Biofuel components change the ecology of bacterial volatile petroleum hydrocarbon degradation in aerobic sandy soil

o dy ll J ne N ya, In Fuel biodegradability that e bl scar n d relative ease of ethanol compared to methyl ester biodegradation. Denaturing gradient gel electropho- ncreas eliance lso kn tion where fuels are spilled or leaked. Understanding the impact of a biofuel components on the microbial degradation of harmful petroleum compounds such as benzene, toluene, ethyl-benzene and xylenes (BTEX) is therefore important (Dakhel et al., 2003). While the biofuel component of blended fuels is typically less toxic Less is known about the microbial ecology of blended fuel degradation in soils (Osterreicher-Cunha et al., 2009). The biodegradation of petroleum hydrocarbons may effectively protect groundwater resources from diffuse pollution via the vapor phase or leachate in the case of smaller fuel spills retained in the unsaturated zone (Christophersen et al., 2005; Pasteris et al., 2002). Lysimeter (Dakhel et al., 2003) and field studies (Freitas and Barker, 2011) with ethanol containing fuels suggest that ethanol is retained in the unsaturated zone following small * Corresponding author. Contents lists available at Environment .e ls Environmental Pollution 173 (2013) 125e132 E-mail address: [email protected] (D. Werner). Energy Directive, has a target for renewables in the transport fuel mix, of 10% by 2020, subject to the sustainability of production. Renewable fuel usage has also increased because of environmental concerns about the fuel oxygenate MTBE which has been elimi- nated in the U.S. from 2005 onwards (2001a; 2001b). Because of current fuel standards and car manufacturer warranty limitations, the most likely scenario of the near-future fuel composition in Europe and the US is the blending of a low proportion of bioethanol (5e15%) with petroleum or biodiesel (up to 20%) with diesel. The ability of environmental microorganisms to degrade and metabolize fuel components is an intrinsic defense against pollu- petroleum hydrocarbon biodegradation in groundwater (Da Silva and Alvarez, 2002; Ma et al., 2011), and this may cause prolonged BTEX plumes (Ruiz-Aguilar et al., 2002). In groundwater, ethanol degradation was shown to result in microbial community structure changes (Nelson et al., 2010) which may decrease the relative abundance of BTEX degraders (Da Silva and Alvarez, 2002). The presence of ethanol may also reduce the rate of BTEX degradation per cell of known BTEX degrader because BTEX compounds are degraded by inducible enzymes that can be repressed when a more easily degradable substrate such as ethanol is available (Lovanh et al., 2002). Soil pollution Microbial ecology 1. Introduction Demand for renewable fuels is i new legislation seeks to reduce our r instance, EU directive 2009/28/EC, a 0269-7491/$ e see front matter � 2012 Elsevier Ltd. http://dx.doi.org/10.1016/j.envpol.2012.10.010 fuel types, ranging from 0.40 � 0.16 to 0.51 � 0.22 g of biomass carbon per gram of fuel carbon degraded. Inorganic nutrient availability had a greater impact on petroleum hydrocarbon biodegradation than fuel composition. � 2012 Elsevier Ltd. All rights reserved. ing worldwide because on fossil resources. For own as the Renewable and more readily biodegraded than petroleum hydrocarbons, there are concerns about an inhibition of the microbial petroleum hydrocarbon degradation in the presence of ethanol (Da Silva and Alvarez, 2002) or biodiesel (Lapinskiene et al., 2006). Ethanol degradation consumes electron acceptors and thereby impacts Biofuels Petroleum overall effects on VHP biodegradationwere minor, and average biomass yields were comparable between Keywords: resis (DGGE) of bacterial 16S rRNA genes revealed that each fuel mixture selected for a distinct bacterial community, each dominated by Pseudomonas spp. Despite lasting impacts on soil bacterial ecology, the Biofuel components change the ecology hydrocarbon degradation in aerobic san Abdulmagid Elazhari-Ali a, Arvind K. Singh b, Russe a School of Civil Engineering and Geosciences, Newcastle University, Newcastle upon Ty bDepartment of Biochemistry, North-Eastern Hill University, Shillong 793022, Meghala a r t i c l e i n f o Article history: Received 23 July 2012 Received in revised form 17 September 2012 Accepted 19 October 2012 a b s t r a c t We tested the hypothesis sandy soil is affected by th inorganic nutrients were monoaromatic hydrocarbo journal homepage: www All rights reserved. f bacterial volatile petroleum soil . Davenport a, Ian M. Head a, David Werner a,* E1 7RU, England, United Kingdom dia the biodegradation of volatile petroleum hydrocarbons (VPHs) in aerobic ending with 10 percent ethanol (E10) or 20 percent biodiesel (B20). When ce, competition between biofuel and VPH degraders temporarily slowed egradation. Ethanol had a bigger impact than biodiesel, reflecting the SciVerse ScienceDirect al Pollution evier .com/locate/envpol of the most favorable electron acceptor and focused instead on 550 mg/kg total phosphorus and 1.6% dry weight total organic carbon (all data from results and also with other reports in the literature (Dakhel et al., 2003; Pasteris each nutrient amended batch 1.8� 10�3 g of N in the form of NH4Cl and 1.8� 10�4 g enta of P in the form of KH2PO4 was added to give a C:N:P ratio of approximately 100:10:1. To keep nutrient amended batches aerobic, about 40.0 mL of headspace air were replaced with fresh air using gas-tight syringes on day 14. The amount of CO2 and VPH vapors removed was measured and considered in the yield calculations. 2.4. Mini-lysimeter experiments Stainless steel mini-lysimeter buckets (height 20 cm, diameter 15 cm) were filled with soil to a height of 5 cm. A volume of 35 mL of PP, E10, or B20 fuel was mixed into the soil to obtain a source zone with a residual non-aqueous phase liquid (NAPL) saturation qNAPL of 4% (v/v), and the buckets were quickly filled with soil to a height of 15 cm, leaving 5 cm as headspace. The mini-lysimeters were closed with stainless steel lids with a clay seal and had sampling ports to withdraw gas samples from the headspace, and an inlet and outlet in the lid which could be sealed with a nut or connected to a multi-channel pump (Watson-Marlow Bredel Pumps, UK) to purge the headspace. The soil-filled volume of the mini-lysimeters had a volumetric water content of 0.23� 0.01 and a total porosity of 0.49� 0.02. The headspace of the three mini-lysimeters was simultaneously flushed on day 11, 21, 24, 30, 38, 44, 58, et al., 2002). 2.3. Batch experiments Batch experiments were performed in triplicate 65 mL amber vials for each treatment (Jencons, a VWR Division, Bedfordshire, UK) closed with Teflon Mininert valves (SigmaeAldrich Company Ltd. UK) containing 15.07 � 0.12 g wet weight of sandy soil with water content of 8.5 � 1.3% by weight. VPH concentrations were monitored after injecting 0.030 mL of the fuels as liquid (approximately 0.022 g). To Derwentside Environmental Testing Services Limited, Durham, UK). 2.2. Fuel mixtures A mixture of 12 typical fuel compounds (Table S1 in Supporting information) was prepared from chemicals of at least 99% purity, all obtained from Sigmae Aldrich, Gillingham, UK. The hydrocarbon mixture, made up of major constitu- ents of gasoline or kerosene, closely resembles that used in previous studies (Dakhel et al., 2003; Pasteris et al., 2002), and will be referred to as pure petroleum hydrocarbons, PP. The PP was blended with 10% (v/v) ethanol (Sigmae Aldrich, Gillingham, UK) to create ethanol blended fuel, E10, or with 20% (v/v) biodiesel (100% rapeseed oil methyl ester, 8.2 � 0.13% (w/w) palmitic acid methyl ester C16:0, 3.5 � 0.02% (w/w) stearic acid methyl ester C18:0, 40.3 � 0.32% (w/ w) oleic acid methyl ester C18:1, 38.9 � 0.20% (w/w) linoleic acid methyl ester C18:2, and 9.0 � 0.59% (w/w) linolenic acid methyl ester C18:3, from East Durham Biodiesel Ltd.) to create biodiesel blended fuel, B20. The PP composition is more typical for gasoline than diesel, but was also used as the petroleum hydrocarbon components of B20 to allow for cross-comparison with PP and E10 inorganic nutrient consumption and carbon substrate induced catabolic repression as shaping factors of soil bacterial community structure and activity. 2. Materials and methods 2.1. Soil characterization Sandy soil was obtained from a construction site on Newcastle University campus. It consisted of 87.3% sand, 11.3% silt and 1.4% clay. The soil contained by dry weight 7.3 mg/kg ammoniacal nitrogen, enta A. Elazhari-Ali et al. / Environm profiles were calculated for all samples using Pearson correlations between relative band intensities whichwere calculated as the intensity of an individual band divided by the sum of all band intensities of a sample. Cluster analysis, followed by Analysis of Similarity (ANOSIM) of DGGE profiles were performed using PRIMER-6 (Primer-E Ltd, UK). ANOSIM compares the rank similarity between pairs of replicates within a group (treatments) to the rank similarity of all pairs of replicates between groups (treatments) giving an R value: an R¼ 1 when all replicates within a group are more similar to each other than any replicates from different groups, and an R ¼ 0 if the similarities between and within groups (treatments) are on average the same. In the ANOSIM analysis, if treatment effects are studied based on triplicates there are only 10 distinct permutations and a significance level better than 10% could never be Fig. 1. (a,c,e) Extractable residual at the end of each experiment as percentage of the amo experiments; (c,d) Batch experiments with added inorganic nutrients; (e,f) Mini-lysimeter ex analyzed for the mini-lysimeter headspace toluene concentrations, n.d. ¼ not detected. l Pollution 173 (2013) 125e132 127 attained. Therefore, the difference level 10% was considered a significant level, when the R-value was high (close to 1). 3. Results and discussion 3.1. VPH degradation For the batch experiments without nutrient addition, the residual amounts of each VPH compound as a percentage of the unt added; and (b,d,f) toluene and ethanol headspace concentrations for (a,b) batch periments. Error bars represent the standard deviation of triplicates. No replicates were 3.3. Similarity of predominant bacterial communities The bacterial population from different batch treatments showed distinct DGGE profiles (Fig. 3) based on the number, rela- tive intensity and position of migration of the PCR products or bands. The relative intensity of the PCR product is suggested to be proportional to the abundance of the template and therefore the abundance of each population (Ovreas et al., 1997). Hence, the appearance and disappearance of bands, and changes in the relative intensity of the bands, in the DGGE profiles approximate shifts in the predominant bacterial community structure. DGGE band rich- ness increased in response to fuel addition, due to growth of new Table 1 Biomass and CO2 production and yield calculation for the batch incubation studies. Cell count (cells per g of soil) Cell size (mm3) Biomass C (% of C added) CO2 C (% of C added) Yield (Biomass C to biomass þ CO2 C) No nutrients added PP (8.3 � 2.8) $ 107 0.36 � 0.25 0.87 � 0.10 0.6 � 0.12 0.51 � 0.22 E10 (9.9 � 3.4) $ 107 0.36 � 0.25 0.90 � 0.37 1.2 � 0.18 0.40 � 0.16 B20 (1.4 � 0.1) $ 108 0.43 � 0.17 1.6 � 0.64 1.7 � 0.49 0.47 � 0.12 Nutrients added PP (4.8 � 1.2) $ 108 0.97 � 0.56 12.2 � 2.7 18.6 � 8.6 0.37 � 0.18 E10 (5.3 � 0.5) $ 108 1.00 � 0.12 13.83 � 2.5 25.7 � 7.8 0.35 � 0.04 ental Pollution 173 (2013) 125e132 amount added was statistically indistinguishable on day 24 between fuel treatments, except for m-xylenewhich was present in higher residual amounts in batches spiked with E10 compared to batches spiked with PP (ManneWhitney test, P < 0.05). Since E10 and B20 contain less petroleum hydrocarbons per unit volume due to dilution by the biofuel component, the remaining total petro- leum hydrocarbon mass was actually the highest for the PP mixture. Toluene was the only hydrocarbon no longer detected after 24 days in solvent extracts of batches without nutrient addi- tion (Fig.1a), but monitoring of headspace concentrations indicated that toluene was most rapidly degraded in batches with PP and least rapidly in batches with E10 (Fig. 1b), in which ethanol head- space concentrations decreased below the detection limit within 9 days. This trend is consistent with results reported by Lovanh et al. (2002) who have demonstrated in chemostat experiments that the metabolic flux of ethanol hinders that of toluene and other BTEX compounds. Addition of N and P to achieve a C:N:P ratio of approximately 100:10:1 allowed faster biodegradation to occur as is evident from the lower residual amounts found in solvent extracts after 24 days (Fig. 1c), and more rapid decline of vapor concentrations (shown for toluene as an example in Fig. 1d). This demonstrates that VPHs biodegradation was limited by inorganic nutrient availability. In comparison with the batch studies, the amount of fuel added per g of soil was much higher in the mini-lysimeter source zones at 0.0158, 0.0160 and 0.0158 g/g of soil for PP, E10 and B20 respectively, and the monitoring period was extended to 92 days. Occasional purging of the headspace kept the mini-lysimeter soils aerobic and allowed for some volatilization. Consequently, the most volatile VPHs were relatively more depleted in comparison with the batch study (Fig. 1a and e), but otherwise the findings of the batch study were confirmed. The residual amounts of each VPH compound as a percentage of the amount added was again statistically indistinguishable between fuel treatments, except for the percentage of m-xylene remaining after 92 days which was again significantly higher for E10 as compared to PP and this time also B20 (ManneWhitney test, P < 0.05). Headspace concentrations confirmed that toluene biodegradation was again fastest for PP and slowest for E10 (Fig. 1f). Ethanol headspace concentrations in the mini-lysimeter study fell to below the detection limit over a period of 18 days, which coincided with an apparent lag phase prior to the decrease of toluene concentration in the headspace (Fig. 1f). An approxi- mately 20-day lag phase followed by an intensive BTEX degra- dation phase during the next 9 days was also observed in soil slurry experiments with E10, E50 and E90 treatments reported by Lawrence et al. (2009). A biodiesel residual of 18.3 � 9.6% of the amount added remained in the B20 treated mini-lysimeter after 92 days, indicating that biodiesel is more slowly degraded than ethanol. 3.2. Cell growth, size and biomass yield Microbial growth on the added fuels was evident in all batches by greater biomass in fuels treated soils on day 24 as compared to the batches without fuel addition (controls) with no statistically significant difference between PP, E10 and B20 (Fig. 2). The average biomass yield (g cell C per g petroleum or biofuel C degraded) is comparable for PP, E10 and B20 (Table 1), and consistent with typical heterotrophic growth yields (Chiellini et al., 2007). Nutrient addition increased both total cell counts and average cell size (Table 1). The average yield with nutrient addition was again comparable for PP, E10 and B20, and somewhat lower thanwithout A. Elazhari-Ali et al. / Environm128 nutrient addition. A significant increase in the biomass can also be noticed in the mini-lysimeter study for all fuel treatments in comparison to the live soil without fuel addition which served as the control (non- parametric test KeW, P< 0.005). Total microbial cell carbon per g of soil on day 92 was significantly higher in the mini-lysimeter with E10 as compared to the lysimeter with PP (Fig. 2; t-test, P < 0.01). This difference, which was not observed in the 24 days batch study, suggests that E10 may sustain greater biomass than PP in the long- term. Österreicher-Cunha et al. (2004) also found higher hetero- trophic colony-forming unit (CFU) counts in gasoline-ethanol (22% by volume) contaminated, biovented (e.g. aerobic) soil as compared to pure gasoline contaminated soil one hundred days after contamination. Fig. 2. Cell carbon per gram of soil calculated from total cell counts, average cell size and an assumed carbon content of 310 fg per mm3. Cont ¼ live soil without fuel addition, d ¼ day sampled. B20 (7.5 � 1.0) $ 108 0.61 � 0.19 11.92 � 2.2 32.2 � 2.9 0.27 � 0.08 enta A. Elazhari-Ali et al. / Environm predominant bacterial community members (Table S2 in Supporting information). For the batch study, analysis of DGGE banding patterns gave separate clusters for the all the controls without fuel addition, and for PP without, B20 without and E10 with nutrient addition fuel treatments (Fig. 3). ANOSIM (details provided in Table S3 and S4 in Supporting information) of PP, B20 and E10 fuel treatments (i.e. excluding controls) revealed a significant global effect of nutrient addition (R ¼ 0.855, P ¼ 0.001) and to a lesser extent, fuel type (R ¼ 0.617, P < 0.005). Pairwise comparison shows that predomi- nant bacterial communities from all fuel treatments with and without nutrient addition were significantly different from their respective control treatments with no fuel addition (all R ¼ 1, all P ¼ 0.5 vs. > 0.1 respec- tively). Thus, scarcity of inorganic nutrients resulted in greater distinction amongst predominant bacterial communities growing on the different fuel types. For the lysimeter study, sampled on day 92, analysis of DGGE banding patterns grouped each fuel treatment in a separate cluster (Fig. 4). Additionally, source zone (SZ, liquid fuel and fuel vapors present) and top zone (TZ, only fuel vapors present) sub-clusters dy, with reference numbers inserted above the bands which were sequenced, and the enta A. Elazhari-Ali et al. / Environm130 were distinguished for the PP and E10 fuel treatments, and day 0 and day 92 sub-clusters were distinguished for the controls without fuel addition. Pairwise ANOSIM comparison (Table S5 in Supporting information) confirmed that each fuel treatment selected for a distinct predominant bacterial community (all R � 0.91, all P < 0.005 for (PP, E10), (E10, B20) and (PP, B20) respectively), which was also distinct from the predominant bacterial community in the unpolluted control soil (all R � 0.73, all P < 0.005). Fig. 4. DGGE image of PCR-amplified bacterial 16S rRNA gene fragments from the lysimeter NAPL contaminated layer, SZ ¼ source zone containing NAPL, d ¼ day sampled. l Pollution 173 (2013) 125e132 3.4. Predominant bacterial community members Searches of GenBank with the BLAST program (Altschul et al., 1997) were performed to determine the closest known relatives of the sequences recovered from the DGGE bands from the batch experiments. The closest matching sequences and their classifica- tion using the RDP classifier, are shown in Table 2 together with their GenBank accession numbers. All the recovered sequences had more than 98% identity with previously identified bacterial 16S study, and the corresponding cluster analysis. Abbreviations: TZ ¼ top zone above the s (re sest el udom ultu ultu udom ultu ultu ncsic udom ultu ncsic udom udom udom ultu ncsic udom ultu ultu ultu ultu teriu udom enta Table 2 Summary of the relative DGGE band intensities, nearest neighbor of cloned sequence from the batch experiments. Banda Percent band intensity relative to the total sample Clo No nutrients added Nutrients added PP E10 B20 PP E10 B20 Cut and sequenced strong bands which migrated to the same position in the g 1 31% Pse 2 21% Unc Unc 3 17% Pse Unc 4 34% Unc (Ta Pse 5 51% Unc (Ta Pse 6 39% Pse Pse 7 40% Unc (Ta Pse 8 27% Unc Unc 9 46% Unc Unc Additional cut and sequenced strong bands 10 28% Bac Pse A. Elazhari-Ali et al. / Environm rRNA gene sequences from uncultured organisms or cultured isolates. A similarity cut-off of 97% is often used for 16S rRNA genes/ gene fragments to identify taxa belonging to different species (Quince et al., 2009). All sequences were also classified using the RDP Classifier (Cole et al., 2009) as belonging to either the family Pseudomonadaceae (>95% confidence) or the genus Pseudomonas (>52% confidence), which are common and ubiquitous environ- mental bacteria, many of which possess a range of enzymes for petroleum hydrocarbon degradation (Tancsics et al., 2010; Zhang et al., 2011) which may be regulated by carbon catabolite repres- sion (Morenoet al., 2009). Aftermultiple alignment, sequences from bands 1e9, which migrated to the same position on the gel (Fig. 3), were shown to share 100% identity in regions of shared sequence (116 bp). The read length obtained from different bands varied slightly (122e169 bp), and these differences in sequence length led to the identification of different closest matching sequences in the BLAST search (Table 2). Interestingly, the most common identity of the closest relative of these sequences was an uncultured cloned partial sequence from the clone library of a petroleum hydrocarbon (BTEX) contaminated groundwater dominated by Pseudomonas spp. (Tancsics et al., 2010). Analysis using the RDP Classifier (Wang et al., 2007) identified this sequence as belonging to the genus Pseudo- monas (mean confidence level of 52%). Without nutrient addition, microbial communities from batches containing ethanol or biodiesel are characterized by the appear- ance of additional and/or stronger bands in the DGGE gels compared to the control and PP batches. These additional strong bands (bands 10e12 in Table 2 and Fig. 3) would likely be linked to 11 28% Pseudom Pseudom 12 18% Uncultu (Manes Uncultu a Numbers correspond to band numbers in Fig. 3 in Supporting information. b Closest relative of the sequence obtained from the DGGE band found in the GenBan c Between the sequence obtained from the DGGE band and the sequence of the closes d Classification according to the RPD classifier. For partial sequences of length shorter th to accurately classify sequences at genus level. ference provided if published) and RDP classification for the dominant excised bands relative (accession number)b Percent identityc RDP classificationd onas sp. (HQ652598.1) 100% Pseudomonas 85% red bacterium (DQ103582.1) 100% Pseudomonas 96% red bacterium (JF015032.1) 99% onas sp. (FR749860.1) 100% Pseudomonas 100% red bacterium (JF019062.1) 100% red bacterium (HM447076.1) s et al., 2010) 100% Pseudomonas 52% onas andersonii (HM581514.1) 100% red bacterium (HM447076.1) s et al., 2010) 100% Pseudomonadaceae 95% onas andersonii (HM581514.1) 100% onas andersonii (HM581514.1) 99% Pseudomonadaceae 95% onas andersonii (HM581513.1) 99% red bacterium (HM447076.1) s et al., 2010) 100% Pseudomonadaceae 100% onas andersonii (HM581514.1) 100% red bacterium (GQ389198.1) 98% Pseudomonadaceae 96% red bacterium (GQ389188.1) 98% red bacterium (GU775327.1) 98% Pseudomonas 63% red bacterium (GU763439.1) 98% m ASFP-1 (HQ018734.1) 99% Pseudomonas 100% onas fluorescens (FR749873.1) 99% l Pollution 173 (2013) 125e132 131 species growing on the biofuels. They were also identified as coming from different Pseudomonas spp. (Table 2), which may explainwhy average biomass yields (Table 1) and percent VPHmass removal (Fig. 1a,c,d) were comparable for PP, E10 and B20 fuel treatments, despite their distinct bacterial communities. Given the predominance of Pseudomonas spp. in all the soil bacterial communities, it is plausible that repression of BTEX degradation in the presence of ethanol may have played a role. Lovanh et al. (2002) showed for Pseudomonas putida mt-2, Pseudomonas putida F1, and Pseudomonas mendocina KR1 strains, that the metabolic flux of ethanol hinders that of toluene. 4. Conclusions Ethanol was the most rapidly degraded compound for the E10 fuel treatment, whereas toluene was the most rapidly degraded compound in PP and B20 fuel treatments. When inorganic nutri- ents are scarce the discriminating effect of competition between the biofuel degraders and the VPH degraders appears to tempo- rarily slow monoaromatic hydrocarbon degradation in the pres- ence of biofuels, in particular ethanol. However, toluene biodegradation set in quickly once the ethanol in E10 fuel treat- ments had been biodegraded and inorganic nutrient availability had a greater overall impact on petroleum hydrocarbon biodegra- dation than fuel composition. Hence, the inhibitory effect of biofuel components on VPH biodegradability would likely be minor or transitional if small amounts of fuel are spilled and retained in the unsaturated zone. onas sp. (HQ710831.1) 100% Pseudomonas 100% onas mosselii (HQ202559.1) 100% red bacterium (GU326513.1) et al., 2011) 99% Pseudomonas 100% red bacterium (HM445212.1) 99% k database with reference, if published. t relative found in the GenBank database. an 250 bps (longer than 50 bps) a bootstrap cut-off of 50%was shown to be sufficient Appendix A. Supporting information Supporting information related to this article can be found at http://dx.doi.org/10.1016/j.envpol.2012.10.010. References Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.H., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25, 3389e3402. Benson, D.A., Karsch-Mizrachi, I., Lipman, D.J., Ostell, J., Wheeler, D.L., 2008. Gen- Bank. Nucleic Acids Research 36, D25eD30. Bill S.265, 2001a. Introduced to the U.S. Senate on May 15th. Bill S.892, 2001b. Introduced to U.S. Senate on February 6th. Chiellini, E., Corti, A., D’Antone, S., Billingham, N.C., 2007. Microbial biomass yield and turnover in soil biodegradation tests: carbon substrate effects. Journal of Polymers and the Environment 15, 169e178. Christophersen, M., Broholm, M.M., Mosbæk, H., Karapanagioti, H.K., Burganos, V.N., Kjeldsen, P., 2005. Transport of hydrocarbons from an emplaced fuel source experiment in the vadose zone at Airbase Værløse, Denmark. Journal of Contaminant Hydrology 81, 1e33. Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., Kulam-Syed- Mohideen, A.S., McGarrell, D.M., Marsh, T., Garrity, G.M., Tiedje, J.M., 2009. The Ma, J., Xiu, Z., Monier, A.L., Mamonkina, I., Zhang, Y., He, Y., Stafford, B.P., Rixey, W.G., Alvarez, P.J., 2011. Aesthetic groundwater quality impacts from a continuous pilot-scale release of an ethanol blend. Ground Water Monitoring and Reme- diation 31, 47e54. Manes, C.L.D., West, N., Rapenne, S., Lebaron, P., 2011. Dynamic bacterial commu- nities on reverse-osmosis membranes in a full-scale desalination plant. Biofouling 27, 47e58. Moreno, R., Marzi, S., Romby, P., Rojo, F., 2009. The Crc global regulator binds to an unpaired A-rich motif at the Pseudomonas putida alkS mRNA coding sequence and inhibits translation initiation. Nucleic Acids Research 37, 7678e7690. Muyzer, G., de Waal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Applied and Environ- mental Microbiology 59, 695e700. Nelson, D.K., LaPara, T.M., Novak, P.J., 2010. Effects of ethanol-based fuel contamination: microbial community changes, production of regulated compounds, and methane generation. Environmental Science & Technology 44, 4525e4530. Österreicher-Cunha, P., Vargas Jr., E.p.d.A., Guimarães, J.R.D., de Campos, T.M.P., Nunes, C.M.F., Costa, A., Antunes, F.d.S., da Silva, M.I.P., Mano, D.M., 2004. Evaluation of bioventing on a gasolineeethanol contaminated undisturbed residual soil. Journal of Hazardous Materials 110, 63e76. Osterreicher-Cunha, P., Vargas, E.D., Guimaraes, J.R.D., Lago, G.P., Antunes, F.D., da Silva, M.I.P., 2009. Effect of ethanol on the biodegradation of gasoline in an unsaturated tropical soil. International Biodeterioration & Biodegradation 63, 208e216. A. Elazhari-Ali et al. / Environmental Pollution 173 (2013) 125e132132 Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Research 37, D141eD145. Da Silva, M., Alvarez, P.J., 2002. Effects of ethanol versus MTBE on benzene, toluene, ethylbenzene, and xylene natural attenuation in aquifer columns. Journal of Environmental Engineering ASCE 128, 862e867. Dakhel, N., Pasteris, G., Werner, D., Höhener, P., 2003. Small-volume releases of gasoline in the vadose zone: impact of the additives MTBE and ethanol on groundwater quality. Environmental Science and Technology 37, 2127e2133. Davenport, R.J., Curtis, T.P., 2004. Quantitative Fluorescence in Situ Hybridisation (FISH): Statistical Methods for Valid Cell Counting, Molecular Microbial Ecology Manual. Kluwer Publishing. Freitas, J.G., Barker, J.F., 2011. Oxygenated gasoline release in the unsaturated zone e part 1: source zone behavior. Journal of Contaminant Hydrology 126, 153e166. Fry, J.C., 1990. Direct methods and biomass estimation. Methods in Microbiology 22, 41e85. Hale, S.E., Meynet, P., Davenport, R.J., Martin Jones, D., Werner, D., 2010. Changes in polycyclic aromatic hydrocarbon availability in River Tyne sediment following bioremediation treatments or activated carbon amendment. Water Research 44, 4529e4536. Höhener, P., Dakhel, N., Christophersen, M., Broholm, M., Kjeldsen, P., 2006. Biodegradation of hydrocarbons vapors: comparison of laboratory studies and field investigations in the vadose zone at the emplaced fuel source experiment, Airbase Vaerlose, Denmark. Journal of Contaminant Hydrology 88, 337e358. Lapinskiene, A., Martinkus, P., Rebzdaite, V., 2006. Eco-toxicological studies of diesel and biodiesel fuels in aerated soil. Environmental Pollution 142, 432e437. Lawrence, A., Jonsson, S., Börjesson, G., 2009. Ethanol, BTEX and microbial community interactions in E-blend contaminated soil slurry. International Biodeterioration & Biodegradation 63, 654e666. Lovanh, N., Hunt, C.S., Alvarez, P.J.J., 2002. Effect of ethanol on BTEX biodegradation kinetics: aerobic continuous culture experiments.Water Research 36, 3739e3746. Ovreas, L., Forney, L., Daae, F.L., Torsvik, V., 1997. Distribution of bacterioplankton in meromictic Lake Saelenvannet, as determined by denaturing gradient gel electrophoresis of PCR-amplified gene fragments coding for 16S rRNA. Applied and Environmental Microbiology 63, 3367e3373. Pasteris, G., Werner, D., Kaufmann, K., Hohener, P., 2002. Vapor phase transport and biodegradation of volatile fuel compounds in the unsaturated zone: a large scale lysimeter experiment. Environmental Science & Technology 36, 30e39. Quince, C., Lanzen, A., Curtis, T.P., Davenport, R.J., Hall, N., Head, I.M., Read, L.F., Sloan, W.T., 2009. Accurate determination of microbial diversity from 454 pyrosequencing data. Nature Methods 6, 639e641. Ruiz-Aguilar, G.M.L., Fernandez-Sanchez, J.M., Kane, S.R., Kim, D., Alvarez, P.J.J., 2002. Effect of ethanol and methyl-tert-butyl ether on monoaromatic hydro- carbon biodegradation: response variability for different aquifer materials under various electron-accepting conditions. Environmental Toxicology and Chemistry 21, 2631e2639. Selinummi, J., Seppala, J., Yli-Harja, O., Puhakka, J.A., 2005. Software for quantifi- cation of labeled bacteria from digital microscope images by automated image analysis. Biotechniques 39, 859e863. Tancsics, A., Szabo, I., Baka, E., Szoboszlay, S., Kukolya, J., Kriszt, B., Marialigeti, K., 2010. Investigation of catechol 2,3-dioxygenase and 16S rRNA gene diversity in hypoxic, petroleum hydrocarbon contaminated groundwater. Systematic and Applied Microbiology 33, 398e406. Wang, Q., Garrity, G.M., Tiedje, J.M., Cole, J.R., 2007. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology 73, 5261e5267. Zhang, Z.Z., Hou, Z.W., Yang, C.Y., Ma, C.Q., Tao, F., Xu, P., 2011. Degradation of n- alkanes and polycyclic aromatic hydrocarbons in petroleum by a newly iso- lated Pseudomonas aeruginosa DQ8. Bioresource Technology 102, 4111e4116. Biofuel components change the ecology of bacterial volatile petroleum hydrocarbon degradation in aerobic sandy soil 1. Introduction 2. Materials and methods 2.1. Soil characterization 2.2. Fuel mixtures 2.3. Batch experiments 2.4. Mini-lysimeter experiments 2.5. Soil gas sampling and analysis 2.6. Non-aqueous phase liquid (NAPL) residuals quantification 2.7. Total microbial cell number 2.8. Bacterial community analysis 2.9. Statistical methods 3. Results and discussion 3.1. VPH degradation 3.2. Cell growth, size and biomass yield 3.3. Similarity of predominant bacterial communities 3.4. Predominant bacterial community members 4. Conclusions Appendix A. Supporting information References