164. Australia contends that this second "warning signal" is effectively no different in character from the first "warning signal" and should therefore be discounted.¹²⁵ We note, however, that in this case the degree of difference in the levels of protection (prohibition versus tolerance) is indeed, as the Panel stated, "rather substantial". We, therefore, consider it legitimate to treat this difference as a separate warning signal.

165. The third "warning signal" the Panel considered was the inconsistency of the SPS measure at issue with Articles 5.1 and 2.2 of the SPS Agreement. The Panel considered that its earlier finding of inconsistency with Articles 5.1 and 2.2:

166. Australia objects to the use of this inconsistency as a warning signal in the context of the third element of Article 5.5.¹²⁷ It argues that inconsistency with Article 5.1 cannot "presume" or pre-empt a finding under Article 5.5. We note that a finding that an SPS measure is not based on an assessment of the risks to human, animal or plant life or health - either because there was no risk assessment at all or because there is an insufficient risk assessment - is a strong indication that this measure is not really concerned with the protection of human, animal or plant life or health but is instead a trade-restrictive measure taken in the guise of an SPS measure, i.e., a "disguised restriction on international trade". We, therefore, consider that the finding of inconsistency with Article 5.1 is an appropriate warning signal for a "disguised restriction on international trade".

167. The first "additional factor" considered by the Panel is the fact that the two substantially different SPS measures that Australia applies (import prohibition versus import tolerance) lead to discrimination between salmon, on the one hand, and herring used as bait and live ornamental finfish on the other. In the Panel's view, the concept of "disguised restriction on international trade" in Article 5.5 includes, among other things, restrictions constituting arbitrary or unjustifiable discrimination between certain products.¹²⁸

168. Australia contends that this first "additional factor" is merely a combination of the first two "warning signals" and does not, therefore, constitute additional "evidence". Furthermore, Australia argues that this first "additional factor" is based on an inappropriate analogy to Article III of the GATT 1994 and a wrong concept of discrimination which, in the context of Article 5.5, means, in its view, discrimination between countries.¹²⁹ According to Australia, the first "additional factor" should therefore be excluded.

169. We believe that the first "additional factor" should indeed be excluded from the examination of the third element of Article 5.5. All "arbitrary or unjustifiable distinctions" in levels of protection will lead logically to discrimination between products, whether the products are the same (e.g., discrimination between imports of salmon from different countries or between imported salmon and domestic salmon) or different (e.g., salmon versus herring used as bait and live ornamental finfish). The first "additional factor" is therefore not different from the first warning signal, and should not be taken into account as a separate factor in the determination of whether an SPS measure results in a "disguised restriction on international trade".

170. The second "additional factor" considered by the Panel was the substantial, but unexplained change in conclusion between the 1995 Draft Report (which recommended allowing the importation of ocean-caught Pacific salmon under certain conditions) and the 1996 Final Report (which recommended continuing the import prohibition). The Panel suggested that the decisive reason for the reversal of the 1995 draft recommendation "might well have been inspired by domestic pressures to protect the Australian salmon industry against import competition".¹³⁰

171. Australia argues that the Panel erred in considering this difference as a factor to be taken into account in the examination of the third element of Article 5.5. Australia contends that the Panel has incorrectly accorded a draft recommendation the status of an SPS measure and that no provision of the SPS Agreement requires WTO Members to implement draft recommendations absent new scientific evidence. Moreover, Australia argues that the Panel refused to consider its arguments and evidence on the role of draft reports and recommendation in the decision-making process of governments. Australia contends that the Panel mischaracterized the reasons for the introduction of QP86A. In Australia's view, the Panel also erred in speculating about the presence and role of lobbying in Australia's decision to adopt the 1996 Final Report.¹³¹

172. We consider Australia's arguments to be without merit. First, we note that paragraph 1 of Annex A of the SPS Agreement defines a sanitary measure of the type relevant in this dispute as a measure applied to protect animal life or health within the territory of a Member from risks arising from the entry, establishment or spread of diseases. In the light of this definition, the Panel was correct to consider the recommendation of the 1995 Draft Report to allow under certain conditions the importation of ocean-caught Pacific salmon to be a recommendation of an SPS measure.

173. Second, we note that the Panel did not at any point state that WTO Members are obliged to implement draft recommendations absent new scientific evidence. It did not introduce such obligation. We note that the Panel explicitly acknowledged that the substantial but unexplained reversal of the 1995 draft recommendation does not constitute, in itself, sufficient proof that the measure results in a disguised restriction on trade.¹³² The Panel merely considered that this factor "can be taken into account cumulatively with other factors" in the examination under the third element of Article 5.5. We agree with the Panel. We do not share Australia's criticism on the Panel's use of the 1995 Draft Report, which the Panel used correctly as "part of the architecture" or "part of a process"¹³³ leading to the 1996 Final Report. We also do not see the relevance of the historical reasons for the introduction of QP86A in 1975 to the examination of the substantial changes in conclusion between the 1995 Draft Report and the 1996 Final Report.

174. The third "additional factor" considered by the Panel was the absence of controls on the internal movement of salmon products within Australia compared to the prohibition of the importation of ocean-caught Pacific salmon.¹³⁴ The Panel did not come to a conclusion on the existence or nature of this alleged difference, but considered that its doubts whether Australia applies similarly strict sanitary standards, "though probably not conclusive as such, can also be taken into account, cumulatively with other factors, in [its] decision on whether the measure at issue results in a 'disguised restriction on international trade'."¹³⁵

175. Australia contends that the Panel erred in implying that consistency with Article 5.5, requires either restrictions on the internal movement of salmon products within Australia or, alternatively, that Australia apply import zoning to grant access to Australia for ocean-caught Pacific salmon.¹³⁶

176. We note that, as acknowledged by Australia¹³⁷, the Panel did not conclude that the alleged absence of internal controls constituted a violation of Article 5.5 or any other provision of the SPS Agreement. The Panel merely stated its doubts on whether Australia applies similarly strict sanitary standards on the internal movement of salmon products within Australia as it does on the importation of salmon products and considered that as a factor which can be taken into account in the examination under the third element of Article 5.5. We consider that these doubts do not carry much weight, but we agree with the Panel that they can nevertheless be taken into consideration.¹³⁸

177. In the above analysis, we have upheld the Panel's findings on the first, second and third "warning signals" as well as its findings on the second and third "additional factors". We have only reversed the Panel's finding on the first "additional factor". We consider, however, that this reversal does not affect the validity of the Panel's conclusion in paragraph 8.159 of its Report, that the "warning signals" and "other factors", considered cumulatively, lead to the conclusion that the distinctions in the levels of protection imposed by Australia result in a disguised restriction on international trade.

178. We, therefore, uphold the Panel's finding that, by maintaining the measure at issue, Australia has acted inconsistently with its obligations under Article 5.5, and, by implication, Article 2.3 of the SPS Agreement.¹³⁹

D. Article 5.6 of the SPS Agreement

179. We next consider the question of whether the Panel erred in finding that the SPS measure at issue is "more trade-restrictive than required" to achieve Australia's appropriate level of protection, and, therefore, that Australia has acted inconsistently with Article 5.6 of the SPS Agreement.

180. After noting that Article 5.6 must be read in context and, in particular, in light of Article 2.2 of the SPS Agreement¹⁴⁰, the Panel turned its attention to the footnote to Article 5.6.¹⁴¹ In the Panel's view, this footnote defines an SPS measure to be "more trade-restrictive than required" only if there is another SPS measure which:

182. The Panel then examined whether any of the latter four quarantine policy options (options 2 to 5, "the four options") meet the three elements of the test under the footnote to Article 5.6. The Panel excluded option 1 since it speaks of the heat-treatment requirement which the Panel considered to be the SPS measure against which the other four options are to be examined.

183. On the first element of the test under Article 5.6, i.e., whether there is an alternative SPS measure which is "reasonably available, taking into account technical and economic feasibility", the Panel noted that the four options were described in the 1996 Final Report as options "which merit consideration", and found that "this implies that the 1996 Final Report put forward the four alternatives ... as technically or economically feasible policy options". The Panel, therefore, concluded that the first element of the test under Article 5.6 is met.¹⁴⁴

184. On the second element, i.e., whether any of these four quarantine policy options "achieves Australia's appropriate level of sanitary protection", the Panel started from the premise that:

On the basis of this premise, it stated that:

185. Following this reasoning and focusing its attention on the second quarantine policy option, i.e., certification, inspection, evisceration and filleting, the Panel found that there are alternative SPS measures which would meet Australia's appropriate level of protection.¹⁴⁸ It, therefore, concluded that the second element of the test under Article 5.6 is met.¹⁴⁹

186. On the third element, i.e., whether any of the four quarantine policy options is significantly less restrictive to trade than the SPS measure currently applied, the Panel pointed out that these four alternative quarantine policy options would "allow imports" of ocean-caught Pacific salmon, albeit under specific conditions, whereas the SPS measure currently applied amounts to an "outright prohibition". The Panel, therefore, concluded that the third element of the test under Article 5.6 is also met.¹⁵⁰

187. Having found that all three elements of the test under Article 5.6 are fulfilled, the Panel reached the conclusion that Australia, by maintaining the measure at issue, has acted inconsistently with Article 5.6.¹⁵¹

188. We note that, in the examination of Canada's claim under Article 5.6, the Panel referred to its earlier ruling¹⁵² that "the measure in dispute is inconsistent with Article 2.2", and, therefore, the Panel decided that it "shall not further address the legal relationship between Articles 5.6 and 2.2".error on the SPS measure at issue, Australia argues that the Panel erred in considering that heat treatment "constitutes" the level of protection for fresh, chilled or frozen salmon. According to Australia, its appropriate level of protection is "explicit in the measure currently imposed, i.e., the import prohibition on fresh, chilled or frozen salmon."¹⁵³

189. Australia appeals from the finding of inconsistency with Article 5.6. Australia does not disagree with the three-pronged legal test under Article 5.6 as set out by the Panel. Australia argues, however, that the Panel erred in law, particularly in the way in which it examined the second element of the test under Article 5.6, i.e., whether any of the four options achieves Australia's appropriate level of protection.¹⁵⁴

190. According to Australia, the Panel incorrectly characterized the existing measure which currently achieves Australia's appropriate level of protection. Australia asserts that the Panel has effectively determined that Australia's appropriate level of protection for fresh, chilled or frozen salmon can be met by certain heat-treatment conditions.¹⁵⁵ In line with its claim of error on the SPS measure at issue, Australia argues that the Panel erred in considering that heat treatment "constitutes" the level of protection for fresh, chilled or frozen salmon. According to Australia, its appropriate level of protection is "explicit in the measure currently imposed, i.e., the import prohibition on fresh, chilled or frozen salmon."¹⁵⁶

191. We recall that we have already concluded that the SPS measure at issue is not the heat-treatment requirement, but rather is the import prohibition on fresh, chilled or frozen salmon.¹⁵⁷ As we observed in the context of our examination of Article 5.1 of the SPS Agreement, the measure to be examined under the SPS Agreement is, therefore, the import prohibition. The Panel had to examine under Article 5.6 whether the import prohibition is "not more trade-restrictive than required" to achieve Australia's appropriate level of protection. Instead, the Panel examined whether the heat-treatment requirement is "not more trade-restrictive than required". Because of this error, we reverse the Panel's finding in paragraph 8.183 that Australia, by maintaining the SPS measure at issue, has acted inconsistently with Article 5.6.

To continue with Article 5.6 of the SPS Agreement

¹²⁵Australia's appellant's submission, para. 217.

¹²⁶Panel Report, para. 8.151.

¹²⁷Australia's appellant's submission, paras. 218 and 273 - 276.

¹²⁸Panel Report, para. 8.153.

¹²⁹Australia's appellant's submission, paras. 219 and 277 - 280.

¹³⁰Panel Report, para. 8.154.

¹³¹Australia's appellant's submission, paras. 281 - 289.

¹³²Panel Report, para. 8.154.

¹³³Panel Report, footnote 418.

¹³⁴Panel Report, para. 8.155.

¹³⁵Panel Report, para. 8.158.

¹³⁶Australia's appellant's submission, para. 290.

¹³⁷Ibid.

¹³⁸That such doubts can be taken into consideration in the examination of the existence of discrimination in the context of Article 5.5 does not prejudice the question of whether or not there is a violation of Article 2.3, first sentence. This point will be taken up in Section F of Part V of this Report.

¹³⁹Panel Report, para. 8.160.

¹⁴⁰Panel Report, para. 8.165.

¹⁴¹Panel Report, para. 8.167.

¹⁴²Panel Report, para. 8.167.

¹⁴³Panel Report, para. 8.168, which incorporates the list of quarantine policy options found in the 1996 Final Report, p.62. "IRA" refers to "Import Risk Analysis".

¹⁴⁴Panel Report, para. 8.171.

¹⁴⁵Panel Report, para. 8.173.

¹⁴⁶Panel Report, para. 8.173. See also para. 8.175, where the Panel said:

¹⁴⁷Panel Report, para. 8.173.

¹⁴⁸Panel Report, para. 8.176.

¹⁴⁹Panel Report, para. 8.181.

¹⁵⁰Panel Report, para. 8.182.

¹⁵¹Panel Report, para. 8.183.

¹⁵²Panel Report, para. 8.99.

¹⁵³Panel Report, para. 8.165.

¹⁵⁴We note that Australia also claims with regard to the first element of the test under Article 5.6 that the Panel misapplied the rules on burden of proof. We deal with this claim in Part VI of this Report.

¹⁵⁵Australia's appellant submission, para. 306.

¹⁵⁶Australia's appellant submission, para. 311.

¹⁵⁷Supra., para. 105.