Technique of systems identification applied to estimating copepod production

Abstract

Systems identification offers some advantages over other methods of estimating zooplankton production. The technique involves specification of a dynamic population model with unknown parameters. By using least squares analysis to fit the model to data series the parameters and production are estimated. Two cohort population models are described and tested against data generated by a simulation model with known production. A third popular method developed by Winberg was also applied for comparison. The systems identification technique was found to be least sensitive to varying levels of stage analysis and sampling interval. The models were also applied to population data from an enclosed water column (CEPEX). Recommendations are made on the degree of stage aggregation and sampling interval required to obtain good estimates of production for populations with identifiable cohorts.