Seasonal and inter-annual scales of variability in phytoplankton assemblages: comparison of phytoplankton dynamics in three peri-alpine lakes over a period of 28 years

1. A method based on hierarchical clustering and Bayesian probabilities is used to identify phytoplankton assemblages and analyse their pattern of occurrence and temporal coherence in three deep, peri-alpine lakes. The hierarchical properties of the method allowed ranking by order of importance of the effects of changes related to climate and to human activity on the phytoplankton structure. 2. The three deep, peri-alpine lakes (the Lower Zurich, Upper Zurich and Walen lakes) investigated in this study have been monitored since 1972. During that period they have undergone oligotrophication as a result of management programmes and they have been subject to similar meteorological effects that have led to higher water temperatures since 1988. 3. The phytoplankton assemblages of the most eutrophic lake (Lower Zurich) differ strongly from those observed in the two meso-oligotrophic lakes. Local environmental conditions appear to be the main factor responsible for species composition and change in climate characterised by the warmer water temperatures observed since 1988 have had a major impact on the winter composition of the lower basin of Lake Zurich by promoting Planktothrix rubescens. 4. Some phytoplankton assemblages are found in all the lakes. Their patterns of occurrence display strong synchrony at the annual and/or inter-annual scales. However, temporal coherence between the lakes sometimes also involves different assemblages. 5. The reduction in phosphorus had a great influence on long-term changes in composition. In all three lakes, decreases in phosphorus are associated with a community characterised by some mixotrophic species or species adapted to low nutrient concentrations or sensitive to transparency. In the Lower Lake Zurich the decrease in phosphorus has also led to the development of species adapted to low light intensities. 6. Seasonal meteorological forcing has also induced synchronous changes, but the same assemblages are not necessarily involved, because the pool of the well-placed candidate taxa that may develop is determined by the local environmental conditions, and mainly by phosphorus concentrations. In the most eutrophic lake, the seasonal pattern is characterised by a succession of more stages. However, the seasonal assembly dynamics involve the succession of species sharing common selective advantages that make them relatively stronger under these nutrient and light conditions.