The explicit engineering of bacterial populations requires that we know which organisms perform which tasks. The comparison of the bacterial diversity of activated sludge plants may give important information about the functions of different bacteria. This difficult task may be made easier by the use of technologies based on 16S rRNA based techniques. In this study we have used denaturing gradient gel electrophoresis (DGGE) to determine the optimal sampling regime for comparative studies and used cluster analysis to show how plants may be quantitatively compared. We sought evidence of spatial, diurnal and intrasample variation in a number of sites. No evidence for variation was found in the plants studied and we concluded that a single sample of an activated sludge plant was sufficient for a plant to plant comparison. The cluster analysis was able to distinguish between plants, though further work is required to find the most appropriate basis for such comparisons. We found organisms from raw sewage in the mixed liquor samples, these organisms may have no functional significance in the treatment process and thus complicate plant to plant comparisons as will the probable presence of heteroduplex rDNA products. Nevertheless we believe that these drawbacks do not outweigh the advantages of being able to take and compare relatively large numbers of samples.
Optimal sampling of dynamical large deviations via matrix product states