B. Questions to the Experts and their Compiled Responses (Summarized) (Cont.)

Question 4: In the US first submission¹⁷⁷, it is stated that "the accepted international practice in the area of pest quarantine treatment by virtually every other country in the world is to require testing by commodity, not by variety". To your knowledge, is it common practice for governments to require variety-by-variety testing for high-risk quarantine pests? ¹⁷⁸ How common is the variety-by-variety testing requirement compared to, for example, testing by commodity?

34. Dr. Ducom stated that he had no knowledge of any variety-by-variety testing requirement for quarantine except the Japanese requirements.
35. Dr. Heather noted that although the United States submission contended that international practice was to test by commodity and not by variety, at least two countries, Japan and New Zealand had, in the past, adopted the practice of limiting acceptance to varieties tested although their prohibitions might be listed by commodity. ¹⁷⁹ That the United States had maintained the general policy of accepting the variety used for testing as fully representative of the commodity was evidenced by their own proclaimed schedules. ¹⁸⁰
36. Although the term "international practice" was used, many countries that imported fruit did not actively enforce phytosanitary barriers. It was thus difficult to generalize on the basis of past policy. Before the SPS Agreement came into force, decisions made by countries importing commodities subject to quarantine considerations would have been more influenced by government policies, procedures and precedents.
37. Mr. Taylor noted that as far as he was aware, it was not common practice for governments to require variety-by-variety testing, and it was more usual for testing to be undertaken on a commodity basis.

Question 5: Considering the current Japanese varietal testing guidelines, how long would it technically take (1) to conduct these tests and (2) once these tests are conducted, to come to an administrative decision on acceptability of a new variety of an already tested commodity?

38. Dr. Ducom noted that a varietal test according to the Japanese guidelines was a time-consuming procedure. The exact amount of time it would take would differ greatly depending on whether there was already any active research on the insect at issue. If a test site was permanently set up for mass rearing, the preparation of the insect stages for the specific harvest date was not a major problem. It would, however, demand a sufficient number of insects to take into account unavoidable incidents. There were two series of tests to conduct that would necessitate a one-year interval because there was only one harvest per year:

1. the dose-mortality test which required two to three months (the treatment itself having to be carried out during the harvest); and,
2. the confirmatory test which was best conducted the following year (this test would entail three more months of work).

39. In a case were no mass rearing had been previously set up, it would have to be established. It would take at least 1 to 2 years to domesticate the codling moth.
40. In respect of the administrative decision, in theory, the results of the confirmatory test were immediate and authorization could be given soon thereafter. However, an additional year would be reasonable. The whole process could take at least 3 to 4 years.
41. Dr. Heather noted that the answer to the question depended on whether a new pest and commodity were involved, whether a research infrastructure already existed and whether the same or a similar pest and commodity had been the subject of comparable research elsewhere.
42. A reasonable expectation for research and preparation of a submission where there was an established rearing method for the pest would be between 2 to 3 seasons. After that, consideration by Japanese experts could take up to 1 year or longer depending on their backlog of submissions. Questions from the Japanese experts to be answered, possibility involving further research, could take a further season or longer. After acceptance of a proposal, there would be a need for a (Japanese) confirmatory trial which, depending on the availability of the commodity, might not be possible before the next production season. Further time would be required for Japanese authorities to prepare for a Public Hearing and subsequent regulatory amendment. Dr. Heather noted that for Australia, the time for the above procedure would usually span over 3 to 7 years.
43. Mr. Taylor noted that the test would need to be conducted over more than one season and also with fruit of varying maturity, so that a period of two years would be necessary. He maintained that it ought to be possible to reach an administrative decision regarding acceptance of tests on a new variety of a commodity already tested within a period of twelve months.

Question 6: In the submissions before the Panel, which documents do you consider to contain the fullest statement of the scientific rationale - if any such rationale exists - behind Japan's current varietal testing requirements in respect of apples, cherries, nectarines and walnuts? Is such scientific rationale (if any) linked to the commodity which is examined, or does it apply equally across the commodities?

44. Dr. Ducom noted that Japan's concern was reflected in the statement that, in a few cases, the sensitivity of a given insect stage apparently differed according to the variety that it was on. Turning this into a general principle amounted to a precautionary principle more than any scientific rationale. Japan based its precautionary principle on the fact that too high sorption risked resulting in an insufficient CxT value. According to Dr. Ducom, this argument was not without merit, but he questioned why Japan did not then demand as a criterion for approval the obtainment of a defined CxT value, rather than the setting of initial fumigation conditions.
45. Dr. Heather noted that the rationale advanced by Japan for "varietal testing" was given in section III.E of the first written submission of Japan and defended in II.A of their second written submission. Japan therein identified sorption of the fumigant gas as the major reason for differences in LD₅₀ and CxT parameters between varietal samples and attributed these to "physical and chemical properties of the fruits, which are then attributable to varietal characters" (paragraph 4.111). In Dr. Heather's view, neither parameter was ideal for showing that there were consistent realistic differences in efficacy of a treatment between varieties of a commodity, yet no alternatives appeared more practicable. The CxT product was an average of the fumigant concentrations measured over the fumigation time and the true LD₅₀ was modified by MB sorption by the commodity. Also, individual LD value comparisons did not take into account the slope of the response line and hence did not measure the overall or direct response of the insect population.
46. Essentially, the Japanese argument hinged on whether the differences in test samples that affected these parameters were predominantly varietal characteristics and whether they were of sufficient magnitude to realistically affect treatment efficacy. It was the Japanese view that they did both and this was believed to be the basis of their rationale.
47. The Japanese preferred model first required dose-mortality testing of the pest stage and commodity variety(s) to ensure that differences did not influence treatment efficacy. By using the stage/variety combination, if any, for which the pest was most difficult to kill for the remainder of testing, the greatest risk was deemed to be covered. On this basis subsequent testing of additional varieties should only need to be comparative unless greater difficulty of kill of the pest was demonstrated. The Japanese model acknowledged this but required a further confirmatory test, the reason for which was not readily apparent.
48. It was theoretically correct that results of an experiment were only proven for that exact set of conditions, but this overlooked the purpose of experiments, which was to provide guidance for wider use. The major problem was to determine the extent to which differences were varietal in origin.
49. Mr. Taylor noted that the documents contained within the second written submission of Japan contained the fullest statement of the scientific evidence behind Japan's contention that testing by variety was necessary to establish the efficacy of quarantine treatments. The submission contained references and data to show that differences in sorption and CxT value had been shown between different varieties of cherries, nectarines and walnuts. No data were presented with reference to apple varieties.

Question 7: In respect of the risk of entry, establishment or spread of codling moth due to differences in varieties that could affect the efficacy of the quarantine treatment, what are the technical/scientific factors relevant in an assessment of risk? To what extent has Japan taken these factors into account?

50. Dr. Ducom stated that he did not know risk assessment techniques adequately enough to respond to this question.
51. Dr. Heather noted that risk of entry, establishment or spread of codling moth due to differences in varieties that could affect the efficacy of the quarantine treatment would relate predominantly to:

1. interaction between physical or physiological characteristics of the commodity and the fumigant resulting in higher sorption (fumigant inactivation) in one variety than another, and
2. higher susceptibility of the commodity to the pest for a range of reasons resulting in consistently higher levels of infestation risk in one variety than another.

Japan had concentrated on the first, paying close attention to LD₅₀ and CxT values.

52. The practice of quarantine was one of risk management. A key criterion adopted by Japan for quarantine treatments for otherwise prohibited commodities was that the treatment had to provide a measure of protection equivalent to import prohibition. While import prohibition could effectively exclude commercially traded commodities, it still had to deal with traveller-carried contraband, so in a sense, exclusion of a pest was always less than absolute. Quality requirements normally ensured that commercially traded fruit was essentially free of codling moth infestation even where a quarantine disinfestation treatment was not obligatory.
53. A important factor was the low pre-treatment incidence of codling moth in commercially marketable host commodities from the United States compared to near 100 per cent artificially infested experimental material, i.e., walnuts less than 0.03 per cent; nectarines, less than 0.0003 per cent; cherries, less than 0.00007 per cent; and apples less than 0.00008 per cent as shown in exhibits (these estimates needed to be considered in conjunction with their variance).
54. Mr. Taylor noted that in 1995, the FAO adopted new guidelines on pest risk assessment (PRA), the purpose of which was to ensure that pests were defined as quarantine pests on the basis of scientific principles in order to prevent unfair restrictions on trade. In the case of codling moth, Japan, in its first submission to the Panel (pages 9-15) reported on the PRA conducted in 1996, the assessment being conducted together with that for other pests. The Japanese assessment concluded the following:

1. Codling moth was not present in the country.
2. Environmental conditions in Japan were such as to give a grade 'a' for the potential for establishment of codling moth within Japan.
3. In relation to the potential for the pest to spread within Japan, this was graded as 'b' taking into account the insect's relatively low reproductive capacity.
4. Economic consequences to Japan of the establishment of codling moth were assessed to be of particular significance because host plants such as apples and cherries were produced in great quantity. As a consequence, the PRA analysis resulted in an 'a' grading.

55. In summarizing the assessment, Japan concluded that codling moth presented a very high risk and an overall grade 'A' was given. It was noted that the analysis with respect to codling moth incorporated all the guidelines for PRA (recommended and adopted by FAO).
56. It would appear that Japan did take into account all of the technical and scientific factors necessary in making a proper risk assessment for the entry of codling moth.

Question 8: In the US first submission ¹⁸¹ it is stated that "[e]very article published on the efficacy of methyl bromide and/or methyl bromide and cold storage for disinfection of codling moth has demonstrated that there are no differences among varieties that affects efficacy of a quarantine treatment". Could you comment on this? On the basis of the scientific evidence before the Panel, to what degree does the mortality of codling moth differ between varieties of the same commodity, of either apples, cherries, nectarines or walnuts, when treated with methyl bromide (MB) or MB and cold storage?

57. Dr. Ducom noted that the conclusions of authors who had published articles on the efficacy of methyl bromide had always been that the differences in sensitivity that might exist among different varieties was insignificant. The major argument was that the 20 per cent buffer was sufficient to exceed the limit of a possible difference in varietal sensitivity. ¹⁸² Dr. Ducom noted that there was a need for research on the factors which influenced sorption. Reported differences between varieties were significant with traditional probit statistical methods (the LD₅₀ values were different). However, it was very difficult from that to derive any practical conclusions. A great part of the present problem lay therein.
58. Dr. Heather noted that the US view, that there were no published differences among varieties that affected efficacy, was true if the criterion used was large-scale test results. If the criterion used was statistically significant differences between experimental samples of different varieties there were differences as identified by Japan, but there was no certainty that they were attributable to unique varietal characteristics. In subsequent research on additional varietal samples these differences were too small to cause the efficacy of a treatment, based on varieties used in initial trials, to fail in further testing. This was equally true for commodity-specific treatments with methyl bromide or methyl bromide and cold storage, for apples, nectarines, cherries or walnuts. For each of these, the Japanese minimum efficacy test requirement of no survivors from tests on more than 30,000 insects had been met (3 x >10,000).
59. The treatments for apples, cherries, nectarines and walnuts differed, in part because the fumigations had to be done at different handling temperatures and this affected the toxicity of methyl bromide to the insects. Also, for apples that were cold tolerant in storage, fumigation was supplemented with cold storage treatment. The actual tolerance of codling moth eggs to methyl bromide was unlikely to differ between these commodities. If all were fumigated at the same temperature, the treatments could be expected to be much closer.
60. Japanese concerns were valid, that a low treatment based on dose-mortality testing followed by large-scale trials on a single varietal batch could have resulted in a treatment that would fail on other varieties. It would largely depend on the amount of the "buffer" increase in relation to the difference between varietal samples. Each instance of a very low threshold of susceptibility in a variety in one season�s experiments ¹⁸³ was shown to be anomalous as testing progressed, but some varieties were lower than others in both LD and CxT parameters.

Question 9: Can a difference in the "sorption" level of MB during MB fumigation between different varieties of the same commodity affect the efficacy of the quarantine treatment? If so, do such differences indicate differences in varietal characteristics or are they partly/mostly/completely due to other variables? Could you list the factors which you consider contribute to differences in sorption levels.

61. Dr. Ducom stated that sorption on cereal grains had been covered by Banks 1992, showing its importance in the success or failure of fumigation¹⁸⁴, but there was little understanding of the sorption of methyl bromide on fresh fruits.
62. Sorption levels had a direct influence on the efficacy of a treatment, but this influence was not usually measured directly. It was measured using the CxT value. If the sorption was too high, then the resulting CxT value would be too low, and there was a risk of insufficient efficacy.
63. It was, however, possible to find several cases where the influence of sorption on the CxT value was nonexistent: for example in Yokoyama 1994¹⁸⁵, the set of sorption values (sorption having diverse origins: variety, annual climatic differences, packing material, but with the same concentration of 48 g/m³) was organized in ascending order¹⁸⁶; the corresponding CxT values were distributed at random but within narrow limits. Generally, the inverse was found, this was the case related in Question 3 by the Research Division of Yokohama Plant Protection Station, where the sorption - CxT value relationship was linear. ¹⁸⁷
64. Dr. Ducom stated that to his knowledge, no complete study specific to fresh fruit had been conducted in the United States on the variety-sorption relationship. This would have been a good way to respond to the question in a precise manner or to model possible interactions. Intuitively, it was conceivable that variety could be an intrinsic factor in sorption variation, as was the case, for example, in a study of sorption in raisins. ¹⁸⁸ This study was based on the partition coefficient of methyl bromide between the gaseous phase (the only part of gas that is active on the insect) and the solid phase, i.e. the fruit (the part of sorbed gas that is not utilizable by the insect.) This coefficient was directly linked to sorption and allowed it to be modeled. The study showed that this coefficient could vary considerably between varieties and thus influenced the efficacy of a given general "raisin" dose. The varietal aspect could be the result of interactions like the physiological state of the fruit, its ripeness, etc.
65. Dr. Ducom noted that there was a lack of precise studies on this subject. A priori, one could cite almost anything, the size of the fruit, the nature of the epidermis, the average sugar content, the ripeness of the fruit, its physiological condition, the time between harvest and fumigation, etc. However, in respect of the issue at hand, there was a case which merited study.
66. Dr. Heather noted that differences in sorption levels attributable to varietal characteristics, together with differences in sorption levels attributable to other causes between batches of a commodity, would cause gas concentrations to decay differentially and hence affect the efficacy of fumigation against the pest. Sorption was an inherent characteristic of fumigation as a treatment method. Factors contributing to differences in sorption levels justified detailed comment by an expert in fumigation chemistry.
67. Mr. Taylor noted that difference in the level of sorption of methyl bromide between different varieties of the same commodity could affect the efficacy of quarantine treatment if the level of sorption of a particular variety was of sufficient magnitude to reduce the concentration of methyl bromide gas below the level required to cause insect mortality. However, this would require significant differences between varieties to cause this difference in the level of sorption. Three types of sorption of a fumigant could occur: adsorption, absorption and chemisorption.

1. Adsorption of a gas such as methyl bromide was a physical surface effect and resulted from the attraction of molecules to the surface of a commodity being treated. The larger the surface area the larger may be the adsorption effect. An example would be the case where a commodity had a very rough and therefore greater surface area compared to another with a much smoother surface. Adsorption would be expected to be less in the latter, but the differences in the magnitude of adsorption would probably be small and likely to be insufficient to affect the efficacy of quarantine treatment.
2. Absorption of methyl bromide was also a physical process, but here the chemical entered into the commodity and was held in either solids or liquids. Methyl bromide was absorbed particularly by oils and fats in which it dissolved and, in commodities with a high oil or fat content such as nuts, application rates of methyl bromide were very much affected by this factor. It had been shown that the level of absorption could be affected by the commodity moisture content; the higher the moisture level the higher the sorption.
3. Chemisorption was a third type of sorption and, being chemical in nature, was an irreversible reaction in which residues were left in the fumigated commodity. Methyl bromide reacted in particular with proteins and amino acid groups in a reaction known as methylation, which led to a splitting of the methyl bromide molecule and resulted in inorganic bromide residues. The rate of chemical reactions increased with increasing temperature and for this reason chemisorption took place more readily the higher the temperature.

68. Mr. Taylor noted that both adsorption and absorption were reversible reactions and were affected by the temperature, sorption being greater at lower temperatures. For this reason, methyl bromide application rates had to be greater at lower temperatures.
69. Sorption of methyl bromide was of particular importance for durable commodities because different types of commodities, depending on their chemical constituents, required more or less fumigant to achieve the level of treatment required. Durable commodities could be conveniently placed in groups according to the dosage rates required. These commodity groups were very much connected with oil and fat content, although other facts such as fineness of the commodity which could affect the rate of gas penetration were important and could even affect the length of exposure necessary for effective treatment.

To continue with Questions to the Experts and their Compiled Responses

¹⁷⁷ US first submission, paragraph 96.

¹⁷⁸ The footnotes to paragraph 27 of Japan's first submission identify pests considered by Japan to fall within this category.

¹⁷⁹ Dr. Heather referred to Japan, Exhibit 24: New Zealand MAF Regulatory Authority Standard 155.02.03 (1994).

¹⁸⁰ Dr. Heather referred to USDA APHIS Treatment Manual Section 2.15 and T101. (1992 or subsequent).

¹⁸¹ US first submission, paragraph 83.

¹⁸² Dr. Ducom noted that the amelioration of statistical methods allowed for a reworking of the first results and invalidated the hypothesis of a statistical difference despite the large difference interpreted as significant with traditional methods. (US Exhibit 15)

¹⁸³ Supra note (US Exhibit 14); 1992 New Zealand study on cherries for Sam cherries (US, Exhibit 4 and Japan, Exhibit 21).

¹⁸⁴ Dr. Ducom referred to PD Exhibit 2.

¹⁸⁵ Dr. Ducom referred to US Exhibit 36.

¹⁸⁶ Dr. Ducom referred to PD Exhibit 3.

¹⁸⁷ Dr. Ducom referred to US Exhibit 13.

¹⁸⁸ Dr. Ducom referred to PD Exhibit 4.