Notes: Abstract Species composition and diversity of phytoplankton were studied for several years in two lakes which differ with respect to mixing conditions and nutrient limitation: Schlachtensee regularly stratifies very stably. In contrast, size and wind-exposure predispose Lake Tegel to deeper mixing; additionally, stratification is artificially destabilized by aeration. As the duration of aeration was varied, the study period includes interannual changes in mixing conditions. For both lakes, it also covers trophic change due to restoration; this was especially pronounced in Schlachtensee. Results show that mixing conditions affect species composition on two levels: on a superordinate level, lake morphology or hydrology govern stability of stratification and susceptibility to perturbation, and hence the extent to which motile species can develop. In Schlachtensee, species with some means of actively seeking preferred depths usually dominated during summer stratification: Planktothrix agardhii during the hypertrophic phase, and flagellates since restoration. In contrast, in Lake Tegel deeper mixing as a generally prevailing condition favored non-motile species. Their seasonal pattern was remarkably constant from year to year. Although changes in the extent of mixing were pronounced during the four years studied (1987–1990), these were within a range that affected species composition only slightly: in summer, cyanobacteria and diatoms represented ‘climax species’ whose dominance was not offset by additional, weather-induced increases of turbulence. On a subordinate level, and within the constraints set by nutrient limitation as well as by grazing pressure, small-scale changes in mixing conditions caused by meteorological cycles were shown to strongly affect species composition and in consequence diversity: Results for the fouryear post-restoration study period at Schlachtensee show that considerable interannual variations of species composition and diversity can be attributed to variations in the frequency of meteorological changes. In accordance with the ‘intermediate disturbance hypothesis’ (IDH), diversity was lowest during 1989, the year with the longest and most pronounced cycles of fair weather (14 to 27 fair days on end). However, the mechanism for this was rarely a decline of diversity caused by competitive exclusion within single long phases of stable conditions, as conceived by the ‘intermediate disturbance hypothesis’. Instead, diversity responded to changes in mixing conditions with a variety of patterns — often with low values during phases of increased mixing and with high values under quiescent conditions, especially during the first calm days just after increased mixing. Thus, not disturbance as such, but rather the rate of change between phases of disturbance and quiescence appears to determine the frequency of high diversity indices. In Lake Tegel, high diversity indices were somewhat more frequent in 1989, the year during which thermal stratification was most stable. For species adapted to frequent or continuous mixing, interjected calm phases with unusually high stability of thermal stratification may represent a ‘disturbance’ of accustomed conditions. Thus, in turbulent Lake Tegel, meteorological cycles appear to act in reversal to the IDH, in a sense which may be termed ‘intermediate quiescence hypothesis’. Phosphorus limitation due to successful restoration was found to decrease winter and vernal diversity in Schlachtensee. Presumably, without nutrient constraints, new populations could grow more rapidly in response to the rapid changes of physical parameters during this season. In contrast, restoration has increased summer diversity, as phosphorus concentrations no longer allow the virtual monocultures of Planktothrix agardhii which prevailed previously. In Lake Tegel, the reduction of phosphorus concentration down to 60 µg/l P in 1989 limited biomass, but this level was still too high to significantly alter species composition or diversity.