Seasonal and interannual variability of picocyanobacteria in Lake Constance (1987-1997).

The ecology of chroococcoid picocyanobacteria was studied from 1987 to 1997 in large, deep, mesotrophic Lake Constance in relation to various abiotic and biotic factors that may influence their population dynamics. Picocyanobacteria dominated the autotrophic picoplankton (APP) numerically in this lake at all depths and times. Their abundances did not respond unequivocally to the decline of winterly phosphorus concentrations by a factor of 2.5 during the decade of investigation. They showed a recurrent seasonal pattern with peaks in spring and late summer, interspersed by a pronounced minimum during and after the clear-water phase around June. The magnitude, timing, and number of peaks and troughs which varied interannually could in part be related to weather conditions or the impact of other plankton groups. Larger phytoplankton and picocyanobacteria exhibited a distinct and predictable response to the vertical mixing intensity during early spring. Except for 1993, picocyanobacteria and larger phytoplankton decreased simultaneously during the mass development of daphnids in late May or June which gave rise to the clear-water phase. As the daphnid development depends more strongly on surface water temperature than on vertical mixing intensity, an early onset of stratification may imply a longer spring development which contributed to a higher seasonal average of picocyanobacterial abundances in 1989-1991. The decline in picocyanobacteria around the clear-water phase was often more pronounced and lasted longer than did the decline in larger algae. The rate of decrease may be related to daphnid abundance, however, no such relationship existed in respect to its duration. Summer peaks of picocyanobacteria were recorded despite the presence of relatively high densities of daphnids. We conclude that with the exception of the clear-water phase grazing control by nano- and microzooplankton may be more important for controlling picocyanobacterial numbers than is grazing by daphnids. Picocyanobacteria declined in autumn prior to or concomitant with larger algae without any obvious relationship to phytoplankton biovolume or the extent of vertical mixing within the uppermost 20 m. The as yet unexplained variation in the population dynamics of picocyanobacteria points to the significance of species-specific protist grazing and to shifts in picocyanobacterial species composition which should be tackled in future studies