Morphological variations of Asterionella Formosa were studied both in field samples from Bautzen Reservoir, Germany, and in culture experiments. It was observed that this diatom exhibited considerable variation in the number of cells per colony, ranging from unicells to colonies with more than 16 cells. The winter period of two consecutive years which differed with respect to ice- and snow cover were compared (1995: no ice cover; 1996: all ice cover lasting 4 months). A. formosa was obviously inedible for Daphnia in winter 1995 whereas in the same period in 1996 this species represented a main food source for Daphnia galeata, as was demonstrated by microscopic observation of the gut content of D. galeata during periods of Asterionella blooms. From feeding experiments with C-14-labelled A. formosa of different colony sizes and from microscopic observations of Daphnia feeding in a suspension of Asterionella Colonies, I confirmed that only Asterionella colonies with up to 7 cells were effectively ingested by D. galeata. Above this threshold, the ingestion efficiency decreased to less than 30%. Steady-state populations of A. formosa in culture experiments showed a marked dependence on the number of cells per colony on light intensities. The mean colony size was usually less than 5 when light was limited (fell below a threshold of 20 mu E m(-2)s(-1)), and greater than 8 when light intensities exceeded 110 mu Em(-2)s(-1). In Bautzen Reservoir, low light penetration owing to snow-covered ice in winter 1996 resulted, similar to culture experiments, in Asterionella colonies with 4 or less than 4 cells. Contrary to 1996, non-limited light penetration corresponded to 8-celled colonies in winter 1995. Light limitation during periods of ice cover should therefore be a strong environmental predictor of the colony size of A. formosa and, consequently, of the ingestion efficiency of a Daphnia population, thereby strongly modifying trophic interactions in pelagic food webs. (C) 2008 Elsevicr GmbH. All rights reserved.