A simulation study of the feedback of phytoplankton on thermal structure via light extinction


Karsten Rinke, University of Konstanz, Limnological Institute, Mainaustrasse 252, D-78464 Konstanz, Germany. E-mail: karsten.rinke@uni-konstanz.de


1. A simulation study of the feedback of phytoplankton biomass on temperature stratification in the large, monomictic Lake Constance was undertaken. Phytoplankton biomass affects the light extinction coefficient (LEC) of the water and, in turn, the vertical distribution of short-wave radiation, which shapes the temperature stratification in the lake.

2. A sensitivity study of the variation in LEC using the hydrodynamic model DYRESM showed that a high LEC is associated with stronger stratification, shallower thermoclines, higher surface temperatures and reduced heat content during the heating phase of the lake. During the cooling phase, a shallower thermocline at high LEC leads to a faster decrease in water temperature so that during autumn, a high LEC is associated with lower surface temperatures. Thermal structure was particularly sensitive to changes in LEC when its value was below 0.5 m−1.

3. When LEC is simulated dynamically with the coupled hydrodynamic–ecological model DYRESM-CAEDYM, its value is a function of phytoplankton dynamics that change vertically and temporally. Comparing simulations with and without dynamic LEC (i.e. with and without phytoplankton dynamics) produced a complex picture: during the vegetation period, we often found a warmer surface layer and colder water beneath in the simulations with dynamic LEC, as expected from the higher LEC when phytoplankton is abundant. However, since phytoplankton biomass (as LEC) fluctuates and because of occasional cooling phases, the patterns were comparatively weak and not consistent over the whole growing season.

4. The most obvious patterns emerged by comparing simulations of oligotrophic and eutrophic conditions. In the eutrophic state, with its higher LEC, stratification was stronger and characterized by higher surface water temperatures, a shallower thermocline and colder water temperatures between 5 and 10 m depth.

5. Statistical analysis of long-term data of water temperatures in Lake Constance, corrected for external forcing by air temperature, revealed a significant tendency towards warmer temperatures at 7.5 and 10 m depths with decreasing LECs over the course of reoligotrophication. This finding is consistent with our model results.