Microbial population changes during bioremediation of an experimental oil spill.

Three crude oil bioremediation techniques were applied in a randomized block field experiment simulating a coastal oil spill. Four treatments (no oil control, oil alone, oil plus nutrients, and oil plus nutrients plus an indigenous inoculum) were applied. In situ microbial community structures were monitored by phospholipid fatty acid (PLFA) analysis and 16S rDNA PCR-denaturing gradient gel electrophoresis (DGGE) to (i) identify the bacterial community members responsible for the decontamination of the site and (ii) define an end point for the removal of the hydrocarbon substrate. The results of PLFA analysis demonstrated a community shift in all plots from primarily eukaryotic biomass to gram-negative bacterial biomass with time. PLFA profiles from the oiled plots suggested increased gram-negative biomass and adaptation to metabolic stress compared to unoiled controls. DGGE analysis of untreated control plots revealed a simple, dynamic dominant population structure throughout the experiment. This banding pattern disappeared in all oiled plots, indicating that the structure and diversity of the dominant bacterial community changed substantially. No consistent differences were detected between nutrient-amended and indigenous inoculum-treated plots, but both differed from the oil-only plots. Prominent bands were excised for sequence analysis and indicated that oil treatment encouraged the growth of gram-negative microorganisms within the alpha-proteobacteria and Flexibacter-Cytophaga-Bacteroides phylum. alpha-Proteobacteria were never detected in unoiled controls. PLFA analysis indicated that by week 14 the microbial community structures of the oiled plots were becoming similar to those of the unoiled controls from the same time point, but DGGE analysis suggested that major differences in the bacterial communities remained.