Microbial diversity in contaminated soil

Little is known about soil population dynamics: how microbial communities are affected by environmental stress; whether they respond adversely, or adjust to it; what the extent and duration of these adjustments might be; which genera of bacteria are the most reliable indicators of soil quality. Inevitably, it is beyond the scope of this thesis to resolve all these issues, but it nonetheless sets out to address specific objectives which should cumulatively enhance our total understanding of the soil microbial environment. A three year study of abiotic environmental stress, using a distinctive multi-disciplinary methodology, examined how pseudomonad communities react to chronic metal pollution from sewage sludge. It deployed three profiling methods: population size, catabolic and genetic diversity, across three sequential sampling times, and processed large numbers of bacterial isolates to facilitate meaningful data analysis. This process required innovative methodologies. Efficient analysis of numerous pseudomonad isolates necessitated semi-automation, by adapting ERIC profiling to run on Genescan, a specialist application for analysing sequence data, hitherto unused to fingerprint soil isolates. Specialist computer programmes were designed to compare multiple isolates. A database system was built to gather ERIC profiles and convert them to generate standardised data for subsequent analysis. The main body of this thesis focuses on the interpretation of genetic relationships, largely derived from ERIC profiles, for which a series of programs was built. Alongside them, the BIOLOG™ technique is deployed to profile catabolic diversity and thus compare the function of environmental samples at different sampling times. Population composition, a third dimension, was examined using conventional plate counts. Initial findings that population diversity might be affected by metal stress were not, in the end, corroborated by the extended study. Both catabolic and population studies proved inconclusive, highlighting only subtle differences between metal-contaminated plots. Soil pseudomonads, in contrast to rhizobia, failed to prove sensitive indicators of low level metal contaminants.