Resumen

The evaluation of a biological aerated filter for the bioremediation of fuel oxygenates MTBE, ETBE and TAME, was the goal of the study at hand. In order to comply with the study, a set of pre-steps were needed, thus the steps followed were: selection of fuel ether biodegrading strains, set-up and design of a pilot-biofilter, and finally, the evaluation of seeded biofilters with theoretically recent contaminated groundwater. We selected a total of 9 bacterial strains, and tested their biotransformation capacities on individual ethers, and assayed their microbial count up to 21 days. Good growth curves were not obtained, but we chose Acinetobacter calcoaceticus M10, Rhodococcus ruber E10 and Gordonia amicalis T3 for further studies.
An evaluation of the GC method used was performed, since conventional methods (purge and trap or headspace autosamplers), were unavailable in our lab. We concluded from a storage test with abiotic and biotic samples, that the ether was hydrolysed more quickly in abiotic samples than their biotic counterparts. We were also able to obtained LODs and MDLs similar and in some cases better than those in the literature with headspace autosamplers.
In addition, we also evaluated the most adequate DNA extraction method from 0.22 ¿m nitrocellulose filter after filtration of groundwater. Varying DNA extraction protocols were tested in order to understand the different outcomes and reproducibility in samples examined through TGGE. Mechanical lysis with the FastDNA kit was one of the most reproducible protocols, exclusive bands appeared in with this method, which did not appear in the others, and thus was this protocol was used for biofilm and filter samples.
The three chosen strains were evaluated further for growth and biotransformation of metabolites TBA and TAA. Their physiological activity was also tested using flow cytometric analysis in order to evaluate whether they were viable and active, even when no growth or poor growth had been observed. The highest activity was observed with strain E10 in the presence of MTBE for 24 h.
EPS production was assessed on all three strains, however, only M10 produced a significant amount of EPS, therefore we decided on performing attachment assays on the biofilter designed with the individual strains, and the dual strain consortia M10-E10 and M10-T3, since strain M10 was the only observed to attach with several layers to the support material. The consortia M10-E10 attached successfully to the support material (Bioflow 9®), therefore we decided on using strain M10 and consortium M10-E10 further to study their ability in removing MTBE, using the designed downflow pilot-scale biofilter.
Three biofilters were tested at the same time in order to evaluate the oxygenate removal with different inocula, but with the same influent. Biofilter D was seeded with A. calcoaceticus M10, biofilter E with the consortium M10-E10, and a control biofilter (F) was mounted to evaluate the performance of a non-seeded biofilter. Among the biofilters, only biofilter E showed oxygenate removal tendency. Toxicity levels were also lowered when removal of MTBE was shown, thus no toxic metabolites were being accumulated.
The functionality of the biofilter in removing MTBE was sought by amplifying the genomic DNA of all biofilters with primers specific to alkB and ethB, previously found to biotransform MTBE. Though no ethB genes were found, alkB amplification was positive in all biofilters and in strains M10 and E10. Bacterial isolates from the biofilms of the biofilters were cultivated, identified taxonomically and tested for the alkB gene. Seven of the 13 isolates were found to contain the gene, however, none of the isolates were found among the TGGE bands from RNA extracts except for strains E10 and M10. We believe the consortium A. calcoaceticus M10 and R. ruber E10 was the only bacterium biodegrading MTBE in biofilter E, and should therefore be used further, after seeking optimization of the environmental conditions for growth and MTBE degradation to improve its function.