On the one hand, for biologists, a stock comprises all the individuals of fish in an area, which are part of the same reproductive process. It is self-contained, with no emigration or immigration of individuals from or to the stock. It occupies a well-defined spatial range and is independent of other stocks of the same species. Random dispersal and directed migrations due to seasonal or reproductive activity can occur. Some species form a single stock (e.g. southern bluefin tuna) while others are composed of several stocks (e.g. albacore tuna in the Pacific Ocean comprises separate Northern and Southern stocks). The impact of fishing on a species cannot be determined without knowledge of this stock structure.
On the other hand, and for practical reasons, a fraction of the unit stock may be considered a "stock" for management purposes (e.g. as management unit), as long as the entity is distinct enough for the results of the assessments and management to remain close enough to what they would be on the unit stock.
Of course, collecting catch data on a given stock is easiest when it is harvested by only one country, but this rarely happens because stocks often cross political boundaries and many countries are generally involved in their exploitation. For example, many fish follow migratory routes that take them through the waters of several countries or out into international waters. The situation may be further complicated by the life history of the target species. In many marine species, the eggs and larvae drift away from the spawning grounds into the open sea. In others, young and adult fish spend most of their life in the open ocean, returning to coastal waters or even rivers only to breed.
One of the first requirements to identify a stock is to establish the distribution and biological 'unity' of the resource that is being caught.. The greatest source of error in stock assessment and management is to underestimate the extent of a unit stock. Fish can move over considerable distances. So the success of a fishery or the outcome of a management scheme in one area can be affected by catches of another some distance away. For example, attitudes to an expanding salmon fishery off Greenland changed rapidly when it was discovered that the fish being caught there came from, and could have been expected to return to, American and European rivers.
Differences between stocks may be established by taxonomic, physiological and biochemical features. For example, the analysis of blood proteins, the structure of which is determined genetically, has been used to identify herring stocks, sub-populations of sardines and groups of tuna. It has also been used to differentiate between the many cod stocks that exist in the north Atlantic. Tagging and fin clipping are also widely used to track the migration and distribution of fish.
From a practical standpoint, deciding on the practical boundaries of a "stock" and of a management scheme is less critical when the pattern of fishing is relatively uniform over a wide area. In such a situation, the fish of a given species may be treated as a single unit stock if they do not differ greatly in reproduction, growth and mortality rates. On the other hand, if the intensity of fishing differs markedly within the area occupied by the species and the fish populations in different parts of the area rarely mix then the fish population in each part would have to be treated as a unit stock.
While there is a substantial amount of information on important stocks exploited by large scale industrial fishing, there is little or no information on a very large number of smaller coastal stocks (e.g. exploited by small-scale fisheries) or very deep sea stocks exploited without control far away from the coasts (e.g. on sea-mounts).
FAO has started the development of a worldwide inventory of "stocks" in the sense of distinguishable, pragmatic resource units identified by governments as management units and often, as a consequence, used as a basis for stock assessment.