Part II: Some Observations on the Agreement between Greece and Egypt on the Delimitation of the Exclusive Economic Zone

Where to from here? The Greece-Turkey maritime boundary dispute

The dispute over the maritime boundary between Greece and Turkey has been ongoing since 1974. As I wrote elsewhere, ‘bilateral consultations and negotiations have been unable to prevent occasional flare-ups and, as tensions remain significantly high, it may not be long before one of the not infrequent confrontations spirals out of control’. Only a few days after the signing of the Greece-Egypt Agreement, Greece and Turkey came close to an armed incident. Turkey sent a survey vessel, escorted by warships, and the standoff resulted in a minor collision between two frigates.

The existing status quo is unstable and does not favour either side. It poisons bilateral and regional relations and holds hostage the multiple benefits that could be generated from the exploitation of the seabed resources in the disputed maritime area. Given the recurring tensions and vastly overlapping maritime claims, the next maritime boundaries to be determined should be between Greece and Libya, on the one hand, and Greece and Turkey, on the other.

UNCLOS, to which the vast majority of States are parties, does not provide a single delimitation method. Yet, multiple maritime boundary litigations and arbitrations have taken place since UNCLOS’ adoption. Courts, tribunals and State practice have come to articulate specific delimitation methods and approaches. Going back to square one in every delimitation situation is thus no longer necessary, as Elferink and others explain. The three-stage delimitation approach, which involves a provisional equidistance line drawn from the nearest base points of two adjacent or opposite States, adjusted for equity in light of the relevant circumstances and proportionality requirements, has now become the standard approach, Murphy explains. The approach has been adopted by the ICJ, the International Tribunal for the Law of the Sea (ITLOS), and several arbitral tribunals exercising jurisdiction under UNCLOS. Cases involving islands in the delimitation area, where the applicable law was customary international law (not UNCLOS), have also followed this approach, such as in the Nicaragua v Colombia ICJ case (see Burke’s analysis).

Applying a three-stage delimitation in the delimitation of the maritime boundary between Greece and Turkey is not straightforward. A key element in the negotiations between Greece and Turkey (and arguably between Greece and Libya) will be the role of Greek islands in the delimitation exercise. The exact course of the final boundary line may be completely different depending on the effect accorded to these islands in the delimitation process. Greece’s position is that its islands may generate maritime areas of their own; they should thus be taken as base points and a median line of delimitation be drawn between the Greek islands and the Turkish mainland coasts. Neither UNCLOS nor customary law addresses this issue directly. I have argued that international practice on the treatment of islands in the delimitation process is ‘unhelpfully inconsistent’. In Bangladesh/Myanmar, ITLOS cited Bowett in holding that the treatment of islands in relation to delimitation is ‘so diverse that any generalization as to their effect will be hazardous’ (para 77).

In several instances, islands were given full effect in the delimitation process. This was the case in the delimitations between Cyprus and neighbouring States. Also, in St Pierre and Miquelon, the tribunal granted the small islands of St Pierre and Miquelon (totalling about 240 square kilometres with a total population of 5,800) a full seaward projection of 200 nautical miles towards the south where ‘the French islands have a coastal opening…which is unobstructed by any opposite or laterally aligned Canadian coast’ (para 45).

Small, usually uninhabited, islands have been ignored altogether for delimitation purposes. Islands were not considered as base points in Libya/Malta (Maltese island of Filfla – about 0.06 square kilometres); Guinea/Guinea-Bissau (Guinea island of Alcatraz – about 0.19 square kilometres); Qatar/Bahrain (Qit’at Jaradah – about 0.04 square kilometres); the Black Sea case (Serpents’ Island – 0.6 square kilometres); and Nicaragua v Colombia (where the barren and uninhabited islands of Quitasueño and Serrana were not used as base points in drawing the EEZ and continental shelf delimitation line). In Bangladesh/Myanmar, St Martin’s island (87 square kilometres) was given full effect in the delimitation of the territorial sea, but was discounted as a base point in the delimitation of the EEZ and the continental shelf between the two adjacent States.

Islands were used as base points but ultimately received less ocean space compared to their opposing mainland in other instances, such as the Continental shelf (Libya/Malta) delimitation, in the Anglo-French Arbitration (Isles of Scilly, 6.58 square kilometres), Jan Mayen (Jan Mayen, square kilometres), and the Gulf of Maine (Seal Island, 3.2 square kilometres). Moreover, islands were fully or partially disregarded in the Anglo/French Arbitration (Channel Islands, 198 square kilometres), the Dubai/Sharjah arbitration (Abu Musa island, 12 square kilometres), and Eritrea/Yemen (Hanish islands, 300 square kilometres).

These previous delimitations involving islands offer important lessons but they should be taken cautiously. The Greek-Turkish delimitation context is unique. The topography is not one where a few tiny, uninhabited and isolated maritime features potentially interfere with the maritime entitlements of a large continental mainland. Greece is comprised of substantially-sized islands, all home to large populations, with an economic life of their own. Rhodes is about 1500 square kilometres and home to 51,000 permanent residents. Crete is about 9000 square kilometres and home to 635,000 permanent residents.

Moreover, the vast majority of Greek islands are closely knit, forming groups that represent a geographical unity; their combined coastline size, exceeds 7,000 km. Most of the islands in the Sporades group on the western sector lie at a distance of less than 3 nautical miles from each other (e.g. Yura, Kyra Panagia, Alonissos, Peristera, Skopelos and Skiathos). The eastern Aegean islands of Samos, Furni and Ikaria lie at a distance of less than 10 nautical miles from the northernmost islands of the Dodecanese group, Patmos and Lipsi. Within the Dodecanese group, distances are on average less than 12 nautical miles. Islands of the Cyclades group, in the south-west sector, are tightly grouped together with distances not exceeding 10 nm. The tribunal in the Eritrea/Yemen arbitration emphasised the importance of ‘geographical unity’ of islands in maritime boundary delimitation. It gave full weight to Eritrea’s Dahlak islands (a group of 350 islands and islets of which only the largest has a population of 3000 people). It did not discount the South West Rocks, finding that islands situated less than 24 nautical miles apart from each other could generate a continuous belt of territorial sea, regardless of their distance from the mainland.

Turkey relies on the ‘equitable character’ of any delimitation solution, whereby no delimitation method, including equidistance, has priority because equity can only be defined in light of the specific circumstances of the area to be delimited. Given the geography at hand, any future delimitation in the Eastern Mediterranean Sea and the Aegean should strike a reasonable balance between the parties’ competing maritime entitlements. Equal sacrifices must be made to reach an equitable result. Whilst equitable adjustments are necessary to avoid a grossly disproportionate result, the maritime projections generated by the Turkish coasts would inevitably be encroached upon. Inasmuch as the various Greek islands would encroach upon the seaward projection of the Turkish coasts, the cutting-off of the seaward projections of the Greek islands would also result in inequitable effects.

Pending a delimitation agreement, unilateral coercive actions contravene both Articles 74(3) and 83(3) UNCLOS and the fundamental legal obligations under Article 2(3) and 2(4) of UN Charter to refrain from the threat or the use of force, or any other means of forcible coercion for the settlement of disputes; refrain from interference in other states’ internal affairs; and have recourse at all times to peaceful means of dispute resolution. Such actions raise questions of State responsibility. They also raise questions of possible collective action to steer recalcitrant States towards meeting their legal obligations under the UN Charter.

Unilaterally-defined maritime boundaries are not legally opposable vis-à-vis third States (Fisheries case 1951) – they are albeit useful in clarifying the scene of overlapping maritime regional claims. Creative, practical, maritime delimitation solutions are urgently needed to resolve the situation between Greece and Turkey. I have suggested some proposals. It remains to be seen whether solution will come through negotiations or through third-party adjudication or arbitration. Greece and Turkey could also refer their disputes to a conciliation commission. Decisions would not be binding, but may allow the parties to negotiate a solution based on the conciliation commission’s report (see Timor-Leste and Australia for example, discussed by Tomuchat and Kohen).