The term shared fish stocks is understood by FAO (see, in particular, the FAO, Code of Conduct for Responsible Fisheries, Article 7 (FAO, 2003b)) to include the following:
(a) fish resources crossing the EEZ boundary of one coastal State into the EEZ(s) of one, or more, other coastal States - transboundary stocks;
(b) highly migratory species, as set forth in Annex 1 of the 1982 UN Convention on the Law of the Sea (UN, 1982), consisting, primarily, of the major tuna species (being highly migratory in nature, the resources are to be found, both within the coastal State EEZ, and the adjacent high seas);
(c) all other fish stocks (with the exception of anadromous/catadromous stocks) that are to be found, both within the coastal State EEZ and the adjacent high seas - straddling stocks;
(d) fish stocks to be found exclusively in the high seas - discrete high seas fish stocks.
Clearly, these categories are not mutually exclusive. One can find many examples of fish stocks that fall into Category (b), or Category (c), which also fall into Category (a).
John Caddy (1997) provides a definition of transboundary stocks, which, with minor modification, can be extended to cover straddling and highly migratory (but not discrete high seas) stocks. The modified Caddys definition is as follows:
...a group of commercially exploitable organisms, distributed over, or migrating across, the maritime boundary between two or more national jurisdictions, or the maritime boundary of a national jurisdiction and the adjacent high seas, whose exploitation can only be managed effectively by cooperation between the States concerned...
There is not universal agreement, by the way, that cooperation between and among relevant States/entities is required for the conservation and management of all of these resources. Be that as it may, a fundamental issue pertaining to shared fish stocks, which will be the recurrent theme throughout this paper, is the development of appropriate management and conservation regimes (which could include non-cooperative regimes as an option) for these resources.
The 1982 UN Convention on the Law of the Sea (1982 UN Convention, hereafter) (UN, 1982), which came into force in 1994, and the 1995 UN Fish Stocks Agreement (UN, 1995), which came into force in 2001, will be seen to provide the basic legal framework for such regimes. The 1982 UN Convention calls upon States, be they coastal States or distant water fishing States (DWFSs), to cooperate, or at the very least to negotiate, with respect to the management and conservation of all categories of shared stocks. In so doing, the 1982 UN Convention sets forth the basis upon which States are to negotiate or cooperate. Such negotiation and cooperation may be effected through bilateral or other agreements, or may be effected through appropriate subregional and regional organizations.
One of the earliest analyses of the problems of managing and conserving shared fishery resources was that prepared by John Gulland of the FAO, prior to the close of the UN Third Conference on the Law of the Sea (Gulland, 1980). In his paper, Gulland focuses on transboundary stocks, because of their then perceived importance. His analysis can, however, be readily extended to straddling and highly migratory stocks.
In any event, Gulland presents a biological/geographical categorization of transboundary fish stocks, which is useful in setting the stage for the discussion of the problems of managing the resources. He makes the following distinctions:
I. stocks occurring within two or more EEZs, but showing no clear migratory pattern;
II stocks occurring within two or more EEZs, and displaying a clear pattern of movement:
1) resulting from seasonal migration
2) according to development stages.
Change stocks occurring within two or more EEZs, in I and II, to stocks occurring within the EEZ and the adjacent high seas, and one has a description of straddling and highly migratory fish stocks.
In the case of (I), Gulland contends, it is not always clear that exploitation on one side of the relevant boundary will necessarily have a significant effect upon harvesting opportunities on the other side of the boundary (Gulland, ibid.). Munro (1987) provides such an example in the form of the rich Georges Bank scallop fishery, shared by Canada and the United States. The resource was, and is, clearly a transboundary fish stock. It was, however, questionable whether Canadian (American) harvesting of scallops would have any significant impact upon American (Canadian) harvesting opportunities. Adult scallops are more or less stationary. Moreover, while there is some transboundary movement at the larval stage, there were, in 1987, extensive beds of larvae producing scallops, which were free from exploitation due to the sea bed terrain.
These facts led to the argument that, since Americans and Canadians could harvest the resource without affecting one anothers harvest opportunities, cooperative fisheries management of this shared resource was largely beside the point (Munro, 1987). Whether the situation, which prevailed in 1987, continues to prevail today is not known to these writers. Nonetheless, the point remains, and leads to our first conclusion, namely that cooperative fisheries management of a shared fishery resource is not, in all cases, necessarily required, or desirable. The conclusion is reinforced by the fact that establishing a cooperative fisheries management regime is not a costless undertaking. As the Report of the Norway-FAO Expert Consultation points out, realizing the benefits of cooperation will require a significant investment in the form of financial resources. The investment should properly be assessed in a cost/benefit framework (FAO, 2002a). If the assessment reveals that the expected net benefits from the investment will be negative, then obviously the case for cooperation collapses.
The implication is that, in the case of a Georges Bank scallop type of resource, the relevant States could, in keeping with the 1982 UN Convention, enter into negotiations and conclude that a satisfactory conservation and management regime would consist of each State managing its segment of the resource to the best of its ability, without reference to the other State(s). It will be seen that the 1982 UN Convention does, in fact, allow for just such an outcome.
Suppose now that the harvesting activities of one State, sharing a fishery resource, do impinge significantly upon the harvesting opportunities of the one, or more, other States sharing the resource, so that a prima facie case for cooperation does, in fact, exist. The question, which then arises, is the appropriate level of cooperation. There are, as Gulland points out, at least two levels of cooperation (Gulland, 1980). The first level, or what we might term the primary level, consists of cooperation in research alone, without reference to coordinated management programmes. Since all parties should stand to benefit from improved information and data, the cooperation should be relatively easy to achieve. The emphasis is on the word relative, however. It is still possible that one or more parties may suspect that research information, which it shares, will serve to benefit its rival exploiters of the resource, at its own expense.
In any event, if it is not possible to achieve cooperation at this primary level, it certainly will not be possible to achieve cooperation in active management of the resource. In actual cooperative management regimes, which have proven to be successful, cooperation in research alone is often seen, in retrospect, to have been the precursor to cooperation in active management.
The significance of this primary stage of cooperation was made apparent at the Norway-FAO Expert Consultation. Scientific cooperation, although not cooperation in active management, was seen to exist between Kuwait and Iran in the Persian Gulf, between and among the Pacific Latin American states of the Southeast Pacific, and among the coastal states of Northwest Africa (Al-Hussaini, 2003; Zuzunaga, 2003; Samb, 2003). In each of these cases, there are prospects that cooperation at the scientific level will provide the foundation for more extensive cooperation in resource management, in the future.
The Norway-FAO Expert Consultation also provided evidence that inadequate scientific cooperation can serve to undermine the development of cooperation in active resource management. Although there are clear benefits to South Africa, Namibia and Angola in cooperating in the management of a major shared hake stock, lack of scientific knowledge of the resource is acting as a barrier to the development of an effective cooperative arrangement (Sumaila, Ninnes and Oelofsen, 2003). On the other side of the South Atlantic, a decline in cooperative scientific research, due to lack of funding, is serving as a hindrance to effective cooperation in the management of fishery resources shared by Argentina and Uruguay (Chaluleu, 2003).
What we might call cooperation at the secondary level - active management - involves, almost by definition, the establishment of coordinated joint management programmes. As Gulland (1980) informs us, this will require:
(a) allocation of harvest shares among the participating states (or entities);
(b) determination of an optimal management strategy through time, including inter alia, the determination of optimal global harvests over time;
(c) implementation and enforcement of coordinated management agreements.
The Report of the Norway-FAO Expert Consultation maintains that, in order to achieve all of this, it will be necessary to have:
a cooperative management authority;
a detailed joint management plan;
a set of agreed upon common objectives;
agreed upon tools for managers, including indicators and reference points to monitor performance;
a joint scientific body to provide advice.
The detailed joint management plan should be expected, at a minimum, to contain: (i) a description of the fishery, (ii) objectives of management, (iii) measures to achieve the objectives, (iv) indicators and reference points to be used to measure actual performance against objectives, (v) decision rules on how to change management, when the objectives are not being reached, and (vi) information needs and research required to support management (FAO, 2002a).
Obviously, establishing cooperation in shared stock management at the secondary level is a much more formidable undertaking than is establishing such cooperation at the primary level. Outside of the expense and administrative complexities involved in establishing a mechanism for secondary level cooperation, numerous other problems and difficulties emerge. To begin, even cooperation in research may lose its benign character. Research findings can influence harvest allocations, and thus can easily become tools of combat in negotiations between and among relevant states.
Secondly, consider (b), in the Gulland list, optimal management strategies. While we shall come to talk, for ease of exposition, in terms of the management of single shared stocks, the reality in many, if not most, cases is likely to be far less simple. In many cases, what will be shared might better be referred to as ecosystems, in which there are sets of shared stocks interrelated in ways, both complex and intricate (e.g. through predator-prey relations).
Next, under the heading of optimal management, is the fact that there is no assurance that the relevant states will have identical resource management goals. The FAO recognized this fact, as early as 1979, with reference to transboundary stocks, through its Advisory Committee on Marine Resources Research (FAO, 1979). The Committee pointed out that, if two coastal States share a fishery resource, one might favour low long run TACs, but a large stock and high catch rates, while the other might favour large long run TACs, and accept with good grace low catch rates. If management goals are not identical, then one is faced with the burden of developing a mutually acceptable compromise resource management programme, or so it would seem (FAO, ibid.).
Thus, establishing cooperative conservation and management at the secondary level can prove to be frustrating and costly. One can add that the anticipated cost might appear not only in economic terms, but may also appear, as far as coastal States are concerned, in the form of perceived loss of sovereignty. If, the net economic benefits from cooperative management appear not to be substantial, the relevant states, with the possible loss of sovereignty in mind, may conclude, to use an old English expression, that the game is not worth the candle.
Certainly in the case of transboundary stocks, each relevant coastal State could, upon assessing the benefits of cooperation, conclude the following. If it were to manage alone its segment of the resource in a rational manner, and if its fellow states sharing the resource were to do the same, the overall resource management results, while not ideal, would be adequate, even though the coastal States harvesting opportunities are affected by the harvesting activities of its fellow coastal State(s). One of the central questions to be addressed is whether or not this comfortable view of the world is, in fact, reasonable.
Difficulties of achieving effective cooperation in resource management to one side, the significance of the issue of cooperative management of shared fishery resources is dependent ultimately upon the importance of shared fishery resources in terms of world fisheries. No precise measures exist, but there is enough evidence to indicate that the significance of shared fish stock in world capture fisheries is decidedly non-trivial.
To commence, John Caddy (1997) has undertaken a thorough investigation of transboundary stocks. Caddy first observes, in his 1997 paper (Caddy, 1997), that, in 1982, as the world Exclusive Economic Zone (EEZ) regime was emerging, he had pointed out that a significant proportion of fishery resources then being encompassed by EEZs would be found to be shared with other coastal States. He subsequently proceeds, with the aid of the Geographical Information System database, to estimate the number of contiguous EEZ maritime boundaries. Then, making a very conservative estimate of the number of fishery resources crossing these boundaries on average, he comes forth with an estimate of 1 000 to 1 500 transboundary fishery resources (Caddy, ibid.). In short, transboundary fish stocks are ubiquitous, worldwide.
As a contribution to the UN Fish Stocks Conference, the FAO prepared a survey of world highly migratory stocks, and of straddling stocks, both actual and potential (i.e. Categories (b) and (c)) (FAO, 1994). In the publication, the FAO presents estimates of the total annual harvests of the two categories of stocks, as of 1991. Upon updating these estimates to the year 2001, we conclude that the total annual harvests of highly migratory, plus actual and potentially straddling stocks, may be as high as 15 million tonnes. This, in turn, is roughly equal to one fifth of the total harvests from world marine capture fisheries for that year (FAO, 2002b; 2003a). One can only guess at what the proportion would be, if it were possible to add in harvests of transboundary stocks, which are not also straddling or highly migratory stocks, and harvests of discrete high seas stocks. For example, many important small pelagic stocks (e.g. Southeast Pacific, West and South Africa) are transboundary, but not straddling, in nature. Including such stocks could raise the total catches from shared fish stocks to close to one third of global marine capture fishery harvests.
As well, we shall point out (and as should come as no surprise), the management of transboundary fish stocks is, by and large, a considerably easier undertaking than is the management of highly migratory, straddling and discrete high seas stocks. Nonetheless, John Caddy, on the basis of his survey, argues that only a small percentage of the 1 000-1 500 relatively easy to manage transboundary fish stocks are subject to effective cooperative management (Caddy, 1997). To this, we can add the observation that the UN Fish Stocks Agreement emerged in response to the manifest inadequacy of the management of many of the worlds highly migratory and straddling fish stocks (as well as the limited guidance offered by the 1982 UN Convention on the Law of the Sea with respect to the management of these stocks).
The information, which we have at our disposal, is, admittedly incomplete, and somewhat fragmentary. It does, nonetheless, allow us to advance the following propositions. If it is indeed the case that cooperative management is important for the long term stability and productivity of most shared fish stocks, then the scope for improved management of these resources is immense. Secondly, the potential significance of such improved management for the sustainability of world capture fisheries in general is indeed very high.
With the overview of shared fish stocks complete, we turn to the question of the conservation and management of these stocks, commencing with Category (a), transboundary fish stocks.
 There is not uniform
agreement on the categorization of these fish stocks (see: Van Houtte, 2003, for
a thorough discussion). While there is no disagreement on the definitions of
straddling, highly migratory and discrete high seas stocks, what we might refer
to as a second school of thought prefers to use the term "transboundary stocks"
as the generic term, and to use the term "shared stocks" to denote those fish
stocks crossing the EEZ boundary of one coastal State into the EEZ(s) of one, or
more, other coastal States.|
 The full title of the Agreement is: Agreement for the Implementation of the Provisions of the United Nations Convention on the Law of the Sea of 10 December 1982 Relating to the Conservation and Management of Straddling Fish Stocks and Highly Migratory Fish Stocks.
 Although admittedly not discrete high seas stocks, which we shall, at a later point, refer to as "orphan" stocks.
 The authors are grateful to Mr Luca Garibaldi, FAO Department of Fisheries (FIDI), for his assistance in preparing these estimates.