Modelling zooplankton population dynamics with the extended Kalman filtering technique

Abstract

The extended Kalman filter is a mathematical method for simultaneous state and parameter estimation, originally developed for use in engineering science. We applied the technique for modelling zooplankton population dynamics in nature. We described population dynamics by a stage-classified matrix projection model, where vital rates were allowed to vary between stages and over time. We tested the technique with simulated rotifer data and with field data of a Filinia longiseta (Rotifera) population from a sewage treatment pond in Hungary. Very quick changes in model parameters were typical for the population examined. However, the extended Kalman filter was capable of tracking parameter changes in the varying environment. The technique was also effective in filtering moderate sampling noise. The Kalman filter seems to be a very promising method for zooplankton population analysis.