD). The documentation of the analytical part of the trial was not very detailed. In EU proficiency tests it was found that pirimiphosmethyl is normally an analyte for which the participating laboratories are able to provide good results. However, the EU Reference Laboratory for cereals pointed out that before extracting cereals, it is important to add water, otherwise the recovery would not be satisfactory. Despite the deficiencies of the documentation EFSA accepted the results provided by the stakeholder as indicative results, but notes that they contribute to the overall uncertainty of the assessment. Study 2: A similar study was performed by Arvalis (2011). 25 tons of wheat treated at a dose rate of 3.4 mg/kg were stored in a metal-walled cylindrical silo with a capacity of 40 m³. After 7 days of storage the wheat was removed from the silo which was then refilled with untreated maize (20 t). After 7 days of storage in the silo, samples of the maize were taken by means of static sampling from the cell periphery and a dust sample was collected from the cell walls. In addition, when unloading the silo, a series of 10 dynamic samples were collected, representing 2 tons of maize each. In the figure below the results of the sub samples described are presented. Figure 3-1: Overview of the results of the residue analysis of the storage trial (storage of maize after unloading silos filled with treated wheat) (source: Arvalis, 2011) The residues in the treated wheat samples at the end of the storage in the silo accounted for 0.53 mg/kg, corresponding to 15.6% of the applied dose rate. The residue concentrations of each of the primary samples of maize, corresponding to two tons of grain ranged from 0.01 to 0.08 mg/kg. The overall mean of these 10 sub samples (sub lot) was mg/kg. The residue concentration measured with the static sampling, the dust sample and the concentration at the bottom was 0.06 mg/kg, 22.4 mg/kg and 0.46 mg/kg, respectively. EFSA Journal 2011;9(11):

17 Also this study demonstrated that residues above the LOQ are expected in maize stored in a silo which previously hosted wheat which was treated at a dose rate of ca. 4 g/t. Using dynamic sampling devices, an inhomogeneous distribution of the residues in the individual sub lots was observed. The highest residue in a sub sample was 2.7 times higher compared with the mean residue concentration of the total batch. In contrast to the study provided by Euromaisiers, in this study the highest residue concentration was observed in the first sub sample. Thus, the dynamic of unloading a silo may be different, depending on the geometry of the silo. Further studies would be necessary to understand this phenomenon. It is also recognized that the residue concentration in the last twenty kilograms of the silo was significantly higher than the residues in the individual sub lots. c. Cross-contamination of maize via handling circuit Arvalis (2011) also investigated the possible contamination of maize which is handled on conveyor belts/handling circuit which was used for transport of treated wheat. After having treated a batch of wheat (25 t) with pirimiphos-methyl at a dose rate of 4 g/t (flow treatment at the level of the grain elevator head) a batch of 10 t of untreated maize was transferred from a silo using the same handling circuit. During this transfer operation, the batch of maize was dynamically sampled to study the overall residue content and the possible decline throughout the procedure. The concentration of pirimiphos-methyl was on average 0.04 mg/kg. The concentration in the 10 sub samples ranged from 0.13 mg/kg in the first sample to 0.02 mg/kg in the last three sub samples. The results demonstrate that the contamination is decreasing over time, but according to the study author a complete decontamination of the handling circuit could not be achieved. d. Contamination of maize after treatment of empty silos (cell-wall treatment) A study provided by Arvalis (2011) investigated the residue concentration in maize stored in metalwalled silos with a capacity of 40 m³ which was treated with a dose rate of 0.2 g/m². Seven days after the treatment the silo was filled with 20 t of untreated maize (pirimiphos-methyl residues below LOQ). Maize samples were taken 7 days after filling by means of static sampling devices taking samples along the walls and by means of dynamic sampling during emptying the silo. In addition, dust samples of the silo walls and the remaining 20 kg at the cell bottom were analysed separately. In the assay for dynamic sampling the sub samples were pooled to obtain 10 primary samples representative for 2 tons each. In the figure below the results are reported for the different sub samples analysed. The residue content in each of the 10 primary samples, corresponding to two tons of grains, range between the LOQ of 0.01 mg/kg and 0.12 mg/kg. The overall mean of the 10 samples was mg/kg. The highest residue concentration was measured in the last of the 10 sub samples (0.12 mg/kg). The residue measured in the remaining 20 kg left at the bottom of the silo after unloading the maize, the static sample and the dust sample were mg/kg, 0.27 mg/kg and 417 mg/kg, respectively. Considering that the authorised application rate for empty storage rooms ranges between 4 and 140 g/100 m² the trial does not reflect the worst case situation. For identifying the worst-case residue situation, not only the dose rate but also the surface (metal, concrete) and the shape of the silo needs to be taken into account (ratio of surface to capacity in terms of stored grain). The trial should therefore be considered as an indicative study. Before deriving general conclusions on the residue levels resulting from contamination via treatment of empty silos further studies would be necessary. However, the study demonstrates that an inhomogeneous residue distribution may occur when unloading cereals stored in silos which underwent a surface treatment. In the trial the highest residue observed in the last fraction was ca. 3 times higher than the overall mean (dynamic sampling). EFSA Journal 2011;9(11):

18 Figure 3-2: Overview of the results of the residue analysis of the storage trial (source: Arvalis, 2011) e. Contamination of wheat grain after treatment of empty storage rooms (structural treatment) In the framework of the peer review, the applicant submitted 4 residue trials to assess the residues expected in wheat grain stored in storage rooms which were treated with a EC formulation at a dose rate of 100 g a.s./100 m 2 using a directed spray. The treated stores were of different material (3 steel silos, 1 concrete storage room). The stores were filled between 3 and 16 days after treatment. Samples of untreated wheat grain were taken immediately after the store was filled, at 7, 14 and 30 days after filling, then 2, 3 and 4 months after filling. Samples were analysed for residues of pirimiphos-methyl. Residue levels of pirimiphos-methyl in the wheat grain were all below the limit of quantification (<0.05 mg/kg) which was seen as a proof that the treatment of the empty storage rooms does not result in measurable residues above the LOQ of 0.05 mg/kg in grain (United Kingdom, 2003, EFSA, 2005). Taking into account the findings of the trial submitted by Euromaisiers and Arvalis and described in the previous section, it becomes questionable whether the study design used in the 4 residue trials reported in the DAR is appropriate to cover the phenomenon of residue transfer from surface treatment to stored grains. EFSA would propose to develop more guidance on how to perform residue trials for empty storage rooms to cover the possibility that cross-contamination may occur. It would be more appropriate to take samples by means of dynamic sampling device rather than static samples to reflect better the fact that residues are accumulating while unloading silos. f. Cross-contamination data of maize (monitoring data) The European Starch Industry Association (AAF) submitted a list of 644 monitoring results on maize % of the results exceeded the proposed MRL of 0.05 mg/kg (EFSA; 2009), 2.8% exceeded a level of 0.5 mg/kg. The highest residue measured was 2.4 mg/kg. The 95 th percentile was calculated to be 0.3 mg/kg. Lacking a detailed description of the samples (year of the sampling, origin of the sample, confirmation that the sample was not subject to a pirimiphos-treatment, analytical method) EFSA is of EFSA Journal 2011;9(11):

19 the opinion, that the decision on the MRL modification cannot be based on this data set. The data however, can be considered as supporting information g. Cross-contamination in oil-seeds (data on rapeseed and sunflower seeds) Considering that oilseeds can potentially be handled and stored in a similar way as untreated maize as described in the studies provided by Euromaisiers and Intercéréales (see previous section), it is envisaged that cross-contamination could also occur for oilseeds. The EU Oil and Proteinmeal Industry represented by Fediol has submitted a study investigating the reasons for residues in untreated oilseeds (Fediol, 2011). A survey was performed, covering 58 samples of sunflower seeds taken in 2007 and 2009 and 54 samples of rapeseed taken in 2008 and 2009 when the crops were delivered to the storage facilities. The samples were analysed for pirimiphos-methyl residues before and after storage in the silos, using a method based on Soxhlet extraction with hexane, clean-up of the extract with acetonitrile and freezing, purification with solid phase extraction and determination with GC-ECD and GC-NPD. The limit of quantification that could be achieved was reported as 0.01 mg/kg. In case residues above the LOQ were detected, the source of contamination was traced. In table 3-3 the results are summarised. Table 3-3: Overview of the monitoring results in oilseeds after storage/handling in premises that also store cereals (results for pirimiphos-methyl expressed in mg/kg) Sampling year Number of samples Mean Median Standard deviation P90 Maximum % samples LOQ Sunflower < % 4% < % 7% Rapeseed < % 32% < % 9% % samples > MRL By means of questionnaires which had to be completed by the store managers, the reasons for the presence of pirimiphos-methyl residues in sunflower seeds and rapeseed should be identified. The analysis of the questionnaires revealed the following situations as being critical with view of contamination: Treatment of cereals on delivery, just before delivering oilseeds Storage of treated cereals in the same bin just before storage of oilseeds Treatment of empty bin and of handling equipment before receiving oilseeds The highest risk of contamination appeared when cereals were systematically treated on delivery, just before delivering oilseeds, using the same conveyor circuits. In the case of rapeseed which is harvested during the same period as cereals, the highest risk was identified if oilseeds are delivered to the storage facilities at the same time as cereals which underwent treatment at receipt. The Fediol study also reported monitoring results generated by the oilseed industry. French monitoring data and data from other EU Member States which were collected between 2005 and 2010 are summarised in Table 3-4 and 3-5. EFSA Journal 2011;9(11):

20 These data demonstrate that a MRL of 0.05 mg/kg is exceeded in 6-15% of cases. If the MRL is raised to the proposed level of 0.5 mg/kg, more than 95% of the samples would be within the permitted residue concentration. Table 3-4: French monitoring data on rapeseed and sunflower seed ( ) Pirimiphos-methyl Rapeseed Sunflower seed Number % Number % Number of samples Samples < MRL (0.05 mg/kg) % % Samples MRL (0.05 mg/kg) % 56 15% Minimum < LOQ < LOQ Maximum Statistics on data MRL Median th percentile th percentile Table 3-5: European monitoring data (France excluded) on rapeseed and sunflower seed ( ) Pirimiphos-methyl Rapeseed Sunflower seed Number % Number % Number of samples Samples < MRL (0.05 mg/kg) % % Samples MRL (0.05 mg/kg) 18 7% 13 6% Minimum < LOQ < LOQ Maximum Statistics on data MRL Median th percentile th percentile Summary of the studies presented in a to g: The studies investigating the direct post-harvest treatment in cereals ( a) were used to derive a MRL proposal and risk assessment input values for for wheat, barley, millet, oats and sorghum (MRL 5 mg/kg, STMR 2.65 mg/kg, HR 3.2 mg/kg). Since the manufacturer does not support the use in buckwheat, maize, rye and rice the MRL proposal should not be extrapolated to these crops. The possibility of segregation of residues in directly treated crops was not assessed. EFSA is of the opinion that this aspect should be further investigated by performing a trial where dynamic sampling methods are applied which allow to monitor the residue concentrations in subsamples taken during the process of unloading silos. The studies on cross-contamination of maize stored in silos which previously hosted treated cereals ( b) demonstrate that residues above the LOQ are likely to occur (study 1 (2N treatment of EFSA Journal 2011;9(11):