Court of Justice of the European Communities (including Court of First Instance Decisions)

You are here: BAILII >> Databases >> Court of Justice of the European Communities (including Court of First Instance Decisions) >> Commission v Spain (Environment and consumers) [2006] EUECJ C-235/04 (14 September 2006)
URL: http://www.bailii.org/eu/cases/EUECJ/2006/C23504.html
Cite as: [2006] EUECJ C-235/4, [2006] EUECJ C-235/04

[New search] [Help]

IMPORTANT LEGAL NOTICE - The source of this judgment is the web site of the Court of Justice of the European Communities. The information in this database has been provided free of charge and is subject to a Court of Justice of the European Communities disclaimer and a copyright notice. This electronic version is not authentic and is subject to amendment.

OPINION OF ADVOCATE GENERAL

KOKOTT

delivered on 14 September 2006 (1)

Case C-235/04

Commission of the European Communities

Kingdom of Spain

(Conservation of wild birds - Special protection areas - Autonomous communities in Spain - IBA 98)

I - Introduction

1. In this case, the Commission is once again bringing proceedings against a Member State alleging the inadequate classification of special protection areas for birds ("SPAs") in accordance with Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds (2) ("the Birds Directive"). It has already obtained judgment in respect of similar infringements against the Netherlands, (3) France, (4) Finland (5) and Italy. (6) Proceedings are also pending against Greece (7) and Ireland. (8) The Commission is preparing another case against Portugal. (9)

2. The central issue in each of these cases is that of establishing that the relevant Member State has not yet classified as SPAs all areas requiring designation as such. In the present case, the Commission bases its claim on a list of ornithologically important areas in Spain which was published by the Spanish Ornithological Society (Sociedad Española de Ornitología, "SEO/BirdLife") in 1998 ("the IBA 98"; IBA(s) being the abbreviation for "Important Bird Area" or "Important Bird Areas). (10) Spain questions the quality of that inventory.

II - Legal background

3. Article 4(1) and (2) of the Birds Directive determines which areas Member States are to classify as SPAs, while Article 4(3) governs the information on classification to be sent to the Commission:

"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;

(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. The ninth recital in the preamble to the Birds Directive explains this rule:

"... the preservation, maintenance or restoration of a sufficient diversity and area of habitats is essential to the conservation of all species of birds; ... 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; ... such measures must also take account of migratory species and be coordinated with a view to setting up a coherent whole".

5. Article 10 of the Birds Directive provides that the Member States are to encourage ornithological research:

"1. Member States shall encourage research and any work required as a basis for the protection, management and use of the population of all species of bird referred to in Article 1.

2. Particular attention shall be paid to research and work on the subjects listed in Annex V. Member States shall send the Commission any information required to enable it to take appropriate measures for the coordination of the research and work referred to in this Article."

6. Annex V refers to individual areas of research to which particular attention is to be paid.

7. Under the second subparagraph of Article 3(1) of Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora (11) ("the Habitats Directive"), the Natura 2000 network set up by that directive is also to include the SPAs classified by the Member States pursuant to the Birds Directive.

III - Pre-litigation procedure and forms of order sought

8. On 26 January 2000, the Commission invited the Spanish Government to submit observations pursuant to Article 226 EC (letter of formal notice). The Commission complained that Spain had classified too few areas as SPAs in accordance with Article 4 of the Birds Directive. According to the information provided by the Commission, at that time Spain had classified as SPAs 175 areas covering 33 582 square kilometres.

9. The Commission relied on IBA 98 as evidence of the inadequacy of areas classified as SPAs. That inventory lists 391 sites covering 15 862 567 hectares, or 31.5% of Spain's land area.

10. During 2000, Spain rejected the Commission's complaint but at the same time classified further areas as SPAs.

11. Accordingly, the reasoned opinion of 31 January 2001 referred to 262 SPAs in Spain, covering an area of 53 674 square kilometres. In that reasoned opinion, the Commission gave Spain a final period of two months within which to carry out the additional classifications requested. At the request of the Spanish Government, the Commission extended this period to 3 May 2001. In March 2001 Spain notified to the Commission a further 13 SPAs covering 402 272 hectares.

12. In the following years, Spain increased, in several stages, the number of SPAs to 427, covering around 79 778 square kilometres, or 15.8% of Spain's land area.

13. The Commission did not consider this progress sufficient and it therefore brought an action on 4 June 2004.

14. The Commission claims that the Court should:

1. declare that, by failing to classify areas of a sufficient number and size as special protection areas for birds in order to provide protection for all the species of birds listed in Annex I to Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds and for the migratory species not referred to in Annex I, the Kingdom of Spain has failed to fulfil its obligations under Article 4(1) and (2) of that directive;

2. order the Kingdom of Spain to pay the costs.

15. The Kingdom of Spain contends that the Court should:

1. dismiss the action;

2. order the applicant institution to pay the costs.

16. Although the forms of order sought in the application relate to Spain as a whole, the pleadings are limited to the Autonomous Communities of Andalusia, the Balearics, Extremadura, the Canaries, Castilla-La Mancha, Catalonia, Galicia and Valencia.

17. Since the action was brought, the number of Spanish SPAs has risen to 512. They cover around 91 803 square kilometres. The land-based SPAs are equivalent to 18.2% of Spain's land area. In addition, 20 SPAs contain marine sections covering 574 square kilometres. (12)

18. At the hearing, the Commission withdrew its application in respect of the Autonomous Community of Extremadura.

IV - Appraisal

19. The Commission complains that Spain has classified too few areas as SPAs. However, the subject-matter of the action is limited to seven of Spain's autonomous communities. (13)

A - Legal bases for the obligation to classify

20. The legal bases for the obligation to classify are not disputed between the parties.

21. Pursuant to the fourth subparagraph of Article 4(1) of the Birds Directive, Member States are to classify the most suitable territories in number and size as SPAs for the conservation of the species mentioned in Annex I, taking into account their protection requirements in the geographical sea and land area where the directive applies. It is not possible to avoid this obligation by adopting other special conservation measures. (14)

22. Under Article 4(2), Member States are to take similar measures for regularly occurring migratory species not listed in Annex I as regards their breeding, moulting and wintering areas and staging posts along their migration routes. To that end, Member States are to pay particular attention to the protection of wetlands and particularly to wetlands of international importance.

23. It is clear from Article 4(3) of the Birds Directive and the ninth recital in the preamble thereto, that the SPAs are intended to form a coherent whole, which meets the protection requirements of those species in the geographical sea and land area where the directive applies.

24. It is settled case-law that although Member States do have a certain margin of discretion with regard to the choice of SPAs, a decision on the classification and delimitation of those areas must nevertheless be made solely on the basis of the ornithological criteria determined by the directive. Other considerations, particularly those of an economic or social nature, may play no role in the classification of the area. (15)

25. The Court has already ruled that Spain should have fulfilled the obligation to classify areas as SPAs in full by the date of its accession, that is to say, 1 January 1986. (16) On the other hand, in the present case the material date for the assessment of the infringement is, as is well known, the date of expiry of the period which the Commission laid down in the reasoned opinion. (17) Since the Commission retrospectively extended that period to 3 May 2003, it is necessary to examine whether Spain had classified sufficient areas as SPAs by that date.

B - Partial acceptance of the complaint

26. It is true that Spain vigorously disputes the complaint, in particular as regards the relevance of the IBA 98 inventory to the assessment of the obligation to classify. However, since 3 May 2003 it has classified further areas as SPAs, or extended existing SPAs, in Andalusia, the Balearics, Castilla-La Mancha, Catalonia, Galicia and Valencia. In respect of the Canaries, Spain has notified no new classifications since that date, but in its pleadings it accepts, as a minimum, that a further 16 areas should be classified as SPAs and that 11 existing SPAs should be extended in respect of that archipelago. Further areas are also to be classified as SPAs and SPAs are to be extended in Andalusia, Catalonia, Galicia and Valencia.

27. In some earlier cases, the Court has relied on similar circumstances to rule that insufficient areas have been classified as SPAs by the material time. (18) By classifying an area as an SPA, a Member State acknowledges that that area is one of the most suitable areas for the conservation of birds. (19) Consequently, the view could be taken that by classifying further areas as bird conservation areas after the expiry of the period laid down in the reasoned opinion Spain has acknowledged that it is obliged to make such a classification.

28. However, in all those cases there were additional circumstances which made the failure appear particularly serious. In the case of France, no areas had been classified as SPAs in respect of six species listed in Annex I; (20) in the case of Finland, a total of only 15 areas had been classified as SPAs, which, it was not disputed, was inadequate; (21) and in the case of Italy, that Member State conceded that economic and recreational grounds had been taken into account in the selection of areas. (22) Furthermore, it was common ground that the Italian classifications fell far short of the IBA 89. (23) Therefore, to rule against Spain on the basis alone that areas had been classified as SPAs in the meantime would, by comparison, be weakly supported.

29. However, above all, a judgment on this basis would not be justified having regard to the continuing dispute between the parties. The action has not lost its point by reason of the classifications and notifications made in the meantime. Spain denies that it is obliged to classify further areas on the basis of the IBA 98, whilst the Commission is clearly still not satisfied with the concessions made thus far. The dispute over the extent to which further areas must be classified would, in the event of a ruling against Spain on the basis of those classifications and notifications, be carried over unresolved into the area covered by Article 228 EC and could again come before the Court in that context. (24)

30. The burden of this uncertainty would essentially rest on Spain, since it would remain uncertain as to the scope of its obligations under a judgment that might be delivered but would have to fear being ordered to pay a periodic penalty payment and/or a fixed amount in a further case. That disadvantage would be disproportionate, unjustified and unwarranted in particular because a ruling against Spain on the basis of classifications or classification notifications which it had made in the meantime would be based solely on the fact that it had made efforts to fulfil its obligations during the infringement proceedings.

31. It is therefore necessary also to resolve the remainder of the dispute between the parties. (25)

C - The remainder of the dispute between the parties

32. In support of its complaint, the Commission relies principally on the fact that the Spanish classifications in the autonomous communities referred to in the application do not cover large parts of the areas referred to in the IBA 98 and also on the fact that Spain has not classified all the wetlands which it had recognised as falling within the scope of the Ramsar (Iran) Convention on Wetlands of International Importance. (26) It also cites the inadequate coverage of species requiring special protection which come under Annex I to the Birds Directive.

33. The Commission repeatedly cites the inadequate protection of species requiring special protection but in almost all cases fails to state any express basis for doing so. In any event, the lists of the species concerned, which are to be found in respect of all the autonomous communities, and the partial indication of the habitats concerned (steppe, wetlands or mountains) are not sufficient conclusively to demonstrate inadequate protection. The only, implicit, grounds for this complaint are to be found in the IBA 98, which refers to as yet unclassified territories which are most suitable for the conservation of these species. Consequently, this complaint contains no new argument in comparison with the reference to the IBA 98, but merely helps to demonstrate the evidential value of that inventory as regards the inadequacy of the areas classified. Consequently, it is not necessary to assess this argument separately.

34. As regards the "Ramsar areas", the Commission refers in the application to two areas in Andalusia and one in Galicia which had still not been classified as SPAs at the material time. Since Spain does not contradict this claim, this point is conceded.

35. Consequently, the examination can be limited to the evidential value of the IBA 98.

36. The Commission submits that of all the available reference documents the IBA 98 is the best documented and most accurate in terms of determining the areas most suitable for the conservation, and in particular the survival and breeding, of important species. That inventory is based on balanced ornithological criteria which make it possible to indicate which sites are most appropriate to ensure the conservation of all the species listed in Annex I and other migratory species and to classify the bird protection areas in Spain which require classification as a matter of priority.

37. A comparison between the IBA 98 data and the SPAs designated by the Kingdom of Spain demonstrates, both as regards the territory of Spain as a whole and in the case of an examination broken down by the individual autonomous communities, that the territories classified as SPAs are fewer in number and are smaller in size than the areas which, on the basis of scientific evidence, are most suitable for providing appropriate protection for the birds covered by Article 4 of the directive.

38. Spain objects in principle to the reliance on the IBA 98 on the ground that the designation of a network of SPAs must be based on information available to the competent authorities. The IBA 98, however, cannot constitute such a reference.

39. This view is only partially correct. The Member States bear sole responsibility for the classification of SPAs. They cannot relinquish their responsibility by simply adopting and implementing the findings of other bodies, including those of organisations for the protection of birds. Rather, for an area to be classified it must number among the most suitable areas for the protection of birds, as viewed by the competent authorities on the basis of the best available scientific facts. (27)

40. However, it does not follow that the obligation to classify does not apply in general where the competent authorities have failed fully to examine and verify new scientific findings. Rather, it should be recalled that the obligation to classify has existed since the expiry of the period for transposing the Birds Directive, that is to say, since 1 January 1986 in the case of Spain. (28) Moreover, the obligation to classify is not limited by the state of scientific knowledge at any given time. (29)

41. That obligation included a further requirement, namely to identify the most suitable areas. Thus, Article 10 of the Birds Directive, in conjunction with Annex V thereto, requires the Member States to support the necessary research and work. Consequently, by 1986 Spain ought itself to have carried out a comprehensive scientific survey of the presence of birds in its territory and classified the resulting areas as SPAs. Had it fulfilled that obligation in full, either the IBA 98 would contain only SPAs or Spain would be able easily to reject any further calls for the classification of areas as SPAs. Further requirements to classify can arise only if the presence of birds alters. In the present case, no party has claimed that this has occurred.

42. In the light of the foregoing, the relevance of the IBA 98 inventory to the present dispute is clear. It does not - even in the view of the Commission - describe definitively the network of SPAs to be designated. The Commission does not expect Spain to classify each one of the territories listed in the inventory in their entirety as SPAs. Accordingly, the Commission has abandoned the complaint concerning inadequate classifications in relation to some autonomous communities, even though their classifications fall short of the IBA 98, (30) and has also accepted the scientifically based boundaries of SPAs within the IBAs even though in some cases they fall, in terms of size, well short of the IBA 98 data. (31) Rather, the IBA 98 serves merely as an indication that the classifications to date fall substantially short of the requirements of Article 4 of the Birds Directive.

43. In principle it is possible to use an inventory of areas in this way. In relation to the earlier IBA 89 inventory, the Court has held, in view of its scientific value, and of the absence of any scientific evidence to show that the obligations flowing from Article 4(1) and (2) of the Birds Directive could be satisfied by classifying as SPAs sites other than those appearing in that inventory and covering a smaller total area, that that inventory, although not legally binding on the Member State concerned, could be used by the Court as a basis of reference for assessing whether that Member State has classified a sufficient number and area of sites as SPAs for the purposes of the abovementioned provisions of the Directive. (32)

44. The Commission no longer relies on this older inventory, which the Court has recognised, but on the IBA 98. It takes the view that the IBA 98 now provides the best available scientific information on the areas most suitable for the conservation of birds.

45. Spain could refute that information by providing better scientific data showing that the areas classified as SPAs fully satisfy the obligations flowing from Article 4 of the Birds Directive. (33) It would appear that the autonomous communities which no longer form the subject-matter of the present case have furnished such evidence to the satisfaction of the Commission. Furthermore, Spain submits in relation to various other autonomous communities that relevant information is available or being compiled. However, in general, (34) this submission is not substantiated to such an extent that the Court is able to examine whether the IBA 98 is rebutted in that regard.

46. As regards the present case, Spain instead pursues a different strategy. It seeks to undermine the scientific value of the IBA 98 inventory at the outset, in order to prevent it from being used as evidence of inadequate classification. To that end, it puts forward a number of fundamental criticisms, contests the quality of the data used and finally objects to some of the criteria for determining the areas most suitable for the conservation of birds.

1. Fundamental criticism of the IBA 98

47. The Spanish Government firstly expresses some fundamental criticisms of the IBA 98. It claims that the inventory is of an entirely different quality than the IBA 89. The IBA 89 was drawn up for the Commission by the International Council for Bird Preservation, whilst the IBA 98 was drawn up solely by SEO/BirdLife, a Spanish organisation for the protection of birds. BirdLife International, the successor organisation of the International Council for Bird Preservation, expressly declined any responsibility for the content of the IBA 98.

48. It is inexplicable how, only nine years after the IBA 89, the IBA 98 could cover 16 million instead of 9.5 million hectares. In the view of the Commission and the Court, the IBA 89 was already stringent, precise and exhaustive in this respect. Apparently, after the judgment against the Netherlands (35) SEO/BirdLife intended unilaterally to increase the number and size of important bird areas in order to create arguments on which infringement proceedings could be based. The IBA 98 was published at the beginning of 1999 and the present infringement proceedings were initiated just one year later.

49. However, as the Commission correctly points out, the IBA 98 was drawn up by the same organisation that had produced the Spanish part of the IBA 89, namely SEO/BirdLife. Although the IBA 98 still contains the proviso that BirdLife International, the international umbrella organisation for organisations for the protection of birds, declines any responsibility, that organisation has since incorporated the IBA 98 into the IBA 2000, the European inventory of important areas for the protection of birds, for which it is responsible. (36)

50. SEO/BirdLife is a recognised authority on ornithological matters relating to Spain. Spain expressly concedes this. The instructions to produce expert reports, listed by the Commission, which the Spanish authorities have made to this organisation, the involvement of national Spanish authorities in the IBA 98, (37) and the additional classifications made on the basis of the IBA 98 in some of Spain's autonomous communities confirm SEO/BirdLife's authority in scientific matters. Finally, the Court also recently relied expressly on an SEO/BirdLife report on the capture of thrushes. (38)

51. The new inventory is very similar to its predecessor as regards its methodology and the involvement of the Commission. Like its predecessor, the IBA 89, the IBA 98 inventory results from the application of a series of ornithological criteria to identify the most suitable areas to carry out a survey of the presence of birds.

52. As regards the criteria for the selection of areas, the IBA 89 and IBA 98 are largely at one. (39) The Commission's involvement in the IBA 89, emphasised by Spain, was essentially limited to monitoring the ornithologists" work on the criteria. Since the criteria largely continue to be applied, the Commission is, at least indirectly, also responsible for the IBA 98. On the other hand, the Commission was hardly able to monitor the collection of data in the case of the IBA 89, since it could not verify the existence and extent of each individual bird presence that was referred to. Consequently, in this regard also there is no significant difference between the IBA 89 and the IBA 98.

53. Furthermore, the Spanish Government's fundamental objections to the IBA 98 are mere suppositions - almost suspicions - and therefore cannot, seen in isolation, undermine the authority of the inventory. It may be that increasing scientific knowledge of the presence of birds will make it possible to identify further areas which are most suitable for the conservation of birds. For an organisation devoted to bird conservation such as SEO/BirdLife, it is also a sensible strategy to seek to update and complete the inventory of areas of importance now that the Court has recognised its value as evidence that insufficient areas are classified as SPAs. A similar strategy has also been pursued by BirdLife International in collaboration with its other partners in the countries of Europe, as a new European inventory was published in 2000. (40)

54. Finally, Spain's objection that it has already made a disproportionately large contribution towards the areas classified as SPAs in the Community cannot be accepted. It is true that the proportion of land area covered by SPAs was highest in Spain when the action was brought. Spain accounted at that time for 35% of the areas in the Community classified as SPAs but had only 16% of the Community's land area.

55. However, this argument has no legal weight. Article 4 of the Birds Directive does not require that each Member State classify areas as SPAs in accordance with its size, but rather that the Member States classify the most suitable areas. The Member States have a different proportion of such areas, depending on their geographical and biological situation. Thus, Slovenia and Slovakia have since classified substantially larger proportions of their land area as SPAs. (41) If a Member State were to make a disproportionately large contribution towards the establishment of the Natura 2000 network by classifying areas as SPAs, the Community would be required, in accordance with the principle of cooperation between Community institutions and Member States, to take accordingly greater account of that Member State as regards Community support for the network. (42)

56. Therefore, Spain's fundamental objections to the IBA 98 cannot undermine the scientific quality of the inventory. Logically speaking, the more recent inventory must, by reason of its origin, be deemed to be of a scientific quality that is equivalent to that of its predecessor.

2. The data used

57. Further Spanish objections relate to the substantive quality of the IBA 98. They require an examination of the data collection and the resulting quality of the IBA 98 data.

58. Spain objects that the inventory contains no sources in respect of the assessment of each of the areas listed. Therefore, it is not possible to verify the underlying information and the extent of the areas. In that respect, the Spanish Government fails to appreciate that scientific surveys onsite, in the individual IBAs, can verify at any time the presence of birds and the areas used. Moreover, the situation regarding sources was no better as regards the earlier IBA 89 inventory. That did not prevent the Court from giving recognition to that inventory.

59. Spain also objects that SEO/BirdLife refused to grant the autonomous communities, which have sole responsibility for the classification of areas as SPAs, unlimited access to its IBA database. That database can contain different information on each IBA which is not reproduced in the inventory, in particular references to textbooks and details of the quality of the data relating to the definition of areas" boundaries. (43)

60. It is not known whether similar instruments existed which could make it easier for Member States to classify areas as SPAs. It would be most regrettable if the situation were different today and the competent authorities were nevertheless unable to access such instruments. That gives the impression that SEO/BirdLife is not doing its utmost to support its own objective of the complete classification of areas as SPAs. However, at the same time it should be noted that it was only completely free access to the database that SEO/BirdLife refused to grant, and it is clearly prepared to negotiate conditions of access.

61. In any event, the denial of access does not undermine the scientific quality of the published inventory. The inventory is not limited to an assessment of the literature referred to in that database, but rather is based essentially on otherwise unpublished observations and experiences of SEO/BirdLife local groups, regional authorities, biologists, nature protection groups, ornithologists, researchers, naturalists, university teachers, forestry workers and others. (44) The information that is relevant to the selection of IBAs is therefore contained in the publication, whilst the references to textbooks contained in the database are of less relevance to the identification of those areas.

62. Consequently, it would be helpful to identify the persons responsible for the collection of data in the individual IBAs. Given their expertise, those persons could assist the competent authorities in defining and classifying areas as SPAs. However, the Commission correctly pointed out at the hearing that those persons are entitled to confidential treatment on the part of SEO/BirdLife if they have to fear reprisals for their commitment to the classification of areas as SPAs. That possibility cannot be ruled out in particular where they work for national authorities which take an unfavourable view as to the classification of an area as an SPA. In the case of disputed IBAs, problems can also arise at local level.

63. Nor can the Spanish Government object to the application on the grounds that it was denied the right to a fair hearing as regards the information contained in the database. The application is not based on that information but solely on the published IBA 98 inventory.

64. Similarly, Spain cannot plead lack of support on the Commission's part. Although the Commission is required to assist the Member States as far as possible in classifying areas as SPAs, that does not allow it to provide the Member States with information in the possession of private parties. That is precisely the case as regards this database.

65. Irrespective of the extent of that denial of access, the Spanish Government is in any event not prevented from taking the necessary measures to meet its classification obligations in full. On the contrary, SEO/BirdLife has made completion of the classification considerably easier by publishing the inventory. Therefore, the Spanish authorities do not have to examine the whole of the national territory and can instead concentrate on the IBAs.

66. The Spanish Government further casts doubt on the quality of the data used. SEO/BirdLife classifies the data quality as 38% high, 44% medium, and 18% low. (45) In respect of each protected bird species in the individual areas, the IBA 98 places the data quality in the categories A, B, C or U. U means unknown, C stands for scarce information, B for incomplete information, and category A alone describes exact information. (46)

67. Spain considers that only category A information should be taken into account in infringement proceedings. However, that objection ultimately amounts to saying that Spain could fail to act until the Commission or private organisations had recorded systematically and at the highest scientific level the presence of birds in its entire national territory. However, that is Spain's responsibility under the Birds Directive. Therefore, as long as no better information is available, information of medium or low quality can serve as an indication of the existence of areas most suitable for the conservation of birds.

68. Consequently, the Spanish Government's objections regarding the data quality of the IBA 98 must also be rejected.

3. The criteria for determining the areas most suitable for the conservation of birds

69. Finally, the Spanish Government objects to some of the criteria for determining the areas most suitable for the protection of birds and the application thereof by SEO/BirdLife.

70. The criteria are set out in the IBA 98 and IBA 2000 inventories. (47) The IBAs generally accommodate a significant number of a globally threatened species (C.1) (48) or at least 1% of the relevant overall population of a species threatened in the EU, that is to say, a species listed in Annex I to the Birds Directive (C.2), (49) or another migratory bird species (C.3). (50) The IBAs also cover large congregations of over 20 000 waterbirds or over 10 000 pairs of seabirds (C.4) and bottleneck sites where over 5 000 storks and/or over 3 000 raptors or 3 000 cranes regularly pass (C.5). (51) Finally, criterion C.6 provides that the five most important areas for species listed in Annex I are to be regarded as the most important bird area in the relevant European region (the "top five criterion"). (52) Criterion C.7 covers areas designated as SPAs or selected as candidate SPAs on the basis of other criteria.

71. In addition, BirdLife International relies on general principles to define the boundaries of areas. An IBA either differs clearly from its surroundings, that is to say, it is already an area whose boundaries are defined for the purposes of environmental protection, or fulfils all the requirements of the significant species, on its own or in conjunction with other IBAs, at the time they are present. (53)

72. Since the criteria under the IBA 89 and IBA 98 for determining the areas most suitable for the conservation of birds are largely the same, the Court has already recognised them in principle when it found, in relation to the earlier IBA 89 inventory, that it was (at the time material to the relevant case) the only document containing scientific evidence making it possible to assess whether the defendant State had fulfilled its obligation to classify as SPAs the most suitable territories in number and size for conservation of the protected species. (54)

73. However, there is nothing to prevent the Member States themselves from developing criteria for the identification of the areas most suitable for the conservation of birds and classifying their SPAs on the basis thereof. (55) Nevertheless, those criteria must, from an ornithological point of view, be at least equivalent to the criteria under the IBA inventories.

74. However, Spain has not developed its own criteria but still objects to the application of selection criteria C.1 and C.6 under the IBA 98 and to the boundaries of many of the areas identified.

(a) Criterion C.1

75. Spain maintains that, contrary to the definition of criterion C.1, in certain cases SEO/BirdLife has selected areas which clearly support less than 1% of the reproducing population of the protected species. This is true in particular in the case of the great bustard. Spain is home to around 23 000 birds and consequently the limit value is 230 birds. SEO/BirdLife uses instead a limit value of 50 birds.

76. In making this criticism, the Spanish Government fails to understand the definition of this criterion. In the case of globally threatened species, it is sufficient that a significant number are present in the area concerned. This significant number can be substantially less than the threshold value of 1% which applies in the case of criterion C.2, that is to say, in relation to species which are threatened only in the EU, but not globally.

77. In respect of species which are globally threatened, BirdLife International uses, in order to counter this threat, substantially lower limit values, for example the value of 30 birds in respect of the great bustard referred to by Spain. (56) On the other hand, as the Commission points out, SEO/BirdLife set the IBA 98 threshold value higher than BirdLife International in respect of the European inventory on account of the comparatively greater presence in Spain.

78. Furthermore, this criterion was not tightened, but on the contrary relaxed, in comparison also with the IBA 89 inventory which has been recognised by the Court. Criterion No 5 in the IBA 89 also covered all breeding areas of rare or threatened species.

79. Consequently, this objection by Spain must be rejected.

(b) Criterion C.6

80. Spain further objects to the application of criterion C.6, that is to say, the "top five criterion". According to this criterion, the five best areas of each region are regarded, in respect of each species, as most suitable for the conservation of birds. Within each Member State, the boundaries of the regions are demarcated in accordance with the statistical zoning in Europe established by Eurostat (nomenclature des unités territoriales statistiques - NUTS). In different Member States, BirdLife International selected different levels of this zoning in order to establish European regions of comparable size. (57) In Spain, level 2 was selected, which corresponds to the autonomous communities and thus results in a division into 17 regions.

81. Spain refuses to recognise the autonomous communities as European regions within the meaning of this criterion. It maintains that the boundaries of the autonomous communities are not demarcated in accordance with ornithological criteria but are instead administrative units. Accordingly, there are 17 such regions in Spain, 20 in Italy and 95 in France. (58) It claims that the "biogeographical regions" used in connection with the Habitats Directive should be used instead.

82. For the purposes of the Habitats Directive, the entire territory of the Community is currently divided into seven biogeographical regions: the Continental, Mediterranean, Alpine, Atlantic, Macaronesian, Boreal and Pannonic regions. Parts of four of these regions are to be found in Spain, namely parts of the Atlantic, Mediterranean, Alpine and Macaronesian biogeographical regions.

83. If only the Spanish parts of the biogeographical regions were recognised as being "European regions" within the meaning of criterion C.6, the number of areas which comply with this criterion would be substantially lower. Instead of a maximum of 85 most suitable areas for each protected species, there would be at most only 20 most suitable areas for each protected species in Spain.

84. This objection raised by Spain cannot be dismissed out of hand. Of no relevance in this context is the Commission's submission that thus far Spain has always insisted that fulfilment of the classification obligations must be assessed separately in relation to the autonomous communities, which have sole competence in this regard. The competences of the autonomous communities cannot be the decisive factor in determining which areas are to be classified. As is known, the selection of areas must be based on scientific criteria. (59)

85. The Commission's argument that the biogeographical regions referred to in the Habitats Directive are based not on bird populations but primarily on the distribution of habitat types is clearly based much more on ornithology but nevertheless not decisive. A comparison can hardly be drawn between birds and the species and habitats protected by the Habitats Directive. However, even though the biogeographical regions are not targeted specifically at the conservation of birds, they would appear to have a stronger connection with the protection of birds than administrative and statistical territorial divisions.

86. BirdLife International itself concedes that the use of the NUTS zoning as a guide is not ideal for ornithological purposes since many bird species prefer sparsely populated areas. NUTS, on the other hand, is guided by population. The lower the population density, the larger the regions. Therefore, different NUTS levels have been selected for each Member State in order to ensure that the regions are comparable.

87. These reasons demonstrate the real value of the reference to NUTS, which does not differentiate in ornithological terms. It creates a reference size which is comparable in all Member States for the application of the top five criterion. This in turn ensures that SPAs are distributed in a reasonably uniform manner across the Community. Such a distribution is required to ensure that the SPAs are not concentrated in particular areas, but instead together form a network which covers the Community in a more or less uniform manner.

88. Such uniform coverage is necessary in particular for ornithological reasons, since it ensures that species are protected throughout their area of distribution. The importance of geographic distribution is demonstrated by the definition of the conservation status of species in Article 1(i) of the Habitats Directive. Conservation status means the sum of the influences acting on the species concerned that may affect the long-term distribution and size of its populations. Although this definition is not directly applicable to the Birds Directive, it nevertheless illustrates the scientific consensus which must also prevail when selecting areas under the Birds Directive, which must be justified on ornithological grounds.

89. If, on the other hand, larger regions were to be selected as reference sizes, for example the biogeographical regions proposed by Spain, the result would be a smaller number of SPAs in relation to an identical number of areas to be selected per region. The network would then be less dense. Furthermore, there would be a danger that those SPAs would be distributed in a less uniform manner than on the basis of smaller regions. The danger would also exist even if the number of regions to be selected were increased in the case of larger regions. In both cases, the possibility could not be ruled out that the most suitable areas would be concentrated in particular territories, whilst other territories, which, as independent regions, would have five new areas each, would not be covered.

90. Therefore, if it were desired to use reference regions other than the zoning according to NUTS, the resulting criterion would have to be designed in such a way that it produced a similarly finely-linked network of SPAs. Such a division of Spanish territory on an ornithological basis could certainly be carried out with the necessary scientific effort and subsequently be used to identify SPAs. (60) However, Spain has made no attempt to do so, but merely referred to the biogeographical regions, which do not provide a basis comparable to the autonomous communities for the establishment of a uniform network.

91. Consequently, in respect of criterion C.6 Spain has also failed to show that the IBA 98 is not the best scientific document for identifying the areas most suitable for the conservation of birds.

92. Finally, Spain is dissatisfied with the definition of the IBA boundaries. It claims that often they are incorrect since clearly unsuitable areas, such as urban areas, are covered. Furthermore, the IBAs have often turned out to be too large, on average much larger than in other Member States.

93. As with the selection of protection areas for birds, the definition of the boundaries thereof must also be based on the ornithological criteria laid down in Article 4(1) and (2) of the Birds Directive. (61) Spain criticises the boundaries of the IBAs, but not the criteria used to define them. Consequently, it objects only to the application thereof or the data used to that end. However, the generalised references, for example to urban areas, are not sufficient to enable these criticisms to be examined. Certain species can, in some circumstances, be dependent precisely on such habitats. For example, it is not disputed that colonies of the globally threatened lesser kestrel (Falco naumanni) also nest within urban areas.

94. None the less, Spain rightly emphasises that as regards the IBAs covered only partially by the SPAs the Commission contradicts its own arguments at another point of the application. (62) At that point, it submits that it accepted the scientific arguments put forward by the Autonomous Communities of Catalonia, Valencia, Galicia and Castilla-La Mancha that the boundaries of their SPAs are capable, in respect of each one ("cada una") of their IBAs, of ensuring compliance with the Birds Directive. The Commission mentioned this submission by Spain in its reply, (63) but did not explain the contradiction either in the reply or in response to questioning at the hearing.

95. This summary of the pleas in law on which the application is based does not meet the requirements of Article 38(1)(c) of the Rules of Procedure. The summary required under this provision must be sufficiently clear and precise to enable the defendant to prepare its defence and the Court to rule on the application. It is therefore necessary for the basic legal and factual particulars on which a case is based to be indicated coherently and intelligibly in the application itself. (64) In relation to SPAs within IBAs in the abovementioned autonomous communities, it is unclear, on account of this contradiction in the application, whether the Commission is actually raising the complaint. Therefore, neither an appropriate defence nor an assessment by the Court is possible.

96. Therefore, the application is inadmissible in this regard. This deficiency concerns the SPAs in the Autonomous Communities of Castilla-La Mancha, Catalonia, Valencia and Galicia which were classified at the time the action was brought and which were notified to the Commission.

97. Therefore, the complaint that certain areas of an insufficient size were classified as SPAs needs to be further examined only in relation to Andalusia, the Balearics and the Canaries.

D - The individual autonomous communities

98. Finally, both parties make further submissions regarding the individual communities. These submissions must be examined in connection with the relevant region.

1. Andalusia

99. On the basis of a comparison between the IBA 98 and the classifications in Andalusia, the Commission submits that upon the expiry of the period laid down in the reasoned opinion 37 of 60 IBAs were not covered at all by SPAs, that the 22 areas classified as SPAs only partially matched 23 IBAs, and that various species were covered to only an inadequate degree, in particular the Spanish imperial eagle (Aquila adalberti), the black stork (Ciconia nigra), the red-knobbed coot (Fulica cristata), the marbled teal (Marmaronetta angustirostris), the lesser kestrel, the purple gallinule (Porphyrio porphyrio), Audouin's Gull (Larus audouinii), the little bustard (Tetrax tetrax), the great bustard (Otis tarda), Montagu's harrier (Circus pygargus) and Bonelli's eagle (Hieraaetus fasciatus). The classified area of around one million hectares is equivalent to only around one third of the areas covered by the IBAs.

100. Spain does not contest this complaint. Therefore, it must be deemed to have been accepted. Consequently, the form of order sought in respect of Andalusia should be granted in full.

2. The Balearics

101. Forty areas covering 121 015 hectares in the Balearics had been classified as SPAs at the material time. However, the 20 IBAs covering 131 243 hectares were only 54% covered. In particular, the Commission complained that the red kite (Milvus milvus) was covered to only an inadequate degree.

102. Spain objects that the areas identified in the IBA 98 are of no use on account of the 1:2 500 000 scale and are, moreover, obsolete, but recognises that the red kite must be covered better in SPAs and to this end has classified new areas as SPAs as part of a species conservation plan and drawn new boundaries around existing SPAs in such a way that 70% of the species is covered.

103. Therefore, in principle, Spain has accepted the Commission's complaints also in respect of the Balearics. It is not possible to assess whether or not the new findings of the competent authorities actually undermine the evidential value of the IBA 98, since these findings have not been presented to the Court.

104. Consequently, the form of order sought in respect of the Balearics should be granted in full.

3. The Canaries

105. Twenty-eight areas covering 211 598 hectares in the Canaries had been classified at the material time - and, it must be answered, are so classified at present. They cover around 59.5% of the 65 IBAs and extend to 133 443 hectares. Twenty-three IBAs are not covered at all, and some of the remaining IBAs are only partially covered. The Commission highlights in particular the inadequate coverage of the following species: the Houbara (Chlamydotis undulata), the Egyptian vulture (Neophron percnopterus), the Fuerteventura chat (Saxicola dacotiae), the cream-coloured courser (Cursorius cursor) and the Bulwer's petrel (Bulweria bulwerii).

106. Spain recognises that further areas must be classified as SPAs and certain SPAs must be extended. However, the competent authorities will not approve all IBAs in their entirety on a scientific basis. In support of this submission, the Spanish Government submitted with its defence a detailed study of the Canaries" IBAs not yet fully covered, which sets out the ornithological considerations on the basis of which IBAs should or should not be classified as SPAs. (65) The Commission has not challenged this study. Therefore, the content thereof must be deemed to have been accepted and constitutes, in comparison with the IBA 98, more up-to-date and precise evidence of the failures to classify which exist. Consequently, the complaint relating to the inadequacy of areas classified as SPAs should be retained only in so far as areas must be classified as SPAs or SPAs must be extended in accordance with the Canaries study.

107. Subject to that qualification, the form of order sought in respect of the Canaries should therefore also be granted.

4. Castilla-La Mancha

108. In the case of Castilla-La Mancha too, the Commission compares the areas classified as SPAs with the data in the IBA 98. Ten of 39 IBAs covering 261 000 hectares are not covered at all, whilst the rest are covered only 32.3% overall. The lesser kestrel and little bustard in particular are not yet covered adequately.

109. As already stated, the complaint regarding the boundaries of SPAs within IBAs is inadmissible. (66) On the other hand, as regards the 10 IBAs not covered at all by SPAs, Spain has since acknowledged in three cases that classification is necessary and has probably since also carried out such classification. (67) It is not possible to assess whether the boundaries which differ from the IBA 98 are based on ornithological grounds in each case as Spain has provided no details in this regard. Seven IBAs, however, remain at issue.

110. Spain rejects the idea of classifying areas as SPAs in respect of five of these IBAs, since only small sections are located in the territory of Castilla-La Mancha. It claims that these sections have no independent ornithological value and therefore do not need to be classified as SPAs.

111. However, that submission cannot be accepted. The fact that an area most suitable for the conservation of particular species extends over the territory of various regions does not justify excluding certain sections thereof. If those sections form an integral part of the area as a whole, they must also be classified as SPAs, since otherwise measures adversely affecting the area as a whole could be taken in an uncontrolled manner. Furthermore, the Commission submits, without being challenged, that species listed in Annex I breed in at least two of the abovementioned sections, that is to say, the globally threatened imperial eagle, the black stork, Bonelli's eagle, the golden eagle (Aquila chrysaetos), the Eurasian griffon (Gyps fulvus), the Griffon vulture and the peregrine falcon (Falco peregrinus).

112. In relation to the IBA No 185, "San Clemente-Villarrobledo", Spain objects that its area of 103 000 hectares makes it too large. The covered populations of lesser kestrel, great bustard, pin-tailed sandgrouse (Pterocles alchata) and little bustard are comparatively small.

113. This submission cannot be accepted, because the area was included in the IBA 98 not only because it was one of the top five areas, but also because it supported significant populations of species which are threatened globally and in Europe. Therefore, the objections could possibly justify different boundaries but not the complete failure to classify the area as an SPA.

114. During the course of the proceedings before the Court, Spain accordingly classified an area as an SPA of the same name which covers at least 10 677.81 hectares and in doing so basically acknowledged that IBA No 185 also had to be classified. It is not possible to assess whether or not the demarcation of the boundaries is correct in ornithological terms, since Spain has put forward no grounds in support of it.

115. Finally, Spain refuses to classify IBA No 199 "Torrijos" covering 28 600 hectares, on the ground that the population of 150 to 200 great bustards is insubstantial in comparison with the overall population of the autonomous community of around 3 000. However, in this respect the European importance of this presence of a globally threatened species must be emphasised. It exceeds by far the threshold for a significant population, namely 50 birds, which has already been increased for Spain. In addition, the IBA 98 also lists a significant presence of the equally globally threatened little bustard, namely 1 200 birds, whilst the threshold value is 200 birds. The classification of this area would therefore appear to be necessary.

116. In so far as the Commission complains of the deficiencies, in the light of the IBA 98, regarding the conservation of the lesser kestrel and the little bustard, it is clear from the foregoing that further areas must be classified as SPAs in respect of both species.

117. Precisely in view of the threat to the little bustard, Spain likewise cannot successfully rely on the difficulties which result from the bird's great mobility, its minimal attachment to particular locations, its different habitat requirements during different seasons and the lack of precise population estimates. Spain must deal with these difficulties by means of further research and the classification of sufficiently large areas with correspondingly flexible management, since otherwise there must be a concern that this species will become extinct.

118. The view that colonies of the lesser kestrel within urban areas cannot be classified as SPAs is also incorrect. If lesser kestrels rely on these habitats, their protection by classification of an area is required precisely at that place. That is the only way of ensuring that urban development, for example, does not drive out the lesser kestrel. Nor does the promotion of urban development make it possible not to classify areas since, as is known, it is not possible to cite economic and social reasons for failing to do so. If such considerations were to override the interest in protecting the lesser kestrel, they would have to be implemented under Article 6(4) of the Habitats Directive, that is to say, where no alternative solutions exist and with the adoption of the required compensatory measures.

119. Consequently, Spain's objections to the IBAs in Castilla-La Mancha which are still not covered at all by SPAs must be rejected in their entirety. The form of order sought by the Commission in this regard should be granted.

5. Catalonia

120. In relation to Catalonia too, the Commission compares the IBA 98 with the actual classifications and concludes that 10 of 21 IBAs are not covered at all and the remainder are only partially covered and therefore that some of the 62 species of bird listed in Annex I which breed in Catalonia are protected in SPAs to only an inadequate degree, namely the lesser grey shrike (Lanius minor), the capercaillie (Tetrao urogallus), the shag (Phalacrocorax aristotelis), the little bustard, the Calandra lark (Melanocorypha calandra), the Dupont's lark (Chersophilus duponti), the European roller (Coracias garrulus), the short-toed lark (Calandrella brachydactyla) and the collared pratincole (Glareola pratincola).

121. As already stated, the complaint regarding the boundaries of SPAs within IBAs is inadmissible. (68) In respect of the remaining 10 IBAs which are not covered at all by SPAs, the IBA 98 does not list the lesser grey shrike, the capercaillie, the Dupont's lark, or the collared pratincole. Therefore, it has not been established that the classification of areas is inadequate in respect of these species.

122. Furthermore, Spain disputes the admissibility of the action in relation to the listed species of bird on the ground that it is unclear in respect of which species further areas should have been classified as SPAs. The Commission names 62 species listed in Annex I, whilst 73 species must be protected under Catalan law. Moreover, the Commission did not always refer to the same species during the course of the pre-litigation procedure.

123. This objection must, however, be rejected. Firstly, Spain misunderstands the reference to species of birds. Although the Commission puts the number of species listed in Annex I in respect of Catalonia at 62, it provides evidence of classification shortcomings only in respect of the species specifically mentioned above. Those species at least implicitly formed the subject-matter of the pre-litigation proceedings, since the Commission always cited the IBA 98 and the species mentioned result from the differences between the IBA 98 and the SPAs in Catalonia.

124. Secondly, Spain submits that some of the IBAs have already been proposed as sites of Community importance in connection with the Habitats Directive. However, seen from a legal perspective, such proposals cannot substitute for deficiencies in the classification of areas as SPAs, since the classification of areas as SPAs is based on Article 4 of the Birds Directive, whereas proposals for sites are based on Article 4 of the Habitats Directive, that is to say, on different rules with different legal consequences.

125. In its rejoinder, Spain pleads that most of the habitats which have not yet been classified as SPAs are protected as part of the plan for sites of natural interest or as specially protected nature areas under regional law. Irrespective of whether or not such protection satisfies SPA protection requirements, it does not justify the failure to classify an area as an SPA. Article 4(1) and (2) of the Birds Directive requires classification so that the areas referred to in Article 4(3) can be incorporated into a European network. Within this network, the quality of the protection is safeguarded by European rules. Therefore, the Court has already ruled that it is not possible to avoid the obligation to designate areas as SPAs by adopting other measures. (69) Purely national protection categories would also constitute such other measures which cannot be a substitute for classification. (70)

126. Finally, in its rejoinder Spain takes a scientific approach to the IBA 98 for the first time in respect of Catalonia. It compares the IBA 98 data on the presence of species in IBAs with the data on the relevant overall population in Catalonia provided by an atlas of breeding birds based on observations in the period from 1999 to 2002. However, these figures are comparable to only a limited degree, since the overall populations of the species in Catalonia as a whole will normally be greater than the populations in the IBAs.

127. Accordingly, the more recent data show only in the case of four species of relevance to IBAs still at issue in this case that the figures in the IBA 98 are higher than the subsequent observations, namely in respect of the Calcamar storm petrel (Hydrobates pelagicus), with 0 to 10 pairs instead of 5 to 15, the great bittern (Botaurus stellaris), with 1 to 5 pairs instead of 8 to 11, Montagu's harrier, with 5 to 10 pairs instead of 15 to 20, and the black-bellied sandgrouse (Pterocles orientalis), with 5 to 10 pairs instead of 10 pairs. It does not therefore follow that the data basis for the IBA 98 was inadequate but at most that these species, which are already very rare in Catalonia, are on the verge of disappearing and therefore require special protection.

128. In addition, the overall populations of the species represented in the 10 IBAs still at issue in this case are higher than the IBA populations. In some cases, only a fraction of the overall population is covered in IBAs, for example in the case of the kingfisher (Alcedo atthis), of which 20 to 30 pairs are present in IBAs but have a population of 1 009 to 1 420 pairs, or in the case of the Thekla lark (Galerida theklae), of which there are 100 to 200 pairs within the IBAs compared with a population of 7 300 to 18 400 pairs. Therefore, this comparison does not show that there is no need to classify further areas, but rather indicates that areas beyond the IBA 98 must be classified.

129. Spain also presents a table showing the coverage of the species in the SPAs. Spain points out that the SPAs in Catalonia in certain cases cover larger sections of the population than the IBA 98. However, Spain also declares that in the case of 16 species it falls short of the coverage by the IBA 98. These species are also represented in 9 of the 10 IBAs still at issue. The final IBA (No 138, "Islas Medas") would, furthermore, make it possible to protect at least two very rare species which have not been previously represented at all in Catalonia's SPAs, namely the storm petrel and the shag. Therefore, this submission also cannot invalidate the Commission's argument.

130. Accordingly, Spain has been unable to rebut the Commission's complaints regarding the IBAs not yet covered at all by SPAs in Catalonia. The form of order sought by the Commission in this respect should be granted.

6. Galicia

131. In relation to Galicia, the Commission states that 3 of 11 IBAs have not been classified as SPAs and some of the others have been only partially classified. In total, only 10% of the areas included in IBAs are SPAs. Therefore, the great bittern, Montagu's harrier, Bonelli's eagle, the capercaillie, the little bustard, the honey buzzard (Pernis apivorus) and the Iberian subspecies of the partridge (Perdix perdix hispaniensis) are protected to only an inadequate degree.

132. As already stated, the complaint regarding the boundaries of SPAs within IBAs is inadmissible. (71) In respect of the three remaining IBAs which are not covered at all by SPAs, the IBA 98 does not list the great bittern, Bonelli's eagle or the honey buzzard. Therefore, there is no evidence that the classification of areas is inadequate in respect of these species.

133. One of the three remaining IBAs has since been designated an SPA as regards 86% of the site and the two others are to be partially designated as SPAs on the basis of studies which have since been carried out. Therefore, Spain has accepted the remaining Commission criticism in principle and has not substantively disputed the boundaries of the IBAs. Therefore, the form of order sought in relation to the IBAs in Galicia which are still not covered at all by SPAs should be granted.

7. Valencia

134. In relation to Valencia, the Commission refers to 5 of 21 IBAs which are not covered at all by SPAs and states that some of the others are partially unclassified. Therefore, the great bittern, the Squacco heron (Ardeola ralloides), the marbled teal, the white-headed duck (Oxyura leucocephala), Audoin's gull, Montagu's harrier and Bonelli's eagle are not protected adequately in SPAs.

135. As already stated, the complaint relating to the boundaries of SPAs within IBAs is inadmissible. (72) In respect of the remaining five IBAs, which are not covered at all by SPAs, the IBA 98 lists only Bonelli's eagle. Therefore, there is no evidence that the classification of areas is inadequate in respect of the other species.

136. In respect of Valencia, Spain does not contradict the form of order sought but rather announces the classification of further areas. Therefore, the form of order sought in relation to the IBAs in Valencia which are still not covered at all by SPAs should be granted.

E - Conclusion

137. In the Autonomous Communities of Andalusia, the Balearics, the Canaries, Castilla-La Mancha, Catalonia, Galicia and Valencia, Spain has failed to classify territories of a sufficient number as SPAs and to classify certain territories of a sufficient size in the Autonomous Communities of Andalusia, the Balearics and the Canaries.

V - Costs

138. Under Article 69(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for by the successful party. In the present case, the Commission withdrew the action in relation to one autonomous community after Spain complied with the Commission's requests. In this respect, as in respect of the three autonomous communities in relation to which the Commission is entirely successful, Spain should be ordered to pay the costs.

139. On the other hand, in respect of four other autonomous communities, a substantial part of the action must be dismissed. Therefore, the costs should be shared in that regard.

140. Accordingly, Spain should pay three quarters of its own costs and three quarters of the Commission's costs and the Commission should pay one quarter of Spain's costs and one quarter of its own costs.

VI - Conclusion

141. I therefore propose that the Court should declare that:

(1) the Kingdom of Spain has failed to fulfil its obligations under Article 4(1) and (2) of Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds by failing:

- to classify territories of a sufficient number as special protection areas for birds in the Autonomous Communities of Andalusia, the Balearics, the Canaries, Castilla-La Mancha, Catalonia, Galicia and Valencia, and

- to classify certain territories of a sufficient size as special protection areas for birds in the Autonomous Communities of Andalusia, the Balearics and the Canaries,

in order to provide protection for all the species of birds listed in Annex I to the directive and for the migratory species not mentioned in Annex I;

(2) the remainder of the application is dismissed;

(3) the Kingdom of Spain is to pay three quarters of its own costs and three quarters of the Commission's costs. The Commission is to pay one quarter of its own costs and one quarter of the Kingdom of Spain's costs.

1 - Original language: German.

2 - OJ 1979 L 103, p. 1.

3 - Case C-3/96 Commission v Netherlands [1998] ECR I-3031.

4 - Case C-202/01 Commission v France [2002] ECR I-11019.

5 - Case C-240/00 Commission v Finland [2003] ECR I-2187.

6 - Case C-378/01 Commission v Italy [2003] ECR I-2857.

7 - See my Opinion delivered today in Case C-334/04 Commission v Greece.

8 - See my Opinion delivered today in Case C-418/04 Commission v Ireland.

9 - Commission press release IP/05/45 of 14 January 2005.

10 - Carlota Viada (ed.), Áreas importantes para las aves en España, Madrid, 1998.

11 - OJ 1992 L 206, p. 7.

12 - Figures taken from the Commission's Natura barometer, as at June 2006, ec.europa.eu/environment/nature/nature_conservation/useful_info/barometer/barometer.htm.

13 - For example, the Court similarly confined to Gibraltar the action brought against the United Kingdom as a whole in Case C-6/04 Commission v United Kingdom [2005] ECR I-9017, paragraph 79, and the fourth indent of paragraph 1 of the operative part.

14 - Commission v Netherlands (cited in footnote 3), paragraph 55 et seq..

15 - Case C-355/90 Commission v Spain (Santoña Marshes) [1993] ECR I-4221, paragraph 26; Case C-44/95 Royal Society for the Protection of Birds (Lappel Bank) [1996] ECR I-3805, paragraph 26; and Commission v Netherlands (cited in footnote 3), paragraph 59 et seq.

16 - Case C-355/90 Commission v Spain (cited in footnote 15), paragraph 11.

17 - Case C-173/01 Commission v Greece [2002] ECR I-6129, paragraph 7; Case C-114/02 Commission v France [2003] ECR I-3783, paragraph 9; and Case C-221/04 Commission v Spain (hunting with snares) [2006] ECR I-4515, paragraph 23.

18 - Case C-202/01 Commission v France (cited in footnote 4), paragraph 19 et seq.; Commission v Finland (cited in footnote 5), paragraph 28 et seq.; and Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 16.

19 - Case C-57/89 Commission v Germany (Leybucht) [1991] ECR I-883, paragraph 20.

20 - Case C-202/01Commission v France (cited in footnote 4), paragraph 20.

21 - Commission v Finland (cited in footnote 5), paragraph 31.

22 - Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 17.

23 - Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 18.

24 - The cases against France and Italy have been at the stage of the reasoned opinion under Article 228 EC for over a year (Commission press releases IP/05/29 of 12 January 2005 regarding France and IP/05/56 of 18 January 2005 regarding Italy). However, the Commission is examining at present whether France has since fulfilled its obligations by classifying further areas (Commission press release IP/06/907 of 3 July 2006).

25 - See the Opinion of Advocate General Alber in Case C-202/01 Commission v France (cited in footnote 4), point 25.

26 - Convention of 2 February 1971, UNTS, Volume 996, p. 245.

27 - See Case C-157/89 Commission v Italy (hunting periods) [1991] ECR I-57, paragraph 15, and Case C-60/05 WWF Italia and Others [2006] ECR I-5083, paragraph 27.

28 - Case C-355/90 Commission v Spain (cited in footnote 15), paragraph 11.

29 - Case C-209/04 Commission v Austria (Lauteracher Ried) [2006] ECR I-2755, paragraph 44.

30 - Paragraph 38 of the reply.

31 - Paragraph 183 of the application. See, in that respect, point 94 et seq. below.

32 - Commission v Netherlands (cited in footnote 3), paragraphs 68 to 70, and Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 18.

33 - See Commission v Netherlands (cited in footnote 3), paragraph 66, and Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 18, both concerning the IBA 89, and the Opinion of Advocate General Léger in Commission v Finland (cited in footnote 5), point 42.

34 - The beginnings of a scientifically substantiated argument can be seen only in respect of the Autonomous Communities of Catalonia and the Canaries. See, in this respect, points 106 and 126 et seq. below.

35 - Cited in footnote 3.

36 - Heath, M.F. and Evans, M.I., Important Bird Areas in Europe. Priority sites for conservation. Volume 2: Southern Europe, BirdLife Conservation Series No 8, Volume II, Cambridge, 2000, p. 515 et seq.

37 - See the list on p. 14 of the IBA 98.

38 - Case C-79/03 Commission v Spain [2004] ECR I-11619, paragraph 19.

39 - See point 70 et seq. below.

40 - IBA 2000, cited in footnote 36.

41 - According to the Commission's Natura barometer, as at June 2006 (cited in footnote 12), the proportion is 23% of the land area in Slovenia and 25.2% of the land area in Slovakia.

42 - Although the financing rules contained in Article 8 of the Habitats Directive do not cover SPAs under the Birds Directive, the sixth environment action programme refers to "establishing the Natura 2000 network and implementing the necessary technical and financial instruments and measures required for its full implementation and for the protection, outside the Natura 2000 areas, of species protected under the Habitats and Birds Directives" as a priority action (seventh indent of Article 6(2)(a) of Decision No 1600/2002/EC of the European Parliament and of the Council of 22 July 2002 laying down the sixth Community environment action programme (OJ 2002 L 242, p. 1)).

43 - IBA 98, p. 363 et seq.

44 - IBA 98, p. 39.

45 - IBA 98, p. 43, Figure 5.

46 - IBA 98, p. 364 et seq.

47 - The criteria underlying the IBA 89 are set out in the IBA 2000, Volume 2, p. 776 et seq., and compared with the criteria used in relation to the new inventories. See also the IBA 98, p. 368.

48 - This criterion is equivalent to criterion No 5 in the IBA 89.

49 - This criterion is equivalent to criterion No 1 in the IBA 89.

50 - This criterion has no equivalent in the IBA 89.

51 - Criteria C.4 and C.5 develop criterion No 9 in the IBA 89. The latter covers the staging areas of over 20 000 waterbirds and/or over 5 000 raptors.

52 - Criterion C.6 was not referred to specifically in the list of criteria for the IBA 89 but was referred to specifically in the explanatory notes thereto. See Annex 7, p. 2, to the application in Case C-3/96 Commission v Netherlands and Annex 16 to the application in Case C-378/01 Commission v Italy. Accordingly, this criterion was developed and applied in connection with the Corine biotope project.

53 - IBA 2000, p. 13

54 - Commission v Netherlands (cited in footnote 3), paragraph 69, and Case C-378/01 Commission v Italy (cited in footnote 6), paragraph 18.

55 - See the references cited in footnote 33.

56 - IBA 2000, p. 13.

57 - IBA 2000, Volume 2, p. 18. Thus, it is clear from p. 778 that in respect of Belgium, Greece and Ireland a higher level than that of the IBA 89 was selected and the number of regions was thereby reduced.

58 - The Spanish Government is wrong in this regard. According to the IBA 2000, Volume 2, p. 778, the second level of NUTS is also used in France, which has 22 regions.

59 - See the references cited in footnote 15.

60 - For example, in Germany so-called "principal units of natural area" are used as subdivisions of the German parts of the biogeographical regions to identify proposed areas in connection with the Habitats Directive.

61 - Royal Society for the Protection of Birds (cited in footnote 15), paragraph 26, and Case C-191/05 Commission v Portugal (Moura, Mourão, Barrancos) [2006] ECR I-6853, paragraph 10.

62 - Paragraph 183 of the application.

63 - Paragraph 33 of the reply.

64 - Case C-178/00 Italy v Commission [2003] ECR I-303, paragraph 6. See also my Opinion in Case C-221/04 Commission v Spain (cited in footnote 17), point 25.

65 - Annex II to the defence.

66 - See point 94 et seq. above.

67 - The IBAs involved are IBA No 78 "Puebla de Beleña", IBA No 183 "Hoces del Río Mundo y del Río Segura", and IBA No 189 "Parameras de Embid-Molina".

68 - See point 94 et seq. above.

69 - See point 21 above.

70 - Commission v Austria (cited in footnote 29), paragraph 48.

71 - See point 94 et seq. above.

72 - See point 94 et seq. above.