A stochastic model of phytoplankton dynamics in lake balaton

Deterministic models (mainly systems of differential equations) are usually used to model lake ecosystems; however, they are inadequate for reflecting the inherently-random behaviour of ecosystems. A stochastic version of the Balaton Eutrophication Model, a Markovian jump process has been devised consisting of the same components and subprocesses as the deterministic one, but the changes of state proceed as follows: the system stays at a state for a random period of time, then one of the subprocesses (selected at random) takes place. Relations between the two approaches are discussed showing that in this case the stochastic approach provides an explanation for large fluctuations in measured data (especially in the case of blue-green algae) while the deterministic approach gives none.