A diesel fuel-contaminated aquifer was bioremediated in situ by the injection of oxidants (O 2 and NO 3 ؊) and nutrients in order to stimulate microbial activity. After 3.5 years of remediation, an aquifer sample was excavated and the material was used (i) to isolate bacterial strains able to grow on selected hydrocarbons under denitrifying conditions and (ii) to construct a laboratory aquifer column in order to simulate the aerobic and denitrifying remediation processes. Five bacterial strains isolated from the aquifer sample were able to grow on toluene (strains T 2 to T 4 , T 6 , and T 10), and nine bacterial strains grew on toluene and m-xylene (strains M 3 to M 7 and M 9 to M 12). Strains T 2 to T 4 , T 6 , and T 10 were cocci, and strains M 3 to M 7 and M 9 to M 12 were rods. The morphological and physiological differences were also reflected in small sequence variabilities in domain III of the 23S rRNA and in the 16S rRNA. Comparative sequence analyses of the 16S rRNA of one isolate (T 3 and M 3) of each group revealed a close phylogenetic relationship for both groups of isolates to organisms of the genus Azoarcus. Two 16S rRNA-targeted oligonucleotide probes (Azo644 and Azo1251) targeting the experimental isolates, bacteria of the Azoarcus tolulyticus group, and Azoarcus evansii were used to investigate the significance of hydrocarbon-degrading Azoarcus spp. in the laboratory aquifer column. The number of bacteria in the column determined after DAPI (4,6-diamidino-2-phenylindole) staining was 5.8 ؋ 10 8 to 1.1 ؋ 10 9 cells g of aquifer material ؊1. About 1% (in the anaerobic zone of the column) to 2% (in the aerobic zone of the column) of these bacteria were detectable by using a combination of probes Azo644 and Azo1251, demonstrating that hydrocarbon-degrading Azoarcus spp. are significant members of the indigenous microbiota. More than 90% of the total number of bacteria were detectable by using probes targeting higher phylogenetic groups. Approximately 80% of these bacteria belonged to the ␤ subdivision of the class Proteobacteria (␤-Proteobacteria), and 10 to 16% belonged to the ␥-Proteobacteria. Bacteria of the ␣-Proteobacteria were present in high numbers (10%) only in the aerobic zone of the column. Diesel fuel-contaminated soils and aquifers can be partially remediated by pumping hydrocarbons occurring in free phase back to the soil surface or by stripping the subsurface with air (7). Residual hydrocarbons, however, are often trapped in cracks and pores of the subsurface, and they may be removed by in situ bioremediation. This technique is usually based on the infiltration of water supplemented with oxidants (e.g., O 2 and NO 3 Ϫ) and/or nutrients (e.g., NH 4 ϩ and PO 4 3Ϫ) to stimulate the catabolic activity of microorganisms in the subsurface and thereby the biodegradation of the hydrocarbons (18, 23-25, 32). An in situ bioremediation process was applied in a diesel fuel-contaminated aquifer in Menziken, Switzerland (23). Groundwater supplemented with O 2 (329 M) and NO 3

The in situ bioremediation of aquifers contaminated with petroleum hydrocarbons is commonly based on the infiltration of groundwater supplemented with oxidants (e.g., 0 2 , N0 3 ) and nutrients (e.g., NH1, Pol-). These additions stimulate the microbial activity in the aquifer and several field studies describing the resulting processes have been published. However, due to the heterogeneity of the subsurface and due to the limited number of observation wells usually available, these field data do not offer a sufficient spatial and temporal resolution. In this study, flow-through columns of 47-cm length equipped with 17 sampling ports were filled with homogeneously contaminated aquifer material from a diesel fuel contaminated in situ bioremedia tion site. The columns were operated over 96 days at 12°C with artificial groundwater supple mented with 0 2 , N0 3 and POJ-. Concentration profiles of 0 2 , N0 3 , N0 2 , dissolved inorganic and organic carbon (DIC and DOC, respectively), protein, microbial cells and total residual hydrocarbons were measured. Within the first 12 cm, corresponding to a mean groundwater residence time of < 3.6 h, a steep 0 2 decrease from 4.6 to < 0.3 mg 1-1 , denitrification, a production of DIC and DOC, high microbial cell numbers and a high removal of hydrocarbons were observed. Within a distance of 24 to 40.5 cm from the infiltration, 0 2 was below 0.1 mg 1-1 and a denitrifying activity was found. In the presence and in the absence of 0 2 , n-alkanes were preferentially degraded compared to branched alkanes. The results demonstrate that: (1) infiltra tion of aerobic groundwater into columns filled with aquifer material contaminated with hydrocar bons leads to a rapid depletion of 0 2 ; (2) 0 2 and N0 3 can serve as oxidants for the mineralization of hydrocarbons; and (3) the modelling of redox processes in aquifers has to consider denitrifying activity in presence of 0 2 . ' Corresponding author.

Bacterial and protozoan communities were examined in three cores (A, B and C) from an aquifer located at an abandoned refinery near Hünxe, Germany. Cores were removed along a transect bordering a plume containing various monoaromatic hydrocarbons. Monoaromatic hydrocarbons could not be detected in the unsaturated zone in any core but were present in the saturated zones of core C (between 280 and 42 600 μmol kg−1 of core material [dry wt.]) and cores A and B (between 30 and 190 μmol kg−1 of core material [dry wt.]). Xylene isomers accounted for 50–70% of monoaromatic hydrocarbons in all cores. The number of DAPI‐stained bacteria was found to increase from the low‐contaminated cores A and B (approx. 0.1×108 cells and 0.2×108 cells g−1 of core material [dry wt.], respectively) to the high‐contaminated core C (2.4×108 cells g−1 of core material [dry wt.]). The higher bacterial numbers in core C were found to coincide with a higher detection rate obtained by in situ hybridization using probe Eub338 to target the domain Bacteria (13–42% for core C as compared to 3–25% for cores A and B, respectively). Proteobacteria of the δ‐subdivision (which includes many sulfate‐reducing bacteria) were the most predominant of the groups investigated (7–15% of DAPI‐stained bacteria) and were followed by Proteobacteria of the γ‐ and β‐subdivisions (4% and 1% of DAPI‐stained bacteria, respectively). The total numbers of protozoa and bacteria determined by direct counting occurred in a ratio of approx. 1:103, which was independent of depth or core examined. Most probable number analysis combined with a subsequent classification of the culturable protozoa revealed nanoflagellates as the major component of the protozoan community. Naked amoebae became increasingly more encysted with depth, except in the high‐contaminated core C where vegetative trophozoites were present in the saturated zone. The co‐occurrence of bacteria and protozoa in association with high concentrations of monoaromatic hydrocarbons suggests the involvement of trophic interactions in the process of biodegradation.

No abstract