Microbial Processes Influencing the Attenuation and Impacts of Fuel

Abstract

Microbial processes influencing the attenuation and impacts of ethanol blend

fuel releases

Jie Ma

Fuel releases that impact groundwater are a common occurrence, and the growing use of

ethanol as a transportation biofuel is increasing the likelihood of encountering ethanol in such releases. Therefore, it is important to understand how such releases behave and affect public safety and environmental health, and how indigenous microorganisms respond and affect their migration, fate, and overall impacts.

Vapor intrusion risk (i.e., methane explosion and enhanced fuel hydrocarbon vapor intrusion) associated with ethanol blend releases is a potential concern. Using both experimental measurements and mathematical model simulations, this thesis shows that methane is unlikely to build up to pose an explosion hazard (5% v:v) if diffusion is the only mass transport pathway through the unsaturated zone. However, if methanogenic activity near the source zone is sufficiently high to cause advective gas transport, the methane indoor concentration may exceed the flammable threshold. As a group of widely distributed microorganisms, methanotrophs can significantly attenuate methane migration through the vadose zone, and thus alleviate the associated explosion risk. However, methane biodegradation could consume soil oxygen that i

would otherwise be available to support biodegradation of volatile hydrocarbons, and increase

their vapor intrusion potential.

The release of an ethanol blend solution (10 % v:v ethanol solution mixed with 50 mg/L benzene and 50 mg/L toluene) experiment into a pilot-scale (8 m3) aquifer tank produced a large

amount of volatile fatty acids (VFAs). The accumulation of VFAs (particularly butyric acid)

exceeded the secondary maximum contaminant level value for odor, which represents a

previously unreported aesthetic impact. After the release was shut off, ethanol anaerobic

degradation was temporarily stimulated when the dissolved ethanol concentration decreased

below its toxicity threshold (~2,000 mg/L for this system). Methane generation persisted for

more than 100 days after the disappearance of dissolved ethanol. The persistent methane was

likely generated from ethanol degradation byproducts (e.g., acetate) and solid organic carbon in

aquifer materials. Ethanol blend releases stimulate the microbial growth and increased the

organic carbon content in the aquifer.

Microorganisms play a critical role in the fate of ethanol-blended fuel releases, often

determining their region of influence and potential impacts. This thesis used advanced molecular

tools including 454 pyrosequencing and real-time PCR (qPCR) to characterize changes in

structure of indigenous microbial communities in response to 1) a pilot-scale ethanol blend

release and to 2) the shut-off of such release. This thesis shows that the ethanol blend release

stimulated microbial growth and significantly changed the microbial community structure by

enriching microbial groups involved in the fermentative degradation process. The growth of

putative hydrocarbon degraders and commensal anaerobes, and increases in degradation rates ii

suggest an adaptive response that increases the potential for natural attenuation of ethanol blend

releases. After the release was shut off, the microbial community returned towards the pre-

contaminated state; however, restoration was relatively slow and far from complete even one

year later.

Overall, this thesis advanced current understanding of vapor intrusion risks and

groundwater quality impacts associated ethanol blend releases and microbial ecology in the

impacted aquifer. The integration of this knowledge with site-specific information on pertinent hydrogeological processes will undoubtedly enhance engineering practices such as site investigation, risk assessment, and bioremediation implementation and maintenance to deal with releases of current and future biofuel blends.

iii

Acknowledgements

I would like to thank my advisor Dr. Pedro Alvarez for his unwavering support, unconditional trust and outstanding example of how to become a good scientist. I wish to thank

Dr. Bill Rixey for his long-term collaborations and constant help in this study. My thanks also go to other members of my dissertation committee, Dr. Qilin Li and Dr. George Bennett for their valuable inputs and suggestions on the development of this dissertation. My research collaborators Yi Zhang, Dr. Carlos W. Nossa, Dr. George DeVaull, Dr. Hong (Emma) Luo and

Dr. Brent Stafford provide a lot of time and effort on this study. I wish to thank Dr. Qiyou Jiang for his help with the shared computing facilities at Rice University and Dr. .Jan Hewitt for her help on the dissertation revisions. My office mate Dr. Zongming Xiu gave me a lot of guidance in doing research and beyond. I am also grateful to all other labmates and colleagues who helped make my time at Rice. Most importantly, I wish to thank my parents, who have always supported and loved me.

This work was funded by the American Petroleum Institute. I also received a stipend from the China Scholarship Council.

Based on this research, the following papers are published or submitted or in preparation:

Ma, J., Xiu, Z., Monier, A., Mamonkina, I., Zhang, Y., He, Y., Stafford, B., Rixey, W. and Alvarez, P. (2011) Aesthetic Groundwater Quality Impacts from a Continuous Pilot-Scale Release of an Ethanol Blend. Ground Water Monitoring & Remediation 31(3), 47–54.

Ma, J.,Rixey, W.G., DeVaull, G.E., Stafford, B.P. and Alvarez, P.J.J. (2012) Methane bioattenuation and implications for explosion risk reduction along the groundwater to soil surface pathway above a plume of dissolved ethanol. Environmental Science & Technology 46(11), 6013–6019.

Ma, J., Rixey, W.G., and Alvarez, P.J.J. Microbial processes influencing the transport, fate and groundwater impacts of fuel ethanol releases. Current Opinion in Biotechnology 24(3): 457-466.

Ma, J., Nossa, C.W., Xiu, Z., Rixey, W.G. and Alvarez, P.J.J. Adaptive changes in microbial community structure in response to a continuous pilot-scale release of an ethanol blend. Environmental Pollution 178(0): 419-425.

Ma, J., Luo, H., DeVaull, G.E., Rixey, W.G.,Alvarez, P.J.J.A numerical model investigation for potential methane explosion and benzene vapor intrusion associated with high-ethanol blend releases. Environmental Science & Technology (minor revision)

Ma, J., et al., Response to the shut-off of a pilot-scale ethanol blended release: increased ethanol degradation activities and persistent methanogenesis. (in preparation)

Ma, J., et al., Pyrosequencing-based investigation for microbial response to a 2-years ethanol blended release and the shut-off of such release. (in preparation)

Table of Contents

Abstract ...... i

Acknowledgements ...... iv

List of Figures ...... xi

List of Tables ...... xv

Chapter 1 Introduction...... 1 1.1 Problem statement ...... 1 1.2 Objectives and hypothesis ...... 4 1.3 Dissertation outline ...... 6 1.4 Significance and potential benefits of this study ...... 9

Chapter 2 Literature Review ...... 10 2.1 Physical behavior of ethanol-blended fuel releases ...... 10 2.2 Biodegradation of ethanol-blended fuel ...... 12 2.3 How would ethanol affect BTEX degradation? ...... 14 2.3.1 Gene expression ...... 16 2.3.2 Metabolic flux dilution ...... 17 2.3.3 Thermodynamic inhibition ...... 17 2.3.4 Cell physiology ...... 18 2.3.5 Community structure ...... 19 2.3.6 Overall effect of ethanol on BTEX plume dynamics ...... 22 2.4 Existing knowledge gaps ...... 25 2.4.1 Methane explosion risk ...... 25 2.4.2 Enhanced potential for benzene vapor intrusion ...... 26 2.4.3 VFAs accumulation ...... 26 2.4.4 Microbial ecology ...... 26

Chapter 3 Methodology ...... 28 3.1 Pilot-scale aquifer tank ...... 28 3.2 Releases experiment ...... 34 3.3 Chemical analytical methods ...... 35 3.3.1 Methane ...... 36 3.3.2 Ethanol degradation byproducts ...... 37 vi

3.3.3 Ethanol ...... 38 3.3.4 Benzene and toluene ...... 38 3.3.5 Bromide tracer ...... 38 3.4 Groundwater geochemical parameters monitoring ...... 39 3.5 Quantitative real-time PCR analysis ...... 39 3.6 Pyrosequencing ...... 41 3.7 Geochip ...... 45 3.7 Biovapor 1-D analytic vapor intrusion model ...... 47 3.8 Abreu and Johson 3-D numerical vapor intrusion model ...... 49

Chapter 4 Experimental and 1-D analytic model investigation on CH4 explosion and benzene vapor intrusion ...... 51 4.1 Introduction ...... 51 4.2 Materials and Methods ...... 53 4.2.1 Pilot-scale aquifer system ...... 53 4.2.2 Flux chamber description ...... 54

4.2.3 Sampling and analysis methods for CH4 and O2 ...... 55

4.2.4 Assessment of CH4 oxidation activity at different depths ...... 56 4.2.5 qPCR assays for pmoA gene ...... 56 4.2.6 Biovapor model simulation ...... 58 4.3 Results and Discussion ...... 61

4.3.1 CH4 accumulation in the flux chamber...... 61

4.3.2 Aerobic biodegradation of CH4 in the pilot-scale aquifer ...... 63

4.3.3 CH4 accumulation simulation ...... 67

4.3.4 Impacts of CH4 oxidation on benzene vapor intrusion ...... 69 4.4 Conclusion ...... 71

Chapter 5 A 3-D numerical model investigation for methane explosion and benzene vapor intrusion potential associated with ethanol-blended fuel releases ...... 73 5.1 Introduction ...... 73 5.2 Materials and Methods ...... 76 5.2.1 3-D numerical model ...... 76 5.2.2 Simulated scenarios and model input parameters...... 77 5.2.3 Assumptions and limitations of this modeling study...... 84 5.3 Results and Discussion ...... 86

vii

5.3.1 The explosion risk for diffusion-driven CH4 migration is negligible ...... 86

5.3.2 The explosion risk increases significantly for advection-driven CH4 migration ...... 89 5.3.3 Oxygen consumption during CH4 biodegradation in the vadose zone increases benzene vapor intrusion potential ...... 95 5.3.4 Implications for site assessment and remedial action ...... 103

Chapter 6 Seasonal variation of ethanol fermentative degradation and aesthetic impact of volatile fatty acids generation ...... 106 6. 1 Introduction ...... 106 6.2 Materials and Methods ...... 107 6.2.1 Pilot-scale aquifer system ...... 107

6.2.2 CH4 and VFAs analysis ...... 108 6.3 Results and Discussion ...... 109 6.3.1 Effect of groundwater temperature on VFAs production ...... 109 6.3.2 VFAs odor generation ...... 111

6.3.3 Effect of temperature on CH4 production ...... 115 6.3.4 Ethanol, benzene, and toluene attenuation ...... 117 6.4 Conclusion ...... 118

Chapter 7 Response to the shut-off of a pilot-scale ethanol blended release: increased ethanol degradation activities and persistent methanogenesis ...... 119 7. 1 Introduction ...... 119 7.2 Materials and Methods ...... 121 7.2.1 Pilot-scale aquifer system ...... 121 7.2.2 Chemical analysis methods ...... 122 7.2.3 Microcosm experiment ...... 122 7.2.4 Sand organic carbon content ...... 123 7.2.5 Sand sample collection and DNA extraction ...... 123 7.2.6 qPCR analysis ...... 123 7.2.7 Microarray ...... 124 7.2.8 Microarray Data processing ...... 125 7.3 Results ...... 126 7.3.1 Microcosm experiments for ethanol toxicity ...... 126 7.3.2 Higher ethanol degradation activity following source removal ...... 129 7.3.3 Persistence of dissolved methane and methane metabolism genes ...... 131

viii

7.3.4 Ethanol releases stimulate microbial EPS production ...... 135 7.4 Discussion ...... 137 7.4.1 Higher ethanol degradation activity following source shutoff is due to lower ethanol toxicity ...... 137 7.4.2 Persistent methane accumulation may be due to anaerobic degradation of remaining acetate and black slime ...... 139

Chapter 8 Adaptive microbial population shifts in response to a continuous ethanol blend release increases natural attenuation potential ...... 141 8.1 Introduction ...... 141 8.2 Materials and Methods ...... 142 8.2.1 Pilot-scale aquifer system ...... 142 8.2.2 Chemical and geochemical analyses ...... 143 8.2.3 Sand sampling and qPCR analysis ...... 144 8.2.4 Pyrosequencing ...... 146 8.2.5 Pyrosequencing data analysis ...... 147 8.3 Results and Discussion ...... 148 8.3.1 Summary of sequencing data ...... 148 8.3.2 Impacts of ethanol-blend release on the abundance of selected functional genes ...... 148 8.3.3 Impacts of ethanol-blend release on microbial community structure...... 151 8.3.4 Effect of seasonal fluctuations in groundwater temperature ...... 154 8.3.5 Overall effects on biodegradation of the continuous ethanol-blend release ...... 156 8.4 Conclusion ...... 159

Chapter 9 Pyrosequencing-based investigation for microbial response to a 2-years ethanol blended release and the shut-off of such release ...... 160 9.1 Introduction ...... 160 9.2 Materials and Methods ...... 162 9.2.1 Pilot-scale tank and different experimental stages ...... 162 9.2.2 Groundwater chemical and geochemical analysis ...... 165 9.2.3 Sand sampling and analysis ...... 165 9.2.4 DNA extraction and pyrosequencing ...... 165 9.2.5 Sequence data processing ...... 166 9.2.6 Statistical analysis...... 167 9.3 Results and Discussion ...... 168

9.3.1 Dissimilarities in microbial community structure among different stages ...... 168 9.3.2 Taxonomic composition of archaeal communities ...... 170 9.3.3 Taxonomic composition of bacterial communities ...... 173 9.3.4 Species richness, evenness, and diversity ...... 179 9.4 Conclusions ...... 182

Chapter 10 Conclusions, engineering significance and recommendations for future research . 183 10. 1 Conclusions ...... 183 10. 2 Engineering significance ...... 185 10.3 Recommendations for future research...... 187

References ...... 189

Appendix I: Supporting information for Chapter 7 ...... 204

Appendix II. Supporting information for Chapter 9 ...... 208

List of Figures

Figure 1.1 World’s biofuel production (Year 2000-2012)...... 2 Figure 1.2 Annual production of fuel ethanol in US (Year 1980-2012)...... 3 Figure 2.1 Fate, transport, and potential impacts of ethanol-blended fuel releases...... 12 Figure 2.2 Simulated benzene plume dynamics (centerline reach) resulting from a 30-gal release of regular gasoline or various fuel alcohol blends ...... 24 Figure 3.1 A photo of the pilot-scale aquifer tank ...... 29 Figure 3.2 A photo of the pilot-scale tank ...... 29 Figure 3.3 Plan view of the aquifer tank...... 31 Figure 3.4 Profile view of the aquifer tank...... 31 Figure 3.5 Photos of tap water inlet (a) and ethanol/benzne/toluene injection port (b) of the aquifer tank...... 32 Figure 3.6 Groundwater sampling ports and monitoring wells (a) and groundwater outlet (b) of the aquifer tank...... 34 Figure 3.7 Timeline for the tank release experiment...... 35 Figure 3.8 HP5890 gas chromatograph for methane measurement...... 36 Figure 3.9 HP5890 gas chromatograph for the measurement of ethanol degradation byproducts...... 37 Figure 3.10 Applied Biosystems 7500 real-time PCR system ...... 41 Figure 3.11 Workflow of 454 pyrosequencing analysis ...... 43 Figure 3.12 Pyrosequencing reaction in PTP wells...... 44 Figure 3.13 454 Genome Sequencer Junior System...... 45 Figure 3.14 Schematic presentation of GeoChip development and operations for analyzing environmental DNA samples ...... 47 Figure 3.15 Conceptual model of “Biovapor”...... 48 Figure 3.16 Biovapor software interface...... 49 Figure 4.1 Plan view (a) and profile view (b) of the pilot-scale aquifer system...... 53 Figure 4.2 The structure schematic (left) and the photo of the flux chamber (right)...... 55

Figure 4.3 CH4 accumulation inside the flux chamber in different seasons...... 63

Figure 4.4 Vertical concentration profiles of CH4 and O2 ...... 64

Figure 4.5 Vertical distribution of pmoA gene concentration and CH4 biodegradation rate ...... 65 Figure 4.6 Correlation between methane biodegradation rate and pmoA gene concentration ...... 66

Figure 4.7 Methane biodegradation activity in microcosms prepared with soil samples from different depths ...... 67 Figure 4.8 Simulated methane indoor concentrations (a) with and (b) without methane biodegradation ...... 68 Figure 4.9 Simulated benzene indoor concentrations and the aerobic zone thickness ...... 71 Figure 5.1 Cross-sectional view of the model domain and the perimeter crack on the building foundation ...... 79 Figure 5.2 Plan view of the foundation with perimeter crack distribution ...... 79

Figure 5.3 Simulated CH4 indoor concentrations for different CH4 source concentrations and source depths ...... 86

Figure 5.4 Changes in CH4 mass flux emitted from the source (Jsource), flux into the building

(Jbuilding), flux across the soil surface (Jsurface), and flux biodegraded (Jbio) with different

CH4 source concentrations...... 88

Figure 5.5 Simulated CH4 indoor concentrations for different source depths and source gas pressures...... 90 Figure 5.6.Pressure field distribution with different source pressure for source depth of 5 m. .... 91

Figure 5.7 Simulated benzene indoor concentrations for different CH4 source concentrations and source depths...... 97

Figure 5.8. Changes in benzene flux emitted from source (Jsource), flux into the building basement

(Jbuilding), flux across the soil surface (Jsurface), and flux biodegraded (Jbio) for different

CH4 source concentrations...... 98

Figure 5.9 Changes in O2 consumption by aerobic degradation of benzene (B) and CH4 with

different CH4 source concentrations ...... 99

Figure5.10 Normalized benzene concentration distribution with different CH4 source concentration for a source depth of 8 m...... 100

Figure 5.11 Normalized CH4 concentration distribution with different CH4 source concentration for a source depth of 8 m...... 101

Figure 5.12 Normalized O2 concentration distribution with different CH4 source concentration for a source depth of 8 m...... 102 Figure5.13 Simulated indoor benzene concentrations for different source gas pressures and source depths...... 103 Figure 6.1 Plan view of the aquifer system...... 108 Figure 6.2 Acetic acid concentrations at sampling wells A1 ...... 110

xii

Figure 6.3 Correlations between acetic acid concentrations (measured at A1) and groundwater temperature...... 110 Figure 6.4. Butyric acid concentrations at sampling wells A1 ...... 111

Figure 6.5 CH4 concentration at sampling well A1 ...... 116

Figure 6.6 Correlations between CH4 concentrations (measured at A1) and groundwater temperature...... 116 Figure 6.7 Ethanol, benzene, and toluene attenuation at sampling well A1 ...... 118 Figure 7.1 Plan view (a) and profile view (b) of the aquifer tank...... 121 Figure 7.2 Methane accumulations in the headspace of microcosms...... 127 Figure 7.3 Dissolved concentrations of ethanol degradation byproducts in each microcosm after 133 days of incubation...... 129 Figure 7.4 Changes in the concentrations of ethanol and its degradation byproducts ...... 131

Figure 7.5 Comparison of CH4 concentrations with ethanol and its degradation byproducts concentrations ...... 133 Figure 7.6 Changes in the copy numbers of methane metabolism functional gene following the shutoff ethanol release...... 134 Figure 7.7 Sand total organic carbon content...... 135 Figure 7.8 Normalized signal intensity of detected functional genes for extracellular polymeric substance (EPS) production ...... 137 Figure 8.1 Plan view (a) and profile view (b) of the pilot-scale aquifer system...... 143 Figure 8.2 Seasonal changes in groundwater geochemical parameters following the release. .. 149 Figure 8.3 A comparison of the absolute (a) and relative (b) abundance of selected genes between baseline sample (t=0 day) and the sample following 10 months of continuous ethanol-blend release (t=318 days)...... 150 Figure 8.4 Changes in bacterial communities (genus level) before (a) and after (b) 10 months exposure to the release...... 153 Figure 8.5 Seasonal changes in archaeal community (genus level) following 10 months of exposure to the continuous ethanol-blend release...... 154 Figure 8.6 Seasonal variations of the abundance of functional genes mcrA (a) and fhs (b) with

groundwater temperature fluctuations, and the corresponding CH4 and acetate groundwater concentrations...... 155 Figure 8.7 Correlation between methane concentration in the groundwater and mcrA gene abundance...... 156

xiii

Figure 8.8 Correlation between acetate concentration in the groundwater and fhs gene abundance...... 156 Figure 8.9 Linear regression of normalized ethanol concentration versus time...... 157 Figure 8.10 Linear regression of normalized benzene concentration versus time...... 158 Figure 8.11 Linear regression of normalized toluene concentration versus time...... 158 Figure 8.12 Bio-attenuation of ethanol, benzene and toluene during the continuous release of ethanol-blended solution...... 159 Figure 9.1 Plan view (a) and profile view (b) of the aquifer tank...... 162 Figure 9.3 Two-dimensional Principal Coordinate Analysis (PCoA) plot based on weighted UniFrac distance matrix for both bacterial communities...... 169 Figure 9.4 Two-dimensional Principal Coordinate Analysis (PCoA) plot based on weighted UniFrac distance matrix for both archaeal communities...... 170 Figure 9.6 Changes in the relative abundance of five dominant archaeal taxa: ...... 172 Figure 9.7 Taxonomic composition of the bacterial community at Stage1...... 174 Figure 9.8 Taxonomic composition of the bacterial community at Stage2...... 175 Figure 9.9 Taxonomic composition of the bacterial community at Stage3...... 176 Figure 9.10 Taxonomic composition of the bacterial community at Stage 4b...... 178 Figure 9.11 Changes in the relative abundance of ten families that were dominant at Stage 1. 178 Figure 9.12 Alpha diversity indices of bacterial and archaeal communities...... 181

xiv

List of Tables

Table 2.1 Mechanisms by which ethanol affects BTEX degradation ...... 16 Table 2.2 qPCR primer/probe sets for BTEX degradation ...... 21 Table 2.3 Fate and transport models ...... 23 Table 4.1 Primers, target group, standard DNA and annealing temperature for qPCR ...... 57 Table 4.2 Model inputs for methane explosion simulation ...... 59 Table 4.3 Model inputs for benzene vapor intrusion simulation ...... 60

Table 4.4 Measured CH4 concentration and calculated surface flux ...... 62 Table 4.5 CH4 concentrations in groundwater ...... 62 Table 5.1 Simulated scenarios for Table 5.3-5.5 and Figure 5.3-5.13 ...... 81 Table 5.2 Model input Parameters ...... 82

Table 5.3 Simulated air flow rate and CH4 flux with ifferent source pressures for a source depth of 5 m ...... 92 Table 5.4 Simulated air flow rate (L/min) with different source pressures for depth of 3, 8, and 15 m ...... 93

Table 5.5 Simulated CH4 flux (g-CH4/m2/s) with different source pressures for depth of 3, 8, and 15 m ...... 94

Table 5.6 Measured CH4 flux in natural and contaminated environments ...... 95 Table 6.1. Calculated VFAs gas phase concentrations ...... 113 Table 6.2 VFAs threshold odor number ...... 114 Table 7.1 Primers and probes for qPCR analysis ...... 124 Table 7.2 Fitted methane generation rates in the microcosm experiment ...... 128 Table 8.1 Primers and probes for qPCR analysis ...... 145 Table 8.2 Sequencing result summary ...... 148 Table 9.1 General information for each experimental stage ...... 164

Chapter 1

Introduction

1.1 Problem statement

As an alternative energy source, biomass-derived fuel (i.e., biofuel) is increasingly being incorporated into the world’s energy grid. The potential benefits of using biofuel include alleviating dependence on imported oil and enhance energy security, mitigating air pollution and reducing greenhouse gas emissions by fossil fuel combustion, (FAPRI-ISU, 2012). Currently, there are two major types of commercialized biofuel: bioethanol and biodiesel. Bioethanol holds a much larger global market share than biodiesel (22,715 vs. 5,670 million gallons/year). Since year 2000, the world production of ethanol and biodiesel have increased 5- and 26-fold respectively and will continue to increase in future (Figure 1.1) (FAPRI-ISU, 2012).

2.5x104 Ethanol Biodiesel 2.0x104

1.5x104

1.0x104 Million gallons

5.0x103

0.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year

Figure 1.1 World’s biofuel production (Year 2000-2012).

In United States, ethanol is the most commonly used biofuel which accounts 3.4% of the total transportation fuel consumption. The most commonly used ethanol blend fuel in US is E10

(i.e., gasoline blended with 10% (v:v) ethanol) which has become standard at most of the nation's

180,000 gas stations. In 2012, E15 (i.e., gasoline blended with 15% (v:v) ethanol) was approved by the U.S. Environmental Protection Agency (EPA) for light duty vehicles model year 2001 and newer, and all flex-fuel vehicles (FFVs) and has been sold in several states in US since then.

Designed for flexible fuel vehicles, E85 was sold at 2414 public gasoline stations located in 1701 cities by October 2010 in US. The annual production of fuel ethanol in US has increased from

2,130 million gallons in 2002 to 10,600 million gallons in 2009, with an average growth rate of

25.8% per year (Figure 1.2). Studies have estimated that ethanol and other biofuels could replace

30% or more of U.S. gasoline demand by 2030. 2

15000

12000

9000

6000 Million gallons

3000

0 1980 1985 1990 1995 2000 2005 2010

Year

Figure 1.2 Annual production of fuel ethanol in US (Year 1980-2012). Data are obtained from Renewable Fuels Association (www.ethanolrfa.org).

Accidental and incidental release of petroleum fuels that impact the soil and groundwater environments is a common occurrence and the likelihood of encountering biofuels (mainly ethanol) in such releases is increasing. Therefore, it is crucial to understand how ethanol blend fuel behaves and how such release affects soil and groundwater environment, human health, and public safety. Since microorganisms play a key role in the attenuation of fuel or ethanol-blend fuel constituents (including both ethanol and BTEX), often determining their region of influence and fate (Powers et al., 2001). It is important to characterize the structure, function, diversity of indigenous microbial population in aquifer impacted by ethanol blend fuel. This information is critical to optimize site characterization, risk assessment and remediation practices when dealing with releases of current and future biofuels blends. 3

1.2 Objectives and hypothesis

The general goal of this study is to elucidate the potential impacts to groundwater quality

and overlying enclosed space associated with fuel ethanol releases, and to determine how

microorganisms adapt to attenuate such release. Specifically, we seek to:

1) Characterize CH4 generation, upward migration and the associated explosion risk in a

pilot aquifer system impacted by a fuel ethanol release. This includes investigating the spatial distribution and the effect of methanotrophic activity on the fate and transport of subsurface CH4.

Hypothesis: methanotrophs can significantly attenuate CH4 flux through the vadose zone, thus alleviating its explosive hazard.

2) Investigate the effects of CH4 biodegradation on the fate and transport of benzene in subsurface environments and determine whether and to what extent methanotrophic activity could increase the potential for benzene vapor intrusion.

Hypothesis: aerobic degradation of CH4 could compete for available O2 in the vadose zone and reduce the near-source attenuation for benzene, increasing benzene vapor intrusion potential.

3) Quantitatively investigate the potential impacts of pressure-driven advective soil-gas flow

(due to generation and accumulation of ethanol-derived CH4 and CO2 gas near the subsurface

source zone) on the migration of CH4 gas and benzene vapor and the associated vapor intrusion

risks.

Hypothesis: if the fermentative degradation of ethanol near the source zone is sufficiently high to

cause advective soil-gas flow, CH4 and benzene flux and the associated vapor intrusion risks

could be increased significantly.

4) Monitor changes in the concentrations of ethanol, benzene, toluene and ethanol

degradation byproducts (e.g., CH4, acetate, propionate and butyrate) and groundwater

geochemical footprints (e.g., pH and dissolved O2) following an ethanol blend release and the shut-off of such release. Quantify the volatile fatty acids (VFAs) accumulations and their seasonal variation within the context of potential aesthetic impacts to groundwater quality.

Hypothesis: Ethanol fermentation may result in high concentrations of VFAs which may cause aesthetic impacts on groundwater quality due to their associated odor and taste. Warm temperature would accelerate VFAs generation, thus such aesthetic impact would be much more severe in the summer. After the ethanol blend release was shut off, concentrations of benzene, toluene, ethanol and its degradation byproducts will gradually decrease to zero and the groundwater dissolved O2 and pH will increase.

5) Characterize the microbial response to ethanol blend releases and how such changes influence the biodegradation and attenuation of ethanol blend fuel releases.

Hypothesis: ethanol blend releases will enrich certain taxonomic group as well as functional

genotypes that are involved the biodegradation process and significantly change both taxonomic

composition and functional structure of indigenous microbial community. The growth of putative

hydrocarbon degraders and commensal anaerobes may enhance biodegradation of an ethanol-

blended fuel.

6）Characterize the microbial population shifts following the shut-off of ethanol blend release and determine whether and how quickly the indigenous microbial community is able to restore back to the pre-contaminated conditions.

Hypothesis: as residual contaminants are attenuated completely and the geochemistry of impacted groundwater changes back to the pre-release conditions, the structure of microbial community will tend to revert to initial conditions, but this process is relatively slow compared to the population shifts in response to the release.

1.3 Dissertation outline

Chapter 1 provides the background knowledge, objectives and hypothesis, and the rational and significance of this study. Chapter 2 reviews the recent studies and present understandings on the physical migration, biodegradation, and environmental impacts of ethanol blend fuel releases. Existing knowledge gaps and new research frontier are also included in Chapter 2. This chapter has been published as a critical review paper in Current Opinion in Biotechnology (Ma et al., 2013a). In Chapter 3, all major research methods used in this study are described in details.

The remaining chapters address the following research objectives of this study:

Chapter 4, entitled “Experimental and 1-D analytic model investigation on CH4 explosion

and benzene vapor intrusion”, used flux chamber measurements (emplaced on the pilot-scale 6

aquifer tank) and 1-D analytic model simulations (Biovapor) to addresses: 1) the explosion risk

associated with methanogenesis in the aquifer tank impacted by fuel ethanol; 2) the vertical

distribution and effect of methanotrophs on the upward migration and fate of CH4 in the tank;

and 3) the potential effect of oxygen consumption by methanotrophs on the benzene vapor

intrusion pathway. The results of this Chapter have been published in Environmental Science &

Technology (Ma et al., 2012).

Chapter 5, entitled “A 3-D numerical model investigation for methane explosion and benzene vapor intrusion potential associated with ethanol-blended fuel releases”, used a 3-D numerical model (Abreu and Johnson) to investigate: 1) the explosion risk for diffusion-driven

CH4 migration; 2) the explosion risk for advection-driven CH4 migration; 3) effects of

methanotrophs on vapor intrusion potential for CH4 and benzene; 4) benzene vapor intrusion risk

for advection-driven soil gas flow; 5) how classic site conceptual model for petroleum

hydrocarbon vapor intrusion might change when strong methanogenic activities is present in the

subsurface; 6) how current regulations and guidance to manage vapor intrusion risk should be

modified to deal with emerging ethanol-based biofuel releases. The results of this Chapter have

been submitted to Environmental Science & Technology.

Chapter 6, entitled “Seasonal variation of ethanol fermentative degradation activity and aesthetic impact of volatile fatty acids generation” shows the changes in concentrations of ethanol, benzene, toluene and ethanol degradation byproducts including CH4, acetate, propionate,

and butyrate) in response to a pilot-scale (8 m3) ethanol blend releases (10 % v:v ethanol solution 7

mixed with 50 mg/L benzene and 50 mg/L toluene). The volatile fatty acids (VFAs) accumulations and their seasonal variation were quantified within the context of potential aesthetic impacts to groundwater quality. The results of this Chapter have been published in

Ground Water Monitoring & Remediation (Ma et al., 2011).

Chapter 7, entitled “Response to the shut-off of a pilot-scale ethanol blended release: increased ethanol degradation activities and persistent methanogenesis”, shows two unexpected phenomena (i.e., temporal stimulation of ethanol fermentative degradation and persistent methanogenic activity) following the shutoff of the ethanol blend release in the model aquifer tank. The engineering implications of these findings are also discussed. The results of this

Chapter are used for a manuscript in preparation.

Chapter 8, entitled “Adaptive microbial population shifts in response to a continuous ethanol blend release increases natural attenuation potential”, uses 16S rRAN gene pyrosequencing and quantitative real-time PCR (qPCR) analyses to characterize changes in microbial community structure and selected functional gene abundance in response to a 10- month continuous release of an ethanol blend solution. The results of this Chapter have been published in Environmental Pollution (Ma et al., 2013b)

Chapter 9, entitled “Pyrosequencing-based investigation for microbial response to a 2- years ethanol blended release and the shut-off of such release”, uses 16S rRNA gene pyrosequencing and latest pyrosequencing data analysis algorithms (including the removal of

sequencing noise and chimeric sequences) to characterize changes in microbial taxonomic

composition following the shut-off of ethanol blend release and determine whether and how

quickly the indigenous microbial community is able to restore back to the pre-contaminated conditions. The results of this Chapter are used for a manuscript in preparation.

Finally，Chapter 10 wraps up the dissertation with a discussion of conclusion and recommended future research.

1.4 Significance and potential benefits of this study

This study is aim to advance our understanding on how vapor intrusion risk and groundwater quality associated with ethanol blend releases, and how indigenous microorganisms respond and affect the fate and impacts of such releases. The integration of this knowledge with site-specific information on pertinent hydrogeological processes will undoubtedly enhance

engineering practices such as site investigation, risk assessment, and bioremediation implementation and maintenance to deal with releases of current and future biofuel blends.

Chapter 2

Literature Review

[Modified from a review paper published in Current Opinion in Biotechnology (Ma et al., 2013a)]

Fuel releases that impact groundwater are a common occurrence, and the growing use of ethanol as a transportation biofuel is increasing the likelihood of encountering ethanol in such releases. Thus, it is important to understand how ethanol-blend releases behave and affect groundwater geochemistry, and how indigenous microorganisms respond and affect their migration, fate, and overall impact. This information is critical to optimize site characterization, risk assessment and remediation practices when dealing with releases of current and future biofuel blends. . This chapter summarizes current understanding on the biogeochemical footprint of ethanol blend fuel releases and the factors that influence their natural attenuation.

2.1 Physical behavior of ethanol-blended fuel releases

When an ethanol-blended fuel release occurs, it infiltrates as a non-aqueous phase liquid

(NAPL) through the unsaturated zone to the water table and forms a floating NAPL pool at the

water table when a sufficient volume is spilled (Figure 2.1). Ethanol will partition into pore

water throughout the unsaturated zone (McDowell et al., 2003; McDowell and Powers, 2003;

Freitas and Barker, 2011) and will tend to accumulate at the water table interface and the

capillary fringe due to its buoyancy (McDowell and Powers, 2003; Capiro et al., 2007; Molson et

al., 2008; Stafford et al., 2009; Yu et al., 2009b). For high content ethanol fuels (e.g., E95, which

has 95% ethanol and 5% gasoline by volume), the fuel will likely migrate through this interface,

initially as a water miscible phase, and then separate into two phases as the fuel becomes diluted,

precipitating a new NAPL phase along its path (Capiro et al., 2007; Stafford et al., 2009; Stafford

and Rixey, 2011). Pore water containing high ethanol concentrations will also be enriched in

hydrocarbons due to their enhanced solubility in the presence of ethanol (cosolvent effect)

(Heermann and Powers, 1998; Corseuil et al., 2004; Chen et al., 2008a; He et al., 2011). Thus, different domains of microbial activity are likely to develop: a region of anaerobic activity in the core of a contaminant plume in the saturated zone (where the biochemical oxygen demand [BOD] exerted by the release exceeds the available dissolved oxygen) with aerobic degradation occurring at the fringes of the plume; a second region of high anaerobic activity in the capillary zone (except in cases when ethanol concentrations are sufficiently high to be toxic to microbial processes); and a third region in the unsaturated zone where aerobic degradation of methane

(emanating from the anaerobic fermentation of ethanol in the capillary zone) is predominant

(Figure 2.1).

Figure 2.1 Fate, transport, and potential impacts of ethanol-blended fuel releases.

2.2 Biodegradation of ethanol-blended fuel

Direct exposure to ethanol in drinking water has minimal adverse impacts on human health, but ethanol may increase the exposure potential of toxic fuel constituents (i.e., benzene, toluene, ethylbenzene and xylenes [BTEX]) by hindering their biodegradation and increasing their region of influence (Powers et al., 2001). Because ethanol generally biodegrades faster than

BTEX, the latter tend to form larger and more persistent plumes than ethanol. Therefore,

substrate interactions during ethanol and BTEX degradation and their effect on plume dynamics

(range and longevity) have received considerable attention (Corseuil et al., 1998; Da Silva and

Alvarez, 2002; Deeb et al., 2002; Molson et al., 2002; Mackay et al., 2006; Zhang et al., 2006;

Chen et al., 2008b; Feris et al., 2008; Freitas et al., 2011a; Da Silva and Corseuil, 2012).

During transport in groundwater, ethanol and BTEX can undergo a series of biotransformations which can be performed by a variety of microorganisms in aerobic or anaerobic environments (Powers et al., 2001). The relatively high concentration of ethanol found in recently-impacted groundwater exerts a high BOD that rapidly consumes the available dissolved oxygen and other terminal electron acceptors in the vicinity of the source zone, which results in the development of strongly anaerobic, fermentative methanogenic conditions (Figure

2). Nevertheless, aerobic microbial activity might be important for the natural attenuation of the leading edge of the plume.

Under aerobic conditions, BTEX are activated by oxygenases to form catechol or structurally related compounds, which subsequently undergo ring fission to byproducts such as acetyl-CoA, acetaldehyde and pyruvic acid that enter central metabolic pathways such as Krebs’ cycle (for final mineralization to CO2) or glycolysis (Alvarez and Illman, 2005). Ethanol can also

be aerobically metabolized to the pivotal intermediate acetyl-CoA via acetaldehyde and acetate

(Powers et al., 2001).

Under anaerobic conditions, BTEX are initially transformed via different pathways

(fumarate addition, O2-independent hydroxylation, and carboxylation) to a common aromatic

intermediate, benzyl-CoA, which subsequently undergoes ring reduction followed by hydrolytic

cleavage (Fuchs et al., 2011). Further anaerobic transformations in anaerobic (methanogenic) food webs eventually produce acetate, which is finally mineralized by acetoclastic methanogens to produce CH4 and CO2. BTEX fermentation also generates H2, which is consumed by different commensal anaerobes, including hydrogenotrophic methanogens. Ethanol is similarly transformed to acetate and H2, which are subsequently metabolized by methanogens to produce

CH4 and CO2 (Powers et al., 2001). Depending on the available electron acceptors, sulfate reducers, iron reducers, and denitrifiers could also participate in the anaerobic degradation of ethanol-blended fuel, and spatially distinctive redox zones could form in plume (Figure 2).

2.3 How would ethanol affect BTEX degradation?

The major impact from ethanol may be related to its inhibitory effect on BTEX biodegradation (Table 2.1), which (depending on the release scenario) may increase the likelihood of BTEX to reach receptors (longer plumes) as well as the potential duration of exposure (more persistent plumes).

Benzene, which is the most toxic compound of the BTEX and often drives the need for cleanup action, is relatively resistant to degradation under anaerobic conditions (Foght, 2008), while ethanol and its degradation byproducts (e.g., volatile fatty acids [VFAs]) are easier to degrade under both aerobic and anaerobic conditions (Powers et al., 2001). The preferential degradation of ethanol and its degradation byproducts may deplete available O2 that would otherwise be available for aerobic benzene degraders, hindering their activity. Therefore,

accelerated oxygen depletion is one of the most important inhibitory mechanisms of ethanol on

benzene degradation (Corseuil et al., 1998; Da Silva and Alvarez, 2002; Deeb et al., 2002).

Although the initial steps of ethanol and BTEX degradation (including both aerobic and anaerobic pathways) are catalyzed by different enzymes under different pathways, their degradation may eventually converge to common intermediates (e.g., acetate and acetyl Co-A)

that enter central metabolic pathways (e.g., Krebs cycle) for final mineralization. Fast

degradation of ethanol may result in the accumulation of acetyl-CoA (inside the cell) and acetate

(mainly secreted in groundwater), which may hinder BTEX degradation by both intracellular

mechanisms (e.g., catabolite repression and/or metabolic flux dilution) and abiotic constraints

(decreased pH and/or thermodynamic inhibition) as discussed below.

Table 2.1 Mechanisms by which ethanol affects BTEX degradation Effects on BTEX Mechanisms System Affected degradation Catabolite repression Gene expression - Metabolic flux dilution Metabolism - pH decrease Cell physiology - Ethanol toxicity Physiology & metabolism - Fortuitous growth of BTEX degraders Community structure + Genotypic dilution Community structure - Growth of syntrophic microorganisms Community structure + Increase richness and diversity Community structure + Electron acceptor/nutrients depletion Metabolism, kinetics - Thermodynamic inhibition due to VFAs Metabolism, kinetics - accumulation

2.3.1 Gene expression

Ethanol is metabolized by constitutive enzymes through a central metabolic pathway, while the initial step of BTEX degradation is usually catalyzed by inducible enzymes. To save the energy associated with the synthesis of inducible catabolic enzymes, which are not needed when ethanol is available, microorganisms are likely to consume ethanol preferentially (Madigan and

Martinko, 2006). Therefore, the presence of ethanol could repress the synthesis of inductive enzymes required for BTEX degradation (Powers et al., 2001), thus hindering BTEX degradation at the transcription level (Lovanh and Alvarez, 2004). Two independent experiments using different detection methods reported that ethanol (or its byproduct acetate) could repress the tod gene (coding for toluene dioxygenase) and that the degree of tod repression increased with the ethanol concentration (Lovanh and Alvarez, 2004; Da Silva and Alvarez, 2010). It should be

noted that catabolite repression is unlikely to occur under carbon-limiting conditions that are

conducive to simultaneous utilization of multiple substrates (Egli, 1995).

2.3.2 Metabolic flux dilution

Ethanol could hinder BTEX degradation by “metabolic flux dilution” (Lovanh et al., 2002;

Lovanh and Alvarez, 2004). The metabolic flux of a specific compound is analogous to the specific degradation rate and can be defined as the rate at which the compound is metabolized per unit biomass (Egli, 1995). Metabolic flux dilution is a form of non-competitive inhibition in which the utilization rate of one substrate decreases due to the metabolism of another that is not necessarily degraded by the same enzymes. For example, BTEX and ethanol are initially transformed by different pathways that eventually converge into common metabolic intermediates (e.g., acetyl-CoA). This could create a bottleneck that exerts feedback inhibition and decreases the degradation rate of a target compound (e.g., benzene). Whereas the utilization of ethanol would decrease the specific BTEX degradation rates, this does not preclude a potential enhancement in overall degradation rates due to additional (fortuitous) growth of BTEX degraders on ethanol (Lovanh et al., 2002). To illustrate simplistically, ten bacteria degrading

BTEX at 20% capacity would be faster than one bacterium working at 100% capacity.

2.3.3 Thermodynamic inhibition

The build-up of ethanol-derived acetate could thermodynamically hinder benzene degradation under methanogenic (Corseuil et al., 2011) and sulfate reducing (Rakoczy et al.,

2011) conditions. The degradation of BTEX under anaerobic fermentative conditions is

endergonic under standard conditions (Corseuil et al., 2011; Rakoczy et al., 2011), as illustrated

- + -1 for benzene: C6H6 + 6H2O → 3CH3COO + 3H + 3H2; ∆Gº´ = + 190.19 kJ.mol .

Therefore, syntrophic consumption of acetate and hydrogen is needed for the reaction to proceed, and the accumulation of ethanol-derived acetate at concentrations greater than about 64 mg/L makes this reaction thermodynamically unfavorable (Corseuil et al., 2011). The effects of ethanol on the dynamics of commensal populations that produce and consume acetate and hydrogen remain poorly understood.

2.3.4 Cell physiology

High concentrations of ethanol are toxic to microorganisms. Ethanol could dissolve phospholipids and disintegrate the cell membrane (Ingram and Buttke, 1984). As the cell membrane loses its structural integrity, ethanol could enter the cell and denature enzymes. High concentrations of ethanol could inhibit the synthesis of DNA (Osztovics et al., 1980), RNA

(Mitchell and Lucaslenard, 1980) and proteins (Haseltin.Wa et al., 1972), thus leading to loss of functions or even cell death (Ingram and Buttke, 1984). The inhibitory threshold of ethanol ranges between 10,000 and 100,000 mg/L for various microorganisms (Ingram and Buttke,

1984). Concentrations in this range are possible only near the source of relatively recent releases.

In poorly buffered aquifers, ethanol-derived VFAs could significantly decrease groundwater pH (pH < 5 in the core of the plume) (Ma et al.). Some microorganisms are very sensitive to pH changes. For example, the growth of methanogens is generally inhibited at pH < 6 (Powers et al.,

2001). Because methanogens consume thermodynamically inhibitory byproducts (e.g., H2 and

acetate) and play a key role in the fermentative/methanogenic mineralization pathway, low pH

could adversely affect anaerobic BTEX degradation.

2.3.5 Community structure

Ethanol could be consumed by a wide variety of microorganisms, including some BTEX

degraders. Thus, ethanol could fortuitously stimulate the growth of BTEX degraders and enhance

the potential for BTEX degradation (Ma et al.; Lovanh et al., 2002; Ruiz-Aguilar et al., 2002;

Capiro et al., 2008). Increases in the abundance of catabolic genes for aromatic hydrocarbons

degradation, such as bssA (coding for benzylsuccinate synthase) (Beller et al., 2008) and PHE

(coding for phenol hydroxylase) (Capiro et al., 2008; Ma et al., 2013b), were reported in systems

exposed to ethanol-blended fuel. However, more microbial species can feed on ethanol than on

BTEX, which is conducive to a greater proliferation of commensal microorganisms and a

decrease in relative abundance of BTEX degraders (genotypic dilution) (Da Silva and Alvarez,

2002). While genotypic dilution decreases specific BTEX degradation rates, overall degradation rates may increase due to higher total concentration of BTEX degraders (Lovanh et al., 2002;

Gomez and Alvarez, 2009), especially after ethanol is removed and its inhibitory effects have waned while a higher concentration of BTEX degraders remains.

Ethanol could also influence BTEX degradation kinetics by affecting the growth and activity

of syntrophic microorganisms. Anaerobic biodegradation of organic compounds is usually a

syntrophic process which involves the interaction and cooperation of different microbial groups.

Ethanol blend releases could stimulate the growth of commensal syntrophs that consume

inhibitory fermentation byproducts (e.g., H2 and acetate), thereby enhancing anaerobic bioremediation (Ma et al., 2013b).

Pristine aquifer ecosystems usually have very low biomass concentration because substrates are scarce (Griebler and Lueders, 2009). Ethanol blend releases increase substrate concentrations and the available metabolic niches, thus stimulating the growth of diverse species (Ma et al.,

2013b). Ecological resilience is generated by diverse but functionally overlapping species

(Peterson et al., 1998), and phylogenetic diversity and ecological resilience are usually positively correlated (Botton et al., 2006). Thus, the resulting increases in phylogenetic diversity enhance the resilience of groundwater ecosystems to bioremediate hydrocarbons remaining after ethanol is consumed as well as for recurring releases (Ma et al., 2013b).

Table 2.2 summarizes the most widely used quantitative real-time PCR (qPCR) primer sets for catabolic genes involved in aerobic and anaerobic degradation of BTEX. As a sensitive and reliable method to detect and quantify genes, qPCR may be very useful in establishing the presence of specific biodegradation potential and assessing biodegradation activities and bioremediation performance (Alvarez and Illman, 2005).

Table 2.2 qPCR primer/probe sets for BTEX degradation

Target Primer Sequence Function Reference enzyme Toluene 5’-ACCGATGARGAYCTGTACC-3' Aerobic degradation of (Baldwin et TOD dioxygenase 5’-CTTCGGTCMAGTAGCTGGTG-3' BTEX al., 2003) Xylene 5’-TGAGGCTGAAACTTTACGTAGA-3' Aerobic degradation of (Baldwin et TOL monooxygenase 5‘ -CTCACCTGGAGTTGCGTAC-3' xylene al., 2003) Toluene 5’-TCTCVAGCATYCAGACVGACG-3' Aerobic degradation of (Baldwin et RMO monooxygenase 5’-TTKTCGATGATBACRTCCCA-3' toluene al., 2003) Aerobic degradation of (Baldwin et PHE Phenol 5’-GTGCTGACSAAYCTGYTGTTC-3' BTEX in O limited al., 2003) monooxygenase 5’-CGCCAGAACCAYTTRTC-3' 2 environment Aerobic degradation of (Lillis et Catechol 2,3- 5’-AAGAGGCATGGGGGCGCACCGGTTCGATCA-3’ cat23 BTEX in O limited al., 2010) dioxygenase 5’-AACAAADGCGCSGTCATGCGG-3’ 2 environments Catechol 2,3- 5’-ATCGAGGCCTGGGGTGTGAAGACCACCATGCT-3’ (Yeh et al., cat23 The same as above dioxygenase 5’-CTCGTTGCGGTTGCCGCTSGGGTCGTCGAAGAAGT-3 2010) Catechol 2,3- 5’-AGGTGCTCGGTTTCTACCTGGCCGA-3’ (Phillips et cat23 The same as above dioxygenase 5’-ACGGTCATGAATCGTTCGTTGAG-3’ al., 2012) Anaerobic degradation (Beller et Benzylsuccinate 5’-ACGACGGYGGCATTTCTC-3’ bssA 5’-GCATGATSGGYACCGACA-3’ of toluene and xylene al., 2002) synthase FAM-5’CTTCTGGTTCTTCTGCACCTTGGACACC 3’-TAMRA (denitrifying) Anaerobic degradation (Staats et Benzylsuccinate 5’-TCGAYGAYGGSTGCATGGA-3’ bssA of toluene and xylene al., 2011) synthase 5’-TTCTGGTTYTTCTGCAC-3’ (iron-reducing) Anaerobic degradation (Beller et Benzylsuccinate 5’-GTSCCCATGATGCGCAGC-3’ bssA of toluene and xylene al., 2008) synthase 5’-CGACATTGAACTGCACGTGRTCG-3’ (sulfate-reducing) Benzylsuccinate 5’-CCTATGCGACGAGTAAGGTT-3’ (Winderl et bssA 5’-TGATAGCAACCATGG AATTG-3’ The same as above al., 2008) synthase FAM-5’TCCTGCAAATGCCTTTTGTCTCAA3’-TAMRA Benzylsuccinate 5’-GGCTATCCGTCGATCAAGAA-3’ (Winderl et bssA 5’-GTTGCTGAGCGTGATTTCAA-3’ The same as above al., 2008) synthase FAM-5’CTACTGGGTCAATGTGCTATGCATG3’-TAMRA 6-oxocyclohex- Anaerobic degradation (Staats et 1-ene-1- 5’-GCAGTACAAYTCCTACACSACYGABATGGT-3’ of aromatic al., 2011) bamA carbonyl-CoA 5’-CCRTGCTTSGGRCCVGCCTGVCCGAA-3’ hydrocarbons including hydrolase BTEX Note: This table uses standard code for mixed base sites: R = A,G; Y = C,T; M = A,C; K = G,T; S = G,C; W = A,T; H = A,C,T; B = G,T,C; V = G,C,A; D = G,A,T; N = A,C,G,T

2.3.6 Overall effect of ethanol on BTEX plume dynamics

Several laboratory (Corseuil et al., 1998; Da Silva and Alvarez, 2002; Deeb et al., 2002;

Ruiz-Aguilar et al., 2002; Lovanh and Alvarez, 2004) and field studies (Ruiz-Aguilar et al., 2003;

Mackay et al., 2006; Corseuil et al., 2011) showed that ethanol could inhibit BTEX degradation and result in longer plumes. However, results from other laboratory (Beller et al., 2002; Schaefer et al., 2010), pilot-scale (Dakhel et al., 2003; Ma et al., 2011), and field studies (Freitas et al.,

2011a) indicate that BTEX plume elongation may be insignificant under certain site conditions

(e.g., small volume of spill, significant retention of ethanol in the unsaturated zone, high replenishment rate of electron acceptors and nutrients, and fortuitous proliferation of BTEX degraders).

Several mathematical models have simulated the fate and transport of BTEX and ethanol as well as potential effects of ethanol on BTEX plume dynamics (Table 2.3). These model simulations predict that the presence of ethanol would elongate benzene plumes by 17%-150%

(Deeb et al., 2002; Molson et al., 2002; Gomez et al., 2008; Gomez and Alvarez, 2009, 2010;

Freitas et al., 2011a; Freitas et al., 2011b). However, the risk of exposure depends not only on plume length but also on persistence (plume lifespan), both of which can be affected by the content of ethanol in the fuel blend. Simulations for higher ethanol content blends yielded shorter-lived benzene plumes because of decreased mass of benzene present in the source zone

NAPL and increased benzene degradation rates associated with fortuitous growth (and higher concentration) of BTEX degraders (Gomez and Alvarez, 2009) (Figure 2.2). Accordingly, a

release of a high ethanol content blend (e.g., E85) may pose a lower overall risk than a comparable size release of a low ethanol content blend (e.g., E10) (Gomez and Alvarez, 2009).

Table 2.3 Fate and transport models Model name Conceptual model Mathematic Biodegradation Increased benzene Reference al model plume length (Heermann 2D (X-Z); Focus on Analytical Not included ≤ 10% for xylene, and Powers, cosolvent and interphase not benzene 1996) mass transfer. (McNab et al., 3D aqueous transport from a Analytical First-order decay of +100% 1999) finite source zone ethanol and benzene;

(Molson et al., 3D; Consider microbial Numerical Monod kinetics; ≤ +150% 2002) growth (Monod kinetics) and Fermentation pathway. O2 competition; BOD comes from ethanol Cosolvency is not included and its degradation byproducts (Deeb et al., 2D (X-Y) transport from a Numerical Firs-order decay of ethanol 17-34% 2002) gasoline pool and benzene Benzene is not biodegraded when

Cethanol > 3mg/L (Gomez et al., 3D model based on RT3D; Numerical Multiplicative Monod ≤ 40% 2008; Gomez Consider O2 competition, kinetics and Alvarez, catabolic repression, 2009) metabolic flux dilution and microbial population shifts

(Freitas et al., 3D multi component NAPL Numerical Firs-order, Mond (O2 E95 inhibits 2011b) dissolution with dissolved- limited) kinetics, Monod benzene phase reactive transport partial mineralization degradation while E10 does not. (Freitas et al., The same model with Numerical The same with (Freitas et 40% 2011a) (Freitas et al., 2011b). al., 2011b) Consider ethanol retention in the unsaturated zone

250 a) Regular Gasoline 200

150 Benzene Plume Contour (5ppb) Alcohol Plume Contour (5ppb) 100 Anaerobic Zone Contour (100ppb DO) 50

250 b) 10% Ethanol c) 85% Ethanol 200

150

100

Benzene Plume Length Plume Benzene (m) 250 d) 10% e) 85% 200 n-Butanol n-Butanol

150

100

0 0 5 10 15 20 0 5 10 15 20 Time (Year) Time (Year)

Figure 2.2 Simulated benzene plume dynamics (centerline reach) resulting from a 30-gal release of regular gasoline or various fuel alcohol blends. Adapted from Gomez and Alvarez 2010. The anaerobic zone was arbitrarily defined at the 0.1 ppm dissolved oxygen (DO) contour.

Overall, the effects of ethanol on BTEX degradation and plume length are complex and site-specific. In most cases, this adverse impact may not be a serious threat to drinking water resources because BTEX plume elongation would unlikely exceed a few hundred feet (Deeb et

al., 2002; Molson et al., 2002; Ruiz-Aguilar et al., 2003; Mackay et al., 2006; Gomez et al., 2008;

Gomez and Alvarez, 2009), while drinking water wells are often miles away from gasoline

stations.

2.4 Existing knowledge gaps

A primary concern about ethanol blend releases is exacerbating the potential impact of

co-occurring or pre-existing BTEX contamination. Ethanol (and other biofuels) could increase potential exposure to BTEX in groundwater (i.e., causing longer BTEX plumes), either by

enhancing BTEX dissolution and migration or by hindering biodegradation. The significance of

these complex effects will be site-specific, and insufficient data are available to determine how

ethanol might affect BTEX remediation time and costs or the number of sites that will require

corrective action. In most cases, the presence of ethanol should not pose a serious threat to

drinking water resources because BTEX plume elongation is unlikely to exceed a few hundred

feet, while drinking water wells are often located beyond one mile from fuel stations.

2.4.1 Methane explosion risk

Recently, there has been an increased focus on vapor exposure pathways. The

biodegradation of ethanol could result in relatively high CH4 concentrations and in subsurface

deep soil gas. However, no study has quantified CH4 accumulation in overlaying enclosed spaces,

and to assess the potential for bioattenuation by methanotrophs that consume CH4 along the

groundwater to ground surface pathway.

2.4.2 Enhanced potential for benzene vapor intrusion

Another important knowledge gap is the effect that the generated CH4 has on the fate and transport of benzene vapors through the unsaturated zone. Previous research have showed that aerobic biodegradation can significantly attenuate benzene flux and reduce its vapor intrusion potential (Hers et al., 2000; Davis et al., 2005; Abreu and Johnson, 2006; Davis et al., 2009;

Kristensen et al., 2010). However, the effect of CH4 on the biodegradation of benzene vapors is not fully understood.

2.4.3 VFAs accumulation

Besides CH4, other ethanol degradation byproducts such as VFAs could also be problematic. Whether VFAs accumulations could compromise groundwater aesthetic quality has not been investigated. VFAs accumulate and cause a decrease in pH (Ma et al., 2011; Ma et al.,

2013b), which may promote the dissolution of redox- and/or pH-sensitive metals from the aquifer matrix (e.g., iron, manganese, and arsenic), thus exacerbating groundwater contamination

(Deutsch, 1997; Brown et al., 2010). Although no studies of metal mobilization by ethanol-blend releases has been reported in the literature, elevated arsenic concentrations have been detected in groundwater contaminated by petroleum hydrocarbons (Brown et al., 2010).

2.4.4 Microbial ecology

Microorganisms play a key role in the attenuation of ethanol-blend fuel constituents

(including both ethanol and BTEX), often determining their region of influence and fate (Powers

et al., 2001). Previous microbial ecology studies focused on individual (phylogenetic or catabolic)

genotypes associated with aromatic hydrocarbon degradation (Da Silva and Alvarez, 2004;

Beller et al., 2008; Capiro et al., 2008; Feris et al., 2008; Nelson et al., 2010). However, contaminant biodegradation is usually carried out by a complex microbial food web rather than a single degrading strain or catabolic gene (de Lorenzo, 2008). Few studies have used advanced metagenomic tools to investigate biodegradation processes and microbial responses to fuel ethanol releases from a community level.

Chapter 3

Methodology

3.1 Pilot-scale aquifer tank

An 8 m3 (3.7 m×1.8 m ×1.2 m) pilot-scale continuous-flow metal tank was used in this study

(Figure 3.1 and 3.2). This tank was constructed for a previous joint research project to investigate the source behavior, migration, biodegradation and microbial impacts of E100 and E95 releases

(Capiro, 2007; Stafford, 2007). Details about tank design and construction can be found in

Stafford (2007) and Natalie (2006). For this experiment, the tank was modified to have two channels and was repacked with new fine grain masonry sand (Circle Sand, Houston, TX, US).

The sand packing method was the same with the previous study and details about sand packing method can be found in Stafford (2007).

Figure 3.1 A photo of the pilot-scale aquifer tank

Figure 3.2 A photo of the pilot-scale tank

A plan view of the tank is shown in Figure 3.3. Two parallel channels separated by an acrylic barrier were equipped with independent tap water inlet. Tap water was added at 170

L/day (average seepage velocity of 0.75 m/day) to obtain a water table elevation of about 70 cm from the bottom of the tank. A photo of the tap water inlet is showed in Figure 3.5 (a). The total aquifer thickness was 115 cm and the depth of the water table was 45 cm below sand surface

(Figure 3.4). The top 5 cm of the soil was air-dried as previously described (Stafford et al., 2009).

A 10-cm layer above the water table was saturated with groundwater due to capillary action.

Because of the small variation in groundwater flow rate, the depths of the water table (as well as the upper boundary of saturated capillary fringe) varied between 35 cm and 45 cm below sand surface (Figure 3.4).

Figure 3.3 Plan view of the aquifer tank.

Figure 3.4 Profile view of the aquifer tank.

Figure 3.5 Photos of tap water inlet (a) and ethanol/benzne/toluene injection port (b) of the aquifer tank.

In Channel 1, a municipal water feed with added 10 % (v/v) ethanol, 50 mg/L benzene, 50

mg/L toluene (E/B/T) and 24,000 mg/L sodium bromide (NaBr) was injected at a depth of 22.5

cm below the water table (67.5 cm blow sand surface) at a rate of 0.4 L/d. A photo of the ethanol/benzene/toluene injection port is showed in Figure 3.5 (b). The NaBr was added as a conservative tracer, and to maintain a solution density to reach a neutral buoyancy condition with the flowing groundwater. Although high salt concentrations can be inhibitory to bacteria due to osmotic stress, the added bromide salt was diluted by the tank flow to less than 5,000 mg/L, which is within the typical tolerance range of soil bacteria (Ronald M. Atlas, 1993). The density of the ethanol/NaBr solution injected relative to water was measured as 1.002 at 20 oC. Channel

2 served as a control with the same injection depth and injection rate of water mixture containing

50 mg/L benzene, 50 mg/L toluene (B/T) and 24,000 mg/L NaBr with an estimated density relative to water of 1.019 at 20 oC.

A total of 28 groundwater sampling ports (steel tubes screened on the bottom outlet) were installed to collect groundwater samples for chemical measurements (Figure 3.3). All groundwater sampling ports were placed at the same depth as the injection port of ethanol blend solution (22.5 cm below water table, Figure 3.4). Four monitoring wells (PVC) were also installed to: 1) monitor water table depth in each channel; 2) install (at M2) a in-line YSI600XLM groundwater geochemical probe (YSI, Yellowspring, OH, US); 3) observe free-phase NAPL migration (Figure 3.6).

Figure 3.6 Groundwater sampling ports and monitoring wells (a) and groundwater outlet (b) of the aquifer tank.

3.2 Releases experiment

This pilot-scale release experiment lasted 4 years, which could be divided into 4 different stages. Figure 3.7 shows the timeline for the entire experiment and for timeline for each chapter in this thesis. On August 7th 2009, sand and groundwater samples were collected as pre- contaminated baseline samples (Stage 1). Ten days later (August 17th 2009), the ethanol blend release (10% v:v ethanol, 50 mg/L benzene, 50 mg/L toluene) began at a flow rate of 0.4 L/day.

On September 5th 2011, sand and groundwater samples were collected as the sample exposed to

the ethanol blend release for 2 years (Stage 2). After the Stage 2, ethanol was removed from the

blend solution and the solution containing 50 mg/L benzene and 50 mg/L toluene continued to be released at the same flow rate (0.4 L/day) for another 8 months. On May 4th 2012, sand and groundwater samples were collected as the sample exposed to the benzene and toluene release for 8 months (Stage 3). After the Stage 3, the release of benzene and toluene mixture was shut off, but the tap water was still injected into the tank at the same flow rate (170 L/day). Two samples were taken during this recovery stage: at 4 months (September 2th 2012, Stage 4a) and

12 months (May 23th 2013, Stage 4b) after exposure to clean water.

Figure 3.7 Timeline for the tank release experiment.

3.3 Chemical analytical methods

Groundwater samples were collected from groundwater sampling ports using 60 mL plastic syringes (Thermo Fisher Scientific, Waltham, MA, US). The analysis of ethanol, benzene, toluene and bromide tracer was performed by our collaborators in University of Houston.

3.3.1 Methane

For CH4 analysis, water samples (50 ml) were injected into glass serum bottles (125 ml)

capped with a Teflon-lined septa and aluminum crimps. Bottles were shaken on an Orbit 300

Multipurpose Vortexer (Labnet international inc., Edison, NJ) at 35 rpm for 1.5 hours.

Headspace samples (100 μL) were injected into a HP5890 gas chromatograph (Hewlett Packard,

MN, US) quipped with a flame ionization detector (FID) and a packed column (6 ft × 1/8 in o.d.

60/80 carbopack B/1% SP-1000, Supelco, Bellefonte, PA, US, Figure 3.8). The GC heating

program was 60°C for 1 min, injection port temperature 250 °C, and FID temperature 250 °C.

The detection limit was 0.1 mg/L.

Figure 3.8 HP5890 gas chromatograph for methane measurement.

3.3.2 Ethanol degradation byproducts

For the analysis of ethanol degradation byproducts (acetate, propionate, butyrate and

butanol), 2.7 mL water samples were collected and mixed with 0.3 mL of 0.3-M oxalic acid (to

acidify the samples and protonate the VFAs; Capiro et al. 2008) .Mixtures were then filtered into

1-mL screw-cap vials followed by 1 μL injections into a HP5890 gas chromatograph (Figure 3.9) equipped with a flame ionization detector (Hewlett Packard, MN, US) equipped with a FID and a glass column (2 m×2 mm inner diameter) containing 80/120 Carbopack B-DA*/4% Carbowax

20 M (Supelco, Bellefonte, PA, US). The GC heating program was 175°C for10 min, injection

port temperature 200 °C, and FID temperature 200 °C. The method detection limit was 1 mg/L

for acetate and propionate acid, and is 2 mg/L for butyrate.

Figure 3.9 HP5890 gas chromatograph for the measurement of ethanol degradation byproducts.

3.3.3 Ethanol

For ethanol analysis, supernatants were collected in 2-mL gas-tight glass vials with polypropylene caps and PTFE septa (Sun SRI, Rockwood, TN, US) was injected directly into a gas chromatograph (GC) (HP 6890, Santa Clara, CA, US) equipped with a capillary column

(Supelco, model SPB-5, 30 m length, 0.53 mm diameter, 5 m film thickness, St. Louis, MO, US)

and a flame ionization detector (OI Analytical, College Station, TX, US). The detection limit was

1 mg/L.

3.3.4 Benzene and toluene

For benzene and toluene analysis, supernatants (5 ml) were placed into a Tekmar P&T autosampler (Model No. 2016, Mason, Ohio, US) and measured by GC (Agilent 6890N, Santa

Clara, CA, US) equipped with a 5973N Mass Selective Detector (J&W Scientific, Model DB-

624, 20 m length, 0.130 mm diameter, Santa Clara, CA, US). The detection limit was 1.0 g/L for

both benzene and toluene.

3.3.5 Bromide tracer

Bromide samples were collected separately in 125 mL field sampling bottles (Fisher

Scientific; Pittsburgh, PA, US) and analyzed using a bromide ion selective electrode (Cole-

Parmer, Vernon Hills, Illinois) as described in Cápiro et al. (2007). The detection limit was 1 mg/L.

3.4 Groundwater geochemical parameters monitoring

Temperature, pH, oxidation reduction potential (ORP), dissolved oxygen (DO) and

conductivity of groundwater were monitored in Channel 1 by a YSI 600XLM groundwater geochemical probe (YSI, Yellow Springs, Ohio, US) installed at M2 (Figure 1). The probe was

programmed to take readings at 0:00 am and 12:00 pm daily. Sensors were calibrated per

manufacturer protocols.

Groundwater temperature and pH were also measured our collaborators from University of

Houston. pH was measured using Oakton Waterproof EcoTestr™ pH 2 Pocket pH Tester (Davis

Instruments, Vernon Hills, IL, US) and groundwater temperature was measured using Taylor

9848 2 3/4" Dishwasher Proof Antimicrobial Digital Pen-Style Thermometer (Taylor Precision

Products, Oak Brook, IL, US). The YSI 600XLM groundwater geochemical probe does not work

in the winter of 2011. For the groundwater sample collected at Stage 3, 4a and 4b, dissolved

oxygen was measured using Dissolved Oxygen AccuVac® Ampules (Hach, Loveland, CO, US).

Groundwater conductivity and oxidation reduction potential (ORP) are not measured for Stage 3,

4a and 4b.

3.5 Quantitative real-time PCR analysis

Quantitative real-time PCR analysis (qPCR) was conducted using Applied Biosystems

7500 Real Time PCR System (Applied Biosystem, Foster City, CA, US). In this study, absolute

quantification method was used to enumerate the gene copy number (DNA). Genomic DNA was

used as the standard for the calibration curve. For qPCR analysis, DNA was extracted in

triplicate from 0.25 g sand using PowerSoil DNA Kit (MOBIO, Carlsbad, CA, US). Depending on different target gene, both TaqMan and SYBR Green methods were used. For TaqMan method, the mixtures contained 12.5μL TaqMan® Environmental Master Mix 2.0 (Applied

Biosystems; Foster City, CA, US), 500 nM each primers, 200 nM probe and 2μL template DNA in a total volume of 25μL. For SYBR Green Ι method, the mixture contained 12.5μL Power

SYBR® Green PCR Master Mix (Applied Biosystems; Foster City, CA, US), 500 nM each primers and 2μL template DNA in a total volume of 25μL. AB 7500 Sequence Detector was used to perform qPCR reactions with the following temperature program: 50°C for 2 min, followed by 95°C for 10 min and 40 cycles at 95°C for 15 s, and 30s at annealing temperature for each primers, 40s at 72 °C for extension. For SYBR green detection, melting curve analysis was conducted after thermal cycle complete to make sure no nonspecific PCR products were generated. Primer sequences, reaction chemistry, target gene and standards DNA for calibration curves can be found in the Materials and methods section in each chapter.

Figure 3.10 Applied Biosystems 7500 real-time PCR system

3.6 Pyrosequencing

The DNA sequencing technology is developing very fast during recent 10 years. Several next generation DNA sequencing instrument has been commercially available at present:

Rochae/454 FLX, the Illumina/Solexa Genome Analyzer and the Applied Biosystems SOLiDTM

System (Mardis, 2008b). Rochae/454 is the first commercialized next generation sequencing instrument. A 454 Genome Sequencer Junior System (454 Life Sciences, Branford, CT) in

Department of Ecology and Evolutionary Biology of Rice University was used for in this study.

For the workflow of 454 pyrosequencing, genomic DNA is fragmented by nebulization and ligated to adaptor (Figure 3.11 a). The library fragments are then mixed with a population of

agarose beads (Figure 3.11 b). The bead surface has oligonucleotides complementary to the 454-

specific adapter sequences, so each bead connected with only one single fragment.(Mardis,

2008b) Each of these fragment:bead complexes is isolated into individual oil:water micelles

Inside the micelles, DNA fragment is amplified to approximately >106 copies on the surface of

each bead by emulsion PCR (emPCR, Figure 3.11 b). After emPCR, several hundred thousand such beads (each contain on unique 106-fold amplified single strand DNA template) are added to the surface of 454 picotiter plate (PTP) that hold one single bead in each of several hundred thousand single well. Then pyrosequencing reaction occurs in each well (Figure 3.11 c) (Mardis,

2008b).

Figure 3.11 Workflow of 454 pyrosequencing analysis. (a) describes the DNA library preparation steps. (b) describes that emPCR amplification and (c) describes pyrosequencing process. This figure is adapted with permissions from (Mardis, 2008b).

Pyrosequencing is a sequencing method based on "sequencing by synthesis" principle. To

sequence the single strand DNA template in each PTP well, this technique synthesizes the

complementary strand along the target DNA template, one base pair at a time, and detecting which base was actually added at each step. The incorporation of a nucleotide (synthesis) is accompanied by release of pyrophosphate (PPi) which initiates a series of downstream reactions

(ATP is used by luciferase to converse luciferin to oxyluciferin) that ultimately produce light by the firefly enzyme luciferase. (Owen-Hughes and Engeholm, 2007) The amount of light produced is proportional to the number of nucleotides incorporated. (Mardis, 2008b). Figure 3.12 illustrates the pyrosequencing reaction in PTP wells.

Figure 3.12 Pyrosequencing reaction in PTP wells. This figure is adapted with permissions from (Mardis, 2008a). Two independent experiments (Chapter 6 and 8) were conducted for different DNA samples using different PCR primers and reaction chemistry. Details on DNA library preparation 44

methods can be found in the materials and methods section in corresponding chapters. The

pyrosequencing was conducted s by a Genome Sequencer Junior System (Figure 3.13, 454 Life

Sciences, Branford, CT, US).

Figure 3.13 454 Genome Sequencer Junior System.

3.7 Geochip

Geochip is a comprehensive high-throughput functional gene microarray technique, which targets a variety of key functional genes involved in major biogeochemical processes including carbon, nitrogen, phosphorus, and sulfur cycling, organic contaminant remediation, metal resistance and reduction, virulence genes, stress response, and bacterial phage genes. This study used the latest Geochip version (Geochip 4.6), that contains more than 120,000 50-mer

oligonucleotide probes, targeting more than 200,000 gene variants from 539 functional genes

families derived from 4,123, 183, and 396 bacterial, archaeal, and fungal species/strains

respectively (He et al., 2012b).

For the general process of Geochip analysis, DNA is extracted from environmental samples and labeled with a fluorescent dye. Purified DNA samples are then hybridized on the Geochip

microarray at 42 to 50°C with 40% to 50% formamide (He et al., 2012a). When the labeled DNA

molecules come to contact with appropriate gene probe, the DNA will bind with the probe

(hybridization). After hybridization, unbinding DNA molecules are washed away, and the signal

intensity of the fluorescently labeled and captured DNA molecules is scanned by a laser scanner.

Theoretically, the fluorescent signal intensity is proportioned to DNA abundance. Geochip has

been used to analyze microbial functional gene diversity in a variety of environmental samples

including soil, water, sediment, contaminated sites, extreme environments and bioreactors (He et

al., 2012a). Figure 3.14 show a simplified diagram for the Geochip development, analysis and

data processing methods.

Figure 3.14 Schematic presentation of GeoChip development and operations for analyzing environmental DNA samples. This figure is adapted with permissions from http://www.glomics.com/t_Geochip_Operations.html.

3.7 Biovapor 1-D analytic vapor intrusion model

“Biovapor” is an analytic vapor intrusion model which is based on the widely used

Johnson and Ettinger’s model,(Johnson and Ettinger, 1991) and it additionally includes oxygen- limited biodegradation.(DeVaull, 2007; DeVaull et al., 2010) “Biovapor” incorporates a steady- state vapor source, diffusion-dominated soil vapor migration in a homogeneous soil layer, and mixing within a building enclosure. An illustrative conceptual model assumed in “Biovapor” is presented in Figure 3.15. The soil is divided into a shallow aerobic layer including biodegradation and a deeper anaerobic layer in which biodegradation is omitted. Oxygen demand

is attributed to a sum of baseline respiration of soil organic matter and biodegradation of multiple chemicals assuming first-order degradation rates. The model is solved by iteratively varying the aerobic depth to match oxygen demand to oxygen supply. The website of American Petroleum

Institute (API) provides a user-friendly Excel spreadsheet implementation of the Biovapor model

(http://www.api.org/environment-health-and-safety/clean-water/ground-water/vapor- intrusion/biovapor-form). The software interface of Biovapor is showed in Figure 3.16.

Figure 3.15 Conceptual model of “Biovapor”. This figure is adapted with permissions from (DeVaull et al., 2010).

Figure 3.16 Biovapor software interface.

As a simplified one dimensional (1-D) analytic model, “Biovapor” is not expected to provide highly accurate predictions using a limited number of input parameter. Rather, the model is used in this study to have a preliminary understanding of methane explosion and benzene vapor intrusion problems associated with ethanol blend fuel releases.). A much more powerful three dimensional (3-D) numerical models that consider multispecies transport, reaction and phase partitioning processes are used in Chapter 5 to investigate scenarios with complicated model domain and boundary conditions. Details about this 3-D model are described in next section.

3.8 Abreu and Johson 3-D numerical vapor intrusion model

The 3-D numerical model used in this study was developed by Lilian Abreu and Paul

Johnson at Arizona State University. This finite difference model simultaneously solves 1)

continuity equations that govern the soil gas pressure distribution and the resulting soil gas velocity field, 2) chemical reactive transport equations that account for diffusion, advection, and biodegradation in the subsurface, 3) air flow and chemical transport through foundation cracks, and 4) chemical mixing with indoor air (Abreu, 2005). The soil characteristics (e.g., porosity, water content, permeability, bulk density and organic carbon content) can be modeled as homogeneous, layered (up to 10 layers) or heterogeneous. This model also allows for different biodegradation kinetics (non-biodegradation, 0-order, 1st-order, Monod 2nd-order ) and user- defined building characteristics (e.g., foundation cracks position and size, air exchange rate and building depressuration). Model outputs include soil gas pressure, soil gas velocity and chemical concentration fields, chemical flux and indoor air concentration. Details about mathematical model development can be found in Abreu (2005). This model has been used in several studies

(Abreu and Johnson, 2005, 2006; Abreu et al., 2009) and for establishing the US EPA’s guidance on vapor intrusion (EPA, 2012b, 2013).

Chapter 4

Experimental and 1-D analytic model investigation on

CH4 explosion and benzene vapor intrusion

[Modified from a research paper published in Environmental Science and Technology(Ma et al., 2012)]

4.1 Introduction

Fuel ethanol releases can stimulate methanogenesis in impacted aquifers, which could pose an explosion risk if CH4 migrates into enclosed spaces where ignitable conditions exist. Several recent studies have reported relatively high methane concentrations in groundwater (23 to 47 mg/L) (Freitas et al., 2010; Spalding et al., 2011) and subsurface deep soil gas (68% v:v) (Jewell and Wilson, 2011) at sites impacted by fuel ethanol releases. Whereas these studies contribute to the understanding of potential CH4 intrusion pathways, a comprehensive assessment of the associated explosion risk needs to consider multiple processes that affect the rate and extent of 51

CH4 accumulation in buildings overlying contaminated groundwater, such as phase partitioning, diffusion and advection, biodegradation, attenuation across building foundations, building ventilation and indoor mixing (Patterson and Davis, 2009; Rivett et al., 2011).

In particular, there is a need for studies that quantify CH4 accumulation in overlaying enclosed spaces, and to assess the potential for bioattenuation by methanotrophs that consume

CH4 along the groundwater to ground surface pathway. Methanotrophs are widely distributed in the environment (Hanson and Hanson, 1996), but their vertical distribution profile and activity have not been investigated in aquifer systems impacted by fuel ethanol.

Another important knowledge gap is the effect that the generated CH4 would have on the fate and transport of benzene vapors through the unsaturated zone. Previous research on benzene vapor intrusion focused on the fate and transport of benzene alone, and showed that aerobic biodegradation can significantly attenuate benzene flux and reduce its vapor intrusion potential

(Hers et al., 2000; Abreu and Johnson, 2006). However, the effect of CH4 on the biodegradation of benzene vapors is not fully understood. Aerobic biodegradation of CH4 consumes O2 that would otherwise be available for benzene biodegradation. Since high concentrations of benzene and methane can coexist in the vicinity of the source zone (Corseuil et al., 2011; Spalding et al.,

2011), it is important to investigate whether aerobic benzene degradation in the vadose zone would be inhibited by competition for O2 by methanotrophs, thus increasing benzene vapor intrusion.

4.2 Materials and Methods

4.2.1 Pilot-scale aquifer system

The plan view and profile view of the sampling port for this study are showed in Figure 4.1.

Details on tank construction and release scenario can be found in Chapter 3.

Figure 4.1 Plan view (a) and profile view (b) of the pilot-scale aquifer system.

4.2.2 Flux chamber description

A stainless steel flux chamber was emplaced above sand surface in the tank to measure

methane surface accumulation. The structure schematic of the flux chamber was showed in

Figure 4.2 The lower part of the chamber was a cylinder with the diameter of 60 cm and the

height of 25 cm and the bottom is open. The top part of the chamber was a cone with the height

of 15 cm. The chamber had an enclosed surface area of 2.8×103 cm2 and total enclosed volumes

of 8.5 ×104 cm3. When emplaced on the soil surface, 8 cm of the chamber bottom was buried in the soil to make sure no gas leaked out from the bottom. We used this chamber as a “static-

chamber” which means no anthropogenic introduction of gas into the chamber during the

incubation period. There were three gas sampling ports (A, B and C). Our test showed that the

gas samples collected from these three ports have same results, thus we used top sampling port

(A) for our experiment. The sampling port contained one steel tube connected with an 8 cm

rubber tube clamped by an open ended clamp (Pentair Technical Products, Anoka, MN). Testing

for potential leaks from the chamber (including sampling ports), was conducted by inverting the

chamber and filling with water. No leaks were observed.

Figure 4.2 The structure schematic (left) and the photo of the flux chamber (right).

4.2.3 Sampling and analysis methods for CH4 and O2

To measure CH4 accumulation in the flux chamber, 30 mL headspace gas samples were

collected from the top sampling port using VICI Series A-2 Precision Sampling Syringes (VICI

Instruments, Baton Rouge, LA, US). To measure the vertical concentration profiles of CH4 and

O2 in the unsaturated zone, 100 μL of soil pore gas samples at different depths (5, 10, 15, 20, 25

and 30 cm BGS) were collected in six replicates using VICI Series A-2 Precision Sampling

Syringes and analyzed immediately in the lab. Details on CH4 measurement can be found in

Chapter 3. O2 was analyzed with an Agilent 7890 gas chromatograph equipped with thermal

conductivity detector (GC-TCD).

4.2.4 Assessment of CH4 oxidation activity at different depths

To assess the vertical distribution of the CH4 oxidation activity and the spatial variability

of the concentration of a representative methantrophic functional gene (pmoA), soil samples were

collected from different depths in the pilot-scale aquifer (5 to 10 cm below ground surface (BGS) and 15 to 20 cm BGS for the unsaturated zone; 30 to 40 cm BGS for the saturated capillary fringe; 40 to 50 cm BGS cm for the region across the water table, 60 to 70 cm BGS for the anaerobic saturated zone near the centerline of ethanol plume. Microcosms were prepared to measure CH4 biodegradation rate in soil samples. Triplicate soil samples (15 g) were mixed with

10 mL sterile H2O and placed in sterile 125-mL serum bottles. Each bottle was then supplemented with 1 ml of CH4 and incubated in a rotary shaker at 150 RPM and 37 ºC.

Headspace CH4 was analyzed using same method as described in Chapter 3.

4.2.5 qPCR assays for pmoA gene

pmoA gene is a widely used functional biomarker for methanotrophs. qPCR analyses were

performed to quantify the abundance of pmoA in the same soil samples used in the microcosms.

SYBR Green methos was used for this study and details on qPCR method can be found in

Chapter 3. The target groups, primer sets and annealing temperatures for each assay are

summarized in Table 4.1.

Table 4.1 Primers, target group, standard DNA and annealing temperature for qPCR Target group Annealing Calibration standard Assay Primer Primer sequences (genus) temp. (ºC) (genomic DNA) * MBAC A189F 5’-GGNGACTGGGACTTCTGG-3’ Methylobacter and Methylomicrobium 54 Mb601R 5’-ACRTAGTGGTAACCTTGYAA-3’ Methylosarcina album A189F 5’-GGNGAC TGGGACTTCTGG-3’ Methylococcus MCOC Methylococcus 60 Mc468R 5’-GCSGTGAACAGGTAGCTGCC-3’ capsulatus A189F 5’-GGNGACTGGGACTTCTGG-3’ MCAP Methylocapsa 60 Methylocapsa acidiphila Mcap630R 5’-CTCGACGATGCGGAGATATT-3’ A189F 5’-GGNGACTGGGACTTCTGG-3’ DGGE band affiliated FOREST Forest clones 60 Forest675R 5’-CCYACSACATCCTTACCGAA-3’ with forest clones II223 F 5’-CGTCGTATGTGGCCGAC-3’ TYPEII Methylosinus 60 Methylocystis parvus II664R 5’-CGTGCCGCGCTCGACCATGYG-3’ 2 *Except for the assay “FOREST” which uses DGGE band affiliated with forest clones

4.2.6 Biovapor model simulation

Biovapor is an analytic vapor intrusion model which is based on the widely used Johnson

and Ettinger’s model, and it additionally includes oxygen-limited biodegradation (DeVaull, 2007;

DeVaull et al., 2010). Detail on this model can be found in Chapter 3 and De Vaull 2007.

Biovapor was used to calculate the CH4 indoor concentrations under different scenarios (e.g.,

different source concentrations, source depths, with and without biodegradation). This model

was also used to simulate benzene vapor intrusion under different site conditions. Model inputs

for simulations for methane explosion and simulations for benzene vapor intrusion under strong

methanogenic conditions are listed in Table 4.2 and Table 4.3 respectively.

Table 4.2 Model inputs for methane explosion simulation Building parameters:

O2 concentration below the foundation: 5% (v:v) Aquifer parameters: 3 3 Soil Porosity (θT-soil): 0.38 cm -void/cm -soil 3 3 Soil Water Content (θw-soil): 0.05 cm -water/cm -soil -3 3 3 Soil Organic Carbon Fraction (foc): 5×10 cm -void/cm -soil 3 Soil Density-Bulk (ρs): 1.7 g-soil/cm -soil 3 Airflow Under Foundation (Qf): 83 cm -air/sec Annual Median Soil Temperature (T): 10ºC Depth of Water Table (LT): change from 1m to 20m

Minimum O2 Concentration For Aerobic Biodegradation: 1% Chemicals: Methane Concentration in Groundwater: change from 0 to 20 mg/L 1st order Biodegradation Rate of Methane: 71 hr-1 Attenuation Factor: 1

Table 4.3 Model inputs for benzene vapor intrusion simulation Building parameters:

Indoor Mixing Height (Lmix): 244cm Air Exchange Rate (ER): 6 day-1 Foundation Thickness: 15cm Foundation Area: 1060,000 cm2 Foundation Crack Fraction (η): 3.77×10-4 cm2-cracks/ cm2-total 3 3 Total Porosity (Soil-filled Cracks, θT-cracks): 1 cm -void/cm -soil 3 3 Water Filled Porosity (Soil-filled Cracks, θw-cracks): 0 cm -void/cm -soil 3 Airflow Through Basement Foundation (QS): 83 cm -air/sec Aquifer parameters: 3 3 Soil Porosity (θT-soil): 0.38 cm -void/cm -soil 3 3 Soil Water Content (θw-soil): 0.05 cm -water/cm -soil -3 3 3 Soil Organic Carbon Fraction (foc): 5×10 cm -void/cm -soil 3 Soil Density-Bulk (ρs): 1.7 g-soil/cm -soil 3 Airflow Under Foundation (Qf): 83 cm -air/sec Annual Median Soil Temperature (T): 10ºC Depth of Water Table (LT): 5m

Minimum O2 Concentration For Aerobic Biodegradation: 1% Chemicals: Benzene Concentration in Groundwater: 5mg/L Methane Concentration in Groundwater: Change from 0 to 20mg/L 1st order Biodegradation Rate: 1st order Biodegradation Rate of Benzene: 0.79 hr-1 1st order Biodegradation Rate of Methane: 71 hr-1 Attenuation Factor: 1

4.3 Results and Discussion

4.3.1 CH4 accumulation in the flux chamber

Methane emissions from the soil surface were measured using a static flux chamber of

internal volume (V) =8.5×104 cm3 and surface area (A) =2.8×103 cm2. Four measurements events

were made in different seasons (Figure 4.3). Methane concentrations inside the chamber

increased exponentially (k= 0.26 h-1) and reached an asymptotic concentration 30 to 80 h after

the chamber was emplaced. With a presumed constant emission flux of methane from the soil

surface (during the sampling period) and low methane concentrations in ambient air, this implied

an effective passive air flow rate (Q) through the chamber of Q= V • k = 2.21×104 cm3/hr. Thus,

the surface methane emission flux (J) was estimated as J =Q • Ca / A (Table 4.4), where Ca is the

average asymptotic chamber concentration. The seasonal variation in J (2.11×10-5 to 9.74×10-5

mg/cm2-hr) reflects differences in methane generation rates at different groundwater temperatures,(Ma et al., 2011) with higher values observed during summer months when

groundwater was saturated with methane (Table 4.5). The solubility of methane is 21.4 mg/L at

28 °C.(Yamamoto et al., 1976) The maximum concentration of methane in the headspace of the flux chamber was 21 ppmv, a value far below the methane vapor concentrations in equilibrium

6 with saturated groundwater (i.e., 10 ppmv), and also far below the lower explosion limit (LEL,

50,000 ppmv) for methane in ambient air.(Bjerketvedt et al., 1997)

Table 4.4 Measured CH4 concentration and calculated surface flux

Sampling date Average asymptotic CH4 concentration in Calculated surface the flux chamber, Ca emission flux, J 3 2 ppmv mg/cm mg/cm -hr October 2010 14.2 9.47×10-6 7.47×10-5 February 2011 4.0 2.67×10-6 2.11×10-5 April 2011 18.5 1.23×10-5 9.74×10-5 June 2011 16.5 1.10×10-5 8.69×10-5

Table 4.5 CH4 concentrations in groundwater * Winter (February 2011) Summer (June 2011) Sampling port Before (mg/L) After (mg/L) Before (mg/L) After (mg/L) C1 8.9 7.6 21.6 23.7 C2 5.5 8.3 20.6 19.2 C3 6.3 5.8 22.5 20.0

* Groundwater methane was measured before and after the flux chamber was installed. CH4 concentrations in groundwater remained stable during the flux chamber measurement period.

2010-October 25 2011-February

) 2011-April v 2011-May 20 ppm (

10 concentration 4 5 CH

0 0 30 60 90 120 Time (hour)

Figure 4.3 CH4 accumulation inside the flux chamber in different seasons. Warmer summer temperatures stimulated methanogenesis to a greater extent than methanotrophic activity, and CH4 accumulated to a greater extent.

4.3.2 Aerobic biodegradation of CH4 in the pilot-scale aquifer

The vertical methane concentration profile shows that more than 99% of the methane was

attenuated before reaching the unsaturated zone (30 cm BGS; Figure 4.4). The average methane

3 3 concentration at 30 cm BGS was 4.9×10 ± 2.7×10 ppmv, which is only 0.5% of the equilibrium

6 methane concentration for the saturated groundwater (10 ppmv). The low methane

concentrations in the unsaturated zone represent a low biochemical oxygen demand and no

significant oxygen depletion occurred in that zone (Figure 4.4). The relative contribution of

biodegradation to methane attenuation in the unsaturated zone (15 to 30 cm BGS) likely exceeds

99%, as estimated by a one-dimensional steady-state diffusion model with first-order reaction

(see the supporting information in Ma et. 2012).

O2 concentration (%) 0 5 10 15 20 0

-10 CH4 O2 -20

-30

-40 Depth (cm) -50

-60

-70 100 102 104 106 CH4concentration (ppmv)

Figure 4.4 Vertical concentration profiles of CH4 and O2 in the soil gas near the groundwater sampling port C2 (Figure 1). CH4 concentration at 67.5 cm BGS was calculated based on measured groundwater concentration using Henry’s law. The dissolved O2 concentration at 67.5 cm BGS was measured by groundwater geochemical monitoring probe (Figure 4.1).

Microcosm assays and pmoA analysis (Figure 4.5) show that the saturated capillary fringe

(30 to 40 cm BGS) exhibited the highest methanotrophic activity (0.51 ± 0.028 μg CH4/hr/g soil and 2.2×107 ± 4.8×106 pmoA gene copies/g soil). Furthermore, methane degradation rate and pmoA copy numbers were significantly correlated (p < 0.05, Figure 4.6), corroborating the usefulness of this biomarker to assess methane bioattenuation potential. Apparently, the co-

3 existence of high concentrations of oxygen (21 % v:v at 30 BGS) and methane (> 2.9×10 ppmv) in the capillary fringe favored the proliferation and activity of methanotrophs. Relatively high aerobic biodegradation activity of hydrocarbon vapors in the capillary fringe has also been reported.(Lahvis and Baehr, 1996) In addition to biodegradation, the slower diffusion of

methane through the capillary fringe resulted in higher dispersion and dilution, which would

also contribute to decreasing the concentration of methane reaching the surface.

CH4 biodegradation rate (μg CH4/ hr/ g soil) 0.0 0.2 0.4 0.6 0 pmoA gene -15 CH4 biodegradation Unsaturated rate Zone -30 Saturated

Depth (cm) Capillary Fringe -45

Saturated -60 Zone 104 105 106 107 108 pmoA copies number/g soil

Figure 4.5 Vertical distribution of pmoA gene concentration and CH4 biodegradation rate in the pilot-scale aquifer. Degradation rates and pmoA copy numbers were significantly correlated (r2 = 0.997, p < 0.05). The designed water table was at 45 cm BGS (red dotted line). A 10 cm layer above the water table was usually saturated with groundwater due to capillary action (blue dotted line). Due to the fluctuation of the water table, the actual upper boundaries of saturated zone and capillary fringe were often several centimeters higher than the designed levels.

0.6

) R = 0.977 0.4 p = 0.003 /hr/gsoil 4 0.2 gCH μ biodegradation rate ( 4

CH 0.0

0 1x107 2x107 3x107 pmoA copies number / g soil

Figure 4.6 Correlation between methane biodegradation rate and pmoA gene concentration in the soil samples collected from different depths in the pilot-scale aquifer.

The absence of lag phases during the biodegradation assays (Figure 4.7) indicates that the methanotrophs were already adapted. The maximum methane biodegradation rate (0.51±0.028

μg CH4/h/g soil, in capillary fringe) was comparable to some reported biodegradation rates for

landfill cover soils (e.g., 0.65 μg CH4/hr/g soil (Jones and Nedwell, 1993) and 0.75 μg CH4/hr/g

soil (Schuetz et al., 2003)), although much higher biodegradation rates have been reported for

similar systems (e.g., 112 μg CH4/hr/g soil (Rivett et al., 2011)).

5-10 cm BGS 15-20 cm BGS Saturated capillary fringe Water table 1.2x104 Saturated zone Sterile control

) 4 v 1.0x10 ppm ( 8.0x103

6.0x103

3 concentration 4.0x10 4

CH 2.0x103

0.0 0 50 100 150 200 250 Time (Hours)

Figure 4.7 Methane biodegradation activity in microcosms prepared with soil samples from different depths. Error bars depict ± one standard deviation from the mean of triplicate microcosms, and solid lines are linear fits to estimate zero-order degradation rates.

4.3.3 CH4 accumulation simulation

“Biovapor” simulations corroborate the non-existence of explosion risk in overlying

confined spaces associated with diffusion-driven methane migration under more generic

conditions. Simulated methane indoor concentrations increase as the source concentration increases and the source depth decreases (Figure 4.8). However, even under the worst-case scenario (i.e., high methane source concentration, shallow source depth and no biodegradation), the simulated methane indoor concentration is still much lower than the lower explosion limit for methane (50,000 ppmv). Model simulations also corroborate that aerobic biodegradation

significantly reduces the methane flux into the enclosure (“Jf” in Figure 3.15 of Chapter 3) by 78%

to 99%, depending on the source concentration and depth. If the methane source concentration is high (e.g., 20 mg/L in groundwater) and the source is shallow (e.g., 1 m), methane oxidation would be limited by oxygen availability, but biodegradation still decreases Jf by 78%. If the methane source concentration is low (e.g., 1 mg/L) and the source is deep (e.g., 20 m), more methane would be biodegraded (99% of Jf) and the simulated concentrations with biodegradation would be much lower (e.g., 3%) than those simulated without biodegradation.

105 105 ) ) v v

4 4

ppm 10

103 103

102 Depth of WT=1m 102 Depth of WT =1m Depth of WT=3m Depth of WT=3m Depth of WT=5m Depth of WT=5m 101 Depth of WT=10m 101 Depth of WT=10m indoor concentration ( Depth of WT=20m indoor concentration ( Depth of WT=20m 4 CH explosion limit 4 CH explosion limit 0 4 0 4 CH 10 CH 10 0 5 10 15 20 0 5 10 15 20 CH4 concentration in groundwater (mg/L) CH4 concentration in groundwater (mg/L) (a) (b)

Figure 4.8 Simulated methane indoor concentrations (a) with and (b) without methane biodegradation under different source concentrations and depths to water table. Simulation parameters are given in Table 4.2.

“Biovapor” assumes that diffusion is the only vapor transportation pathway in the vadose zone.(DeVaull et al., 2010) This assumption is appropriate for most contaminated sites.(Johnson and Ettinger, 1991; McHugh and McAlary, 2009) However, we cannot exclude the possibility that in some scenarios methanogenesis could be strong enough to increase the pore pressure near

the source and produce significant vertical advective flow in the vadose zone.(Kruger and

Frenzel, 2003; Schuetz et al., 2003) Further research is needed to address the possible advective contribution to methane fluxes in the vadose zone overlying ethanol blend releases.

4.3.4 Impacts of CH4 oxidation on benzene vapor intrusion

Experimental conditions (e.g., shallow water table with open surface without overlying structures and sandy porous medium that facilitate aeration, and high methanotrophic activity in the capillary zone) precluded significant oxygen consumption in the unsaturated zone of this pilot aquifer system. However, oxygen depletion has been reported in the vadose zone of many fuel contaminated sites (Lundegard and Johnson, 2006; Lundegard et al., 2008; Molins et al.,

2010) and landfill cover soil.(Molins et al., 2008) Therefore, simulations were conducted using

“Biovapor” to investigate how oxygen consumption by methanotrophs in the vadose zone might affect hydrocarbon vapor intrusion pathways under broader release scenarios. Benzene, which is commonly the selected risk driver in vapor intrusion risk assessments for fuel impacted sites,(EPA, 2002a) was chosen in this modeling effort.

Model simulations indicate that under more generic conditions, methane oxidation could deplete oxygen that would otherwise be consumed in benzene degradation, thereby increasing potential benzene vapor intrusion. When methane is absent in the groundwater, extensive aerobic biodegradation of benzene vapors occurs in the vadose zone and the simulated benzene indoor concentration is more than six orders of magnitude lower than the EPA screening level (0.31

µg/m3) (Figure 4.9). Benzene indoor concentrations increase with methane groundwater 69

concentrations. If the methane groundwater concentration reaches 20 mg/L, the benzene indoor concentration reaches 8.5 µg/m3, which is 27 times higher than the EPA screening level.

Competition for oxygen is the major reason that benzene vapor intrusion is enhanced. Oxygen consumption and aerobic zone thickness were calculated by “Biovapor”. The aerobic zone is conservatively defined as the soil region with oxygen concentration higher than 1% (v:v), which is conservatively assumed to be the minimum oxygen level under which aerobic biodegradation can occur.(DeVaull et al., 2010) As methane groundwater concentrations increase, more oxygen is consumed by methane oxidation and the aerobic zone thickness decreases sharply (Figure 4.9).

We simulated a worse-case scenario (high benzene groundwater concentration (10 mg/L), high methane groundwater concentration (20 mg/L), and low depth to the water table (3 m)). The simulated benzene indoor concentration was 1.6×103 µg/m3, which is 5,300 times higher than the

EPA screening level. However, for the same conditions without methane, the simulated benzene indoor concentration was only 1.2×10-2 µg/m3, which is significantly lower than the EPA screening level. Methanotrophic activity increases the simulated benzene flux into the enclosure by 1.3×105 times from 2.2×10-4 to 30 μg/s.

Benzene indoor concentration Aerobic zone yhickness EPA screening level for benzene ) 3 250

g/m 0 μ

( 10 200

10-2 150

10-4 100

-6 10 50 Aerobic zone thickness (cm)

10-8 0 Benzene indoorBenzene concentration 0 5 10 15 20 CH4 concentration in groundwater (mg/L)

Figure 4.9 Simulated benzene indoor concentrations and the aerobic zone thickness under different methane groundwater concentrations. Simulation parameters are given in Table 4.3.

These results indicate that methanotrophs would consume oxygen and significantly enhance the vapor intrusion potential for benzene. However, it should be noted that this inference is based on several simplified and idealized assumptions that may apply to sites experiencing strong methanogenic conditions with associated high methanotrophic activity, but are unlikely to be representative of all sites impacted by releases of fuel ethanol blends.

4.4 Conclusion

Overall, whereas fuel ethanol releases can stimulate significant methanogenic activity in groundwater, both model simulations and flux chamber measurements indicate that methane is

unlikely to build up to explosive levels in overlying confined spaces. Methanotrophs can

significantly attenuate methane migration through the vadose zone, particularly in the capillary 71

zone where adequate moisture and oxygen availability to favor methanotrophic activity.

Nevertheless, aerobic biodegradation of methane may have a negative effect. Depending on the release scenario, methanotrophs could deplete the available oxygen and reduce the near-source attenuation for other volatile compounds such as benzene, increasing their vapor intrusion potential.

Chapter 5

A 3-D numerical model investigation for methane explosion and benzene vapor intrusion potential associated with ethanol-blended fuel releases

[Extract from a manuscript in submitted to Environmental Science and Technology]

5.1 Introduction

Recent US legislation promoting a higher percentage of ethanol in blended fuel will further

stimulate the production and consumption of fuel ethanol. (EPA, 2012a) Vapor intrusion risk associated with high-ethanol blend releases (E20 up to E95) has been increasingly recognized as a potential concern. (Ma et al., 2013a) Fuel ethanol releases often stimulate methanogenic activity,(Freitas et al., 2010; Nelson et al., 2010; Jewell and Wilson, 2011; Ma et al., 2012;

Wilson et al., 2012; Ma et al., 2013a; Sihota et al., 2013) which may pose an explosion hazard when methane accumulates in a confined or poorly ventilated space at 5% to 15% (v:v). Ethanol- derived methane may also increase the vapor intrusion potential of toxic fuel hydrocarbons by 73

stimulating the depletion of oxygen by methanotrophs and thus inhibiting aerobic biodegradation

of hydrocarbon vapors. (Jewell and Wilson, 2011; Ma et al., 2012; Jourabchi et al., 2013)

Relative high concentrations of methane have been reported in groundwater (23 to 47 mg/L)

(Buscheck et al., 2011; Ma et al., 2011; Spalding et al., 2011) and soil gas (15% to 58%

v:v)(Spalding et al., 2011; Wilson et al., 2013) that have been impacted by fuel ethanol spills.

Although these studies contribute to the understanding of methane generation and migration in

the subsurface, none of them directly assessed methane intrusion and accumulation in overlying

buildings. Because of flame quenching within the soil matrix, methane explosion will not occur

in-situ in the soil, but may happen when methane accumulates in a confined space above ground

where ignitable conditions exist. (Wilson et al., 2012)

Flux chambers have been used to measure methane intrusion and accumulation in confined

spaces above fuel ethanol-impacted sites. (Ma et al., 2012; Sihota et al., 2013) However, flux

chamber measurements are not representative of actual vapor flow into buildings, because 1) flux

chambers do not have foundations, thus may overestimate the vapor flux into buildings, and 2)

flux chambers cannot mimic “building effects” (e.g., depressurization) that play an important

role in the vapor subsurface-to-indoor air pathway. (Patterson and Davis, 2009) Therefore, direct measurements of methane concentrations in the indoor air or model simulations that consider both attenuation across foundations and building effects are necessary to assess potential explosion risk associated with high ethanol blend releases.

CH4 gas release by ebullition and advection has previously been noted in a variety of

environments, including saturated peat(Baird et al., 2004), rice fields(Jain et al., 2004),

sediment(Amos and Mayer, 2006), landfills(Molins et al., 2008), and aquifers contaminated by petroleum spills.(Amos et al., 2005; Molins et al., 2010) Recent field studies show the

importance of CH4 migration by ebullition and advection at sites impacted by ethanol-blend

releases. (Sihota et al., 2013; Wilson et al., 2013) Soil gas advection would enhance the upward migration of methane, thus increasing potential explosion risk. Another knowledge gap is the impact of biogenic methane on the fate and transport of fuel hydrocarbons in the subsurface.

Anaerobic biodegradation processes are relatively slow and do not significantly attenuate hydrocarbon vapor migration through the vadose zone. (Lundegard et al., 2008) If sufficient oxygen (e.g., >1 %) is present in the unsaturated zone, biodegradation could reduce hydrocarbon concentrations by several orders of magnitude within a relatively short distance (1-2 m). (Hers et al., 2000; Fitzpatrick and Fitzgerald, 2002) However, the consumption of methane by methanotrophs may deplete available soil oxygen, thus inhibiting aerobic hydrocarbon degradation. (Jewell and Wilson, 2011; Ma et al., 2012) These processes are still poorly understood.

With improved understanding of vapor intrusion processes, various mathematical models have been developed to assess potential impacts to indoor air quality. (Johnson and Ettinger,

1991; Hers et al., 2000; Abreu and Johnson, 2005; DeVaull, 2007; Mills et al., 2007; Pennell et al., 2009; Yu et al., 2009a; Verginelli and Baciocchi, 2011) Although 1-D analytical models such

as the Johnson and Ettinger model16 are simple, fast, and widely used for screening purposes, 3-

D numerical models that consider multispecies transport, reaction and phase partitioning are

more accurate and applicable to describe scenarios with complex model domain and boundary

conditions.(Abreu and Johnson, 2005, 2006; Abreu et al., 2009; Bozkurt et al., 2009; Pennell et

al., 2009; Yao et al., 2011) To our knowledge, such 3-D models have not been used to assess the

vapor intrusion risk associated with ethanol-blended fuel releases.

In this study, a 3-D numerical vapor intrusion model (Abreu and Johnson, 2005) was used to simulate various scenarios and quantitatively address: 1) the potential for methane accumulation in buildings overlaying ethanol-blended fuel impacted sites, and the associated explosion hazard; 2) the effect of methane (and associated methanotrophic activity) on the vapor intrusion pathway of benzene; and 3) the impact of gas advection on the vapor intrusion pathway of methane and benzene.

5.2 Materials and Methods

5.2.1 3-D numerical model

The 3-D numerical model used in this study was developed by Abreu and Johnson. (Abreu

and Johnson, 2005) This finite difference model solves 1) continuity equations that govern the

soil gas pressure distribution and the resulting soil gas velocity field, 2) chemical reactive

transport equations that account for diffusion, advection, and biodegradation in the subsurface, 3)

air flow and chemical transport through foundation cracks, and 4) chemical mixing with indoor

air. The soil characteristics (e.g., porosity, water content, permeability, bulk density and organic

carbon content) can be modeled as homogeneous, layered (up to 10 layers) or heterogeneous.

This model also allows for different biodegradation kinetics (non-biodegradation, 0-order, 1st-

order, Monod) and user-defined building characteristics (e.g., foundation cracks position and size, air exchange rate and building depressuration). Model outputs include soil gas pressure, soil gas velocity and chemical concentration fields, chemical flux and indoor air concentration. Details about mathematical model development can be found in Abreu (2005).(Abreu and Johnson, 2005)

This model has been used in several studies,(Abreu and Johnson, 2005, 2006; Abreu et al., 2009)

including a US EPA guidance on vapor intrusion.(EPA, 2012b, 2013)

5.2.2 Simulated scenarios and model input parameters

This study simulated a symmetrical scenario that includes a single building (10 m × 10 m)

located at the center of an open field (48 m × 48 m) (Figure 5.1). The building has a 2 m deep

basement and a perimeter crack (1 mm wide) around the entire foundation of the basement

(Figure 5.2). Similar to other numerical modeling studies,(Abreu and Johnson, 2005, 2006;

Abreu et al., 2009; Bozkurt et al., 2009; Yao et al., 2011) the building is assumed to have a

disturbance pressure of -5 Pa. The disturbance pressure is the difference in gas pressure between

building indoor air and atmosphere, which is generated by a building depressurization effect. The

contaminant source zone is located at the bottom of the unsaturated zone and spreads across the

entire model domain (48 m × 48 m, Figure 5.1). Four different source depths (3, 5, 8 and 15 m)

were chosen. The 3 m depth represents a shallow vapor source case and 15 m depth represents a

deep vapor source case. Since the basement has a depth of 2 m, actual intervals between source zone and building foundation are 1, 3, 6 and 13 m. All simulations were conducted for homogeneous and steady-state scenarios and the associated model input parameters related to

CH4 and benzene source concentrations, source gas pressure and source depths are listed in

Table 5.1. Other model input parameters regarding building and foundation, soil properties, contaminant properties and biodegradation rate coefficients were selected based on previous studies(Abreu and Johnson, 2005, 2006; Abreu et al., 2009) (Table 5.2).

Figure 5.1 Cross-sectional view of the model domain and the perimeter crack on the building foundation (blue dashed circle). The vapor mass fluxes used in following figures includes flux emitted from the source (Jsource), flux into the building (Jbuilding), flux across the soil surface (Jsurface), and flux biodegraded (Jbio).

Figure 5.2 Plan view of the foundation with perimeter crack distribution (red dash line) used in the simulations presented in this study.

Table 5.1 Simulated scenarios for Table 5.3-5.5 and Figure 5.3-5.13

CH4 source TPH source Benzene source Source gas Source depth concentration concentration concentration pressure Tab le 5.3 75 % 200 g/m3 NA 0 to 200 Pa 5 m Tab le 5.4 75 % 200 g/m3 NA 0 to 200 Pa 3, 8 and 15 m Tab le 5.5 75 % 200 g/m3 NA 0 to 200 Pa 3, 8 and 15 m Figure 5.3 0.015% to 75% 200 g/m3 NA 0 Pa 3, 5, 8 and 15 m Figure 5.4 0.015% to 75% 200 g/m3 NA 0 Pa 8 m Figure 5.5 75 % 200 g/m3 NA 0.1 to 200 Pa 3, 5, 8 and 15 m Figure 5.6 75 % 200 g/m3 NA 0 to 200 Pa 5 m Figure 5.7 0.076% to 75% NA 0.1 g/m3 0 Pa 3, 5, 8 and 15 m Figure 5.8 0.015% to 75% NA 0.1 g/m3 0 Pa 8 m Figure 5.9 0.015% to 75% NA 0.1 g/m3 0 Pa 8 m Figure 5.10 0 % to 75% NA 0.1 g/m3 0 Pa 8 m Figure 5.11 0 % to 75% NA 0.1 g/m3 0 Pa 8 m Figure 5.12 0 % to 75% NA 0.1 g/m3 0 Pa 8 m Figure 5.13 75 % NA 0.1 g/m3 0.1 to 200 Pa 3, 8 and 15 m

Table 5.2 Model input Parameters

Building and foundation Vapor source zone Length: 10 m Location: base of vadoez zone Width: 10 m Size: entire domain footprints Depth in soil: 2 m Source depth: 3 m to 15 m Foundation type: basement Foundation thickness: 0.15 m Methane properties Enclosed space volume: 174 m3 Henry's Law constant:29.9 cm3-water/cm3-gas -1 Air exchange rate: 0.5 h Organic Carbon Water Partitioning Coefficient (Koc):0 Crack width: 0.001 m Diffusivity in air: 0.195 cm2/s Total crack length: 38.64m Diffusivity in water: 1.71×10-5 cm2/s Crack location: perimeter 1st-order degradation rate: 82 h-1 Disturbance pressure: 5 Pa Benzene and TPHproperties Soil properties Henry's Law constant:0.228 cm3-water/cm3-gas Soil bulk density: 1700 kg/m3 Organic Carbon-Water Partitioning Coefficient (Koc):61.7 Fraction of organic carbon: 0.001 Diffusivity in air: 8.8×10-2 cm2/s Total porosity: 0.35 Diffusivity in water: 9.8×10-6 cm2/s Water-filled porosity; 0.07 1st-order degradation rate: 0.18 h-1 Soil permeability: 10-11 m2 Air dynamic viscosity: 1.8×10-4 g/cm/s Oxygen properties Henry's Law constant:31.6 cm3-water/cm3-gas Algorithm parameters Organic Carbon Water Partitioning Coefficient (Koc):0 Numerical scheme: implicit Diffusivity in air: 0.2 cm2/s Pressure subroutine: Diffusivity in water: 2.41×10-5 cm2/s Variable time step: 0.01 s-40 h Minimum O2 concentration for degradation to occur: 1% Percent change allowed: 10% Concentration subroutine: Variable time step: 0.1 s-80 h Percent change allowed: 5%

Based on the overall stoichiometry of ethanol degradation under fermentative methanogenic conditions (2 CH3CH2OH = CO2 + 3 CH4), the degradation of ethanol could

2 3 produce gas with up to 75 % (v:v) CH4 content. Therefore, 75 v:v % (4.91×10 g/m ) was chosen as the maximum CH4 source concentration.

Benzene, the main risk driver in vapor intrusion assessment at fuel-impacted sites, was

chosen as a representative volatile fuel hydrocarbon. To simulate the impacts of CH4 generation

on benzene vapor intrusion potential, 0.1 g/m3 was chosen as benzene source concentration as a

representative value from a petroleum vapor intrusion database. (Lahvis et al., 2013) For

simplicity, benzene was assumed to be the only hydrocarbons present in the vadose in these

simulations. To simulate the CH4 explosion risk, we considered aerobic CH4 biodegradation by

methanotrophs as it migrates upward through the vadose zone. Since the available oxygen can

also be consumed by the biodegradation of other compounds associated with the release, the model assumes that 200 g/m3 of total petroleum hydrocarbons (TPH) were present at the source

zone (Table 5.1). This concentration is representative of non-aqueous phase liquid (NAPL)

sources (EPA, 2013) and has been used in previous modeling studies. (Abreu and Johnson, 2005,

2006; Abreu et al., 2009) The properties of these hydrocarbons were assumed to be the same as

for benzene vapor (Table 5.2) because model computation slows significantly when running

multiple fuel constituents.

To simulate the impact of gas advection on migration and intrusion of subsurface CH4

and benzene, different source gas pressures (0.1, 1, 5, 10, 50, 100, 150, 200 Pa) were selected

(Table 5.1). Note that the higher simulated pressures may be rare in the high-permeability sandy

soils assumed here, but were considered to delineate the potential effects of high-ethanol blend

releases under a wide range of conditions. To our knowledge, data on soil gas pressures in the

source zone of fuel ethanol releases are not available. However, according to a numerical

simulation of soil gas data at a crude oil release site, methanogenesis in the source zone could

generate about 1 Pa of source pressure.(Molins et al., 2010) Landfill sites could have as high as

several thousand Pa of source pressure.(Nastev et al., 2001) As a readily degradable compound,

the release of large volumes of fuel ethanol usually stimulates much stronger methanogenic

activity than petroleum hydrocarbons, but is unlikely to be stronger than methanogenesis at

landfills. Therefore, the source gas pressure at a fuel ethanol impacted aquifer is likely to be

higher than that at a petroleum spill site and smaller than that at a landfill site.

5.2.3 Assumptions and limitations of this modeling study

This modeling study has following assumptions: 1) the source zone is non-depleting and

infinite; 2) the building is modeled as a perfectly mixed continuously stirred tank reactor (CSTR);

3) no NAPL phase is present in the transport domain (it could be present at the source zone) and

chemicals partition among gas, dissolved and adsorbed phases only; and 4) density-driven

advective transport is neglected.

Similar to previous studies (Abreu and Johnson, 2005, 2006; Abreu et al., 2009; Yao et al.,

2011), all simulations reported herein assume homogeneous and steady-state conditions.

However, real site geologic conditions are usually heterogeneous, complex and site- specific.(Bozkurt et al., 2009) Non-steady state environmental factors such as barometric pressure fluctuations and wind load on buildings could affect the migration and intrusion of gas contaminant into buildings.(Robinson and Sextro, 1997; Riley et al., 1999) Although a modified version of the Abreu and Johnson model is capable of simulating vapor intrusion during transient 84

wind load and barometric pressure fluctuations,(Luo, 2009) such simulations are time consuming

and require significant computational resources. Therefore, this study simulates quasi-steady scenarios representing short-term average indoor concentrations during periods of net airflow is into the building enclosure.

Diffusion in this model follows Fick’s law, which presumes low concentrations for each soil gas constituent and that their individual fluxes are independent. This is addressed in these model scenarios by specifying, indirectly, through pressures at the model boundaries, a constituent-independent Darcy velocity field for the soil gas. This is a reasonable assumption for the nearly equimolar (and nearly equal volume) aerobic reactions included in the model domain.

Also, the effect of displaced, passive non-reactive soil gases (mainly N2) on flux estimates is not

addressed; this would require a different approach, such as a ‘dusty gas’ model.(Cunningham

and Williams, 1980; Mason and Malinauskas, 1983)

Finally, the model presumes a separating and attenuating layer of vadose-zone soil between

the methanogenic source and the building enclosure. Direct entry of methane gas into a building,

such as gas evolution through an untrapped or unvented well, sump, or sewer connected to a

building, is not evaluated.

5.3 Results and Discussion

5.3.1 The explosion risk for diffusion-driven CH4 migration is negligible

If diffusion is the major mass transport process in the deeper vadose zone (advection may

occur in the vicinity of the basement due to building depressurization), CH4 is unlikely to build

up to the lower flammable level (5% v:v) in overlying buildings. CH4 indoor concentrations are

simulated for different CH4 source concentrations and source depths, with and without

biodegradation (Figure 5.3). Simulated CH4 indoor concentrations increase as source

concentrations increase and source depth decreases. However, even under the worse-case

scenario examined here (i.e., CH4 source concentration at 75% v:v, the source depth is 3 m, and

no biodegradation), the simulated CH4 indoor concentration is still much lower (> 20-fold) than

the lower flammable level for CH4 (5% v:v) (Figure 5.3).

0 0 10 Flammable limit 10 Flammable limit 10-3 10-3

10-6 10-6

10-9 Source depth 10-9 Source depth 3 m 3 m -12 5 m -12 5 m 10 8 m 10 8 m indoor concentration v:v) (% 15 m indoor concentration v:v) (% 15 m 4 10-15 4 10-15 CH 0.01 0.1 1 10 100 CH 0.01 0.1 1 10 100 CH source concentration (% v:v) CH source concentration (% v:v) (a) 4 (b) 4

Figure 5.3 Simulated CH4 indoor concentrations for different CH4 source concentrations and source depths, (a) with and (b) without CH4 biodegradation. The source was assumed to contain 200 g/m3 TPH that contribute to the biochemical oxygen demand in the vadose zone. The source pressure is 0 Pa. 86

Methanotrophic bacteria could significantly attenuate the mass flux of CH4 in the unsaturated zone and reduce indoor concentrations and the associated explosion risk (compare

Figure 5.3(a) and 5.3(b)). Figure 3 shows the simulated scenarios with the source depth of 8 m

3 and a benzene source concentration of 200 g/m . If the CH4 source concentration is lower than

1.5 % (v:v), more than 99% of CH4 flux emitted from the source (Jsource) is degraded before reaching the ground surface, and the percent of source flux that is biodegraded (Jbio/Jsource) does not change with increasing CH4 source concentrations (Figure 5.4). If the CH4 source concentration is higher than 1.5 % (v:v), the percentage of source flux that is biodegraded

(Jbio/Jsource) decreases as the CH4 source concentration increases. However, even when the CH4 source concentration reaches 75 % (v:v), biodegradation still attenuates more than 92% of the upward CH4 flux (Figure 5.4). Methanotrophic bacteria are widespread in natural environments

(they are especially abundant in soil).(Hanson and Hanson, 1996) Our modeling results corroborate previous pilot-scale experimental results(Ma et al., 2012) and indicate the importance of methanotrophic activity to attenuate CH4 generated from ethanol plumes and reduce its potential to reach the surface.

Jsource Jbuilding Jbio / Jsource Jbio Jsurface 10-3 100

-5 /s) 2 10 98 , %) -7 10 96 source flux (g/m 4 -9 / J

CH 10 94 bio

-11 (J 10 92

0.01 0.1 1 10 100 Percentage of source flux degraded CH source concentration (% v:v) 4

Figure 5.4 Changes in CH4 mass flux emitted from the source (Jsource), flux into the building (Jbuilding), flux across the soil surface (Jsurface), and flux biodegraded (Jbio) with different CH4 source concentrations. The right Y-axis shows the percent of source flux degraded (Jbio/Jsource). The source also contains 200 g/m3 TPH that contribute to the biochemical oxygen demand in the vadose zone. The source soil gas pressure is 0 Pa. The source depth is 8 m.

CH4 explosion risk in non-pressurized flow can also be assessed using the attenuation

factors assembled from the U.S. EPA vapor intrusion database, which contains indoor air

measurements of toxic vapors coupled with subslab soil gas, exterior soil gas groundwater, or

crawlspace measurements for 913 buildings at 41 sites in 15 U.S. states.(EPA, 2012c) By

compiling this database, a recent EPA report recommends 0.01 as a conservative subslab

attenuation factor (the ratio of indoor air concentration to subslab soil gas concentration, AFsubslab

= Cindoor/Csubslab) for chlorinated volatile compounds. (EPA, 2012c) Using this attenuation factor

and assuming that the source of CH4 concentration is 75% (v:v), the corresponding CH4 indoor

concentration is only 0.75% (v:v), which is much lower than the lower flammable limit. This

calculation is conservative and the methane attenuation factor from source to indoor is likely to

be smaller than 0.01, because 1) methane degrades much faster than chlorinated solvents in an

aerobic vadose zone; and 2) additional attenuation occurs when methane transports from the

source zone to the subslab area.

5.3.2 The explosion risk increases significantly for advection-driven CH4 migration

A recent field study showed that the accumulation of ethanol-derived methane and carbon

dioxide in the source zone could generate a pressure gradient and cause significant advective gas

transport in the subsurface.(Sihota et al., 2013) They reported up to 6.3×103 g/m2/s of surficial

CH4 efflux and 6 % to 22 % (v:v) of CH4 concentration in the flux chamber emplaced at a site

impacted by the release of a large volume of denatured fuel-grade ethanol. (Sihota et al., 2013)

Our simulations corroborate this field study, indicating that if methanogenic activity near the source zone is sufficiently high to cause significant advective gas transport, CH4 could build up

to flammable levels (> 5% v:v) in overlying buildings. Scenarios with different source gas

pressures (0, 0.1, 1, 5, 10, 30, 50, 100, 150, and 200 Pa) and source depths (3, 5, 8, and 15 m)

were simulated (Figure 5.5). Figure 5.6 shows that increases in source pressure significantly

change the soil gas pressure field distribution, which may change the subsurface air flow and

contaminant mass flux. Table 5.3 lists the simulated air flow rate and CH4 flux for the source

depth of 5 m. The simulated air flow rate and CH4 flux for other source depths are listed in

Tables 5.4 and 5.5 respectively. Simulated CH4 indoor concentrations for all source depths are

shown in Figure 5.5. As source gas pressures increase from 0 Pa to 200 Pa, simulated air flow

rates from the source (Qsource) for a source depth of 5 m increased from 0 to 757 L/min (Table

5.3), resulting in a 44-fold increase in the CH4 flux emitted from the source (Jsource) (Table 5.3).

As a result, the simulated CH4 flux into buildings (Jbuilding) increases by 56-fold (Table 5.3), which causes the CH4 indoor concentrations to increase by more than 60-fold for the source depth of 5 m (Figure 5.5). When the source depth is equal to or less than 5 m and the source gas pressure is higher than 100 Pa, simulated CH4 indoor concentrations exceed the 5% v:v

flammable level, resulting in a potential explosion risk (Figure 5.5).

1 10 Flammable limit

100

10-1 Source depth

indoor concentration v:v) (% 3m 5m 4 -2 8m 15m

CH 10 0.1 1 10 100 Source soil gas pressure (Pa)

Figure 5.5 Simulated CH4 indoor concentrations for different source depths and source gas 3 pressures. The source contains 75 % (v:v) CH4 and 200 g/m TPH that contribute to the biochemical oxygen demand in the vadose zone. The CH4 indoor concentration for the source soil gas pressure of 0 Pa can be found in Figure 5.3a.

(m) pth pth De

X (m) 1 2 Figure 5.6.Pressure field distribution with different source pressure for source depth of 5 m. The values are gauge pressure which is 3 relative to atmosphere pressure (unit: Pa). Negative values (such as in the plot for the source soil gas pressure of 0 Pa) indicate that the 4 soil gas pressure is lower than atmospheric pressure

Table 5.3 Simulated air flow rate and CH4 flux with ifferent source pressures for a source depth of 5 m 2 Air flow rate (L/min) CH4 flux (g-CH4/m /s)

(1) (2) (3) Pressure (Pa) Qsource Qsurface Qbuilding Jsource Jsurface Jbuilding 0 0 -0.9(4) 0.9 2.5×10-4 1.2×10-5 1.8×10-4 0.1 1.1 -0.1 1.2 2.5×10-4 1.2×10-5 2.4×10-4 1 4.5 3 1.3 2.8×10-4 1.9×10-5 2.8×10-4 5 20 18 1.9 4.0×10-4 6.9×10-5 4.7×10-4 10 39 36 3 6.0×10-4 1.9×10-4 7.4×10-4 30 114 109 6 1.6×10-3 9.2×10-4 1.8×10-3 50 190 183 9 2.7×10-3 1.8×10-3 2.9×10-3 100 379 366 16 5.4×10-3 4.0×10-3 5.3×10-3 150 568 549 24 8.1×10-3 6.1×10-3 7.8×10-3 200 757 732 31 1.1×10-2 7.9×10-3 1.0×10-2 (1) 6 Jsource was calculated by dividing mass flow rate emitted from the source (g-CH4/s) by the 7 source area (48 m × 48 m). (2) 8 Jsurface was calculated by dividing mass flow rate across the soil surface (g-CH4/s) by the soil 9 surface area (48 m × 48 m -10 m × 10 m). (3) 10 Jbuilding was calculated by dividing mass flow rate into buildings (g-CH4/s) by the building 11 foundation area (10 m × 10 m). (4) 12 The negative value for Qsurface means the air flows from atmosphere into the soil due to 13 building depressurization effect.

Table 5.4 Simulated air flow rate (L/min) with different source pressures for depth of 3, 8, and 15 m(1)

Source depth 3 m 8 m 15 m

Pressure (Pa) Qsource Qsurface Qbuilding Qsource Qsurface Qbuilding Qsource Qsurface Qbuilding 0 0.0 -0.8 0.8 0.0 -1.0 1.0 0.0 -1.0 1.0 0.1 1.7 0.4 1.4 0.7 -0.4 1.1 0.4 -0.7 1.0 1 7 6 2 3 2 1.2 1.5 0.5 1.1 5 33 30 3 12 11 2 7 5 1.3 10 64 61 4 24 22 2 13 12 2 30 190 183 9 72 68 4 38 36 3 50 316 305 13 119 114 6 64 61 4 100 631 610 25 238 229 11 128 122 6 150 946 915 37 357 344 16 191 183 9 200 1261 1221 49 476 458 21 255 245 12

(1) 3 15 The source contains 75 % (v:v) CH4 and 200 g/m TPH

(1) Table 5.5 Simulated CH4 flux (g-CH4/m2/s) with different source pressures for depth of 3, 8, and 15 m

Source depth 3 m 8 m 15 m

Pressure (Pa) Jsource Jsurface Jbuilding Jsource Jsurface Jbuilding Jsource Jsurface Jbuilding 0 4.1×10-4 6.5×10-5 2.1×10-4 1.5×10-4 2.0×10-6 1.2×10-4 8.3×10-5 5.1×10-8 5.6×10-5 0.1 4.2×10-4 6.8×10-5 3.9×10-4 1.6×10-4 2.5×10-6 1.4×10-4 8.6×10-5 7.0×10-8 6.1×10-5 1 4.6×10-4 8.3×10-5 4.6×10-4 1.7×10-4 4.3×10-6 1.7×10-4 9.3×10-5 1.4×10-7 7.3×10-5 5 6.7×10-4 2.2×10-4 7.7×10-4 2.5×10-4 1.6×10-5 2.9×10-4 1.4×10-4 1.4×10-6 1.4×10-4 10 1.0×10-3 4.3×10-4 1.2×10-3 3.8×10-4 6.4×10-5 4.7×10-4 2.0×10-4 9.2×10-6 2.4×10-4 30 2.7×10-3 1.8×10-3 2.8×10-3 1.0×10-3 4.4×10-4 1.3×10-3 5.5×10-4 1.5×10-4 7.0×10-4 50 4.5×10-3 3.3×10-3 4.3×10-3 1.7×10-3 9.8×10-4 2.0×10-3 9.1×10-4 3.8×10-4 1.1×10-3 100 9.0×10-3 6.7×10-3 8.2×10-3 3.4×10-3 2.4×10-3 3.6×10-3 1.8×10-3 1.1×10-3 2.1×10-3 150 1.3×10-2 9.3×10-3 1.2×10-2 5.1×10-3 3.7×10-3 5.2×10-3 2.7×10-3 1.9×10-3 3.0×10-3 200 1.8×10-2 1.1×10-2 1.6×10-2 6.8×10-3 5.1×10-3 6.9×10-3 3.6×10-3 2.6×10-3 3.9×10-3

(1) 3 18 The source contains 75 % (v:v) CH4 and 200 g/m TPH 19

To put the simulated CH4 fluxes data into context, measured CH4 surficial efflux

data in natural and impacted environments are listed in Table 5.6. The simulated Jsurface

for the 75% (v:v) CH4 source concentration varied by six orders of magnitude with

different source depths and source pressures; e.g., 5.1×10-8 g/m2/s for 0 Pa source

pressure and 15 m source depth to 1.1×10-2 g/m2/s for 200 Pa source pressure and 3 m

source depth (Table 5.5). Methane mass fluxes lower than 10-3 g/m2/s are typical for

natural methanogenic environments (Table 5.6). Values greater than 10-3 g/m2/s have

been observed at landfills and sites impacted by ethanol blend fuel releases (Table 5.6).

Table 5.6 Measured CH4 flux in natural and contaminated environments

2 Environments Flux (g-CH4/m /s) Reference

Rice field 5.7×10-7 to 1.2×10-5 (Singh et al., 1999; Datta et al., 2013) Wetlands 8.0×10-8 to 4.5×10-5 (Elberling et al., 2011) (Alford et al., 1997) Lagoon 3.1×10-9 to 6.8×10-5 (Hirota et al., 2007) (Deborde et al., 2010) Peat 1.0×10-7 to 3.6×10-5 (Smemo and Yavitt, 2006; Bohdalkova et al., 2013) E95(1) impacted aquifer 2.2×10-5 to 6.3×10-3 (Sihota et al., 2013)

Landfill 2.4×10-8 to 4.6×10-2 (Bogner et al., 1997) (Cardellini et al., 2003) (1) Ethanol denatured with 5% gasoline.

5.3.3 Oxygen consumption during CH4 biodegradation in the vadose zone increases benzene vapor intrusion potential

Baseline simulations indicate that if there is no CH4 generation in the source zone,

0.1 g/m3 of benzene will not cause a vapor intrusion problem even for a shallow source

(e.g., 3 m). However, CH4 generation near the source could significantly increase

benzene indoor concentrations (Figure 5.7). As CH4 source concentrations increase from

0 to 75% (v:v), benzene indoor concentrations increase by 15-, 1.0×104-, 6.2×107-, and

6.3×1016-fold for source depths of 3, 5, 8 and 15 m, respectively. Although previous studies indicate a low benzene vapor intrusion potential when the source depth is larger than 10 m,(Abreu et al., 2009; EPA, 2013; Lahvis et al., 2013) our simulations infer that if very high CH4 concentrations (e.g., 75 % v:v) are generated and subsequently

consumed in the vadose zone (resulting in O2 depletion), even a separation distance of 13

m (15 m of source depth) may result in some benzene vapor intrusion and possibly

exceed the EPA indoor air screening level of 3.1×10-5 g/m3 that corresponds to a 10-4

lifetime risk (EPA, 2002a) (Figure 5.7). Note that very high CH4 concentrations (e.g., 75 %

v:v) are seldom sustained at sites impacted by leaking underground storage tanks, and are

used in these simulations to delineate potential effects under a wide range of conditions.

) 3 10-4 EPA standard

10-7

Source depth 10-10 3 m 5 m 8 m 15 m 10-13

Benzene indoorBenzene concentration (g/m 0.1 1 10 100 CH source concentration (% v:v) 4 Figure 5.7 Simulated benzene indoor concentrations for different CH4 source concentrations and source depths. The EPA indoor air standard for benzene (3.1×10-5 g/m3) corresponds to a 10-4 lifetime risk (EPA, 2002a). The source contains 0.1 g/m3 benzene, with a source soil gas pressure of 0 Pa.

Inhibition of benzene biodegradation as methanotrophs consume vadose-zone oxygen is the major reason for the increase in benzene vapor intrusion. Biodegradation attenuates more than 90% of the benzene flux emitted from the source (Figure 5.8). As

CH4 source concentrations increase from 0.015 % (v:v) to 75 % (v:v), the benzene flux

that is bio-attenuated (Jbio) decreases 25-fold, and the percent of benzene source flux that

is biodegraded (Jbio/Jsource) decreases from >99.99 % to 90.5 %. This leads to a significant

7 increase (>-10 -fold) in the benzene flux intrusion into buildings (Jbuilding) and transport

across the soil surface (Jsurface), despite a decrease in the benzene flux emitted from the

source (Jsource) by at least 22-fold due to alower benzene concentration gradient between

the source and the subslab (Figure 5.8).

Jsource Jbuilding Jbio / Jsource Jbio Jsurface

10-5 100

/s) -8 2 10 98 , %) 96 10-11 source J / -14 94

10 bio (J

Benzene fluxBenzene (g/m 92 10-17

90 Percent of source flux degraded 0.01 0.1 1 10 100 CH source concentration (% v:v) 、 4

Figure 5.8. Changes in benzene flux emitted from source (Jsource), flux into the building basement (Jbuilding), flux across the soil surface (Jsurface), and flux biodegraded (Jbio) for different CH4 source concentrations. The right Y-axis shows the percent of source flux 3 that is biodegraded (Jbio/Jsource). The source contains 0.1 g/m benzene, with a source soil gas pressure of 0 Pa. The source depth is 8 m.

Depletion of O2 is the major reason for decreases in benzene biodegradation. Under aerobic conditions, CH4 degrades faster than benzene. As the CH4 source concentration increases from 0.015 % (v:v) to 75 % (v:v), the O2 consumed by CH4 degradation increases by 180-fold (Figure 5.9), thus leading to a rapid decrease (25-fold) in the O2 that is used for benzene degradation, even though the total O2 flux entering into the system increases by more than 50-fold due to a lower O2 concentration gradient between the soil surface and the vapor source (Figure 5.9). To better illustrate these processes, changes in the concentration distribution of benzene, CH4, and O2 with different CH4 source concentrations for the source depth of 8 m are shown in Figures 5.10, 5.11, and

5.12. 98

O2 flow into the subsurface

O2 consumption by CH4 degradation 100 O2 consumption by B degradation 10-1

10-2

10-3

10-4

-5

flow or consumption rate (g/s) 10 2

O 0.01 0.1 1 10 100

CH4 source concentration (% v:v)

Figure 5.9 Changes in O2 consumption by aerobic degradation of benzene (B) and CH4 3 with different CH4 source concentrations. The source contains 0.1 g/m benzene, with a soil gas pressure of 0 Pa. The source depth is 8 m.

(m) Depth Depth

X (m)

Figure5.10 Normalized benzene concentration distribution with different CH4 source concentration for a source depth of 8 m. The source also contains 0.1 g/m3 benzene. The source soil gas pressure is 0 Pa. Benzene contours were normalized to the benzene source concentration (0.1 g/m3).

100

(m) Depth Depth

X (m)

Figure 5.11 Normalized CH4 concentration distribution with different CH4 source concentration for a source depth of 8 m. The 3 source also contains 0.1 g/m benzene. The source soil gas pressure is 0 Pa. CH4 contours were normalized to the CH4 source concentration..

101

(m) Depth Depth

X (m)

Figure 5.12 Normalized O2 concentration distribution with different CH4 source concentration for a source depth of 8 m. 3 The source also contains 0.1 g/m benzene. The source pressure is 0 Pa. O2 contours were normalized to the atmospheric concentration .

102

Benzene vapor intrusion could be exacerbated if methanogenic activity is sufficiently high to cause significant advective gas transport. Benzene indoor concentrations were simulated for different source gas pressures (Figure 5.13). As the source gas pressure increases from 0.1 Pa to 200 Pa, the advective gas transport strips more benzene from the source zone and thus increases benzene indoor concentrations by at least 40-fold for all four source depths (Figure

5.13). )

3 Source depth 10-2 3 m 5 m 8 m 15 m 10-3

10-4

EPA standard 10-5 0.1 1 10 100 Benzene indoorBenzene concentration (g/m

Source soil gas pressure (Pa) Figure5.13 Simulated indoor benzene concentrations for different source gas pressures and source depths. The EPA indoor air standard for benzene (3.1×10-5 g/m3) corresponds to a 10-4 lifetime risk. The source contains 0.1 g/m3 benzene. The benzene indoor concentration for the source soil gas pressure of 0 Pa can be found in Figure 5.

5.3.4 Implications for site assessment and remedial action

Guidance for assessing health risks associated with vapor intrusion of petroleum hydrocarbon and chlorinated solvents is relatively well established, but guidance for assessing

103

explosion risks associated with methane vapor intrusion from ethanol fuels is limited. Using a 3-

D numerical model, this study indicates that methane is unlikely to reach flammable levels in overlying buildings if diffusion is the major mass transfer process in the deeper vadose zone.

However, our simulations show that if methanogenic activity is sufficiently strong (as might occur for releases with high ethanol content) to increase gas pressure and cause advective gas transport near the source zone, CH4 could build up to potentially flammable levels (> 5% v:v) in overlying buildings.

The US EPA’s guidance document for petroleum vapor intrusion (EPA, 2013) is based on field measurement data at retail service station sites, including sites for which E10 (10% ethanol fuel) would have been used for decades. According to this document, regular gasoline or E10 releases are unlikely to cause a flammability hazard, unless the gasoline is in the building or in direct contact with the foundation. Therefore, the inferences of our simulations are mainly applicable to releases of high-ethanol content fuels, including E20 up to E95.

Conceptual models of vapor intrusion usually assume that diffusion is the major vapor transport mechanism in the deeper vadose zone, and that advection plays an important role only in the vicinity of the building basement (due to building depressurization This study indicates that advective soil gas transport generated from the accumulation of fermentative biogas could play an important role in the subsurface vapor transport. Therefore, gas advection should be considered for fuel ethanol impacted sites or other sites where strong fermentation activities exist.

Conditions, that are conducive to advective gas migration through the vadose zone include 1)

104

high soil moisture content that inhibits diffusion, 2) a shallow source zone, 3) a soil surface that is paved or covered by a large building foundation that inhibits O2 inflow, 4) release of a high

ethanol blends (e.g., E85), and 5) a large volume release where the source is not removed.

The US. EPA is trying to establish vertical separation criteria for screening vapor intrusion

risk at petroleum release sites. Using available field measurement data, recent EPA guidance

indicates that 5.5 to 6.1 m of separation distance could reduce soil benzene concentration below

a defined soil gas threshold (100 µg/m3) at 95% of sites that have a NAPL source present. (EPA,

2013) Although that document is based on comprehensive site data, our modeling results indicate

that under our simulated conditions the presence of high concentrations of CH4 originating from

releases of high ethanol blends may deplete the available soil O2 and inhibit benzene aerobic

degradation, thus resulting in a higher benzene vapor intrusion potential than suggested in the

current EPA guidance for gasoline fuel release sites. If methanogenic activity is sufficiently high

to generate advective gas transport, the benzene intrusion rate would be even higher. Therefore,

further studies are needed to establish appropriate vertical separation criteria for sites impacted

by high ethanol blends, including E20 up to E95.

105

Chapter 6

Seasonal variation of ethanol fermentative degradation and aesthetic impact of volatile fatty acids generation

[Modified from a research paper published in Ground Water Monitoring and Remediation (Ma et al., 2011)

6. 1 Introduction

Groundwater is one of major drinking water sources in United States. Malodor is one of the main reasons for consumers to complaint about their drinking water quality. Many odorous compounds are not toxic to human body, but they affect the public’s perception of the safety of drinking water. Therefore, US EPA includes odor as one of fifteen contaminants in National

Secondary Drinking Water Regulations (EPA, 2002b). In groundwater, ethanol biodegradation rapidly consumes oxygen and other electron acceptors creating an anaerobic environment. Under these anaerobic conditions, ethanol can be fermented to volatile fatty acids (VFAs) such as acetic, propionic, butyric acids, which can be further syntrophically transformed to hydrogen (H2) and

106

methane (CH4) (Powers et al., 2001). The intermediate degradation products are ultimately mineralized (to H2O and CO2) under oxidizing conditions. Transient presence of VFAs, however, may cause aesthetic impacts to potable groundwater due to their odor and taste. To our knowledge, no study has ever investigated the aesthetic impact of ethanol degradation intermediates on groundwater quality.

Temperature is an important environmental factor that affects microbial growth and activities (Alvarez and Illman, 2005). The seasonal variation in groundwater temperature is between 5°C to 10 °C near the surface (Heath, 1983). However, no research has discussed about impacts of groundwater temperature on the biodegradation process of fuel ethanol. Through this study, we intend to show that variations in groundwater temperature with seasonal changes should be considered when assessing an aquifer’s capacity for natural attenuation of ethanol blends releases and characterizing impacts from byproducts of ethanol degradation.

6.2 Materials and Methods

6.2.1 Pilot-scale aquifer system

The plan view and profile view of the sampling port for this study are showed in Figure 6.1.

Details on tank construction and release scenario can be found in Chapter 3.

107

Glass window

A1 B1 Inlet Outlet Channel 1 EtOH+B/T M1 M2

E/B/T Sonde

A2 B2 Inlet Outlet Channel 2 B/T M3 M4 B/T

54 “ 34 “ 23“ 11“

Groundwater inlets/outlets E/B/T injection ports Monitoring wells Groundwater sampling ports

Figure 6.1 Plan view of the aquifer system.

6.2.2 CH4 and VFAs analysis

Groundwater samples were collected from A1, A2, B1, and B2. For CH4 analysis, 50 mL

groundwater sample was injected into a 125 mL serum bottles capped with a Teflon-lined septa

and aluminum crimps. 100 μL headspace samples were injected into a HP 5890 GC-FID (Agilent

Technologies Inc., Santa Clara, CA). For VFAs analysis, 2.7 mL groundwater sample was mixed with 0.3 mL of 0.3 mol/L oxalic acid. Mixtures were then filtered into 1-mL screw-cap vials

followed by 1 μL injection into a HP 5890 GC-FID (Agilent Technologies Inc., Santa Clara, CA).

Detailed information on chemical analysis can be found in Chapter 3. 108

6.3 Results and Discussion

6.3.1 Effect of groundwater temperature on VFAs production

Acetic acid concentrations remained below 5 mg/L in the control channel throughout the

monitoring period (Figure 6.2). However, in the channel exposed to the ethanol, acetic acid

concentrations (A1) increased from <1 mg/L (August 7th, 2009, 29.9 ºC) to 95.7 mg/L

(December 8th, 2009, 14.6 ºC), followed by a concentration decrease to below 40 mg/L in

January. (< 10 ºC) From February to June, with the subsequent increase in temperature (from 8.0

ºC to 30.0 ºC), the acetic acid concentration increased again to 131 mg/L (April 29th, 2010). A similar trend was observed at the sampling well B1. The maximum concentration measured was

226 mg/L at B1 (May 10th, 2010, 23.9 C). This indicates that acetic acid production was

significantly influenced by temperature variations. A significant correlation was found between

acetic acid production (A1) and temperature (p = 0.000024, Figure 6.3), indicating that acetic

acid generation from the fuel ethanol blends were significantly influenced by the seasonal

variation of temperature. The annual average temperature of shallow ground water (10-25m depth) in the U.S. ranges from 4°C in the north central areas to approximately 25°C in southern

Florida. The seasonal variation in groundwater temperature is greatest near the surface, amounting 5°C to 10 °C (Heath, 1983).

109

A1 150 A2 35 Groundwater 30 Temperature 120 ) C o

25 (

90 20

60 15 Temperature Acetic acid (mg/L) 10 30 5

0 0 0 100 150 200 250 300 Days after B/T/(E) release

Figure 6.2 Acetic acid concentrations at sampling wells A1 (in Channel 1, exposed to ethanol, benzene, and toluene) and A2 (in Channel 2, exposed to only benzene and toluene). Day 0 corresponds to August 17th, 2009.

Figure 6.3 Correlations between acetic acid concentrations (measured at A1) and groundwater temperature.

110

Unlike acetic acid, butyric acid remained at a relatively low level (< 20 mg/L) from August

7th, 2009 until late February, and then increased steadily to 280 mg/L (A1, May 29th, 2010)

(Figure 6.4). The initial lag in butyric acid production was expected as butyric acid was a

daughter-product of acetic acid biotransformation (Barker et al., 1945; Gibson, 1965). Since acetic acid is a direct precursor for butyric acid formation, its higher abundance is conducive to higher butyric acid accumulation.

A1 300 A2 35 Groundwater 30

) 250 Temperature ) C o 25 (

mg/L 200 20 150 15

100 Temperature 10 n-butyric acid ( 50 5

0 0 0 100 150 200 250 300 Days after B/T/(E) release

Figure 6.4. Butyric acid concentrations at sampling wells A1 (in Channel 1, exposed to ethanol, benzene, and toluene) and A2 (in Channel 2, exposed only to benzene and toluene). Day 0 corresponds to August 17th, 2009.

6.3.2 VFAs odor generation

The standard odor criteria (Secondary Maximum Contaminant Level or SMCL) for the US

EPA National Secondary Drinking Water Regulations is a threshold odor number (TON) = 3.

The TON is defined as the greatest dilution of sample with odor-free water yielding a definitely

111

perceptible odor (Greenberg et al., 1992). We determined the TON for each VFAs species according to equation (1):

Odorant concentration (C ) Threshold odor number = gas (1) Odor threshold value for that odorant

The “odor threshold value” is the lowest concentration of a specific odorant detectable by human olfaction. The “odorant concentration” is the gas phase concentration (Cgas) of a specific odorant (e.g., VFAs), which can be estimated based on the measured aqueous concentration (Caq).

Note that Caq is the total concentration comprising both the weak acid (i.e., HA, the protonated form susceptible to volatilization) and its conjugated base (i.e., A-, which is charged and not susceptible to volatilization). The concentration of the protonated form that can undergo volatilization (and thus generate odor), CHA, can be calculated based on the measured Caq, the pH of the solution, and the corresponding acid/base equilibrium constant (Ka) and molecular weight (MW) according to equation (2):

C (mg / L )× 10−3 ( g / mg ) = aq C(/)HA mol L − pH (2) MW (g / mol )×+ (1 Ka / 10 )

Cgas can be calculated using Henry’s law (equation (3)), where KH is Henry’s law constant:

3 3 3 C HA (mol / L) ×10 (L / m ) × K H (atm ⋅ m / mol) 6 Cgas ( ppmv ) = ×10 (3) 1 atm

Two representative samples of different seasons (A1, Jan 8th and A1, May 29th) were chosen to assess the seasonal variation of odor generation. The groundwater temperature and pH for these two samples were 6.6 ºC, pH 6.6 for (A1, Jan 8th) and 26.9 ºC, pH 4.6 for (A1, May

29th). Table 6.1 summarizes the calculated Cgas values, and Table 6.2 depicts the odor threshold

112

value for each VFAs and the TON values for each sample. Specific odor occurrence and impact will vary between direct testing methods and specific use scenarios (drinking, cooking, washing, showering, etc.).

Table 6.1. Calculated VFAs gas phase concentrations

Measured CHA Henry’s law constant Cgas VFAs pKa 3 (b) Caq (mg/L) (mol/L) (atm m /mol) (ppmv)

Summer (sampled at A1; May 29th;26.9 ºC) Acetic acid 116 4.75(a) 1.02×10-3 1.08×10-7 1.10×10-1 Propionic acid 7 4.87(a) 5.64×10-5 4.42×10-7 2.49×10-2 Butyric acid 280 4.85(a) 1.86×10-3 5.62×10-7 1.04 Winter ( sampled at A1; Jan 8th; 6.6 ºC) Acetic acid 25 4.75(a) 5.80×10-6 4.58×10-8 2.66×10-4 Propionic acid 4 4.87(a) 9.87×10-7 2.17×10-7 2.15×10-4 Butyric acid 3 4.85(a) 5.95×10-7 3.23×10-7 1.92×10-4 Source: (a) (Schwarzenbach et al., 2002) (b) Henry’s constants were obtained from (Howard, 1990) for acetic acid, and from (Howard, 1997) for propionic and butyric acids. These constants were corrected for the corresponding temperature using the Van’t Hoff equation, using standard entropy values from Haynes (2010)

113

Table 6.2 VFAs threshold odor number

Odor threshold value C Threshold odor VFAs gas (ppmv) (ppmv) number (TON)

Summer (sampled at A1; May 29th;26.9 ºC) Acetic acid 1(a) 1.10×10-1 0.1 Propionic acid 0.0057(b) 2.29×10-2 4.4 Butyric acid 0.001(a) 1.04 1045 Winter (sampled at A1; Jan 8th; 6.6 ºC) Acetic acid 1(a) 2.66×10-4 < 0.1 Propionic acid 0.0057(b) 2.15×10-4 < 0.1 Butyric acid 0.001(a) 1.92×10-4 0.2 Source: (a)(Cheremisinoff, 1999) (b) (Nagata, 2003)

For simplicity, we assumed that only acetic acid, propionic acid and butyric acid contribute to the odor in the groundwater sample. The threshold odor number of the summer sample (A1,

May 29th, 2010) (1,045 TON) was much larger than SMCL, and butyric acid was the major

contributor to odor generation. The threshold odor number of the winter sample (A1, Jan 8th)

(<0.4 TON), however, was lower than the SMCL. As discussed previously, lower temperature

decreased microbial activities (including transformation of ethanol into VFAs) which mitigated

odor generation. Overall, the results indicate that near a source, ethanol releases to groundwater

can generate odor problems that compromise water quality, but the level of impact would likely

vary seasonally. Therefore, seasonal variation of odor generation should be considered at sites

contaminated with fuel ethanol blends.

114

6.3.3 Effect of temperature on CH4 production

Within the channel exposed to the ethanol, dissolved CH4 in A1 increased from <0.1 mg/L

(August 7th, 2009, 29.9 ºC) to 6.8 mg/L (December 18th, 2009, 10.8 ºC) and then decreased to less than 0.5 mg/L concomitantly with the lower temperatures in January and February (<10 ºC,

th Figure 6.5). CH4 concentration then increased 0.2 mg/L (March 29 , 2010) to 12.9 mg/L (June

9th, 2010) with the increasing temperatures (from 16.0 ºC to 30.0 ºC). A similar trend was observed at the B1 sampling well. The maximum CH4 concentration was 17.9 mg/L (B1, May

29th , 2010, 26.9 °C), representing 81% of the solubility limit at the corresponding temperature

(Yamamoto et al., 1976). CH4 was not detected in the control channel amended with only benzene and toluene (Channel 2) over the 11 month period. Similar to acetic acid, a strong correlation existed between CH4 production (A1) and water temperature (p = 0.00075) (Figure

6.6). Methanogenesis is known to be enhanced at higher temperatures and inhibited by low temperatures (Cullimore et al., 1985; Conrad et al., 1987; Westermann, 1993).

115

15 A1 35 A2 Groundwater 30 12 temperature )

25 C o ( 9 20 (mg/L) 4 6 15 CH

10 Temperature 3 5

0 0 0 100 150 200 250 300 Days after B/T/(E) release

Figure 6.5 CH4 concentration at sampling well A1 (in Channel 1, exposed to ethanol, benzene, and toluene) and A2 (in Channel 2, exposed only to benzene and toluene). Day 0 corresponds to August 17th, 2009.

Figure 6.6 Correlations between CH4 concentrations (measured at A1) and groundwater temperature.

A BX 168 portable combustible gas detector (Henan Hanwei Electronics Co. Ltd, China)

(detection limit: 1% of CH4 lower explosive limit or 400 ppmv CH4) was used to analyze for CH4 116

concentrations in the air just above the sand surface of the ethanol-amended channel. No CH4

was detected, probably due to dilution by air movement as CH4 reaches the surface as well as to

some possible CH4 biodegradation by methanotrophs in the vadose zone (Chapter 4).

6.3.4 Ethanol, benzene, and toluene attenuation

Attenuation of ethanol, benzene, and toluene in Channel 1 was also affected by temperature

(Figure 6.7). The data in Figure 6.7 are plotted as normalized solute concentrations (C/Co)i

divided by the normalized bromide concentrations (C/Co)Br. When plotted in this way,

attenuation due to dilution is separated from attenuation resulting from biodegradation and volatilization. For ethanol, a short acclimation period with negligible attenuation was followed by significant removal; then, the low temperature winter conditions occurred and little ethanol degradation was observed. Benzene and toluene similarly experienced lower attenuation during the winter. Significant attenuation for ethanol, benzene, and toluene returned in the spring as temperatures increased. Attenuation of toluene was generally one order of magnitude greater than that for benzene.

Because the injected mixtures for both channels were the same except for the ethanol concentration, the absence of the lighter ethanol in Channel 2 could have resulted in a denser solute plume. Additional sample points from different depths were collected and analyzed, but a

solute plume was not identified in this channel. Since monitoring did not identify the location

and fate of the benzene and toluene plume in Channel 2, a comparison of attenuation of benzene

and toluene in the presence versus the absence of ethanol was not possible. 117

Ethanol Benzene 10 Toluene 35 Groundwater Temperature 30 Br C) ) 1 o 0 25 C/C (

/ 20 i )

0 0.1 15 C/C (

10 Temperature ( 0.01 5

0.001 0 0 100 150 200 250 300 350 Days after B/T/(E) release Figure 6.7 Ethanol, benzene, and toluene attenuation at sampling well A1 (in channel 1, exposed to ethanol, benzene, and toluene). Day 0 corresponds to August 17th, 2009.

6.4 Conclusion

A strong correlation was observed between water temperature and CH4/VFAs

concentrations (p < 0.05) and associated odor generation within the channel exposed to

continuously released ethanol. The main contributor to water odor was butyric acid, which

accumulated at levels that exceeded the SMCL stipulated by National Secondary Drinking Water

Regulations. The production of methane up to an aqueous concentration of 17.9 mg/L did not

result in detectable concentrations at the surface (45 cm above the water table). Overall, these

results show that groundwater temperature fluctuations can influence methane and VFAs

generation. Therefore, seasonal variation of odor generation in the subsurface should be

considered at sites contaminated with fuel ethanol blends.

118

Chapter 7

Response to the shut-off of a pilot-scale ethanol blended release: increased ethanol degradation activities and persistent methanogenesis

[Extract from a manuscript in preparation]

7. 1 Introduction

As higher ethanol blend fuels (E15 up to E95) are introduced to the market and the probability of related releases increase, there is growing interest in discerning potential impacts from ethanol degradation byproducts. Several ethanol degradation byproducts (e.g., volatile fatty acids, butanol, and methane) may have negative impacts on public safety, groundwater quality or natural attenuation processes. For example, high concentration of acetate could hinder the thermodynamic feasibility of anaerobic benzene degradation, thus undesirably increasing the length of benzene plume (Corseuil et al., 2011). Volatile fatty acids (particularly butyric acid) generate odor that could compromise groundwater aesthetic quality (Chapter 6). Butanol is a regulated compound in drinking water standards of several states in U.S. (Nelson et al., 2010).

119

High concentration of methane may cause vapor intrusion problems (Chapter 4 and 5). Whereas several studies have quantified the formation of such degradation byproducts following discrete or continuous release of ethanol blends (Capiro et al., 2007; Corseuil et al., 2011; Ma et al.,

2011), very few studies (if any) have considered how the system responds following source removal. Therefore, we conducted a comprehensive investigation of changes in the concentrations of ethanol degradation byproducts as well as in the abundance of methane production and oxidation genes in following the shut-off of the continuous pilot-scale release.

It was expected that the concentrations of ethanol degradation byproducts would decrease with decreasing ethanol influx, however, the pilot release experiment revealed two unexpected phenomena. First, after the release was shut off, anaerobic degradation of ethanol was temporarily stimulated. Apparently, this was related to a decrease in the dissolved ethanol concentration below its toxicity threshold (~2,000 mg/L for this system). Second, methane generation persisted for more than 100 days after the disappearance of dissolved ethanol, despite clean water being continuously injected and flowing through the tank at a relative high seepage velocity of 0.76 m/day. The possible reason that leads to these phenomena and the associated engineering significance are discussed in this chapter.

120

7.2 Materials and Methods

7.2.1 Pilot-scale aquifer system

The plan view and profile view of the sampling port for this study are showed in Figure 7.1.

Details on tank construction and release scenario can be found in Chapter 3.

Figure 7.1 Plan view (a) and profile view (b) of the aquifer tank.

121

7.2.2 Chemical analysis methods

Groundwater samples were collected from four centerline groundwater sampling ports

(Figure 7.1) using 60 mL plastic syringes (Thermo Fisher Scientific, Waltham, MA, US).

Ethanol, methane, acetate, propionate, butyrate and butanol were measured by GC-FID. Benzene and toluene were measured by GC-MS. The sodium bromide tracer was measured with a

bromide ion selective electrode. Details on groundwater sampling and chemical measurement

can be found in Chapter 3.

7.2.3 Microcosm experiment

A microcosm study was conducted to assess potential inhibitory effects of high ethanol

levels and to characterize the distribution of degradation products at different ethanol

concentrations. To set up the microcosm, ~1.3L tank water was collected from the sampling port

D and ~ 800 g sand was collected from a depth of 50-75 cm below ground surface near the

groundwater sampling port D in October of 2012 (Figure 7.1). Microcosms were set up in an

anaerobic chamber. The sand was homogenized before use. 15 g sand were mixed with 30 mL

tank water and various concentrations of ethanol (0, 500, 1,000, 1,500, 2,000, 3,400, 10,000,

40,000, 80,000 mg/L). The mixture was placed in sterile 125-mL serum bottles and sealed with

gas-tight butyl rubber stoppers and aluminum crimp caps. Four replicate microcosms were

prepared for each ethanol concentration. The microcosms were incubated in room temperature

(24 ºC) for 133 days. Methane concentrations in the headspace were monitored using GC-FID.

On the last day of microcosm experiment (day 133), water samples in the microcosm were 122

collected to measure the concentrations of ethanol degradation byproducts including acetate,

propionate, butyrate and butanol.

7.2.4 Sand organic carbon content

The total organic carbon content of sand samples from different experimental stages (five replicates for each stage) was measured by Soil, Water and Forage Testing Laboratory at Texas

A&M University. The analytical method could be found on their website

(http://soiltesting.tamu.edu/webpages/swftlmethods1209.html). A variety of other sand chemical

characteristics including soil pH, conductivity, nitrate-nitrogen, P, K, Ca, Mg, S, Na, Fe, Zn, Mn,

Cu were also measured in the same lab. These data could be found in Table A2.1 in Appendix II.

7.2.5 Sand sample collection and DNA extraction

Sand samples were collected from a depth of 50-75 cm below ground surface (Figure 1) every month during the entire experimental period. Details on sand sampling can be found in

Chapter 3. For qPCR analysis, DNA was extracted in triplicate from 0.25 g of sand using a

PowerSoil DNA Isolation Kit (MOBIO, Carlsbad, CA, US). Details on DNA extraction can be found in Chapter 8.

7.2.6 qPCR analysis

Absolute quantification was used to enumerate the gene copy number for two functional genes associated with methane metabolism: 1) mcrA for methanogenesis and pmoA for

123

methanotrophs. The target genes, primer sequences, and DNA standards for calibration are given

in Table 7.1. Details of the qPCR method can be found in Chapter 3.

Table 7.1 Primers and probes for qPCR analysis

Annealing Standard Gene Primer Sequence temp. (ºC) genomic DNA

ML-f 5’-GGTGGTGTMGGATTCACACARTAYGCWACAGC-3’ Methanocaldoccus mcrA 55 °C ML-r 5’- TTCATTGCRTAGTTWGGRTAGTT-3’ jannaschii

MBAC-f 5’-GGNGACTGGGACTTCTGG-3’ Methylomicrobium pmoA 54 °C MBAC-r 5’-ACRTAGTGGTAACCTTGYAA-3’ album

7.2.7 Microarray

The microarray study used the same DNA samples extracted that were used in Chapter 9,

except that Stage 5 in the Chapter 9 was not analyzed in this study. Details on sand sampling and

DNA extraction method can be found in Chapter 9 and Chapter 3.

Geochip version 4.6 was used in this study. For each sample, 800 ng DNA was labeled

with the fluorescent dye Cy3 by random priming and the Klenow fragment of DNA polymerase I.

The labeled DNA was purified with QIAquick purification kit (Qiagen, Valencia, CA, US),

measured by NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE,

US) to check the label efficiency, and dried in a SpeedVac (ThermoSavant, Milford, MA, US) at

45 ºC for 45 min.

Dried DNA was rehydrated with 2.68 uL sample tracking control (NimbleGen, Madison,

WI, US) to confirm sample identity. The mixture was incubated at 50 ºC for 5 min. 7.32 µL

hybridization solution, which contains 40% formamide, 25% SSC, 1%SDS, 2% Cy5-labeled

124

common oligo reference standard target, and 2.38% Cy3-labeled alignment oligo (NimbleGen,

Madison, WI, US) was added to each sample. The mixture was incubated at 95 ºC for 5 min and

then maintained at 42 ºC until ready for hybridization.

Geochip 4.6 was synthesized by NimbleGen for 12 arrays per slide. The slide was

preheated to 42 ºC on MUNI hybridization station (BioMicro Systems, Salt Lake City, UT, US)

for at least 5 min. 6.8 µL DNA sample was then loaded onto the array hybridized at 42 ºC for 16

h. After washing and drying, the microarrays were scanned by using a MS200 Microarray

Scanner (NimbleGen, Madison, WI, US) with a laser power of 100% and 100% PMT.

7.2.8 Microarray Data processing

Spots on the microarray were scored as positive if the signal-to-noise ratio (SNR) was ≥

2.0 and the coefficient of variations (CV) of the background was < 0.8. The gene that was detected in only one sample was removed as noise. The microarray data was then normalized in two steps. First, the average Cy5 intensity of the universal standard in each sub-grid was used to normalize the Cy3 intensity for samples in the same sub-grid. Second, the Cy3 intensity of the sample after the first normalization was normalized by the average Cy5 intensity of universal standards in all slides. Hybridization data are available at the web site of the Institute for

Environmental Genomics (http://ieg.ou.edu).

125

7.3 Results

7.3.1 Microcosm experiments for ethanol toxicity

A series of microcosms with different initial ethanol concentrations were prepared to assess changes in degradation byproducts distribution and potential inhibitory effects at high concentrations. Four parameters were inferred from microcosm data, including 1) the acclimation time for methane generation, 2) methane accumulation after 133 days of incubation,

3) fitted methane generation rate, and 4) total dissolved concentrations of ethanol degradation

byproducts after 133 days of incubation (Figure 7.2, Figure 7.3 and Table 7.2).

The microcosm containing 1,000-2,000 mg/L ethanol had the highest ethanol

biodegradation activity. The microcosm containing higher initial ethanol concentrations

experienced longer acclimation time (Figure 7.2b). The descending order of methane

accumulation after 133 days of incubation is 2,000 mg/L > 1,500 mg/L ≈3,400 mg/L > 10,000

mg/L > 1,000 mg/L > 500 mg/L >> 40,000 mg/L > 80,000 mg/L (Figure 7.2a). The descending

order of fitted methane generation rate is 1,000 mg/L > 500 mg/L > 2,000 mg/L > 1,500 mg/L >

3,400 mg/L > 10,000 mg/L > 40,000 mg/L > 0 mg/L (Table 7.2). The descending order of total

concentration of ethanol metabolites after 133 days of incubation is 2,000 mg/L > 1,500 mg/L >

10,000 mg/L > 3,400 mg/L > 1,000 mg/L > 500 mg/L (Figure 7.3). These data show that ethanol

degradation activities of sand samples in the tank are partially inhibited by 3,400 to 40,000 mg/L

ethanol and are strongly inhibited by > 40,000 mg/L ethanol.

126

5x104 Ethanol concentration 1x104 Ethanol concentration (mg/L) (mg/L) ) 0 ) v

v 0 4x104 500 500 1000 1000 1500 1500 4 2000 2000 3x10 3400 3400 10000 10000 4 40000 40000 2x10 80000 80000 concentration (ppm

concentration (ppm 3 4 4 1x10 4 CH

1x10 CH

0 0 10 20 30 40 50 60 120 0 2 4 6 8 10 (a) Days (b) Days Figure 7.2 Methane accumulations in the headspace of microcosms. (a) is the whole experimental period (133 days) and (b) is the first 12 days (methane concentrations are on log scale). Curves represent different microcosms with different initial ethanol concentrations. The microcosm prepared with 2,000 mg/L of ethanol had the highest level of methane accumulation (a). The microcosm containing 500 mg/L of ethanol had the shortest acclimation time.

127

Table 7.2 Fitted methane generation rates in the microcosm experiment Initial ethanol Fitted methane generation rate concentration (mg /L) (g/day) *

0 2.9 × 10-4

500 7.8 × 10-3

1,000 1.0 × 10-2

1,500 5.2 × 10-3

2,000 5.8 × 10-3

3,400 3.6 × 10-3

10,000 3.5 × 10-3

40,000 1.1 × 10-3

80,000 0 * The time range that was used to fit methane generation rate was: day 5-20 for 0 mg/L, day 5-15 for 500 mg/L, day 5-20 for 1,000 mg/L, day 8-50 for 1,500 mg/L, day 8-60 for 2,000 mg/L, day 8-60 for 3,400 mg/L, day 8-40 for 10,000 mg/L, day 60-133 for 40,000 mg/L, and day 0-133 for 80,000 mg/ L. The regression fitting plots can be found in Appendix I.

128

Butanol 800 Butyrate Propionate Acetate 600

400

200 Degradation products (mg/L) 0 0 500 1000 1500 2000 3400 10000 40000 80000

Ethanol (mg/L) Figure 7.3 Dissolved concentrations of ethanol degradation byproducts in each microcosm after 133 days of incubation. The microcosm containing 2,000 mg/L ethanol has the highest total concentration of dissolved ethanol degradation byproducts.

7.3.2 Higher ethanol degradation activity following source removal

After the ethanol release was shut off, dissolved ethanol concentrations gradually decreased to zero within 60 days at sampling ports C and D (Figure 7.4). Unlike ethanol, the concentration of acetate at the port D first increased 2-2.5 fold and then decreased to zero within

90 days (Figure 7.4). Similar to acetate, butyrate and butanol concentrations at the port D also

temporarily increased and their peak concentrations appeared following acetate peak

concentration (Figure 7.4a). In the ethanol anaerobic degradation pathway, ethanol is first

metabolized to acetate, which can then be transformed to butyrate and butanol (Chapter 2). The successive increases in the concentrations of acetate, butyrate and butanol corroborate that the fermentative biodegradation of ethanol at the sampling port D was temporarily stimulated

129

following the shut-off of the ethanol release. The transient accumulation of acetate was also observed at the sampling port C, but there were no obvious increases in butyrate and butanol concentrations following the increase of acetate concentration. It may be because the port C is closer to the source zone (ethanol/benzene/toluene inject port) than the port D. When the residual ethanol plume passed by the port C, the reaction time is not long enough to have obvious accumulation of butyrate and butanol.

130

Ethanol 3500 300 Acetate Butyrate 300 2800 Butanol 60 200 2100 200 40 1400

Ethanol (mg/L) 100 Acetate (mg/L) Butanol (mg/L) 100 Butyrate (mg/L) 20 700

0 0 0 0 0 10 20 30 40 50 60 70 80 (a) Days Ethanol Acetate 250 60 3500 Butyrate Butanol 150 200 2800 40 150 2100 100

1400 100 Ethanol (mg/L) 20 Acetate (mg/L) Butyrate (mg/L)

50 Butanol (mg/L) 700 50

0 0 0 0 0 10 20 30 40 50 (b) Days

Figure 7.4 Changes in the concentrations of ethanol and its degradation byproducts at the groundwater sampling port D (a) and C (b).

7.3.3 Persistence of dissolved methane and methane metabolism genes

Methane persisted in the tank aquifer even several months after the apparent disappearance of ethanol and its metabolites (i.e., acetate, butyrate and butanol). At the sampling

131

port C, ethanol and its metabolites disappeared after day 47, while methane persisted in the

groundwater until day 131 (Figure 7.5 b). Similarly, at the sampling port D ethanol and its

metabolites disappeared after day 89, while methane persisted until day 212 (Figure 7.5 a). The

persistence of dissolved methane has also been reported in some field studies (Buscheck et al.,

2001; Corseuil et al., 2011; Spalding et al., 2011; Sihota et al., 2013). At a site impacted by an

accidental fuel ethanol spill, the ethanol concentration had decreased from ~ 20,000 mg/L to

below detection limit (0.1 mg/L) within the first 3 years, but 20 mg/L of dissolved methane was

detected even 6 years after the spill (Spalding et al., 2011; Sihota et al., 2013). At an experimental site impacted by a pulse injection of Brazilian gasoline which contains 24% (v:v) of ethanol, oversaturated dissolved methane (> 24 mg/L) was continuously detected in the groundwater even 6.5 years after the spill (Corseuil et al., 2011). Based on these field studies and this well-controlled pilot study, it could be concluded that methane persistence is likely to be a common phenomenon at sites impacted by ethanol blend fuel releases.

132

5x103 8x104

CH 4x103 4 Acetate 6x104 Butyrate 3x103 Butanol Ethanol 4x104 2x103 CH solubility

4 Ethanol (mg/L) 2x104 1x103 , acetate, butyrate, butanol (umol/L) 4

CH 0 0 0 20 40 60 80 100 (a) Days

5x103 8x104

3 CH4 4x10 4 Acetate 6x10 Butyrate 3x103 Butanol Ethanol 4x104 2x103 CH solubility

4 Ethanol (umol/L) 2x104 1x103 , acetate, butyrate, butanol (umol/L) 4

CH 0 0 0 50 100 150 200 (b) Days Figure 7.5 Comparison of CH4 concentrations with ethanol and its degradation byproducts concentrations from groundwater sampling port D (a) and C (b).

In addition to methane accumulation, the functional genes associated with methane metabolism also persisted in the aquifer. Two functional genes were selected for this study: 1)

133

methyl coenzyme-M reductase gene (mcrA) for methanogens and 2) methane monooxygenase

gene (pmoA) for methanotrophs (Figure 7.6). Ethanol releases stimulated strong methanogenic

activity. Accordingly, the abundance of mcrA gene increases for four orders of magnitude from

9.8×102 copy/g dry sand (pre-contamination baseline, August 7st, 2009) to 6.4×106 copy/g dry sand (exposed to ethanol blended solution for 2 years, September 5th, 2011). Even one year after

the ethanol solution was shut off, a high abundance of mcrA gene (~106 gene copy/g dry sand)

still existed in the aquifer, thus indicating the persistence of methanogens in this system. The

pmoA gene showed a similar temporal trend as the mcrA gene. The persistence of methanogenic

and methotrophic microorganisms was corroborated by 16S rRNA amplicon pyrosequencing.

Details on pyrosequencing results can be found in Chapter 9.

108 Shutoff of ethanol release 107 106 105 104 mcrA pmoA 103

Gene copy number / copy number g dryGene sand 10 -760 0 100 200 300 Days

Figure 7.6 Changes in the copy numbers of methane metabolism functional gene following the shutoff ethanol release. mrcA is the functional gene for methanogenesis and pmoA is the functional gene for methanotrophs.

134

7.3.4 Ethanol releases stimulate microbial EPS production

Fuel ethanol releases stimulate the growth of microbes that extrude extracellular

polymeric substance (EPS) and increase the organic carbon content in the impacted aquifer

materials. The total organic carbon content of the sand samples from differential experimental

stages (Stage 1, 2, and 4a) were measured (Figure 7.7). The timeline for each experimental stage

can be found in Figure 3.7 of Chapter 3. As the sand sample exposed to ethanol blend release for

2 years, Stage 2 has 85% higher organic carbon content than Stage 1, which was the pre-

contaminated sand sample (Figure 7.7). As the sand sample that was collected 1 year after the

shut-off of ethanol release, Stage 4a has 34% less organic carbon content than Stage 2, thus

indicating a net consumption of sand organic carbon after the shut-off of ethanol release.

0.15

0.12

0.09

0.06

0.03 Soil organic carbon (%)

0.00 Stage 1 Stage 2 Stage 4a

Figure 7.7 Sand total organic carbon content. Stage 1 is the pre-contaminated sand sample. Stage 2 is the sand sample exposed to a continuous ethanol blend release for 2 years. Stage 4a is the sand sample collected 1 year after the ethanol was removed from release solution. Detailed information about the timeline for each experimental stage can be found in Figure 3.7 of Chapter 3. 135

In addition to organic carbon content data, several functional genes involved in extracellular polymeric substance (EPS) production were significantly (p < 0.05) enriched in the sand sample exposed to ethanol blend release. Functional gene of sand samples from four different experimental stages (Stage 1-4a) were analyzed by a comprehensive functional gene microarray (Geochip), which contains 11 EPS production genes. Compared to Stage 1, seven

EPS production genes were enriched at Stage 2, including gene encoding for mannose-1- phosphate guanylyltransferase, LPS heptosyltransferase, NAD dependent epimerase dehydratase family protein, UDP-3-O-[3-hydroxymyristoyl] glucosamine N-acyltransferase, capsular polysaccharide biosynthesis protein, glycosyl transferase, tyrosine protein kinase (Figure 7.8).

The detection of EPS production genes corroborates that fuel ethanol releases could stimulate microbial EPS production, thus increasing the sand organic carbon content.

136

a 6 a Stage 1 - Clean sand b Stage 2 - Exposed to ethanol blend for 2 years Stage 4a - 1 year after the shutoff of ethanol solution 5

baa 2

baa aa aa aa a b b Normalized signal intensity 1 b bb

ADP heptose synthase glycosyl transferasetyrosine protein kinase LPS heptosyltransferaseNAD-dependent epimerase undecaprenyl diphosphate synthase polysaccharide biosynthesis protein mannose-1-phosphate guanylyltransferase capsular polysaccharide biosynthesis protein NAD dependent epimerase dehydratase family protein UDP-3-O-[3-hydroxymyristoyl] glucosamine N-acyltransferase Figure 7.8 Normalized signal intensity of detected functional genes for extracellular polymeric substance (EPS) production at different experimental stages. Gene number is the protein ID number for each gene as listed in the GenBank database. Different letters indicate statistical differences at a p value of < 0.05 among treatments by Fisher’s least-significant-difference (LSD) test. No letter was labeled if there was no statistical difference (p > 0.05).

7.4 Discussion

7.4.1 Higher ethanol degradation activity following source shutoff is due to lower ethanol toxicity

Previous studies reported that 3.1%-6% (v:v) ethanol could strongly inhibit microbial growth and activities (Ingram and Vreeland, 1980; Nelson et al., 2010), which is consistent with our results (the toxicity threshold we found is 40,000 mg/L, or 5.1 % v:v). However, such a high

concentration could only be found near the non-aqueous phase liquid (NAPL) phase ethanol source, while the commonly reported concentration of dissolved ethanol in impacted 137

groundwater is usually lower than 10,000 mg/L (Corseuil et al., 2011; Spalding et al., 2011).

Since the active zone for biodegradation is usually located within dissolved plume (especially at

the plume edge) rather than in NAPL phase source zone, toxicity information for < 10,000 mg/L

ethanol is of more practical significance to predict the degradation and environmental behavior

of ethanol release. This study shows that 3,400 to 10,000 mg/L of ethanol has partial inhibitory

effect on ethanol biodegradation in this pilot aquifer system. Note that this partial inhibitory

threshold may vary for different soil/sand/sediment samples.

Results from this microcosm experiment are supported by previous biochemical and

microbiological studies. The activity of succinate dehydrogenase (enzyme in Kreb’s cycle)

shows 20% inhibition with 3,350 mg/L ethanol and 80% inhibition with 8,500 mg/L ethanol

(Eaton et al., 1982). The activities of the lactose, glutamate, proline, and leucine permease (cell

membrane transport) shows 10%-30% inhibition with 3,350 mg/L ethanol and 60%-80%

inhibition with 8,500 mg/L ethanol (Eaton et al., 1982). Bringmann and Kuhn reported that 6,500

mg/L ethanol had significant negative impact on the growth of Pseudomonas putida, an

important aromatic hydrocarbon degrader (Bringmann and Kuhn, 1980). 8,000 mg/L ethanol

could reduce the growth rate of P. putida S12 by 30% (Heipieper and Debont, 1994). These biochemical and microbiological studies corroborate our microcosm results, indicating that depending on site conditions the common ethanol concentration detected in the impacted groundwater (3,000 to 10,000 mg/L) could have partial inhibitory effects on microbial growth and activities.

138

The temporary stimulation of ethanol degradation in the tank can be explained by ethanol

toxicity. The initial concentration of ethanol detected at port C and D were 3,300-3,500 mg/L

(Figure 3), a level that may pose partial inhibitory effects on microbial population. When ethanol

concentration dropped below a certain level (~2,000 mg/L) following the source removal, such

inhibitory effect disappeared and indigenous microbial population was able to exert their full

metabolic capacity, thus leading to a temporary increase in the concentrations of ethanol

metabolites. Note that the sand used in the microcosm experiments were collected one year after

the ethanol was removed from the release solution (October 2012), thus the microbial population

in the microcosm may not be exactly the same with the indigenous microbial population in the

tank when the ethanol solution was just shut off (September 2011).

7.4.2 Persistent methane accumulation may be due to anaerobic degradation of remaining

acetate and black slime

Although this study and several other field studies corroborate that methane persistence may be a common phenomenon at fuel ethanol impacted sites, previous studies did not clarify

the reason that causes this phenomenon. In this model tank, tap water was continuously injected

into and flowed out of the tank, with an average seepage velocity of 0.76 m/day. Since methane is unlikely to be adsorbed and retained by aquifer materials, methane accumulating post shut off must have been generated from anaerobic degradation of residual organic byproducts. Figure 7.5 shows that ethanol metabolites (acetate, butyrate and butanol) remained in the groundwater for a longer time than ethanol after the ethanol solution was shut off. Therefore, it is very likely that

139

ethanol metabolites (especially acetate) contributed to the persistence of dissolved methane in the tank.

However, methane still persisted for more than 80 days after the apparent disappearance of acetate, butyrate and butanol (Figure 4). There must be an alternative carbon source to support the activity of methanogens and the only available alternative carbon source is the solid organic carbon in aquifer materials. Both Geochip data and the sand organic carbon content data show that the ethanol blend release could stimulate the growth of microorganisms that excrete EPS and increase the organic carbon content in the impacted sand. Therefore, it can be inferred that the methanogens may use the solid organic carbon in the sand to survive and produce methane, thus resulting in the persistence of dissolved methane in the tank.

140

Chapter 8

Adaptive microbial population shifts in response to a continuous ethanol blend release increases natural attenuation potential

[Modified from Ma et al. 2013a-published in Environmental Pollution (Ma et al., 2013b)]

8.1 Introduction

Microorganisms play a key role in the transformation of ethanol blend fuels, thus determining their fate and influential area. Previous studies investigating the microbial response

to fuel ethanol releases focused on individual (phylogenetic or catabolic) genotypes associated with aromatic hydrocarbon degradation (Da Silva and Alvarez, 2004; Beller et al., 2008; Capiro et al., 2008; Feris et al., 2008; Nelson et al., 2010). However, such approaches generally detect only predominant or well characterized genotypes whereas contaminant biodegradation is

usually carried out by a complex microbial food web (de Lorenzo, 2008), and smaller populations that fill important niches may remain undetected (Osborn and Smith, 2005).

Therefore, the characterization of microbial population shifts in response to ethanol blend

141

releases would benefit from using high-throughput tools that enable high sensitivity and

resolution analysis of the community structure and its functional diversity.

Recently, pyrosequencing technologies have enhanced the study of microbial communities

with unprecedented coverage and resolution (Cardenas and Tiedje, 2008). However, the

application of pyrosequencing to study the natural attenuation of ethanol-blended fuel releases

has not been reported. In this chapter, pyrosequencing and quantitative real-time PCR (qPCR)

analyses are conducted simultaneously to characterize changes in microbial community structure

in response to a 10-month continuous (pilot-scale) release of an ethanol-blended solution.

Metabolic byproducts (e.g., methane and acetate) and several environmental variables

(temperature, pH, redox potential and dissolved oxygen) were monitored to enhance our understanding of the relationship between bioremediation processes, biogeochemical indicators, and microbial population shifts.

8.2 Materials and Methods

8.2.1 Pilot-scale aquifer system

The plan view and profile view of the sampling port for this study are showed in Figure 8.1.

Details on tank construction and release scenario can be found in Chapter 3.

142

Figure 8.1 Plan view (a) and profile view (b) of the pilot-scale aquifer system.

8.2.2 Chemical and geochemical analyses

Groundwater geochemical footprints including temperature, pH, oxidation-reduction

potential (ORP), dissolved O2 (DO) and conductivity were monitored 1 by YSI 600XLM water

quality probe (YSI, Yellow Springs, Ohio, US) installed at M2 in channel 1(Figure 8.1).

Groundwater samples were collected from a groundwater sampling port 150 cm down gradient from the source. Ethanol, methane and volatile fatty acids (VFAs) including acetic, propionic, and butyric acid in the groundwater were measured by GC-FID. The groundwater sampling and

143

chemical analysis method were the same with Chapter 6, except that a different groundwater

sampling port was used in this study. (B1 in Figure 6.1 of Chapter 6)

8.2.3 Sand sampling and qPCR analysis

Sand core in the saturated zone were collected for 16S rRNA metagenomic pyrosequencing analysis and qPCR analysis. For qPCR analysis, DNA was extracted in triplicate from 0.25 g

sand using PowerSoil DNA Kit (MOBIO, Carlsbad, CA, US). qPCR analysis quantified both

phylogenetic genes including Bacteria/Archaea 16s rRNA gene and functional genes including

methanogenesis (mcrA), acetogenesis (fhs) and BTEX aerobic degradation (PHE). Details on

qPCR methods can be found in Chapter 3.

144

Table 8.1 Primers and probes for qPCR analysis

Target qPCR Primer Annealing Calibration Primer/probe sequences Reference gene reagent probe temperature DNA standard Bacterial TaqMan 1055f 55 °C 5’-ATGGCTGTCGTCAGCT3’ Escherichia (Harms et al., 16S 1392r 5’- ACGGGCGGTGTGTAC-3’ coli 2003) rRNA 16STaq1115 5‘-FAM-CAACGAGCGCAACCC-TAMRA-3’ Archaeal TaqMan ARCH1-1369F 59 °C 5’-CGGTGAATACGTCCCTGC-3’ Methanococcus (Suzuki et al., 16S ARCH2-1369F 5’-CGGTGAATATGCCCCTGC-3’ maripaludis 2000; Da Silva rRNA PROK1541R 5’-AAGGAGGTGATCCTGCCGCA-3’ and Alvarez, TM1389F 5’-FAM-CTTGTACACACCGCCCGTC0-BHQ-3’ 2004) mcrA SYBR Green MLf 55 °C 5’- GGTGGTGTMGGATTCACACARTAYGCWACAGC-3’ Methanocaldoc (Luton et al., MLr 5’- TTCATTGCRTAGTTWGGRTAGTT-3’ cus jannaschii 2002) fhs SYBR Green fhs1 55 °C 5’-GTWTGGGCWAARGGYGGMGAAGG-3’ Morrella (Xu et al., 2009) FTHFS-r 5’-GARGAYGGWTTTGAYATYAC-3’ thermoacetia PHE SYBR Green PHE-F 55 °C 5 -GTGCTGACSAAYCTGYTGTTC-3' Pseudomonas (Baldwin et al., PHE-R 5 -CGCCAGAACCAYTTRTC-3' putida CF600 2003) Bacterio- TaqMan 60 °C 5’-ACGCCACGCGGGATG-3’ Bacteriophage (Beller et al., phage λ 5’-AGAGACACGAAACGCCGTTC-3’ λ DNA 2002) 5’-TET-ACCTGTGGCATTTGTGCTGCCG-TAMRA-3’

145

8.2.4 Pyrosequencing

Four sand samples (t=0 day, t=123 days, t=184 days, and t=318 days) were used for

16S rRNA amplicon pyrosequencing analysis. DNA was extracted from 5 g sand using a

modified Qiagen DNeasy kit (Qiagen, Valencia, CA, US): three frozen (-80°C) and thaw

(60°C) cycles and a lysozyme treatment (incubated the sand sample in 20 mg/L lysozyme

(EMD Biosciences, San Diego, CA, US) solutions at 37 °C for 30 min) were added before kit extraction to increase DNA yield.

For the pyrosequencing library preparation, the V4 region of bacterial 16S rRNA gene was amplified using bacterial primers 563F (5’-AYTGGGYDTAAAGVG-3’) and 802R (5’-

TACNVGGGTATCTAATCC-3’) (Cardenas et al., 2010) and the V4 region of archaeal 16S rRNA gene was amplified using primers U519F (5’-CAGYMGCCRCGGKAAHACC-3’) and Arch 806R (5’-GGACTACNSGGGTMTCTAAT-3’) (Porat et al., 2010). PCR reactions

(50μL) were performed using FastStart High Fidelity PCR System (Roche Applied Science,

Indianapolis, IN, US) containing 0.4 μM each primer, 1U Reaction Buffer with 18 mM

MgCl2, 1U of each dNTP, 1U of FastStart high-fidelity PCR system enzyme blend, and 5 μL template DNA. The PCR conditions were as follows: 95°C for 3 min; 40 cycles of denaturation (95°C; 45 s), annealing (bacterial 16S rRNA gene: 57°C for 45 s and archaeal

16S rRNA gene: 55°C for 45 s) and extension (72°C; 1 min); followed by a final extension

(72°C; 4 min).

PCR products were separated in a 1% (wt/vol) gel. The bands between 270 and 300

bp were excised and DNA was recovered using a QIAquick gel extraction kit (Qiagen,

146

Valencia, CA, US). DNA samples were quantified using PicoGreen (Invitrogen, Eugene, OR,

US) and pooled in normalized quantities for direct sequencing by a Genome Sequencer

Junior System (454 Life Sciences, Branford, CT, US).

8.2.5 Pyrosequencing data analysis

A total of 51,924 raw sequences was obtained, deposited in the NIBI Sequence Read

Archive (BioSample accession ID: SAMN01889068, SAMN01889069, SAMN01889070,

SAMN01889071) and processed by the Ribosomal Database Project (RDP) Pyrosequencing

Pipeline (http://pyro.cme.msu.edu/) (Cole et al., 2009). Raw sequences were sorted by

barcode, and fusion primers were removed. The quality filter removed low quality sequences

with lengths less than 150 bases or having any ambiguity (N) or those with any change in

forward or reverse primers. After removing low quality sequences, 1,139-1,814 bacterial

sequences per sample and 2,024-11,570 archaeal sequences per sample were obtained (Table

8.2). Although deeper sequence efforts can yield more information on microbial diversity,

several studies have shown that similar sequencing efforts (even of lower depth, with only a

few hundred sequences per sample) can reliably discriminate microbial communities from different environments (e.g. saline versus nonsaline environments (Lozupone and Knight,

2007), soil samples from Florida versus Hawaii (Lemos et al., 2011), and gastrointestinal microflora in individuals with inflammatory bowel disease versus healthy individuals (Frank et al., 2007). Since our goal was to compare the microbial community in aquifer material before and after continued exposure to an ethanol-blend release, we deemed our level of sequencing analysis appropriate. Sequences were aligned by MUSCLE 3.5 (Multiple

147

sequence comparison by log-expectation) (http://www.drive5.com/muscle/) (Edgar, 2004) and were classified using RDP II classifier with a 50% bootstrap confidence (Wang et al.,

2007).

8.3 Results and Discussion

8.3.1 Summary of sequencing data

The bacterial 16S rRNA library and archaeal 16S rRNA library were pooled together and

sequenced on one 454 Jr sequencing run. A total of 51,924 raw sequences were obtained and

a total of 32,298 high quality sequences pass the quality filter. Among these high quality

sequences, 26,090 sequences were identified by archaeal primers and 6,208 sequences were

identified by bacterial primers. The number of sequences and the average length for each

sample are summarized in Table 8.2.

Table 8.2 Sequencing result summary Bacterial primer Archaeal primer Sequence number Average length (bp) Sequence number Average length (bp) t=0 day 1,139 239 2,024 276 t=123 days 1,249 239 2,791 276 t=184 days 1,814 239 8,182 276 t=318 days 1,654 239 11,570 276

8.3.2 Impacts of ethanol-blend release on the abundance of selected functional genes

To discern the effect of the release, two aquifer material samples were collected from the

same location (Figure 8.1) at different times: 1) a baseline (pre-release) sample (t=0 day); and

2) a sample exposed to the ethanol blend for 10 months (t=318 days). The corresponding

groundwater temperatures during sample collection were very similar (29.9°C for t=0 and

148

28.6°C for t=318 days, Figure 8.2) so that the temperature was not a confounding factor in

comparing the microbial communities in these samples. As a readily degradable compound,

ethanol represents a favorable carbon and energy source that stimulates microbial growth.

Accordingly, the ethanol-blend release increased total Bacteria and Archaea populations by

14- and 110-fold, respectively (Figure 8.3).

35 Release started 9 Release started

28 8

21 7 C o 14 6

7 5

0 4 Aug Sep Dec Feb Apr Jun Aug Sep Dec Feb Apr Jun (a) Temperature (b) pH

200 Release started 6 Release started

0 4 -200 mV mg/L 2 -400

-600 0 Aug Sep Dec Feb Apr Jun Aug Sep Dec Feb Apr Jun (c) Redox potential (d) Dissolved O 2 Figure 8.2 Seasonal changes in groundwater geochemical parameters following the release. (a) temperature, (b) pH, (c) redox potential, and (d) dissolved O2.

149

9 10 18 t=0 day (2009-8-7) t=0 day (2009-8-7) 8 t=318 days (2010-6-21) t=318 days (2010-6-21) 10 15

7 10 12

6 10 9

5 10 Percentage (%) 6 Gene CopiesGene /g Sand 104 3

103 0 Bacteria Archaea mcrA fhs PHE mcrA fhs PHE (a) Absolute abundance (b) Relative abundance

Figure 8.3 A comparison of the absolute (a) and relative (b) abundance of selected genes between baseline sample (t=0 day) and the sample following 10 months of continuous ethanol-blend release (t=318 days). The relative abundance is expressed as a ratio of target functional genes to the corresponding 16S rRNA genes. Specifically, fhs and PHE were normalized to bacterial 16S rRNA and mcrA was normalized to archaeal 16S rRNA. Both samples were collected during the summer season.

The ethanol-blend release also increased the relative abundance of functional genotypes

involved in the (syntrophic) anaerobic degradation of ethanol-blended fuel, such as the

formyltetrahydrofolate synthetase gene (fhs) involved in acetogenesis (Xu et al., 2009) and

the methyl coenzyme-M reductase gene (mcrA) that is critical for methanogenesis (Luton et

al., 2002) (Figure 8.3). The absolute abundance of fhs and mcrA increased 87- and 1,939-fold

respectively, and the relative abundance increased 18- and 6-fold, respectively. The fhs and

mcrA sequences are highly conserved, and are widely used as functional biomarkers for

acetogenesis and methanogenesis respectively (Leaphart and Lovell, 2001; Luton et al., 2002;

Friedrich, 2005; Xu et al., 2009). The high abundance of mcrA and fhs after 318 days is

consistent with the high CH4 and acetate groundwater concentrations measured in the

summer (Figure 6.5 and Figure 6.2 in Chapter 6), reflecting the strong methanogenic and

150

acetogenic activity stimulated by the release. The accumulation of ethanol fermentation byproducts (e.g., H2 and acetate) could thermodynamically inhibit (i.e., make endergonic) the anaerobic biodegradation of aromatic hydrocarbons such as benzene (Corseuil et al., 2011).

Therefore the enrichment of methanogens that consume H2 and acetate is conducive to enhanced anaerobic bioremediation of ethanol-blended fuel.

The abundance of gene PHE, which codes for phenol hydroxylase (an enzyme that catalyzes the oxidation of hydroxylated metabolites of monoaromatic hydrocarbons (Baldwin et al., 2003), also increased (43-fold, Figure 8.3). This gene is often detected under low or fluctuating dissolved oxygen conditions (Duetz et al., 1994; Leahy and Olsen, 1997; Baldwin et al., 2008; Capiro et al., 2008; Baldwin et al., 2009; Nebe et al., 2009; Da Silva and

Corseuil, 2012). The increase in PHE abundance is likely due not only to microbial growth on benzene and toluene, but also to the fortuitous growth on ethanol or its byproduct acetate

(Capiro et al., 2008). The growth of aromatic hydrocarbon degraders on ethanol would potentially enhance the rate of aerobic bioremediation in shallow aquifers after ethanol and its degradation by-products are degraded or flushed out.

8.3.3 Impacts of ethanol-blend release on microbial community structure

The ethanol blend release changed the indigenous bacterial community. Before the release, the bacterial community was mainly composed of aerobes such as Acinetobacter,

Pseudomonas, Acidovorax, Turneriella, Leptospira and Comamonas (Figure 8.4). Following

10 months of release, several genera that are likely involved in the degradation of ethanol blends were enriched, including Pseudomonas (from 5.4 to 13.0 % of the total number of

151

classified sequences), Bellilinea (from non-detected (ND) to 8.3 %), Prosthecobacter (from

ND to 6.5 %), Geothrix (from ND to 5.6 %), Opitutus (from ND to 5.3 %) and Xanthobacter

(from ND to 5.1 %) (Figure 8.4). The genus Pseudomonas includes many species that are commonly associated with the aerobic biodegradation of benzene and toluene (Cantwell et al.,

1978; Grimberg et al., 1996; Whyte et al., 1997; Reardon et al., 2000; Gupta et al., 2008; Lee et al., 2010). Similarly, Xanthobacter is associated with the degradation of a wide range of organic compounds (Zhou et al., 1999; Torz et al., 2007), including benzene and toluene at low oxygen concentrations (<0.2 ppm) (Hirano et al., 2004). Although ethanol-blended fuels rapidly induce anaerobic conditions (Capiro et al., 2008), oxygen entrainment in such open systems and aerobic micro-niches (Brune et al., 2000) may facilitate the aerobic biotransformation. Opitutus is a group of obligate anaerobes that produce acetate and propionate (Chin et al., 2001). Bellilinea is another group of obligate anaerobic–fermenters that produce volatile fatty acids (VFAs) (Yamada et al., 2007). Thus, Opitutus and Bellilinea may be predominant fermenters that initiate the anaerobic degradation of ethanol blends and interact syntrophically with methanogens. Geothrix is a group of strict anaerobes that grows on various simple organic acids such as acetate, propionate, and lactate (Coates et al., 1999).

Therefore, Geothrix may have participated in the consumption of volatile fatty acids (mainly acetate and butyrate (Ma et al., 2011)) generated during the anaerobic degradation of the release.

152

Figure 8.4 Changes in bacterial communities (genus level) before (a) and after (b) 10 months exposure to the release. The category “other” contains all genera with relative abundance less than 1%. Both samples were collected during the summer season.

Compared to the bacterial community, the archaeal community was much less diverse.

In this aquifer system, only four archaeal genera (i.e., Methanobacterium, Methanosarcina,

Thermogymnomonas, and Halalkalicoccus) were detected and methanogens were the

predominant groups (Figure 8.5).

153

Figure 8.5 Seasonal changes in archaeal community (genus level) following 10 months of exposure to the continuous ethanol-blend release. The release started at t=10 days.

8.3.4 Effect of seasonal fluctuations in groundwater temperature

We previously reported that the biodegradation of ethanol, benzene and toluene proceed faster in this system during the summer than in the winter (Ma et al., 2011). Here, we show a corroborating impact of temperature on functional gene abundance and groundwater geochemical footprints (i.e., pH, redox potential and dissolved oxygen). In winter, low temperature inhibited microbial fermenting activities. Correspondingly, the abundance of methangenesis (mcrA) and acetogenesis (fhs) functional genes decreased to the lowest level

(Figure 8.6). Significant correlations (p < 0.05) between CH4 concentrations and mcrA gene abundance (Figure 8.7) and between acetate concentrations and fhs gene abundance (Figure

8.8) were observed. Slower microbial activities in the winter also resulted in less oxygen consumption and less organic acid production, which led to the rebound of dissolved oxygen

154

concentration, redox potential and pH (Figure 8.2). Higher dissolved oxygen solubility at lower temperatures also likely contributed to the rebound of dissolved oxygen concentrations.

mcrA gene (a) 7 10 CH4 25 32 Temperature

6 20

10 24 ) C o 15 105 16 10

4 concentration (mg/L)

4 8 10 Temperature (

Gene copiesGene / g sand 5 CH

103 0 0 0 50 150 200 250 300 350 Days

(b) fhs gene 107 Acetate 300 32 Temperature 250 106 24 ) C

200 o

105 150 16

100 4

10 8 Temperature ( Gene copiesGene /g sand 50 Acetate concentration (mg/L) 103 0 0 0 50 150 200 250 300 350 Days Figure 8.6 Seasonal variations of the abundance of functional genes mcrA (a) and fhs (b) with groundwater temperature fluctuations, and the corresponding CH4 and acetate groundwater concentrations. The release started at t=10 days on August 17, 2009.

155

10 (mg/L) 4 2 CH R =0.782 5 p=0.0003

0 1x106 2x106 3x106 4x106 5x106 mcrA (copy number / g sand)

Figure 8.7 Correlation between methane concentration in the groundwater and mcrA gene abundance.

250

200

150

100 R2=0.446 p=0.025 Acetate (mg/L) 50

0 1x106 2x106 3x106 4x106 5x106 fhs (copy number / g sand)

Figure 8.8 Correlation between acetate concentration in the groundwater and fhs gene abundance.

8.3.5 Overall effects on biodegradation of the continuous ethanol-blend release

Whether microbial adaptation and associated changes in microbial community structure resulted in increased rate of attenuation was determined by linear regression of normalized concentration versus time data (Figure 8.9, Figure 8.10 and Figure 8.11). This normalization approach factors out dilution effects and discerns removal processes (mainly biodegradation)

156

(Alvarez and Illman, 2005). Despite temperature fluctuations that are known to affect biodegradation rates, a statistically significant trend (p < 0.05) of decreasing normalized concentrations was observed for ethanol, benzene and toluene (Figure 8.12). Since influent concentrations remained relatively constant, this suggests that biodegradation rates increased over the 10-month experiment. This observation is consistent with the adaptive shift in microbial community structure (indicated by higher microbial diversity and greater abundance of related functional genes), which are conducive to enhanced bioremediation of an ethanol-blended fuel.

1.0 R2=0.185 0.8 p=0.00662 Br ) 0 0.6 C/C ( / 0.4 ethanol ) 0 0.2 C/C ( 0.0 0 70 140 210 280 350 Days following the release

Figure 8.9 Linear regression of normalized ethanol concentration versus time. Initial concentrations (before t = 5 days) were not considered since the (constant source) plume was just beginning to break through the monitoring well at that time.

157

0.6 R2=0.185

Br )

0 p=0.0332 0.4 C/C ( /

Benzene 0.2 ) 0 C/C ( 0.0

0 70 140 210 280 350 Days following the release

Figure 8.10 Linear regression of normalized benzene concentration versus time. Initial concentrations (before t = 22 days) were not considered since the (constant source) plume was just beginning to break through the monitoring well at that time.

0.15 R2=0.185309

Br p=0.00639

) 0.10 0 C/C ( / 0.05 Toluene ) 0 0.00 C/C (

-0.05 0 70 140 210 280 350 Days following the release

Figure 8.11 Linear regression of normalized toluene concentration versus time. Initial concentrations (before t = 22 days) were not considered since the (constant source) plume was just beginning to break through the monitoring well at that time.

158

Ethanol 1 Benzene Toluene Br ) 0 0.1 C/C ( / i ) 0

C/C 0.01 (

0.001 Aug Oct Dec Feb Apr Jun Days

Figure 8.12 Bio-attenuation of ethanol, benzene and toluene during the continuous release of ethanol-blended solution.

8.4 Conclusion

The enrichment of acetogens and methanogens following an ethanol-blend release reflects an adaptive response for anaerobic bioremediation. By influencing functional gene abundance and microbial community structure, groundwater temperature can significantly affect biodegradation activities and their geochemical footprints (e.g., pH, redox potential, and dissolved oxygen). Overall, this study demonstrates adaptive microbial population shifts that enhance biodegradation of an ethanol-blended fuel, and underscores the importance to consider seasonal changes in groundwater temperature for assessing and predicting the performance of natural attenuation.

159

Chapter 9

Pyrosequencing-based investigation of the microbial response to a 2-year continuous release of ethanol blended fuel and its subsequent shut-off

[Extract from a manuscript in preparation]

9.1 Introduction

Since microorganisms are crucial to the functioning of all known ecosystems, there is considerable interest in understanding the response of microbial communities to anthropogenic disturbances or environmental changes, including how these changes affect the diversity, resistance and resilience of indigenous microorganisms (Allison and Martiny, 2008;

Baho et al., 2012; Shade et al., 2012; Griffiths and Philippot, 2013). The term resistance is used by microbial ecologists to denote the tendency of a microbial community structure to remain unchanged after being disturbed (Shade et al., 2012). Resilience refers to the tendency of a microbial community to return to the pre-disturbance structure after the disturbance disappears (Pimm, 1984).

160

Fuel releases that impact groundwater represent a common disturbance to subsurface

ecosystems, and the growing use of ethanol as transportation biofuel is increasing the

likelihood of encountering ethanol in such releases (Ma et al., 2013a). Several studies have

shown that ethanol blend releases significantly change the structure of indigenous microbial

communities (i.e., microorganisms are not "resistant" to such release) (Capiro et al., 2008;

Feris et al., 2008; Nelson et al., 2010; Ma et al., 2013b). However, no study has investigated

whether the altered microbial community is able to revert to the pre-release state after the contaminant is removed (i.e., whether microbial community is "resilient" to such release).

Furthermore, little is known about the evolution of the microbial community structure throughout the life cycle of a release, from initial plume expansion through its stabilization and eventual disappearance.

This study used 16S rRNA pyrosequencing to characterize both bacterial and archaeal responses to a 2-year continuous release of an ethanol blend solution, and its subsequent shut- off for another 2 years. Chemical concentrations (e.g., ethanol, benzene, toluene, and degradation byproducts methane, acetate, butyrate, and butanol) and several environmental master variables (e.g., temperature, pH, redox potential and dissolved oxygen) were also monitored to enhance our understanding of the relation between the geochemical footprint of biofuel spills and microbial community stability (resistance and resilience) during natural attenuation of the release.

161

9.2 Materials and Methods

9.2.1 Pilot-scale tank and different experimental stages

The plan view and profile view of the sampling port for this study are shown in Figure

9.1. Details on tank construction and the release scenario can be found in Chapter 3.

Figure 9.1 Plan view (a) and profile view (b) of the aquifer tank.

This pilot-scale release experiment lasted for 4 years, which could be divided into 4 different experimental stages. Figure 9.2 shows the timeline for this experiment. General information for each stage can be found in Table 9.1. Stage 1 is the pre-release baseline.

Stage 2 begins with the ethanol blend release (10% ethanol+50 mg/L benzene+50 mg/L toluene) and lasts for 2 years. Stage 3 follows the removal of ethanol from the continuous

162

release, and continuous exposure to 50 mg/L benzene and 50 mg/L toluene continues for 8

months. Stage 4 is the return to initial conditions (benzene and toluene removed from the tank

influent) and the aquifer material was exposed to clean water for one year. Two samples were taken during this recovery stage: at 4 months (4a) and 12 months (4b) after exposure to clean water. Detailed information for each stage can be found in Chapter 3.

Figure 9.2 Five stages of the pilot-scale release experiment.

163 Table 9.1 General information for each experimental stage

Stage Stage 1 Stage 2 Stage 3 Stage 4a Stage 4b Sampling date 8/7/2009 9/5/2011 5/4/2012 9/2/2012 5/23/2013 Exposed to ethanol, Exposed to benzene and Exposed to clean water Exposed to clean water for Action Pre-release baseline benzene and toluene toluene mixture for 8 for 8 months after Stage 4 months after Stage 3 mixture for 2 years months 4 Chemical concentration in the groundwater Ethanol (mg/L) 0.0 3269.5 0.0 0.0 0.0 Benzene (µg/L) 0.0 83.1 1121.0 0.0 0.0 Toluene (µg/L) 0.0 38.3 6.4 0.0 0.0 Methane (mg/L) 0.0 28.5 0.0 0.0 0.0 Acetate (mg/L) 0.0 116.0 0.0 0.0 0.0 Butyrate (mg/L) 0.0 267.0 0.0 0.0 0.0 Butanol (mg/L) 0.0 72.0 0.0 0.0 0.0 General environmental conditions pH (1) 7.2 4.7 7.8 7.6 7.2 Dissolved oxygen (mg/L) (2) 5.1± 0.4 0.0 ± 0.0 1.6 ± 0.2 2.1 ± 0.3 3.2 ± 0.3 Temperature (° C) (3) 29.1 28.5 26.3 27.9 27.1 Redox potential (mV) (4) 108.3± 6.1 -402.3 ± 7.6 NA NA NA

Prevailing electron Aerobic Methanogenic Aerobic Aerobic Aerobic accepting conditions (1) Groundwater pH was measured above ground using a pH meter (Davis Instruments, Vernon Hills, IL, US) after the in situ probe malfunctioned during Stage 2. (2) DO concentrations during Stage 1 and 2 were measured by a groundwater geochemical probe and DO concentrations during Stage 3, 4a and 4b were measured by the Dissolved Oxygen AccuVac® Ampules kit using 4 replicates for each date. The standard error for DO measured by the probe was calculated using consecutive values measured every 12 hours around each date. (3) Groundwater temperature was measured by a Pen-Style Thermometer (Taylor Precision Products, Oak Brook, IL, US). (4) The redox potential was measured by the groundwater geochemical probe. Since the probe did not work since the winter of 2011, no redox data was available for Stage 3, 4a, and 4b. The standard error for redox potential measurements was calculated by consecutive values measured obtained 12 hours around each date.

164

9.2.2 Groundwater chemical and geochemical analysis

At the end of each of the experimental stages (Figure 9.2), groundwater samples were

collected from the groundwater sampling port (Figure 9.1 a) for chemical analysis. The methods

used for ethanol, methane, acetate, propionate, butyrate, butanol, benzene, toluene, and sodium

bromide tracer analyses can be found in Chapter 3.

9.2.3 Sand sampling and analysis

Sand samples were collected from a depth of 5-30 cm below water table (50-75 cm below

the sand surface, Figure 9.1) on the same day with the groundwater sampling. The sand sampling

method can be found in Chapter 3. Dry sand samples were sent to the Soil, Water and Forage

Testing Laboratory at Texas A&M University to measure various sand chemical properties,

including soil pH, total organic carbon content, conductivity, nitrate-nitrogen, P, K, Ca, Mg, S,

Na, Fe, Zn, Mn, Cu. Soil analysis methods are listed on the website of Soil, Water and Forage

Testing Laboratory: (http://soiltesting.tamu.edu/webpages/swftlmethods1209.html). The sand

chemical characteristics data can be found in Table A2.1 in Appendix II.

9.2.4 DNA extraction and pyrosequencing

DNA was extracted from 10 g sand using a modified PowerMax Soil DNA isolation kit

(MOBIO, Carlsbad, CA, US). The modifications include: 1) after step 2 of the MOBIO protocol,

1 ml of 100g/L lysozyme (EMD Biosciences, San Diego, CA, US) was added and the mixture

was incubated at 37 °C for 30 min; 2) after step 3 of MOBIO protocol, the mixture was

incubated at 65 °C for 10 min before bead beating; 3) after step 4, 0.5 mL 20 g/L proteinase K

(QIAGEN, Valencia, CA, US) was added and the mixture was incubated at 56 °C for 16 h; 4) 2

µL of 5,000 mg/L RNase was added after the proteinase K treatment to remove all RNA; 5) step

165

7-10 was modified as: add 2 mL of Solution C2 and C3 together; 6) at step 18, 3 mL (instead of

5 mL) of Solution C6 was added. DNA was extracted from two replicates (10 g) of each sand sample and the two DNA extracts were pooled together, cleaned up and concentrated to 80 μL using Zymo Genomic DNA Clean & Concentrator™ kit (Zymo, Irvine, CA, US).

For pyrosequencing library preparation, the bacterial 16S rRNA gene was amplified using bacterial primers 347F (5’-GGAGGCAGCAGTRRGGAAT-3’) and 803R (5’-

CTACCRGGGTATCTAATCC-3’); the archaeal 16S rRNA gene was amplified using primers

A571F (5’-GCYTAAAGSRICCGTAGC-3’) and UA1204R (5’-TTMGGGGCATRCIKACCT-

3’). PCR reactions (50μL) were performed using FastStart High Fidelity PCR System (Roche

Applied Science, Indianapolis, IN, US) containing 0.4 μM each primer, 1U Reaction Buffer with

18 mM MgCl2, 1U of each dNTP, 1U of FastStart high-fidelity PCR system enzyme blend, and 5

μL template DNA. The PCR conditions were as follows: 95°C for 3 min; 30 cycles of denaturation (95°C; 45 s), annealing (55°C; 45 s) and extension (72°C; 1 min); followed by a final extension (72°C; 4 min). PCR products were separated in a 1% (wt/vol) gel. DNA of the correct size (527 bp for bacterial library and 704 bp for archaeal library) was recovered using a

QIAquick gel extraction kit (Qiagen, Valencia, CA, US). Recovered PCR products were quantified using PicoGreen (Invitrogen, Eugene, OR, US) and pooled in normalized quantities for direct sequencing by a Genome Sequencer Junior System (454 Life Sciences, Branford, CT,

US).

9.2.5 Sequence data processing

Pyrosequencing data were processed using the Quantitative Insights Into Microbial

Ecology (QIIME) pipeline (Caporaso et al., 2010b). Raw sequences were sorted by barcode, and

166

fusion primers were removed. All reads had to satisfy the following criteria to pass the quality filter: 1) a minimum length of 200 bp; 2) no ambiguous bases; 3) no homopolymer of 6 bp and above; 4) no error in forward or reverse primers; 5) a minimum quality score of 25. After initial filtering, sequences were denoised by QIIME Denoiser (Reeder and Knight, 2010). The denoised sequences were further filtered by UCHIME to remove chimeric sequences and singletons

(Edgar et al., 2011). Similar sequences were clustered into operational taxonomic units (OTUs) based on a cutoff similarity of 0.03 using USEARCH (Edgar, 2010). The most abundance sequence in the cluster of each OTU was selected as a representative sequence for that OTU.

Representative sequences were aligned using PYNAST with the Greengenes database with a minimum alignment length of 150 bp and a minimum sequence identity of 70% (Caporaso et al.,

2010a). The taxonomic identity of each representative OTU was assigned using RDP classifier with a minimum confidence level of 0.80 (Wang et al., 2007). The aligned sequences were filtered by removing columns comprised of only gaps and positions known to be excessively hyper-variable using a lane mask file (the Greengenes core set in the QIIME workflow)

(DeSantis et al., 2006). The filtered alignment file was then used to build a phylogenetic tree using the FastTree method (Price et al., 2010). The resulting phylogenetic tree was used for downstream phylogenetic-based diversity analysis. The sequence number for each sample can be found in Table A2.1 in Appendix II.

9.2.6 Statistical analysis

Since sample heterogeneity would bias the comparison of samples with different sequence numbers, all bacterial samples were rarefied to 12,000 sequences/sample and all archaeal samples were rarefied to 8,000 sequences/sample. Alpha diversity measures the diversity within a particular area or ecosystem, and beta diversity is a comparison of diversity

167

between ecosystems (Whittaker, 1972). A variety of alpha diversity indices, that describe species richness (Chao1, Observed Species), evenness (Equitability), and diversity (Shannon, Simpson and Phylogenetic Diversity), were calculated for both bacterial and archaeal libraries. Principal

Coordinate Analysis (PCoA) was conducted to assess beta diversity (dissimilarity) of microbial communities among different stages. Weighted Unifrac distances were also used to create a

PCoA plot.

9.3 Results and Discussion

9.3.1 Dissimilarities in microbial community structure among different stages

After the ethanol blend release was shut off, both bacterial and archaeal communities were shifting back towards the pre-release conditions, but this restoration process was relatively slow compared to the population shifts in response to the release. The dissimilarity (beta diversity) of both bacterial and archaeal communities from 4 different stages were illustrated through a Principal Coordinate Analysis (PCoA), which was conducted based on a weighted

Unifrac distance matrix (Figure 9.3 and 9.4). The general information for each experimental stage are summarize in Table 9.1. For both bacterial (Figure 9.3) and archaeal communities

(Figure 9.4), the distance between Stage 4b and Stage 1 was smaller than that between Stage 2 and Stage 1, suggesting that both bacterial and archaeal communities may have been reverting towards the pre-release states after the ethanol blend release was shut off. If so, the bacterial community structure at Stage 4b was still far from that at Stage 1 (Figure 9.3), indicating that the reversal process was relatively slow compared to the population shifts in response to the release.

Alternatively, the microbial populations may have been shifting towards another meta-stable community structure where many populations that prevailed before the release were regaining

168

prominence (e.g., Thermoprotei, Rhodocyclaceae, Rhizobiales, Rhodospirillales, Nitrospira,

GP3, GP6, Rhodospirillaceae). Detailed taxonomic information will be discussed in the following text.

0.2 Stage 4a

0.1 Stage 4b 0.0 Stage 3 -0.1 Stage 1 -0.2

PCo2 (32.7% Variation)PCo2 -0.3 Stage 2 -0.4 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 PCo1 (38.7% variation)

Figure 9.3 Two-dimensional Principal Coordinate Analysis (PCoA) plot based on weighted UniFrac distance matrix for both bacterial communities. The average weighted UniFrac distance matrix was calculated from 12,000 sequence even rarefactions for each sample. The generation information for each stage is summarized in Table 9.1. The PCo1 axis explains 38.7% of the total variation of the bacterial community structure and the PCo2 axis explains 32.7% of the total variation.

169

0.10 Stage 4b Stage 4a 0.05

0.00 Stage 1 -0.05 Stage 3 -0.10 PCo2 (5.9% variation)PCo2 Stage 2 -0.15 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2

PCo1 (89.5% variation)

Figure 9.4 Two-dimensional Principal Coordinate Analysis (PCoA) plot based on weighted UniFrac distance matrix for both archaeal communities. The average weighted UniFrac distance matrix was calculated from 8,000 sequence even rarefactions for each sample. The PCo1 axis explains 89.5% of the total variation of the archaeal community structure and the PCo2 axis explains 5.9% of the total variation.

9.3.2 Taxonomic composition of archaeal communities

The taxonomic composition of the archaeal community was significantly changed by the ethanol blend release and was shifting back towards the pre-release states after such release was shut off. At Stage 1, the archaeal community was dominated by the class Thermoprotei and unclassified Archaea (Figure 9.5 and Table A2.3 in Appendix II). Thermoprotei has been reported to be abundant in desert sands, and most of the species in this class are chemolithotrophs (Andrew et al., 2012). Before the release, the sand had low organic carbon content (0.063% ± 0.011%, Table A2.1 in Appendix II) the influent tap water lacked organic carbon and other nutrients, which likely facilitated the establishment of chemolithrophs. At Stage

2, the dissolved oxygen (DO) had been depleted by biodegradation processes (Table 9.1), strong anaerobic conditions had developed (Redox potential was -402.3 ± 7.6 mV), and two

170

methanogenic genera Methanobacterium and Methanosarcina proliferated (Figure 9.5). This is

consistent with previous findings (Chapter 8), showing that ethanol blend release enriched

methanogens and significantly changed the archaeal community.

100

80 Unclassified Methanosarcinales (order) Unclassified Methanobacteriaceae (family) Unclassified Thermoplasmatales (order) 60 Unclassified Methanosarcinaceae (family) Unclassified Euryarchaeota (Phylum) 40 Methanocella (genus)

Percentage (%) Unclassified Thermoprotei (Class) Unclassified Archaea (Domain) 20 Methanobacterium (genus) Methanosarcina (genus) 0 Stage1.rep1 Stage1.rep2 Stage1.rep3 Stage2.rep1 Stage2.rep2 Stage2.rep3 Stage3.rep1 Stage3.rep2 Stage3.rep3 Stage4a.rep1 Stage4a.rep2 Stage4a.rep3 Stage4b.rep1 Stage4b.rep2 Stage4b.rep3 Figure 9.5 Taxonomic compositions of archaeal communities. Only the 10 most abundant taxa are included.

At Stage 4a, dissolved O2 concentrations rebounded to 2.1 ± 0.3 mg/L (Table 9.1), but

methanogens still persisted in the tank aquifer. Methanogens comprised > 85% of archaeal 16S

rRNA sequences this stage (Figure 9.6). This result agrees with our previous qPCR analysis

(Figure 8.6 in Chapter 8) for the methanogenic functional gene (mcrA), corroborating the

persistence of methanogens following the shut-off of the ethanol blend release. However, the

structure of the methanogenic community at Stage 4a was different from that at Stage 2. Stage 4a

had a higher proportion of Methanosarcina and Methanocella and a lower proportion of

Methanobacterium than Stage 2 (Figure 9.6). Although methanogenesis is generally thought to

occur only in highly reducing anoxic environments, two recent studies found that viable

Methanosarcina and Methanocella are present in many types of aerated upland soils sampled

globally (Angel et al. 2012) and even in desert sand (Angel et al., 2011; Angel et al.,

171

2012).These two groups of methanogens are known for their ability to withstand O2 stress;

apparently they possess several (6 to 7) genes encoding enzymes that detoxify reactive oxygen

species (Erkel et al., 2006; Angel et al., 2011). Because of spatial heterogeneity and the existence

of anaerobic micro-niches in generally aerobic soils (Renault and Sierra, 1994; Renault and

Stengel, 1994; Eyre and Ferguson, 2009), anaerobic microorganisms could survive in soil

matrices. The capability of Methanosarcina and Methanocella to survive in aerobic

environments facilitates the enrichment of these two methanogenic taxa at Stage 4a, in the

presence of 2.1 ± 0.3 mg/L DO (Table 9.1). Shifts from O2-sensitive methanogen (e.g.,

Methanobacterium) to the methanogens that are not so sensitive to O2 (e.g., Methanosarcina and

Methanocella) may contribute to the persistence of methanogens in the oxic samples (e.g., at

Stage 4a).

unclassifiedThermoprotei (class) unclassified Archaea 100 Methanobacterium 8 Methanosarcina Methanocella 80 6

60 4 and Methanosarcina (%) unclassified Archaea , 40 Methanocella (%)

2 20 Thermoprotei, 0 0 Methanobacterium Stage 1 Stage 2 Stage 3 Stage 4a Stage 4b

Figure 9.6 Changes in the relative abundance of five dominant archaeal taxa: Thermoprotei (class), unclassified Archaea, Methanobacterium (genus), Methanosarcina (genus) and Methanocella (genus).

172

At Stage 4b, DO concentrations increased to 3.2 ± 0.3 mg/L and no contaminant or its

degradation byproducts were detected. Some dominant taxa during Stage 1 (e.g.,

chemolithotrophic Thermoprotei and unclassified Archaea) rebounded (Figure 9.6), but the

relative abundance of methanogens (dominant taxa during Stage 2) decreased at Stage 4b (Figure

9.6). Such changes in the archaeal taxonomic composition corroborate the PCoA analysis (Figure

9.4), indicating that the archaeal community was returning towards the pre-release structure. The

persistence of methanogens at Stage 4b also suggested that this reversal was far from complete,

even two years after the removal of the ethanol source.

9.3.3 Taxonomic composition of bacterial communities

The taxonomic composition of the bacterial community was significantly changed by the

ethanol blend release, and was shifting back towards the pre-release state after source removal.

At (pre-release) Stage 1, the phylum Acidobacteria and Proteobacteria dominated the bacterial

community. Dominant Acidobacteria groups included subdivision Gp6 (12.9%), Gp3 (5.1%),

Gp16 (4.2%), Gp2 (1.7%), and Gp12 (1.2%) (Figure 9.7 and Table A2.4 in Appendix II).

Dominant Proteobacteria groups included Rhizobiales (order, 7.0%), Rhodospirillales (order,

2.6%), Rhodospirillaceae (family, 2.1%), Sphingomonas (genus, 1.8%), Rhodocyclaceae (family,

1.8%), and Hyphomicrobiaceae (family, 1.5%) (Figure 9.7). The phylum Acidobacteria is

abundant in many soil and sediment environments, but most of its members are very difficult to

isolate and culture; therefore, little is known about their physiology and ecological functions

(Liles et al., 2010). As a group of strictly aerobic heterotrophs, (Krieg, 2010), Sphingomonas are

able to survive in low concentrations of nutrients (Stolz, 2009). The family Rhodocyclaceae contains mainly aerobic or denitrifying bacteria and has been found in oligotrophic environments

(Fahrbach et al., 2006; Bae et al., 2007; Chou et al., 2008). Nitrospira, a group of nitrite-

173

oxidizing chemolithoautotrophs (Luecker et al., 2010), was the only dominant taxon (2.6%) of

Stage 1 that did not belong to Acidobacteria or Proteobacteria (Figure 9.7).

GP7 (genus) GP1 (genus) 80 GP12 (genus) Unclassified Proteobacteria (pylum) Unclassified Hyphomicrobiaceae (family) Gp2 (genus) Unclassified Actinobacteria (class) 60 Unclassified Rhodocyclaceae (family) Sphingomonas (genus) Unclassified Deltaproteobacteria (class) Unclassified Rhodospirillaceae (family) 40 Nitrospira (genus) Unclassified Rhodospirillales (order) Percentage (%) Unclassified Betaproteobacteria (class) Gp16 (genus) Gp3 (genus) 20 Unclassified Gammaproteobacteria (class) Unclassified Alphaproteobacteria (class) Unclassified Rhizobiales (order) Gp6 (genus) 0 Unclassified Bacteria Replicate1 Replicate2 Replicate3

Stage 1 Figure 9.7 Taxonomic composition of the bacterial community at Stage1. Only the dominant taxa with the relative abundance higher than 1% are included.

The release drove the system to anaerobic conditions and enriched a variety of anaerobic heterotrophs (Stage 2). These include Veillonellaceae (family 13.8%), Clostridium (genus,

12.4%), Clostridiales (order, 9.2%), Desulfovibrio (genus, 5.6%), Syntrophomonas (genus,

4.8%), Desulfuromonas (genus, 3.9%), Geobacter (genus, 2.3%), Methylococcaceae (family,

1.9%), Desulfuromonadales (order, 1.7%), Longilinea (genus, 1.4%), Clostridiaceae (family,

1.3%), Coriobacteriaceae (family, 1.2%) and Methylosarcina (genus, 1.2%) (Figure 9.8 and

Table A2.4 in Appendix II). Clostridium is a group of obligate anaerobes which includes many species of fermenters that could produce acetate, butyrate and butanol (Krieg, 2010). Some members of Clostridium degrade aromatic hydrocarbons (Mechichi et al., 1999; Winderl et al.,

2010). Desulfovibrio is a group of anaerobic heterotrophs that are able to degrade aromatic

174

hydrocarbons (Allen et al., 2008). Syntrophomonas is a group of anaerobic syntrophic volatile fatty acids degraders (Harms et al., 2003; Luecker et al., 2010). Desulfuromonas is a group of anaerobes that could degrade volatile fatty acids (Coates et al., 1995; Vandieken et al., 2006).

Longilinea is a group of anaerobic fermenters that produce volatile fatty acids (Yamada et al.,

2007). Veillonellaceae (Marchandin et al., 2010), Clostridiales (Paredes-Sabja et al., 2011;

Smith et al., 2012), Clostridiaceae (Krieg, 2010), Geobacter (Lovley et al., 2011),

Desulfuromonadales (Vandieken and Thamdrup, 2013), and Coriobacteriaceae (Harmsen et al.,

2000) are also obligate anaerobes.

80 Unclassified Rhodocyclaceae (family) Unclassified Coriobacteriaceae (family) Unclassified Actinobacteria (class) Methylosarcina (genus) Uncalssified Clostridiaceae 1 (family) 60 Gp16 (genus) Longilinea (genus) Unclassifed Rhizobiales (order) Unclassifed Desulfuromonadales (order) Unclassified Firmicutes (phylum) 40 Unclassified Methylococcaceae (family) Unclassified Gammaproteobacteria (class) Geobacter (genus) Percentage (%) Unclassified Betaproteobacteria (class) Unclassified Bacteria 20 Desulfuromonas (genus) Syntrophomonas (genus) Desulfovibrio (genus) Unclassified Clostridiales (order) Clostridium (genus) 0 Unclassified Veillonellaceae (family) Replicate1 Replicate2 Replicate3

Stage 2 Figure 9.8 Taxonomic composition of the bacterial community at Stage2.Only the dominant taxa with the relative abundance higher than 1% are included.

At Stage 3 which was hypoxic with dissolved O2 concentration of 1.6 ± 0.2 mg/L, anaerobes still dominated the bacterial community, but the relative abundance of aerobes increased. The enriched aerobes included Methylococcus (genus, 2.7%) Methylococcaceae

(family, 2.2%), Thiobacillus (genus, 1.6%), Methylocystis (genus, 1.2%) (Figure 9.9 and Table

175

A2.4 in Appendix II). Thiobacillus (genus, 1.6%) is a group of aerobic autotrophs and some

facultative members of this genus could also grow anaerobically (Allen et al., 2006).

Methylococcus (genus, 2.7%) Methylococcaceae (family, 2.2%), Methylocystis (genus, 1.2%) are

aerobic methanotrophs, possibly feeding on the generated methane. The dominant anaerobes

included Anaerolineaceae (family, 4.9%), Desulfobulbaceae (family, 3.0%), Syntrophomonas

(genus, 2.3%), Longilinea (genus, 2.2%), Desulfuromonas (genus, 2.2%), Ignavibacterium

(genus, 1.6%), Clostridium (genus, 1.6%), Geobacter (genus, 1.3%) (Figure 9.9 and Table A2.4

in Appendix II). The family Anaerolineaceae consists mainly of anaerobic fermenters (Yamada

et al., 2006; Yamada et al., 2007). As a novel genus, Ignavibacterium currently contains only one

species, Ignavibacterium album, which is a anaerobic heterotroph (Iino et al., 2010). Several dominant groups of Stage 1, including Rhizobiales, Rhodocyclaceae, Gp3, were also enriched at

Stage 3 (Figure 9.9).

80 Methylocystis (genus) Geobacter (genus) Unclassified Deltaproteobacteria (class) Thiobacillus (genus) Clostridium (genus) 60 Unclassified Actinobacteria (class) Ignavibacterium (genus) Desulfuromonas (genus) Unclassified Methylococcaceae (family) Longilinea (genus) 40 Syntrophomonas (genus) Unclassified Gammaproteobacteria (class) Gp3 (genus) Percentage (%) Methylococcus (genus) Unclassified Desulfobulbaceae (family) 20 Unclassified Rhodocyclaceae (family) Unclassified Rhizobiales (order) Unclassified Anaerolineaceae (family) Unclassified Bacteroidetes (phylum) Unclassified Bacteria 0 Unclassified Betaproteobacteria (class) Replicate1 Replicate2 Replicate3

Stage 3 Figure 9.9 Taxonomic composition of the bacterial community at Stage3. Only the dominant taxa with the relative abundance higher than 1% are included.

176

A variety of dominant taxa during Stage 1 rebounded at Stage 4b, which suggests that the

bacterial community may have been reverting towards the pre-release structure after the removal of the release. The recovering taxa included Rhodocyclaceae (family, 3.7%), Rhizobiales (order,

3.1%), Rhodospirillales (order, 2.2%), Nitrospira (genus, 2.2%), GP3 (genus, 2.1), GP6 (genus,

1.7%), Rhodospirillaceae (genus, 1.1%) (Figure 9.10 and Table A2.4 in Appendix II). Figure

9.11 shows the temporal changes in the relative abundance of 10 dominant families from Stage 1.

These 10 families comprised 33.6% of bacterial 16S rRNA sequences of Stage 1. The ethanol blend release significantly changed the bacterial community, and the total abundance of these 10 families decreased to the lowest level (3.2%, Figure 9.10). After the release was shut off, the abundance of these 10 families steadily increased to 9.5% at Stage 4b (Figure 9.11). However,

9.5% at Stage 4b was still much lower than the 33.6% at Stage 1, thus suggesting that the recovery was far from complete. It is possible that the microbial community would shift to another meta-stable structure instead of completely reverting to the pre-release conditions after the contaminant source is removed, but longer recovery monitoring (perhaps a few years) would be needed to determine so.

177

80 Unclassified Rhodospirillaceae (family) Unclassified Bacteroidetes (phylum) Unclassified Desulfobulbaceae (family) Gp6 (genus) 60 Gp3 (genus) Unclassified Proteobacteria (phylum) Nitrospira (genus) Unclassified Rhodospirillales (order) Unclassified Alphaproteobacteria (class) 40 Clostridium (genus) Percentage (%) Unclassified Anaerolineaceae (family) Unclassified Rhizobiales (order) Unclassified Actinobacteria (class) 20 Unclassified Rhodocyclaceae (family) Unclassified Deltaproteobacteria (class) Unclassified Gammaproteobacteria (class) Unclassified Bacteria (genus) Unclassified Betaproteobacteria (class) 0 Replicate1 Replicate2 Replicate3

Stage 4b Figure 9.10 Taxonomic composition of the bacterial community at Stage 4b. Only the dominant taxa with the relative abundance higher than 1% are included.

Gp7 35 Gp1 30 Gp12 Hyphomicrobiaceae 25 Gp2 Rhodospirillaceae 20 Nitrospiraceae Gp16 15 Gp3 10 Gp6

Relative abundance (%) 5 0 Stage 1 Stage 2 Stage 3 Stage 4a Stage 4b

Figure 9.11 Changes in the relative abundance of ten families that were dominant at Stage 1.

178

9.3.4 Species richness, evenness, and diversity

Microbial diversity can be characterized by two parameters: richness (How many species are there?) and evenness (How equal are the abundances of the species?) The microbial community with higher richness and higher evenness is more diverse. Biodiversity could be described by a variety of diversity indices. The diversity indies used in this study included

“Chao1” and “Observed species” for richness, “Equitability” for evenness, and “Shannon” and

“Simpson” for diversity.

Changes in archaeal evenness and diversity (indicated by “Equitability”, “Shannon”, and

“Simpson” indices) were not statistically significant (p> 0.05, Figure 12). However, ethanol blend releases decreased archaeal richness, which recovered to the pre-release level after the release was shut off. The archaeal community at Stage 2 had significantly lower “Chao1” and

“Observed species” indices (p< 0.05) than Stage 1, and these two richness indices rebounded at

Stage 4b (Figure 9.12). The rebound of richness indices provide another evidence for the restoration of the archaeal community after the ethanol blend release was shut off.

Changes in bacterial richness throughout all five stages (indicated by “Chao1” and

“Observed species”) were not statistically significant (p> 0.05, Figure 9.12). In contrast, bacterial evenness and diversity decreased following the ethanol blend releases (Stage 2) and recovered to the pre-release level after the release was shut off (Stage 4b). The release enriched certain groups that were involved in the degradation process, thus resulting in significant lower (p< 0.05) species evenness at Stage 2 (indicated by “Equitability”, Figure 9.12). Decreased evenness resulted in a decrease in diversity at Stage 2 (indicated by “Shannon” and “Simpson”). At Stage

4b, “Equitability”, “Shannon”, and “Simpson” indices rebounded (Figure 9.12), indicating that the evenness and diversity of the bacterial community reverted to the pre-release levels. These

179

results corroborate the PCoA analysis and taxonomic classification results, indicating that the bacterial community was reverting towards the pre-release state following the shut-off of the ethanol blend release.

180

Bacteria Archaea Bacteria Archaea 2700 600 1800 500 a A a A a a a a 2400 a 400 AB 400 1500 2100 300 a AB Bacteria

a Archaea Bacteria

a Archaea 1800 BC BC 200 1200 200 B C B 1500 B 100 0 900

Stage 1Stage 2Stage Stage3 4aStage 4b Stage 1Stage 2Stage Stage3 4aStage 4b (a) Chao1 (b) Observed species Bacteria Archaea Bacteria Archaea 0.9 0.6 9 a 5 A a a a a aAB ab A 0.8 0.5 8 AB 4 AB AB b bAB b B AB Bacteria Archaea Bacteria 0.7 0.4 7 B 3 Archaea

0.6 0.3 6 2

Stage 1Stage 2Stage Stage3 4aStage 4b Stage 1Stage 2Stage Stage3 4aStage 4b (c) Equitability (d) Shannon Bacteria Archaea 1.00 0.9 a A ab abcA A AB c 0.98 bc 0.8

B Bacteria 0.96 0.7 Archaea

0.94 0.6

Stage 1Stage 2Stage Stage3 4aStage 4b (e) Simpson

Figure 9.12 Alpha diversity indices of bacterial and archaeal communities. “Chao1” and “Observed species” are two species richness indices. “Equitability” is an evenness index. “Shannon” and “Simpson” are two non-phylogeny based diversity indices. Letters above each column indicate statistical differences between different stages (at a p value of < 0.05) by Fisher’s least-significant-difference (LSD) test. Statistical difference existed between two stages that do not have the same letter. Capital letters describe archaeal communities and small letters describe bacterial communities.

181

9.4 Conclusions

The ethanol blend fuel release significantly changed the microbial structure, by enriching certain microbial groups involved in the degradation of ethanol blend fuel (e.g., fermenters, acetogens, methanogens and hydrocarbon degraders). This release also significantly (p< 0.05) decreased the richness of the archaeal community and the evenness and diversity of the bacterial community. This suggests that the microbial community was not resistant to long-term (e.g., 2 years) continuous releases of ethanol blend fuel. After such release was shut off, the relative abundance of the anaerobic heterotrophs decreased and the relative abundance of taxa that were used to be dominant at Stage 1 increased, thus indicating that the microbial community was reverting towards the pre-release conditions (is resilient). The resilience of microbial community to the ethanol blend release is corroborated by the rebound of alpha diversity indices for both bacterial and archaeal communities. Note that such restoration process was relatively slow compared to the microbial community shifts in response to the start of the release. The structure of both bacterial and archaeal communities were still quite different from the pre-release sample even two years after the release was shut off. Therefore, future research with longer recovery monitoring is needed to investigate whether the indigenous microbial community could completely return to the pre-release conditions or reach another meta-stable structure after the contaminant source is removed.

182

Chapter 10

Conclusions, engineering significance and recommendations for future research

10. 1 Conclusions

This study had two primary objectives: 1) to assess the potential vapor intrusion risk (i.e., methane explosion hazard, and enhanced vapor intrusion potential of fuel hydrocarbons) and groundwater quality impacts associated with ethanol blend releases; and 2) to characterize the adaptation and response of the microbial community to a pilot-scale ethanol blend release and the shut-off of such release. The following points highlight the primary conclusion of this study.

 Methane is unlikely to build up to pose an explosion hazard (5% v:v) if diffusion is the only

mass transport pathway through the unsaturated zone.

 If methanogenic activity near the source zone is sufficiently high to cause advective gas

transport, the methane indoor concentration may exceed the flammable threshold (5% v:v).

183

 As a group of widely distributed microorganisms, methanotrophs can significantly attenuate

methane migration through the vadose zone, and thus alleviate the associated explosion risk.

 Methanotrophic activities could consume soil oxygen that would otherwise be available to

support hydrocarbon degradation, and increase the vapor intrusion potential for toxic fuel

hydrocarbon such as benzene.

 Benzene vapor intrusion would also be exacerbated if methanogenic activity results in

sufficiently high pressure to cause advective gas transport in the unsaturated zone.

 Ethanol biodegradation rapidly consumes oxygen, creating an anaerobic environment in the

impacted aquifer. Under the anaerobic conditions, ethanol can be fermented to volatile fatty

acids (VFAs). The accumulation of VFAs (particularly butyric acid) in the tank exceeded the

secondary maximum contaminant level value for odor, which represents a previously

unreported aesthetic impact. However, this aesthetic problem was relatively short-lived due

to the fast biodegradability of VFAs and their slower production in cooler seasons.

 Methane generation in the impacted aquifer may continue for a long time even after the

disappearance of dissolved ethanol. The persistent methane was likely generated from

ethanol degradation byproducts (e.g., acetate) and solid organic carbon in aquifer materials.

Ethanol blend releases stimulate the microbial growth and increased the organic carbon

content in the aquifer material.

 The anaerobic degradation of ethanol was temporarily stimulated when ethanol source was

removed and the dissolved ethanol concentration decreased below its toxicity threshold

(~2,000 mg/L for this system).

184

 Fuel ethanol release significantly changed the structure of indigenous microbial

communities and enriched the microbial groups involved in the biodegradation of ethanol

blends (e.g., fermenters, acetogenes, methanogens, and aromatic hydrocarbon degraders).

Such changes reflect adaptive changes in microbial communities that are conducive to

enhanced biodegradation of ethanol blend fuels.

 By influencing functional gene abundance and microbial community structure, seasonal

changes in groundwater temperature can significantly affect biodegradation activities and

their geochemical footprints (e.g., pH, redox potential, and dissolved oxygen).

 After the ethanol releases was shut off, the microbial community returned back towards the

pre-contaminated conditions, but the restoration process is relatively slow compared to the

population shifts in response to the release.

10. 2 Engineering significance

The following points highlight the engineering significance of this study.

 Recently, there have been increasing concerns over the methane vapor intrusion risk

associated with ethanol blend fuel releases (Jewell and Wilson 2011, Jourabchi et al. 2012,

Ma et al. 2012, Sihota et al. 2013, Wilson et al. 2012, 2013). This study shows that methane

generation in the groundwater may continue for a long time (~100 days) even after the

disappearance of ethanol and its degradation products (e.g., acetate, butyrate, butanol). Note

that the groundwater velocity (0.76 m/day) in this sandy aquifer is relative fast compared to

real aquifers, thus at real contaminated sites with slower groundwater flow, methane

persistence time could be longer. Based on these results, for site closure, long-term

185

monitoring for methane should be required even when the ethanol source is removed and

residual ethanol was attenuated.

 Besides methane, other byproducts of anaerobic ethanol degradation may also be

problematic (e.g., acetate may inhibit benzene anaerobic degradation (Corseuil et al. 2011);

butyrate may compromise groundwater aesthetic quality (Chapter 6); butanol is a regulated

compound in the drinking water standard of several states in U.S (Nelson et al. 2010)). This

study shows that after the ethanol source was removed, the fermentative degradation of

ethanol may be temporarily stimulated when the ethanol concentration decreases below its

toxicity threshold (e.g., ~2,000 mg/L for this system). The transient stimulation of ethanol

degradation would lead to accumulations of problematic ethanol degradation byproducts

(e.g., acetate, butyrate, and butanol). Therefore, monitoring for these compounds is

recommended particularly in a shrinking plume after the ethanol source is removed.

 This study indicates that advective soil gas transport generated from the accumulation of

fermentative biogas could play an important role in the subsurface vapor transport.

Therefore, conceptual models of vapor intrusion may need be modified to consider gas

advection for sites where strong fermentation activities exist.

 The release of ethanol-blended may result in lower aerobic attenuation of BTEX vapors

through the vadose zone and thus, higher potential for BTEX vapor intrusion into overlying

buildings. Therefore, monitoring fuel hydrocarbons in soil gas and the corresponding vapor

intrusion risk should be considered for sites impacted by ethanol blend fuel.

 This study shows the adaptive changes (enrichment of fermenters, acetogenes, methanogens,

and aromatic hydrocarbon degraders) in microbial communities that are conducive to

186

enhanced biodegradation activity, thus corroborating that monitored natural attenuation is an

effective remediation technology for ethanol blend fuel release. This study also shows that

seasonal changes in groundwater temperature can significantly affect biodegradation

activities. Therefore, changes in groundwater temperature should be considered for assessing

and predicting the performance of bioremediation and natural attenuation system.

10.3 Recommendations for future research

 Although the 3-D numerical model used in this study is a powerful tool that is based on a

sound theoretical framework regarding current understanding of vapor intrusion pathway, it

may or may not adequately simulate the complexities at real sites. Therefore, further field

studies are needed to have a complete understanding of the vapor intrusion hazard associated

with ethanol-blended fuel releases.

 The US. EPA is trying to establish vertical separation criteria for screening vapor intrusion

risk at petroleum release sites. This study the presence of high concentrations of CH4

originating from releases of high ethanol blends may deplete the available soil O2 and inhibit

benzene aerobic degradation, thus resulting in a higher benzene vapor intrusion potential

than suggested in the current EPA guidance for gasoline fuel release sites. Therefore, further

studies are needed to establish appropriate vertical separation criteria for sites impacted by

high ethanol blends, including E20 up to E95.

 Depending on the amount released, ethanol-blended fuels can greatly alter groundwater

geochemistry. The high BOD exerted by ethanol creates strongly anaerobic (reducing)

conditions (Mackay et al., 2006) under which VFAs accumulate and cause a decrease in pH

(Ma et al.). These conditions promote the dissolution of redox- and/or pH-sensitive metals

187

from the aquifer matrix (e.g., iron, manganese, and arsenic), thus exacerbating groundwater contamination (Deutsch, 1997; Brown et al., 2010). Although no studies of metal mobilization by ethanol-blend releases has been reported in the literature, elevated arsenic concentrations have been detected in groundwater contaminated by petroleum hydrocarbons

(Brown et al., 2010). Because of higher dissolved concentrations and faster anaerobic degradation, ethanol is more likely to induce reducing and acidic conditions that mobilize metals than petroleum hydrocarbons. Therefore further studies are needed to investigate whether ethanol-blended fuel could cause more drastic changes in groundwater geochemistry and higher risk for metal mobilization than regular fuel.

188

References

Abreu, L.D.V. (2005) A transient three-dimensional numerical model to simulate vapor intrusion into building. In. Tempe, AZ: Arizona State University. Abreu, L.D.V., and Johnson, P.C. (2005) Effect of vapor source - building separation and building construction on soil vapor intrusion as studied with a three-dimensional numerical model. Environmental Science & Technology 39: 4550-4561. Abreu, L.D.V., and Johnson, P.C. (2006) Simulating the effect of aerobic biodegradation on soil vapor intrusion into buildings: Influence of degradation rate, source concentration, and depth. Environmental Science & Technology 40: 2304-2315. Abreu, L.D.V., Ettinger, R., and McAlary, T. (2009) Simulated soil vapor intrusion attenuation factors including biodegradation for petroleum hydrocarbons. Ground Water Monitoring and Remediation 29: 105-117. Alford, D.P., Delaune, R.D., and Lindau, C.W. (1997) Methane flux from Mississippi River deltaic plain wetlands. Biogeochemistry 37: 227-236. Allen, T.D., Lawson, P.A., Collins, M.D., Falsen, E., and Tanner, R.S. (2006) Cloacibacterium normanense gen. nov., sp nov., a novel bacterium in the family Flavobacteriaceae isolated from municipal wastewater. International Journal of Systematic and Evolutionary Microbiology 56: 1311-1316. Allen, T.D., Kraus, P.F., Lawson, P.A., Drake, G.R., Balkwill, D.L., and Tanner, R.S. (2008) Desulfovibrio carbinoliphilus sp nov., a benzyl alcohol-oxidizing, sulfate-reducing bacterium isolated from a gas condensate-contaminated aquifer. International Journal of Systematic and Evolutionary Microbiology 58: 1313-1317. Allison, S.D., and Martiny, J.B.H. (2008) Resistance, resilience, and redundancy in microbial communities. Proceedings of the National Academy of Sciences of the United States of America 105: 11512-11519. Alvarez, P.J., and Illman, W.A. (2005) Bioremediation and Natural Attenuation: Process Fundamentals and Mathematical Models. Hoboken, New Jersey, USA: Wiley- Interscience. Amos, R.T., and Mayer, K.U. (2006) Investigating ebullition in a sand column using dissolved gas analysis and reactive transport modeling. Environmental Science & Technology 40: 5361-5367. Amos, R.T., Mayer, K.U., Bekins, B.A., Delin, G.N., and Williams, R.L. (2005) Use of dissolved and vapor-phase gases to investigate methanogenic degradation of petroleum hydrocarbon contamination in the subsurface. Water Resources Research 41. Andrew, D.R., Fitak, R.R., Munguia-Vega, A., Racolta, A., Martinson, V.G., and Dontsova, K. (2012) Abiotic Factors Shape Microbial Diversity in Sonoran Desert Soils. Applied and Environmental Microbiology 78: 7527-7537. Angel, R., Matthies, D., and Conrad, R. (2011) Activation of Methanogenesis in Arid Biological Soil Crusts Despite the Presence of Oxygen. Plos One 6. Angel, R., Claus, P., and Conrad, R. (2012) Methanogenic archaea are globally ubiquitous in aerated soils and become active under wet anoxic conditions. Isme Journal 6: 847-862. Bae, H.-S., Rash, B.A., Rainey, F.A., Nobre, M.F., Tiago, I., da Costa, M.S., and Moe, W.M. (2007) Description of Azospira restricta sp nov., a nitrogen-fixing bacterium isolated

189

from groundwater. International Journal of Systematic and Evolutionary Microbiology 57: 1521-1526. Baho, D.L., Peter, H., and Tranvik, L.J. (2012) Resistance and resilience of microbial communities - temporal and spatial insurance against perturbations. Environmental Microbiology 14: 2283-2292. Baird, A.J., Beckwith, C.W., Waldron, S., and Waddington, J.M. (2004) Ebullition of methane- containing gas bubbles from near-surface Sphagnum peat. Geophysical Research Letters 31. Baldwin, B.R., Nakatsu, C.H., and Nies, L. (2003) Detection and enumeration of aromatic oxygenase genes by multiplex and real-time PCR. Applied and Environmental Microbiology 69: 3350-3358. Baldwin, B.R., Nakatsu, C.H., and Nies, L. (2008) Enumeration of aromatic oxygenase genes to evaluate monitored natural attenuation at gasoline-contaminated sites. Water Research 42: 723-731. Baldwin, B.R., Nakatsu, C.H., Nebe, J., Wickham, G.S., Parks, C., and Nies, L. (2009) Enumeration of aromatic oxygenase genes to evaluate biodegradation during multi-phase extraction at a gasoline-contaminated site. Journal of Hazardous Materials 163: 524-530. Barker, H.A., Kamen, M.D., and Bornstein, B.T. (1945) The synthesis of butyric and caproic acids from ethanol and acetic acid by clostridium kluyveri. Proceedings of the National Academy of Sciences of the United States of America 31: 373-381. Beller, H.R., Kane, S.R., Legler, T.C., and Alvarez, P.J.J. (2002) A real-time polymerase chain reaction method for monitoring anaerobic, hydrocarbon-degrading bacteria based on a catabolic gene. Environmental Science & Technology 36: 3977-3984. Beller, H.R., Kane, S.R., Legler, T.C., McKelvie, J.R., Lollar, B.S., Pearson, F. et al. (2008) Comparative assessments of benzene, toluene, and xylene natural attenuation by quantitative polymerase chain reaction analysis of a catabolic gene, signature metabolites, and compound-specific isotope analysis. Environmental Science & Technology 42: 6065- 6072. Bjerketvedt, D., Bakke, J.R., and van Wingerden, K. (1997) Gas explosion handbook. Journal of Hazardous Materials 52: 1-150. Bogner, J., Meadows, M., and Czepiel, P. (1997) Fluxes of methane between landfills and the atmosphere: natural and engineered controls. Soil Use and Management 13: 268-277. Bohdalkova, L., Curik, J., Kubena, A.A., and Buzek, F. (2013) Dynamics of methane fluxes from two peat bogs in the Ore Mountains, Czech Republic. Plant Soil and Environment 59: 14-21. Botton, S., van Heusden, M., Parsons, J.R., Smidt, H., and van Straalen, N. (2006) Resilience of microbial systems towards disturbances. Critical Reviews in Microbiology 32: 101-112. Bozkurt, O., Pennell, K.G., and Suuberg, E.M. (2009) Simulation of the Vapor Intrusion Process for Nonhomogeneous Soils Using a Three-Dimensional Numerical Model. Ground Water Monitoring and Remediation 29: 92-104. Bringmann, G., and Kuhn, R. (1980) Comparison of the toxicity thresholds of water pollutants to bacteria, algae, and protozoa in the cell multiplication inhibition test. Water Research 14: 231-241. Brown, R.A., Zimmerman, M.D., and Ririe, G.T. (2010) Attenuation of naturally occurring arsenic at petroleum hydrocarbon–impacted sites. In 7th International Conference on Remediation of Chlorinated and Recalcitrant Compounds. Monterey, California.

190

Brune, A., Frenzel, P., and Cypionka, H. (2000) Life at the oxic-anoxic interface: microbial activities and adaptations. FEMS Microbiology Reviews 24: 691-710. Buscheck, T.E., Reilly, K.O., Koschal, G., and Davis, G. (2001) Ethanol in groundwater at the Pacific Northwest terminal. In Proceedings of the NGWA/APl Petroleum Hydrocarbons Conference. Westerville, Ohio, US: National Ground Water Association. Buscheck, T.E., Reilly, K.O., Koschal, G., and Regan, G.O. (2011) Ethanol in groundwater at the Pacific Northwest terminal. In NGWA/APl Petroleum Hydrocarbons Conference. Westerville, Ohio, US: National Ground Water Association. Cantwell, S.G., Lau, E.P., Watt, D.S., and Fall, R.R. (1978) Biodegradation of acyclic isoprenoids by pseudomonas species. Journal of Bacteriology 135: 324-333. Capiro, N.L. (2007) Evaluation of fuel ethanol releases in a pilot-scale aquifer tank: Source dynamics, NAPL migration and microbial community response. In. Houston, TX, US: Rice University. Capiro, N.L., Stafford, B.P., Rixey, W.G., Bedient, P.B., and Alvarez, P.J.J. (2007) Fuel-grade ethanol transport and impacts to groundwater in a pilot-scale aquifer tank. Water Research 41: 656-664. Capiro, N.L., Da Silva, M.L.B., Stafford, B.P., Rixey, W.G., and Alvarez, P.J.J. (2008) Microbial community response to a release of neat ethanol onto residual hydrocarbons in a pilot-scale aquifer tank. Environmental Microbiology 10: 2236-2244. Caporaso, J.G., Bittinger, K., Bushman, F.D., DeSantis, T.Z., Andersen, G.L., and Knight, R. (2010a) PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics 26: 266-267. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K. et al. (2010b) QIIME allows analysis of high-throughput community sequencing data. Nat Meth 7: 335-336. Cardellini, C., Chiodini, G., Frondini, F., Granieri, D., Lewicki, J., and Peruzzi, L. (2003) Accumulation chamber measurements of methane fluxes: application to volcanic- geothermal areas and landfills. Applied Geochemistry 18: 45-54. Cardenas, E., and Tiedje, J.M. (2008) New tools for discovering and characterizing microbial diversity. Current Opinion in Biotechnology 19: 544-549. Cardenas, E., Wu, W.M., Leigh, M.B., Carley, J., Carroll, S., Gentry, T. et al. (2010) Significant association between sulfate-reducing bacteria and uranium-reducing microbial communities as revealed by a combined massively parallel sequencing-indicator species approach. Applied and Environmental Microbiology 76: 6778-6786. Chen, C.S., Lai, Y.W., and Tien, C.J. (2008a) Partitioning of aromatic and oxygenated constituents into water from regular and ethanol-blended gasolines. Environmental Pollution 156: 988-996. Chen, Y.D., Barker, J.F., and Gui, L. (2008b) A strategy for aromatic hydrocarbon bioremediation under anaerobic conditions and the impacts of ethanol: A microcosm study. Journal of Contaminant Hydrology 96: 17-31. Cheremisinoff, N.P. (1999) Handbook of industrial toxicology and hazardous materials. New York, NY, CRC Press; 1 edition Chin, K.J., Liesack, W., and Janssen, P.H. (2001) Opitutus terrae gen nov, sp nov, to accommodate novel strains of the division 'Verrucomicrobia' isolated from rice paddy soil. International Journal of Systematic and Evolutionary Microbiology 51: 1965-1968.

191

Chou, J.-H., Jiang, S.-R., Cho, J.-C., Song, J., Lin, M.-C., and Chen, W.-M. (2008) Azonexus hydrophilus sp nov., a nifH gene-harbouring bacterium isolated from freshwater. International Journal of Systematic and Evolutionary Microbiology 58: 946-951. Coates, J.D., Ellis, D.J., Gaw, C.V., and Lovley, D.R. (1999) Geothrix fermentans gen. nov., sp nov., a novel Fe(III)-reducing bacterium from a hydrocarbon-contaminated aquifer. International Journal of Systematic Bacteriology 49: 1615-1622. Coates, J.D., Lonergan, D.J., Philips, E.J.P., Jenter, H., and Lovley, D.R. (1995) Desulfuromonas palmitatis sp nov, a marine dissimilatory Fe III reducer that can oxidize long-chain fatty acids. Archives of Microbiology 164: 406-413. Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J. et al. (2009) The ribosomal database project: improved alignments and new tools for rRNA analysis. Nucleic Acids Research 37: D141-D145. Conrad, R., Schutz, H., and Babbel, M. (1987) Temperature limitation of hydrogen turnover and methanogenesis in anoxic paddy soil. FEMS Microbiology Ecology 45: 281-289. Corseuil, H.X., Kaipper, B.I.A., and Fernandes, M. (2004) Cosolvency effect in subsurface systems contaminated with petroleum hydrocarbons and ethanol. Water Research 38: 1449-1456. Corseuil, H.X., Hunt, C.S., Dos Santos, R.C.F., and Alvarez, P.J.J. (1998) The influence of the gasoline oxygenate ethanol on aerobic and anaerobic BTX biodegradation. Water Research 32: 2065-2072. Corseuil, H.X., Monier, A.L., Fernandes, M., Schneider, M.R., Nunes, C.C., do Rosario, M., and Alvarez, P.J.J. (2011) BTEX plume dynamics following an ethanol blend release: geochemical footprint and thermodynamic constraints on natural attenuation. Environmental Science & Technology 45: 3422-3429. Cullimore, D.R., Maule, A., and Mansuy, N. (1985) Ambient-temperature methanogenensis from pig manure waste lagoones-thermal-gradient incubator studies. Agricultural Wastes 12: 147-157. Cunningham, R.E., and Williams, R.J.J. (1980) Diffusion in Gases and Porous Media. New York, NY, US: Plenum Press. Da Silva, M.L.B., and Alvarez, P.J.J. (2002) Effects of ethanol versus MTBE on benzene, toluene, ethylbenzene, and xylene natural attenuation in aquifer columns. Journal of Environmental Engineering-Asce 128: 862-867. Da Silva, M.L.B., and Alvarez, P.J.J. (2004) Enhanced anaerobic biodegradation of benzene- toluene-ethylbenzene-xylene-ethanol mixtures in bioaugmented aquifer columns. Applied and Environmental Microbiology 70: 4720-4726. Da Silva, M.L.B., and Alvarez, P.J.J. (2010) Indole-based assay to assess the effect of ethanol on Pseudomonas putida F1 dioxygenase activity. Biodegradation 21: 425-430. Da Silva, M.L.B., and Corseuil, H.X. (2012) Groundwater microbial analysis to assess enhanced BTEX biodegradation by nitrate injection at a gasohol-contaminated site. International Biodeterioration & Biodegradation 67: 21-27. Dakhel, N., Pasteris, G., Werner, D., and Hohener, P. (2003) Small-volume releases of gasoline in the Vadose Zone: Impact of the additives MTBE and ethanol on groundwater quality. Environmental Science & Technology 37: 2127-2133. Datta, A., Yeluripati, J.B., Nayak, D.R., Mahata, K.R., Santra, S.C., and Adhya, T.K. (2013) Seasonal variation of methane flux from coastal saline rice field with the application of different organic manures. Atmospheric Environment 66: 114-122.

192

Davis, G.B., Patterson, B.M., and Trefry, M.G. (2009) Evidence for instantaneous oxygen- limited biodegradation of petroleum hydrocarbon vapors in the subsurface. Ground Water Monitoring and Remediation 29: 126-137. Davis, G.B., Rayner, J.L., Trefry, M.G., Fisher, S.J., and Patterson, B.M. (2005) Measurement and modeling of temporal variations in hydrocarbon vapor behavior in a layered soil profile. Vadose Zone Journal 4: 225-239. de Lorenzo, V. (2008) Systems biology approaches to bioremediation. Current Opinion in Biotechnology 19: 579-589. Deborde, J., Anschutz, P., Guerin, F., Poirier, D., Marty, D., Boucher, G. et al. (2010) Methane sources, sinks and fluxes in a temperate tidal Lagoon: The Arcachon lagoon (SW France). Estuarine Coastal and Shelf Science 89: 256-266. Deeb, R.A., Sharp, J.O., Stocking, A., McDonald, S., West, K.A., Laugier, M. et al. (2002) Impact of ethanol on benzene plume lengths: Microbial and modeling studies. Journal of Environmental Engineering-Asce 128: 868-875. DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K. et al. (2006) Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Applied and Environmental Microbiology 72: 5069-5072. Deutsch, W.J. (1997) Groundwater geochemistry: fundamentals and applications to contamination. Boca Raton, FL, USA: CRC Press. DeVaull, G., McHugh, T.E., and Newberry, P. (2010) Users Manual "Biovapor: a 1-D Vapor Intrusion Model with Oxygen-Limited Aerobic Biodegradation". In. Washington, DC American Petroleum Institute. DeVaull, G.E. (2007) Indoor vapor intrusion with oxygen-limited biodegradation for a subsurface gasoline source. Environmental Science & Technology 41: 3241-3248. Duetz, W.A., Dejong, C., Williams, P.A., and Vanandel, J.G. (1994) Competition in chemostat culture between Pseudomonas strains that use different pathways for the degradation of toluene. Applied and Environmental Microbiology 60: 2858-2863. Eaton, L.C., Tedder, T.F., and Ingram, L.O. (1982) Effects of fatty acid composition on the sensitivity of membrane functions to ethanol in Escherichia coli. Substance and alcohol actions/misuse 3: 77-87. Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792-1797. Edgar, R.C. (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460-2461. Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., and Knight, R. (2011) UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27: 2194-2200. Egli, T. (1995) The ecological and physiological significance of the growth of heterotrophic microorganisms with mixtures of substrates. In Advances in Microbial Ecology, Vol 14. Jones, J.G. (ed). New York, NY, USA: Plenum Press, pp. 305-386. Elberling, B., Askaer, L., Jorgensen, C.J., Joensen, H.P., Kuhl, M., Glud, R.N., and Lauritsen, F.R. (2011) Linking Soil O-2, CO2, and CH4 Concentrations in a Wetland Soil: Implications for CO2 and CH4 Fluxes. Environmental Science & Technology 45: 3393- 3399. EPA (2002a) OSWER draft guidance for evaluating the vapor intrusion to indoor air pathway from groundwater and soils (subsurface vapor intrusion guidance). In. Washington, DC: U.S. Environmental Protection Agency.

193

EPA (2002b) National Secondary Drinking Water Regulations. In. Washington, DC, USA: U.S. Environmental Protection Agency. EPA (2012a) Renewable Fuels: Regulations and Standards In. Washington, DC, US: U.S. Environmental Protection Agency. EPA (2012b) Conceptual Model Scenarios for the Vapor Intrusion Pathway. In. Washington, DC: Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. EPA (2012c) EPA’s vapor intrusion database: evaluation and characterization of attenuation factors for chlorinated volatile organic compounds and residential buildings. In. Washington, DC: Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. EPA (2013) Evaluation Of Empirical Data To Support Soil Vapor Intrusion Screening Criteria For Petroleum Hydrocarbon Compounds. In. Washington, DC: Office of Underground Storage Tanks, U.S. Environmental Protection Agency. Erkel, C., Kube, M., Reinhardt, R., and Liesack, W. (2006) Genome of Rice Cluster I archaea - the key methane producers in the rice rhizosphere. Science 313: 370-372. Eyre, B.D., and Ferguson, A.J.P. (2009) Denitrification efficiency for defining critical loads of carbon in shallow coastal ecosystems. Hydrobiologia 629: 137-146. Fahrbach, M., Kuever, J., Meinke, R., Kaempfer, P., and Hollender, J. (2006) Denitratisoma oestradiolicum gen. nov., sp nov., a 17 beta-oestradiol-degrading, denitrifying betaproteobacterium. International Journal of Systematic and Evolutionary Microbiology 56: 1547-1552. FAPRI-ISU (2012) World Agricultural Outlook. In. Ames, Iowa, USA: Food And Agricultural Policy Research Institute, Iowa State University Feris, K., Mackay, D., de Sieyes, N., Chakraborty, I., Einarson, M., Hristova, K., and Scow, K. (2008) Effect of ethanol on microbial community structure and function during natural attenuation of benzene, toluene, and o-xylene in a sulfate-reducing aquifer. Environmental Science & Technology 42: 2289-2294. Fitzpatrick, N.A., and Fitzgerald, J.J. (2002) An evaluation of vapor intrusion into buildings through a study of field data. Soil & Sediment Contamination 11: 603-623. Foght, J. (2008) Anaerobic biodegradation of aromatic hydrocarbons: Pathways and prospects. Journal of Molecular Microbiology and Biotechnology 15: 93-120. Frank, D.N., Amand, A.L.S., Feldman, R.A., Boedeker, E.C., Harpaz, N., and Pace, N.R. (2007) Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proceedings of the National Academy of Sciences of the United States of America 104: 13780-13785. Freitas, J.G., and Barker, J.F. (2011) Oxygenated gasoline release in the unsaturated zone - Part 1: Source zone behavior. Journal of Contaminant Hydrology 126: 153-166. Freitas, J.G., Fletcher, B., Aravena, R., and Barker, J.F. (2010) Methane production and isotopic fingerprinting in ethanol fuel contaminated sites. Ground Water 48: 844-857. Freitas, J.G., Doulatyari, B., Molson, J.W., and Barker, J.F. (2011a) Oxygenated gasoline release in the unsaturated zone, Part 2: Downgradient transport of ethanol and hydrocarbons. Journal of Contaminant Hydrology 125: 70-85. Freitas, J.G., Mocanu, M.T., Zoby, J.L.G., Molson, J.W., and Barker, J.F. (2011b) Migration and fate of ethanol-enhanced gasoline in groundwater: A modelling analysis of a field experiment. Journal of Contaminant Hydrology 119: 25-43.

194

Friedrich, M.W. (2005) Methyl Coenzyme M Reductase Genes: Unique Functional Markers for Methanogenic and Anaerobic Methane[hyphen (true graphic)]Oxidizing Archaea. In Methods in Enzymology. Jared, R.L. (ed): Academic Press, pp. 428-442. Fuchs, G., Boll, M., and Heider, J. (2011) Microbial degradation of aromatic compounds - from one strategy to four. Nature Reviews Microbiology 9: 803-816. Gibson, D.M. (1965) Biosynthesis of fatty acids. Journal of Chemical Education 42: 236-&. Gomez, D.E., and Alvarez, P.J.J. (2009) Modeling the natural attenuation of benzene in groundwater impacted by ethanol-blended fuels: Effect of ethanol content on the lifespan and maximum length of benzene plumes. Water Resources Research 45: W03409. Gomez, D.E., and Alvarez, P.J.J. (2010) Comparing the effects of various fuel alcohols on the natural attenuation of Benzene Plumes using a general substrate interaction model. Journal of Contaminant Hydrology 113: 66-76. Gomez, D.E., de Blanc, P.C., Rixey, W.G., Bedient, P.B., and Alvarez, P.J.J. (2008) Modeling benzene plume elongation mechanisms exerted by ethanol using RT3D with a general substrate interaction module. Water Resources Research 44: W05405. Greenberg, A.E., Clesceri, L.S., and Eaton, A.D. (1992) Standard Methods for the Examination of Water and Wastewater 18 th edition: American Public Health Association. Griebler, C., and Lueders, T. (2009) Microbial biodiversity in groundwater ecosystems. Freshwater Biology 54: 649-677. Griffiths, B.S., and Philippot, L. (2013) Insights into the resistance and resilience of the soil microbial community. Fems Microbiology Reviews 37: 112-129. Grimberg, S.J., Stringfellow, W.T., and Aitken, M.D. (1996) Quantifying the biodegradation of phenanthrene by Pseudomonas stutzeri P16 in the presence of a nonionic surfactant. Applied and Environmental Microbiology 62: 2387-2392. Gupta, S.K., Kumari, R., Prakash, O., and Lal, R. (2008) Pseudomonas panipatensis sp nov., isolated from an oil-contaminated site. International Journal of Systematic and Evolutionary Microbiology 58: 1339-1345. Hanson, R.S., and Hanson, T.E. (1996) Methanotrophic bacteria. Microbiological Reviews 60: 439-471. Harms, G., Layton, A.C., Dionisi, H.M., Gregory, I.R., Garrett, V.M., Hawkins, S.A. et al. (2003) Real-time PCR quantification of nitrifying bacteria in a municipal wastewater treatment plant. Environmental Science & Technology 37: 343-351. Harmsen, H.J.M., Wildeboer-Veloo, A.C.M., Grijpstra, J., Knol, J., Degener, J.E., and Welling, G.W. (2000) Development of 16S rRNA-based probes for the Coriobacterium group and the Atopobium cluster and their application for enumeration of Coriobacteriaceae in human feces from volunteers of different age groups. Applied and Environmental Microbiology 66: 4523-4527. Haseltin.Wa, Block, R., Weber, K., and Gilbert, W. (1972) MSI and MSII made on ribosome in idling step of protein synthesis. Nature 238: 381-&. He, X.H., Stafford, B.P., and Rixey, W.G. (2011) Ethanol-enhanced dissolution of a residually trapped synthetic gasoline source. Ground Water Monitoring and Remediation 31: 61-68. He, Z.L., Van Nostrand, J.D., and Zhou, J.Z. (2012a) Applications of functional gene microarrays for profiling microbial communities. Current Opinion in Biotechnology 23: 460-466. He, Z.L., Deng, Y., and Zhou, J.Z. (2012b) Development of functional gene microarrays for microbial community analysis. Current Opinion in Biotechnology 23: 49-55.

195

Heath, R.C. (1983) Basic ground-water hydrology: U.S. Geological Survey, Water Supply Paper 2220. Heermann, S.E., and Powers, S.E. (1996) The dissolution of BTEX compounds from oxygenated gasoline. Abstracts of Papers of the American Chemical Society 211: 104-ENVR. Heermann, S.E., and Powers, S.E. (1998) Modeling the partitioning of BTEX in water- reformulated gasoline systems containing ethanol. Journal of Contaminant Hydrology 34: 315-341. Heipieper, H.J., and Debont, J.A.M. (1994) Adaptation of Pseudomonas putida S12 to ethanol and toluene at the level of fatty acid composition of membranes. Applied and Environmental Microbiology 60: 4440-4444. Hers, I., Atwater, J., Li, L., and Zapf-Gilje, R. (2000) Evaluation of vadose zone biodegradation of BTX vapours. Journal of Contaminant Hydrology 46: 233-264. Hirano, S.I., Kitauchi, F., Haruki, M., Imanaka, T., Morikawa, M., and Kanaya, S. (2004) Isolation and characterization of Xanthobacter polyaromaticivorans sp nov 127W that degrades polycyclic and heterocyclic aromatic compounds under extremely low oxygen conditions. Bioscience Biotechnology and Biochemistry 68: 557-564. Hirota, M., Senga, Y., Seike, Y., Nohara, S., and Kunii, H. (2007) Fluxes of carbon dioxide, methane and nitrous oxide in two contrastive fringing zones of coastal lagoon, Lake Nakaumi, Japan. Chemosphere 68: 597-603. Howard, P.H. (1990) Handbook of Environmental Fate and Exposure Data for Organic Chemicals: CRC Press. Howard, P.H. (1997) Handbook of Environmental Fate and Exposure Data for Organic Chemicals: CRC Press. Iino, T., Mori, K., Uchino, Y., Nakagawa, T., Harayama, S., and Suzuki, K.-i. (2010) Ignavibacterium album gen. nov., sp nov., a moderately thermophilic anaerobic bacterium isolated from microbial mats at a terrestrial hot spring and proposal of Ignavibacteria classis nov., for a novel lineage at the periphery of green sulfur bacteria. International Journal of Systematic and Evolutionary Microbiology 60: 1376-1382. Ingram, L.O., and Vreeland, N.S. (1980) Differential effects of ethanol and hexanol on the Escherichia coli cell envelope. Journal of Bacteriology 144: 481-488. Ingram, L.O., and Buttke, T.M. (1984) Effects of alcohols on microorganisms. Advances in Microbial Physiology 25: 253-300. Jain, N., Pathak, H., Mitra, S., and Bhatia, A. (2004) Emission of methane from rice fields - A review. Journal of Scientific & Industrial Research 63: 101-115. Jewell, K.P., and Wilson, J.T. (2011) A new screening method for methane in soil gas using existing groundwater monitoring wells. Ground Water Monitoring and Remediation 31: 82-94. Johnson, P.C., and Ettinger, R.A. (1991) Heuristic model for predicting the intrusion rate of contaminant vapors into buildings. Environmental Science & Technology 25: 1445-1452. Jones, H.A., and Nedwell, D.B. (1993) Methane emission and methane oxidation in land-fill cover soil. FEMS Microbiology Ecology 102: 185-195. Jourabchi, P., Hers, I., Mayer, K.U., DeVaull, G., Kolhatkar, R., and B., B. (2013) Numerical Modeling Study of the Influence of Methane Generation from Ethanol-Gasoline Blends on Vapor Intrusion. In The Second International Symposium on Bioremediation and Sustainable Environmental Technologies. Jacksonville, FL, US. Krieg, N.R. (2010) Bergey's Manual of Systematic Bacteriology. New York: Springer.

196

Kristensen, A.H., Poulsen, T.G., Mortensen, L., and Moldrup, P. (2010) Variability of soil potential for biodegradation of petroleum hydrocarbons in a heterogeneous subsurface. Journal of Hazardous Materials 179: 573-580. Kruger, M., and Frenzel, P. (2003) Effects of N-fertilisation on CH4 oxidation and production, and consequences for CH4 emissions from microcosms and rice fields. Global Change Biology 9: 773-784. Lahvis, M.A., and Baehr, A.L. (1996) Estimation of rates of aerobic hydrocarbon biodegradation by simulation of gas transport in the unsaturated zone. Water Resources Research 32: 2231-2249. Lahvis, M.A., Hers, I., Davis, R.V., Wright, J., and DeVaull, G.E. (2013) Vapor Intrusion Screening at Petroleum UST Sites. Groundwater Monitoring & Remediation 33: 53-67. Leahy, J.G., and Olsen, R.H. (1997) Kinetics of toluene degradation by toluene-oxidizing bacteria as a function of oxygen concentration, and the effect of nitrate. FEMS Microbiology Ecology 23: 23-30. Leaphart, A.B., and Lovell, C.R. (2001) Recovery and analysis of formyltetrahydrofolate synthetase gene sequences from natural populations of acetogenic bacteria. Applied and Environmental Microbiology 67: 1392-1395. Lee, D.H., Moon, S.R., Park, Y.H., Kim, J.H., Kim, H., Parales, R.E., and Kahng, H.Y. (2010) Pseudomonas taeanensis sp nov., isolated from a crude oil-contaminated seashore. International Journal of Systematic and Evolutionary Microbiology 60: 2719-2723. Lemos, L.N., Fulthorpe, R.R., Triplett, E.W., and Roesch, L.F.W. (2011) Rethinking microbial diversity analysis in the high throughput sequencing era. Journal of Microbiological Methods 86: 42-51. Liles, M.R., Turkmen, O., Manske, B.F., Zhang, M., Rouillard, J.-M., George, I. et al. (2010) A phylogenetic microarray targeting 16S rRNA genes from the bacterial division Acidobacteria reveals a lineage-specific distribution in a soil clay fraction. Soil Biology and Biochemistry 42: 739-747. Lillis, L., Clipson, N., and Doyle, E. (2010) Quantification of catechol dioxygenase gene expression in soil during degradation of 2,4-dichlorophenol. FEMS Microbiology Ecology 73: 363-369. Lovanh, N., and Alvarez, P.J.J. (2004) Effect of ethanol, acetate, and phenol on toluene degradation activity and tod-lux expression in Pseudomonas putida TOD102: Evaluation of the metabolic flux dilution model. Biotechnology and Bioengineering 86: 801-808. Lovanh, N., Hunt, C.S., and Alvarez, P.J.J. (2002) Effect of ethanol on BTEX biodegradation kinetics: aerobic continuous culture experiments. Water Research 36: 3739-3746. Lovley, D.R., Ueki, T., Zhang, T., Malvankar, N.S., Shrestha, P.M., Flanagan, K.A. et al. (2011) Geobacter: The Microbe Electric's Physiology, Ecology, and Practical Applications. In Advances in Microbial Physiology, Vol 59. Poole, R.K. (ed), pp. 1-100. Lozupone, C.A., and Knight, R. (2007) Global patterns in bacterial diversity. Proceedings of the National Academy of Sciences of the United States of America 104: 11436-11440. Luecker, S., Wagner, M., Maixner, F., Pelletier, E., Koch, H., Vacherie, B. et al. (2010) A Nitrospira metagenome illuminates the physiology and evolution of globally important nitrite-oxidizing bacteria. Proceedings of the National Academy of Sciences of the United States of America 107: 13479-13484.

197

Lundegard, P.D., and Johnson, P.C. (2006) Source zone natural attenuation at petroleum hydrocarbon spill sites - II: Application to a former oil field. Ground Water Monitoring and Remediation 26: 93-106. Lundegard, P.D., Johnson, P.C., and Dahlen, P. (2008) Oxygen transport from the atmosphere to soil gas beneath a slab-on-grade foundation overlying petroleum-impacted soil. Environmental Science & Technology 42: 5534-5540. Luo, H. (2009) Field and modeling studies of soil vapor migration into buildings at petroleum hydrocarbon impacted sites. In. Phoenix, AZ, US: Arizona State University, p. 324. Luton, P.E., Wayne, J.M., Sharp, R.J., and Riley, P.W. (2002) The mcrA gene as an alternative to 16S rRNA in the phylogenetic analysis of methanogen populations in landfill. Microbiology-Sgm 148: 3521-3530. Ma, J., Rixey, W.G., and Alvarez, P.J.J. (2013a) Microbial processes influencing the transport, fate and groundwater impacts of fuel ethanol releases. Current Opinion in Biotechnology 24: 457-466. Ma, J., Nossa, C.W., Xiu, Z., Rixey, W.G., and Alvarez, P.J.J. Adaptive changes in microbial community structure in response to a continuous pilot-scale release of an ethanol blend. Environmental Microbiology (minor revision). Ma, J., Rixey, W.G., DeVaull, G.E., Stafford, B.P., and Alvarez, P.J.J. (2012) Methane bioattenuation and implications for explosion risk reduction along the groundwater to soil surface pathway above a plume of dissolved ethanol. Environmental Science & Technology 46: 6013–6019. Ma, J., Nossa, C.W., Xiu, Z., Rixey, W.G., and Alvarez, P.J.J. (2013b) Adaptive microbial population shifts in response to a continuous ethanol blend release increases biodegradation potential. Environmental Pollution 178: 419-425. Ma, J., Xiu, Z., Monier, A., Mamonkina, I., Zhang, Y., He, Y. et al. (2011) Aesthetic groundwater quality Impacts from a continuous pilot-scale release of an ethanol blend. Ground Water Monitoring & Remediation 31: 47–54. Mackay, D.M., De Sieyes, N.R., Einarson, M.D., Feris, K.P., Pappas, A.A., Wood, I.A. et al. (2006) Impact of ethanol on the natural attenuation of benzene, toluene, and o-xylene in a normally sulfate-reducing aquifer. Environmental Science & Technology 40: 6123-6130. Madigan, M.T., and Martinko, J.M. (2006) Brock Biology of Microorganisms Upper Saddle River, NJ, USA: Pearson/Prentice Hall. Marchandin, H., Teyssier, C., Campos, J., Jean-Pierre, H., Roger, F., Gay, B. et al. (2010) Negativicoccus succinicivorans gen. nov., sp nov., isolated from human clinical samples, emended description of the family Veillonellaceae and description of Negativicutes classis nov., Selenomonadales ord. nov and Acidaminococcaceae fam. nov in the bacterial phylum Firmicutes. International Journal of Systematic and Evolutionary Microbiology 60: 1271-1279. Mardis, E.R. (2008a) The impact of next-generation sequencing technology on genetics. Trends in Genetics 24: 133-141. Mardis, E.R. (2008b) Next-generation DNA sequencing methods. Annual Review of Genomics and Human Genetics 9: 387-402. Mason, E.A., and Malinauskas, A.P. (1983) Gas Transport in Porous Media: The Dusty-Gas Model (Chemical Engineering Monographs). New York, NY, US: Elsevier Science Ltd.

198

McDowell, C.J., and Powers, S.E. (2003) Mechanisms affecting the infiltration and distribution of ethanol-blended gasoline in the vadose zone. Environmental Science & Technology 37: 1803-1810. McDowell, C.J., Buscheck, T., and Powers, S.E. (2003) Behavior of gasoline pools following a denatured ethanol spill. Ground Water 41: 746-757. McHugh, T.E., and McAlary, T. (2009) Important physical processes for vapor intrusion: a literature review. in Proceedings of AWMA Vapor Intrusion Conference, San Diego, CA. McNab, W., Heermann, S.E., and Dooher, B. (1999) Screening model evaluation of the effects of ethanol on benzene plume lengths In Health and Environmental Assessment of the Use of Ethanol as a Fuel Oxygenate—Report to the California Environmental Policy Council in Response to Executive Order D-5-99. Rice, D.W., and Cannon, G. (eds). Livermore, Calif., USA: Lawrence Livermore Nattionl Lab. Mechichi, T., Labat, M., Patel, B.K.C., Woo, T.H.S., Thomas, P., and Garcia, J.L. (1999) Clostridium methoxybenzovorans sp nov., a new aromatic o-demethylating homoacetogen from an olive mill wastewater treatment digester. International Journal of Systematic Bacteriology 49: 1201-1209. Mills, W.B., Liu, S., Rigby, M.C., and Brenner, D. (2007) Time-variable simulation of soil vapor intrusion into a building with a combined crawl space and basement. Environmental Science & Technology 41: 4993-5001. Mitchell, J.J., and Lucaslenard, J.M. (1980) The effect of alcohols on guanosine 5'-diphosphate- 3'-diphosphate metabolism in stringent and relaxed Escherichia coli. Journal of Biological Chemistry 255: 6307-6313. Molins, S., Mayer, K.U., Scheutz, C., and Kjeldsen, P. (2008) Transport and reaction processes affecting the attenuation of landfill gas in cover soils. Journal of Environmental Quality 37: 459-468. Molins, S., Mayer, K.U., Amos, R.T., and Bekins, B.A. (2010) Vadose zone attenuation of organic compounds at a crude oil spill site - Interactions between biogeochemical reactions and multicomponent gas transport. Journal of Contaminant Hydrology 112: 15- 29. Molson, J., Mocanu, M., and Barker, J. (2008) Numerical analysis of buoyancy effects during the dissolution and transport of oxygenated gasoline in groundwater. Water Resources Research 44: W07418. Molson, J.W., Barker, J.F., Frind, E.O., and Schirmer, M. (2002) Modeling the impact of ethanol on the persistence of benzene in gasoline-contaminated groundwater. Water Resources Research 38: 1003. Nagata, Y. (2003) Measurement of Odor Threshold by Triangle Odor Bag Method Odor Measurement Review 118-127, Japan Ministry of the Environment. Nastev, M., Therrien, R., Lefebvre, R., and Gelinas, P. (2001) Gas production and migration in landfills and geological materials. Journal of Contaminant Hydrology 52: 187-211. Nebe, J., Baldwin, B.R., Kassab, R.L., Nies, L., and Nakatsu, C.H. (2009) Quantification of aromatic oxygenase genes to evaluate enhanced bioremediation by oxygen releasing materials at a gasoline-contaminated site. Environmental Science & Technology 43: 2029-2034. Nelson, D.K., Lapara, T.M., and Novak, P.J. (2010) Effects of ethanol-based fuel contamination: microbial community changes, production of regulated compounds, and methane generation. Environmental Science & Technology 44: 4525-4530.

199

Osborn, M., and Smith, C. (2005) Molecular microbial ecology London, UK: Taylor & Francis. Osztovics, M., Igali, S., Antal, A., and Veghelyi, P. (1980) Alcohol is not mutagenic. Mutation Research 74: 247-247. Owen-Hughes, T., and Engeholm, M. (2007) Pyrosequencing positions nucleosomes precisely. Genome Biology 8. Paredes-Sabja, D., Setlow, P., and Sarker, M.R. (2011) Germination of spores of Bacillales and Clostridiales species: mechanisms and proteins involved. Trends in Microbiology 19: 85- 94. Patterson, B.M., and Davis, G.B. (2009) Quantification of vapor intrusion pathways into a slab- on-ground building under varying environmental conditions. Environmental Science & Technology 43: 650-656. Pennell, K.G., Bozkurt, O., and Suuberg, E.M. (2009) Development and application of a three- dimensional finite element vapor intrusion model. Journal of the Air & Waste Management Association 59: 447-460. Peterson, G., Allen, C.R., and Holling, C.S. (1998) Ecological resilience, biodiversity, and scale. Ecosystems 1: 6-18. Phillips, L.A., Greer, C.W., Farrell, R.E., and Germida, J.J. (2012) Plant root exudates impact the hydrocarbon degradation potential of a weathered-hydrocarbon contaminated soil. Applied Soil Ecology 52: 56-64. Pimm, S.L. (1984) The complexicty and stability of ecosystems. Nature 307: 321-326. Porat, I., Vishnivetskaya, T.A., Mosher, J.J., Brandt, C.C., Yang, Z.M.K., Brooks, S.C. et al. (2010) Characterization of Archaeal Community in Contaminated and Uncontaminated Surface Stream Sediments. Microbial Ecology 60: 784-795. Powers, S.E., Hunt, C.S., Heermann, S.E., Corseuil, H.X., Rice, D., and Alvarez, P.J.J. (2001) The transport and fate of ethanol and BTEX in groundwater contaminated by gasohol. Critical Reviews in Environmental Science and Technology 31: 79-123. Price, M.N., Dehal, P.S., and Arkin, A.P. (2010) FastTree 2-Approximately Maximum- Likelihood Trees for Large Alignments. Plos One 5. Rakoczy, J., Schleinitz, K.M., Muller, N., Richnow, H.H., and Vogt, C. (2011) Effects of hydrogen and acetate on benzene mineralisation under sulphate-reducing conditions. FEMS Microbiology Ecology 77: 238-247. Reardon, K.F., Mosteller, D.C., and Rogers, J.D.B. (2000) Biodegradation kinetics of benzene, toluene, and phenol as single and mixed substrates for Pseudomonas putida F1. Biotechnology and Bioengineering 69: 385-400. Reeder, J., and Knight, R. (2010) Rapidly denoising pyrosequencing amplicon reads by exploiting rank-abundance distributions. Nature Methods 7: 668-669. Renault, P., and Stengel, P. (1994) Modeling oxygen diffusion in aggregated soils: I. Anaerobiosis inside the aggregates. Soil Science Society of America Journal 58: 1017- 1023. Renault, P., and Sierra, J. (1994) Modeling oxygen diffusion in aggregated soils: II. Anaerobiosis in topsoil layers. Soil Science Society of America Journal 58: 1023-1030. Riley, W.J., Robinson, A.L., Gadgil, A.J., and Nazaroff, W.W. (1999) Effects of variable wind speed and direction on radon transport from soil into buildings: model development and exploratory results. Atmospheric Environment 33: 2157-2168.

200

Rivett, M.O., Wealthall, G.P., Dearden, R.A., and McAlary, T.A. (2011) Review of unsaturated- zone transport and attenuation of volatile organic compound (VOC) plumes leached from shallow source zones. Journal of Contaminant Hydrology 123: 130-156. Robinson, A.L., and Sextro, R.G. (1997) Radon entry into buildings driven by atmospheric pressure fluctuations. Environmental Science & Technology 31: 1742-1748. Ronald M. Atlas, R.B. (1993) Microbial Ecology: Fundamentals and Applications Benjamin/Cummings Publishing Company, Inc. Ruiz-Aguilar, G.M.L., O'Reilly, K., and Alvarez, P.J.J. (2003) A comparison of benzene and toluene plume lengths for sites contaminated with regular vs. ethanol-amended gasoline. Ground Water Monitoring and Remediation 23: 48-53. Ruiz-Aguilar, G.M.L., Fernandez-Sanchez, J.M., Kane, S.R., Kim, D., and Alvarez, P.J.J. (2002) Effect of ethanol and methyl-tert-butyl ether on monoaromatic hydrocarbon biodegradation: Response variability for different aquifer materials under various electron-accepting conditions. Environmental Toxicology and Chemistry 21: 2631-2639. Schaefer, C.E., Yang, X.M., Pelz, O., Tsaob, D.T., Streger, S.H., and Steffan, R.J. (2010) Anaerobic biodegradation of iso-butanol and ethanol and their relative effects on BTEX biodegradation in aquifer materials. Chemosphere 81: 1111-1117. Schuetz, C., Bogner, J., Chanton, J., Blake, D., Morcet, M., and Kjeldsen, P. (2003) Comparative oxidation and net emissions of methane and selected mon-methane organic compounds in landfill cover soils. Environmental Science & Technology 37: 5150-5158. Schwarzenbach, R.P., Gschwend, P.M., and Imboden, D.M. (2002) Environmental Organic Chemistry, 2rd edition Hoboken, New Jersey: Wiley-Interscience. Shade, A., Peter, H., Allison, S.D., Baho, D.L., Berga, M., Burgmann, H. et al. (2012) Fundamentals of microbial community resistance and resilience. Frontiers in microbiology 3: 417-417. Sihota, N.J., Mayer, K.U., Toso, M.A., and Atwater, J.F. (2013) Methane emissions and contaminant degradation rates at sites affected by accidental releases of denatured fuel- grade ethanol. Journal of Contaminant Hydrology 151: 1-15. Singh, S., Singh, J.S., and Kashyap, A.K. (1999) Methane flux from irrigated rice fields in relation to crop growth and N-fertilization. Soil Biology & Biochemistry 31: 1219-1228. Smemo, K.A., and Yavitt, J.B. (2006) A multi-year perspective on methane cycling in a shallow peat fen in central New York State, USA. Wetlands 26: 20-29. Smith, R.J., Jeffries, T.C., Roudnew, B., Fitch, A.J., Seymour, J.R., Delpin, M.W. et al. (2012) Metagenomic comparison of microbial communities inhabiting confined and unconfined aquifer ecosystems. Environmental Microbiology 14: 240-253. Spalding, R.F., Toso, M.A., Exner, M.E., Hattan, G., Higgins, T.M., Sekely, A.C., and Jensen, S.D. (2011) Long-term groundwater monitoring results at large, sudden denatured ethanol releases. Ground Water Monitoring and Remediation 31: 69-81. Staats, M., Braster, M., and Roling, W.F.M. (2011) Molecular diversity and distribution of aromatic hydrocarbon-degrading anaerobes across a landfill leachate plume. Environmental Microbiology 13: 1216-1227. Stafford, B.P. (2007) Impacts to Ground Water from Releases of Fuel Grade Ethanol: Source Behavior. Ph.D. Dissertation. In. Houston, TX: University of Houston. Stafford, B.P., and Rixey, W.G. (2011) Distribution of fuel-grade ethanol near a dynamic water table. Ground Water Monitoring and Remediation 31: 55-60.

201

Stafford, B.P., Capiro, N.L., Alvarez, P.J.J., and Rixey, W.G. (2009) Pore water characteristics following a release of neat ethanol onto pre-existing NAPL. Ground Water Monitoring and Remediation 29: 93-104. Stolz, A. (2009) Molecular characteristics of xenobiotic-degrading sphingomonads. Applied Microbiology and Biotechnology 81: 793-811. Suzuki, M.T., Taylor, L.T., and DeLong, E.F. (2000) Quantitative analysis of small-subunit rRNA genes in mixed microbial populations via 5 '-nuclease assays. Applied and Environmental Microbiology 66: 4605-4614. Torz, M., Wietzes, P., Beschkov, V., and Janssen, D.B. (2007) Metabolism of mono- and dihalogenated C1 and C2 compounds by Xanthobacter autotrophicus growing on 1,2- dichloroethane. Biodegradation 18: 145-157. Vandieken, V., and Thamdrup, B. (2013) Identification of acetate-oxidizing bacteria in a coastal marine surface sediment by RNA-stable isotope probing in anoxic slurries and intact cores. Fems Microbiology Ecology 84: 373-386. Vandieken, V., Mussmann, M., Niemann, H., and Jorgensen, B.B. (2006) Desulfuromonas svalbardensis sp nov and Desulfuromusa ferrireducens sp nov., psychrophilic, Fe(III)- reducing bacteria isolated from Arctic sediments, Svalbard. International Journal of Systematic and Evolutionary Microbiology 56: 1133-1139. Verginelli, I., and Baciocchi, R. (2011) Modeling of vapor intrusion from hydrocarbon- contaminated sources accounting for aerobic and anaerobic biodegradation. Journal of Contaminant Hydrology 126: 167-180. Wang, Q., Garrity, G.M., Tiedje, J.M., and Cole, J.R. (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology 73: 5261-5267. Westermann, P. (1993) Temperature lregulation of methanogenesis in wetlands. Chemosphere 26: 321-328. Whittaker, R.H. (1972) Evolution and Measurement of Species Diversity. Taxon 21: 213-251. Whyte, L.G., Bourbonniere, L., and Greer, C.W. (1997) Biodegradation of petroleum hydrocarbons by psychrotrophic Pseudomonas strains possessing both alkane (alk) and naphthalene (nah) catabolic pathways. Applied and Environmental Microbiology 63: 3719-3723. Wilson, J.T., Toso, M., Mackay, D., Sieyes, N.d., and DeVaull, G.E. (2012) What's the Deal with Methane at LUST Spill Sites? Part 1. Lowell, MA, US: New England Interstate Water Pollution Control Commission. Wilson, J.T., Toso, M., Mackay, D., Sieyes, N.d., and DeVaull, G.E. (2013) What's the Deal with Methane at LUST Spill Sites? Part 2: Vapor Intrusion. Lowell, MA, US: New England Interstate Water Pollution Control Commission. Winderl, C., Anneser, B., Griebler, C., Meckenstock, R.U., and Lueders, T. (2008) Depth- resolved quantification of anaerobic toluene degraders and aquifer microbial community patterns in distinct redox zones of a tar oil contaminant plume. Applied and Environmental Microbiology 74: 792-801. Winderl, C., Penning, H., von Netzer, F., Meckenstock, R.U., and Lueders, T. (2010) DNA-SIP identifies sulfate-reducing Clostridia as important toluene degraders in tar-oil- contaminated aquifer sediment. Isme Journal 4: 1314-1325.

202

Xu, K.W., Liu, H., Du, G.C., and Chen, J. (2009) Real-time PCR assays targeting formyltetrahydrofolate synthetase gene to enumerate acetogens in natural and engineered environments. Anaerobe 15: 204-213. Yamada, T., Sekiguchi, Y., Hanada, S., Imachi, H., Ohashi, A., Harada, H., and Kamagata, Y. (2006) Anaerolinea thermolimosa sp nov., Levilinea saccharolytica gen. nov., sp nov and Leptolinea tardivitalis gen. nov., so. nov., novel filamentous anaerobes, and description of the new classes anaerolineae classis nov and Caldilineae classis nov in the bacterial phylum Chloroflexi. International Journal of Systematic and Evolutionary Microbiology 56: 1331-1340. Yamada, T., Imachi, H., Ohashi, A., Harada, H., Hanada, S., Kamagata, Y., and Sekiguchi, Y. (2007) Bellilinea caldifistulae gen. nov., sp nov and Longilinea arvoryzae gen. nov., sp nov., strictly anaerobic, filamentous bacteria of the phylum Chloroflexi isolated from methanogenic propionate-degrading consortia. International Journal of Systematic and Evolutionary Microbiology 57: 2299-2306. Yamamoto, S., Alcauskas, J.B., and Crozier, T.E. (1976) Solubility of methane in distilled water and seawater. Journal of Chemical and Engineering Data 21: 78-80. Yao, Y.J., Shen, R., Pennell, K.G., and Suuberg, E.M. (2011) Comparison of the Johnson- Ettinger Vapor Intrusion Screening Model Predictions with Full Three-Dimensional Model Results. Environmental Science & Technology 45: 2227-2235. Yeh, C.-H., Lin, C.-W., and Wu, C.-H. (2010) A permeable reactive barrier for the bioremediation of BTEX-contaminated groundwater: Microbial community distribution and removal efficiencies. Journal of Hazardous Materials 178: 74-80. Yu, S., Unger, A.J.A., and Parker, B. (2009a) Simulating the fate and transport of TCE from groundwater to indoor air. Journal of Contaminant Hydrology 107: 140-161. Yu, S., Freitas, J.G., Unger, A.J.A., Barker, J.F., and Chatzis, J. (2009b) Simulating the evolution of an ethanol and gasoline source zone within the capillary fringe. Journal of Contaminant Hydrology 105: 1-17. Zhang, Y., Khan, I.A., Chen, X.H., and Spalding, R.F. (2006) Transport and degradation of ethanol in groundwater. Journal of Contaminant Hydrology 82: 183-194. Zhou, N.Y., Jenkins, A., Chion, C., and Leak, D.J. (1999) The alkene monooxygenase from Xanthobacter strain Py2 is closely related to aromatic monooxygenases and catalyzes aromatic monohydroxylation of benzene, toluene, and phenol. Applied and Environmental Microbiology 65: 1589-1595.

203

Appendix I: Supporting information for Chapter 7

These seven figures shows the regression fitting plots for methane generation rates of the microcosm experiments in Chapter 7. The methane generation rates can be found in Table 7.2.

1.0x103 Ethanol: 2 8.0x10 0 mg/L v 6.0x102 ppm 4.0x102 Equation y = a + b*x Adj. R-S 0.890 2 2.0x10 Value Standard ppmv Interce 161.4 78.74268 ppmv Slope 35.35 6.09329 0.0 0 5 10 15 20

Days

Figure A1.1 Regression fitting plots for methane generation rates of the microcosm containing 0 mg/L ethanol.

1.0x104 Ethanol: 8.0x103 500 mg/L

3 v 6.0x10 ppm 4.0x103 Equatio y = a Adj. R- 0.976 2.0x103 Value Standar ppmv Interc -4669. 870.695 ppmv Slope 935.89 83.4932 0.0 0 5 10 15 20 25

Days

Figure A1.2 Regression fitting plots for methane generation rates of the microcosm containing 500 mg/L ethanol.

204

4 2.0x10 Equatio y = a Adj. R-S 0.956 Value Standard ppmv Interc -6319.0 1646.63 4 1.5x10 ppmv Slope 1205.8 127.420 v 1.0x104 ppm

5.0x103 Ethanol: 1000 mg/L 0.0 0 5 10 15 20

Days

Figure A1.3 Regression fitting plots for methane generation rates of the microcosm containing 1000 mg/L ethanol.

3x104 Ethanol: 1500 mg/L 4

v 2x10 ppm

1x104 Equatio y = a Adj. R-S 0.924 Value Standard ppmv Interc -238.2 1849.42 ppmv Slope 626.54 59.4275 0 0 10 20 30 40 50 60

Days

Figure A1.4 Regression fitting plots for methane generation rates of the microcosm containing 1500 mg/L ethanol.

205

Equation y = a + Adj. R-S 0.974 4 Value Standard 2.0x10 ppmv Interce -3452.4 777.4624 ppmv Slope 431.439 21.08837 1.5x104 v

4 ppm 1.0x10

5.0x103 Ethanol: 3400 mg/L 0.0 0 10 20 30 40 50 60 Days

Figure A1.5 Regression fitting plots for methane generation rates of the microcosm containing 3400 mg/L ethanol.

4 1.5x10 Equatio y = a Adj. R-S 0.985 4 Value Standard 1.2x10 ppmv Interc -2817.3 485.849 ppmv Slope 415.99 19.1303 9.0x103 v

3 ppm 6.0x10 Ethanol: 3.0x103 10000 mg/L

0.0 0 10 20 30 40

Days

Figure A1.6 Regression fitting plots for methane generation rates of the microcosm containing 10000 mg/L ethanol.

206

4 1.2x10 Equation y = a + b*x Adj. R-S 0.369 Value Standard 3 ppmv Interce -9937.3 8378.832 9.0x10 ppmv Slope 133.952 80.67954 v 6.0x103 ppm Ethanol: 3 3.0x10 40000 mg/L

0.0 0 30 60 90 120 Days Figure A1.7 Regression fitting plots for methane generation rates of the microcosm containing 40000 mg/L ethanol.

207

Appendix II. Supporting information for Chapter 9

Table A2.1 Sand chemical characteristics

Sample ID Stage 1 Stage 2 Stage 3 Stage 4a Stage 4b Sampling date 8/7/2009 9/5/2011 5/4/2012 9/5/2012 5/23/2013 Sand pH 8.83 ± 0.04 8.50 ± 0.34 8.60 ± 0.05 8.60 ± 0.09 8.4 ± 0.11 Sand Conductivity (umhos/cm) 26.0 ± 1.0 52.7 ± 6.4 40.3 ±4.0 39.3 ± 3.8 31.0 ± 5.6 Organic carbon content 0.063% ± 0.011% 0.118% ± 0.020% 0.081% ± 0.032% 0.078% ± 0.018% 0.079% ± 0.021% - NO 3-nitrogen (mg/g) 3.3 ± 0.4 4.9 ± 1.8 3.4 ± 0.5 3.0 ± 0.4 1.0 ± 0.8 P (mg/g) 2.8 ± 0.3 3.1 ± 0.8 4.0 ± 1.0 1.9 ± 0.3 0.7 ± 0.6 K (mg/g) 16.2 ± 2.0 18.0 ± 1.3 21.5 ± 1.2 18.2 ± 1.3 8.7 ± 1.8 Ca (mg/g) 1122.5 ±125.2 438.4 ±430.0 967.3 ± 799.0 314.3 ± 88.3 238.7 ± 128.2 Mg (mg/g) 20.8 ± 1.8 17.0 ± 1.6 27.5 ± 7.2 13.3 ± 1.5 25.7 ± 2.3 S (mg/g) 14.9 ± 1.4 20.4 ± 4.9 14.6 ± 0.5 18.2 ± 2.7 12.3 ± 3.5 Na (mg/g) 115.8 ± 8.9 109.4 ± 7.5 89.5 ± 7.6 96.4 ± 14.0 38.3 ± 32.3 Fe (mg/g) 1.6 ± 0.1 16.8 ± 3.4 1.9 ± 0.3 7.6 ± 0.8 1.4 ± 0.3 Zn (mg/g) 6.0 ± 0.1 0.1 ± 0.1 0.1 ± 0.1 0.2 ± 0.1 0.1 ± 0.1 Mn (mg/g) 0.2 ± 0.0 0.1 ± 0.0 0.1 ± 0.0 0.0 ± 0.0 0.3 ± 0.0 Cu (mg/g) 2.7 ± 0.1 0.2 ± 0.1 0.1 ± 0.0 0.1 ± 0.0 0.0 ± 0.0

208

Pyrosequencing data

A total of 402,625 were obtained from 15 bacterial, and a total of 244,567 raw sequences were obtained from 15 archaeal 16S rRNA libraries (Table 9.2). Approximately 17%

(bacteria) and 21 % (archaea) of these raw sequences were removed after quality filtering

through QIIME pipeline. The filtered sequences were denoised by QIIME Denoiser.

Chimeric sequences and singletons were removed by UCHEMIC. After that, 18,922 ± 8,680

(bacteria) and 12,420 ± 3,868 (archaea) of high quality sequences were obtained for

downstream analysis (Table 9.2).

209

Table A2.2 Sequence numbers for raw, quality filtered, denoised, and chimera/singleton removed samples

Bacteria Archaea Denoised and Denoised and Sample ID Raw Quality filtered chimera/singleton Raw Quality filtered Chimera/singleton removed removed Stage1-1 27,949 18,268 13,987 16,691 15,253 14,858 Stage1-2 31,224 25,365 23,697 7,918 7,134 6,984 Stage1-3 21,062 16,764 13,544 14,780 11,033 10,106 Stage2-1 27,356 20,097 15,033 21,236 19,936 19,563 Stage2-2 22,800 20,516 19,175 11,396 10,292 9,796 Stage2-3 23,125 20,225 18,399 15,093 13,861 13,247 Stage3-1 29,496 20,877 13,775 16,394 15,417 14,908 Stage3-2 34,653 30,834 26,748 9,802 9,146 8,908 Stage3-3 23,127 20,102 17,406 19,403 18,285 17,425 Stage4a-1 27,008 19,434 14,713 16,761 15,697 15,470 Stage4a-2 50,993 47,994 46,308 13,192 12,401 12,063 Stage4a-3 29,680 25,111 21,355 17,386 16,482 16,062 Stage4b-1 15,897 14,416 12,231 16,513 7,374 7,175 Stage4b-2 19,427 17,654 14,346 23,401 9,171 8,596 Stage4b-3 18,828 16,851 13,109 24,601 11,714 11,142 Average 26,842 ± 8,437 22,301 ± 8,186 18,922 ± 8,680 16,304 ± 4,667 12,880 ± 3,925 12,420 ± 3,868

210

211

Table A2.3 Taxonomic classification of archaeal communities

Stage 1 Stage 2 Stage 3 Stage 4

Phylum Class Order Family Genus S1-1 S1-2 S1-3 S2-1 S2-2 S2-3 S3-1 S3-2 S3-3 S4a-1 S4a-2 S4a-3 S4b-1 S4b-2 S4b-3

Euryarchaeota Methanomicrobia Methanosarcinales Methanosarcinaceae Methanosarcina 0.0% 5.0% 0.1% 47.3% 30.1% 72.0% 70.6% 56.1% 66.4% 82.8% 71.2% 71.8% 21.7% 37.2% 71.7%

Euryarchaeota Methanobacteria Methanobacteriales Methanobacteriaceae Methanobacterium 0.0% 0.9% 0.0% 46.0% 65.1% 24.0% 11.1% 35.6% 23.6% 8.3% 16.9% 19.8% 14.3% 22.4% 8.1%

Other Other Other Other Other 31.5% 32.8% 30.8% 4.3% 2.3% 1.8% 11.0% 5.2% 4.4% 2.3% 2.0% 2.5% 53.1% 27.4% 11.9%

Crenarchaeota Thermoprotei Other Other Other 60.3% 55.3% 42.5% 0.3% 0.2% 0.1% 2.5% 1.0% 2.0% 1.1% 1.1% 1.3% 7.4% 5.0% 2.9%

Other Other Other Other Other 5.6% 4.1% 15.5% 0.2% 0.1% 0.2% 0.2% 0.1% 0.4% 0.2% 0.2% 0.4% 0.9% 2.1% 1.8%

Other Other Other Other Other 2.0% 1.1% 10.8% 0.2% 0.1% 0.1% 0.1% 0.1% 0.2% 0.3% 0.1% 0.1% 0.4% 1.4% 2.4%

Euryarchaeota Methanomicrobia Methanocellales Methanocellaceae Methanocella 0.0% 0.1% 0.0% 0.1% 0.0% 0.0% 0.4% 0.1% 0.4% 3.6% 6.4% 2.4% 1.0% 0.4% 0.1%

Euryarchaeota Other Other Other Other 0.5% 0.5% 0.2% 0.5% 1.1% 1.0% 1.8% 0.8% 0.9% 0.3% 0.6% 0.4% 0.6% 2.7% 0.7%

Euryarchaeota Methanomicrobia Methanosarcinales Methanosarcinaceae Other 0.0% 0.0% 0.0% 0.2% 0.3% 0.3% 0.3% 0.2% 0.4% 0.1% 0.2% 0.2% 0.0% 0.1% 0.1%

Euryarchaeota Thermoplasmata Thermoplasmatales Other Other 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.9% 0.3% 0.2% 0.2% 0.1% 0.1% 0.1% 0.3% 0.1%

Euryarchaeota Methanobacteria Methanobacteriales Methanobacteriaceae Other 0.0% 0.0% 0.0% 0.1% 0.4% 0.2% 0.3% 0.1% 0.2% 0.1% 0.1% 0.1% 0.1% 0.2% 0.0%

Euryarchaeota Methanomicrobia Methanosarcinales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.1% 0.1% 0.1% 0.2% 0.3% 0.2%

Euryarchaeota Methanomicrobia Other Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.3% 0.3% 0.8% 0.3% 0.1% 0.1% 0.0% Thermoplasmatales incertae Euryarchaeota Thermoplasmata Thermoplasmatales Thermogymnomonas 0.0% 0.0% 0.0% 0.5% 0.0% 0.1% 0.3% 0.0% 0.1% 0.2% 0.1% 0.2% 0.1% 0.3% 0.1% sedis Euryarchaeota Halobacteria Halobacteriales Halobacteriaceae Halalkalicoccus 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Euryarchaeota Methanobacteria Methanobacteriales Methanobacteriaceae Methanosphaera 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% Methanomicrobiales incertae Euryarchaeota Methanomicrobia Methanomicrobiales Methanoregula 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% sedis Methanomicrobiales incertae Euryarchaeota Methanomicrobia Methanomicrobiales Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% sedis Euryarchaeota Methanomicrobia Methanomicrobiales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0%

Euryarchaeota Methanomicrobia Methanosarcinales Methanosaetaceae Methanosaeta 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.1% 0.2% 0.1% 0.0% 0.0%

Euryarchaeota Methanomicrobia Methanosarcinales Methanosarcinaceae Methanomethylovorans 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

212

Table A2.4 Taxonomic classification of bacterial communities

Stage 1 Stage 2 Stage 3 Stage 4

Phylum Class Order Family Genus S4a-1 S4a-1 S4a-1 S4a-1 S4a-1 S2-3 S3-1 S3-2 S3-3 S4a-1 S4a-2 S4a-3 S4b-1 S4b-2 S4b-3 11.3 13.7 16.3 36.9 21.6 19.4 16.4 26.7 Proteobacteria Betaproteobacteria Other Other Other 4.7% 3.9% 3.4% 2.8% 2.0% 3.9% 8.9% % % % % % % % % 22.5 23.7 20.9 10.4 13.8 16.5 11.6 13.4 12.4 11.4 Other Other Other Other Other 3.4% 3.8% 4.0% 8.5% 9.7% % % % % % % % % % % Proteobacteria Gammaproteobacteria Other Other Other 6.5% 5.5% 4.5% 3.0% 1.0% 2.3% 3.4% 2.5% 1.3% 2.8% 3.4% 3.1% 8.0% 6.9% 7.8% Clostridium sensu 15.1 12.7 Firmicutes Clostridia Clostridiales Clostridiaceae 1 0.1% 0.1% 0.1% 9.3% 0.6% 2.1% 2.0% 2.7% 2.1% 2.0% 5.0% 1.6% 1.4% stricto % % 15.4 10.9 Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Other 0.1% 0.1% 0.1% 1.2% 0.3% 0.9% 1.9% 4.2% 8.5% 1.6% 0.9% 4.2% 3.3% % % Proteobacteria Alphaproteobacteria Rhizobiales Other Other 6.4% 7.0% 7.5% 2.1% 1.1% 1.8% 8.9% 3.3% 1.8% 1.1% 1.3% 1.7% 2.6% 4.6% 2.1%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Other 2.1% 1.4% 1.8% 1.1% 0.9% 1.2% 4.1% 3.4% 2.4% 6.2% 9.0% 5.7% 4.0% 3.8% 3.4% 11.9 12.1 14.8 Acidobacteria Acidobacteria_Gp6 Gp6 Gp6 Gp6 0.4% 0.5% 0.7% 0.6% 0.4% 0.3% 0.3% 1.5% 0.9% 2.2% 1.8% 1.2% % % % 10.9 18.1 12.3 Firmicutes Negativicutes Selenomonadales Veillonellaceae Other 0.1% 0.0% 0.2% 0.1% 0.5% 0.4% 0.9% 0.8% 0.4% 0.1% 0.1% 0.2% % % % 11.7 Firmicutes Clostridia Clostridiales Other Other 0.1% 0.1% 0.1% 7.9% 8.1% 0.2% 1.4% 1.2% 1.9% 0.7% 1.3% 0.9% 0.3% 0.7% % Acidobacteria Acidobacteria_Gp3 Gp3 Gp3 Gp3 5.6% 4.3% 5.5% 0.3% 0.3% 0.5% 1.5% 2.6% 3.4% 1.1% 1.3% 1.1% 2.9% 1.9% 1.5% 12.9 Bacteroidetes Other Other Other Other 0.3% 0.2% 0.2% 0.8% 0.6% 0.5% 1.8% 5.5% 1.9% 0.7% 1.4% 0.7% 0.7% 2.6% % Proteobacteria Alphaproteobacteria Other Other Other 5.8% 6.2% 5.0% 0.8% 0.5% 0.8% 1.4% 0.9% 0.5% 0.6% 0.5% 0.9% 2.7% 1.9% 2.3%

Proteobacteria Deltaproteobacteria Other Other Other 2.7% 2.0% 1.2% 0.5% 0.3% 0.4% 1.4% 1.6% 1.3% 2.6% 2.4% 2.6% 5.5% 4.8% 2.4%

Actinobacteria Actinobacteria Other Other Other 1.8% 2.1% 1.3% 1.4% 1.0% 1.1% 2.3% 1.1% 1.4% 1.3% 2.9% 1.7% 5.2% 2.8% 3.0% Syntrophomonadacea Firmicutes Clostridia Clostridiales Syntrophomonas 0.0% 0.0% 0.0% 4.5% 5.2% 4.6% 1.2% 3.2% 2.4% 1.2% 0.6% 2.0% 0.0% 0.1% 0.2% e Acidobacteria Acidobacteria_Gp16 Gp16 Gp16 Gp16 2.3% 3.7% 6.5% 0.9% 1.4% 1.5% 1.4% 0.5% 0.3% 0.4% 2.3% 0.8% 0.6% 0.7% 0.5%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobulbaceae Other 0.0% 0.0% 0.0% 1.1% 0.6% 1.0% 2.7% 2.7% 3.6% 2.4% 2.9% 1.9% 1.8% 1.3% 1.1%

Proteobacteria Other Other Other Other 0.9% 1.5% 1.8% 0.4% 0.5% 0.5% 1.6% 1.0% 0.9% 2.1% 2.4% 1.7% 2.5% 1.6% 2.4%

Proteobacteria Deltaproteobacteria Desulfovibrionales Desulfovibrionaceae Desulfovibrio 0.0% 0.0% 0.1% 4.0% 6.8% 6.0% 0.2% 1.4% 0.9% 0.7% 0.1% 0.3% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Rhodospirillales Other Other 2.2% 3.2% 2.5% 0.6% 0.3% 0.5% 1.1% 0.9% 0.2% 0.4% 0.6% 0.7% 2.6% 1.9% 2.2%

Proteobacteria Deltaproteobacteria Desulfuromonadales Desulfuromonadaceae Desulfuromonas 0.0% 0.0% 0.1% 1.5% 2.5% 7.7% 3.1% 2.8% 0.6% 0.2% 0.1% 0.1% 0.0% 0.1% 0.2%

Nitrospira Nitrospira Nitrospirales Nitrospiraceae Nitrospira 2.6% 1.7% 3.4% 0.3% 0.1% 0.4% 1.0% 0.6% 0.2% 0.1% 0.3% 0.6% 2.7% 2.3% 1.7%

213

Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Longilinea 0.0% 0.0% 0.0% 2.2% 0.6% 1.3% 0.5% 2.8% 3.3% 1.8% 0.3% 1.6% 0.2% 0.3% 1.1%

Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Methylococcus 0.0% 0.0% 0.0% 1.0% 0.2% 0.9% 5.5% 2.2% 0.4% 0.4% 0.1% 0.7% 0.1% 2.3% 0.3%

Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Other 0.0% 0.0% 0.0% 3.1% 1.2% 1.4% 3.7% 2.1% 0.8% 0.1% 0.1% 0.1% 0.0% 1.4% 0.3%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Other 2.2% 2.0% 2.1% 0.3% 0.2% 0.2% 0.5% 0.3% 0.2% 1.0% 0.6% 0.9% 1.7% 1.0% 0.7%

Proteobacteria Deltaproteobacteria Desulfuromonadales Geobacteraceae Geobacter 0.0% 0.0% 0.0% 1.9% 1.7% 3.4% 1.7% 1.6% 0.5% 0.4% 0.2% 0.6% 0.0% 0.9% 0.5%

Chlorobi Ignavibacteria Ignavibacteriales Ignavibacteriaceae Ignavibacterium 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.6% 2.0% 2.3% 0.7% 0.6% 1.1% 0.4% 0.6% 0.5%

Firmicutes Other Other Other Other 0.1% 0.0% 0.1% 1.0% 3.5% 1.0% 0.1% 0.5% 0.5% 0.7% 0.3% 0.3% 0.2% 0.1% 0.1%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Sphingomonas 2.1% 1.6% 1.8% 0.1% 0.1% 0.1% 0.5% 0.1% 0.1% 0.1% 0.1% 0.2% 0.5% 0.6% 0.3%

Actinobacteria Actinobacteria Coriobacteriales Coriobacteriaceae Other 0.0% 0.0% 0.0% 0.9% 1.2% 1.4% 0.5% 0.4% 0.6% 0.9% 0.1% 0.5% 0.0% 0.2% 0.3%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Other 1.5% 1.6% 1.4% 0.4% 0.1% 0.2% 0.5% 0.2% 0.1% 0.2% 0.1% 0.3% 0.3% 0.6% 0.5%

Proteobacteria Alphaproteobacteria Rhizobiales Methylocystaceae Methylocystis 0.0% 0.0% 0.1% 1.1% 0.3% 0.8% 2.4% 1.0% 0.3% 0.1% 0.0% 0.2% 0.1% 0.9% 0.3%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobulbaceae Desulfobulbus 0.0% 0.0% 0.0% 1.2% 1.1% 0.8% 0.4% 0.3% 1.4% 0.6% 0.4% 0.4% 0.0% 0.2% 0.1%

Proteobacteria Deltaproteobacteria Desulfuromonadales Other Other 0.0% 0.0% 0.0% 1.7% 1.6% 1.9% 0.6% 0.5% 0.2% 0.0% 0.0% 0.1% 0.0% 0.6% 0.1%

Proteobacteria Deltaproteobacteria Myxococcales Other Other 0.7% 0.6% 0.5% 0.3% 0.1% 0.3% 0.6% 0.3% 0.8% 0.6% 0.8% 0.4% 0.6% 0.7% 0.4%

Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Methylosarcina 0.1% 0.1% 0.1% 1.9% 0.7% 1.0% 1.6% 1.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.4% 0.2%

Acidobacteria Acidobacteria_Gp1 Gp1 Gp1 Gp1 1.1% 0.9% 1.0% 0.0% 0.0% 0.1% 0.1% 1.6% 0.4% 0.0% 0.1% 0.1% 0.2% 0.2% 0.3%

Acidobacteria Acidobacteria_Gp2 Gp2 Gp2 Gp2 1.5% 1.9% 1.7% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0%

Firmicutes Clostridia Clostridiales Clostridiaceae 1 Other 0.0% 0.0% 0.0% 1.5% 1.5% 0.8% 0.0% 0.1% 0.2% 0.5% 0.3% 0.1% 0.4% 0.1% 0.1%

Proteobacteria Betaproteobacteria Hydrogenophilales Hydrogenophilaceae Thiobacillus 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 1.1% 0.9% 2.7% 0.1% 0.3% 0.3% 0.2% 0.3% 0.2%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobacteraceae Other 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.5% 0.6% 0.6% 1.1% 0.3% 1.1% 0.2% 1.4% 0.2%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Pseudomonas 0.0% 0.0% 0.0% 1.6% 0.6% 0.7% 0.8% 0.8% 0.4% 0.3% 0.0% 0.4% 0.0% 0.2% 0.4%

TM7 genera TM7 genera incertae TM7 genera incertae TM7 genera TM7 0.0% 0.0% 0.0% 0.5% 0.4% 0.5% 0.4% 0.2% 0.0% 0.2% 0.5% 0.2% 1.1% 0.4% 1.1% incertae_sedis sedis sedis incertae sedis

Acidobacteria Acidobacteria_Gp12 Gp12 Gp12 Gp12 1.3% 1.4% 1.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.1% 0.1%

Acidobacteria Acidobacteria_Gp7 Gp7 Gp7 Gp7 1.3% 1.1% 0.6% 0.1% 0.1% 0.1% 0.5% 0.2% 0.3% 0.1% 0.1% 0.3% 0.0% 0.4% 0.1%

Actinobacteria Actinobacteria Acidimicrobiales Other Other 0.2% 0.1% 0.1% 0.4% 0.3% 0.4% 0.6% 0.3% 0.2% 0.1% 0.3% 0.5% 0.8% 0.4% 0.4% Bacteroidetes_incertae_s Bacteroidetes Ohtaekwangia Ohtaekwangia Ohtaekwangia 0.4% 0.4% 0.3% 0.1% 0.0% 0.1% 0.5% 0.6% 0.5% 0.0% 0.1% 0.1% 0.6% 0.3% 0.4% edis Gemmatimonadete Gemmatimonadetes Gemmatimonadales Gemmatimonadaceae Gemmatimonas 0.4% 0.5% 0.4% 0.2% 0.1% 0.2% 0.2% 0.1% 0.2% 0.0% 0.1% 0.2% 0.6% 0.3% 0.4% s

214

Proteobacteria Alphaproteobacteria Rhodospirillales Acetobacteraceae Other 0.5% 0.7% 0.6% 0.1% 0.1% 0.1% 0.3% 0.1% 0.1% 0.1% 0.1% 0.1% 0.6% 0.4% 0.3%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophaceae Desulfomonile 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.8% 0.6% 0.5% 0.3% 0.3% 0.6% 0.4% 0.5%

Proteobacteria Gammaproteobacteria Legionellales Coxiellaceae Aquicella 0.8% 0.9% 0.6% 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.1% 0.2% 0.1% 0.5% 0.1% 0.2%

Acidobacteria Acidobacteria_Gp13 Gp13 Gp13 Gp13 0.5% 0.4% 0.3% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.6% 0.3% 0.3%

Acidobacteria Acidobacteria_Gp17 Gp17 Gp17 Gp17 0.5% 0.5% 0.5% 0.0% 0.0% 0.1% 0.4% 0.2% 0.1% 0.1% 0.1% 0.2% 0.0% 0.2% 0.1%

Actinobacteria Actinobacteria Actinomycetales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.4% 0.3% 0.1% 0.2% 0.2% 0.2% 0.2% 0.3% 0.4%

Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Bellilinea 0.2% 0.0% 0.2% 0.0% 0.0% 0.0% 0.1% 0.1% 0.3% 0.3% 0.0% 0.4% 0.1% 0.3% 0.1%

Firmicutes Clostridia Clostridiales Gracilibacteraceae Other 0.0% 0.0% 0.0% 0.2% 0.4% 0.2% 0.0% 0.4% 0.2% 0.2% 0.0% 0.1% 0.2% 0.2% 0.3% Desulfitobacteriu Firmicutes Clostridia Clostridiales Peptococcaceae 1 0.0% 0.0% 0.0% 0.7% 0.5% 0.3% 0.0% 0.1% 0.1% 0.1% 0.1% 0.3% 0.2% 0.1% 0.1% m Firmicutes Clostridia Clostridiales Ruminococcaceae Other 0.0% 0.0% 0.0% 0.4% 0.5% 0.4% 0.2% 0.3% 0.2% 0.3% 0.1% 0.2% 0.0% 0.1% 0.1% Syntrophomonadacea Firmicutes Clostridia Clostridiales Other 0.0% 0.0% 0.0% 0.2% 0.6% 0.4% 0.1% 0.1% 0.6% 0.1% 0.3% 0.2% 0.0% 0.0% 0.0% e

Thermoanaerobacterale Thermoanaerobactera Firmicutes Clostridia Other 0.0% 0.0% 0.0% 0.6% 1.5% 0.4% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% s ceae

Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Other 0.1% 0.0% 0.0% 0.1% 0.1% 0.5% 1.1% 0.5% 0.3% 0.1% 0.0% 0.3% 0.0% 0.2% 0.2%

Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Other 0.1% 0.0% 0.0% 0.7% 0.3% 0.7% 0.1% 0.3% 0.2% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1%

Proteobacteria Betaproteobacteria Burkholderiales Other Other 0.0% 0.1% 0.1% 0.2% 0.1% 0.2% 0.4% 0.2% 0.1% 0.0% 0.0% 0.2% 0.0% 0.7% 0.1%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobulbaceae Desulfocapsa 0.0% 0.0% 0.0% 1.8% 0.5% 0.2% 0.1% 0.1% 0.1% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfuromonadales Desulfuromonadaceae Other 0.0% 0.0% 0.0% 0.8% 0.7% 0.7% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Kofleriaceae Kofleria 0.1% 0.2% 0.2% 0.1% 0.1% 0.1% 0.2% 0.1% 0.1% 0.3% 0.9% 0.3% 0.1% 0.1% 0.1%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophaceae Other 0.0% 0.0% 0.0% 0.2% 0.1% 0.1% 0.2% 0.3% 0.5% 0.5% 0.2% 0.3% 0.0% 0.1% 0.1%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophobacteraceae Other 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.4% 0.2% 0.1% 0.1% 0.1% 0.3% 0.1% 1.0% 0.3%

Proteobacteria Deltaproteobacteria Syntrophorhabdaceae Syntrophorhabdaceae Syntrophorhabdus 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.1% 0.1% 0.3% 0.9% 0.4% 0.4% 0.0% 0.1% 0.0% Gammaproteobacteria Proteobacteria Gammaproteobacteria Other Other 0.6% 0.4% 0.4% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.1% 0.1% 0.1% 0.2% 0.4% 0.3% ncertae sedis Acidobacteria Acidobacteria_Gp1 Other Other Other 0.4% 0.4% 0.4% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp18 Gp18 Gp18 Gp18 0.3% 0.3% 0.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp22 Gp22 Gp22 Gp22 0.4% 0.4% 0.5% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.1% 0.0%

Acidobacteria Acidobacteria_Gp23 Gp23 Gp23 Gp23 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.7% 0.1% 0.1% 0.1% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp4 Gp4 Gp4 Gp4 0.2% 0.2% 0.3% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.0% 0.1% 0.0% 0.2% 0.3%

215

Acidobacteria Holophagae Holophagales Holophagaceae Geothrix 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% 0.1%

Acidobacteria Holophagae Holophagales Holophagaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.4% 0.1% 0.0% 0.1% 0.0% 0.1% 0.0%

Acidobacteria Other Other Other Other 0.1% 0.6% 0.5% 0.1% 0.0% 0.1% 0.1% 0.0% 0.0% 0.1% 0.3% 0.1% 0.1% 0.1% 0.0%

Actinobacteria Actinobacteria Acidimicrobiales Acidimicrobiaceae Ilumatobacter 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.2% 0.1% 0.1% 0.0% 0.0% 0.1% 0.3% 0.1% 0.8% Acidimicrobineae Actinobacteria Actinobacteria Acidimicrobiales Aciditerrimonas 0.2% 0.5% 0.3% 0.0% 0.0% 0.0% 0.1% 0.2% 0.1% 0.0% 0.0% 0.2% 0.1% 0.2% 0.1% incertae sedis Actinobacteria Actinobacteria Acidimicrobiales Iamiaceae Iamia 0.1% 0.1% 0.1% 0.0% 0.2% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Chitinophagaceae Other 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.5% 0.2% 0.0% 0.0% 0.1% 0.1% 0.1% 0.7% 0.1%

Bacteroidetes Sphingobacteria Sphingobacteriales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.3% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1%

Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Anaerolinea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.0% 0.2% 0.1% 0.1% 0.1%

Chloroflexi Other Other Other Other 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.4% 0.2% 0.4% 0.1% 0.1% 0.1% Cyanobacteria Chloroplast Chloroplast Chloroplast Bacillariophyta 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.3% 0.0% 0.0% 0.0% 0.3% 0.0% 0.0% 0.3% /Chloroplast Firmicutes Bacilli Bacillales Bacillaceae 1 Bacillus 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.2% 0.8%

Firmicutes Clostridia Clostridiales Peptococcaceae 1 Desulfosporosinus 0.0% 0.0% 0.0% 0.3% 0.2% 0.2% 0.0% 0.2% 0.2% 0.2% 0.1% 0.1% 0.1% 0.1% 0.1%

Firmicutes Clostridia Clostridiales Peptococcaceae 1 Other 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.1% 0.3% 0.3% 0.0% 0.1% 0.1% 0.1% 0.3% Desulfotomaculu Firmicutes Clostridia Clostridiales Peptococcaceae 2 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.0% 0.1% 0.1% 0.1% 0.0% 0.0% 0.1% 0.1% 0.1% m Firmicutes Clostridia Clostridiales Peptococcaceae 2 Pelotomaculum 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.1% 0.1% 0.1% 0.0% 0.1% 0.1% 0.0% 0.1% 0.0%

Firmicutes Clostridia Clostridiales Ruminococcaceae Clostridium III 0.0% 0.0% 0.0% 0.2% 0.2% 0.2% 0.1% 0.2% 0.1% 0.1% 0.1% 0.1% 0.1% 0.2% 0.1%

Firmicutes Clostridia Clostridiales Ruminococcaceae Ethanoligenens 0.0% 0.0% 0.0% 0.2% 0.1% 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0%

Firmicutes Clostridia Other Other Other 0.0% 0.0% 0.0% 0.1% 0.3% 0.1% 0.1% 0.1% 0.1% 0.7% 0.1% 0.1% 0.2% 0.0% 0.1%

Firmicutes Negativicutes Selenomonadales Veillonellaceae Sporomusa 0.0% 0.0% 0.0% 0.2% 1.2% 0.3% 0.0% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Bradyrhizobiaceae Bradyrhizobium 0.2% 0.2% 0.1% 0.1% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Hyphomicrobium 0.2% 0.1% 0.2% 0.2% 0.0% 0.1% 0.3% 0.2% 0.1% 0.2% 0.1% 0.1% 0.1% 0.2% 0.1%

Proteobacteria Alphaproteobacteria Rhizobiales Phyllobacteriaceae Other 0.0% 0.0% 0.0% 0.1% 0.0% 0.1% 0.4% 0.5% 0.1% 0.0% 0.0% 0.1% 0.0% 0.1% 0.1%

Proteobacteria Alphaproteobacteria Rhizobiales Rhizobiaceae Rhizobium 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.3% 0.1% 0.3% 0.2% 0.2% 0.1% 0.0% 0.2% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Xanthobacteraceae Xanthobacter 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.2% 0.4% 0.2% 0.2% 0.3% 0.0% 0.1% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Defluviicoccus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.3% 0.0% 1.2% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Dongia 0.1% 0.1% 0.1% 0.1% 0.0% 0.1% 0.3% 0.2% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1%

216

Proteobacteria Alphaproteobacteria Sphingomonadales Erythrobacteraceae Porphyrobacter 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.5%

Proteobacteria Alphaproteobacteria Sphingomonadales Other Other 0.4% 0.3% 0.2% 0.0% 0.1% 0.0% 0.1% 0.0% 0.0% 0.1% 0.1% 0.0% 0.2% 0.3% 0.3%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Novosphingobium 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.1% 0.2% 0.1% 0.0% 0.0% 0.1% 0.0% 0.1% 0.1%

Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Ramlibacter 0.0% 0.0% 0.0% 0.3% 0.3% 0.4% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Oxalobacteraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 1.4%

Proteobacteria Betaproteobacteria Hydrogenophilales Hydrogenophilaceae Sulfuricella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.5% 0.0% 0.0% 0.1% 0.2% 0.2% 0.2%

Proteobacteria Betaproteobacteria Methylophilales Methylophilaceae Other 0.0% 0.0% 0.0% 0.4% 0.1% 0.1% 0.2% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Azospira 0.0% 0.0% 0.0% 0.4% 0.2% 0.3% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Sulfuritalea 0.1% 0.0% 0.0% 0.1% 0.0% 0.1% 0.3% 0.1% 0.1% 0.1% 0.0% 0.1% 0.2% 0.1% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Thauera 0.0% 0.0% 0.0% 0.1% 0.1% 0.5% 0.2% 0.1% 0.1% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfobacterales Other Other 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.1% 0.1% 0.6% 0.0% 0.0% 0.2% 0.0% 0.1% 0.1%

Proteobacteria Deltaproteobacteria Desulfovibrionales Desulfovibrionaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.2% 0.1% 0.4% 0.1% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Cystobacteraceae Other 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.2% 0.1% 0.1% 0.1% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Polyangiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.3% 0.0% 0.0% 0.1% 0.0% 0.1% 0.0% 0.2% 0.0%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophobacteraceae Desulfovirga 0.0% 0.0% 0.0% 0.3% 0.1% 0.2% 0.3% 0.3% 0.2% 0.1% 0.0% 0.1% 0.3% 0.2% 0.2%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophobacteraceae Syntrophobacter 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.2% 0.3% 0.1% 0.2% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Legionellales Legionellaceae Legionella 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1% 0.1% 0.3% 0.3% 0.2% 0.0% 0.2% 0.1%

Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Methylomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.9% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Azotobacter 0.0% 0.0% 0.0% 0.2% 0.1% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.2% 0.0% 0.1% 0.1%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 0.0% 0.0% 0.0% 0.3% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Sinobacteraceae Steroidobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.2% 0.1% 0.0% 0.2% 0.1%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Other 0.2% 0.2% 0.2% 0.0% 0.0% 0.0% 0.4% 0.1% 0.7% 0.0% 0.1% 0.1% 0.0% 0.1% 0.1%

Acidobacteria Acidobacteria_Gp1 Edaphobacter Edaphobacter Edaphobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp1 Terriglobus Terriglobus Terriglobus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp10 Gp10 Gp10 Gp10 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp11 Gp11 Gp11 Gp11 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp15 Gp15 Gp15 Gp15 0.2% 0.3% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp20 Gp20 Gp20 Gp20 0.2% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

217

Acidobacteria Acidobacteria_Gp3 Bryobacter Bryobacter Bryobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp3 Other Other Other 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp5 Gp5 Gp5 Gp5 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Acidobacteria_Gp9 Gp9 Gp9 Gp9 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Acidobacteria Holophagae Holophagales Holophagaceae Holophaga 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Acidimicrobiales Acidimicrobiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Cellulomonadaceae Cellulomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Corynebacteriaceae Corynebacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Geodermatophilaceae Blastococcus 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Geodermatophilaceae Modestobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Geodermatophilaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Intrasporangiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Intrasporangiaceae Terrabacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Kineosporiaceae Kineosporia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Microbacteriaceae Agromyces 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Microbacteriaceae Leifsonia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Microbacteriaceae Microcella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Microbacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Micrococcaceae Arthrobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2%

Actinobacteria Actinobacteria Actinomycetales Micrococcaceae Nesterenkonia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Micrococcaceae Rothia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Micromonosporaceae Actinoplanes 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Micromonosporaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Mycobacteriaceae Mycobacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1%

Actinobacteria Actinobacteria Actinomycetales Nakamurellaceae Nakamurella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Nocardioidaceae Marmoricola 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Nocardioidaceae Nocardioides 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Actinobacteria Actinobacteria Actinomycetales Nocardioidaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

218

Actinomycetospor Actinobacteria Actinobacteria Actinomycetales Pseudonocardiaceae 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% a Actinobacteria Actinobacteria Actinomycetales Pseudonocardiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Pseudonocardiaceae Pseudonocardia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Actinomycetales Sporichthyaceae Sporichthya 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Thermomonosporacea Actinobacteria Actinobacteria Actinomycetales Actinomadura 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% e Actinobacteria Actinobacteria Bifidobacteriales Bifidobacteriaceae Gardnerella 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Coriobacteriales Coriobacteriaceae Olsenella 0.0% 0.0% 0.0% 0.0% 0.1% 0.2% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Euzebyales Euzebyaceae Euzebya 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Nitriliruptorales Nitriliruptoraceae Nitriliruptor 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Rubrobacterales Rubrobacteraceae Rubrobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Solirubrobacterales Conexibacteraceae Conexibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Solirubrobacterales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Actinobacteria Actinobacteria Solirubrobacterales Solirubrobacteraceae Solirubrobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Armatimonadetes Armatimonadetes Armatimonadetes gp2 Armatimonadetes gp2 Armatimonadetes gp2 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% gp2 Armatimonadetes_gp Armatimonadetes Armatimonadetes Armatimonadetes gp4 Armatimonadetes gp4 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4 gp4 Armatimonadetes Armatimonadetes Armatimonadetes gp5 Armatimonadetes gp5 Armatimonadetes gp5 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% gp5 Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Arenibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Cloacibacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Flavobacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Bacteroidetes Flavobacteria Flavobacteriales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Chitinophagaceae Flavisolibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Chitinophagaceae Lacibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Chitinophagaceae Segetibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Chitinophagaceae Terrimonas 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cyclobacteriaceae Aquiflexum 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cyclobacteriaceae Belliella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cyclobacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

219

Bacteroidetes Sphingobacteria Sphingobacteriales Cytophagaceae Fibrella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cytophagaceae Hymenobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cytophagaceae Litoribacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cytophagaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Cytophagaceae Spirosoma 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Flammeovirgaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Flammeovirgaceae Roseivirga 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Saprospiraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Sphingobacteriaceae Mucilaginibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Bacteroidetes Sphingobacteria Sphingobacteriales Sphingobacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Chlamydiae Chlamydiae Chlamydiales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Chlamydiae Chlamydiae Chlamydiales Parachlamydiaceae Neochlamydia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0%

Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Leptolinea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae Levilinea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Chloroflexi Caldilineae Caldilineales Caldilineaceae Caldilinea 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0%

Chloroflexi Dehalococcoidetes Dehalogenimonas Dehalogenimonas Dehalogenimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Chloroflexi Thermomicrobia Sphaerobacterales Sphaerobacteraceae Sphaerobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Cyanobacteria Chloroplast Chloroplast Chloroplast Chlorophyta 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% /Chloroplast Cyanobacteria Chloroplast Chloroplast Chloroplast Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.1% 0.0% /Chloroplast Cyanobacteria Cyanobacteria Other Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% /Chloroplast Elusimicrobia Elusimicrobia Elusimicrobiales Elusimicrobiaceae Elusimicrobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Fibrobacteres Fibrobacteria Fibrobacterales Fibrobacteraceae Fibrobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Alicyclobacillaceae Alicyclobacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Alicyclobacillaceae Tumebacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Bacillaceae 2 Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Bacillaceae 2 Vulcanibacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Firmicutes Bacilli Bacillales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5%

220

Firmicutes Bacilli Bacillales Paenibacillaceae 1 Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Paenibacillaceae 1 Paenibacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Planococcaceae Lysinibacillus 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Planococcaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Bacillales Planococcaceae Sporosarcina 0.1% 0.0% 0.0% 0.4% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sporolactobacillac Firmicutes Bacilli Bacillales Sporolactobacillaceae 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% eae incertae sedis Thermoactinomycetac Firmicutes Bacilli Bacillales Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% eae 1 Firmicutes Bacilli Lactobacillales Streptococcaceae Streptococcus 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Bacilli Other Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Clostridiaceae 1 Anaerobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Clostridiaceae 1 Oxobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Clostridiaceae 1 Sarcina 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Clostridiales Incertae Firmicutes Clostridia Clostridiales Finegoldia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sedis XI Clostridiales Incertae Firmicutes Clostridia Clostridiales Parvimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sedis XI Clostridiales Incertae Firmicutes Clostridia Clostridiales Tissierella 0.1% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sedis XI Clostridiales Incertae Firmicutes Clostridia Clostridiales Anaerovorax 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sedis XIII Clostridiales Incertae Firmicutes Clostridia Clostridiales Sulfobacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Sedis XVII Firmicutes Clostridia Clostridiales Eubacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Gracilibacteraceae Gracilibacter 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Incertae Sedis XI Clostridium XII 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Lachnospiraceae Clostridium XlVa 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Lachnospiraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Peptococcaceae 2 Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Peptostreptococcacea Firmicutes Clostridia Clostridiales Clostridium XI 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% e Peptostreptococcacea Firmicutes Clostridia Clostridiales Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% e Firmicutes Clostridia Clostridiales Ruminococcaceae Acetivibrio 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Clostridia Clostridiales Ruminococcaceae Oscillibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

221

Saccharofermenta Firmicutes Clostridia Clostridiales Ruminococcaceae 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% ns Syntrophomonadacea Firmicutes Clostridia Clostridiales Pelospora 0.0% 0.0% 0.0% 0.1% 0.2% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% e Firmicutes Negativicutes Selenomonadales Other Other 0.0% 0.0% 0.0% 0.1% 0.4% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Negativicutes Selenomonadales Veillonellaceae Desulfosporomusa 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Negativicutes Selenomonadales Veillonellaceae Propionispora 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Firmicutes Negativicutes Selenomonadales Veillonellaceae Psychrosinus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Alphaproteobacteria Proteobacteria Alphaproteobacteria Elioraea Elioraea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% incertae sedis Alphaproteobacteria Proteobacteria Alphaproteobacteria Geminicoccus Geminicoccus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Alphaproteobacteria Proteobacteria Alphaproteobacteria Rhizomicrobium Rhizomicrobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% ncertae sedis Proteobacteria Alphaproteobacteria Caulobacterales Caulobacteraceae Asticcacaulis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Caulobacterales Caulobacteraceae Brevundimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Caulobacterales Caulobacteraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Caulobacterales Caulobacteraceae Phenylobacterium 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Hyphomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Beijerinckiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Bradyrhizobiaceae Agromonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Bradyrhizobiaceae Bosea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Bradyrhizobiaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Blastochloris 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Devosia 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Pedomicrobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Rhodoplanes 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Methylobacteriaceae Methylobacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Rhizobiales Methylobacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Methylocystaceae Methylosinus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Methylocystaceae Other 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0%

222

Proteobacteria Alphaproteobacteria Rhizobiales Methylocystaceae Pleomorphomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Phyllobacteriaceae Mesorhizobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Rhizobiaceae Ensifer 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Rhizobiaceae Kaistia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Rhizobiales ncertae Proteobacteria Alphaproteobacteria Rhizobiales Bauldia 0.1% 0.2% 0.1% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% sedis Rhizobiales ncertae Proteobacteria Alphaproteobacteria Rhizobiales Vasilyevaea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% sedis Proteobacteria Alphaproteobacteria Rhizobiales Rhodobiaceae Parvibaculum 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Xanthobacteraceae Azorhizobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhizobiales Xanthobacteraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Amaricoccus 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Paracoccus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Rhodobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2%

Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Rubellimicrobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Acetobacteraceae Acidisoma 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1%

Proteobacteria Alphaproteobacteria Rhodospirillales Acetobacteraceae Roseomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.1% 0.1% 0.0% 0.1% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Rhodospirillales Acetobacteraceae Stella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Azospirillum 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Inquilinus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Rhodospirillales Rhodospirillaceae Magnetospirillum 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Sphingomonadales Erythrobacteraceae Erythrobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Sphingomonadales Erythrobacteraceae Other 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Blastomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Other 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Sphingobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Sphingopyxis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Sphingosinicella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Alcaligenaceae Azohydromonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Alcaligenaceae Derxia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

223

Proteobacteria Betaproteobacteria Burkholderiales Burkholderiaceae Burkholderia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Burkholderiaceae Chitinimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Burkholderiaceae Cupriavidus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Burkholderiaceae Pandoraea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Burkholderiaceae Ralstonia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Burkholderiales Proteobacteria Betaproteobacteria Burkholderiales Aquabacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Burkholderiales Proteobacteria Betaproteobacteria Burkholderiales Inhella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Burkholderiales Proteobacteria Betaproteobacteria Burkholderiales Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Acidovorax 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Curvibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Hydrogenophaga 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Comamonadaceae Polaromonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Oxalobacteraceae Naxibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Burkholderiales Oxalobacteraceae Oxalicibacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Methylophilales Methylophilaceae Methylobacillus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Neisseriales Neisseriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Neisseriales Neisseriaceae Uruburuella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Nitrosomonadales Nitrosomonadaceae Nitrosospira 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Azoarcus 0.0% 0.0% 0.0% 0.2% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Dechloromonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Georgfuchsia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Methyloversatilis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Betaproteobacteria Rhodocyclales Rhodocyclaceae Zoogloea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Bdellovibrionales Bacteriovoracaceae Bacteriovorax 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.3% 0.1%

Proteobacteria Deltaproteobacteria Bdellovibrionales Bacteriovoracaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Bdellovibrionales Bacteriovoracaceae Peredibacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Bdellovibrionales Bdellovibrionaceae Bdellovibrio 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.0% 0.1% 0.0%

224

Proteobacteria Deltaproteobacteria Bdellovibrionales Bdellovibrionaceae Vampirovibrio 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobacteraceae Desulfobacterium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobacteraceae Desulforegula 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobacteraceae Desulfosalsimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfovibrionales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfuromonadales Desulfuromonadaceae Pelobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Desulfuromonadales Geobacteraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Cystobacteraceae Anaeromyxobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Cystobacteraceae Cystobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Nannocystaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Phaselicystidaceae Phaselicystis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Myxococcales Polyangiaceae Byssovorax 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.1% 0.0%

Proteobacteria Deltaproteobacteria Syntrophobacterales Other Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophaceae Desulfobacca 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophaceae Smithella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophaceae Syntrophus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophobacteraceae Desulforhabdus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Epsilonproteobacteria Campylobacterales Campylobacteraceae Sulfurospirillum 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Epsilonproteobacteria Campylobacterales Helicobacteraceae Sulfuricurvum 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Aeromonadales Aeromonadaceae Zobellella 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Alteromonadales Alteromonadaceae Haliea 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Alteromonadales Alteromonadaceae Marinimicrobium 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Alteromonadales Alteromonadaceae Marinobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Alteromonadales Idiomarinaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Chromatiales Chromatiaceae Rheinheimera 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Enterobacteriales Enterobacteriaceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Gammaproteobacteria Proteobacteria Gammaproteobacteria Alkalimonas Alkalimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Gammaproteobacteria Methylohalomona Proteobacteria Gammaproteobacteria Methylohalomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis s

225

Gammaproteobacteria Proteobacteria Gammaproteobacteria Methylonatrum Methylonatrum 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Gammaproteobacteria Proteobacteria Gammaproteobacteria Solimonas Solimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Gammaproteobacteria Proteobacteria Gammaproteobacteria Thioprofundum Thioprofundum 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% incertae sedis Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Methylobacter 0.0% 0.0% 0.0% 0.3% 0.1% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Methylococcales Methylococcaceae Methylosoma 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Oceanospirillales Halomonadaceae Halomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Pseudomonadales Moraxellaceae Acinetobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Pseudomonadales Moraxellaceae Alkanindiges 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Azomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Azorhizophilus 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Cellvibrio 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Sinobacteraceae Nevskia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Sinobacteraceae Other 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Aquimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Arenimonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Lysobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Rhodanobacter 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Proteobacteria Gammaproteobacteria Xanthomonadales Xanthomonadaceae Stenotrophomonas 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Subdivision3 Subdivision3 Subdivision3 Verrucomicrobia Subdivision3 Genera incertae 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Genera incertae sedis Genera incertae sedis sedis WS3 WS3 WS3 WS3 WS3 Genera incertae 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Genera incertae sedis Genera incertae sedis Genera incertae sedis sedis

226