(c) CxT Values (Cont.)

124. Japan contended that among the exogenous factors to which the United States attributed the differences in varieties (as set out in paragraph 4.115), the load factor had been controlled in the second and third of the three empirical cases described in paragraph 4.110 encima. Other factors, such as leakiness of fumigation chambers or the measurement error, called for further empirical confirmation. They by no means vindicated the US position that varietal differences were irrelevant to the statistically significant differences in CxT values. In addition, the factor of "the amount of toxicant taken up by the product", which the United States admitted affected the variation in CxT values, was exactly what Japan believed was a key determinant of the values which could be attributable to varietal differences. Japan contended that even though codling moth larvae burrowed into the pulp from the exterior of the fruit, it respired the surrounding, fumigated air. It was therefore reasonable to assume that that CxT was indicative of some interaction with the fumigant.
125. In respect of the US argument in paragraph 4.117, regarding the 1992 New Zealand study on cherries, Japan noted that the cited results indicated possible interaction of variables, and by no means proved the absence of a link between a significant difference of CxT values and varietal differences. Japan reiterated that it was the scientists' responsibility to identify variables, account for them and devise ways to alleviate statistical problems, unless one were to discard the CxT value in phytosanitary experiments. ¹¹⁰ After all the efforts to discredit the statistical significance of variation of the CxT values, the United States had nevertheless admitted that they had value as "tools in estimating and establishing a treatment level that results in mortality" (paragraph 4.120). If the values could vary for "any number of reasons", and if scientists were incapable of controlling these reasons, Japan failed to understand how the values could "estimate" or "establish" any parameter.
126. Furthermore, in respect of the same study, Japan noted that the authors' conclusion applied only to the five varieties tested for efficacy confirmation: "[b]ecause the five cultivars of cherry that we considered have different sorption patterns, different injected doses are required to achieve a given level of mortality. ... Where complete kill is the objective, the commercial rate used by cherry exporters is 64 g/m³, 12 C, 2 hours, and 40 per cent load. This treatment controls codling moth eggs potentially infesting any of the cultivars considered". ¹¹¹ Japan stressed that the authors had, in fact, acknowledged the presence of varietal difference; for they had stated: "We found differences in sorption of methyl bromide between cultivars and between seasons" ¹¹² and, "For the range of mortalities that we considered, the relationship between the c x t sum and the injected dose is close to linear, with a non zero constant term that varies between cultivars and between season. Factors that may affect the sorption pattern of any one cultivar require further investigation." ¹¹³ Japan reiterated that with regard to nectarines, the same authors had suggested that the CxT value "would be a useful measurement to help maintain treatment security for control of codling moth on all nectarine cultivars".¹¹⁴
127. With regard to walnuts, the reason why Japan did not request different quarantine schedules for the three varieties was because, under the concept of a representative variety discussed before, Japan had accepted the large-scale efficacy data on Hartley and found that the proposed treatment would disinfest codling moth to the satisfactory level. It was not because varietal differences were irrelevant.
128. Japan noted, in respect of the US citation of the Hartsell et al., 1991 walnut study on MB treatment (paragraph 4.118), that the study did not include comparative tests of varieties and the cited conclusion was drawn from the result of 100 per cent mortality which was achieved for all varieties stored in the same fumigation chamber at the same time. Absence of varietal differences could not be proved under such conditions. In fact, in a preliminary study it was stated the "[t]here were no significant differences due to cultivar except Eureka, which had higher residue at all three temperatures tested and is indicative of the higher oil content found in this cultivar", and that "[a]ll cultivars desorb MB at about the same rate with the exception of Eureka which due to its higher oil content has a slightly longer retention time". ¹¹⁵ Japan had concluded from these results that the Eureka variety tended to retain MB at a higher level due to its oil content and could show differences in efficacy, and that further comparative tests would be needed on the variety. In the end, the United States had dropped the Eureka variety from the request for approval, and import prohibition was lifted for the other three varieties in 1986. Japan stressed that the issue of MB residue in Eureka was not related to concern over food safety.
129. With regard to Bond's study (paragraph 4.121), Japan had cited an authoritative view on the utility of the CxT value as an indicator of efficacy of fumigation. The Japanese argument was simply that if CxT value varied depending on the variety, efficacy of a treatment might vary according to varieties.
130. The United States maintained that Japan had misinterpreted the scientific data. Japan was alleging that the primary basis for its assumption (that variety mattered) was that differences in CxT values meant that there were differences in sorption between varieties that affected efficacy of treatment. The fact that CxT values had been observed to be different in tests involving different varieties appeared to be the sole basis of Japan's theory that variety mattered. Yet the experts advising the Panel had stated that there was no evidence to support Japan's suggestion that sorption differences had been shown to be large enough to affect treatment efficacy. Such differences, would, in their expert opinion, have to be "significant" or "large" in order to create sorption differences of sufficient magnitude to affect treatment efficacy. Furthermore, Dr. Ducom had emphasized that there was a lack of precise studies on factors contributing to differences in sorption levels and that the "notion of CxT has not been studied enough". In respect of sorption and apples, the United States pointed out that Dr. Heather had explained that while sorption might have an impact on MB treatment, it would have no impact on cold treatment. Thus, there were no grounds to even speculate that sorption differences among apples could affect treatment efficacy.
131. Moreover, the United States noted that, according to Dr. Ducom, if Japan were serious about developing sufficient scientific evidence to support its theory that CxT variations might be related to varietal differences affecting treatment efficacy, Japan could conduct specific studies on the matter. Yet no effort had been undertaken in this respect. In fact, Japan had not put its own theory into practice. In regard to those products that Japan indicated exhibited differences in CxT values (nectarines and walnuts), the fact remained that Japan had never attempted to confirm its hypothesis, to explore why CxT values varied, or to link differences in CxT values with differences in products. In regard to other products subject to this measure (apples, apricots, cherries, pears, peaches, plums and quince) Japan had not made any such observations or assertions regarding variations in CxT. Therefore, presumably there was no basis for applying its measure a priori to these products.
132. The United States reiterated that the measure of efficacy was not CxT values, but whether the exporting country could eliminate the pest of concern at a sufficient rate to achieve Japan�s level of phytosanitary protection. Differences in CxT values were the result of "real world" variables typically seen in testing procedures and did not indicate that varietal differences affected the efficacy of quarantine treatment. Under the SPS Agreement, Japan was required to have a scientific basis for its assumption that variety did matter, but its embrace of CxT values as the basis of such a theory fell short of that responsibility.
133. The United States claimed that if Japan truly believed that CxT values were the indicator of efficacy, there would be no need to show that the dose of fumigant resulted in a particular level of mortality of the pest. Yet the ability of the United States to gain access for a variety of fruit into Japan depended upon its ability to show that the requisite amount of codling moth had been killed. CxT values could vary from Payne walnuts to Franquette walnuts, but the MB treatment was the same because that treatment killed codling moth uniformly in both varieties. There could be differences in CxT values between May Glo and May Diamond nectarines, but the MB treatment was identical for these two varieties because that treatment killed the codling moth uniformly. It was evident that CxT was not a direct indicator of efficacy of treatment. The indicator of efficacy of quarantine treatment was the ability to uniformly achieve the required level of mortality of codling moth. In every instance, regardless of CxT values, and regardless of variety, the level of phytosanitary protection required by Japan had been achieved with a treatment that was uniform for all varieties of a product.
134. The United States also noted that the arbitrary manner in which the measure was applied further called into question its scientific basis. As Japan had explained it, if an exporting country had ten varieties of a commodity available for export, a so-called representative variety could be used in a confirmatory test. Yet if those same ten varieties were proposed in sets of five over two different years, a confirmatory test that previously would have sufficed for all ten varieties had now to be applied twice. There was no logic to this application. Moreover, the United States noted that Dr. Heather had pointed out that if one were to accept that dose-response tests could be used to establish varietal differences, why then would it be necessary to follow-up these tests on newly tested varieties with a confirmatory test?
135. Japan noted that it had never claimed that varietal differences always resulted in differences in CxT values. What was argued was that there existed studies which detected a statistically significant difference of CxT values between varieties. Furthermore, replication-to-replication variation was in itself irrelevant; testing in replicates was exactly for the purpose of eliminating experimental bias. Presence of replication-to-replication variation by no means impaired the utility of the CxT value as an indicator of varietal difference. Japan pointed out that Dr. Ducom had endorsed the presence of the link between varietal differences and differences in the CxT value. In any event, it was the United States' responsibility to identify relevant variables which they claimed existed and establish a treatment which would satisfactorily incorporate them so that Japan's level of protection would be achieved despite such variations.

(d) Comparison with other products

136. Japan noted that while the main focus of the dispute at issue related to fumigation, there were cases where certain differences in efficacy of thermal treatment between different varieties were observed. Japan recalled that the United States had established standards which treated mango varieties differently for hot water treatments. ¹¹⁶ The immersion time was set differently according to varieties with physical differences (the size and shape) of the mango. The treatment time was shorter for fruits with a flat and long shape, or for smaller fruits.
137. Japan did not doubt that the United States had a genuine reason to differentiate the varieties as they related to physical characteristics of the fruits, in light of their impact on the efficacy of the treatment. Under the same logic, Japan's policy of differentiating varieties was justifiable as they related to physical/chemical characteristics of the fruits, which might impact on efficacy of treatment. Moreover, in 1993, when additional varieties of Thai mangoes (Nam Dokmai, Rad and Pimsen Daeng) were investigated, Japan had found by a mortality comparison that the existing vapor heat treatment (46.5° Celsius for 10 minutes, developed for the Nang Klamguan variety) was not sufficient. This resulted in a new vapour heat treatment (the temperature to be raised by a fixed rate to 43 degrees Celsius and held at over 47 degrees for 20 minutes). ¹¹⁷
138. Japan noted that other countries also considered varietal differences in their design of quarantine treatment. According to a Japanese survey, in addition to the Republic of Korea, both Canada and Australia considered relevant issues surrounding varietal differences. Japan noted that Australia had indicated in a communication to Japan that when new varieties were proposed for approval into Australia, their "likely procedure would involve re-evaluation of the relevant pest risk analysis (PRA)," and that "[t]he PRA process would consider any relevant technical information on varietal differences". Hence, contrary to the United States' assertion, Australia would perceive the risk of varietal differences and act accordingly. ¹¹⁸ In the case of New Zealand, the authorities� manual for development of fruit fly disinfestation treatment stated:

"A disinfestation treatment may need to be developed for each variety of fruit separately. While a variety may be described formally in the procurement of proprietary rights, where a variety is not formally described or where varieties can be shown to be morphologically and physiologically similar, definition of the distinctive fruit types must be provided (for example, ovoid eggplant as opposed to oblong eggplant, or long green chilli as opposed to small yellow chilli)." ¹¹⁹

139. The United States noted that Japan, in referring to thermal treatments as used on mangoes from Thailand, was attempting to infer that the US practice of treatment for mangoes made distinctions based on variety; this was distorting the scope of the dispute by introducing alternative treatments which were not within its scope. It was clear to the United States that the scope of the dispute related to products that could be host to codling moth and for which the quarantine treatment of methyl bromide or methyl bromide and cold storage was used.
140. Nevertheless, the United States was compelled to rebut the assertion made by Japan that the United States differentiated quarantine treatment for pests on a variety-by-variety basis. US scientists were aware that such objective morphological characteristics as fruit size and shape, irrespective of variety, were important parameters in determining the schedule for heat treatments. The treatment schedule for mangoes in the United States was not by variety but by size and shape, which could vary within the same variety. As Dr. Heather had noted, size and shape of the fruit "are a varietal characteristic but not exclusively so". The larger or rounder the fruit, the longer it took to heat the fruit to the centre. Thus the heating rate for fruit that was small or flat was much faster than for fruit that was large or round. This was evident from the heat treatment schedules for mangoes found in the APHIS Plant Protection and Quarantine Treatment Manual. ¹²⁰ Moreover, the United States did not require testing of new mango varieties, but assigned a treatment based on the size and shape of the proposed import variety. Furthermore, the United States noted that the experts advising the Panel had stated that reactions of different products to different treatments "have nothing to do with each other" and that fumigation bore little relationship to other treatments. ¹²¹ It was therefore not relevant to compare thermal treatment on mangoes with fumigation on the products at issue.
141. The United States also noted that Japan had indicated that Canada, Australia and New Zealand all took into account varietal differences in quarantine treatments. Recent communications with these countries clearly refuted this assertion:

1. Australia had indicated in a communication with the United States, that it did not require assessments by variety for fumigation quarantine treatments. ¹²² In respect of the letter from Australia referred to in paragraph 4.138 above¹²³, the United States emphasized that Australia did not require varietal testing nor did Australia apply different quarantine treatments based on variety. The letter in fact indicated that Australia had not prejudged the need for a varietal testing requirement, that it would examine "any relevant scientific or technical information on varietal differences" in deciding whether to extend existing treatments to additional varieties. The United States maintained that the experts had confirmed that there was no scientific evidence that varietal differences affected efficacy.
2. The Government of New Zealand had similarly confirmed that it did not differentiate between varieties of a product¹²⁴; moreover, the "MAF Regulatory Authority Standard 155.02.03 Specification for the Determination of Fruit Fly Disinfestation Treatment Efficacy" addressed the possibility of a statement of varieties for dimethoate dip treatments for fruit fly in tomatoes. ¹²⁵ All other fruit fly treatments were determined on a species basis.
3. The Government of Canada, in a communication dated 17 April 1998, had confirmed that it did not require testing by variety for quarantine treatments. ¹²⁶

Hence, in sum, none of these countries required varietal testing for quarantine treatments.

142. In respect of Australia ((a) above), Japan restated that contrary to the United States' assertion, Australia also considered varietal differences in their design of quarantine treatment and would act accordingly (paragraph 4.138). Japan refuted the US statement that the treatment schedule for mangoes was not by variety. Clearly the US schedule referred to varieties. If, as the United States implied, a "flat, elongated" variety such as Frances was sometimes treated as a "rounded" variety, an almost insurmountable administrative problem would arise. It appeared impractical to implement the schedule without a predetermination of the treatment on a varietal basis. Although this schedule was for hot water treatment and not MB fumigation, it did show that concern over varietal difference was quite common.

To continue with Article 5.1

¹¹⁰ In this respect, Japan noted that it was puzzled by the small number of samples in the 1997 retest (30 individuals compared to 160 in 1987).

¹¹¹ Op. cit., p.1229. (Japan, Exhibit 21 and US Exhibit 4)

¹¹² Ibid., Abstract, p.1222. (Japan, Exhibit 21 and US Exhibit 4)

¹¹³ Ibid., Discussion, p.1228. (Japan, Exhibit 21 and US Exhibit 4)

¹¹⁴ Supra note , p.470. (US Exhibit 12)

¹¹⁵ Supra note , p.107. (US Exhibit 17)

¹¹⁶ PPQ Treatment Manual. (Japan, Exhibit 22)

¹¹⁷ Unahawutti, U. et al., Unpublished, 1991. (Japan, Exhibit 23)

¹¹⁸ Japan, Exhibit 38.

¹¹⁹ MAF, Regulatory Authority Standard., paragraph 3.2.1. (Japan, Exhibit 24)

¹²⁰ United States Department of Agriculture, Animal and Plant Inspection Service, Treatment Manual, V.2, pp. 5.45 - 5.58. (US Exhibit 32)

¹²¹ See Section VI of this report, Question 15.

¹²² Letter to Ms. Audrae Erickson from Mr. Paul Morris, dated 20 April 1998. (US Exhibit 33)

¹²³ Letter to Mr. Takeo Kocha from Mr. Christopher W. Wood, dated 18 June 1998. (Japan, Exhibit 38)

¹²⁴ Letter to Ms. Audrae Erickson from Mr. Tony Pautua, dated 9 April 1998. (US Exhibit 34)

¹²⁵ Dated 22 November 1994. (Japan, Exhibit 24)

¹²⁶ Letter to Ms. Audrae Erickson from Dr. J.E. Hollebone, dated 17 April 1998. (US Exhibit 35)