Court dismisses appeal on SEA/AA requirements but adjourns main issue pending CJEU reference on policy status.

APPROVED NO REDACTION NEEDED

THE COURT OF APPEAL

CIVIL

Court of Appeal Record Number: 2023/300

High Court Record Number: 2021/888 JR

Neutral Citation: [2025] IECA 71

Whelan J. Butler J. O’Moore J.

BETWEEN/

FRIENDS OF THE IRISH ENVIRONMENT CLG

APPELLANT/ APPLICANT

- AND -

THE GOVERNMENT OF IRELAND,

MINISTER FOR AGRICULTURE, FOOD AND THE MARINE,

IRELAND AND THE ATTORNEY GENERAL

RESPONDENTS

JUDGMENT of Ms. Justice Butler delivered on the 25th day of March 2025

Introduction:

1. This appeal raises issues concerning the status of a strategy for the Agri-food sector which was prepared by a committee representative of stakeholders in the area and approved for publication by a Government decision of 21st July 2021. The strategy is called Food Vision 2030 and will be referred to in this judgment as “FV2030” or as “the Strategy”. The Strategy is part of a series of such strategies and replaces an earlier strategy called “Food Wise 2025”. Each strategy has had a nominated duration of 10 years but, typically, each has been replaced midway through its anticipated lifespan.

2. The appellant is an environmental NGO which regularly litigates issues of environmental importance before the Irish Courts. The second respondent (referred to in this judgment as “the Minister”) is the Government Minister responsible for the Department under whose auspices the Strategy was prepared and adopted. I will refer to the respondents collectively unless referring specifically to the actions of the Government or of the Minister.

3. The status of the Strategy has a significant bearing on the extent to which it was required by law to be subject to a strategic environmental assessment (“SEA”) and/or an appropriate assessment (“AA”) prior to its adoption. The appellant says that the Strategy was legally required to undergo both of these assessments and that the assessments conducted were deficient. Notwithstanding the fact that both an SEA and an AA were carried out, the respondents’ position is that such assessments were not legally required and, consequently, that any defect in them does not affect the validity (if that is the correct term) of the Strategy, which the respondents characterise as a policy document.

4. In response, the appellant contends that the respondents cannot be permitted to distance themselves from the stance taken by them during the preparation of the Strategy to the effect that it required both SEA and AA. Both of these assessments were carried out in ostensible compliance with the relevant legal requirements. Whilst the appellant no longer relies on arguments based on legitimate expectation and estoppel which were made before the High Court, it asserts that the screening determinations made, particularly regarding AA, remain legally valid and binding.

5. In addition, the respondents have accepted that if an AA was legally required, then the specific conclusions drawn at both the screening stage and the substantive stage of the appropriate assessment carried out did not satisfy the legal requirements of Article 6(3) of the Habitats Directive 92/43/EEC. The reasons for this concession are not relevant to the issues the court had to decide on this appeal, but it is notable that those reasons are not reflected in the grounds pleaded by the appellant. In any event, if the court determines that an AA was legally required in respect of the Strategy, the respondents then concede that an order of certiorari should be granted in respect of it.

Issues on this Appeal:

6. Very helpfully, the parties were able to net the issues remaining live on this appeal down to three core points. I will set them out in the sequence in which they were dealt with by the appellant but, as will become apparent, I propose to deal with them in a slightly different order.

7. The first issue concerns the screening assessment carried out by a consultant on behalf of the Minister and the formal screening determination made by the Minister purportedly under Regulation 46(2) of the EC (Birds and Nature Habitats) Regulations 2011 (S.I. 477 of 2011). The appellant argues that the respondents cannot resile from that determination for the purposes of defending these proceedings. The screening determination and subsequent determination on the AA were formal decisions which have not been quashed or set aside and therefore remain valid. A similar point is made in respect of the SEA conclusion statement. The respondents counter this by pointing to the fact that the requirement to conduct an SEA and an AA have their origins in EU law. Objectively and as a matter of EU law the Strategy either required such assessments or it did not, and decisions made by the respondents cannot be determinative of that question. On the respondents’ case the validity of the screening determination and other decisions is not in issue, only the legal necessity for them.

8. The second point arises only if the respondents are not bound by the screening determination that an AA was required and asks whether the Strategy is a “plan or project” within the meaning of Article 6(3) of the Habitats Directive such that it was both capable of and, on the facts, required to undergo an AA. The appellant argues that the analysis in the screening determination is correct and that an AA was legally required. It then makes a detailed argument as to why the manner in which the AA was conducted and the findings made do not meet the requirements of the Habitats Directive. In the respondents’ view, the nature of the Strategy is such that it is not a plan or project within the meaning of Article 6(3) of the Habitats Directive, noting that those terms are not actually defined in the Directive. The respondents also rely on some fairly technical textual arguments regarding the interpretation of this phrase which were rejected by the trial judge. In respect of the appellant’s arguments regarding the nature of the AA which was carried out, the respondents make two alternate arguments. The first is to point to the alleged deficiencies identified by the appellant to say that these illustrate why it is not possible to conduct an AA of a highlevel policy such as the Strategy and, by extension, why it should not be regarded as a plan or project. Alternatively, if an AA is legally required, the respondents argue that, for the reasons identified by the appellant, it can only be conducted at a very high strategic level which is what was done and, thus, the assessment was “appropriate”.

9. The third issue concerns SEA. Although the argument was not made at the same level of detail as AA, the appellant again asserts that the respondents cannot contend in defence of these proceedings that the screening decision and the SEA carried out were unnecessary and that SEA was not legally required. Further, the appellant relies on Article 3(2)(b) of the SEA Directive to contend that once the Minister determined that the Strategy required AA, then it also required SEA. The respondents assert that there is an anterior question which must be addressed, namely whether the Strategy constitutes a plan or programme within the meaning of Article 2(a) of the SEA Directive so as to bring the Strategy within its scope. Needless to say, the respondents contend that the answer to that question is “No”.

High Court Judgment:

10. Some of the issues canvassed before the High Court and indeed some of those raised in the Notice of Appeal are no longer live on this appeal. No appeal has been brought against the trial judge’s conclusions regarding the publication of material relating to the adoption of the Strategy under the principles applicable to access to information on the environment nor his conclusions on whether the national regulations (S.I. 133 of 2007) properly transpose the requirements of Directive 2003/4/EC governing this issue. Although grounds of appeal were raised regarding the identity of the decision maker and whether individual members of the stakeholder committee needed to be joined as respondents to the proceedings, it appears that these difficulties were adequately met by amendments to the pleadings to which the respondents consented and, thus, did not need to be pursued further. Subsequent discussion of the High Court judgment will be confined to the issues live on this appeal.

11. The High Court judgment, which necessarily had to consider more issues than those outlined above, dealt firstly with the implications of the fact that assessments were carried out, then with the AA grounds and lastly with the SEA issue.

12. The first issue is characterised as a preliminary issue and headed “Estoppel/Collateral Attack”. The appellant is no longer maintaining an estoppel or legitimate expectation argument. However, it is still relying on arguments based on the continuing validity of decisions made in the course of the assessments which were carried out and the contention that to allow the respondents disregard them and the fact that the assessments were carried out would constitute a collateral attack on those decisions. Some, but not all, of the High Court’s analysis under this heading remains relevant to this appeal.

13. The main reasons the High Court dismissed the estoppel argument is that it had already been rejected by the Supreme Court in FIE v. Government of Ireland [2022] IESC

42 (National Planning Framework) (“the NPF case”), on the basis that the official view as to the necessity of an assessment could not be determinative as to the true legal situation. In any event the High Court held that the legal conditions necessary for an estoppel to arise were not met. The argument on this issue on appeal has taken a different and narrower shape than in the High Court.

14. All of the High Court’s analysis under the AA heading is relevant to this appeal. Much of the High Court judgment under this heading is taken up with rejecting the respondents’ linguistic analysis of the Habitats Directive on the basis that it constituted an “inappropriately literal approach to a context where purposive interpretation is paramount”. The trial judge’s central conclusion on the AA issue is that even if the Strategy is correctly characterised as a Government policy, it is of such generality that it does not relate to any site capable of analysis in terms of its conservation objectives (para. 77 of the judgment) and the potential impacts on a European site do not need to be assessed until the stage where they are identifiable (para. 81). Consequently, an AA process regarding the Strategy is neither practicable nor possible (para. 86) and, by extension, was not legally required.

15. Because the High Court dealt with AA prior to dealing with SEA and decided that an AA was not legally required, the analysis under the SEA heading is heavily predicated on the fact that the court had determined that an AA was not required. The other limb of the argument made to the High Court under Article 3(2)(a) of the SEA Directive 2001/42/EC has not been pursued on appeal. Because the High Court disposed of the SEA argument under Article 3(2)(b) on the basis that AA was not required as a consequence of which SEA was not required, it did not formally determine what the respondents contend is the anterior question of whether the Strategy is a plan or programme within the scope of the SEA Directive. The appellant argues that if the court were to hold that AA was required and it then became necessary to decide if the Strategy is a plan or programme within the meaning of the Directive, this issue should be remitted to the High Court for a first instance decision.

16. I am not minded to adopt this approach. It is clear from the statements of opposition filed on behalf of the respondents that the status of the Strategy as not being a plan or programme under the SEA Directive featured centrally in their approach to these proceedings. Para. 7 of the final version of the statement of opposition summarises the respondents’ position in seven principal grounds the first of which is: - “(i) FV2030 did not require assessment under the SEA Directive, where it is not a “plan or programme” within the meaning of that Directive and/or it does not meet the threshold criteria under Article 2(a) or Article 3 of that Directive;” Further, in section 2 of the Respondents’ Notice on this appeal under the heading “Additional grounds on which the decision should be affirmed” the respondents plead at para. 2(i): - “(i) FV2030 is not a “plan or programme” within the meaning of the SEA Directive at all, where it does not define criteria and detailed rules for the development of land or subject implementation of one or more projects to rules and procedures for scrutiny;

(ii) Even if FV2030 is a “plan or programme” within the meaning of the SEA Directive (which is denied), it does not satisfy the threshold criteria in Article 2(a) and therefore falls outside the scope of the Directive.” It is clear from the appellant’s written submissions (para. 126) that the appellant understood that the respondent was squarely raising this issue in the appeal. The appellant chose not to engage with it, opting instead to contend that it would be inappropriate for this court to deal with the issue in light of the reference made by the Supreme Court to the CJEU in the NPF case which raised somewhat similar issues. That reference has since been determined (see further below). The appellant also had a full opportunity to deal with this issue at the hearing of the appeal. I cannot see any benefit in the proposal that, depending on this court’s determination of the various issued raised, the matter should be remitted to the High Court for a further decision on this discrete issue which would then be subject to a further right of appeal to this court. The issue has always been in the case and, if necessary, this court will determine it.

Preliminary Issues and High Court Procedure:

17. I must confess to some difficulty understanding the progression of this case. Leave to apply for judicial review in respect of all relief and on all grounds pleaded was granted by Meenan J. in the judicial review list on 1st March 2022. That order included directions for the filing of opposition papers and the exchange of affidavits with a view to the case being ready for directions as to legal submissions and the fixing of a hearing date by 21st June 2022. There was undoubtedly some slippage in adherence to this schedule and the respondents’ opposition papers were not filed until June 2022 with replying affidavits being filed on behalf of the appellant over 7 months later in January, February and March 2023.

18. What is surprising is that commencing in February 2023, after most of the affidavits were filed based on the pleadings as they stood, there were a series of amendments to the statement of grounds followed by replying amendments to the statement of opposition. Ultimately, there were three rounds of such amendments leading in all to four iterations of the pleadings. There are no motions seeking such amendments included in the papers furnished to this court nor affidavits explaining why the amendments were necessary. Most unusually the last version of both the statement of grounds and statement of opposition were filed after the hearing of the action had taken place in the High Court. It seems that some items which have been deleted from earlier versions of the statement of grounds were completely omitted from this version rather than appearing on the face of the document with a line struck through them. The numbering of some of the paragraphs has changed completely from earlier versions, presumably because paragraphs have been added and/or deleted. However, the approach taken is not consistent as some of the new paragraphs are inserted in the relevant place and marked “A”, “B”, “C”, etc. rather than being numbered. One of the “core grounds” was deleted (number 6) but the remaining core grounds were not renumbered. Despite this, “core ground 7” is then referred to as “core ground 6” at a later point in the statement of grounds under the heading “particulars of legal grounds” - all of which adds to the general confusion.

19. Some of this confusion may be due to the parties’ attempts to comply with the pleading requirements of Practice Direction 126. The obligation under Order 84 Rule 20(2)(a)(ii) of the Rules of the Superior Courts is to state precisely the grounds on which relief is sought and, under Order 84 Rule 20(3), to identify the facts or matters relied on to support those grounds. Under Schedule 3 of this Practice Direction that obligation is extended into an obligation to state, separately, core grounds, particulars of grounds, jurisdictional grounds and factual grounds. In my view the resulting document is frequently disjointed as the grounds of challenge and the facts on which those grounds are based appear in three different places leading to an amount of repetition which distracts from, rather than adding to, the coherence of the pleadings as a whole.

20. It is extremely difficult for an appeal court to follow this level of rolling amendment. It is also difficult to reconcile the constant tweaking of judicial review pleadings with the fundamental proposition that the order granting leave to apply for judicial review sets the parameters of the case an applicant is permitted to argue. I do, of course, acknowledge that the High Court has a power to amend judicial review pleadings under Order 84 Rule 23(2) and presumably will do so bearing in mind the purpose for which amendments to pleadings are allowed more generally under Order 28 Rule 1, namely for the purpose of determining the real questions in controversy between the parties. However, in an adversarial legal system such as ours it is a matter for the parties to define the issues between them and once judicial review proceedings have been pleaded, their amendment should be exceptional and at the instigation of the parties themselves.

21. Fortunately, on this appeal the parties were able to clearly delineate the matters in issue between them and the appeal was able to proceed largely without reference to the detail of the pleadings.

22. One other procedural matter remains unclear. It seems that on 22nd May 2023 an order was made that “core grounds 1 and 2” would be “modularised” by reason of a reference made by the Supreme Court to the CJEU in the NPF case regarding the status of the National Planning Framework as a plan or programme under the SEA Directive. Core grounds 1 and

2 in the statement of grounds deal with SEA issues. Para. 21 of the judgment indicates that this was revisited at the opening of the hearing and the modularisation was confined to issues dependent on the questions referred by the Supreme Court and that it would be open to the parties to argue that points were (or were not) dependent on those questions. This meant, according to the judgment, that “core grounds 1 and 2” were not fully modularised.

23. The judgment proceeds to deal with core grounds 1 and 2 (between paras. 91 and 98). At para. 92 the trial judge notes that “contrary to what might have appeared superficially to be the case when modularisation was first mooted, the jurisdictional question of whether SEA applies does not in fact depend on the answer to the first question referred by the Supreme Court”. The reason for this was that the question referred was premised on the NPF coming within Article 3(2) of the SEA Directive, which the trial judge states did not apply in this case. It is not clear whether arguments were made by the parties as to whether particular points were or were not dependent on the outcome of NPF case reference. The trial judge goes on to reject the contention that an SEA was required on the basis that Article 3(2) of the SEA Directive was not satisfied. Insofar as it still relevant to this appeal, this was largely on the basis of his earlier finding that an AA was not required.

24. As a result of this, the High Court judgment does not address one of the main arguments made by the respondents on this appeal, i.e. the FV2030 was not a plan or programme under the SEA Directive and did not meet the threshold criteria of Article 2(a) of that Directive. It is unclear if the trial judge regarded it as unnecessary to decide this issue in light of his decision under Article 3(2) or whether he saw the issue as one which would be dealt with at a later stage pursuant to the partial modularisation discussed at para. 21 of the judgment. The text of para. 92 coupled with the fact that the orders made include an order that the proceedings be dismissed – without the reservation of any issues for further trial – suggests the former rather than the latter. However, as previously noted, the appellant now suggests that if it is necessary to decide this issue is should be remitted to the High Court.

25. As with repeated amendment to the pleadings, a rolling process during which a trial is modularised, then the modularisation is downgraded to partial modularisation and then, perhaps, not modularised at all is unhelpful to an appeal court in attempting to ascertain exactly what has and has not been decided by the judgment under appeal. For the record I want to be clear that this court regards all SEA issues raised within the proceedings as being within the scope of the appeal insofar as they are raised in the Notice of Appeal and Respondents’ Notice. Therefore, the decision of this court will be determinative of the proceedings as a whole. Food Vision 2030:

26. I now propose to look briefly at FV2030 before turning to the three broad issues before the court. I will look firstly at those arguments contending that decisions made by the respondents during the assessments carried out, particularly the AA screening determination, have legal consequences which cannot now be avoided. I will then look at the SEA issues and finally at the AA issues. As the appellant’s arguments on SEA are heavily dependent on the proposition that an AA was required, I will look at the relevant legislation under both headings before turning to the legal issues. I will deal with the parties’ arguments and the High Court judgment as part of my consideration of each issue rather than as discrete sections in this judgment.

27. The starting point for considering whether FV2030 requires SEA or AA must be the nature of the Strategy itself. It is the latest of a series of stakeholder 10-year strategies which have been updated every 5 years. The printed version is a glossy, colourful document running to some 187 pages, complete with photographs and graphics, as well as the text of the Strategy itself which also contains case studies and summaries of other documents.

28. As an agri-food strategy, FV2030 covers the broad areas of primary agriculture, fisheries and aquaculture, forestry, food and drink processing and manufacturing and, less obviously, the equine sector. It is divided into four distinct, high-level “missions” each of which has between four and six “goals”. Each of the goals has in turn a number of “actions”, the implementation of which is designed to achieve the goals.

29. The missions cover four broad topics namely environmental sustainability, primary producers, food and competitiveness. The goals are framed in high-level, aspirational terms. For example, within mission 2, the second goal is to “improve the creation and equitable distribution of value”. Actions under that goal include things such as the establishment of a National Food Ombudsman, working with the European Commission on market and price transparency and supporting the establishment of cooperatives and producer organisations. Generally, the actions are narrower and more concrete than the goals in that they focus on identifiable, individual steps.

30. Perhaps of more direct relevance to the appellant’s concerns, the goals under the environmental sustainability mission include the development of a climate neutral agri-food system by 2050. As it happens, many of actions under this goal are either research based (i.e. to understand how greenhouse gas emissions from livestock farming can be addressed) or involve the implementation of other existing policies and strategies (e.g. Ag-Climatise or the Sectoral Climate Change Adaption Plan). The published strategy notes “the actions outlined to address this goal are designed to be built on through more detailed plans, programmes, and initiatives”. The appellant argues that the references to “more detailed plans, programmes” is an implicit acknowledgement that FV2030 is itself a plan or programme. This argument is without merit. Whatever about its precise legal status, FV2030 is clearly a policy document designed to reach a varied and non-legal audience. It cannot and should not be construed as though it were a statute or other legal text.

31. The appellant has not identified any specific action as likely to have a particular adverse environmental impact. Rather, agriculture generally is regarded as being capable of having adverse impacts ranging from the production of greenhouse gas emissions by livestock, the pollution of waters through run-off from agricultural fertilisers to the negative impact on biodiversity of overgrazing or other intensive agricultural practices. In oral argument the appellant emphasised the fact that the Strategy envisages that grass-based livestock production, perceived as having a particularly high environmental impact, will remain at the core of Ireland’s agri-food industry.

32. The parties took opposing stances as to the extent to which FV2030 can be said to represent Government policy but did not disagree that it was a “policy”. The respondents argue that the Strategy was adopted by a stakeholder committee representative of various interests in the agri-food sector. The participation of departmental offices and the provision of secretarial and administrative support by the Department have to be seen in that context. The appellant, on the other hand, spent some time identifying all of the places in the document where the Department of Agriculture is mentioned or where the harp, being the official symbol of the State, is used. I do not regard either of these things as particularly significant. I place greater weight on the fact that the committee was established and appointed by the Minister, the Department of Agriculture instructed the consultants who prepared the SEA and AA documentation, the Minister took the AA screening and final decisions, and the Minister established and chairs the implementation committee. All of these point to a level of involvement such that the Minister, and by extension the Government, cannot realistically or fully dissociate themselves from the policy contained in FV2030.

33. However, it does not follow from this that FV2030 necessarily constitutes Government policy. In support of the contention that it does not, the respondent points to the text of the Government decision of 21st July 2021 in which a number of steps are taken. Firstly, publication of the Strategy is “approved”. Secondly, it was agreed that the Minister would “express government support for the broad policy thrust” of the Strategy, especially the “vision” that Ireland would become a world leader in sustainable food systems and that the Government would utilise the Strategy’s approach in a then-forthcoming UN Food Systems Summit. Finally, a number of things are “noted”. These include the establishment and chairing by the Minister of a high-level implementation committee, the implementation process to be consistent with sectoral targets set for agriculture in various programmes and plans dealing with climate. The Government decision describes the Strategy as “a framework within which more detailed plans and processes will need to be developed”.

34. The respondent observes, correctly, that the Strategy is not formally adopted as Government policy. On the face of it, approval for the publication of the report and expressing support for its broad policy thrust falls somewhat short of its adoption by the Government. On the other hand, notwithstanding the fact that the stakeholder committee was comprised of persons representative of various interests in the agri-food sector, the process whereby the Strategy was adopted was undoubtedly driven by the Department of Agriculture. Further, the Government approved the Minister’s action in expressing support for the broad policy thrust of the Strategy, utilising it at UN level and establishing and chairing the implementation committee all of which indicates a significant degree of Government involvement. Therefore, on balance I am inclined to treat FV2030 as Government policy. In fact, there may not be a material difference in this context between Government policy and a policy to which the Minister has subscribed because, as shall be seen, section 34(2)(a)(iv) of the Planning and Development Act 2000 (“PDA 2000”) treats both in the same manner.

35. As the trial judge points out, although this may have implications under section 34(2)(a) of the PDA 2000, it is not of itself determinative of whether SEA or AA is required. The potential relevance of section 34(2)(a)(iv) is that under that section when a planning authority is making a decision on a planning application it must have regard to a number of matters. These include at para. (iv) “where relevant, the policy of the Government, the Minister or any other Minister of the Government”. This in turn affects the extent to which the Strategy can be said to have a bearing on decisions to grant or refuse development consent for projects falling within its scope. There is of course a difference between a policy which binds downstream decision makers and one which does not bind them but to which they must have regard. At its height, where a decision maker is obliged to have regard to a policy it must have a rational basis for any decision which is not consistent with that policy. I will consider further below whether this level of interaction between the downstream

decision maker and the policy is sufficient to render the policy either a plan or programme under the SEA Directive or a plan or project under the Habitats Directive.

36. Whilst the analysis touched on at the end of the preceding paragraph is relevant to the determination of the legal issues in this case, there is in fact no live issue on this appeal as to the characterisation of FV2030 itself. At para. 29 of the judgment, the trial judge accepted a plea in the respondents’ statement of opposition as to the nature and status of FV2030. (As it happens the judgment describes this plea by reference to the second amended statement of grounds and not the third and final version of that document, but it seems that the plea was retained in a similar form in the final version). The plea is as follows: - “14. … FV2030 is a voluntary stakeholder-led strategy which outlines an agreed vision for the Irish agri-food sector and recommends goals and actions to achieve that vision. It does not have any statutory basis, and it does not create legal obligations, whether for stakeholders, the public, the respondents or otherwise.” The trial judge goes on to comment at para. 30 of his judgment: - “It follows that FV2030 is a voluntary stakeholder-led strategy, outlining an agreed vision for the Irish agri-food sector, and recommending goals and actions for that vision. It has no statutory basis, or mandatory or binding down-stream effects. It does not authorise or limit any activities, nor bind any party (including the Government). Nor does it establish binding (or non-binding) criteria for the granting of development consent, or for the adoption of such grant plans, programmes or strategies.

31. That said, the Government decision broadly approving FV2030 is a form of governmental policy, albeit similarly non-statutory and non-binding. As such policy it is not devoid of legal effects in planning law, even though such effects are not themselves binding.”

37. These paragraphs of the High Court judgment are not the subject of any ground of appeal. Consequently, this characterisation must be taken as a starting point for this court’s analysis of the legal status of FV2030 under the SEA Directive and the Habitats Directive. In fairness to the appellant, the argument on the appeal regarding the need for SEA and AA was not based on a detailed analysis of the Strategy in order to disavow this characterisation. Rather it centred on the AA screening determination made by the Minister on 17th April 2021 which the appellant contended brought the Strategy within the scope of the Habitats Directive and by extension the SEA Directive. To understand this argument, I propose to look briefly at all relevant provisions of both Directives and key elements of the SEA and AA processes carried out.

EU Legal Framework:

38. This case was argued almost exclusively by reference to EU legal instruments although the two directives in issue have been transposed into Irish law. The exception to this is the appellant’s reliance on various formal steps taken by the respondent in the carrying out of an appropriate assessment and its publication under the 2011 Regulations (S.I. 477/2011). For ease of reference, I will deal with the arguments as made under the EU legislation and will identify the relevant provisions of the 2011 Regulations when looking at the appellant’s arguments as to the legal consequences of the formal decisions taken by the respondents under those Regulations.

39. Part of the difficulty in this case lies in the fact that the requirement to conduct an AA in respect of plan or programme brings that plan or programme within the scope of the obligation to conduct an SEA but the matters in respect of which the obligation to conduct an AA under the Habitats Directive relates ( i.e. plans or projects) are described differently to those to which the SEA obligations relate (i.e. plans or programmes). This difficulty is compounded because, although there is a definition of plans or programmes under Article 2(a) of the SEA Directive, there is no equivalent definition of what constitutes a plan or project under the Habitats Directive. The issue here is whether the Strategy which can reasonably be described as a policy (although the parties disagree as to whose policy it is) can also be characterised as a plan or project on the one hand or a plan or programme on the other. The term “plan or project” comes from Article 6(3) of the Habitats Directive which provides as follows: “6(3) Any plan or project not directly connected with or necessary to the management of the site but likely to have a significant effect thereon, either individually or in combination with other plans or projects, shall be subject to appropriate assessment of its implications for the site in view of the site’s conservation objectives. In the light of the conclusions of the assessment of the implications for the site and subject to the provisions of paragraph 4, the competent national authorities shall agree to the plan or project only after having ascertained that it will not adversely affect the integrity of the site concerned and, if appropriate, after having obtained the opinion of the general public.”

40. Although the Habitats Directive contains a definition section at Article 1, it does not include a definition of “plan or project”. The respondents point to the use of different terminology at different parts of that directive to argue that as “policies” are mentioned elsewhere, “plan or project” should be read as excluding policies. In this regard reference is made to a recital which refers to “land-use planning and development policies” and Article

10 which obliges Member States to encourage the management of features of the landscape which are of importance to wild fauna and flora “in their land-use planning and development policies” with a view to improving ecological coherence. (The recitals to the Habitats Directive are, unfortunately, not numbered.)

41. I am not convinced that the meaning of “plan or project” in Article 6(3) can be approached in this way. First, I note another recital, which is also pointed to by the respondents, states that an AA “must be made of any plan or programme likely to have a significant effect on the conservation objectives of a site…”. The contents of this recital are reflected in Article 6(3) save that the reference is to a “plan or project” rather than to a “plan or programme”. If anything, this suggests a looseness of language rather than a conscious effort to delineate a plan or project so as to exclude policies.

42. Second, the reference to policies in Article 10 and the related recital is specifically to land-use planning and development policies and is for the purpose of proposing the inclusion of certain matters in those policies. In my view it would be a stretch to conclude that because the Habitats Directive imposes a requirement to include certain matters in land-use planning and development policies that either these policies specifically or policies more generally are excluded from the term “plan or project” under Article 6(3), such that they could never require an AA.

43. Indeed, land-use policies are formally described as “plans” in Ireland (See Part II of the PDA 2000, as amended, and in particular Chapter I “Development Plans” and Chapter II “Local Area Plans”). It is well established that such plans must be subject to AA if they are likely to have any significant effects on any EU site. In Commission v. United Kingdom Case C-6/04 the CJEU confirmed that AA could be required of land-use plans even though they do not in themselves authorise a particular development to be carried out and where a subsequent development consent will be required. As the statutory scheme under consideration required applications for planning permission to be determined in light of the relevant land-use plan, those plans “may have considerable influence on development decisions and, as a result, on the sites concerned” (para. 55). In the same case Advocate General Kokott noted that implementation of parts of a plan which do not have the potential for adverse effects on an EU site can thwart possible alternatives for other components of the plan which do have the potential for adverse effects. Hence, she identified the need to consider at the level of land-use plans the potential for adverse effects if both components of the plan are to be implemented. Of course, this analysis is focussed specifically on landuse plans. By their nature such plans relate to defined geographical areas and while specific, concrete development proposals might not be in issue, the plan usually contemplates particular types of development taking place at identifiable locations.

44. I will return in due course to the question of whether FV2030 as a policy comes within the scope of “plan or project” under Article 6(3). At this point it is sufficient to observe that I accept the appellant’s contention that that phrase should be given a broad interpretation in light of the objective of the Habitats Directive which is to ensure the conservation of natural habitats, fauna and flora (and, via Article 7, of birds), particularly within areas designated for protection under that Directive.

45. The phrase “plan or programme” is defined under Article 2(a) of the SEA Directive. I note the appellant places some reliance on the observations of AG Kokott in D’Oultremont v. Région Wallonne Case C- 290/15 that this Article does not define the two terms but merely quantifies them for the purposes of the Directive. That may be so in the sense that both “plan” and “programme” are words which are capable of bearing practical meanings outside of the planning and environmental sphere. However, in legal terms, the meaning to be borne by those terms for the purposes of SEA Directive is circumscribed by the definition given to them under the heading “definition” in Article 2. It goes without saying that this definition does not necessarily circumscribe the meaning to be given to either word as used in any other directive and this may be what the Advocate General intended. The definition in Article 2(a) is as follows: “For the purposes of this Directive:

(a) “plans and programmes” shall mean plans and programmes, including those co-financed by the European Community, as well as any modifications to them:

- which are subject to preparation and/or adoption by an authority at national, regional or local level or which are prepared by an authority for adoption, through a legislative procedure by Parliament or Government, and

- which are required by legislative, regulatory or administrative provisions;”.

46. Article 3 of the Directive is headed “Scope” and paragraphs 1 and 2 of that Article provide as follows: “1. An environmental assessment, in accordance with Articles 4 to 9, shall be carried out for plans and programmes referred to in paragraphs 2 to 4 which are likely to have significant environmental effects.

2. Subject to paragraph 3, an environmental assessment shall be carried out for all plans and programmes,

(a) which are prepared for agriculture, forestry, fisheries, energy, industry, transport, waste management, water management, telecommunications, tourism, town and country planning or land use and which set the framework for future development consent of projects listed in Annexes I and II to Directive 85/337/EEC, or

(b) which, in view of the likely effect on sites, have been determined to require an assessment pursuant to Article 6 or 7 of Directive 92/43/EEC.” Thus, the key factor determinative of whether a strategic environmental assessment will be required is the likelihood that a plan or programme will have significant environmental effects. That likelihood is determined under Article 3(2) primarily by reference to whether the plan or programme itself, or projects for which it sets a framework for future development consent, require either EIA or AA. SEA will be mandatory in these cases. Under Article 3(3), plans and programmes which “determine the use of small areas at local level” and minor modifications to existing plans and programmes only require SEA where the Member State determines they are likely to have significant environmental effects. Under Article 3(4) a Member State is obliged to carry out an SEA of a plan or programme which does not come within Article 3(2) if the Member State determines it is likely to have significant environmental effects. Article 3(5) to (7) set out the manner on which the determinations referred to in Articles 3(3) and (4) are to be made and Article 3(8) and (9) contain certain exemptions which are not relevant here.

47. All of this suggests that the SEA Directive takes a broad approach to whether a plan or programme should be subject to environmental assessment. As with Article 6(3) of the Habitats Directive and the sub-threshold provisions of the EIA Directive, the likelihood of significant environmental effects operates as a trigger which mandates the carrying out of an SEA which does not otherwise fall within the mandatory provisions of Article 3(2). As the CJEU pointed out in its judgment in D’Oultremont (above) and confirmed in Terre Wallonne Case C-321/18: - “Consequently, given the objective of Directive 2001/42, which is to provide for a high level of protection of the environment, the provisions which delimit the directive’s scope, in particular those setting out the definitions of the measures envisaged by the directive, must be interpreted broadly…” (Para. 40). Nonetheless there is, as the respondents contend, an anterior question to be addressed before considering whether FV2030 comes within the scope of Article 3(2) which is whether it is a plan or programme under Article 2(a) and thus, whether it comes within the scope of the directive at all. I will return to this question.

Are Decisions Taken in the SEA and AA Processes Binding?

48. A central plank in the appellant’s argument is that because the respondents actually carried out SEA and AA in respect of FV2030 they cannot now resile from the proposition that those assessments were legally necessary. The appellant no longer relies on the principles of estoppel or legitimate expectation in making this claim. The argument now made is a more focused one and largely depends on the screening determination made by the Minister in the AA process. This records that he accepted the recommendation in the AA screening report that Stage 2 AA was required. The rationale for this conclusion was, in part, that FV2030 “both alone and in combination with other plans and programmes could have significant effects on Natura 2000 sites (in view of their conservation objectives) depending on where and how certain agricultural activities are implemented.”

49. In making this argument, the appellant relied on the process through which FV2030 was prepared and adopted. I have touched on some of the matters relied on previously (such as the Department’s actions in establishing the stakeholder committee and the Minister’s chairing of the implementation group) but at this point I intend only to refer to those actions specifically referable to the AA or SEA process. Following the establishment of the stakeholder committee, the Minister engaged consultants in respect of SEA and AA in May 2019. A different approach was taken to each of these assessments.

50. The consultants prepared an AA screening report which included a screening statement dated August 2020. That report concluded that Stage 2 AA was required and on 17th April 2021 the Minister’s screening determination accepting that recommendation was published. In anticipation that the recommendation would be adopted, the same consultants were engaged in respect of the assessment itself and on the same date a draft Natura Impact Statement was published along with a draft Agri-food Strategy (which was later named FV2030). These documents were all made the subject of public consultation between April and June 2021 as a result of which additional mitigation measures were included in the draft NIS and the final version of FV2030. During the public consultation phase submissions were received, inter alia, from An Taisce, the EPA, the NPWS and the appellant. I will return later in this judgment to the main argument made by the appellant to the effect that the AA is inadequate, an argument which is reflected in many of these submissions. The final NIS, which was published in July 2021, concluded that the Strategy would not have significant effects on the integrity of any designated site subject to the implementation of the mitigation measures.

51. On 2nd July 2021, the stakeholder committee finalised FV2030. On 16th July 2021 the Minister made an AA determination with respect to FV2030. That determination is headed “In the matter of Article 6(3) of the Habitats Directive And in the matter of Regulation 42 of the European Communities (Birds and Natural Habitats) Regulations 2011 (S.I. 477/2011)”. The text states that AA screening was undertaken in order to comply with Article 6(3) of the Directive and Regulation 42 of the 2011 Regulations. It recites the screening process, and that the NIS was made available as part of the public consultation process before adopting the reasoning and conclusions of the NIS and determining that the Strategy would not adversely affect the integrity of any EU site. Again, that determination is expressly stated to have been made pursuant to Regulation 42 and for the purposes of Article 6(3).

52. Although the formulation of the Minister’s determination reflects the test set out in Article 6(3), the formulation of the AA conclusion in the NIS does not. As this conclusion was expressly adopted by the Minister, the respondents accept that they cannot stand over the validity of the AA process as a whole. Therefore, if in fact an AA was required, then it is accepted that an order of certiorari should issue. The issue at this point is whether the respondents should be allowed argue that an AA is not required in circumstances where they conducted an AA process ostensibly pursuant to statutory provisions governing the carrying out of a mandatory AA and in which they invited the public to participate without suggesting at any time that the process was a voluntary one.

53. The SEA process was slightly different in that rather than carrying out a screening exercise, the consultants were instructed to carry out a scoping exercise. The distinction between these is that a screening exercise is usually carried out to determine if an environmental assessment is legally required whereas a scoping exercise is carried out on the assumption that an assessment is required in order to determine what the parameters of that assessment should be. There can be a significant overlap between them, particularly in the AA context as, if the outcome of an AA screening process is that a plan or project is likely to have significant effects on an EU site, the site or sites likely to be affected and the pathways through which the effects are likely to occur will usually have been identified in the course of the screening exercise. These in turn dictate the parameters of the stage two assessment required.

54. In this case the scoping report (dated August 2020) records at section 2.3 under the heading “Screening”: - “DAFM has opted to not complete the screening stage and proceed voluntarily to SEA Scoping stage. Therefore, screening is not discussed any further”. The environmental report (dated April 2021) was subject to a public consultation process in which 97 submissions were received. This report contains an introductory section headed “Purpose of this Report”. This describes the SEA process under the SEA Directive and the transposing provisions of EC (Environmental Assessment of Certain Plans and Programmes) Regulations 2004 (S.I. 435/2004 and S.I. 200 of 2011) and the Planning and Development (Strategic Environmental Assessment) Regulations 2004 (S.I. 436/2004 and S.I. 2001/2011). It states that the purpose of the report is “to evaluate the likely environmental effects of implementing the [Strategy] as per the requirements of the Directive and Regulations”. There is no further reference to the SEA being undertaken on a voluntary basis. The SEA process concluded with an SEA statement in the form of a report which was published in July 2021. The principal affidavit sworn on behalf of the respondents in the defence of these proceedings is that of a senior departmental official who acknowledges that these various steps were taken in accordance with Regulations 10, 11, 12, 13, 14, 15 and 16 of the SEA Regulations (S.I. 435 of 2004 as amended by S.I. 200 of 2011).

55. In these circumstances, the appellant makes a number of arguments as to why the respondents should not be allowed to change the position they previously adopted for the purposes of defending the proceedings. Some of these arguments are premised on the simple unfairness of allowing the respondents to change their position in this manner and the lack of legal certainty that results if a public body ostensibly conducts a process on a particular legal basis but later is allowed to assert that it was conducted on a different basis or, voluntarily and without any underlying legal basis.

56. In my view, there is some merit in these arguments. It is undesirable, to put it mildly, that the public should be invited to participate in a process on the basis that it is a statutorily required process in which they have a legal right to participate if that is not in fact correct. As noted, there were 97 submissions made in respect of the environmental report, presumably by people and bodies who thought they were engaging in a statutory process.

57. The appellant also argues that many of the decisions made by the respondents in the course of these processes have a statutory status such that the respondents cannot now resile from them. In the appellant’s view these are the extant and valid decisions of a public

decision maker acting in accordance with statutory powers and duties and they cannot be treated as null and void unless they are formally quashed or set aside by a court. In this regard, the appellant relies on the presumption of validity of such decisions as described by Finlay J. (as he then was) in Comhaltas Ceoltóirí Éireann (unreported, 14th December 1977). It also relies on the immediately following observation of Finlay J. that a challenge to the validity of such decisions can only be made by way of judicial review or by seeking a declaration of invalidity. As no such action has been brought in respect of the AA screening determination, the AA conclusion or the SEA statement, the appellant contends they cannot be disregarded by the respondents. The appellant also relies on comments to similar effect by Charleton J. in Mungovan v. Clare County Council [2020] IESC 17 at para. 19, and comments regarding the need for legal certainty of public decisions by Hogan J. in Krikke v. Barranafaddock [2022] IESC 41.

58. The respondents’ response to these arguments is to say that they are not treating any of the decisions identified by the appellant as having been made in accordance with statutory powers and duties as invalid. Rather they say that the decisions were legally unnecessary. On this argument, the decision of the Minister that AA or SEA was required is not determinative of the legal issue as to whether they were required any more than a decision that they were not required would be determinative if, in fact, they were. Thus, they say the core issue – and a logically anterior issue – is whether FV2030 is a plan or programme on the one hand, or a plan or project on the other, not whether the respondents treated them as such. Further, the respondents say that it is inaccurate to characterise this argument as a change of position on their part. The affidavit of the senior departmental official previously referred to states that the Department, as a general practice, takes a “precautionary position” and carries out environmental assessments not required by the SEA Directive, and that this fact should not preclude the court determining this fundamental issue. In the same affidavit it is noted that the Department proceeded voluntarily to the preparatory stage of the SEA process. Similarly, the Department is described as treating high-level strategy and policy documents as falling under the Habitats Directive (on a precautionary basis) where the scope of application of the Habitats Directive to such documents “has not been definitively determined by the CJEU or the domestic courts”.

59. Certainly, the Department is in an invidious position if it undertakes detailed work on a policy, the validity of which would be undermined if it requires, but does not have, an environmental assessment, where the likely applicability of the assessment requirement to the policy is unclear at the outset. In those circumstances it is not unreasonable that assessments be undertaken on a voluntary basis. If the requirement applies, then it has been satisfied and, if it doesn’t, then there will, in any event, be a benefit by the incorporation of environmental considerations from the outset of the policy making process. The difficulty for the respondents is this is not what was communicated to the public when carrying out processes in which public consultation is a crucial element.

60. I do not regard the statement in the SEA scoping report that the Department had omitted screening and proceeded voluntarily to the SEA scoping stage as a clear indication that the entire process was being undertaken voluntarily. As noted, Art. 3 of the SEA Directive is structured in a way which makes SEA mandatory for certain plans and programmes under Art. 3(2) but requires Member States to conduct a screening exercise under Art. 3(4) to determine whether plans and programmes which do not fall under Art. 3(2) are nonetheless likely to have significant environmental effects. A decision to skip the screening stage could equally reflect an acceptance that a policy is plan or programme which was likely to have significant environmental effects as much as the argument now made that it is not a plan or programme at all. Further, the documents relating to the AA process did not contain even this limited qualification.

61. However, notwithstanding my reservations on these issues, this point has been determined against the appellant by the Supreme Court in previous proceedings taken by it against the Government regarding the adoption of the National Planning Framework (FIE v. Government of Ireland [2022] IESC 42 (“the NPF case”). The main focus of that challenge was whether there had been adequate assessment of alternatives to the NPF in the SEA carried out in respect of it. The High Court (Barr J. [2020] IEHC 225) held that the treatment of alternatives was sufficient and refused the relief claimed. That decision was upheld, albeit on slightly different grounds, by the Court of Appeal (Costello J. [2021] IECA 317). However, in the course of argument before the Court of Appeal an issue was raised for the first time (apparently in response to questions by the court) as to whether the NPF came within the scope of the SEA Directive. The proposition that it did not was “shortly rejected” by the Court of Appeal on the basis that the SEA scoping report advised that the SEA Directive should be applied, the process proceeded on the basis that the NPF was subject to SEA Directive and the respondents had not pleaded otherwise (see judgment of Costello J. at para. 91).

62. I should note in passing that the appellant sought to rely on exchanges which took place in the Supreme Court between its counsel and the court in the NPF case to argue that the issue now raised by it was not raised or pursued in the earlier case. I am not persuaded that the transcript of those exchanges supports the contention that the appellant had not relied on the presumptively valid and unchallenged nature of formal decisions made in the course of the assessment processes for the NPF. However, I do not regard the exchanges in the hearing of that appeal to be material for the purposes of this appeal and will focus instead on the resulting judgment of Baker J.

63. The issue became central to the appeal before the Supreme Court with Baker J. describing it as a “logically prior” question to that of the adequacy of the assessment of alternatives if a SEA was required. In considering the elements of the definition of “plans or programmes” under Art. 2(a), Baker J. identified two discrete elements, the first being that it be prepared or adopted by an authority at a national, regional or local level and the second that it be required by legislative, regulatory or administrative provisions. The first was not in dispute in the NPF case as the NPF had been formally adopted by the Government. Consequently, the analysis focused on whether it was “required”.

64. The respondents accepted that it did not follow merely from the fact that the adoption of the NPF by the Government was discretionary, that it could not also be said to have been “required” within the meaning of the Directive. However, they argued successfully that there were no legislative or regulatory provisions which prescribed the preparation and adoption of the NPF, notwithstanding the reference to its predecessor (which it had replaced) in the definition section (section 2) of the PDA 2000. (It might be noted that subsequent to the adoption of the NPF, the PDA 2000 was amended in 2018 to introduce a statutory basis for the NPF). The key part of this analysis, on which both parties relied to differing effect, is as follows: - “63. Neither party asserts that the classification or characterisation made by the respondents in the recitals to the NPF or in public notices announcing its publication could be determinative. The question of whether an SEA assessment is required is one which must be objectively ascertained in light of the function of each plan.

64. The appellant argues that the State went through the process of preparing an Environmental Report, public consultation, farmers’ boundary consultation, the publication of an SEA statement and of a statutory notice and it therefore acted on the basis that an SEA assessment was required before the NPF was adopted. The appellant derived from that fact the proposition that the respondents had ‘promised’ to follow a certain procedure and that therefore it should implement that promise: Attorney General of Hong Kong v NG Yuen Shiu [1983] 2 AC 629 at p. 635 cited with approval by Keane J. in Wiley v Revenue Commissioners [1994] 2 I.R. 160 at pp. 173 – 174. I do not propose to consider this point further for the purposes of this appeal, as it is well established that as the question of scope is one of European law, and must therefore be determined in a conforming manner and objectively, and not from the fact that Government did take an approach or may be considered to have represented that a particular approach was appropriate…

66. The fact that the Government believed it was adopting the plan by reason of a legislative power or obligation cannot determine its legislative basis, nor can the fact that an SEA assessment was carried out or that public consultation was subsequently conducted on this basis.

67. In the absence of a statutory or regulatory source it must be concluded that the plan was in fact adopted by the executive decision of Government pursuant to its power under Art. 28.2 of the Constitution. …”. At a later point (para. 86 of the judgment) Baker J. notes that “required” in Art. 2(a) cannot be interpreted as if it were a purely domestic measure and that it must be given an autonomous meaning in EU law.

65. Baker J. then proceeded to consider whether the NPF met the threshold test in Art. 3(2)(a) of the SEA Directive namely that it “set the framework for future development consent of projects” listed in the EIA Directive. She started her analysis by accepting that the NPF “was intended to affect decision making at national, regional and local levels”. She described it as an “apex plan” with which other development plans and regional strategies were required by statute to be consistent and concluded that as a result the NPF had “downstream statutory effect, albeit it does not mandate a specific result”. At a later point she concluded that the NPF was “undoubtedly” a framework document.

66. A number of CJEU decisions were considered which demonstrated a broad interpretative approach when looking at the concept of “required” under Art. 2(a). In A & Ors (Wind Turbines at Aalter and Nevele) case C-24/19, the CJEU maintained the views expressed in earlier cases that “a measure must be regarded as ‘required’ where the legal basis of the power to adopt the measure is found in a particular provision, even if the adoption of that measure is not compulsory”. Further, a plan or programme was “required” within the meaning of Art. 2(a) if its adoption was “regulated by national legislative or regulatory provisions, which determine the authorities competent to adopt them and the procedure for preparing them”. More recently in Bund Naturschutz in Bayern v. Landkreis Rosenheim Case C-300/20 the CJEU further clarified that a plan or programme would be regarded as “required” where “there exists, in national law, a particular legal basis authorising the competent authorities to adopt that plan or programme even if such adoption is not mandatory”.

67. In deciding whether the NPF was “required” in that sense, Baker J. concluded as follows at para. 106 of her judgment: - “My conclusion is that the NPF was prepared and published by the Government by virtue of the executive power which is vested in it by Article 28.2 of the Constitution. This is an autonomous self-executing power the exercise of which does not in any sense depend on statutory authority. Against that background it is, I think, not easy to see how – judged at least by the bare language of the text of the Directive – a document of this kind prepared by the Government comes within the scope of Art. 2(a) of the Directive.”

68. However, because the Supreme Court is a court of last resort and the issue was not so clear that the court could comfortably arrive at its own conclusion, it referred a series of questions to the CJEU regarding, inter alia, the interpretation of Art. 2(a) of the SEA Directive. Questions were also referred regarding the assessment of alternatives under Art. 5(1) of the SEA Directive, but these are not relevant to the issues in this case.

69. The first of the questions referred asked whether Art. 2(a) of the SEA Directive, read in conjunction with Art. 3(2)(a) meant that a measure adopted by the Executive “other than by reason of a legislative or administrative compulsion, and not on the authority of any regulatory, administrative or legislative measure, is capable of being a plan or programme” to which the SEA Directive applied “if the plan or programme so adopted sets a framework for the downstream grant or refusal of development consent” and thus satisfied the test under Art. 3(2)(a).

70. At this point it might be recalled that initially the High Court directed the “modularisation” of this case to postpone consideration of core grounds 1 and 2 which concern SEA pending the outcome of this reference. It then revisited this decision and decided that the modularisation would be confined to issues dependent on the questions referred by the Supreme Court and allowed the parties to dispute the extent to which the issues were so dependent during the hearing. The judgment proceeds to deal with core grounds 1 and 2 on the basis that the jurisdictional question of whether SEA applied did not depend on the answer to the first question referred to the CJEU because that question was premised on the plan or programme setting the framework for the downstream grant or refusal of development consent under Art. 3(2)(a) which the trial judge held did not apply here.

71. This latter conclusion was based on the decision of the CJEU in NJ v. An Bord Pleanála Case C-9/22 which held that for a plan or programme to come within the scope of Article 3(2)(a) of the SEA Directive it must, inter alia, envisage “developments distinct from those envisaged in another plan or programme provided, however, that it is at least binding on the authorities with competence to grant development consent”. As FV2030 is not binding on any planning decision maker, the last element of this analysis of the test was not satisfied.

72. I do not disagree with this conclusion and I note that the appellant no longer relies on FV2030 fulfilling Art. 3(2)(a) of the SEA Directive. However, the assumption that the CJEU’s answer to the first question referred in the NPF case would necessarily be constrained by the reference in that question to Art. 3(2)(a) or limited to circumstances where the plan or programme in issue set a framework for the grant or refusal of future development consent proved unwarranted. In a judgment delivered on 4th October 2024 (Case C-727/22) the CJEU commenced its analysis of the question by pointing out (at para. 23) that “only ‘plans and programmes’ which meet the definition of that concept set out in Article 2(a) of Directive 2001/42 fall within the scope of that directive, and thus within that of Article 3 thereof.” In other words, the fact that a policy (to use a neutral word) is capable of satisfying the requirements of Art. 3(2) of the SEA Directive, does not have a bearing on the anterior question of whether it is a plan or programme under Article 2(a) and, thus, whether it comes within the scope of the Directive. I will be returning to this judgment when I consider the SEA grounds.

73. What is relevant at this stage is the fact that the Supreme Court permitted the respondents to argue that a SEA was not necessary even though a SEA of the NPF had been conducted on the ostensible basis that it was necessary. In answering the question referred to the effect that the measure, such as the NPF, adopted solely on the basis of the executive power of the State could not constitute a plan or programme under Art. 2(a), the CJEU was expressly aware that the NPF had been the subject of an assessment under the SEA Directive. Thus, the fact of the assessment did not preclude a legal question being subsequently raised as to the necessity for it nor did it preclude that question being answered in favour of the respondents.

74. The appellant argues that the Supreme Court decision does not deal with the specific argument now made which is based on the undisturbed legal validity of the decisions made in the SEA and AA processes. It points to the fact that Baker J.’s analysis commences (at para. 63) with the observation that neither party asserted that the classification made in the various public documents and notices regarding the NPF could be determinative, whereas in this case it makes precisely that argument. This may be so, but it ignores two things.

75. The first is the fact that Baker J. rejected the estoppel argument which was made in the NPF case on the basis that the scope of an EU measure is a question of EU law (para. 64). The same rationale applies to arguments based on the presumption of validity of decisions as a matter of Irish law, perhaps even more so when the decisions in question are ones relating to the EU measure the scope of which is in issue. Indeed, at para. 66 of her judgment, Baker J. expressly rejects the notion that the legislative basis for a decision could be determined by the Government’s belief, the fact that it acted on foot of that belief or that it engaged the public in a consultation process on foot of that belief.

76. The second point with which the appellant does not engage is the respondents’ argument that it is not challenging the validity of the decisions to which the appellant points, merely the legal need for them. Whilst this is an unattractive argument on some levels, it nonetheless has merit in the context of steps taken in circumstances where the applicability of the directives to the Strategy was, at best, unclear. In principle it would be undesirable for a prudent decision maker who undertakes an environmental assessment when it is unclear whether one is required to be fixed with that assessment as a legal obligation, even if there is no such requirement, when a reckless decision maker in the same circumstances who ignores the possibility of an assessment being required and proceeds without one, has no such obligation. For that reason, I accept that a screening determination, albeit formally made under the Habitats Directive and the 2011 Regulations, can be treated as unnecessary without its validity being called into question and without the need for separate proceedings to set it aside.

77. There is also a distinction to be drawn between the question as to whether a measure falls within the scope of a directive and, when a measure unquestionably falls within the scope of a directive, a decision as to whether it meets the threshold set in the directive for a legal obligation to arise. In the latter circumstances, the screening determination is looking at the potential environmental significance of the measure in order to determine whether it passes the relevant threshold at which formal assessment is required. Thus, under Art. 6(3) of the Habitats Directive the question is whether the measure is likely to have a significant effect on a designated site; under Art. 3(4) of the SEA Directive the question is whether the measure is likely to have significant environmental effects. A similar decision falls to be made regarding the likelihood of significant environmental effects of a sub-threshold project under the Environmental Impact Assessment Directive 2011/92/EU, as amended. A positive answer to that question in a screening assessment will trigger a legal requirement to carry out the substantive assessment.

78. Where a determination of that nature falls to be made regarding a measure which is otherwise within the scope of the relevant directive, it would not be appropriate to allow a subsequent challenge to the assessment process to be defeated by an argument that the measure had been wrongly screened in and was not in fact likely to have significant effects on the environment or on a European site. The position is different, however, when the issue raised is whether the measure is within the scope of the directive in the first place. This latter question is a purely legal one. The former is a qualitative one requiring a calculation of both likelihood and significance which, once a positive decision is made, should not be open to being revisited by the decision maker. I think this is correct as a matter of principle but, in any event, the case law of the CJEU makes it clear that the test for the likelihood of a significant effect on the environment is whether such an effect can be positively excluded. Once a measure has been screened in as being likely to have significant effects, then, in the absence of a full assessment, legally it is very difficult to contend thereafter that the likely effect can be positively excluded.

79. In light of this analysis, I reject the appellant’s contention that the issue of whether FV2030 required either SEA or AA cannot be raised by the respondents because the presumptively valid decisions made in the course of those processes have not been formally quashed or set aside.

80. That said, I think it would be preferable where an environmental assessment process is being embarked upon on a voluntary or precautionary basis, that this should be made clear in the documentation made available for public consultation. Public consultation is a key element of all forms of environmental assessment required by European law. Participation in these processes involves a commitment from the public and entails time, effort and expense. It is only fair that members of the public who choose to engage with such a process are made aware of the basis on which it is being conducted, including any caveats as to the legal requirement for such assessment.

Is FV2030 within the Scope of the SEA Directive?

81. I propose to deal with the question of whether FV2030 is within the scope of the SEA and of the Habitats Directive in reverse order to that of the High Court judgment. The High Court focused on the AA issue, found that FV2030 was not a plan or project requiring assessment under that Directive and, by reason of that conclusion, held that it could not satisfy Art. 3(2)(b) of the SEA Directive. It also held that FV2030 did not satisfy Art. 3(2)(a) of the SEA Directive as it is not binding on the authorities with competence to grant development consent, a point no longer in issue on this appeal. Although counsel for the respondents characterised these conclusions as findings by the High Court that FV2030 was not a plan or programme for the purposes of the application of the SEA Directive, I do not think that this characterisation is accurate. No consideration at all was given by the High Court to the anterior question of whether FV2030 is a plan or programme within Art. 2(a) of the directive and, thus within the scope of the obligation created by Art. 3(1). Instead, the High Court proceeded on the basis that the “test for SEA” was set out in Art. 3(2) and that FV2030 did not satisfy that test.

82. For the reasons explained at paras.16 of this judgment I do not accept the appellant’s contention that, if he does not succeed in this appeal, this issue should be remitted to the High Court for a first instance decision. The issue is one which has been front and central to the respondents’ defence of these proceedings from the outset and was expressly put in issue by the respondents in the appeal notwithstanding the fact that it had not been dealt with in the High Court judgment. The appellant accepted that this court has jurisdiction to decide issues raised in proceedings which have not been expressly determined in the High Court but urged that it not be done in this case because of the “complexity of the area”.

83. The resulting position is unsatisfactory because despite the Art. 2(a) point being clearly raised in the Respondents’ Notice, the appellant chose not to address it in its written submissions, save to state that it would not be appropriate for the point to be determined because of the then-outstanding reference to the CJEU in the NPF case. That reference has since been determined in a way which lends overwhelming support to the respondents’ position. Counsel did not address the issue in his main submission to the court on the appeal, save to urge for its remittal. In his reply he made a brief argument by reference to the respondents’ written submissions, mainly focused on asking the court to “steer clear of this because it is so complicated” and that it was “best left alone”. It goes without saying that a party cannot, through its failure to deal with an issue that is clearly raised in an appeal, deprive the court of jurisdiction to deal with it.

84. In fairness to the appellant there is certainly some confusion arising from the respondents’ written submissions which do not clearly distinguish between the criteria in Art. 2(a) which determine whether a plan or programme is within the scope of the SEA Directive and the criteria in Art. 3(2)(a) that the plan or programme “set the framework for future development consent of projects” in order to be subject to the requirement for an environmental assessment. Further, when the respondents’ written submissions expressly turn to Art. 2(a), they start from the premise that Art. 2(a) does not define “plan or programme” and suggest that there is a further, anterior, issue as to whether something is a plan or programme at all before considering whether the criteria in Art. 2(a) are met. This may be so in a case where the criteria of Art 2(a) are otherwise satisfied but, if they are not, I do not think there is much benefit in attempting to ascertain whether something which is not a plan or programme within the scope of the SEA Directive is nonetheless a plan or programme for other purposes.

85. I say this in circumstances where the passage relied on by the respondents, namely para. 32 of the Opinion of Advocate General Kokott in D’Oultremont (above) which in turn refers to the judgments of the CJEU in Nomarchiaki Case C-43/10 and Inter-Environnement Bruxelles Case C-567/10 is linked by them back to the requirement that a plan or programme set a framework for future decision under Art. 3(2)(a). The passage in question sets out the definition of plan or programme as follows: - “According to that finding, a plan or programme is a measure which defines criteria and detailed rules for the development of land and which subjects implementation of one or more projects to rules and procedures for scrutiny …”. There is some merit in the appellant’s argument that sub-paragraph (b) of Art. 3(2) is intended to capture a range of cases which do not fall within sub-paragraph (a), including by reason of not setting a framework for future development consent for projects, a point recognised by Advocate General Kokott in para. 49 of the same Opinion. Because I am satisfied that FV2030 does not come within the scope of the SEA Directive for the reasons elaborated on below, I will leave for another case the question of whether the test for whether something is a “plan and programme” so far adopted by the CJEU is comprehensive in light of the possibility that Art. 3(2)(b) may apply to measures which do not meet that definition insofar as it suggests that the measure must set a framework for future development decisions.

86. In its judgment on the reference made by the Supreme Court in the NPF case, the CJEU expressly addressed the second limb of Art. 2(a), i.e. that to come within the scope of the SEA Directive, a plan or programme must be required by legislative, regulatory or administrative provisions. In answering the question referred, the CJEU repeated its earlier observations, most recently contained in the Bund Naturschutz judgment (above), emphasising that to be required, a plan or programme must be one the adoption of which is required by national legislative or regulatory provisions which determine the competent authority for adopting them and the procedure for preparing them and that there must be a particular legal basis authorising the adoption of the plan, even if its adoption is not mandatory. In short, the key consideration is that to come within Art. 2(a), the adoption of the plan or programme must be “regulated” within the meaning of the CJEU’s case law.

87. The CJEU then identified the absence of any such legal basis as regards the NPF. The vesting of executive power in the Government under Art. 28.2 of the Constitution did not constitute a legal provision which regulated the adoption of such plans or programmes, it merely established the exercise of executive power by the Government in accordance with the constitutional separation of powers. Even accepting the need for a broad interpretation, the CJEU felt that to treat a measure such as the NPF as coming within the scope of the directive “would go beyond interpreting that directive broadly” and “would be tantamount to rendering meaningless the second of the two cumulative conditions set out in Art. 2(a) of the same directive”.

88. In my view, FV2030 is even less capable of satisfying Art. 2(a) than the NPF was. Although the NPF was not adopted pursuant to statute, it was recognised by s. 2 of the PDA 2000 and formed the apex in a hierarchy of land-use plans under that Act. There is no statutory or other regulatory basis for FV2030 and it is not referred to in the PDA or other similar legislation. The only statutory connection is the likelihood that it would be regarded as the policy of the Minister or the Government under s. 34(2)(a)(iv) of the PDA and, thus, is something to which planning decision makers must have regard in making decisions to grant or refuse planning permission. The High Court characterised FV2030 as not being devoid of legal effects, even though such effects are not in themselves binding (para. 31 of the judgment). The downstream effects of the NPF are undoubtedly greater - and intended to be so - in that other plans in the land-use hierarchy which are binding on planning decision makers must be consistent with its terms.

89. The fact that FV2030 is not binding on decision makers is undoubtedly significant. However, the question of whether a measure is binding on downstream decision makers is relevant to whether it meets with the CJEU’s definition of “plan or programme” referred to above (“Defines criteria and detailed rules for the development of land and which subjects implementation of … projects to rules and procedures”) and to whether it comes within Art. 3(2)(a) (“Set the framework for future development consent of projects”). It does not appear to be relevant to Art. 2(a) which looks instead to the extent to which the preparation and adoption of the measure itself is subject to binding rules and procedures rather than the extent to which it subjects other decisions to binding rules and procedures.

90. For all of these reasons, there does not appear to be any basis for distinguishing between FV2030 and the NPF in a manner which would bring FV2030 within the second limb of Art. 2(a) when the CJEU has held that that the NPF clearly falls outside it.

91. There was some dispute before the High Court as to the authorship of FV2030 and whether ownership of the Strategy belonged to the stakeholder committee or to the Government or the Minister and, if the former, whether the stakeholder committee could or should be properly sued in judicial review proceedings. Indeed, this seems to have prompted both amendment and re-amendment of the statement of grounds. The issue was ultimately resolved by focusing the relief sought on the Government’s decision “to adopt, publish or otherwise approve the strategy”. Therefore, it became unnecessary for the High Court to focus in greater detail on the identity of the decision maker.

92. Nonetheless, this is an issue which could be relevant under the first limb of Art. 2(a) as the plans or programmes referred to are ones which must be prepared or adopted by “an authority”. It seems unlikely that the stakeholder committee would be regarded as an authority per se unless the degree of departmental involvement were to be regarded as determinative of its character. In that context, the text of the Government decision in which the support of the Government for the “broad policy thrust” of FV2030 is expressed and its publication approved, may be crucial to deciding whether it has been “adopted” by the Government. Thus, the role of the stakeholder committee and the degree of involvement by the Department in the preparation of FV2030 could be relevant to whether FV2030 could be said to have been prepared by an “authority”.

93. At this point I do not propose to offer a view on these issues. This is mainly because the requirements of Art. 2(a) are cumulative, and the second limb is clearly not met in this case. In addition, the argument made by the respondents did not squarely address this point and, as noted, the appellant largely chose not to engage with the issue at all even though its’ factual arguments regarding the manner in which and by whom decisions were made in the course of the assessments did touch on matters potentially relevant to this issue.

94. I am satisfied that FV2030 does not satisfy the criteria of Art. 2(a) and, therefore, it does not come within the scope of the SEA Directive. For that reason, it is not necessary to determine whether the criteria of Art. 3(2)(b) are satisfied since, even if they were, it could not bring FV2030 within the scope of the Directive.

95. Before leaving the SEA issue, I should note the nuance of the appellant’s arguments under Art. 3(2)(b). The appellant argued that even if the High Court decision were correct in holding that FV2030 was not a plan or project within the Habitats Directive such that AA was not required, this would not be determinative of the applicability of Art. 3(2)(b). Looking closely at the text of that provision, the appellant argues that the test is not whether an AA is actually required under Articles 6 or 7 of the Habitats Directive but whether the plan or programme has “been determined to require an assessment pursuant to Article 6 or 7”. As a screening determination was made by the Minister that FV2030 required AA, the appellant contends that Art. 3(2)(b) was and remains satisfied even if the Habitats Directive were subsequently held not to apply to it. Needless to say, the respondents reject this argument which they characterised as being based on an overly literal interpretation of Art. 3(2)(b). There are also issues as to the extent to which a screening determination, the validity of which has not been challenged, remains binding on the respondents if FV2030 were found not to fall within the scope of the Habitats Directive.

96. I do not propose to decide these issues as the potential applicability of Art. 3(2)(b) does not arise where the measure in question does not satisfy the criteria in Art. 2(a). Nonetheless there is a point which may arise for determination in another case as to whether it is the actual need for assessment under Art. 6 or 7 of the Habitats Directive which is relevant to Art. 3(2)(b) or whether a determination that such an assessment is required, even if it is not, suffices to bring a plan or programme within the scope of the obligation in Art. 3(1).

Is FV2030 a Plan or Project under the Habitats Directive?

97. The question of whether FV2030 is a plan or project under the Habitats Directive is, in my view, a more complex one than whether it is a plan or programme under the SEA Directive. The Habitats Directive contains no definition of plan or project and no qualification whatsoever on what might constitute a plan or project under Art. 6(3). Further, whilst the SEA Directive requires a plan or programme to be prepared or adopted by “an authority”, there is no restriction imposed by the Habitats Directive on the person or entities who may be responsible for a plan or project. The EU has repeatedly stated that, for obvious reasons relating to the need to ensure a high level of protection for Natura 2000 sites, the provisions of the Habitats Directive must be interpreted broadly.

98. There is some overlap between the terminology used in the three main directives under which environmental assessments are required under European law. The SEA Directive applies to plans and to programmes; the Habitats Directive to plans and projects and the EIA Directive to projects. However, it cannot be assumed that the meanings of these terms are co-extensive as between the various directives in which they appear. In Coöperatie Mobilisation for the Environment UA (“the Dutch Nitrates Case”) Cases C-293/17 and C- 294/17 the question referred to the CJEU asked whether certain activities carried out in the vicinity of a Natura 2000 site (the grazing of cattle and application of fertiliser) could be classified as a project under the Habitats Directive when, because they did not involve a “physical intervention in the natural surroundings”, they did not constitute a project under the EIA Directive.

99. In answering the question, the court noted that the qualifying conditions in the EIA Directive are not to be found in the Habitats Directive which simply requires assessment where a project is likely to have a “significant effect” on a designated site. As the definition of the concept of “project” in the EIA Directive is more restrictive than in the Habitats Directive, any activity coming within the former will automatically come within the latter. However, it does not follow that the meaning of “project” for the purposes of the Habitats Directive is to be limited by reference to the EIA Directive. The court stated at paras. 66 and 67 of this judgment: - “66. It follows that, if an activity is regarded as a ‘project’ within the meaning of the EIA Directive, it may constitute a ‘project’ within the meaning of the Habitats Directive. However, the mere fact that an activity may not be classified as a ‘project’ within the meaning of the EIA Directive does not suffice, in itself, to infer therefrom that the activity may not be covered by the concept of ‘project’ within the meaning of the Habitats Directive.

67. In the second place, in order to determine whether the grazing of cattle and the application of fertilisers on the surface of land or below its surface may be classified as a ‘project’ within the meaning of Article 6(3) of the Habitats Directive, it is important to examine whether such activities are likely to have a significant effect on a protected site.” Two things emerge from this analysis. First, qualifying conditions or criteria which limit the scope of the concept of a plan or project under any of the other directives will not be read into the Habitats Directive. Second, the only criterion by reference to which it will be determined whether something is a plan or project is that found in Art. 6(3) namely its potential to have a significant effect on the designated site. Thus, unlike the SEA Directive and the EIA Directive where there is an anterior question to be asked in all cases as to whether the measure is a plan, programme or project within the scope of the relevant directive before the likelihood of significant environmental effects falls to be considered, under the Habitats Directive the only question to be asked about the measure is whether it is likely to have a significant effect on any designated site.

100. Unsurprisingly, the appellant places significant reliance on the fact that a screening assessment of FV2030 determined that it was likely to have significant effects on designated sites. On one level, that might be seen as conclusive. However, the respondents argue, partly based on the nature of the appellant’s challenge to the AA which was carried out, that FV2030 is a measure of a type that does not come within the scope of the Habitats Directive. It may be useful to look briefly at the manner in which the assessment was carried out and the challenge now made to it.

101. FV2030 is a national policy covering the entire country, the nature and contents of which are considered in some detail at paras. 27 to 37 of this judgment. It is not a land-use plan in that it does not envisage any particular development nor that particular types of development will take place in particular areas. It does not authorize any project, nor does it bind any downstream decision makers in the grant or refusal of development consent. At its height it may be something to which decision makers will have regard when making a development consent decision.

102. The NIS which formed the basis of the AA carried out was updated to incorporate the results of the public consultation process and the final version of that document is dated July 2021. There is also an AA conclusion statement dated July 2021 which is effectively a summary of the final version of the NIS.

103. The NIS identifies the existence and location of 439 special areas of conservation (“SACs”) and 165 special protection areas (“SPAs”) in Ireland and its territorial waters (as of 2020). It does not treat any of these sites individually but instead identifies the predominant habitat types in Ireland’s SACs, the Annex I habitats present in Ireland, the number of SACs in which each habitat type is present, and the overall status of those habitat types (which, unfortunately, is generally “inadequate” or “bad”) and Annex II species (the status of some of which is, fortunately, “favourable”). A similar exercise is carried out in respect of bird species in Ireland’s SPAs.

104. Under the heading “Relevant Natura 2000 Sites” the NIS states at para. 3.5: “The [Strategy] applies to the whole territory of Ireland; agricultural land encompasses much of the country, and fishing includes Ireland’s EEZ, which extends to a 200 mile nautical limit. The missions, goals and actions are therefore applicable to farmers, other land-owners and fisheries (including aquaculture). It is therefore not possible to screen out any Natura 2000 sites on a geographical basis. It should also be noted that it is not feasible nor desirable to assess individual Natura 2000 sites at a national scale.” Consequently, the approach taken to the carrying out of the AA is identified at para. 3.5.1 as follows: - “It is not the place of this NIS to identify impacts on specific Natura sites, rather to identify impact pathways and to discuss their relevance to site ‘types’. For example, diffuse pollution is likely to have detrimental impacts on a range of water-dependent Natura habitats and species. Impacts are therefore discussed at the ‘strategic level’ in relation to a range of similar sites, and are not broken down to individual site level (as would be appropriate at project level Appropriate Assessment).”

105. The NIS then goes on to look at the impact of agriculture on the decline in the conservation status of protected habitats identifying a number of agricultural practices which impact on habitats. These include inappropriate grazing, land abandonment and the generation of water, air and marine pollution. The impact on species was regarded as less severe.

106. The potential effects of the Strategy are dealt with at chapter 4 of the NIS. This chapter comprises a high-level description of key elements of the Strategy under each of the four missions and an equally high-level account of the potential environmental consequences of the implementation of their goals. No reference is made to any particular Natura 2000 site or to any element of the Strategy being implemented at any particular location. Rather, the discussion is on the likely impact of certain agricultural practices on particular types of habitats and/or species – but actually mainly on habitats.

107. The discussion of mitigation measures is at chapter 5 of the NIS and is covered in two pages. At para. 5.1 the NIS states: - “There is clearly a multiplicity of potential impacts on a wide range of Natura habitats and species arising from agricultural activities (including those from forestry and fisheries). This means that any strategy addressing these potential impacts needs to be broad in scope and to include cross-checking of its various missions, goals and actions. Key areas for mitigation in relation to the location and extent of Natura 2000 sites (i.e. where these sites lie in proximity to agricultural activities, managing agricultural activities) in accordance with national and EU legislative mechanisms, managing diffuse and atmospheric pollution and regulation and management of fisheries. The [Strategy] addresses all of these issues in some detail and provides appropriate mechanisms for reducing and counteracting impact pathways.” Seven specific additional mitigation measures of a fairly generic nature are proposed, the first of which requires “the targeting of the most appropriate measures in the most appropriate places”.

108. The appellant’s challenge to the validity of the AA is largely discursive and includes lengthy quotations from the NIS which are subjected to a critical commentary. Insofar as legal grounds of challenge can be extracted these seem to be as follows: the goals contained in FV2030 are described but their implications are not quantified or assessed; the efficacy and utility of the measures proposed by FV2030 is not assessed and it was impossible for the respondents to reach the conclusion that FV2030 would not adversely affect the integrity of any designated site in the absence of quantification and assessment of the measures involved. Overall, the focus of the appellant’s pleading is on the lack of adequate specificity as to the measures included in FV2030 to enable an appropriate assessment to take place. It is also pleaded that in circumstances where Ireland had not prepared site-specific conservation objectives (SSCOs) for all designated sites, it is not possible for the respondents to have concluded that FV2030 would not adversely affect those sites in light of their – nonexistent – conservation objectives. I understand that this latter issue is currently being pursued in other litigation and it did not form a major part of the argument on this appeal. In circumstances where the respondents will consent to an order of certiorari for other reasons if it is held that an AA of FV2030 was required, I do not think it is necessary to deal further with this issue.

109. The respondents’ reply to this is two-fold. Firstly, FV2030 did not require AA because it is not a plan or project under the Habitats Directive precisely because it does not contain specific measures (the absence of which lies at the core of the appellant’s case) but also because it does not relate to or impact on any specific sites. Alternatively, the assessment which was carried out at strategic level was appropriate in respect of a high-level strategy that does not propose specific or quantifiable measures. It was not possible to determine the likely effects of FV2030 on a site-specific level and, consequently, no obligation to do so arose under the Habitats Directive.

110. The duality of the respondents’ position is encapsulated in the plea at para. 95 of the third amended statement of grounds: - “Insofar as the Applicant alleges … that it was impossible to carry out an AA that meets the requirements of the Habitats Directive in the absence of FV2030 containing more specific measures, that allegation is denied. The Applicant cannot simultaneously assert that FV2030 requires assessment under the Habitats Directive

– despite the lack of specificity of the measures proposed and its status as a highlevel strategy document – and that the measures contemplated in FV2030 are not sufficiently specific so that a legal assessment of FV2030 under the Habitats Directive is impossible. Insofar as FV2030 is a ‘plan or project’ within the meaning of the Habitats Directive, which is denied, the obligation to assess FV2030under the Habitats Directive can go no further than the assessment that it is possible to carry out within the confines and having regard to the uncertainties of that plan.”

111. The High Court judgment commences its treatment of this issue by rejecting the interpretive argument made by the respondents to the effect that a “policy” does not fall within the concept of a “plan or project” under the Habitats Directive as policies are dealt with separately and elsewhere in the Directive. The same grounds are raised by the respondents in their Respondents Notice as part of a ground citing additional reasons why the High Court decision that FV2030 is not a plan or project should be upheld. I agree with the High Court that this element of the argument is based on an inappropriately literal approach in a context where purposive interpretation is paramount. There is nothing in the use of the concept “plan or project” in Art. 6(3) (which, as I have noted, is not defined) which would exclude a policy if that policy is capable of having a significant effect on any EU site. The scope of the concept and the question of whether a measure falls within its boundaries is determined by the effects the measure is capable of having and, consequently, the need to assess those effects to ensure that designated sites will not be adversely impacted.

112. However, other elements of the respondents’ plea under this heading are relevant to the issue which the court has to decide, namely whether FV2030 as a high-level policy document representing a statement of general political will only and with no mandatory or binding effects that could influence future plans or projects can itself constitute a plan or project. The High Court seems to have rejected this proposition on the basis that it had not been made out that FV2030 had no effect as distinct from no mandatory or binding effect. The judgment is unclear on this point but it is clear that the trial judge did not regard it as material in light of his other conclusions.

113. If this is what paragraph 71 of the High Court judgment is intended to mean, I am not convinced that it is correct. At an earlier point in the judgment it was held that FV2030 does not bind any party nor establish any binding (or non-binding) criteria for the granting of development consent or for the adoption of subsequent plans, programmes or strategies (para. 30). This is subject to the immediately following qualification that in planning law the Government decision broadly approving FV2030 may not be devoid of legal effects. This is presumably because of the requirement under s. 34(2)(a)(iv) PDA 2000 that planning

decision makers have regard to Government policy. However, the judgment acknowledges that any such effects are not in themselves binding (para. 31).

114. In those circumstances the statement that FV2030 is a policy document with no mandatory or binding effect seems to me to be fundamentally correct. The added status given to it by the Government decision approving its publication and expressing support for its broad policy thrust does not make it either mandatory or binding. That remains so notwithstanding the rejection of the respondents’ argument that FV2030 was not a policy falling within s. 34(2)(a)(iv) PDA 2000. It is of course useful to bear in mind that unlike Art. 2(a) of the SEA Directive, Art. 6(3) of the Habitats Directive does not depend on whether the measure in issue is adopted by a public authority. The only relevance of this distinction (i.e. between the stakeholder committee and the Minister/the Government) is that s. 34(2)(a)(iv) applies to policies adopted by the latter but not the former.

115. The real issue is not whether a policy in the abstract is capable of constituting a plan or project, because clearly an appropriate policy can do so, but whether this particular policy is a plan or project. The High Court held that it was not because Art. 6(3) was “not capable of being operated if the plan or project is of such generality that it does not relate to any site capable of analysis in terms of the site’s conservation objectives.” In support of this

conclusion the judgment cites two decisions of the CJEU.

116. The first of these is Inter-Environnement Wallonie ASBL and Bond Beter Leefmilieu Vlaanderen ASBL v. Conseil des Ministres Case C-411/17. That case concerned consents for the continued operation of existing nuclear reactors in the context of a 2003 law which prohibited construction and commissioning of new nuclear power stations and established a timetable for the phasing out of existing ones. The 2003 law was amended in 2013 to postpone the date on which certain reactors would cease the industrial production of electricity. In 2014 the Belgian Government decided that the production period of certain nuclear reactors would be extended by ten years subject to conditions which were set out in a Royal Order of 2015 and which required upgrading and safety work. The applicant environmental organisations sought the annulment of the 2015 Order and complained that no prior assessment had been carried out under, inter alia, the Habitats Directive. One of the questions referred by the Constitutional Court was whether Art. 6 of the Habitats Directive applied to the postponement of the deactivation date for the reactors and, if so, whether Art. 6(3) applied prior to the adoption of a legislative Act.

117. In answering this question, the CJEU held that the measures in question “with the work inextricably linked thereto” constituted a project under the Habitats Directive. In circumstances where the reactors predated the Habitats Directive but the 2015 Order changed the conditions under which their continued operations were to be carried out, an issue then arose as to when the prior assessment should take place. The CJEU initially pointed to existing case law on environmental impact assessment under which, in the case of multi-stage development consent procedures, the assessment must be carried out “as soon as it is possible to identify and assess all potential effects of the project on the environment” (para. 85). It held by analogy that an appropriate assessment “should, in principle, be carried out as soon as the effects which the project in question is likely to have on a protected site are sufficiently identifiable” (para. 143). It went on to hold that the fact that subsequent acts were required in order to proceed with the project did not justify the failure to conduct an AA of the effects of the project prior to the adoption of legislation which had the characteristics of an agreement by the authorities for the project in question. Of course, the projects in question, which were existing nuclear reactors, comprised operations in fixed, identifiable locations and no issue arose as to whether it was possible to identify the potential effects of continuing those operations under new conditions on designated sites in the vicinity.

118. The other case cited by the High Court on this issue is Terre Wallonne ASBL v. Région Wallonne Case C-321/18 which concerned a decree establishing conservation objectives for the Natura 2000 network adopted by a regional Government after a public inquiry. An environmental organisation sought the annulment of the decree on the grounds that it ought to have been subject to prior assessment under the SEA Directive and Habitats Directive. Whilst the questions referred related to the SEA Directive, the CJEU started its analysis by stating at para. 30 of the judgment: - “In the present case, it is clear from the order for reference that the Decree of 1 December 2016 is directly connected to the management of all sites in the Walloon region.

It therefore does not concern a particular site, for the purposes of Article 6(3) of the Habitats Directive and, accordingly, nor does it require an environmental assessment pursuant to Article 3(2)(b) of the SEA Directive.” The prima facie relevance of this was, of course, that the text of Art. 6(3) specifically excludes plans or projects which are directly connected with or necessary to the management of a site. I have to confess some difficulty in understanding the connection with Art 3(2)(b) of the SEA Directive which is apparently made in this paragraph.

Article 3(2)(b) requires an SEA to be conducted where a plan or programme has been determined to require AA “in view of the likely effect on sites”. It does not, in its terms, distinguish between a likely effect on all sites in a region and on particular sites. Nonetheless, the court clearly took the view that because of a combination of Art. 6(3) of the Habitats Directive and Art. 3(2)(b) of the SEA Directive, the measure in question did not need to be preceded by an environmental assessment. Of more direct relevance here, the court clearly distinguished between a measure applying generally to all Natura 2000 sites and one which applied only to particular sites.

In this regard the conclusion at para. 44 (albeit one directed at Art. 3 of the SEA Directive) usefully illustrates the court’s thinking: - “In the light of the foregoing, the answer to the questions referred is that Article 3(2) and (4) of the SEA Directive must be interpreted as meaning that a decree, such as that at issue in the main proceedings, by which a body of a Member State establishes, at regional level for its Natura 2000 network, conservation objectives which have an indicative value, whereas the conservation objectives at site level have a statutory value, is not one of the ‘plans and programmes’, within the meaning of that directive, for which an environmental impact assessment is mandatory.” This echoes the distinction in para. 30 between a decree which is connected to the management of all sites and a management plan in relation to particular sites and suggests that the further you go from an effect on a particular site and the greater the level of generality of the measure in question, the less likely it is to require an environmental assessment.

119. All of this led the High Court to conclude that “in the absence of any specifically identifiable site being capable of being addressed by reference to FV 2030…, the AA process is just not practicable or even possible” (para. 86) and “FV 2030 is just too general and contains nothing relevant to any identifiable site that would allow an AA to be carried out by reference to the site’s conservation objectives.” (Para. 87)

120. In contending that FV2030 is a plan or project for the purposes of the Habitats Directive the appellant relies on a number of CJEU decisions. First it relies on the opinion of Advocate General Fennelly in Commission v. France Case C-256/98 to the effect that “The term plan must…be interpreted extensively” because of the need to ensure a high level of protection for the EU sites intended to benefit from the regime under the Habitats Directive. The appellant also relied on the reasoning in the Dutch Nitrates Case (above) which held that although the term “project” in the Habitats Directive encompassed projects under the EIA Directive it was not limited by the definition of that term in the latter Directive so that it can, and does, encompass a broader range of projects. The appellant emphasised the possibility of FV2030 having indirect effects in forming the basis for more detailed plans or as a framework for development consents (see Commission v. UK (above)).

121. However, a large part of the appellant’s argument centred on a Commission Notice comprising methodological guidance on the provisions of Art. 6(3) and (4) of the Habitats Directive (2021/C 437/01). Although this document dates from October 2021, it was not opened to the High Court. Instead, reference was made to a slightly earlier guidance document on Art. 6 of the Directive adopted by the Commission in November 2018 and published in January 2019 entitled “Managing Natura 2000 Sites: The provisions of Article

6 of the “Habitats” Directive 92/43/EC”.

122. The 2018 guidance document contains at s. 4.4.2 consideration of sectoral plans, into which category FV2030 falls. It commences by saying that sectoral plans should be considered as covered by the scope of Art 6(3) “in so far as they are likely to have a significant effect on a Natura 2000 site”. It continues: - “However, a distinction needs to be made with ‘plans’ which are in the nature of policy statements, i.e. policy documents which show the general political will or intention of a ministry or lower authority. An example might be a general plan for sustainable development across Member States’ territory or region.

It does not seem appropriate to treat these as ‘plans’ for the purpose of Article 6(3), particularly if any initiatives deriving from such policy statements must pass through the intermediary of a land-use or sectoral plan (C179/06 paragraph 41). However, where the link between the content of such an initiative and likely significantly effects on a Natura 2000 site is clear and direct, Article 6(3) should be applied.” The more recent Commission Notice seems to move away from this approach to suggest that sectoral plans should be subjected to AA even where it is not possible to identify likely effects on individual Natura 2000 sites.

It indicates that the term “plan” should be understood more broadly and can include sectoral plans (para. 3.1). It points to the importance of strategic plans, including ones for sectoral activities, in considering the environmental consequences of new developments early at the strategic planning level which would make it possible to integrate environmental conditions and requirements at an early stage so as to avoid the risk of potential conflicts later at project level (para. 4.1). As regards AA of strategic plans it states: - “For strategic plans where it is not possible to identify effects on individual sites, the analysis should as a minimum focus on potential impacts and major risks; sitespecific effects will then need to be analysed at project level.

In such cases, the appropriate assessment should focus at least on determining the Natura 2000 sites that could be adversely affected as well as any EU protected habitats and species that could be affected (also outside Natura 2000), effects on connectivity, fragmentation and other effects at the network scale. This should serve to orientate the scope and focus of the assessment of individual projects. Where there is uncertainty about adverse effects on relevant features of Natura 2000 sites and their conservation objectives, it may be appropriate to carry out and record a risk assessment…”.

The guidance goes on to suggest that such an assessment might identify activities or elements of the plan that are certain to harm the integrity of Natura 2000 sites even if mitigation is carried out and that such activities or elements should be excluded from the plan at that point. It continues: “The assessment could additionally provide an overview of which other activities may be harmful to protected habitats and species and thus better focus the assessment at project level”.

123. It is possible to see the practical benefits of such an approach by analogy with the SEA Directive as the outcome of an environmental assessment conducted under that directive is not determinative of whether the plan or programme should be adopted but can provide useful environmental guidance to the downstream decision makers responsible for the grant or refusal of development consent for projects envisaged by the plan or programme. It is more difficult to see how this approach could work in relation to the AA of strategic plans under the Habitats Directive. Article 6(3) of that Directive is framed in a way such that the plan or project to be adopted can only be agreed if it is conclusively established that it will not adversely affect the integrity of any Natura 2000 site in light of its conservation objectives. Therefore, if AA is required, the guidance contained in the Commission Notice does not seem consistent with the legal obligation on a national authority not to agree a sectoral plan which encompasses elements which may harm protected habitats or species – and that the absence of such harm must be established to a high level of scientific certainty. This is the case even if it is intended that the actual potential for harm will be more fully assessed at project level. In my view this suggests that a sectoral plan which is primarily political in nature is not capable of being subject to AA, both because it does not propose sufficiently identifiable projects or types of project capable of having identifiable types of effect and because the site or sites that might be affected by its implementation are not sufficiently identifiable at the level of the sectoral plan.

124. The respondents argue that it is the legal nature and effect of a measure which determines whether it is a plan under the Habitats Directive (per Commission v. Italy Case C-179/06). The test, according to that case, is whether the plan is “more than at the stage of preliminary administrative reflection” and carries “a degree of precision in the planning in question which calls for an environmental assessment of their effects”. The respondents also point to the test for a valid AA as articulated in Sweetman v. An Bord Pleanála Case C- 258/11 and the impossibility of that test being satisfied when neither the measures to be adopted under the plan nor their likely effect on any designated site, much less the sites likely to be affected, can be identified at the stage of the plan. The court in Sweetman stated at paras. 40 and 44: - “40. Authorisation for a plan or project, as referred to in Article 6(3) of the Habitats Directive, may therefore be given only on condition that the competent authorities – once all aspects of the plan or project have been identified which can, by themselves or in combination with other plans or projects, affect the conservation objectives of the site concerned, and in the light of the best scientific knowledge in the field – are certain that the plan or project will not have lasting adverse effects on the integrity of that site. That is so where no reasonable scientific doubt remains as to the absence of such effects (see, to this effect, Case C-404/09 Commission v Spain, paragraph 99, and Solvay and Others, paragraph 67).

…

44. So far as concerns the assessment carried out under Article 6(3) of the Habitats Directive, it should be pointed out that it cannot have lacunae and must contain complete, precise and definitive findings and conclusions capable of removing all reasonable scientific doubt as to the effects of the works proposed on the protected site concerned.” The type of high-level risk assessment of a strategic plan suggested in the 2021 Commission Notice could not provide the level of scientific certainty as to the absence of any risk required by Sweetman. Equally, where the sites which might be affected by the downstream implementation of any element of the Strategy cannot be identified, it would seem impossible for an AA at this stage to contain complete, precise and definitive findings capable of removing all scientific doubt as to the effects on such site(s).

125. The alternative argument made by the respondents is illustrative of the problems which arise in attempting to conduct an AA of a high-level strategic plan. As previously noted, part of the appellant’s complaint regarding the AA which was conducted is that it did not identify or assess the likely effects of FV2030 on individual Natura 2000 sites. Instead, it looked generally at the potential environmental effects of agricultural activities encompassed within FV2030 on the habitat types and species for which Natura 2000 sites have been designated in Ireland, the pathways through which environmental effects might impact the sites and species and mitigation measures which might prevent an adverse impact. All of this was done at a very high level of generality. The respondents contend that if an AA was required, the AA of FV2030 which was carried out at a high, strategic level was “appropriate” in the context of the nature of the plan and would satisfy the 2021 Commission Notice. That document suggests that a site-specific analysis – the lack of which is part of the appellant’s central complaint under this heading – can be deferred to project level.

126. I am inclined to agree with the respondents on this issue. FV2030 is a high level, nationwide policy document which does not itself propose any particular development or any type of development for any particular area. It is not a land-use plan. It does not set the framework for the grant or refusal of development consent. It is not mandatory or binding on any downstream decision makers. It is simply not possible to identify the environmental effects of the policy in any meaningful way, much less to relate those effects to potential impacts on any identifiable European site. At its height all that can be said is that FV2030 relates to agriculture which activity has acknowledged and potentially adverse environmental impacts. Consequently, its implementation may entail as of yet unidentifiable development of an unknown size and scale at currently unknown locations which may have effects on unidentified Natura 2000 sites. In practical terms it would be impossible to conduct an appropriate assessment which could make complete, precise and definitive findings without any gaps or lacunae in respect of a plan of this generality.

127. Notwithstanding the practical difficulties in conducting an AA which would meet the requirements of this case law, if there is a legal obligation to conduct an AA for FV2030 then it must be sufficient for that AA to be conducted at a high, strategic level and without including a site-specific analysis of the potential effect on individual sites. This is for two reasons, first in practical terms it is the only way that such an assessment could be conducted and, second, it would appear to satisfy the guidance contained in the 2021 Commission Notice. That said, I have serious concerns as to whether an assessment of that nature can meet the requirements of Art. 6(3) which requires the decision maker to be satisfied as to the absence of any adverse effects before a plan or project is agreed to.

Conclusions on Appropriate Assessment:

128. In conclusion, whilst I am minded to agree with the trial judge that FV2030 is a plan of a type which, in practical terms, is incapable of being the subject of a meaningful AA and, thus, falls outside the scope of the concept of a “plan or project” under the Habitats Directive, the position is not clear cut. The 2021 Commission Notice seems to envisage that a highlevel strategic assessment that leaves the analysis of site-specific effects over to project level assessment would be “appropriate” for such strategic plans. This would not meet the requirements of a valid AA as identified by the CJEU in Sweetman v. An Bord Pleanála (above), but it may be that a different test would have to be applied in such cases. Of course, the reality may be that if a habitats assessment is to be required of such strategic policies, it would require a separate legislative basis in much the same way as the EIA Directive looks at projects and the SEA Directive looks at the plans and programmes which provide the frameworks within which development consent for such projects are granted.

129. Ultimately, I think this is an issue which can only be resolved at an EU level. I note that in its Notice of Appeal the appellant seeks a reference to the CJEU on the basis that the case raises important issues as to what constitutes a plan or project for the purposes of Art. 6(3) and whether the approach taken in the AA screening and Natura Impact Statement is a sufficient means of identifying potentially affected sites for the purposes of Art. 6(3) of the Habitats Directive. I agree with the thrust of the first part of this request. As regards the second part, the question may be not so much whether the approach taken by the respondents in this case was a sufficient means of identifying potentially affected sites for the purposes of the Directive but rather, if the Strategy is a plan or project within the meaning of Article 6(3) of the Habitats Directive, whether it is necessary to conduct a site-specific assessment in respect of a measure of this nature.

130. The respondents suggest that as the trial judge’s characterisation of FV2030, has not been appealed there is no need for a reference. To briefly recap, at paras. 29, 30 and 31 of the High Court judgment FV2030 is characterised as a voluntary stakeholder-led strategy with no statutory basis, mandatory or binding down-stream effects which does not authorise or limit any activities and does not establish binding criteria for the granting of development consent or the adoption of subsequent plans, programmes or strategies, Essentially, the respondents treat these factual findings as determinative of the question of whether FV2030 constitutes a plan or programme. At the same time, the respondents justified the carrying out of an AA of FV2030 on a voluntary basis because the scope of application of the Habitats Directive to such policy documents has not yet been definitively determined by either the national or European courts.

131. The respondents also told the court that they had “grave difficulty” with the suggestion made in the 2021 Commission Notice that an appropriate assessment should be carried out on strategic plans of this nature. Counsel is, of course, entirely correct to say that such guidance is not binding. Nonetheless, when the Commission issues formal guidance of this nature I do not think that it can be readily dismissed. There is clearly a potential conflict between the 2021 Commission Notice and the caselaw of the CJEU which suggests that to fall within Article 6(3) a plan must have identifiable effects on identifiable sites which are capable of being assessed. Therefore, in light of the uncertainty in the area, not least the differences between Commission’s guidance documents in 2018 and 2021, I think this is an appropriate case to make a reference to the CJEU.

132. The reference will be confined to the issues arising under the Habitats Directive and will be in the following terms: -

1. Is a high-level strategic policy such as FV2030 capable of constituting or being deemed a “plan or project” within the meaning of Article 6(3) of the Habitats Directive? In this regard the key characteristics of the policy are that it was adopted on a voluntary basis by a stakeholder committee; it is formally supported by a decision of the Government; it does not authorise or limit any activity; it does not have any mandatory or binding effects on downstream decision makers but may have some influence in that it may be considered by such decision makers; it does not set criteria by reference to which development consent decisions will be made; it does not set criteria by reference to which other plans or programmes will be adopted.

2. If the answer to question 1 is “Yes”, can a policy such as FV2030 constitute a plan or programme within the meaning of Article 6(3) of the Habitats Directive if it does not contain measures of adequate specificity to allow them to be quantified and assessed and it is not possible to identify any specific Natura 2000 site or sites which might be affected by its implementation so as to enable an assessment to be carried out by reference to the site(s) conservation measures?

3. If the answer to question 2 is “Yes”, is it sufficient for an appropriate assessment of such policy to identify generally the environmental risks posed by activities coming within the sector covered by the policy and the effects those activities might have generally on protected habitat types and species in Natura 2000 sites in Ireland without identifying specific risks that might impact upon individually identified sites?

133. In light of the above, this appeal will be dismissed insofar as it relates to the question of whether decisions made in the course of the AA and SEA processes carried out by the respondents preclude the respondents from contending that such assessments were not required and insofar as it relates to the SEA Directive. The balance of the appeal will be adjourned pending the decision of the CJEU on the questions to be referred. The question of the costs of this appeal will be dealt with at the conclusion of all issues.

134. My colleagues Whelan J. and O’Moore J. have read this judgment in advance of its delivery and have indicated their agreement with the contents thereof.