One hundred million tonnes of fish, and fisheries research

The nominal fish harvest of 100 X 10(6) t reported for 1993, while exceeding many early predictions of ''potential yield'' remains well below some other, more optimistic predictions. The key elements of a sample of prediction of marine fish potentials are reviewed, enabling identification of their shared features. One of these is the tendency for initial guesses (e.g. of the relations between optimal catches and production, of conversion efficiencies between trophic levels, or of the trophic level of harvested fish) to become legitimized with age, This has led to debates on potential yields turning into circular arguments. Obviously, a way out of this is also proposed.