Published online by Cambridge University Press: 11 May 2009
Enclosures of various sizes and configurations have been employed to maintain natural planktonic communities and to examine their responses to pollutants (e.g. Menzel & Steele, 1978; Steele, 1979). Studies on feeding, growth, and mortality of larval fishes, at times conducted in conjunction with pollution experiments, have also been successfully conducted in large enclosures (Koeller & Parsons, 1977). Implicit in these and other comparable studies has been the expectation (or hope) that a balance could be achieved between the advantages and difficulties inherent in more traditional field and laboratory studies. Field studies offer the realism of working with the natural assemblage with its many interacting components and links, but also suffer the disadvantage associated with a turbulent and advective system, which generally makes the repetitive sampling of the same populations impossible. Laboratory studies reverse the balance; control and definition of the components are gained at the expense of realism. The use of large containers represents a hybrid approach characterized by a partial control over a moderately realistic ecosystem. In the ideal case, the use of large containers allows sufficient control to permit experimental manipulation of (and the testing of hypotheses concerning) planktonic assemblages which are sufficiently realistic to permit extension of the experimental results to the ‘real world’.