Abstract: The relationship between biodiversity and ecosystem functioning has emerged as a major issue in ecology. Several experiments have provided evidence for the notion that decreasing species diversity may impair ecosystem processes. The interpretation of these experiments, however, has been controversial because two types of mechanisms, “niche complementarity” and “sampling effect”, may operate. By means of microcosm experiments, we investigated the relationship between species diversity and ecosystem functioning. Aquatic microcosms in laboratory conditions are designed to contain different numbers of unicellular algal species, with no overlap in species composition between microcosms with the same number of species. Algal dry mass, an index of community productivity, was measured once a week after four weeks from the start of the experiment. The experiment lasted for a period of eight weeks, and in total, five measurements were obtained. Algal species richness had progressively stronger effects on community productivity. A number of higher-diversity mixtures outperformed the most productive monoculture as the experiment proceeded. Certain species dominated some mixtures in initial stages, but this effect disappeared in later stages. There was no significant correlation between reliability of community productivity and species richness. The experimental results reported here support the recent suggestion that both niche complementarity and sampling effect may be involved in the diversity effect, and show that the relative importance of the two mechanisms may undergo substantial changes through the course of community maturation.