5.1 Evaluation of available information
5.2 Re-evaluation of compounds
5.3 Definition of residues
5.4 Estimation of maximum residue levels
5.5 Expression of Maximum Residue Limits
5.6 Recommendations for Maximum Residue Limits
The Joint Meeting carries out a scientific evaluation and takes into account all available information. Better evaluations result from an understanding of the processes of residue behaviour rather than from only an empirical treatment of data. In addition, the available information varies to a great extent. Therefore the JMPR does not follow rigid rules in its evaluations but considers the submitted information on a case-by-case basis. The basic principles outlined below are followed as far as practical and possible.
As part of the evaluation process the members of the FAO Panel prepare the monographs, including all relevant information concerning the pesticide, and the appraisal summarising the findings, conclusions and recommendations, and giving full explanation and reasoning for them. The monographs and appraisals are prepared in a uniform format, described in the FAO Panel Manual attached as Appendix X, to facilitate access to the required information by the reader. The monographs and appraisals are published by FAO in the series Pesticide Residues in Food - Evaluations Part I. Residues. In addition, a short summary of information evaluated and the recommendations for each compound are included in the Report of the JMPR.
Some years ago the JMPR recognised the need to explain more fully the basis for its recommendations. During the past years, information on GAP and data on supervised residue trials are summarised in detail in the monograph and the reasoning behind the evaluation is explained. The increased volume of the evaluations is largely due to the inclusion of more detailed explanations.
5.1.1 Chemical and physical properties of a pesticide
5.1.2 Metabolism and degradation of pesticides after application on plants, animals and soil
5.1.3 Information received on Good Agricultural Practices
5.1.4 Residue data
5.1.5 Processing studies
5.1.6 Results of national monitoring programmes
The purpose of evaluating data submitted to the JMPR on physical and chemical properties of pure active ingredient is mainly to recognise the influence those properties have on the behaviour of the pesticide during and after its application on crops or animals.
The volatility of the compound and its stability in water and after radiation from ultraviolet light may considerably affect the disappearance after application.
The solubility of the pesticide is especially of great interest, as the ability of the compound to penetrate plant and animal tissues is dependent on its solubility in water and organic materials.
The physical property chosen by the JMPR to represent solubility in fat is the octanol-water partition coefficient, usually reported as log POW.
It should be noted that there are errors in estimates of log POW, with differences of one unit for the same compound being reported. Different approaches to the development of these data often give different results. Interpretations must recognise these differences.
The variable composition of some residues, e.g. where the residue is defined as a mixture of parent and metabolites, presents a problem since the fat-solubilities of the metabolites may be different from those of the parent compound. In this case, information on the log POW of each individual metabolite should be considered if available. The relative concentrations within the mixture are also subject to change and, as a result, the tendency of the mixture to partition into fat will also change.
A further factor which may influence designation as a fat-soluble residue is the nature of the available residue data from supervised trials. Data are frequently reported only on a fat basis or only on a whole commodity basis.
The JMPR recommended that the octanol-water partition coefficient should be the prime indicator of fat-solubility, supplemented by inferences which may be drawn from the distribution of residues between muscle and fat tissues, when the residue consists of a single compound. In cases where the residue is defined as a mixture of the parent compound and metabolites, information on the log POW of the individual compounds should be considered if available.
The JMPR recognised that many compounds which are neither clearly fat-soluble nor clearly water-soluble required special consideration.
In general, when log POW exceeds 4 the compound would be designated fat-soluble and when log POW is less than 3 it would not be so designated. Pesticides with intermediate log POW would be considered on a case-by-case basis.
One major route of disappearance of pesticides after application on plants, animals or soil is chemical degradation and metabolism at rates which are dependent on the chemistry of the compounds and circumstances, such as temperature, humidity, light, surface of the crops, pH of crop liquid, composition of soils and other factors. Metabolism studies provide fundamental information on the fate of the compound and its metabolites, provide a qualitative semi-quantitative picture of the composition of the residues and suggest probable residue behaviour. The results of metabolism studies suggest target tissues in animals; other important guides to the site and level of residues are whether the compound is absorbed by the leaves or roots of crops, whether it is mobile in the plant, and its persistence and mobility in soil.
Data on metabolism are used in evaluating both the toxicological and residue profiles of pesticides. The FAO Panel examines the metabolism in experimental animals and compares it with both that in food-producing farm animals and in plant species on which the pesticide is used. This is required to decide upon the relevance of the toxicological studies to humans, and to define the residues in plants and farm animal products. The ADI estimate, based on toxicological studies in experimental mammalian animals, is valid for foodstuffs only if the metabolite pattern is qualitatively and semi-quantitatively similar. If there are plant or farm animal metabolites (where there are significant residues in feed items) which have not been identified as mammalian metabolites in experimental animals, the ADI does not encompass other metabolites and separate studies dosing with these metabolites may be necessary for assessment of their toxicological properties.
The information on the composition of the terminal residue is used to assess the suitability of the residue analytical methodology for the development of residue data from supervised trials and to decide on the definition of residues.
An essential element to enable the JMPR to estimate maximum residue levels of pesticides is information on Good Agricultural Practices. Normally the FAO Panel gives preference to GAP information provided by national authorities when submitted information is in conflict. The FAO Panel uses the information on national GAPs to identify the likely scenarios which may lead to the highest residues in food or feed, and relates these uses to the conditions prevailing in the execution of supervised trials. Therefore, information on national GAP from those countries in which the supervised trials have been carried out, or from countries in close proximity with similar climatic conditions and agricultural practice is of the utmost importance.
With regard to the required presentation of adequate information on Good Agricultural Practice in the use of a pesticide in a country, the FAO Panel recognised that several countries may apply different pesticide use authorization systems. Some use a rigorous formal product-based registration scheme, while others use less formal authorization approaches. The "authorised safe use" or "approved uses" from the latter countries may still be included in the GAP Table provided that the country involved supplies the information on nationally approved uses or authorised safe use. The terms "approved" and "authorised" are understood as GAP information from countries which do not have a full registration scheme, but where there is a form of authorisation of use. This distinction recognises the different terminologies and approaches to GAP authorisations at the national levels and does not imply that one national system is preferred over another.
Registered and approved use of a pesticide may vary considerably from country to country and the use patterns are often very different, especially in regions with great differences in climate. Growing conditions and, naturally, types of crops may also cause differences in the use pattern. According to the definition of Good Agricultural Practice, a pesticide should be applied in such a way as to leave a residue which is the smallest amount practicable. Residue levels exceeding the smallest amount practicable, due to unnecessarily high application rates ("overdose") or unnecessarily short pre-harvest intervals (PHIs), are contrary to the concept of GAP.
184.108.40.206 Results of supervised trials
220.127.116.11 Results of selective surveys
Estimation of maximum residue levels is mainly based on reliable residue data from supervised trials carried out in such a way that treatments in the trials are equivalent to the uses according to Good Agricultural Practice. The importance of reliable data has already been emphasised in the requirements for information and data from trials.
The principles followed in evaluating supervised trial data are described in detail in section 5.4.
Selective surveys may provide supplementary information which will assist in extrapolations (see also "Data from alternative experiments" in Section 3.1.5). As the number of supervised trials is always small in comparison with actual use situations, the submission and evaluation of alternative data are of importance.
"Processed food" in connection with establishing Maximum Residue Limits for pesticides means products resulting from the application of physical, chemical or biological processes to a "primary food commodity" whereas primary food commodities treated with ionising radiation, washed or submitted to similar treatments are not considered to be processed food in this context.
Originally the main interest for processed foods was the types of processed foods which were important items in international trade, such as milled cereal grains and other grain products, oil from oilseed, juice and dried fruit. There was a marked need for MRLs for those commodities, and a number of MRLs have been established over the years. During the last five to ten years interest has increased considerably in obtaining better information about the residue level in other types of processed food, e.g. primary food commodities which are peeled, cooked or baked. Some of those commodities are usually not moving in international trade, but information on the residue levels is essential to obtain knowledge about the real intake of pesticides from food commodities for which MRLs have been established. As in the case of edible/non-edible residue distributions, this may have the consequence that higher MRLs are acceptable when it is demonstrated that residues found in the whole commodity are dissipated or rendered harmless in food processing. Experience has shown that residue levels usually decrease during processing, such as peeling, cooking and juicing. In other cases the residue level may increase during processing as it often does in the case of oil from oilseeds and olives. However, in some cases the active ingredient can be transformed during processing into metabolites that are more toxic than the parent compound.
The JMPR is aware that there is a considerable trade in manufactured foods based, for example, on fruits, vegetables, cereals and meat. However, the variety of forms under which the products are offered makes it impossible to recommend MRLs for all possible processed foods. For this reason the JMPR has specified that, in the case of processed foods for which no MRLs have been recommended, the maximum residue permitted in the processed food should not be greater than the maximum residue permitted in the equivalent weight of the raw agricultural commodity. The JMPR frequently estimates maximum residue levels for important processed foods and feeds in international trade when residues concentrate in these products at levels higher than in the raw agricultural commodities from which they are derived (e.g. oil, bran, peel, etc.). Even when the estimates are not recommended for use as maximum residue limits or when residues do not concentrate in the processed product, the JMPR will continue to record in its monographs the effect of processing on the level and fate of residues in food. This has been found to be critical for better estimates of dietary intake of pesticides.
The results of national monitoring programmes are considered as supporting information for confirmation of practical applicability of estimated maximum residue levels. They are also useful for estimating dietary intake at national level. Data from national monitoring programmes are essential for estimating ERLs.
5.2.1 Re-evaluation of additional information
5.2.2 Re-evaluation of compounds being in the CCPR periodic review programme
Usually new information on GAP and related data from trials do not cause difficulties, if the data received are of the same type and in agreement with data from earlier evaluations. However, information about new developments in the area of metabolism of the compound may be more problematic. Such information may require that the original residue definition be changed, which means that evaluation of old and new data together may be very complicated. In a similar way, problems may arise when a residue definition originally included two pesticides of which one of the compounds is also a metabolite of the other, and for toxicological or other reasons the decision is taken that the pesticides must subsequently be determined separately. In such a case old residue data are often inapplicable.
Improvements of analytical procedures may also cause difficulties. If the limit of determination is lowered, the old residue data below the original limit of determination are difficult to interpret and may be inapplicable and unavailable for later evaluations. In this context, as for the changes in the metabolism of the compounds, the whole set of data on the compound has to be taken into consideration and decisions have to be taken by the JMPR on a case-by-case basis.
In most of such cases, however, all of the information required for the scientific re-evaluation is not available to the JMPR. Therefore, such complex problems are best and most efficiently handled during the periodic review of the compound for which all relevant original reports are supposed to be re-submitted and can be taken into consideration.
The periodic review programme requires different actions from those for the re-evaluation of additional information, called hereunder normal situation (i.e. other than in the periodic review programme); consequently, those compounds to be evaluated within the periodic review programme must be clearly identified in advance.
As discussed in detail in Chapter 3, data submitters should supply all relevant valid information at the time of the periodic review irrespective of whether it has been supplied previously.
The JMPR evaluates all relevant information on periodic review compounds in terms of identity, metabolism and environmental fate (animal and plant metabolism, environmental fate in soil and in water/sediment systems), methods of residue analysis (analytical methods, stability of residues in stored analytical samples, residue definition), current use patterns (registered and officially authorised uses), supervised residue trials, fate of residues in storage and processing, residues in food in commerce or at consumption and national maximum residue limits, as in the case of a new compound. However, the conclusions and recommendations are somewhat different in periodic and normal reviews.
Comparison of the data evaluation of a periodic review compound with normal re-evaluation (re-evaluation of some particular information made available to the JMPR) clarifies the major differences.
New MRLs. If no MRL exists for the individual commodity or for the relevant commodity group, there is little difference in the treatment of information supplied normally or under the periodic review programme.
Existing MRLs. For an individual commodity in the normal situation, if new data are supplied where an MRL already exists the data are evaluated and the MRL may or may not require revision.
In the periodic review situation, where adequate information is supplied on an individual commodity, the MRL is either revised or confirmed to be relevant to modem GAP.
In the normal situation, when information on a single commodity included in a group commodity MRL is received evaluation would either show that the group MRL could remain or that an individual MRL and a group (with specified exceptions) MRL could be recommended.
In a periodic review when information on only a single commodity included in a group commodity MRL is received it may be necessary to withdraw the group MRL and estimate a single-commodity MRL.
GAP information. Under normal circumstances if no new GAP information is supplied the MRL would remain. New GAP information may allow previously recorded residue data to be reinterpreted to permit estimation of a new maximum residue level.
In the normal situation where new residue data are to be evaluated, judgement is required on a case-by-case basis to decide whether previously recorded GAP is still valid. GAP information recorded many years ago for some compounds may still be acceptable.
Under the periodic review programme the absence of GAP and residue information becomes significant. For example, if no GAP information is supplied for a particular commodity the JMPR reviewer can assume that there is no GAP for that commodity. Only GAP supplied for the purposes of re-evaluation is considered valid. If no GAP information is supplied, withdrawal of the MRL will be recommended.
Supporting Studies. Critical supporting studies, which were previously referred to as residue supporting information (metabolism, animal transfer, processing, analytical methods and storage stability of analytical samples), are evaluated to assist with the interpretation of data from supervised residue trials, influence the residue definition, validate residue and other trials and provide further information on residues in food as consumed. The FAO Panel may not recommend MRLs for new or periodic review chemicals in the absence of critical supporting studies if their omission is not adequately justified.
5.3.1 Principles followed in defining residues for MRLs
For the purposes of an MRL or STMR, a pesticide residue is defined as the combination of the pesticide and/or its metabolites, derivatives and related compounds to which the MRL or STMR apply.
Definitions of residues which have been applied in the past are sometimes arbitrary. For this reason, and because of the various purposes for which they are used, definitions of residues established by national governments often do not agree.
The basic requirements for the definition of residues are that it should:
- be most suitable for monitoring compliance with GAP, and
- include compounds of toxicological interest for dietary intake estimations and risk assessment.
The two requirements are sometimes not compatible and, as a compromise, various definitions of residues are possible. For some compounds it may be necessary to establish separate residue definitions for enforcement and dietary intake purposes. The residue definition for dietary intake purposes should include metabolites and degradation products of toxicological concern irrespective of their source, whereas the residue definition for compliance with MRLs needs to be a simple residue definition (i.e. indicator molecule) suitable for practical routine monitoring and enforcement of the MRL at a reasonable cost.
Although metabolites, degradation products and impurities are included in the definition of pesticide residues, this does not necessarily mean that metabolites or degradation products should always be included in the residue definition for enforcement (MRLs) purposes or for estimation of dietary intake (STMR). Inclusion of metabolites in the residue definition depends on a number of factors, and the decision on whether metabolites should be included is very complex and decisions have to be made on a case-by-case basis.
The metabolites have generally been identified and quantified in metabolism experiments with methods based on the use of labelled compounds. In other cases the methods used for supervised trials are complicated or require sophisticated instrumentation and, therefore, they are unsuitable for regulatory analytical work. Furthermore, some countries may experience extreme difficulty to procure metabolites for use as standards in the analytical work. Therefore, inclusion of, in particular, polar metabolites in the residue definition for monitoring compliance with GAP is not practical.
It should be stressed that in choosing the appropriate analytes and the analytical method for the testing of the residue trials samples, the manufacturer/sponsor must consider the needs of both risk assessment and compliance. In practice this will mean generating the data in such a way as to give the flexibility to establish two separate residue definitions where appropriate. In cases where it is likely that a multi-component residue definition will be required for risk assessment purposes, the manufacturer/sponsor should, in testing field trial samples, either:
(i) analyse separately for the individual components of the residue, where analytical methodology allows, rather than carrying out a total residue analysis
(ii) if total residue methodology is used to produce data for risk assessment, and the suitable "indicator molecule" can be analysed with a multi-reside procedure, a second series of analyses of the field trial samples should be carried out for the indicator molecule (e.g. parent compound).
This approach allows the risk assessment to be carried out on the toxicologically significant residue components whilst ensuring that data are available to allow a different simple residue definition to be established, where appropriate, for compliance with the MRLs.
In cases where the manufacturer/sponsor has submitted residue trials data in which total analytical methodology has been used and it is not possible to identify a suitable simple residue definition for practical routine monitoring and enforcement of the MRL at reasonable cost, the FAO Panel may be unable to estimate MRLs for the compound.
The following examples further illustrate the complexity of the situation.
Several pesticides are metabolised to a compound, which itself is used as a pesticide (example: benomyl ® carbendazim), and in addition, in some such cases, the toxicology is substantially different for the pesticide and the metabolite (example: dimethoate ® omethoate). Whenever possible, the parent pesticide and its metabolite(s) used as pesticides should be subject to separate MRLs. Analysing food commodities in trade for the metabolite may provide no information on which compound was used.
Where it is not possible to set separate MRLs because the parent pesticide is degraded rapidly or analytical method is not available for measuring and distinguishing the parent compounds (examples: ethylene-bis-dithiocarbamates, benomyl ® carbendazim, thiophanate-methyl ® carbendazim), the MRLs applying to the pesticides concerned can only be determined in terms of the metabolite(s) or conversion products.
Another problem occurs when the metabolite from a pesticide may also originate from sources other than use of the pesticide. In this case, a residue of the metabolite present in a sample is of no use as proof of illegal use of the pesticide, and the metabolite should not be included in the residue definition for MRL (example; cyromazine ® melamine, also prometryne ® melamine).
The JMPR considers the following factors when proposing a residue definition:
· The composition of the residues found in animal and plant metabolism studies.
· The toxicological properties of metabolites and degradation products (for risk assessment).
· The nature of the residues determined in supervised residue trials.
· The fat-solubility.
· The practicality of regulatory analytical methods.
· Whether metabolites or analytes common to other pesticides are formed.
· Whether a metabolite of one pesticide is registered for use as another pesticide
· The definitions of residues already established by national governments and long-established and customarily accepted definitions.
The definition of residues for enforcement purposes should be as practical as possible and preferably based on a single residue component - the parent compound, a metabolite or a derivative produced in an analytical procedure - as an indicator of the total significant residue. The selected residue component should reflect the application condition (dosage rate, pre-harvest interval) of the pesticide, and it should be determined with a multi-residue procedure whenever possible. Monitoring for additional residue components only adds to the cost of analyses.
The advantage of this approach is appreciable as overall costs can be reduced and many more samples may be analysed by the regulatory laboratories. In addition, more laboratories can participate in regulatory monitoring of residues, since a relatively simple and rapid analytical procedure may not require the expensive equipment and time necessary for an extensive determination of all components of a residue. Nevertheless, the expression of residues with a single compound does not reduce the data requirement as full information on the composition of total residue, and on the relative ratio of residue components, is needed before it can be determined whether a single compound can be used. Furthermore, such full information is often needed for risk assessment purposes.
As far as possible the same definition of the residue should apply to all commodities, although there are exceptions. For example, if the major residue in animal commodities is a specific animal metabolite a definition which includes that metabolite is needed for regulatory monitoring. However, the animal metabolite is not required in the residue definition for crop commodities if it is not found in the crops. Separate definitions would then be proposed for commodities of plant and animal origin.
Example: residue definition of thiabendazole: thiabendazole or, in the case of animal products, the sum of thiabendazole and 5-hydroxythiabendazole.
It is generally preferable to express a residue in terms of the parent compound. Even if the residue consists mainly of a metabolite, the residue should be expressed in terms of the parent pesticide after molecular weight adjustment. Some examples are given to illustrate the practical application of the principle:
If the parent compound can exist as an acid or its salts, the residue is preferably expressed as the free acid.
If metabolites are known to be present in significant amounts, but the analytical method measures, the total residue as a single compound the residue is expressed as the parent compound. The metabolites included in the residue should be listed.
Example: the residue definition of fenthion includes fenthion, its oxygen analogue and their sulphoxides and sulphones which are measured after oxidation as fenthion sulphone; the reside is expressed as fenthion.
There have been exceptions:
Example: the current residue definition of amitraz is: the sum of amitraz and N-(2,4-dimethylphenyl)-N'-methylformamidine expressed as N-(2,4-dimethylphenyl)-N'-methylformamidine.
Ideally it should be possible to measure the residue as defined, with a limit of determination (LOD) adequate for proposed MRLs, with a high degree of specificity by a multi-residue regulatory analytical method. Although circumstances may warrant exceptions, the definition of a residue should not normally depend on a particular method of analysis, which means that the definition should not contain the words "determined as". However, the only way to produce a practical definition for residues of the dithiocarbamates is to describe the residue as ".... determined and expressed as....".
Example: residue definition of thiram for compliance with MRLs: total dithiocarbamates, determined as CS2 evolved during acid digestion and expressed as mg CS2/kg sample.
Where the residue is defined as the sum of the parent compound and metabolites expressed as the parent, the concentrations of the metabolites should be adjusted according to their molecular weight before being added to produce the total residue. The words "expressed as" in the residue definition signify adjustment for molecular weight.
Example: residue definition of methiocarb: sum of methiocarb, its sulphoxide and its sulphone, expressed as methiocarb.
No allowance was made for molecular weights in the definitions of residues of some older compounds. Because such definitions are widely accepted the need for change should be carefully considered. The best time for the reconsideration of an existing residue definition is during a periodic review.
Examples: (no recalculation for molecular weight)
residue definition of DDT: sum of p,p'-DDT, p,p'-DDE and p,p' TDE
residue definition of heptachlor: sum of heptachlor and heptachlor epoxide
Metabolites arising from different sources should generally be excluded from definitions of residues for enforcement purposes unless the definition is a combined one covering the various sources. For example, p-nitrophenol arises from both parathion and parathion methyl. It is often a major component of aged residues but is not included in the definitions of the residues.
Where a metabolite of one pesticide is registered for use as a second pesticide, separate MRLs would normally be established if the analytes of the two compounds were different. Preferably no compound, metabolite or analyte should appear in more than one residue definition.
Example. Triadimenol is a registered pesticide and a metabolite of triadimefon. The MRLs for triadimefon are for triadimefon only. The MRLs for triadimenol are for triadimenol only, but cover triadimenol residues arising from the use of either triadimefon or triadimenol. (Before 1992 the definition of the triadimefon residue included triadimenol, and an analyst detecting triadimenol in a sample could not be certain whether to compare the residue with an MRL for triadimefon or triadimenol.)
There are cases of pesticides, however, where the chemical instability of the parent compound and/or the limitation of analytical methodology do not allow the application of the above principle. In such cases the residue definition has to be based on the stable common moiety.
Examples: residue definition of benomyl,: thiophanate-methyl and carbendazim
Note: According to the current practice of JMPR, the residue definition would be1: the MRLs cover carbendazim residues resulting from direct use, or occurring as a metabolic product of thiophonate-methyl and benomyl and intact (i.e. unmetabolised) thiophonate-methyl and benomyl expressed as carbendazim."1 Presently the following explanation is given in the Codex table of recommendation: MRLs cover carbendazim residues occurring as a metabolic product of benomyl or thiophanate-methyl, or from the direct use of carbendazim. Source of data (the MRL is based on the use of the specified pesticide): B, benomyl; C, carbendazim: Th, thiophanate-methyl. Note that this definition does not incorporate explicitly the parent residues of benomyl and thiophanate-methyl being present in/on the commodity.
A major part of the residue of some pesticides is bound or conjugated, with the free residue disappearing very quickly. The bound or conjugated residue is then a better indicator for monitoring compliance with GAP. If the residue is defined as bound or conjugated there must be a clear instruction for the regulatory analyst as to how to measure it. The instruction could be to extract samples with a particular solvent under specified conditions, or perhaps to begin with a hydrolysis step, etc.
Example: residue definition of bendiocarb:
commodities of plant origin: unconjugated bendiocarb;
commodities of animal origin: sum of conjugated and unconjugated bendiocarb, 2,2 dimethyl-1,3 -benzodioxol-4-ol and N-hydroxymethyl-bendiocarb, expressed as bendiocarb.
Fat-solubility is a property of the residue; it is primarily assessed from the octanol-water partition coefficient and the partition of the residue between muscle and fat observed in metabolism and animal transfer studies. Section 5.1.1 provides guidelines for deciding whether a pesticide is fat-soluble. Sampling protocols for animal commodities depend on whether a residue is fat-soluble or not.
The JMPR has for many years included the qualification 'fat-soluble' in the definition of the residues of fat-soluble pesticides, using the expression:
"Definition of the residue: [pesticide] (fat-soluble)"
As different definitions of residues may be needed for estimating dietary intake and for assessing compliance with MRLs, however, the 1996 JMPR recommended that 'fat-soluble' should no longer be included in the definition of the residue. In order to avoid confusion while conveying the information that a residue is fat-soluble, the JMPR agreed that the definition of a residue should include only the chemical species of concern and a separate sentence should indicate that the residue is fat-soluble.
(a) Definition of the residue for compliance with MRLs and for estimation of dietary intake: diazinon (The residue is fat-soluble.)
(b) Definition of the residue for compliance with MRLs and for estimation of dietary intake: sum of aldicarb, its sulfoxide and its sulfone, expressed as aldicarb
Separate definitions of residues for enforcement (the estimation of maximum residue levels and compliance with MRLs) and for dietary intake purposes should be established when the requirements are in conflict.
Example: definition of the residue of thiram for compliance with MRLs: total dithiocarbamates, determined as CS2 evolved during acid digestion and expressed as mg CS2/kg
Definition of the residue for estimation of dietary intake: thiram
As the above examples indicate, the definition of residues has not always been consistent. Therefore all residue definitions are re-examined during the periodic review of the compounds.
From 1995 the FAO Panel includes in the Monographs a special section on residue definition with appropriate explanation. It will now be easier for the reader to find an explanation for the definition of the residue. It is located in the monographs under the section on Methods of Residue Analysis.
5.4.1 Specific considerations in estimating maximum residue levels
5.4.2 Estimation of group maximum residue levels
5.4.3 Estimation of extraneous residue levels (ERL)
The JMPR examines the possibility of estimating maximum residue levels based on the submitted information and data, and subsequently proposes Maximum Residue Limits in commodities for pesticides used according to Good Agricultural Practice.
Maximum residue levels are estimated for residues in or on the portion of the commodity to which Codex MRLs apply. For dietary intake purposes the residue levels are estimated on the edible portion of the commodity (see Chapter 6).
In addition to primary and some processed food commodities, when the available information permits, JMPR also recommends MRLs for animal feeds and food processing byproducts which can be used as animal feed. The latter are commodities of trade and therefore require Codex MRLs if pesticide uses result in detectable residues in the feeds. Residues in feeds may also lead to detectable residues in animal tissues, milk and eggs, necessitating MRLs for these commodities. In addition, some by-products of food commodities (e.g. apple pomace, grape pomace) and food commodities themselves (e.g. cereal grains) may be used as feedstuffs for food-producing animals.
In estimating maximum residue levels, the FAO Panel takes into account all relevant information especially the residues arising from supervised trials and the relation of trial conditions to the established GAP. The procedure for estimating and recommending MRLs for Codex purposes may be somewhat different from that applicable at national level. The Codex MRLs should cover residues deriving from authorised uses world-wide, which reflect greatly differing agricultural practices and environmental conditions.
An awareness of the expected variability of residues is necessary. If the data truly reflect the range of conditions, application methods, seasons and cultural practices likely to be encountered commercially, then considerable variation in the resulting residue levels is expected. Where copious data are available, consideration of the spread and variability of the residues helps to avoid misleading interpretations of small differences in estimates of the maximum level. Where only limited data are available, the interpretation of fine differences is not valid. It is not a criticism to say that the data are widely spread and variable. If results have been obtained at a number of places over some years they are likely to be a better approximation to commercial practice and will be widely spread. In addition to the variability of residues within a confined area which can be considered uniform regarding climate, agricultural practices, pest situation and use recommendations, there may be an even greater variation of residues among areas of widely differing conditions (e.g. countries being in temperate, Mediterranean and tropical zones). The differences in use conditions can be so large that they result in different residue populations.
Frequently the situation is complex, even when copious data and information are available. There are alternative interpretations, and judgement is required to arrive at an estimate which is realistic, practical and consistent.
Although supervised residue trials are conducted according to GAP prevailing at the time, GAP is often subsequently modified by changing the rate of application, type of formulation, method of application, number of applications and PHI. Judgement is then required in order to determine whether the trial conditions are still close enough to GAP to be relevant.
The nominal rate of application in a trial would normally be considered still consistent with GAP when it is within ± 25-30% of the GAP rate, which includes the probable variation in commercial practice. When little or no residue is present, data from higher applications may be important.
In many situations different formulations would cause no more variation than other factors, and data derived with different formulations would be considered comparable. Experience from trials demonstrates that EC, WP and SC formulations lead to similar residues. On the other hand, controlled-release formulations would be expected to lead to more persistent residues and would not be comparable to others.
The method of application can be quite influential on residue levels. For example, directed application is not comparable to cover spray, and aerial application may not be comparable to ground application.
For a non-persistent pesticide the number of applications is unlikely to influence residue levels: three, four or five applications could be considered equivalent from a residue point of view. For a persistent pesticide the number of applications would be expected to influence residue levels.
The nature of the crop should also be considered. For example, summer squash may be harvested only a few days after flowering; residues of a non-systemic pesticide applied before flowering would be expected to be low, and the number of applications should have little influence on the residue level.
The pre-harvest interval usually, but not always, influences residue levels. In most cases the PHI in supervised trials should agree with that specified by GAP. When residue levels change rapidly it is not possible to extrapolate from one PHI to another; when they change slowly extrapolation is possible. For example, when residues of non-systemic pesticides in crops from trials are not detectable in samples taken before the prescribed pre-harvest interval, there is no reason to expect residues from samples taken according to the prescribed pre-harvest interval, as the residues in those samples also will be non-detectable. Furthermore, in the case of a relatively flat residue decline and typical horticultural or agricultural crop, data from a PHI range of 12-15 days could be considered equivalent. Normally, the range of PHIs which gives ± 25-30% of the residue level obtained at registered PHI is considered comparable to GAP.. Residue decline data are useful in deciding the relative decline of the residue. Trials that do not conform to GAP can be used to provide information on persistence. For example, there are cases where residues are not significantly influenced by the pre-harvest interval, so results obtained at intervals other than the specified PHI may be useful.
Some pesticide uses, such as seed treatments and pre-emergence herbicide treatments, usually lead to non-detectable residues in the final harvested crop; but when many results are provided residues may be detected in occasional samples. While residues resulting from use according to GAP are most likely to be undetectable, the occasional detectable residues should not be ignored when a maximum residue level is estimated. Phorate on potatoes and residues arising from the pre-planting application of glyphosate are two examples.
Greater certainty that the climatic conditions are properly reflected in the supervised trials is afforded when the trials are carried out in a country with established GAP. Trials conducted in other countries with similar climatic conditions may be acceptable on a case-by-case basis. An assessment of those conditions is difficult, and a critical evaluation is needed as only some difference in conditions, such as temperature, may be of great importance for the persistence of many pesticides and consequently for the residue level.
The trials should be carried out with the same crops as those specified in the national GAPs. The proper description of the crops used in the supervised trials is important for deciding if crops referred to in GAP are in accordance with those for which trials have been carried out. The use of the Codex Classification for describing the harvested commodity facilitates the evaluation.
The CCPR establishes MRLs on commodities as they move in trade to enable the control of compliance with and enforcement of GAP. Consequently, the maximum residue levels are estimated on a whole commodity basis (see Appendix VI) as far as practical.
In addition to residues in/on the whole commodity, the JMPR is also interested in residues in the edible part of the crop. Residues of systemic pesticides may be expected to be present in all parts of the crop, while residues of non-systemic pesticides are not always present or may be present in minor quantities in the edible part of a crop. For each pesticide, information on the distribution between edible and non-edible parts should be available to the JMPR from supervised trials or special experiments. This information is also essential for deciding on the toxicological acceptability of residue levels on or in commodities consumed in substantial amount. For example, MRLs are established for whole bananas including the inedible peel. Some MRLs may appear to be unacceptably high based on residues on the whole commodity. However, information that residues on edible portion are practically non-detectable often alleviates that concern. Another example is oranges where usually most residues are present in the peel, especially for non-systemic pesticides.
When estimating maximum residue levels, the FAO Panel examines all residue data arising from supervised trials supporting/reflecting the reported GAPs. First the uniformity of residue population reflecting GAPs is considered. When there is a large gap in residue values, either from the coefficient of variation of residues in composite samples or other appropriate statistical methods, the presence of different populations may be suspected. The residue data and trial conditions need more stringent analysis before the maximum residue level can be estimated. The maximum residue level estimation is based either on all approved uses or on only those which lead to the highest residues. In each case, a sufficient number of trials (residue data) reflecting a particular use should be available to enable the estimation of maximum residue level. It follows that, in case of suspected multiple residue populations, a few data indicating the high population may not be sufficient to estimate a maximum residue level reflecting that population (and use pattern), and the FAO Panel may estimate a maximum residue level reflecting only those uses for which sufficient residue data are available. On the other hand, it is not possible to reconsider and reduce a previous estimate based on a few new trial data, unless the GAP on which the old recommendation was based has been changed in the meantime or the original trials on which the MRL were estimated are now considered inadequate.
The maximum residue levels are usually estimated taking into account the spread of residues within the selected set of trials. Although the JMPR may use statistical tools, it has not routinely adopted standard statistical methods. There are several reasons for this. One is the lack of internationally accepted standards for such an approach. A more operational reason is the nature and quality of the data usually available. (The number of trials is generally too low to apply statistical methods to render a given probability and confidence level to the estimated limits.) The form in which the residue data are provided, for example, often does not lend itself to statistical analysis, while treatments which involve more subjective judgements, such as taking into account data from similar crops or the consideration of historical data, are difficult to fit into a statistical approach. Whenever the data base is suitable, the FAO Panel may also take into account statistical considerations (e.g. in the 1996 Evaluations, aldicarb residues in potato and DDT residues in meat).
Fruits and vegetables
When estimating maximum residue levels in fruits and vegetables all of the previously described general considerations apply. Applications on fruit and vegetables may take place at any stage of the developments of the plants and in the soil before and after sowing, and the residue levels are highly dependent on the treatment.
The importance of PHI on residue levels has been described in section 5.4.
The whole fruit residue may sometimes be derived from data obtained separately from peel and pulp if the weights of peel and pulp are available.
Grains and seeds
The Maximum Residue Limit for seeds or grains applies to the whole commodity. It is important for the JMPR to be able to distinguish between the forms in which the commodities are present and to describe the raw and processed commodities according to the Codex Commodity Classification, as some grains and seeds are still in the husks and others are without husk. Sometimes residues are reported in polished rice. The residue levels are usually considerably different for those sorts of commodities. The estimation of the maximum residue levels should be based on residues in commodities which may move in international trade.
When grains and seeds are milled, the commodities belong to the processed commodities.
The results of animal transfer studies and residues in animal feed and processing by-products of food serve as a primary source of information for estimating maximum residue levels. In addition, animal metabolism studies may give useful information.
Uptake of pesticides by animals, leading to residues in animal products, can occur following either direct application of the pesticide to the animal or its housing, or ingestion of feed carrying residues of pesticides.
Animal feeds with residues of pesticides may derive from:
· crops produced mainly for animal feed (e.g. pasture, straw, forage),
· crops produced mainly for human food which are fed to animals (e.g. cereal grains),
· waste from crops grown primarily for human food (e.g. skins, pulp, stems, stubble, trash),
· animal feeds that have not themselves been treated, but in which unintentional residues occur, as for example originating from soil treated in an earlier season.
When the animals are fed, the potential for dilution of residues is considerable. Not all producers of the primary crop are likely to have used the same pesticide simultaneously, and the pesticides used are not always used at their highest permitted use rates nor at the nearest time to harvest. Consequently, the calculations based on maximum intake and MRL values give an unrealistic overestimate of residue levels in animal products.
Currently there are differences in national approaches to estimating residue levels in animal products, and the subject is also under reassessment by the JMPR. The recommended method of assessment will be published in future JMPR Reports. In the case of pesticides which are excreted rapidly and where there is no build-up of residues in the animal tissues, probably the best approach for making estimates of maximum residue levels is to use the median value of residues in feed items and the upper limit of the range in which a commodity might reasonably be fed. For other pesticides the estimated maximum residue levels should be taken into account. Examples of maximum proportions of agricultural commodities in animal feed in different regions are given in Appendix IX. In selecting the median residues, it is considered prudent to consider those uses and trials which represent the maximum potential for residues only and not all of the data. The principles of estimating supervised trial median (STMR) values are described in detail in section 6.2. In the case of processed feed items, it is preferable to use residue data from commercial processing of field-treated crops containing residues at about the expected maximum residues.
Certain pesticides are highly soluble in fat and are almost completely transferred into fatty tissues, being distributed throughout the body fat. Assessment of fat-solubility is based on the octanol water partition coefficient (see section 5.1.1) and the results of metabolism and animal transfer studies. If the pesticide is considered fat soluble this should always be indicated in association with the definition of residues.
For pesticides which are not fat-soluble, the MRLs are estimated for the whole product as it moves in trade.
For fat soluble pesticides the maximum residue levels are estimated based on residues in fatty tissues. For those commodities where the adhering fat is insufficient to provide a suitable sample, the whole meat commodity (without bone) is analysed and the maximum residue level is estimated on the whole commodity basis.
The maximum residue levels are estimated on a whole commodity basis.
Milk and milk products
For milk it is known that the fat content varies widely between different breeds of dairy cattle. In addition, there are a large number of milk products with varying fat content and it would be difficult to propose separate MRLs for each of them. It was therefore originally decided to estimate MRLs for fat-soluble compounds for milk and milk products on a fat basis, i.e. the fat content of the sample should be determined and the residue levels expressed as if wholly contained in the extracted fat.
Currently the JMPR follows the CCPR convention of expressing the MRL for fat-soluble compounds in milk on a calculated whole product basis, assuming all milks to contain 4% fat. (The residue is calculated for the whole product based on the residue measured in the fat.) For compounds which are not fat-soluble, the analytical portion for enforcement purposes is whole milk and MRLs are expressed on a whole milk basis.
Details of expressing residues in milk and milk products are given in section 5.5.
For eggs, the maximum residue level is estimated on the basis of the whole commodity after removal of the shell.
The establishment of commodity group MRLs as opposed to MRLs for individual commodities has long been considered an acceptable procedure at both the national and international levels. The use of the approach is a recognition that economics may not justify residue trials on all of the individual crops in a group. It also follows naturally in national registration systems where the registered use is on a crop group such as citrus. In principle the approach recognises that adequate data for the major crops of a group may be sufficient to estimate maximum residue levels for the whole group.
Some pesticides behave differently in different circumstances. Consequently, it is not possible to define precisely those commodities on which trials will always provide data which can lead to a group MRL. If the "highest residue" situation can be identified, however, the relevant data can be extrapolated to other crops with confidence. An acceptable example is extrapolation of residue data from gherkins to cucumber; however, the converse is not possible due to the higher residues that can be expected in gherkins as a consequence of their slower growth rate and the difference in weight/surface ratio. Extrapolation requires a detailed knowledge of agricultural practices and growth patterns. In view of the large differences in texture, shape, growth habits, rate of growth and seasonal cultivation and the significant role played by the surface/weight ratio, the JMPR has emphasised that decisions to extrapolate should be made on a case-by-case basis when adequate relevant information is available.
The JMPR has proposed group MRLs since 1966. Historically the JMPR has always approached the issue of group or individual MRLs on a case-by-case basis and that approach is unchanged. The main reasons for this are the many factors which can affect a decision on whether or not to propose a group MRL and the lack of international consensus on criteria. These considerations have prevented the JMPR from developing specific guidance for estimating group MRLs which might be applied at the international level in all situations.
Although such specific guidance is not yet available, some general guidance has been developed and recorded by the JMPR over the years. The development of the Codex Classification provided an important aid for estimating group maximum residue levels.
The 1996 JMPR took into account its previous considerations of the issue as well as the collective experience of its members, and summarised a number of general principles and observations which reflect the current views of the JMPR on estimating group MRLs.
(a) The JMPR continues to rely on the Codex Classification of Foods and Feeds as the primary basis for recommending MRLs for individual or grouped commodities.
(b) Generally the JMPR now refrains from estimating maximum residue levels for large Codex 'classes' of foods or feeds such as fruits, vegetables, grasses, nuts and seeds, herbs and spices, or mammalian products. Residue data and approved uses are usually more likely to refer to smaller Codex 'groups' such as pome fruits, citrus fruits, root and tuber vegetables, pulses, cereal grains, cucurbit fruiting vegetables, milks, meat of cattle, pigs and sheep, etc. As well as being more likely to be justified by the available data on residues and information on GAP, this is judged to be more in line with national approaches and affords more accurate estimates of dietary intake.
(c) When adequate residue data are available for only a few primary commodities in a food group, separate MRLs should generally be recommended for each commodity on which the data are considered adequate.
(d) In some cases the JMPR may, in the absence of sufficient data for one commodity, use data from a similar crop for which GAP is similar to support estimates of maximum residue levels (e.g. pears and apples or broccoli and cauliflower).
(e) If other considerations permit, data on residues in all or most of the major commodities with the potential for high residues within a group may allow estimates of maximum residue levels to be extrapolated to minor crops in the group. An example of a situation in which other considerations do not permit this is that in which the variability of the residue levels is too great, even though data on the major crops within the group are available. A group limit cannot then be established.
(f) When residue levels in a number of commodities in a group vary widely, separate recommendations should be made for each commodity. A limit for a group 'except one or more commodities' which are known to deviate from the norm may be justified (e.g. citrus fruits, except mandarins); in such cases separate MRLs should be estimated for the exceptional commodities.
(g) In order for a group limit to be proposed, not only must residue levels in the major commodities in the group not be too different, but the physical nature and other characteristics of the crops that might influence residue levels, as well as cultural practices and GAP for the individual commodities, must also be taken into account.
(h) Residue data for a crop growing quickly in summer cannot be extrapolated to the same or related crops growing slowly under less favourable conditions (e.g. from summer to winter squash).
(i) In establishing group MRLs, detailed knowledge of the metabolism or mechanism of disappearance of a pesticide in one or more crops must be taken into account.
(j) Group MRLs recommended by the JMPR that generally appear to be acceptable include those for cereal grains (based on data for maize, wheat, barley, oats and rice without specific data on millet, rye or sorghum), pome fruit, based on apples and pears, extrapolated to quinces; stone fruits, poultry meat, milks, meat from mammals other than marine mammals, and oilseed.
(k) A group MRL is generally preferred in the case of citrus fruits, but care must be used in estimating a maximum level for the group because of the large variations in fruit size and in the ratio of peel to pulp in view of the propensity for residues of many pesticides to concentrate in the peel. Data on major members of the group are especially important.
Historically, as of the time of availability of these guidelines, many more Codex limits have been established for citrus fruits as a group (45 pesticides) than for individual citrus fruits (19 pesticides): lemons (2 pesticides); lemons and limes (1); mandarins (4), sweet and sour oranges (8), sweet oranges (1); shaddocks or pomelos (1); and grapefruit (2).
(l) All else being equal, data on a crop picked when immature may sometimes be extrapolated to a closely related species with a lower surface area/weight ratio at the time of the pesticide application which grows quickly to maturity, resulting in a rapid decrease in the ratio of residue to crop weight (dilution by crop growth). Thus, estimated maximum residue levels can be extrapolated from gherkins to cucumbers, but not vice versa.
(m) Individual MRLs can be extrapolated more readily to groups when there is no expectation that terminal residues will occur and when this is supported by studies of metabolism. Examples are early treatments, seed treatments and herbicide treatments of orchard crops.
While the JMPR generally adheres to these principles on a case-by-case basis, it recognises certain difficulties or limitations in the acceptance of group limits at the international level. A primary weakness is the lack of formal criteria or an agreed mechanism to determine the members of a group for which data are needed before a group MRL can be established. One approach that is sometimes used effectively at the national level is to identify commodities of a group (often botanical) that represent both major crops within the group and those most likely to contain the highest residues. The factors used to determine whether a crop is a major or representative member of the group include its dietary significance as a food or feed.
The premise of this approach is that if data are available for representative crops, and if GAP and cultural practices among the individual members are similar, the residue levels will not vary widely and a maximum residue level can be estimated that will suffice for other members of the group for which no data are available. As noted earlier, this approach constitutes the use of common sense and is more or less dictated by the economics of data generation and evaluation.
While the JMPR recognises real advantages in this approach, there is unfortunately no consensus at the international level on the selection of representative commodities for estimating maximum residue levels for groups. Similarly, while the JMPR bases its recommendations on the Codex Classification of Foods and Feeds, this classification has not been fully adopted at the national level in most countries.
Until there is more international agreement in this area, the JMPR will continue to make judgements on a case-by-case basis, using the general policy summarised above or as it may be subsequently amended.
Chemicals for which ERLs are most likely to be needed are those which have been widely used as pesticides, which are persistent in the environment for a relatively long period after uses have been discontinued and which are expected to occur in foods or feeds at levels of sufficient concern to warrant monitoring.
Predictions of persistence in the environment (and the potential for uptake by food or feed crops) can often be based on a combination of data sources normally available for chemicals previously approved as pesticides. These may include information on their chemical and physical properties, metabolism studies, data on supervised field trials, data on environmental fate, rotational crop data, the known persistence of similar chemicals, and especially from monitoring data.
In order to make reasonable estimates to cover international trade, the JMPR needs all relevant and geographically representative monitoring data (including nil results). Better ERL estimates, taking into account trade concerns, can be made when more extensive data are available. However, typically data are available from only three or four (usually developed) countries at the most. By the nature of national monitoring, data are usually received primarily on those commodities in which residues have been found at the national level and which have the potential to create trade difficulties.
On receipt of the data, to the extent possible the JMPR attempts to take into account a number of factors in estimating an ERL. These include the amount of data, the relative importance of the commodity in international trade, the potential for trade difficulties or accounts thereof, the frequency of positive results, a knowledge of the propensity of a particular crop to take up residues (e.g. the uptake of DDT by carrots), historical monitoring data (e.g. previous monographs), and the level and frequency of residues in similar crops, especially those in the same crop group. In some cases the estimate has turned out to be the highest level reported, especially if a relatively good database is available and the spread of results is reasonably narrow.
In recent years there have been cases where the ERL was estimated below the highest residue found, especially if the higher values occur infrequently. For example, the 1993 JMPR recommended an ERL of 0.2 mg/kg for DDT in carrots, although 2 of 4 imported samples reported from one country were 0.4 and 0.5 mg/kg. The JMPR took into account that only 2 of over 800 imported samples exceeded 0.2 mg/kg. This limit covers > 99% of the residue population with 99% confidence. A similar approach was taken for DDT in the fat of meat by the 1996 JMPR. This approach also recognises that residues gradually decline and that monitoring data are often somewhat out-of-date by the time they reach the JMPR. It is more likely to be used when the higher residues occur infrequently.
In the context of ERLs, the JMPR does not consider extreme values to be outliers in a statistical sense, because high residue levels are usually not true statistical outliers but values on one tail of a large distribution. The challenge is to decide when it is reasonable to discard those values in order to reflect the expected gradual decline in the levels of chemicals that are typically subject to ERL estimates, while not creating unnecessary barriers to trade.
Generally, the JMPR considers that the databases needed for estimating ERLs should be significantly larger than those required for the estimation of MRLs, because the ERL data are very far from normal distribution. For example, samples from 598 randomly selected animals are needed to ensure that the estimated ERLs cover 99.5% of a population, allowing a 0.5% violation rate with 95% confidence (Codex Alimentarius, Vol. II, 2nd Ed., p. 372). On the other hand, if a country had only 100 random samples analysed with a 10% violation rate this is quite significant, despite the small number of samples. As ERL databases are derived from the random monitoring of different populations, the JMPR does not normally consider a 'world' population of data, but gives independent consideration to different populations, e.g. of different geographical regions or of different animals, before deciding which data populations might be combined. Therefore, all relevant monitoring data should be submitted regardless of the number of samples analysed.
The JMPR compares data distribution in terms of the likely percentages of violations that might occur if a given ERL is proposed. Since there is no internationally agreed level of acceptable violation rate, the JMPR estimates ERLs based on the available data. However, violation rates of 0.5 to 1% or greater are generally considered unacceptable.
It is to be expected that there will be a gradual reduction and/or elimination of residues of the chemicals for which ERLs have been proposed. The rate will depend on a number of factors, including the nature of the chemical, the crop, the location and environmental conditions.
The JMPR intends to continue lowering the ERLs as the results of monitoring permit. The aim of lowering ERLs is to discourage unauthorised uses which higher levels might accommodate, and it is hoped to encourage those who might not otherwise submit data to do so.
Because residues gradually decrease, the JMPR recommends reassessment of ERLs about every 5 years. Eventually, the data may indicate that there is no longer a need to monitor the chemical. This view would be based on the conclusion that there is no longer a potential for significant disruption of trade and that the incidence or level of residues is no longer a significant health concern.
Although the JMPR does not use targeted monitoring data for estimating ERLs, it agrees that follow-up studies are important when high residues are found in random monitoring to give a clearer view of the significance of the high levels. If properly conducted, such studies may indicate whether or not the higher residues resulted from intentional unauthorised uses and may allow the identification of areas in which production should be limited or where residue reduction strategies should be implemented.
5.5.1 Expression of Maximum Residue Limits at or about the limit of determination
The estimated maximum residue levels and recommended residue limits are expressed in mg residue (as defined)/kg commodity. The portion of commodity to which Codex MRLs apply is given in Codex Alimentarius Vol. 2 (extracted in Appendix VI).
The residues are expressed on fresh-weight basis or as they enter international trade (as received by the laboratory) in most commodities, with the exception of animal feeds. Because of the great variation of their moisture content, MRLs for animal feeds are recommended on a dry-weight basis, and this implies that the commodity is analysed for pesticide residues as received, that the moisture content of the sample is determined, preferably by a standard method recommended for use on that commodity, and that the residue content is then calculated as if it were wholly contained in the dry matter.
If it is not clear in animal feed residue data submissions whether residues are expressed on a dry weight basis, or the moisture content of the feed is not reported, the data may not be suitable for estimating maximum residue levels (See also 18.104.22.168).
For animal products there are certain special cases which need to be mentioned:
For meat and fat soluble pesticides (see section 5.1.1) the residue limits are proposed on a fat basis (the residue content in fatty tissue or fat cut off meat) which is indicated in brackets (fat) after the residue value. For those commodities where the adhering fat is insufficient to provide a suitable sample, the whole meat commodity (without bone) is analysed and the MRL applies to the whole commodity.
For all other pesticides the MRLs apply to the whole product as it moves in trade.
The MRLs and EMRLs for fat-soluble pesticide residues in milk and milk products are expressed on a calculated whole product basis assuming all milks to contain 4% fat. Milk products with a fat content of 2% or more are expressed on a fat basis. The MRL would be 25 times the MRL for milk. The MRL for milk products with a fat content lower than 2% are considered to be half the value for milk and are expressed on a whole product basis.
Milk MRLs for fat-soluble pesticides are indicated by the letter "F"
When the recommended limit is based on veterinary use (direct animal treatment) this fact is indicated by the letter "V" after the limit.
Examples for recommended MRLs (mg/kg) for diazinon -
MO 0098 Kidney of cattle, pigs and sheep:
MM 0097 Meat of cattle, pigs and sheep:
2 (fat) V
ML 0106 Milks
0.02 F V
MRLs reflecting special uses or conditions are also distinguished by letters after the limit: Currently the following cases are distinguished by the letters indicated below:
The MRL is based on extraneous residues
The MRL accommodates post-harvest treatment of the commodity
The MRL accommodates post-harvest treatment of the primary commodity
The MRL/EMRL is temporary, irrespective of the status of the ADI, until required information has been provided and evaluated.
For the numerical expression of the limits, the following scale is used usually: 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2, 3, 5, 10, 15, 20, 30, 50 and 100 mg/kg. Other values could be used if necessary for special reasons.
In general, it has been shown that the expectable minimum coefficient of variation of residue results of supervised trials is around 30-40% (Ambrus, 1996a); therefore, MRLs below 10 mg/kg should be expressed to only one significant figure to reflect the inevitable variation of analytical results, incorporating the uncertainty of sampling, sample preparation and analysis.
The limit of determination (LOD) is the lowest concentration of a compound that can be determined in a commodity with an acceptable degree of certainty.
The JMPR recognised the difficulties that may arise in regulatory laboratories analysing low level of residues in samples of unknown origin, and stated that it usually estimates a limit of determination which is achievable under those conditions. It is this figure that is proposed as a maximum reside limit "at or about the limit of determination". These limits are indicated with an asterisk (*) after the numerical value (e.g. 0.02*). This limit is often referred to as a "practical limit of determination" to distinguish it from the limits of determination reported in supervised trials.
An MRL so identified does not always necessarily imply that residues of the pesticides do not occur in that commodity. The application of a more sensitive or specific method may reveal detectable residues in some commodities as shown, e.g. in Tables 14 and 26 of the 1995 monograph on quintozene.
In many instances the use of a pesticide according to GAP results in a residue level in crops or commodities that is too low to be measurable by available analytical methods. Setting and enforcing MRLs for residues occurring at or about the limit of determination of analytical procedures may require different approaches depending on the composition and definition of the residues. It is emphasised that all available relevant information should be carefully considered to ensure that an MRL established at a level equivalent to a practical limit of determination of the individual residue components will fully accommodate the levels of these components which could occur in commodities following treatment according to GAP.
As in cases of detectable residues, the definition of residues at or about the limit of determination may also include a single residue component (e.g. fenpropimorph in sugar beet) or several residues components (e.g. aldicarb, its sulphoxide and its sulphone expressed as aldicarb in peanut oil, bentazone, 6-hydroxy bentazone and 8-hydroxy bentazone expressed as bentazone in soya bean; and fenthion, its oxygen analogue and their sulphoxides and sulphones expressed as fenthion in potato).
In cases where several metabolites are included in the definition of the residue two basic situations can be distinguished.
(1) The residue components are or may be converted to a single compound or analyte by the analytical method (e.g. fenthion). The total residue is measured as a single compound and expressed as the parent compound (e.g. fenthion sulfone is measured and expressed as fenthion). The MRL is set and enforced on the basis of the total measured residue. Since, after the conversion of all the residue components a single compound is to be determined, the MRL can be simply enforced either at or above the LOD. This situation is similar to other cases where the residue is defined as a single compound.
(2) The residue components are determined separately by the method. The concentrations of measurable residues are adjusted for molecular weight and summed, and their sum is used for estimating the maximum residue level.
The problem is best illustrated with an example. The residues of bentazone in plant commodities are defined as the sum of bentazone, 6-hydroxybentazone and 8-hydroxybentazone, expressed as bentazone. The LODs reported in supervised trials for each of the three components were generally 0.02 mg/kg, but the practical LODs were regarded as 0.05 mg/kg for regulatory purposes. If an MRL for bentazone was set as the sum of the practical LODs of the three components of the residue, it would have to be established at 0.2 mg/kg (3 times the practical limit of determination to incorporate all three residue components). In this case, any one of the residue components could be present at 0.2 mg/kg, or all of the three at 0.06 mg/kg, without exceeding the MRL. Consequently, individual residue components could be respectively 10 and 3 times those which should arise from the recommended use of the compound but would be within the MRL. Similarly, if the sum of the LODs achieved in the supervised trials was considered, an MRL of 0.1 mg/kg would be needed, which would still allow 5 times the residue that would arise from treatments complying with GAP.
The 1995 JMPR concluded that when residues are undetectable in a commodity an MRL based on the sum of the LODs of the individual residue components is not appropriate for enforcement purposes.
The JMPR considered some of the situations which occur under practical conditions and used the examples to explore the possible actions in order to facilitate the elaboration of an internationally acceptable procedure.
The comments of Codex member countries on the approach recommended by the JMPR and information on their national practice were requested by the Codex Secretariat. The JMPR will consider the comments in the further development of the principles.
5.6.1 Recommendation of Temporary MRLs
5.6.2 Estimation of Guideline Levels
The JMPR recommends to the CCPR that the estimated maximum residue levels be used as MRLs if the toxicological evaluation of the compound enabled the estimation of an ADI.
In those cases where a full ADI could not be estimated or the previously estimated ADI has to be withdrawn, the JMPR does not recommend MRLs. However, the estimated maximum residue levels may be used as Temporary MRLs or Guideline Levels.
A temporary maximum residue limit is a maximum residue limit for a specified, limited period, which is clearly related to required information.
As a general policy, TMRLs will not be introduced for a new compound, a compound in the periodic review programme or when there is no established GAP.
Temporary MRLs may be recommended when some information, which is still lacking, is unlikely to affect the validity of the estimated maximum residue level and there is a clear commitment that the information will be available by a specified date.
The JMPR may recommend a TMRL in some special circumstances, decided on a case-by-case basis, for example:
· The JMPR is informed that experiments are in progress and data from residue or processing trials will be available for a specified Meeting in the future.
· Immediate withdrawal of an MRL may be too disruptive if insufficient opportunity has been given for comment and data submission.
The JMPR may also recommend conversion of an MRL to a TMRL when there is a significant change in GAP which would affect residue levels. The JMPR would require complete information to be supplied by a specified date.
TMRLs for specific commodities may be used to replace group commodity MRLs or "fruit" and "vegetable" MRLs where it is known that residue trials on those specific commodities are in progress. Such a situation has arisen when a group MRL has been scheduled for review and residue data are being developed for some commodities in the group. It would not be correct to withdraw the group MRL without introducing some recognition of the continued validity of maximum residue levels estimated for those commodities while the work is in progress. In the absence of other information, the TMRLs would be recommended at the same level as the group MRL to be withdrawn.
Each recommended TMRL will be directly related to an item of required information. Each such item will have a due date specified. The information is to be available for review at the Joint Meeting in the year specified. If the required information is not supplied by the due date, the TMRL will be withdrawn.
The FAO Panel applies the same principles for estimation of maximum residue levels for TMRLs as those followed for estimating maximum residue levels for recommending MRLs.
A Guideline Level is the maximum concentration of a pesticide residue occurring after use of the pesticide according to Good Agricultural Practice, but for which no Acceptable Daily Intake has been established or it has been withdrawn by the JMPR. There may still be a need to inform regulatory authorities about the residue levels to be expected in food items when these pesticides are used in accordance with Good Agricultural Practice.
Over a number of years the Codex Committee established a list of so-called "Guideline Levels" for pesticides. These Guideline Levels had not been submitted to the Commission for adoption, but were used for the internal reference of the Committee. In 1993 the Codex Alimentarius Commission decided that Guideline Levels would no longer be established. The existing Guideline Levels had been submitted to a review programme in order to delete compounds from the list.
The FAO Panel continues to estimate maximum residue levels for various residue/commodity combinations based on the same principles, regardless of the status of ADI. Thus, the estimated limits can be recommended for use as Codex MRLs when an ADI has been established.