A dredge equipped with a gate which may be opened by a diver-monitor when it is passing over bottom selected for study and closed elsewhere is described. A meter wheel which activates an electric switch, connected with a battery and counter in the boat, is in contact with the bottom when the gate is open and raised from the bottom when it is closed. This permits the taking of sampling units of uniform predetermined size. Field trials demonstrated that the dredge efficiently captured all categories of animals large enough to be retained by its meshes.

The difficulty of obtaining unbiased, representative samples of single populations is compounded by the complexities of aggregates of populations, such as biotic communities, and even more by discontinuities in the dispersion of the constituent species and the habitats in which they live. The probability of taking such unbiased, representative samples of marine communities decreases with the depth of water from which they are obtained. Sampling confidence at a level which may be taken for granted in terrestrial habitats is possible only in the littoral and shallow sublittoral. And this is true only if the investigator is able to assure himself that each sampling unit is excised only from the kind of habitat whose biota he wishes to study. He may achieve this specificity in sampling benthos, for example, by using grabs and discarding material brought up from bottom types not under investigation. Another approach is to engage in a comprehensive, preliminary, systematic exploration so that when he samples for parametric estimates he can place his grab exactly where he knows that the desired habitat exists.