Opinion of the Advocate-General
I - Introduction
1 In the present infringement action under Article 169 of the EC Treaty, the Commission is seeking a declaration that the Kingdom of the Netherlands has not sufficiently complied with the obligation imposed on it by Article 4(1) of Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds (1) to designate special protection areas (hereinafter `SPAs') for endangered species of wild birds. The Netherlands contests the admissibility of the action and rejects the Commission's claims on their merits.
II - The relevant provisions of Community law
2 The general structure and objectives of the Directive are well known to the Court, (2) and I will only reproduce below those provisions which are directly relevant to the present proceedings.
3 After describing the background to the adoption of the Directive, and its general scope, the preamble notes that `the preservation, maintenance or restoration of a sufficient diversity and area of habitats is essential to the conservation of all species of birds ... [that] certain species of birds should be the subject of special conservation measures concerning their habitats in order to ensure their survival and reproduction in their area of distribution ... [and that] such measures must ... be coordinated with a view to setting up a coherent whole' (ninth recital).
4 Article 1 is complemented by Article 2 which reads as follows:
`Member States shall take the requisite measures to maintain the population of the species referred to in Article 1 at a level which corresponds in particular to ecological, scientific and cultural requirements, while taking account of economic and recreational requirements, or to adapt the population of these species to that level.'
5 The principal substantive provisions at issue here are Articles 3 and 4. Member States are required, by Article 3(1), to `take the requisite measures to preserve, maintain or re-establish a sufficient diversity and area of habitats for all the species of birds referred to in Article 1'; this obligation must be carried out `[in] the light of the requirements referred to in Article 2'. Article 3(2) specifies the primary means to attain the objectives of the preceding paragraph, including the `creation of protected areas' and the `upkeep and management in accordance with ecological needs of habitats inside and outside the protected zones'.
6 Article 4, the central provision in the present proceedings, merits citation in full:
`1. The species mentioned in Annex I shall be the subject of special conservation measures concerning their habitat in order to ensure their survival and reproduction in their area of distribution.
In this connection, account shall be taken of:
(a) species in danger of extinction; (b) species vulnerable to specific changes in their habitat;
(c) species considered rare because of small populations or restricted local distribution;
(d) other species requiring particular attention for reasons of the specific nature of their habitat.
Trends and variations in population levels shall be taken into account as a background for evaluations.
Member States shall classify in particular the most suitable territories in number and size as special protection areas for the conservation of these species, taking into account their protection requirements in the geographical sea and land area where this Directive applies.
2. Member States shall take similar measures for regularly occurring migratory species not listed in Annex I bearing in mind their need for protection in the geographical sea and land area where this Directive applies, as regards their breeding, moulting and wintering areas and staging posts along their migration routes. To this end, Member States shall pay particular attention to the protection of wetlands and particularly to wetlands of international importance.
3. Member States shall send the Commission all relevant information so that it may take appropriate initiatives with a view to the coordination necessary to ensure that the areas provided for in paragraphs 1 and 2 above form a coherent whole which meets the protection requirements of these species in the geographical sea and land area where this Directive applies.
4. In respect of the protection areas referred to in paragraphs 1 and 2 above, Member States shall take appropriate steps to avoid pollution or deterioration of habitats or any disturbances affecting the birds, in so far as these would be significant having regard to the objectives of this Article. Outside these protection areas, Member States shall also strive to avoid pollution or deterioration of habitats.'
7 Article 4(4) of the Directive has been amended by Article 7 of Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora, (3) though that amendment is not directly at issue in the present proceedings.
8 In accordance with Article 18, the Member States were obliged to `bring into force the laws, regulations and administrative provisions necessary to comply with this Directive within two years of its notification'. For the Netherlands, this period expired on 6 April 1981. (4)
III - Pre-litigation proceedings
9 On 25 September 1989, the Commission sent the Netherlands a letter of formal notice setting out three alleged infringements of the Treaty and the Directive; of these, only the first, regarding the Netherlands' alleged failure to classify a sufficient number of SPAs, has been maintained in the present proceedings. The Netherlands denied the alleged infringements in its reply of 29 December 1989.
10 A reasoned opinion was sent to the Netherlands on 14 June 1993, repeating the claim that it had not designated sufficient SPAs for the endangered species listed in Annex I to the Directive. This sets a deadline for compliance of two months from the date of notification of the opinion. The Netherlands claims to have replied to the reasoned opinion (letter of 1 December 1993, Annex 1 to the statement of defence); the Commission states that it never received any reply to the reasoned opinion. The present proceedings were initiated by an application registered at the Court on 5 January 1996.
IV - Admissibility
11 The Netherlands contests the admissibility of the action on four separate grounds.
(a) Failure to take account of the Netherlands' reply to the reasoned opinion
12 The Netherlands argues that, by not taking account of its reaction to the reasoned opinion, the Commission has failed to respect the rights of the defence, and that the action is therefore inadmissible in its entirety. The Commission contends that the only new element in the Netherlands' letter of 1 December 1993 is the indication that three new SPAs had been designated - including the Deurnese Peel which had been expressly mentioned in the reasoned opinion - and that it had taken account of the new situation of fact in its application. It further argues that the deadline fixed by the reasoned opinion serves to give the addressee Member State one last chance to comply with the Community rules, rather than to restate its views. The Netherlands retorts that the letter also set out legal arguments, of which the Commission has not taken account, in particular to justify the non-designation of certain individual sites, and that the Commission should at least have enquired of the Netherlands Government why, since it had requested two extensions of the deadline for replying to the reasoned opinion, it had not done so.
13 In order for the Netherlands to succeed on this point, it would need to show that Article 169 should be interpreted as requiring the Commission to take account of any reply a Member State may submit to a reasoned opinion. I do not consider that such a requirement may be read into this provision. It is only where the Member State concerned complies with the opinion within the period laid down by the Commission that the latter is precluded from commencing proceedings before the Court. (5) While it is true that the Court has established that the Commission is obliged to take account in its reasoned opinion of the observations of a defendant Member State on the letter of formal notice, (6) such an obligation is closely based on the text of Article 169 and does not assist the Netherlands in the present case. Equally, though the Netherlands has correctly identified the purpose of the pre-litigation procedure as being to `give the Member State concerned an opportunity, on the one hand, to comply with its obligations under Community law and, on the other hand, to avail itself of its right to defend itself against the objections raised by the Commission', (7) this does not in any way affect the admissibility of the present action. Indeed, if the Netherlands' view were correct, a Member State would in effect be able to prevent the Commission from bringing the matter before the Court by simply refusing to reply to the reasoned opinion. I am therefore of the view that the Netherlands' objection to the admissibility on this point should be rejected.
(b) Nature of the obligation under Article 4(1) of the Directive
14 The Netherlands' second argument concerning the admissibility of the present proceedings is that the breach alleged comprises not a single act or omission but rather a series of failures to make individual classification decisions. It contends that, as the Commission has not identified specific, reasoned breaches of the classification obligation under Article 4(1) of the Directive, it has been unable to answer these charges in replying to the letter of formal notice or reasoned opinion. The matters raised in this plea of admissibility, in my view, concern the proper interpretation of Article 4(1) of the Directive. As these arguments go to the substance of the Commission's complaint, they should therefore be examined along with the merits of the present action.
(c) New pleas in law
15 Thirdly, the defendant contends that the Commission's complaint regarding the insufficiency of the total area of SPAs, and their qualitative insufficiency, as well as specific complaints regarding the Friesian IJsselmeerkust and Hooge Platen on the Western Scheldt, were set out for the first time in the application, and that it was therefore unable to answer these at the pre-litigation stage of the proceedings.
16 In its letter of formal notice, the Commission referred expressly to the Netherlands' obligation to ensure that the number and dimension of classified areas in the Member States are in conformity with Article 4, and cited two examples of areas (the Markermeergebied and the Deurnese Peel) which should, in its view, be classified. These considerations were all repeated in the reasoned opinion. It is my view that, in so far as the application alleges a breach of the obligations imposed on the Netherlands by Article 4(1) of the Directive, because of its failure to classify a sufficient total area of SPAs, it is admissible. Given the general character of the complaint formulated in the application, which asks the Court to find a breach of the Directive and of Articles 5 and 189 of the EC Treaty on the sole ground that the Netherlands has not designated sufficient SPAs, I agree with the Commission that the references to the Friesian IJsselmeerkust and Hooge Platen are merely examples to illustrate the breach alleged, and that the Court is not invited to make specific findings in relation to either area.
17 However, in so far as the Commission's complaint concerns the financing the Netherlands has received in respect of these two areas under Council Regulation (EEC) No 1872/84 of 28 June 1984 on action by the Community relating to the Environment, (8) I am of the opinion that it is inadmissible, as neither of these areas is mentioned in either the letter of formal notice or the reasoned opinion. (9)
(d) Reliance on an ornithological study drawn up after the issue of the reasoned opinion
18 The final question of admissibility raised by the defendant concerns the Commission's reliance on a study listing the important areas for birds in the Netherlands, which was published in December 1994 (hereinafter `IBA94'), viz. some 18 months after the reasoned opinion had been sent to the defendant Member State. It contends that, as it was unable to comment on this list at the pre-litigation stage of these proceedings, IBA94 should be discounted in so far as the Commission relies upon it to show a breach of the Directive.
19 The Commission argues that its allegations are based on IBA89, and that its reliance on IBA94 was unnecessary. It expresses surprise that the Netherlands should object to its citing all the scientific evidence available, and particularly the most recent source whose scientific reliability has not been challenged in these proceedings.
20 In accordance with the established case-law of the Court, `the question whether a Member State has failed to fulfil its obligations must be determined by reference to the situation in the Member State as it stood at the end of the period laid down in the reasoned opinion, and the Court cannot take account of any subsequent changes'. (10) That period expired two months from the date on which the reasoned opinion was notified to the Netherlands, on 14 August 1993. In so far as it related to the situation in the Netherlands prior to that date, IBA94 would, in my view, be admissible as proof of the existence of the breach of the Directive alleged. However, the Commission has not sought to argue that IBA94, or any part thereof, relates to the earlier period. I am therefore of the opinion that the Commission may not rely on IBA94 to prove the breach alleged, as it reports on the situation in the Netherlands in a period subsequent to the period laid down in the reasoned opinion for compliance therewith.
21 The Commission has also argued that the SPAs classified by the Netherlands do not conform to the qualitative criteria set by the Directive. In particular, it alleges that the inclusion of fresh water lakes and marshes and moorland in such areas is insufficient. The only evidence put forward by the Commission in support of this specific aspect of its claim is taken from IBA94, and is hence, in my view, inadmissible.
V - Merits of the application
(a) The arguments of the parties
22 The Commission's contention is general in character, that is, that the Netherlands has not sufficiently complied with the obligation imposed on it by Article 4(1) of the Directive. In its view, this provision requires the Member States to designate sufficient SPAs to offer sufficient protection to all the species listed in Annex I. The fact that the population of certain of these species in a given Member State has diminished allows one to suppose that this obligation has not been properly fulfilled. The Commission relies essentially on two factors to support its view, each of which is contested by the defendant, supported by the Federal Republic of Germany.
(i) Number and total area of SPAs
23 A study completed in 1989 by the International Council of Bird Preservation (hereinafter `IBA89') identified 70 sites in the Netherlands, covering an area of 797 920 hectares, as qualifying for classification on ornithological grounds. IBA94, which is an updated version of IBA89 drawn up by a number of Netherlands organisations published in December 1994, identified 87 sites, covering 1 089 357 hectares, as suitable for classification as SPAs. A list drawn up by the Netherlands Ministry of Agriculture and Fisheries in 1991, the reliability of which is contested by the Commission, identified 53 suitable sites covering 398 180 hectares.
24 According to the Commission, the Netherlands has classified 23 SPAs with a total area of 327 602 hectares. In its view, this falls manifestly below the quantitative obligation arising from Article 4(1). The 23 SPAs classified cover 33 of the sites listed in IBA89, that is, less than half the 70 sites identified, while the area classified is also less than half that which arises from IBA89. Furthermore, as one SPA, the Waddenzee, alone covers some 250 000 hectares, the remaining SPAs cover only 77 602 hectares, which is inadequate to ensure sufficient protection for a large number of the species listed in Annex I. The extent of the Netherlands' failings in this regard appears even more striking in the light of IBA94; only 35 of the 87 sites, and less than one-third of the total area, suitable for classification have been so classified.
25 The Netherlands' main line of defence is that Article 4(1) of the Directive does not require it to classify a given number or total area of SPAs. In its view, the classification of SPAs is only one of the measures a Member State may take in order to comply with Article 4(1); a breach of this provision can only arise if a Member State has not adopted any special conservation measures. It is therefore the whole body of measures adopted in respect of a specific site which is decisive. The defendant provides a list of other conservation measures which are relevant in this regard, such as the 1967 Wet houdende voorziening in het belang van de natuurbescherming (Nature Conservation Law), (11) the purchase of sites by nature conservation organisations, nature management contracts with agricultural organisations, the classification of wetlands under the Convention on Wetlands of International Importance especially as Waterfowl Habitat (`the Ramsar Convention'), (12) and the Netherlands' bird conservation plans.
26 It concludes that 40% of the total area of the territories, and 40 out of the 87 sites (46%), listed in IBA94 benefit from nature conservation measures. Furthermore, in indicating only two individual sites which should have been classified, the Commission has not shown that the Netherlands has exceeded the margin of appreciation it enjoys under the Directive to choose `the most appropriate territories'; the Court has acknowledged that the Member States are better placed than the Commission to determine which of the Annex I species live on their territory. Nor has the Commission challenged the validity of the criteria on the basis of which the Netherlands selects SPAs; that each of the three lists mentioned is different shows that the application of such criteria can give different results and results which vary over time. Supported on this point by Germany, the defendant argues that the rule on which the Commission relied, to wit, that Member States must classify at least half of the suitable sites in their territory, does not appear in the Directive.
27 The Commission argues in reply that Article 4(1) creates a specific obligation to classify SPAs, which is not satisfied by the adoption of other measures. It also contends that the Netherlands has not shown that the measures on which it relies provide a sufficient level of protection for the species concerned.
(ii) Fall in the numbers of bird populations
28 As proof of the insufficiency of the standard of protection provided by the Netherlands, the Commission cites nine endangered bird species whose numbers have fallen by 50% between 1981 and 1990, and which are normally to be found in areas listed in IBA94 but which are not protected as SPAs. While expressly admitting that a fall in bird population numbers does not per se justify the conclusion that a Member State has failed in its obligations under Article 4(1), particularly as regards hibernating species, such a conclusion is justified in relation to sedentary species, such as the Black grouse (Tetrao tetrix) and the Bittern (Botauris stellaris). The Commission relies on Santoña Marshes as establishing that the obligation to protect endangered species pre-exists any fall in their numbers. (13)
29 The defendant argues that bird populations are in their nature subject to fluctuation, and cites eight species whose numbers have greatly increased, and one, the Great egret (Egretta alba), which has been sighted in the Netherlands for the first time. It further argues, as regards the species listed by the Commission, that four of these hibernate in the African Sahel swamps, and that the fall in the numbers may be due to the situation there; the populations of all the species mentioned fell in almost all European countries, and it is unfair to single out the Netherlands as being responsible. In any case, the classification of SPAs does not offer any guarantee against a fall in numbers, as illustrated by the case of the Bittern; though more than 10% of its population in the Netherlands is already to be found in SPAs, its total numbers fell considerably during the reference period. The population of five of the species listed, including the Black grouse, has stabilised in recent years.
(b) Analysis
(i) Interpretation of Article 4(1) of the Directive
30 The correct interpretation of Article 4(1) of the Directive must first be resolved. According to the Netherlands, the classification of SPAs is an important conservation measure, but is not rendered mandatory by this provision; a Member State could only be found to be in breach of Article 4(1) if it had not adopted any special conservation measures. It follows, under this view, that the mere finding that a Member State had classified less than half the territories in number and area would not suffice to establish that the Member State was in breach of its obligations under the Directive. The Netherlands describes a number of other conservation measures it has taken, and contends that these constitute compliance with the Directive.
31 The Commission's interpretation of Article 4(1) is radically different; in its view, this provision creates a specific obligation to establish SPAs in sufficient number, and of a sufficient total area, to ensure the survival and reproduction in their area of distribution of Annex I species. While it considers that the best way of complying with that obligation would be for each Member State to classify all the areas identified in IBA89 and IBA94, it accepts that the obligations imposed by the Directive do not go this far, and that the Member States enjoy a certain margin of discretion in this regard. However, the failure to classify even half, in number and in area, of the areas identified in the inventories of important bird areas in its view manifestly constitutes a breach of Article 4(1).
32 The rather extreme hypothesis propounded by the Netherlands does not appear to me to be justified either by the wording or objectives of the Directive, and is not supported by the Court's case-law in this area. The `preservation, maintenance and re-establishment of biotopes and habitats', including the creation of protected areas, is an obligation which applies in respect of all species of wild birds covered by the Directive, in accordance with Article 3(2)(a). The fourth subparagraph of Article 4(1) requires Member States to `classify in particular the most suitable territories in number and size as special protection areas for the conservation of [Annex I] species'. In my view, the words `in particular' show that this phrase should be interpreted as meaning that, amongst the measures the Member States are required to take to ensure the survival and reproduction of these endangered species, they must, as a minimum, classify the most suitable territories as SPAs. In order fully to comply with the more general obligation imposed by the first subparagraph of Article 4(1), they may be obliged to classify other territories as SPAs, and/or to adopt other special conservation measures. The central point, for this case, is that Member States are under a specific obligation to classify the most suitable territories as SPAs.
33 If the Netherlands' interpretation of its obligations under Article 4(1) were taken to its logical conclusion, a Member State could escape the obligation to classify SPAs, where it took the view that other special conservation measures were sufficient to ensure the survival and reproduction of the endangered species. Member States would thus be able to escape the obligations imposed upon them by Article 4(4) to take appropriate steps to avoid the deterioration of habitats or disturbances affecting the birds in protected areas. Such an interpretation would also render Article 4(3) nugatory, as there would be no special protection areas to form `a coherent whole'.
34 The Netherlands' interpretation also appears to me to be inconsistent with the case-law of the Court. In Santoña Marshes, Spain sought to argue that the classification of (part of) the relevant area as nature reserves was a sufficient fulfilment of its obligations under Article 4(1), and that in any case it had classified a large number of other SPAs on its territory, covering a larger area than in any other Member State. (14) In holding in Santoña Marshes that `the classification of [SPAs] is ... subject to certain ornithological criteria determined by the Directive, such as the presence of birds listed in Annex I, on the one hand, and the designation of the habitat as a wetland area, on the other', (15) the Court in my view clearly interpreted Article 4(1) as creating an autonomous obligation to establish SPAs, and, at the same time, indicated the conditions under which this obligation arises. More generally, in the course of the same judgment the Court held that `Articles 3 and 4 of the Directive require Member States to preserve, maintain and re-establish habitats as such, because of their ecological value', (16) illustrating the central place of habitats protection in the scheme of the Directive.
35 The interpretation of Article 4(1) proposed by the Netherlands also fails to take account of the specificity of the obligation to classify protection areas for Annex I species, and would in this regard apply to them the same regime as applies for other wild bird species (excluding migratory species) by virtue of Article 3. This contention was expressly rejected by the Court in RSPB, where it held that `Article 4 ... lays down a protective regime which is specifically targeted and reinforced both for the species listed in Annex I and for migratory species, an approach justified by the fact that they are, respectively, the most endangered species and the species constituting a common heritage of the Community'. (17)
36 In the light of the foregoing, I am of the view that the fourth subparagraph of Article 4(1) imposes on the Member States an autonomous obligation to classify as SPAs the most suitable territories, taking account of the protection requirements of Annex I species within the territory where the Directive applies. That obligation extends, in my opinion, to all of the `most suitable territories', though not necessarily all the sites which provide suitable living conditions for Annex I species; the Council did not either allow the Member States a discretion not to classify any sites identified as being among the most suitable, nor fix a minimum number of SPAs to be classified, as had been proposed for the Habitats Directive. (18) This seems to me to be consistent with the specificity of the regime for Annex I species adverted to above; as it appears from the ninth recital in the preamble, these are species whose very survival is in question. As we shall see below, the determination of which territories are the most suitable in number and size is to be effected by the Member States, on the basis of ornithological criteria.
37 The Netherlands has also argued that Article 4(1), rather than creating an obligation of a general character, requires it to take a series of discrete decisions on the classification of sites. As the Commission has not been able to demonstrate a breach of the obligation to classify in respect of any individual site, it concludes that the application is unfounded.
38 In my opinion, Article 4(1) creates both obligations of a general character, and specific obligations in relation to individual sites. In particular, the requirement that the most suitable territories `in number' be classified as SPAs can only be judged taking account of the general degree of a Member State's compliance with the fourth subparagraph of Article 4(1); for an individual site, numerical suitability is irrelevant. The English rendition of the criterion `most suitable ... in size' is slightly ambiguous, and appears to be inconsistent with the ninth recital in the preamble and with some of the other language versions. The French version, for example, reads `les plus appropriés ... en superficie', which corresponds to `most suitable ... in area', while the Dutch-language version, which reads `naar ... oppervlakte ... meest geschikte', has a similar connotation. (19) Interpreted as referring to area, this criterion can apply in respect of both general and specific assessments of compliance with Article 4(1). Similarly, the requirement that the Member States take account of trends and variations in population levels, and the protection levels of Annex I species throughout the area to which the Directive applies, also support the view that the Commission can proceed against a Member State for a general, as well as a specific, breach of the obligation to classify SPAs; the population trends or European protection levels are relevant for both types of obligation.
(ii) The Member States' margin of discretion
39 Much argument has been devoted to the scope of Member States' margin of discretion in selecting SPAs. The Netherlands argues that the application of Article 4(1) is based on a concrete appreciation of whether a particular site is amongst the most suitable territories, and points out that the previous cases dealt with by the Court have all concerned the question of whether a Member State should have classified an individual site as an SPA. In its view, no breach of this provision can be shown unless a Member State has exceeded the limits of its margin of discretion, for example, by not classifying as an SPA a site of particular ornithological importance.
40 In its intervention, Germany relies upon the margin of discretion to argue that the choice of SPAs is left by Article 4(1) to the Member States, and that the only determinative factor is that the areas must be, as regards their number and area, suitable for the conservation of the species concerned and for the establishment, along with those classified by the other Member States, of a coherent network of protection areas. In its view, this provision does not require that a particular number of SPAs be classified, but rather obliges Member States to ensure that the SPAs which are classified be appropriate for the conservation of endangered bird species.
41 Though the margin of discretion is not mentioned anywhere in the text of Article 4(1), the Court noted in Leybucht Dykes that `the Member States do have a certain discretion with regard to the choice of the territories which are most suitable for classification as special protection areas'. (20) In Commission v Italy, the Court had explained that `the management of the common heritage is entrusted to the Member States ...' and said:
`It is clear from [the] allocation of responsibilities [under the Directive] that it is for the Member State to identify the species which must be the subject of the special protective and conservation measures required by Article 4(1) of the directive. Moreover, the Member States are better placed than the Commission to ascertain which of the species listed in Annex I to the directive occur in their territory.' (21)
42 In the present case, there was no dispute as to the identification of the wild bird species which require protection on the territory of the Netherlands. As the German Government argued in Leybucht Dykes, the selection of an SPA entails an extremely complex assessment of the most varied facts and requires considerable scientific work. (22) In the present case, the Commission has recognised that Member States are not obliged to classify a separate SPA for each Annex I species. Some species require more protection than others, and the classification of a particular site as an SPA could provide protection for different endangered species at the same time. It appears to me that the discretion of the Member States operates in respect of the evaluation, according to objective ornithological criteria, of the suitability of potential SPAs; once a site has been identified as amongst the most suitable for the conservation of the species in question, its classification as an SPA is mandatory. This arises most clearly from RSPB, where the undisputed character of the Lappel Bank as being amongst `the most suitable territories' led the Court, in effect, to the conclusion that the United Kingdom was obliged to classify it. (23)
43 Whatever the scope of a Member State's discretion concerning the classification of an individual site, I do not see how it can assist the defendant in the present proceedings. The Commission is seeking to prove that the Netherlands has not classified a sufficient number and area of SPAs to comply with its general obligations under Article 4(1) of the Directive; the Netherlands is not arguing that it has a general discretion not to comply with these obligations.
44 Furthermore, the reliance put by the Netherlands on its margin of appreciation in the present case is inconsistent. On the one hand, it argues that the Member States are better placed than the Commission to identify sites which are deserving of protection; on the other hand, it contends that the Directive can only be enforced by the Commission's identifying particular sites which should be classified, and taking infringement proceedings against the Member States in respect of each of those sites individually. As the Commission has observed, apart from giving rise to considerable practical difficulties, the approach suggested by the Netherlands would respect the Member States' margin of discretion less than the approach it has adopted in the present proceedings.
45 The Netherlands adds that the Member States are obliged, when adopting special conservation measures, to take account of the economic and recreational requirements to which Article 2 refers. It modified this affirmation somewhat in its rejoinder, following the judgment of the Court in RSPB, (24) to argue that the scope of the obligations arising from Article 4(1) should be interpreted in the light of Articles 1 and 2 of the Directive. Such a contention is in my view in evident contradiction with the first paragraph of the operative part of the judgment in question, which states that `Article 4(1) or (2) of [the Directive] is to be interpreted as meaning that a Member State is not authorised to take account of the economic requirements mentioned in Article 2 thereof when designating a Special Protection Area and defining its boundaries'. (25)
(iii) The probative value of the inventories of important bird areas
46 The Directive does not establish either a list of the most suitable territories of the Member States which must be classified as SPAs, or detailed criteria for the selection of these sites. Article 4(1) does, none the less, provide a number of guidelines of which the Member States must take account in deciding which potential sites are most suitable. (26) As the Court noted in RSPB, `notwithstanding the divergences between the various language versions of the last subparagraph of Article 4(1), the criteria in question are ornithological criteria'. (27) It follows in my view that ornithological criteria must also be used in evaluating whether or not a Member State has sufficiently complied with its general obligation to classify SPAs.
47 In seeking to demonstrate that the Netherlands has failed sufficiently to implement its obligations in this regard, the Commission relies primarily on IBA89, though referring also to the modified and updated list IBA94. IBA89 is itself an updated version of an inventory drawn up at the Commission's behest by the Muséum National d'Histoire Naturelle, Paris, in 1987. In IBA89, `for the first time individual sites in each European country were evaluated in a standard way, and a continent-wide network of sites was identified that, if protected, would safeguard a significant proportion of the European populations of many species'. (28) The notes on the inventory for the Netherlands, set out in Annex 7 to the Commission's application, identify the following three categories of criteria on the basis of which sites were included in IBA89: numerical criteria, inclusion on the list of the 100 most important sites in the Community for a vulnerable species or subspecies, or inclusion amongst the five most important sites for a vulnerable species or subspecies in a given region of the Community. Five other sites were included in the inventory for other reasons, for example Het Zwin because it is contiguous with an important bird area in Belgium, or Krammer and Volkerak because this site `could, if managed appropriately, be developed into an important freshwater ecosystem'. The seven distinct categories of numerical criteria for breeding sites, and five categories for areas other than breeding areas, are set out in a table annexed to Annex 7 to the application; the former include sites supporting one per cent or more of the breeding pairs of the biogeographical population of a species or subspecies, (29) criteria based on the specific characteristics of dispersion and habitat preference of the species, all regular breeding sites of rare or endangered species or of small and endangered distinct biogeographical populations (2 500 pairs or less), and regular breeding sites for significant numbers of three or more Annex I species.
48 Germany has argued forcefully that IBA89 and IBA94 only contain lists of sites which in accordance with scientific criteria could potentially contribute to the conservation of endangered species; these lists are neither part of the Directive, nor legally binding. Moreover, neither the criteria upon which the lists are based, nor the resulting lists have been agreed upon at the Community level. It adds that the fixing of a lower limit of 50% of sites classified is arbitrary and cannot be scientifically verified.
49 This argument seems to confuse the legal obligation and the evidence required to prove non-compliance. It is, of course, true that IBA89 is not per se binding on the Member States; if it were, the present case could have been disposed of much more briefly. Though drawn up by the Eurogroup for the Conservation of Birds and Habitats in conjunction with the International Council of Bird Preservation (now `BirdLife') rather than through any exclusively Community procedure, IBA89 was prepared for the relevant Directorate-General of the Commission and in cooperation with Commission and national experts; the inventory was at least partly designed to assist the Member States in their implementation of the Directive. In identifying the vulnerable species and subspecies which are taken into account, for example, IBA89 refers explicitly to Annex I to the Directive, as amended by Directive 85/411/EEC, with the addition of those species and subspecies `likely to be added to Annex I to take account of Spain and Portugal's membership of the European Community'.
50 It follows, in my view, that IBA89 not only constitutes scientific evidence, the necessity for which Germany appears to accept in principle, but was expressly designed for use in the application of the Directive. It is not itself conclusive or constitutive of a legal obligation, but can be relied upon in demonstrating the extent of a Member State's compliance therewith, both as regards the general obligation and specific sites. (30) As regards an individual site, it is open to a Member State to produce better scientific evidence to show that it is not amongst the `most suitable' for the conservation of Annex I species. Similarly, it is open to a Member State to produce contrary evidence to prove that the total figures for SPAs, in number and in area, which arise from IBA89, or from any other such list upon which the Commission relies, are erroneous.
51 In the present case, the Netherlands does not directly contest the scientific viability of IBA89, except as regards the matter of the definition of a minimum viable area for an SPA. In its pleadings, it notes that IBA94 gives some indication of the definition of a biotope, which was missing from the earlier inventory. Indeed, it claims that the list drawn up by its Ministry of Agriculture and Fisheries in 1991 was based on the same three criteria as those on which IBA89 and IBA94 were founded. The Netherlands does, however, argue that the application of these criteria does not give unequivocal results, citing, on the one hand, the differences between IBA89 and IBA94, and, on the other hand, the differences between the IBA lists and the Agriculture Ministry list of 1991. It suggests that the difference between the list of sites which qualify for classification according to the IBA inventories, and those which have in fact been classified, can be explained by the nature of ornithological data. It also suggests that the difference in the number of SPAs results from a difference in the delimitation and regrouping of sites, while the difference in area is due to the absence of adequate criteria for defining the boundaries of sites for classification.
52 In the first place, the Netherlands has not demonstrated convincingly why a national list of sites to be designated, which was drawn up after the pre-litigation stage of the present proceedings had commenced, should be more reliable than an inventory drawn up by ornithological experts from different Member States, including the Netherlands, before the pre-litigation stage. In particular, the terms of the first of the three criteria as set out in the Netherlands' defence, which refers solely to the regular presence on a site of at least one per cent of the biogeographical population of species of water birds, seem to me to be much less comprehensive and detailed than the list of categories of numerical criteria on which IBA89 was based (Table 1 to Annex 7 to the Commission's application). Whether or not the restriction of this numerical criterion in the pleadings of the Netherlands to water birds is the result of a clerical error, it seems likely that the significant differences in the numerical criteria applied in IBA89 and the Netherlands' list of 1991 would itself be sufficient to explain the differences between the resulting lists.
53 While the Netherlands has not been able to demonstrate the objective superiority of its own national list, the Commission has questioned the scientific basis of the Netherlands' list. In an annex to its reply, the Commission has presented a table comparing the (theoretical) results of the application of the three criteria on which the Netherlands' list is based with the actual classification of SPAs for 26 Annex I species found in the Netherlands; in no case does the result achieved by the Netherlands in fact correspond with the figure which should have been achieved according to the Netherlands criteria, and in most cases the disparity is very significant. (31) The Netherlands has not explained the discrepancy between these two sets of figures.
54 The Netherlands' list of 1991 contains some 53 sites corresponding, according to the Commission, (32) to all or parts of 57 of the 70 sites identified in IBA89, though only covering approximately half the total area of sites listed as amongst the most suitable in IBA89. Even if this national list were shown to contain the most suitable territories, the defendant has not sought to demonstrate that it has classified these as SPAs, presumably largely because it contests the existence of an obligation to classify them.
55 The Netherlands has also sought to rely on the differences between IBA89 and IBA94 to show that the application of ornithological criteria gives uncertain results. The Commission has vigorously challenged this assertion. Seven of the 12 sites in the later list missing from IBA89 were included to take account of the addition of new species to Annex I to the Directive, while the other five are the result of various objective factors such as a different division or regrouping of sites, an increase in knowledge, or an evolution in the bird populations in the Netherlands. The Commission's explanations on this point seem to be convincing; moreover, the Netherlands has itself pointed out that the situation of bird species is in constant evolution over time.
56 It appears from the Commission's answer to a question from a Member of the European Parliament that `the Commission together with the Member States has developed a method which is an objective means of evaluating the endangered status of different bird species throughout the Community and determining the proportion of each bird population that should be within SPAs in each region'. (33) In reply to a question at the hearing, the agent of the Commission explained that, while the index of vulnerability was one factor to be taken into account in estimating the degree of protection each species required, it was of no assistance in identifying which sites should be classified as SPAs. Furthermore, the obligations arising from Article 4(1) cover all Annex I birds, including those with a lower vulnerability rating.
57 The latter part of the Commission's answer to the parliamentary question raises the issue of the feasibility of determining the proportion of particular species which should be within SPAs in the territory of a given Member State. In the annex to its reply in the written proceedings, cited above, the Commission table shows the percentage of all the Annex I species in the Netherlands which are found in the five most suitable sites, and the percentage of these species which are in SPAs. The figures are telling indeed; for example, six species whose percentage population in the five best sites ranges from 19% to 100% have none (0%) of their population in SPAs. However, while a Member State's failure to include a sufficient proportion of Annex I birds within SPAs in its territory might constitute a readily quantifiable index of its compliance with one aspect of Article 4(1), I am prepared to accept, in the absence of any evidence to the contrary, that it is not in itself a complete guide to compliance with the obligation to classify SPAs under this provision.
(iv) The existence of a breach of Article 4(1)
58 At the time of the Commission's application in the present case, the Netherlands had classified 23 SPAs, covering a total area of 327 602 hectares. As the equivalent figures which arise from IBA89 are 70 sites covering 797 920 hectares, the Commission is of the view that the Netherlands' failure to comply with its obligations to classify SPAs under Article 4(1) is manifest.
59 In seeking to show that this state of affairs constitutes a manifest breach of the Netherlands' obligations under the Directive, the Commission has repeatedly referred to the figure of a half of the number and total area of sites. The Netherlands and Germany have both pointed out that such a figure is not in the Directive, and is arbitrary and unscientific. However, this is to miss the point. The IBA reports, as I have said, are offered by the Commission as scientific evidence of what are the `most suitable territories' for the classification of SPAs in the Netherlands. The Netherlands has not significantly challenged their evidential value. Its failure to classify even fifty per cent of the proposed areas is advanced to enable the inference to be drawn that the Netherlands has failed in its general obligation to classify, while the Commission's reliance on such a figure is a matter of presentation rather than a definition of the obligation imposed by Article 4(1) of the Directive.
60 As I have already stated, the obligation which arises for the Member States by virtue of the fourth subparagraph of Article 4(1) is to classify all the most suitable territories, identified by the application of reliable objective scientific criteria. The Commission's duty in proceedings such as the present is to indicate the total number and total area of SPAs which a Member State should classify on the basis of these criteria, compared to the number and area of SPAs which it has in fact classified. The existence of any discrepancy between the two sets of figures can of course be challenged by the Member State in question; if proven before the Court, the discrepancy is sufficient to establish a breach of the defendant's obligations under the Directive. The Commission's conclusion on the basis of the evidence in the present case that the Netherlands has failed to classify sufficient SPAs in number and area results from the normal process of legal reasoning, and does not rely on any presumption.
61 At the hearing, the Netherlands argued that the Commission had rendered very difficult the task of identifying exactly what it should do to comply with Article 4(1) of the Directive, while Germany wondered how the Court would deal with a second, hypothetical, action against the Netherlands under Article 171 of the Treaty. It is clear from the preceding analysis of the relevant provisions that, in my view, the Netherlands' difficulties regarding compliance stem primarily from its erroneous interpretation of its obligations under Article 4(1), and that sufficient, and sufficiently reliable, ornithological data are available for the Netherlands to identify the action it is required to take in order properly to comply with this provision. I do not consider it either necessary or appropriate to deal in detail with arguments based on hypothetical future proceedings. As the declaration the Commission is seeking is general in character, I am of the opinion that such a declaration, if granted, could not be relied upon to show that the Netherlands was in breach of its obligation to classify a particular site; the declaration alone could not therefore be relied upon to justify Article 171 proceedings in respect of such a site. In any case, Germany's point was primarily directed at challenging the criterion of fifty per cent classification, which I have already dealt with above.
62 It follows, in my view, that the Commission should be granted the declaration which it has requested. The Commission has asked that the Netherlands be required to pay the costs of the present action. As the points on which I recommend that the Court uphold the views of the Netherlands are minor in character and do not affect the substance of the case, I consider that the Commission's request regarding costs should also be granted.
VI - Conclusion
In the light of the foregoing, I recommend to the Court that it:
(1) Declare that, by failing to classify a sufficient number and area of special protection areas in accordance with Article 4(1) of Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds, the Kingdom of the Netherlands has failed to comply with its obligations under the EC Treaty;
(2) Order the Kingdom of the Netherlands to pay the costs.
(1) - OJ 1979 L 103, p. 1, hereinafter `the Directive'.
(2) - A more detailed account is to be found in paragraphs 11 to 23 of my Opinion in Case C-44/95 Royal Society for the Protection of Birds (hereinafter `RSPB') [1996] ECR I-3805.
(3) - OJ 1992 L 206, p. 7, hereinafter the `Habitats Directive'.
(4) - Case 236/85 Commission v Netherlands [1987] ECR 3989, paragraph 2 of the judgment.
(5) - See, for example, Case C-362/90 Commission v Italy [1992] ECR I-2353.
(6) - Case C-266/94 Commission v Spain [1995] ECR I-1975.
(7) - Case C-473/93 Commission v Luxembourg [1996] ECR I-3207, paragraph 19 of the judgment.
(8) - OJ 1984 L 176, p. 1.
(9) - In any case, the Commission's allegation would seem to concern a failure to comply with the terms of its decision of 27 May 1987, rather than any breach of the Directive.
(10) - See, for example, Case C-302/95 Commission v Italy [1996] ECR I-6765, paragraph 13 of the judgment.
(11) - Stb. 572, 1967.
(12) - United Nations Treaty Series Volume 996, p. 245; see also Commission Recommendation 75/66/EEC of 20 December 1974 to Member States concerning the protection of birds and their habitats (OJ 1975 L 21, p. 24).
(13) - Case C-355/90 Commission v Spain [1993] ECR I-4221, `Santoña Marshes', paragraph 15 of the judgment.
(14) - Case C-355/90, cited in footnote 13 above, Opinion of Advocate General Van Gerven, paragraph 14 (p. I-4249).
(15) - Ibid., paragraph 26 of the judgment.
(16) - Paragraph 15.
(17) - Case C-44/95, cited in footnote 2 above, paragraph 23 of the judgment.
(18) - Case C-57/89 Commission v Germany (hereinafter `Leybucht Dykes'), Opinion of Advocate General Van Gerven, paragraph 26, footnote 24, [1991] ECR I-883, at p. I-914.
(19) - The Danish (`til ... udstrækning er bedst egnede'), German (`die ... flächenmäßig geeignetsten'), Italian (`i ... più idonei ... in superficie'), Greek (`ôá ðéï êáôÜëëçëá, óå ... åðéöÜíåéá'), Spanish (`los ... más adecuados ... en superficie'), Portuguese (`os ... mais apropriados ... em extensão'), Finnish (`kooltaan sopivimmat') and Swedish (`storlek är mest lämpade') ... all either refer to `area' or use a term which can mean `area' or `size'.
(20) - Case C-57/89 Commission v Germany, cited in footnote 18 above, paragraph 20 of the judgment.
(21) - Case C-334/89 Commission v Italy [1991] ECR I-93, paragraphs 8 and 9 of the judgment.
(22) - Case C-57/89 Commission v Germany, cited in footnote 18 above, Report for the Hearing, pp. I-896 and I-897.
(23) - Case C-44/95, cited in footnote 2 above, paragraph 26 of the judgment. See also Case C-72/95 Kraaijeveld and Others, where the Court adopted a similar approach to Member States' discretion under Article 4(2) of Council Directive 85/337/EEC of 27 June 1985 on the assessment of the effects of certain public and private projects on the environment (OJ 1985 L 175, p. 40); an analogy between these two types of discretion had been suggested by the Netherlands Raad van State, the referring court ([1996] ECR I-5403), paragraphs 44, 49 and 50 of the judgment.
(24) - Case C-44/95, cited in footnote 2 above.
(25) - Ibid., [1996] ECR I-3805, at pp. I-3856 and I-3857.
(26) - See paragraph 6 of the present Opinion, above.
(27) - Case C-44/95, cited in footnote 2 above, paragraph 26 of the judgment.
(28) - Tucker et al., Birds in Europe: Their conservation status (cited by the Commission in Annex 7 of its application), BirdLife International, Cambridge, 1994, p. 20.
(29) - The Commission expert explained at the hearing that this phrase referred to distinct fly-away populations of bird species from their breeding grounds to their staging and wintering areas, which may include areas outside the territory to which the Directive applies. Measures which favour the protection of one such population presumably have no effect on the other such populations.
(30) - The Commission declared at the hearing that it had relied on IBA89 in Santoña Marshes, though this does not appear from the case report.
(31) - The Commission does not specify the source of the population figures; the table is not, however, relied upon specifically to show that the Netherlands is in breach of Article 4(1), but to challenge the reliance by the Netherlands on its own 1991 list.
(32) - Summary of Annex 9 to the application.
(33) - Written question No 131/93 by Mr Florus Wijsenbeek, OJ 1993 C 258, p. 7.
Full & Egal Universal Law Academy