The ecosystem of a mesotrophic lake-I. Simulating plankton biomass and the timing of phytoplankton blooms

The response of plankton biomass in Lake Mjøsa, Norway, to changes in exogeneous factors during the years 1976–87 is studied by using a simulation model of the lake ecosystem. The model includes mechanisms required to test the Sverdrup hypothesis for the initiation of the spring phytoplankton bloom, and it includes zooplankton grazing and thermocline erosion which is important factors contributing to the formation of a second autumn bloom. The model describes 45% of the observed inter annual variance in chl-a, but only the right order of magnitude for the zooplankton biomass. The model describes 35% of the variance in the timing of the onset of phytoplankton growth (p = 0.03) and 41% of the variance in the timing of the second bloom (p = 0.07). However, 4 of 12 simulated annual time series showed only one bloom. The OECD regression model for chl-a as a function of TP concentration and flushing rate explained 50% of the variance in chl-a, but a zooplankton regression model did not explain the observed variance in zooplankton biomass. A published regression model for the timing of the spring bloom gave a negative correlation with the observed bloom.