Validation of microbial community structure and ecological functional parameters in an aquatic microcosm designed for testing genetically engineered microorganisms.

Microcosms were designed to facilitate studies of the fate, functioning, and ecological effects of microorganisms released into the aquatic environment. The microcosms were three-phase systems (sediment/water/air) with three compartments (a primary producer component, a herbivore grazer component, and intact sediment cores). The microcosms were validated by comparing gross ecological parameters and microbial community structure between the microcosms and the eutrophic Lake Bagsværd, which was simulated in the model. The photosynthetic potential and chlorophyll a concentrations were significantly lower in the microcosms than in the lake, which apparently was due to inorganic nutrient limitation. In the microcosms, total bacterial numbers and metabolic activity by [(3)H]thymidine incorporation were unaffected by the reduced algal biomass and primary production, simulating field conditions closely, with a strong dependence on temperature. Two days after filling the microcosms, the percentage of similarity of the microbial communities in the microcosm and Lake Bagsværd was 40%, measured by hybridizations of total microbial DNA. The similarity increased during the 10-day experimental period to 63-76%. In two experiments, Alcaligenes eutrophus AEO106(pRO101) was released into the microcosms. The release reduced the similarity between microcosms and lake to 2% and 27%, depending on the number of introduced cells. Concomitant to a decline in the A. eutrophus AEO106(pRO101) population, the similarity gradually recovered. It is concluded that the microcosms can simulate a freshwater lake ecosystem, but care has to be taken when extrapolating microcosm results to the source ecosystem because of the possible different selective conditions in the microcosm.