1. Long-term records of air temperature and ice phenology (ice duration), and phyto- and zooplankton time series (1979–1997) were used to study the effects of ice duration on the successional pattern within plankton communities during spring in a shallow polymictic lake.
2. Water temperature in March was significantly lower after cold winters when compared to average or mild winters. Mean water temperature in April was not significantly different after mild, average or cold winters, but showed an overall significant negative correlation with ice duration.
3. Ice duration affected the timing and the magnitude of the peak abundance of diatoms, rotifers and daphnids during spring, but had no direct effects on the timing and maximum of chlorophytes, cryptophytes, cyanobacteria, bosminids and cyclopoid copepods.
4. Plankton groups which appeared first in the seasonal succession (i.e. diatoms, rotifers and daphnids) reached maximum abundance earlier after mild and average winters. The peak abundance of diatoms was negatively correlated with ice duration, whereas that of rotifers and daphnids was independent of the conditions during the preceding winter.
5. Temperature alone was generally a poor predictor of the timing and magnitude of both phyto- and zooplankton maxima. Turbulence may be important in the timing and the magnitude of peaks in diatoms, while total algal biomass was the most important determinant for the timing of the rotifer maximum. The magnitude of the daphnid maxima were significantly influenced by water temperature in March and April, and by rotifer abundance. The magnitude of the bosminid maximum was correlated with food availability and predation, whereas the timing of the maximum was more closely related to water temperature in May.
6. We conclude that, as a result of the low heat storage capacity of shallow lakes, the effects of winter on planktonic communities are short lived, and soon overtaken by the prevailing weather and by biotic interactions.