Intensive pig production in developed countries generates a large amount of pig slurry. This biological effluent contains about 1010 microorganisms per ml (Rivière et al., 1974). Its microbial transformation over the farm management generates air pollution by emission of a great number of gas compounds (ammonia, green house gas, odours) (Zhu et al., 2000). Moreover, it may contain pathogenic microorganisms and present a potential risk to the recipient ecosystem and finally to human populations. The objective of the present work was to monitor the evolution of a pig slurry microbial community throughout its management process. Dominant microbial populations were followed with a molecular typing technique using small-subunit ribosomal DNA analysis: PCR-SSCP (Polymerase Chain Reaction and Single Strand Conformation Polymorphism electrophoresis) (Delbès et al., 2000) while faecal indicators were numerated with culturable techniques. Materials and methods Sampling was carried out on a commercial pig farm, holding 220 sows plus finishers, that produces about 4500 m3 of pig slurry per year. 54 samples were collected from December 2002 to June 2003: faeces, slurries and soils. Analysis of the slurry microbial community was performed by PCR amplification of the microbial 16S rDNA V3 region using either general bacterial primers or specific primers targeting the Clostridiacea, the Bacillus-Lactobacillus-Streptococcus (BSL), and the Cytophaga-Flexibacter-Bacteroides (CFB) groups according to Dabert et al. (2004). PCR products were separated and visualised by SSCP capillary electrophoresis on an ABI 310 genetic analyser. The microbial community appears as a profile of peaks that corresponds to dominant microbial populations of the ecosystem. Peaks migrating at the same position are assumed to correspond to the same species. Faecal indicator were monitored using plating on selective media or MPN (Most Probable Number) technique using normalised culturable media. Number of microorganisms was expressed as a function of the samples dry matter content determined by drying at 105°C. Results and discussion Dynamics of the pig slurry microbial community was followed by aligning the SSCP electrophoregrams obtained from all slurry samples. The bacterial SSCP profiles appeared very complex with a high number of peaks reflecting the pig slurry microbial diversity (black arrows, fig. 1a). The comparison of the SSCP profiles revealed an unexpected stability of the pig manure slurry microbial community during time within the storage pit. From December 2002 to March 2003, the profiles remained about the same with no major apparent changes in peak number and composition. As expected, no change in SSCP profiles was observed when manure went through the auger press. On the opposite, SSCP profiles observed from lagoon samples showed a slow evolution of the community through time (data not shown). Finally, none of the peaks present in the profile from the lagoon liquid used for spreading were found in soil profiles after spreading. This could be explained by the high dilution rate of manure microorganisms within soil at the time of spreading, and shows that we reached the detection limits of the PCR-SSCP approach. Globally, the same results were obtained with the analysis of the specific microbial groups Clostridiacea, BSL, and CFB The dynamics of the faecal indicator microorganisms observed using the cultural approach revealed also a relative stability of the numeration over time for the different sampling locations (data not shown). However, in the same way as before, numeration changed from one step of slurry management to another (figure 1b). Slight variations took place when the faeces were mixed with urine to form slurry, but important numeration decay happened principally between the slurry from the storage tank and the separated liquid and solid phases from the lagoon and compost respectively. Pig slurry spreading did not seem to change significantly soil numeration of the studied groups. Conclusion The microbial community of pig slurry was studied from faeces to soil spreading with two complementary approaches. The molecular approach allows the detection of uncultured dominant microorganisms, originating from the pork faecal flora, which persist through the different manure management steps. The cultural approach allows to monitor faecal indicators. Both approach conducted to the same conclusions: the manure slurry microbial community does not evolve rapidly during passive anaerobic storage. Significant changes occurs primarily when manure move from one step of the management process to another.