A perspective on the use of cohort analysis to obtain demographic data for copepods 1

Plankton ecologists have devised several methods for extracting demographic information on copepod populations from abundance data. The most widely used of these methods was developed by Rigler and Cooley (1974), who noted the failure of earlier methods to take full advantage of the information in instar-specific abundance data. The Rigler-Cooley method provides estimates of development times and survivorship from the mean pulse times of instar abundances. The Rigler-Cooley method is most often used in estimating development times that are to be used in estimating secondary production (Boers and Carter 1978; Burns 1980; Lewis 1979). Hairston and Twombly (1985) recently undertook a critical appraisal of the RiglerCooley method, among others, and concluded that underlying assumptions impose serious limitations on its utility. Hairston and Twombly show that the Rigler-Cooley method is strictly correct only under the unlikely condition that survival is perfect. We believe that this is correct. In addition, however, Hairston and Twombly present a very pessimistic view of the general usefulness of the Rigler-Cooley method. We believe that the Rigler-Cooley method is of much broader usefulness than one might conclude from the work of Hairston and Twombly. In view of the virtual certainty that assumptions of the Rigler-Cooley method will be violated, we have studied the robustness of the method for the estimation of development times of copepods. If the method is robust, violations of its underlying assumptions within certain limits may not be of great practical importance. For example,