(iii) Evaluation by the Panel

42. We carefully reviewed all the evidence before the Panel in accordance with the rules on burden of proof we set out earlier. ²⁸⁰ After this review and referring, in particular, to the expert opinions quoted above²⁸¹, we consider that, to date, it has not been sufficiently demonstrated that there is a rational or objective relationship between the varietal testing requirement and the scientific evidence submitted to the Panel. In our view, the United States established � on the basis of scientific reports and with the support of the opinions of the experts advising the Panel ²⁸² � that so far not a single instance has occurred in Japan or any other country, where the treatment approved for one variety of a product has had to be modified to ensure an effective treatment for another variety of the same product. We acknowledge that this part of the evidence before us, of course, only relates to those products and varieties for which to date an application for approval to import was made. The United States further provided evidence, supported by the opinions received from the experts advising the Panel, which suggests that varietal differences do not matter for quarantine efficacy, at least not to the extent reflected in the current Japanese varietal testing requirement. ²⁸³ Moreover, even though Japan may have some data � taken from several individual studies � possibly hinting at relevant varietal differences, no evidence before this Panel makes the actual causal link between the differences in the test results and the presence of varietal differences. On these grounds and after having carefully weighed the evidence and opinions of the experts advising the Panel submitted to us, we thus consider that the United States has raised a presumption that Japan's varietal testing requirement is maintained without sufficient scientific evidence and that this presumption has not been sufficiently rebutted by Japan.
43. We thus find that Japan maintains the varietal testing requirement without sufficient scientific evidence in the sense of Article 2.2.
44. According to the scientific experts advising the Panel, the statements they provided on varietal differences are � to the best of their knowledge � equally valid for all US products here at issue. ²⁸⁴ At the meeting with the experts, we asked them the following:

"You know that the scope of this dispute does not cover only the four products, apples, cherries, nectarines and walnuts, but also apricots, plums, pears and quince, even if we have not received any material from either party concerning those four other products. So now the Panel wants to ask you the following question: To the best of your knowledge is what you have stated about varietal differences concerning apples, cherries, nectarines and walnuts, would that also be valid for apricots, plums, pears and quince?". ²⁸⁵

We put this question before the experts since the scope of our mandate covers all eight products ²⁸⁶ and taking into account the Appellate Body's view that a panel needs to make findings with respect to all products falling within its terms of reference. ²⁸⁷

45. Dr. Heather answered "yes" to this question and the other two experts concurred. However, the experts did not further elaborate on their answer. Nor did any of the parties provide any additional comments or information which could enlighten us as to the existence or relevance of varietal differences for the four products for which no specific studies are before us. After careful examination we do not consider, therefore, that there is sufficient evidence before us to extend our finding in paragraph 8.43 also to apricots, pears, plums and quince. We only find that Japan maintains the varietal testing requirement without sufficient scientific evidence with respect to apples, cherries, nectarines and walnuts.
46. With respect to two of the four products to which our finding in paragraph 8.43 does apply (apples and walnuts), the experts advising the Panel made certain additional remarks. First, with respect to apples and the cold treatment they undergo before entering Japan (in addition to fumigation), the experts advising the Panel were even more categorical in their opinion that no evidence before the Panel provides a causal link between varietal differences and a divergent efficacy of the required treatment. Since, according to the experts advising the Panel, the efficacy of cold treatment is not linked to the sorption characteristics of the fruit � the allegedly most important factor which could explain possible varietal differences, if any exist ²⁸⁸ � most (if not all) varietal differences which might exist would be offset by the cold treatment. ²⁸⁹ One expert advising the Panel noted, however, that cold treatment kills the codling moth eggs whereas MB fumigation kills the larvae. ²⁹⁰ The quarantine efficacy across varieties of apples due to the cold treatment would thus only seem to apply to the killing of codling moth eggs.
47. Second, with respect to walnuts, the experts advising the Panel pointed at a specific factor which may influence the sorption level of walnuts, namely their oil or fat content. This could also explain the different test results referred to by Japan. ²⁹¹ However, according to the experts advising the Panel, so far no information is available which shows that the total oil content is a varietal characteristic which could � by means of a different sorption level � affect the quarantine efficacy of an MB treatment. ²⁹² Therefore, the fact that no evidence before this Panel makes the causal link between the differences in the test results and the presence of varietal differences, also applies to walnuts.

(c) Is the varietal testing requirement a provisional measure under Article 5.7?

48. At this juncture � and before we can find, following our considerations and finding in paragraphs 8.42 and 8.43, whether or not Article 2.2 is violated in this dispute � we recall that Article 2.2 provides that "Members shall ensure that any � phytosanitary measure � is not maintained without sufficient scientific evidence, except as provided for in paragraph 7 of Article 5" (emphasis added). We note that Japan invokes Article 5.7 in support of its varietal testing requirement. We therefore need to examine next whether the varietal testing requirement is a measure meeting the requirements in Article 5.7. If the varietal testing requirement meets these requirements, we cannot find that it violates Article 2.2.
49. Article 5.7, in relevant parts, provides as follows:

"In cases where relevant scientific evidence is insufficient, a Member may provisionally adopt � phytosanitary measures on the basis of available pertinent information, including that from the relevant international organizations as well as from � phytosanitary measures applied by other Members. In such circumstances, Members shall seek to obtain the additional information necessary for a more objective assessment of risk and review the � phytosanitary measure accordingly within a reasonable period of time" (underlining added).

(i) Arguments by the parties

50. Under Article 5.7, Japan argues that the rationale for its varietal testing requirement is that "available pertinent information" suggests a possible presence of varietal differences in the efficacy of a disinfestation treatment. Once the import prohibition is lifted for a particular variety, Japan expects that new data will be accumulated on the effects of the treatment approved for that variety in order to reach a sufficient level of confidence as to the broader applicability of that treatment to other varieties. Until such level of confidence is achieved, Japan claims to have the right to maintain, on a provisional basis, the import prohibition for all other varieties of the same product. In this respect, Japan further submits that it has not been demonstrated that the present treatment for any of the US products allowed for import will control the risk at the required level with respect to all other varieties.
51. Japan recognizes, however, that under Article 5.7 it is required to "seek to obtain the additional information necessary for a more objective assessment of risk and review the � phytosanitary measure accordingly within a reasonable period of time". In this respect, Japan submits that its obligation to collect information is discharged by Japan's practice of requiring the exporting countries to submit data each time approval of additional varieties is sought, as well as by MAFF at the Yokohama Plant Protection Station (Research Division) which is seeking to collect information and continues to study the effectiveness of the existing treatments for new varieties.
52. In response, the United States submits that Japan's construction of Article 5.7 is patently incorrect. For the United States, Article 5.7 has a threshold requirement that in order for a measure to be provisional, there must be an insufficient amount of relevant scientific evidence to be able to perform a risk assessment. According to the United States, in this case there is a sufficient amount of evidence. For the United States, all evidence in this case, including the success of uniform treatments on different varieties exported to Japan and the absence of failures by commodity-based testing regimes in other countries, indicates that varietal differences do not affect treatment efficacy.
53. The United States further submits that it strains credulity to describe a 50-year old measure as "provisional". The United States argues that, contrary to what is required in Article 5.7, there is no evidence that Japan has undertaken a process to produce a more objective assessment of risk "within a reasonable period of time" so that it can review whether the "provisional measure" should be continued.

(ii) Evaluation by the Panel

54. In our view, the first sentence of Article 5.7 allows Members to provisionally adopt phytosanitary measures if two elements, cumulative in nature, are met:

• the measure is imposed in respect of a situation where "relevant scientific information is insufficient"; and
• the measure is adopted "on the basis of available pertinent information".

However, even if a measure meets both of these elements, the second sentence of Article 5.7 imposes additional obligations on the Member provisionally adopting the measure, namely the obligation to

• "seek to obtain the additional information necessary for a more objective assessment of risk"; and
• "review the � phytosanitary measure accordingly within a reasonable period of time".

55. Therefore, even if we were to assume that the varietal testing requirement is a phytosanitary measure provisionally adopted in accordance with the first sentence of Article 5.7, i.e., even if we were to assume that in this case "relevant scientific information is insufficient" and there is "available pertinent information" before the Panel on which Japan can base the varietal testing requirement, the second sentence of Article 5.7 obliges Japan to "seek to obtain the additional information necessary for a more objective assessment of risk" and to "review the � phytosanitary measure accordingly within a reasonable period of time".
56. As to the obligation imposed on Japan to "seek to obtain the additional information necessary for a more objective assessment of risk", Japan refers to the fact that exporting countries provide additional information when they apply for access. We note, however, that the studies these countries provide are designed and carried out to comply with the varietal testing requirement. They do not examine the appropriateness of the requirement itself. This is also the case for the two reports before the Panel which were carried out by MAFF's Research Division. ²⁹³ No further information or evidence was submitted to us. As pointed out earlier²⁹⁴, not a single study before the Panel actually addresses the specific issue as to whether varietal characteristics cause a divergency in quarantine efficacy. The requirement that the information necessary to review an SPS measure must be specific enough, was referred to by the Appellate Body in EC � Hormones. ²⁹⁵ In this respect, we further recall that the experts advising the Panel stated that a study or research project to determine whether varietal differences do matter for quarantine efficacy � which would mainly involve sorption tests � could be carried out relatively easily. ²⁹⁶
57. Moreover, with respect to the obligation imposed on Japan to "review the � phytosanitary measure accordingly within a reasonable period of time", we note that, according to Japan, testing variety-by-variety for lifting the import ban imposed in the Plant Protection Law, was first applied in 1969 when the ban was lifted for Hawaiian papayas of the Solo variety, i.e., on a variety basis. For the US products at issue, hosts of codling moth, the import ban was first lifted in 1978. The issue of varietal testing, and the question as to whether it can be scientifically justified, has thus been around for almost 30 years and, with respect to the specific products and pest at issue, for 20 years. During this period of time Japan has been in a position to obtain further information on varietal differences and their relevance to quarantine efficacy. Moreover, since the entry into force of the SPS Agreement on 1 January 1995, Japan has been under an explicit obligation to collect additional information to enable it to more objectively review the appropriateness of the varietal testing requirement.
58. On these grounds, we consider that there is no evidence before us which indicates that Japan sought to "obtain the information necessary for a more objective assessment of risk" and reviewed the varietal testing requirement accordingly "within a reasonable period of time". We consider, therefore, that the United States has established a presumption that Japan did not comply with the requirements in the second sentence of Article 5.7. We also consider that Japan has not been able to rebut this presumption.
59. Following the rules on burden of proof we set out earlier²⁹⁷, we thus find that even if the varietal testing requirement were considered as a provisional measure adopted in accordance with the first sentence of Article 5.7²⁹⁸, Japan has not fulfilled the requirements contained in the second sentence of Article 5.7.
60. In its comments on the interim report, Japan noted that the information gathered through successive demonstrations by exporting countries constitutes experience and that experience is a legitimate means to gather information under Article 5.7. We agree with this point of view. Of course, Japan can take into account the evidence submitted so far by exporting countries. However, in our view, this method of collecting information has, to date, not provided the information "necessary for a more objective assessment of risk" and an appropriate review of the varietal testing requirement "within a reasonable period of time".

(d) The Panel's conclusion under Article 2.2

61. We have found above that the varietal testing requirement � in so far as it applies to imports of apples, cherries, nectarines and walnuts � is neither (1) maintained with sufficient scientific evidence in the sense of Article 2.2 ²⁹⁹ nor (2), in the event it were a provisional measure in accordance with the first sentence of Article 5.7, a measure maintained in conformity with the second sentence of Article 5.7. ³⁰⁰ We recall, however, that Article 2.2 requires Japan to ensure that the varietal testing requirement is "not maintained without sufficient scientific evidence, except as provided for in paragraph 7 of Article 5". Consequently, we come to the conclusion that Japan, by maintaining the varietal testing requirement with respect to apples, cherries, nectarines and walnuts, acts inconsistently with its obligations under Article 2.2.
62. Given this conclusion under Article 2.2, we see no need to further examine what is required for a phytosanitary measure to be "based on scientific principles" in the sense of Article 2.2, nor to determine whether in this dispute the varietal testing requirement is so based.

3. Risk assessment

63. Since we have found earlier ³⁰¹ that the varietal testing requirement violates Article 2.2, we see no need to further examine whether it also needs to be based on a risk assessment in accordance with Articles 5.1 and 5.2 nor to determine whether in this dispute it is so based.

G. Measures not more trade-restrictive than required (article 5.6)

1. Arguments by the parties

64. The United States further claims that the Japanese varietal testing requirement is inconsistent with Article 5.6 in that it is significantly more trade-restrictive than required to achieve Japan's appropriate level of phytosanitary protection. The United States submits that because there are no varietal differences that affect the efficacy of a quarantine treatment, the same established treatment will achieve for all varieties of a product the appropriate level of protection. The United States argues that neither it, nor any other country exporting to Japan, has ever had to modify a quarantine treatment for codling moth for additional varieties of the same product. According to the United States, these results conclusively demonstrate that Japan's varietal testing requirement has no value in providing additional quarantine protection.
65. The United States posits testing by product as a reasonable alternative under Article 5.6. The United States accepts that the first variety of a particular product from any source should be subject to the full gamut of testing. However, it is the US view that after such validation, no further testing at all for additional varieties is necessary. Because testing by variety takes a minimum of 2-4 years to complete per variety, is resource intensive and costly to perform, and seriously delays market access of US products, the United States argues that testing by product is also less trade-restrictive.
66. The United States submits that to be required to only conduct confirmatory tests for each additional variety (as referred to in a question by the Panel to the United States), would be virtually as time-consuming and burdensome as the current requirement to do dose-mortality tests for each variety and confirmatory testing on representative varieties.
67. Although the United States pointed out, in its rebuttal submissions, that it considers testing by product to be the only acceptable quarantine measure in the context of this dispute, in its oral statement at the second substantive meeting it also referred to an alternative measure posited by the experts advising the Panel ³⁰² as a confirmation of the fact that Japan's varietal testing requirement is more trade-restrictive than required. ³⁰³ The alternatives submitted by the experts advising the Panel are outlined below. ³⁰⁴
68. Japan responds that its lifting of the import prohibition for specific varieties is a result of the discharge of its obligation under Article 5.6. According to Japan, whenever it finds a measure which achieves its appropriate level of protection and is significantly less restrictive, the import prohibition is replaced with such a measure. In this particular case, however, Japan submits that it found data which suggests the presence of varietal differences in efficacy of MB fumigation, and a hypothesis which explains such a variation. On that basis Japan requires testing by variety. As the United States has not proven product-wide efficacy, Japan concludes that it should not be obliged to accept the US alternative at this stage.
69. Japan points out that it already made efforts to alleviate the burden put on exporting countries. First, Japan accepts the concept of a representative variety. This means, for example, that for an application for access of five varieties of a product, a large-scale confirmatory test will only need to be carried out for one of the five varieties, i.e., the variety which is shown to be the least sensitive to the treatment in dose-mortality tests. This is why, Japan submits, there is no requirement of full-scale testing of each variety. Second, for approval of additional varieties, the number of codling moth insects required in large-scale demonstrations has been reduced from 30,000 to 10,000.

To continue with Elements under Article 5.6

²⁸⁰ See paragraph 8.13.

²⁸¹ See paragraphs 8.32 - 8.41.

²⁸² See paragraph 8.34.

²⁸³ See, for example, paragraphs 8.35, 8.39 and 8.40.

²⁸⁴ See paragraph 8.33 in fine and footnote 263.

²⁸⁵ Transcript, paragraphs 10.223 - 10.225.

²⁸⁶ See paragraph 8.6.

²⁸⁷ See Appellate Body report on Japan � Taxes on Alcoholic Beverages, adopted on 1 November 1996, WT/DS8/AB/R, p. 26.

²⁸⁸ See paragraph 8.33 and footnote 261.

²⁸⁹ In reply to the Panel's question whether "anything in the apples study submitted by Japan [Kawakami, F., et. al., Methyl Bromide Sorption in Fruit Varieties, Research Division, Yokohama Plant Protection Station, Japan Exhibit 36]� change any of your earlier opinions as to the relevance of varietal differences for quarantine efficacy", Dr. Heather replied:

See also Dr. Heather's reply to Panel question 17:

²⁹⁰ In his comments on the Panel's draft findings now contained in paragraphs 8.73 - 8.101 (received according to the procedure outlined in paragraph 6.116), Dr. Ducom clarified that the combined treatments for apples (MB fumigation and cold treatment) affect two different stages of the codling moth: the cold treatment kills the codling moth eggs, the MB fumigation the larvae (see summary in paragraph 6.114). Dr. Ducom noted the following: "There are indeed two treatments for apples, but they apply to two different and separate stages. The cold kills the eggstage and the gas the fifth larval stage" (translation from French)".

²⁹¹ See paragraph 8.21.

²⁹² As outlined in paragraphs 6.112 - 6.115, after the Panel's second substantive meeting, we addressed an additional question to the experts dealing with walnuts (with reference to a study submitted by the United States, Variation in Polyunsaturated Fatty Acids Composition of Persian Walnut, L. Carl Greve, et. al., J. Amer. Soc. Hort. Sci. 117 (3), pp. 518-522, 1992, US Exhibit 40) namely "whether and to what extent the oil or fat content differs between varieties of walnuts because of varietal characteristics" and "whether any such differences are significant enough to affect quarantine efficacy" (emphasis in original). Dr. Heather answered, inter alia:

Dr. Ducom stated, inter alia:

Mr. Taylor stated, inter alia:

See also the statements by Mr. Taylor, Transcript, paragraph 10.140 and by Drs. Heather and Ducom, Transcript, paragraphs 10.290 - 10.292.

²⁹³ All but two of the studies before the Panel were conducted on behalf of exporting countries. The two exceptions are two studies carried out by MAFF's Research Division (1997 tests on three varieties of Japanese nectarines, Research Division, Yokohama Plant Protection Station, MAFF, 1997, Unpublished, Japan Exhibit 16 and Kawakami, F., et. al., Methyl Bromide Sorption in Fruit Varieties, Research Division, Yokohama Plant Protection Station, Japan Exhibit 36).

²⁹⁴ See paragraph 8.42: no evidence before this Panel makes the causal link between the differential in the test results and the presence of varietal differences.

²⁹⁵ Op. cit., paragraph 200.

²⁹⁶ See footnote 278.

²⁹⁷ See paragraph 8.13.

²⁹⁸ See paragraph 8.54.

²⁹⁹ See paragraph 8.42.

³⁰⁰ See paragraph 8.58.

³⁰¹ See paragraph 8.61.

³⁰² An approach to treatment based on a fixed CxT value to be met by monitoring the MB dose in the fumigation chamber. See additional Panel question 9, posed at the meeting with experts (Transcript, paragraph 10.197), and below paragraph 8.76.

³⁰³ In its answer to an additional Panel question at the second substantive meeting, the United States no longer seemed to consider testing by product as the only alternative, stating that "it remains the preferred quarantine measure" (US answers of 24 June 1998, question 1, p.1, emphasis added).

³⁰⁴ See paragraphs 8.76 and 8.77.