UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluating microbial community responses to blended biofuel contamination Arnold, Ashley Celine 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2020_november_arnold_ashley.pdf [ 49.1MB ]
Metadata
JSON: 24-1.0392576.json
JSON-LD: 24-1.0392576-ld.json
RDF/XML (Pretty): 24-1.0392576-rdf.xml
RDF/JSON: 24-1.0392576-rdf.json
Turtle: 24-1.0392576-turtle.txt
N-Triples: 24-1.0392576-rdf-ntriples.txt
Original Record: 24-1.0392576-source.json
Full Text
24-1.0392576-fulltext.txt
Citation
24-1.0392576.ris

Full Text

EVALUATING MICROBIAL COMMUNITY RESPONSES TO BLENDED BIOFUEL CONTAMINATION by  Ashley Celine Arnold  B.Sc., The University of British Columbia, 2015  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Microbiology and Immunology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   July 2020  © Ashley Celine Arnold, 2020  ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis entitled:  Evaluating microbial community responses to blended biofuel contamination  submitted by Ashley Celine Arnold in partial fulfillment of the requirements for the degree of Master of Science in Microbiology and Immunology  Examining Committee: Dr. Steven Hallam, Professor, Microbiology and Immunology, UBC Supervisor  Dr. Ulrich Mayer, Professor, Earth Ocean and Atmospheric Sciences, UBC Supervisory Committee Member  Dr. Erin Gaynor, Professor, Microbiology and Immunology, UBC Additional Examiner   iii Abstract  As biofuel production increases, so too has the likelihood of accidental spills into the environment with important implications for human health and ecosystem functions. Such impacts can be evaluated at the microbial level as microorganisms are fundamental units of metabolism integral to the conversion of hydrocarbon substrates in the environment. Using the small subunit ribosomal (SSU or 16S) rRNA gene, we evaluate microbial community responses to ethanol and methanol blended biofuel contamination in terrestrial environments, through laboratory and field experiments, by assessing both community abundance and potential activity. For ethanol-based biofuel contamination, we observe differential patterns of enrichment in microbial taxa with traits relating to hydrocarbon degradation and fermentation across both field and laboratory experiments as well as changes in the abundance and potential activity of canonical methanogens. Observed differences highlight the role of the physical environment and the availability of organic matter in shaping microbial response patterns. Similar results were obtained for the methanol laboratory experiments, with respect hydrocarbon-degrading and fermentative taxa. However, a stronger methanogen response was observed suggesting that methanol blended fuels are more readily converted to methane under low organic loading conditions. Together, this work provides an initial assessment on the impact blended biofuels has on microbial community structure by identifying microbial taxa most responsive to contamination with implications for the development of remediation and risk assessment strategies.     iv Lay Summary  As the use of ethanol and methanol blended biofuels increases, so too does the risk of accidental spills into the environments. These spills can impact ecosystems at the microbial level, which in turn impacts the services microorganisms provide, such as nutrient cycling and decomposition, that supports the health of the system. Through laboratory and field site experiments, we observe the response of microbial communities to ethanol and methanol blended biofuel spills by evaluating microbial abundance and activity.  We observe the dynamics of microbial communities as certain members experience changes in abundance and activity while others appear unaffected. Together, this work highlights the intricacies of microbial communities in response to biofuel contamination and provides an initial characterization that can help inform potential remediation and risk assessment strategies. v Preface This thesis contains original, unpublished work completed by Ashley C. Arnold and collaborators.   Chapter 1: Ashley C. Arnold wrote the original text with feedback from Steven J. Hallam.   Chapter 2: Ashley C. Arnold and Aria S. Hahn wrote the original text and received input and feedback from Natasha J. Sihota and Steven J. Hallam. Sample collection and processing was performed by Natasha J. Sihota and Melanie Sorensen and subsequent data analysis and interpretation was conducted by Ashley C. Arnold and Aria S. Hahn with feedback from Steven J. Hallam and K. Ulrich Mayer.   Chapter 3: Experimental design and sample collection for the field experiment was performed by Natasha J. Sihota and K. Ulrich Mayer. Sample processing was completed by Natasha J. Sihota and Melanie Sorensen. Ashley C. Arnold performed all data analysis and interpretation and wrote the original text with input from Natasha J. Sihota and editing from Steven J. Hallam.   Chapter 4: Experimental design for the laboratory experiments was conducted by Jarod Devries, Natasha J. Sihota and K. Ulrich Mayer. Sample collection was performed by Ashley C. Arnold and Jarod Devries. Sample processing, data analysis and interpretation were completed by Ashley C. Arnold with input from Jarod Devries and editing from Steven J. Hallam.   vi Chapter 5: Experimental design for the laboratory experiments was conducted by Laura Laurenzi, Natasha J. Sihota and K. Ulrich Mayer. Laura Laurenzi performed all sample collection. Sample processing, data analysis and interpretation, and written text was completed by Ashley C. Arnold with input Laura Laurenzi and K. Ulrich Mayer and editing from Steven J. Hallam.   Chapter 6: Ashley C. Arnold wrote the original text with feedback from Steven J. Hallam vii Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ........................................................................................................................ vii List of Tables ............................................................................................................................... xii List of Figures ............................................................................................................................. xiii List of Abbreviations ............................................................................................................... xxiv Acknowledgements ....................................................................................................................xxv Dedication ................................................................................................................................. xxvi Chapter 1: Introduction ................................................................................................................1 1.1 Microbial communities and ecological stability ............................................................. 2 1.2 Microbial responses to anthropogenic disturbances ....................................................... 3 1.3 Microbial community responses to biofuel contamination ............................................. 4 1.4 Thesis overview .............................................................................................................. 6 Chapter 2: 16S rRNA gene sequencing and data analysis .........................................................9 2.1 Origins and application of 16S rRNA gene sequencing ................................................. 9 2.2 16S sequence data normalization .................................................................................. 11 2.3 Methods......................................................................................................................... 12 2.3.1 Sample collection, DNA extraction and cDNA synthesis .................................... 12 2.3.2 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing .......... 14 2.3.3 Sequence analyses ................................................................................................. 15 viii 2.3.4 Data analyses ........................................................................................................ 15 2.1 Results ........................................................................................................................... 16 2.1.1 Microbial community diversity ............................................................................. 16 2.1.2 Soil microbial community structure ...................................................................... 19 2.1.3 Microbial community abundance and potential activity ....................................... 22 2.1.4 Potential activity ................................................................................................... 28 2.2 Discussion ..................................................................................................................... 35 2.2.1 Similar trends in community and diversity and structure with varying normalization technique ........................................................................................................ 35 2.2.2 Discrepancies in abundance and potential activity between normalization techniques ............................................................................................................................. 36 2.2.3 Considerations for VST normalization ................................................................. 38 2.2.3.1 Diversity indices ............................................................................................... 39 2.2.3.2 Community structure ........................................................................................ 39 2.2.3.3 Community composition ................................................................................... 39 2.2.3.4 Indicator species analyses ................................................................................. 40 2.2.3.5 Correlations and co-occurrence networks ......................................................... 40 2.3 Conclusions ................................................................................................................... 40 Chapter 3: Evaluating microbial community responses to denatured fuel grade ethanol at the Cambria, Minnesota spill site ...............................................................................................42 3.1 Introduction ................................................................................................................... 42 3.2 Methods......................................................................................................................... 43 3.2.1 Cambria DFE spill site characteristics .................................................................. 43 ix 3.2.2 Sample collection, nucleic acid extraction and cDNA synthesis .......................... 44 3.2.3 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing .......... 45 3.2.4 Sequence processing ............................................................................................. 46 3.3 Results ........................................................................................................................... 47 3.3.1 Microbial community diversity ............................................................................. 47 3.3.2 Microbial community structure ............................................................................. 48 3.3.3 Microbial community abundance and activity ...................................................... 53 3.3.4 Responsive taxa .................................................................................................... 57 3.4 Discussion ..................................................................................................................... 62 3.5 Conclusion .................................................................................................................... 64 Chapter 4: Evaluating microbial community responses to ethanol-blended biofuels ...........65 4.1 Introduction ................................................................................................................... 65 4.2 Methods......................................................................................................................... 66 4.2.1 Column design and construction ........................................................................... 66 4.2.2 Sample collection and geochemical monitoring ................................................... 67 4.2.3 Nucleic acid extraction and cDNA synthesis ........................................................ 68 4.2.4 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing .......... 68 4.2.5 Sequence processing ............................................................................................. 71 4.3 Results ........................................................................................................................... 72 4.3.1 Geochemistry ........................................................................................................ 72 4.3.2 Microbial community diversity ............................................................................. 78 4.3.3 Microbial community structure ............................................................................. 81 4.3.4 Microbial community abundance and potential activity ....................................... 85 x 4.3.5 Responsive taxa .................................................................................................... 90 4.4 Discussion ................................................................................................................... 100 4.4.1 Microbial community responses to ethanol blended fuels .................................. 100 4.4.2 Column experiments ........................................................................................... 100 4.5 Conclusion .................................................................................................................. 103 Chapter 5: Evaluating microbial community responses to methanol-blended biofuels .....104 5.1 Introduction ................................................................................................................. 104 5.2 Methods....................................................................................................................... 107 5.2.1 Methanol blended fuel experiments .................................................................... 107 5.2.1.1 Column design and construction ..................................................................... 107 5.2.1.2 Sample collection and geochemical monitoring: ............................................ 108 5.2.2 Methanol microcosm experiments ...................................................................... 110 5.2.2.1 Microcosm design ........................................................................................... 110 5.2.2.2 Sample collection and geochemical monitoring ............................................. 111 5.2.3 Nucleic acid extraction and cDNA synthesis ...................................................... 111 5.2.4 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing ........ 112 5.2.5 Sequence processing ........................................................................................... 113 5.3 Results ......................................................................................................................... 114 5.3.1 Methanol column experiments ............................................................................ 114 5.3.1.1 Geochemical results ........................................................................................ 114 5.3.1.2 Microbial community diversity ....................................................................... 118 5.3.1.3 Microbial community structure ....................................................................... 121 5.3.1.4 Microbial community abundance and activity ................................................ 123 xi 5.3.1.5 Responsive taxa .............................................................................................. 128 5.3.2 Methanol microcosm experiments ...................................................................... 143 5.3.2.1 Geochemistry .................................................................................................. 143 5.3.2.2 Microbial community diversity ....................................................................... 145 5.3.2.3 Microbial community structure and indicator species analysis ...................... 148 5.3.2.4 Microbial community abundance and potential activity ................................. 149 5.3.2.5 Responsive taxa .............................................................................................. 151 5.4 Discussion ................................................................................................................... 155 5.4.1 Column experiments ........................................................................................... 155 5.4.2 Microcosm experiments ...................................................................................... 158 5.4.3 Comparing methanol column and microcosm experiments ................................ 159 5.5 Conclusion .................................................................................................................. 160 Chapter 6: Conclusion ...............................................................................................................162 6.1 Microbial ecology and the use of the 16S rRNA gene ............................................... 162 6.2 Comparing field experiments and column experiments ............................................. 163 6.3 Future directions ......................................................................................................... 167 6.4 Closing ........................................................................................................................ 168 Bibliography ...............................................................................................................................169 Appendix A: Chapter 5 supplemental .....................................................................................177     A.1      Microcosm indicator species analysis table ................................................................ 177  xii List of Tables  Table 3.1 Sampling summary for the Cambria DFE spill site ...................................................... 45 Table 4.1 Sampling summary for the ethanol blended fuel column experiment .......................... 69 Table 4.2 Treatment and sampling summary for the ethanol blended fuel column experiment .. 70 Table 5.1 Sampling summary for methanol blended fuel column and methanol microcosm experiments ................................................................................................................................. 109 Table 5.2 Sampling tally for the methanol blended fuel column and methanol microcosm experiments ................................................................................................................................. 110 Table A.1 Indicator species analysis summary table for the methanol microcosm experiments 177  xiii List of Figures  Figure 2.1 Cambria denatured fuel-grade ethanol (DFE) spill site. A) Spill site schematic indicating the location of soil sampling cores A and B. B) Sampling plan and ethanol concentrations along the soil profile for core B. ........................................................................... 13 Figure 2.2 Rarefaction curves for sequences generated from Core A and Core B from the Cambria denatured fuel grade ethanol spill site. ........................................................................... 17 Figure 2.3 A) Variance estimates using the negative binomial model when each point represents an OTU B) Venn-diagram of the OTUs shared between rDNA and rRNA fractions within and between Cambria DFE soil Core A and Core B ........................................................................... 18 Figure 2.4 Principle component analysis (PCA) plot for Core A and Core B Cambria DFE spill site rDNA and rRNA samples using proportion-normalized data (in pink) and VST-normalized data (in violet) ............................................................................................................................... 20 Figure 2.5 NMDS plot from Core A and Core B Cambria DFE spill site rDNA and rRNA samples using proportion-normalized data (in pink) and VST-normalized data (in violet). ........ 21 Figure 2.6 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core A using proportion-normalized data ................................................................................................. 24 Figure 2.7 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core B using proportion-normalized data ................................................................................................. 25 Figure 2.8 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core A using VST-normalized data .......................................................................................................... 26 Figure 2.9 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core B using VST-normalized data .......................................................................................................... 27 xiv Figure 2.10 Phylum level rRNA:rDNA ratio taxonomic summary for Cambria DFE soil Core A using VST-normalized data .......................................................................................................... 29 Figure 2.11 Phylum level rRNA:rDNA ratio taxonomic summary for Cambria DFE soil Core B using VST-normalized data .......................................................................................................... 30 Figure 2.12 Core A rRNA:rDNA heatmap for the most variable orders identified using proportion-normalized (in pink) and VST-normalized (in violet) data. Panel A) depicts taxonomic orders identified as being the most variable (top 10%) in both proportion-normalized and VST-normalized datasets. Panel B) depicts orders identified in the proportion-normalized dataset only, and Panel C) depicts orders identified in the VST-normalized dataset only. .......... 32 Figure 2.13 Core B rRNA:rDNA heatmap for the most variable orders identified using proportion-normalized (in pink) and VST-normalized (in violet) data. Panel A) depicts taxonomic orders identified as being the most variable (top 10%) in both proportion-normalized and VST-normalized datasets. Panel B) depicts orders identified in the proportion-normalized dataset only, and Panel C) depicts orders identified in the VST-normalized dataset only. .......... 34 Figure 3.1 Cambria denatured fuel grade ethanol spill site. Black dots indicate the sampling location of soil core A, B, C, D and E. ......................................................................................... 44 Figure 3.2 Rarefaction curve for sequences generated from the Cambria denatured fuel grade ethanol spill site ............................................................................................................................ 49 Figure 3.3 Diversity estimates for the Cambria  denatured fuel grade ethanol spill site. A) Shannon diversity index across each sampling cores B)Venn diagram comparing the number of OTUs shared between each sampling core. .................................................................................. 50 Figure 3.4 NMDS ordination plot for rDNA and rRNA samples generated from the Cambria denatured fuel grade ethanol spill site. ......................................................................................... 51 xv Figure 3.5 NMDS ordination plot for samples collected from the Cambria denatured fuel grade ethanol spill site. Plots have been separated by sampling core and colour-coded by zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm. ........................................ 52 Figure 3.6 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site .................................................................................................................. 54 Figure 3.7 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site comparing the reference site to the contamination source site ................ 55 Figure 3.8 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site within each sample depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core. ................................................................... 56 Figure 3.9 rDNA and rRNA taxonomic summary for taxa within the Betaproteobacteria sub-phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core. ....................................................................................... 58 Figure 3.10 rDNA and rRNA taxonomic summary for taxa within the Deltaproteobacteria sub-phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core. ....................................................................................... 59 Figure 3.11 rDNA and rRNA taxonomic summary for taxa within the Firmicutes phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core .................................................................................................................... 60 Figure 3.12 rDNA and rRNA taxonomic summary for taxa within the Euryarchaeota phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core ........................................................................................................ 61 xvi Figure 4.1 Schematic diagram of the column design used in the ethanol blended fuel experiments....................................................................................................................................................... 67 Figure 4.2 O2  gas concentration measurements for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline. ............................................................ 74 Figure 4.3 CO2 gas concentration measurements for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline .................................................... 75 Figure 4.4 CH4  gas concentration measurement for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline .................................................... 76 Figure 4.5 Volatile fatty acid concentrations (ppm) for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Delayed E20,  EtOH4 – Gasoline, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline .................................................... 77 Figure 4.6 Rarefaction curves for sequences generated from the ethanol blended fuel column experiments. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Delayed E20,  EtOH4 – Gasoline, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline. .............. 79 Figure 4.7 Diversity estimates for the ethanol column experiments. A) Shannon diversity index comparing reference and contaminated samples across all ethanol column treatments. B) Shannon diversity index comparing reference and contaminated samples from within each column treatment, C) Venn diagram indicating the number of OTUs shared both within and between the sand and silt material ethanol column treatments. .................................................... 80 xvii Figure 4.8 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments ................................................................................................................................... 82 Figure 4.9 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments. Plots have been separated by column treatment and colour-coded by sampling timepoint. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline .................. 83 Figure 4.10 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments. Plots have been separated by column treatment and colour-coded by sampling depth. EtOH1 - E85, EtOH2 - E20, EtOH3 - Delayed E20, EtOH4 - Gasoline, EtOH5 - E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline. ......................... 84 Figure 4.11 Phylum level rDNA and rRNA taxonomic summary for the ethanol blended fuel column experiments. ..................................................................................................................... 86 Figure 4.12 Phylum level rDNA and rRNA taxonomic summary across all sampling depths for each ethanol blended fuel column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline..................................................................................................................... 87 Figure 4.13 Phylum level rDNA and rRNA taxonomic summary across all timepoints for each ethanol blended fuel column treatment within the saturated zone (35cm) and vadose zone (75cm, 90cm, 145cm). EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline ..................... 88 Figure 4.14 Phylum level rDNA and rRNA taxonomic summary comparing reference and treatment samples for the ethanol blended fuel column experiments. .......................................... 89 xviii Figure 4.15 rDNA and rRNA taxonomic summary comparing Gammaproteobacteria orders from pristine and contaminated samples collected from the ethanol blended fuel column experiments........................................................................................................................................................ 91 Figure 4.16 rDNA and rRNA taxonomic summary for taxa within the Gammaproteobacteria sub-phylum  across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline. ...................................................................................... 92 Figure 4.17 rDNA and rRNA taxonomic summary for taxa within the Gammaproteobacteria sub-phylum across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline ..................................................................................................... 93 Figure 4.18 rDNA and rRNA taxonomic summary comparing Firmicutes orders from pristine and contaminated samples collected from the ethanol blended fuel column experiments. .......... 94 Figure 4.19 rDNA and rRNA taxonomic summary of taxa within the Firmicutes phylum across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline. ........................................................................................................... 95 Figure 4.20 rDNA and rRNA taxonomic summary of taxa within the Firmicutes phylum across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline ..................................................................................................................................... 95 xix Figure 4.21 rDNA and rRNA taxonomic summary comparing Archaeal taxa from pristine and contaminated samples collected from the ethanol blended fuel column experiments. ................. 97 Figure 4.22  rDNA and rRNA taxonomic summary of Archaeal taxa  across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline. ....................................................................................................................................... 98 Figure 4.23 rDNA and rRNA taxonomic summary of Archaeal taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline. ........ 99 Figure 5.1 Schematic diagram of the column design used in the methanol blended fuel experiments ................................................................................................................................. 108 Figure 5.2 O2  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. ............................................................................ 115 Figure 5.3 CO2  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. ............................................................................ 116 Figure 5.4 CH4  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. ............................................................................ 117 Figure 5.5 Volatile Fatty Acid concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. ............................................................... 118 Figure 5.6 Rarefaction curves generated for sequences from the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 – M85. ........................................................................... 119 xx Figure 5.7 Diversity estimates for the methanol column experiments. A) Shannon diversity index comparing reference and contaminated samples across each methanol column treatment. B) Shannon diversity index comparing reference and contaminated samples from within each column treatment. C) Venn diagram indicating the number of OTUs shared both within and between the two column treatments. ........................................................................................... 120 Figure 5.8 NMDS ordination plot for samples collected from the methanol blended fuel column experiments. ................................................................................................................................ 122 Figure 5.9 NMDS ordination plot for samples collected from the methanol blended fuel column experiments. Plots have been separated by methanol column treatment and colour-coded by sampling depth.  MeOH1 – M15, MeOH2 – M85. ..................................................................... 123 Figure 5.10 Phylum level rDNA and rRNA taxonomic summary for the methanol blended fuel column experiments .................................................................................................................... 124 Figure 5.11 Phylum level rDNA and rRNA taxonomic summary for the methanol blended fuel column experiments comparing reference/pristine and contaminated samples .......................... 125 Figure 5.12 Phylum level rDNA and rRNA taxonomic summary across each sampling depths for the methanol blended fuels column experiment. MeOH1-M15, MeOH2 – M85 ....................... 126 Figure 5.13 Phylum level rDNA and rRNA taxonomic summary across all timepoints for each methanol blended fuel column treatment within the saturated zone (35cm) and vadose zone (75cm, 90cm, 145cm). MeOH1 – M15, MeOH2 – M85. ........................................................... 127 Figure 5.14 rDNA and rRNA taxonomic summary comparing Gammaproteobacteria orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 – M85. ........................................................................... 129 xxi Figure 5.15 rDNA and rRNA taxonomic summary of Gammaproteobacteria taxa  across each sampling depth for each methanol column treatment. MeOH1 – M15, MeOH2 – M85. ........... 130 Figure 5.16 rDNA and rRNA taxonomic summary of Gammaproteobacteria taxa  across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, and 145cm) for each methanol column treatment. MeOH1 – M15, MeOH2 – M85 ........................ 131 Figure 5.17 rDNA and rDNA taxonomic summary comparing Betaproteobacteria orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. ................................................................................................................................ 133 Figure 5.18 rDNA and rRNA taxonomic summary of Betaproteobacteria taxa across each sampling depth for each methanol column treatment. MeOH1 – M15, MeOH2 – M85. ........... 134 Figure 5.19 rDNA and rRNA taxonomic summary of Betaproteobacteria taxa  across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, and 145cm) for each methanol column treatment. MeOH1 – M15, MeOH2 – M85 ........................ 135 Figure 5.20 rDNA and rDNA taxonomic summary comparing Firmicutes taxa from pristine and contaminated samples collected from the methanol blended fuel column experiments. ............ 136 Figure 5.21 rDNA and rDNA taxonomic summary comparing Firmicutes taxa orders across each sampling depth for each methanol column treatment. MeOH1- M15 MeOH2 –M85 ............... 137 Figure 5.22 rDNA and rDNA taxonomic summary comparing Firmicutes taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each methanol blended fuel column treatment ........................................................................... 138 Figure 5.23 rDNA and rDNA taxonomic summary comparing Euryarchaeota orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. ..... 140 xxii Figure 5.24 rDNA and rDNA taxonomic summary comparing Euryarchaeota taxa across each sampling depth each methanol blended fuel column treatment. MeOH1 – M15, MeOH2 – M85..................................................................................................................................................... 141 Figure 5.25 rDNA and rDNA taxonomic summary comparing Euryarchaeota taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each methanol blended fuel column treatment. MeOH1 – M15, MeOH2 – M85 ...................... 142 Figure 5.26 Geochemical measurements collected over the course of the methanol microcosm experiments. A) VFA, methanol, and pressure measurements for the 100ppm silt treatment; B) VFA, methanol, and pressure measurements for the 100ppm sand treatment; C) VFA, methanol, and pressure measurements for the 1000ppm silt treatment; D) VFA, methanol, and pressure measurements for the 1000ppm sand treatment; E) CH4, O2, CO2, and pressure measurements for the 100ppm silt sample; F) CH4, O2, CO2, and pressure measurements for the 100ppm sand sample. ........................................................................................................................................ 144 Figure 5.27 Rarefaction curves for sequences generated from the methanol microcosm experiment. .................................................................................................................................. 146 Figure 5.28 Diversity estimates for the methanol microcosm experiments. A) Shannon diversity index for the sand and silt methanol microcosms comparing  reference  start material, 100ppm treatment and 1000ppm treatment samples.  B) Venn diagram comparing OTUs shared between the sand microcosm samples. C) Venn diagram comparing OTUs shared between the silt microcosm samples. D) Venn diagram comparing OTUs shared between the sand and silt microcosm samples. .................................................................................................................... 147 Figure 5.29 NMDS ordination plot for samples collected from the methanol microcosm experiment. .................................................................................................................................. 149 xxiii Figure 5.30 Phylum level rDNA and rRNA taxonomic summary for samples collected from the methanol microcosm experiments .............................................................................................. 150 Figure 5.31 Phylum level taxonomic summary for methanol microcosm experiment across each treatment and starting material. ................................................................................................... 151 Figure 5.32 rDNA and rRNA taxonomic summary for Gammaproteobacteria taxa across each treatment and starting material for the methanol microcosm experiments. ................................ 152 Figure 5.33 rDNA and rRNA taxonomic summary for Acidobacteria taxa across each treatment and starting material for the methanol microcosm experiments ................................................. 153 Figure 5.34 rDNA and rRNA taxonomic summary for Firmicutes taxa across each treatment and starting material for the methanol microcosm experiments. ....................................................... 154 Figure 5.35 rDNA and rRNA taxonomic summary for Archaeal taxa across each treatment and starting material for the methanol microcosm experiments. ....................................................... 155 Figure 6.1 NMDS ordination plot for rDNA and rRNA samples collected from the Cambria denatured fuel grade ethanol spill site, ethanol blended fuel column experiments, methanol blended fuel column experiments, and methanol microcosm experiment. ................................. 165 Figure 6.2 Phylum level taxonomic summary for samples collected from the Cambria denatured fuel grade ethanol spill site, ethanol blended fuel column experiments, methanol blended fuel column experiments, and methanol microcosm experiment. ...................................................... 166  xxiv List of Abbreviations ASV – amplicon sequence variant BTEX – benzene toluene ethylbenzene xylene cDNA – complementary deoxyribonucleic acid DFE – denatured fuel grade ethanol DNA – deoxyribonucleic acid dNTPs – deoxynucleoside triphosphate / deoxynucleotide EDTA – ethylenediaminetetraacetic acid EtOH – ethanol MeOH- methanol NMDS – non-metric multidimensional scaling OTU – operational taxonomic unit PCA – principle component analysis PCR – polymerase chain reaction PERMANOVA – permutational analysis of variance qPCR – quantitative polymerase chain reaction rDNA – ribosomal deoxyribonucleic acid RNA – ribonucleic acid rRNA – ribosomal ribonucleic acid SSU – small subunit TBE – Tris borate EDTA VST – variance stabilizing transformation  xxv Acknowledgements  If there is one lesson I am taking away from my time studying microbial communities, it is that it really does take a community to survive. As cliché as it may sound, most things in life are best done when working with others and scientific research is no exception. So with that, I’d like to acknowledge and thank the various members of my eukaryotic community for helping and supporting me throughout this graduate degree. To my committee members Dr. Sean Crowe and Dr. Ulrich Mayer, thank you for the time and energy you put into my mentorship and for pushing me to think harder and become a better scientist. To the past members of the Hallam lab, thank you all the character and expertise you added to the lab that I have benefitted from in more ways than I can realize. To the current members of the Hallam lab, thank you for the life and intelligence you bring that have made the lab a fun second home for the past few years; I appreciate you all deeply. To Kateryna Ieydokymenko, thank you for constantly sharing your wisdom with me and for literally carrying me through on countless occasions, you are a true friend. To Joe Ho, thank you for kindness and friendship over these past few years, MICB 506 wouldn’t have been as much fun without you. To my mentors and friends, Aria Hahn, Colleen Kellogg, Monica Beltran Torres, Sachia Traving and Maya Bhatia thank you for all the support and guidance you’ve given me that have allowed me to grow as a scientist, I’m lucky to have such passionate scientists to look up to. To Melanie Sorensen and Andreas Mueller, thank you all the lab training you gave me as an undergrad and for teaching me the incredibly valuable skills of pipetting quickly and opening lids with one hand, I will treasure you and those skills always. To Darlene Birkenhead, thank you for helping me actually complete this degree. To my friends and family, thank you for your endless love and support. And finally, to my supervisor Dr. Steven Hallam, thank you for taking me on as a student, believing in my scientific abilities when I often didn’t, and for fostering the positive and inclusive environment where I truly was able to learn about science. See, it really does take a community. xxvi Dedication  To everyone who helped me get to this point, thank you.  1 Chapter 1: Introduction Microorganisms easily outnumber us. With a global population approaching 1030 cells [1], it becomes difficult to fully appreciate their multitude and the ability microorganisms possess to alter our environment by catalyzing fundamental biogeochemical reactions driven by their own metabolic needs [2-4]. Indeed, microbes are Earth’s master chemists [5, 6] and are responsible for processes, such as nutrient cycling and organic matter decomposition, that sustain ecosystems allowing them to respond to and thrive in spite of anthropogenic change [3, 5, 6]. With remarkable genetic potential contained in such miniscule life forms, microorganisms living in soil, ocean and subsurface environments have gained the fascination of microbiologists seeking to understand who is there and what they are doing both in isolation and community. These two questions have sparked an explosion of interdisciplinary research as scientists from different backgrounds have been brought together by a keen interest to ask these questions with specific applications in mind. Given their ubiquity and role in providing ecosystem services, microorganisms are often the first responders whenever there is significant change to an environment [3, 5, 7]. When environments are altered due to anthropogenic activities such as pollution, microbial community structure adjusts accordingly, often resulting in a decrease in species richness [8]. A decrease in microbial diversity is often met with a subsequent increase in specific taxa with traits that enable survival and overall dominance. Shifts in community structure can in turn alter ecosystem function if changes are significant and persistent [9]. A study by Wagg and colleagues evaluating the impact of soil microbial community composition on ecosystem multi-functionality identified a positive linear relationship wherein changes in community composition impacted different microbially driven ecosystem functions [10]. Additionally, studies evaluating the response of microbial communities to anthropogenic 2 hydrocarbon disturbances have revealed functional shifts in community dynamics with genes relevant to degradation pathways being highly expressed compared to pristine samples [11, 12]. Such studies evaluating microbial community responses also introduce the topics of ecological stability and functional redundancy as there can often be discrepancies between changes community structure and function [4, 13].   1.1  Microbial communities and ecological stability The ability of microorganisms to withstand change is a well-known phenomenon. In the environment, microbial communities are capable of thriving in extreme conditions due to persisting genetic diversity enabling resistance, resiliency and functional redundancy [14]. Microbial community resistance is the ability of the community to withstand a disturbance, resulting in no change in structure or function [9, 14]. Resiliency speaks to the ability to rebound from a disturbance, returning to previous community conditions [9, 14]. Both resistance and resiliency are common features of microbial communities given biological characteristics that exist at the cell, population and community level [9, 15] At the cellular level, microbial species have different stress responses resulting in a diversity of metabolic strategies that exist in response to one contaminant. For a population, the high growth and turnover rate of microorganisms provides an ideal opportunity for mutations and genetic transfers to occur, both of which can promote the stability of a community. Subsequently, interactions between community members acts as an insurance policy as interspecies connections can facilitate community survival [15, 16]. Finally, functional redundancy is the phenomenon where microbial communities are able to perform the same function despite changes in community structure due to a disturbance [14]. Together, these three concepts in microbial ecology highlight the 3 importance of studying community responses to different types of disturbances, both natural and human-caused. Indeed, how microbes respond to change has become an increasingly relevant question as human actions are altering the environment, often with unknown long-term consequences.   1.2 Microbial responses to anthropogenic disturbances Ecological responses to anthropogenic contamination events have been studied using a variety of observational, experimental and modeling approaches [4, 17-20].  Consider how the Exxon Valdez or BP Deepwater Horizon oil spills prompted studies to understand the immediate and long-term impacts of oil release and mitigation efforts spanning different trophic levels including microbial communities [21-24].  Such studies indicate that contamination events stemming from fossil fuel extraction and transport in aquatic and terrestrial environments can be directly linked to microbial responses with implications for remediation, monitoring and public policy [4, 7, 22-25] .  The ability of microorganisms to degrade hydrocarbons is well documented and acts as a natural removal and bioremediation strategy [25-31]. Indeed, certain taxa are capable of using different hydrocarbons metabolically, often with preference to simpler alkanes over complex aromatic compounds [29, 30]. Canonical hydrocarbon degraders typically belong to Proteobacteria, Actinobacteria and Firmicutes phyla as these groups contain the required catabolic machinery required for degradation and subsequent tolerance [27, 31]. A study by Militon et al. identified shifts in microbial community structure towards Gammaproteobacteria and Actinobacteria taxa in response to hydrocarbon contamination in soil systems [25]. Additionally, Morais and colleagues identified increases in the relative abundance of 4 Betaproteobacteria and Gammaproteobacteria taxa with traits of using organic compounds as electron donors [32]. Through these studies and others, we have been able to evaluate how microbial communities respond to an influx of hydrocarbons into the surrounding environment. This in turn has led to the establishment of well-documented models of aerobic and anaerobic degradation patterns and identification of the different microbes that fulfill these roles [33]. Indeed, syntrophic relationships between primary degraders and secondary consumers enables the degradation of complex hydrocarbons in anaerobic environments. In these relationships, primary degraders perform the initial remineralization to convert large hydrocarbon molecules into smaller intermediary products such as acetate, formate and H2 for secondary consumers [32, 33]. Under methanogenic conditions, the initial remineralization is thermodynamically unfavourable unless paired with methanogenesis which consumes intermediate products to keep concentrations low [33]. Indeed, under anaerobic conditions, complete degradation of hydrocarbons can only be achieved through collaborative, syntrophic relationships involving different members of the microbial community.   1.3 Microbial community responses to biofuel contamination Extensive research on microbial community responses to traditional hydrocarbon contamination has prompted similar evaluations to other fuel types, specifically blended biofuels. Blended biofuels use has gained in popularity as consciousness of greenhouse gas emissions from traditional fuel sources has increased. Though not entirely carbon-neutral, ethanol and methanol-blended biofuels are already commonly used in the transportation sector [34].  This trend is likely to increase as biofuel use increased by 1,100 million litres from 2010 to 2016 [35]. This increase in use and overall availability has made the likelihood of accidental spills into 5 terrestrial environments a near reality with unconstrained consequences [20, 36]. While these fuels may share similar performance standards with more traditional hydrocarbon fuels, their fate in the environment can differ drastically due to the volatile nature of ethanol and methanol.  As most subsurface terrestrial environments are carbon limited, the introduction of a new, labile carbon source can have significant impacts on the availability of electron donors and acceptors in microbial metabolism [37]. This can affect prevailing redox conditions, promoting changes in microbial community structure and function. Given the labile nature of ethanol and methanol, this often results in preferential degradation of these more accessible carbon sources over recalcitrant hydrocarbons [38-41]. A study by Feris et al. [37] found that the degradation rates of the more recalcitrant components of ethanol-blended biofuels (BTEX - benzene, toluene, ethylbenzene, and xylene) were reduced in the presence of ethanol. Likewise, concentrations of BTEX have remained elevated in the presence of ethanol from blended biofuels [42]. These changes in microbial-mediated degradation patterns in turn influences soil and groundwater toxicity levels. Additionally, the preferential degradation of ethanol can result in the production of propionate and butyrate by-products which can be used later to fuel the production of methane via methanogenesis [20].  The generation of methanogenic conditions is a common concern associated with biofuel use as it can stimulate conditions hazardous to human health [20, 36]. A long-term monitoring study by Spalding et al evaluating geochemical and environmental parameters at a site that had experienced a large scale ethanol blended fuel spill observed elevated dissolved methane concentrations in groundwater samples [43]. This was followed by a subsequent study by Sihota and colleagues who identified significant methane flux from surface soils in the impacted area [36]. Both studies showcase the possibility of hazardous, methanogenic conditions arising as a 6 result of biofuel contamination at time intervals that often exceed standard monitoring thresholds [36, 43]. Developing a thorough understanding of microbial community structure in response to biofuel spills in terrestrial systems is necessary in the development of effective monitoring and mitigation practices.   1.4 Thesis overview  The goal of this thesis is to investigate the subsurface microbial community response to ethanol and methanol blended biofuel releases in both field and laboratory settings using small subunit ribosomal RNA (SSU or 16S rRNA) gene surveys to measure the abundance and activity of different taxonomic groups in a cultivation-independent manner. In the case of laboratory experiments, microbial community structure associated with different ethanol and methanol blend ratios was determined at multiple time points after perturbation. Overall, this work aims to evaluate microbial community traits relevant to both aerobic and anaerobic degradation processes associated with natural attenuation, and develop a standardized method to analyze microbiome data in line with best statistical practices for a more accurate understanding of microbial community dynamics.   The specific aims for the subsequent chapters are as follows:  Chapter 2: 16S rRNA sequencing and data analysis Chapter 2 introduces the data analysis methodology used in subsequent data chapters. An overview of the 16S rRNA gene survey method is provided along with a comparison of two 7 normalization approaches used in 16S rRNA data analysis using a subset of data collected from a site in Cambria, Minnesota that had experienced a large volume ethanol-blended fuel spill.   Chapter 3: Denatured Fuel Grade Ethanol (DFE) blended fuels  In chapter 3, a standardized method to analyze microbiome data is applied to dataset generated from the Cambria, Minnesota denatured fuel grade (DFE) ethanol spill site. For the Cambria DFE spill site, I surveyed microbial community structure across the site with samples collected both within and outside the spill zone.   Chapter 4: Ethanol blended fuels  In chapter 4, a standardized method to analyze microbiome data is applied to data from the ethanol blended biofuel column experiments. For the ethanol blended biofuel column experiments, I surveyed microbial community structure across each treatment to test the hypothesis that different biofuel blend ratios (ratio of ethanol to gasoline) would drive unique shifts in the microbial community structure over the duration of the experiment.   Chapter 5: Methanol blended fuels Similar to the previous chapters, chapter 5 applies a standardized method to analyze microbiome data is applied to data from the methanol blended biofuel column and methanol microcosm experiments. Through the methanol column experiments, the microbial response to two methanol blended biofuel ratios was surveyed to test the hypothesis that each blend would could unique shifts in the structure of the microbial community. Through the methanol 8 microcosm experiments, I evaluated the impact of methanol toxicity on the microbial community.   Chapter 6: Conclusion Chapter 6 summarizes key findings relating microbial community responses to ethanol and methanol blended biofuels. Recommendations on future directions are also provided.   9 Chapter 2: 16S rRNA gene sequencing and data analysis As high-throughput sequencing technologies continue to advance, profiling microbial community structure through the use amplicons targeting different regions of the 16S rRNA gene has gained in popularity and quantitative power. In this chapter, I describe common methodology used in 16S rRNA gene studies and compare two approaches to data normalization using sequence information collected from the Cambria ethanol blended fuel contaminated site. The implications of each normalization approach on data interpretation are also evaluated. I conclude the chapter with recommendations for best practices in 16S rRNA data analyses; recommendations that have been applied to the subsequent data chapters of this thesis.   2.1 Origins and application of 16S rRNA gene sequencing Advances in high-throughput sequencing technologies have increased our capacity to chart microbial community structure and function across different environments revealing deeper insight in the abundance, diversity and functions of the microbial world. One common sequence-based approach to studying microbial communities comes from the use of the 16S rRNA gene to identify organismal phylogeny. The origins of the 16S rRNA gene as a phylogenetic marker began in 1977 when Carl Woese first used the gene to identify and distinguish different cultivated microbial species from one another [44]. This seminal study laid the foundation for other significant work in microbial ecology, including the use of the 16S rRNA gene to study uncultivated microbes from the environment by Norman Pace nearly 20 years later [45] . Fast forward to the present day, and the use of the 16S rRNA gene in microbial ecology is common place, resulting in an abundance of 16S datasets. This in turn has led to the development of computational tools, such as QIIME [46] and MOTHUR [47], to process these datasets 10 efficiently and reproducibly. As a result of the extensive use of the 16S rRNA gene to characterize microbial communities from a diverse set of environments, a standardized workflow has emerged capturing the key steps required to generate these datasets [48].  First, sample material is collected and nucleic acids (DNA and or RNA) are extracted from cellular material to profile microbial community abundance and potential activity. Following nucleic acid extraction, the 16S rRNA gene must be amplified using primers that capture both the hypervariable and conserved regions of the gene. Using appropriate 16S rRNA primers is a key step as the use of ill-designed primers can lead to strong amplification bias against certain taxa resulting in a skewed representation and interpretation of the community [49]. Next, amplified regions are sequenced which can result in anywhere between 2,000 to 80,000 reads per sample with the variation between samples often attributed to sample amplification biases, library prep chemistry, and multiplexing e.g. the use of DNA barcodes to sequence multiple samples in pools. Using computational tools such as QIIME or MOTHUR, sequences are clustered into operational taxonomic units (OTUs) at sequence similarities ranging from 97 – 99% [46, 50]. Recently, alternative methods have been developed to process 16S rRNA sequence data that avoid clustering altogether and instead rely on extensive “denoising” to remove chimeric sequences resulting in the generation of amplicon sequence variants (ASVs) instead of OTUs [51]. However, regardless of whether OTUs or ASVs are used in downstream analyses, abundance is still expressed as the number of sequence reads attributed to a specific OTU or ASV. Subsequently, due to variations in the number of sequence reads produced per sample, data normalization is required prior to any downstream analyses in order to make meaningful comparisons and interpretations.   11 2.2 16S sequence data normalization  Common methods of data normalization in microbiome studies include rarefying (subsampling the data down to a common number of sequences per samples) or converting read counts into proportions [52]. While both methods are easy to perform and likely sufficient in certain contexts, they have recently been re-evaluated from a statistical standpoint due to intrinsic characteristics of these datasets [52-54]. Statistically, 16S rRNA sequence data is classified as count datasets wherein each value represents a discrete number of times an event has occurred which, in this case, is the number of times a unique OTU is identified in a given sample [55]. These datasets are typically overdispersed, meaning there is greater variability in the data then what would be expected from more typical count data e.g. normal distribution. This can contribute to inaccuracies when comparing the relative abundance of OTUs among and between samples [54, 56]. Given this awareness, there has been a push to adopt alternative normalization approaches that incorporate the statistical distribution of the data. One such approach is the variance stabilizing transformation (VST) which uses the negative binomial distribution to normalize the data to account for differences in the library size (number of sequence reads) between samples [54].  Already employed in numerous RNA-seq data analyses [57-60], this normalization approach is advantageous as it does not discard data, minimizes type 1 errors due to overdispersion, and stabilizes the variance across samples thus directly addressing heteroscedasticity (unequal variance) [54]. Moreover, many standard analyses used in microbial ecology, such as principal component analyses (PCA) and non-metric multi-dimensional scaling (NMDS), assume homoscedasticity wherein there is equal variance for all values of a predictor variable [61]. Because VST normalization produces transformed homoscedastic read counts, VST may prove to be more suitable for these statistical analyses as well as for other multivariate 12 analyses [62]. Finally, VST is also a more suitable normalization approach for analyses relying on correlations such as co-occurrence networks as compositional data (wherein each sample sums to the same number) is known to produce spurious correlations [63]. Despite the strong statistical argument for using VST, a direct comparison of data products resulting from more conventional normalization techniques and VST has yet to be completed.  Here, I directly compare the suitability of standard 16S rRNA exploratory analyses using both proportion-normalized and VST normalized data. I use small subunit ribosomal rRNA and rRNA gene profiles to compare microbial abundance (rDNA) and activity based on rRNA:rDNA ratios. The side-by-side comparison was based on samples from a field site affected by the accidental release of denatured fuel-grade ethanol (DFE) extracted from soil cores from a pristine reference location (Core A) and a location affected by DFE (Core B). Finally, I provide a list of practical considerations that can be used to guide statistical analyses after VST normalization has been applied in order to more accurately visualize and interpret 16S rRNA data.   2.3 Methods 2.3.1 Sample collection, DNA extraction and cDNA synthesis  Samples were collected from a site that had been affected by an accidental release of an ethanol blended fuel. Core A and Core B were collected using a GeoProbe in October 2012 in southwestern Minnesota (lat. 44.247083, long. -94.3371059). The site was affected by the accidental release of DFE as a result of a multicar train derailment (Figure 2.1). Subsamples were taken in 10 cm intervals from each sediment core and at each interval approximately 15 g of sediment was collected in falcon tubes and immediately frozen. Samples were shipped on dry 13 ice to the University of British Columbia where DNA and RNA was extracted. DNA and RNA was extracted using the MoBio PowerMax® Soil DNA and PowerSoil RNA Isolation® Kits. RNA was converted to cDNA using the Invitrogen Superscript®III first-strand synthesis kit following manufacturer protocol.  Figure 2.1 Cambria denatured fuel-grade ethanol (DFE) spill site. A) Spill site schematic indicating the location of soil sampling cores A and B. B) Sampling plan and ethanol concentrations along the soil profile for core B.   ABCore ADissolved CH4 Trees Dissolved EtOHCore B2 4 60EtOH [g kg-soil-1]406080100200%VSTrDNArRNA++++Depth [cm]Rail Line14  2.3.2 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing The V6 – V8 region of the small subunit rRNA (SSU or 16S rRNA) gene was amplified from both DNA and cDNA template using the universal primer pair 926F (5′-AAACTYAAAKGAATTGRCGG-3′) and 1392R (5′-ACGGGCGGTGTGTRC-3′) [64] to generate molecular barcodes capturing microbial community diversity.  Primers were modified to include 454 adaptor sequences and barcodes were added to the reverse primers to allow multiplexing during sample sequencing. Duplicate 25µL PCR reactions were performed for each sample with each well containing 2.5µL 10X buffer, 1µL 50mM MgSO4, 0.5µL 10mM dNTPs, 0.5µL 10µM forward and reverse primers, 2µL DNA or cDNA template, 0.2µL Platinum Taq high fidelity polymerase (5U / µL) and 18.3µL of nuclease free water. Negative controls were included for each sample reaction to confirm that no DNA contamination had occurred. Thermal cycler protocol started with denaturation at 95ºC for 3 minutes, followed by 30 cycles of additional denaturation at 95ºC for 30 seconds, annealing at 55ºC for 45 seconds, and extension at 72ºC for 90 seconds. 10 minutes at 72ºC for final extension completed the amplification process. PCR products were visualized on a 1% 1X TBE agarose gel run at 100V for 35 minutes. Successful PCR products were pooled, cleaned using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel), eluted in 30μL of 5mM Tris buffer pH 8.5, and quantified following the Quant-iT Picogreen dsDNA kit (ThermoFisher) protocol. Each barcoded amplicon was pooled at 100ng DNA prior to sequencing. Genome Quebec (Montreal, Canada) performed emulsion PCR and sequencing on the Roche 454 GS FLX Titanium platform according to manufacturer protocol.   15 2.3.3 Sequence analyses  275,606 and 127,187 total sequences were obtained from the pooled rDNA and rRNA amplicons, respectively. All trimming, clustering, and classifications were performed in QIIME (version 1.4.0 software package) [46]. Quality control removed sequences that failed to meet the following criteria: minimum length = 200 bp, average quality score = 25, maximum number of Ns = 0, maximum homo- polymers = 6, resulting in 224,599 and 102,550 sequences for the pooled rDNA and rRNA amplicons, respectively. Using UCLUST [65], sequences were clustered into operational taxonomic units (OTUs) at the 97% identity threshold with a maximum e-value cut-off of 1e-10. Samples containing less than 400 OTUs in either the rDNA or rRNA fraction were excluded from the datasets resulting in a total of 36 samples from reference (Core A) and ethanol contaminated (Core B) soil cores. Sequences were clustered with an additional 165 core samples (including both rRNA and rDNA) collected at different distances from the contamination source. Although OTUs were generated from an ongoing time-series we focused on the first time point for which there was data from a reference and contaminated core. This approach increased the sensitivity in our study by providing a larger dataset from which to select representative sequences for each OTU [66, 67]. OTUs occurring in less than 5% of the samples were removed in order to reduce the signal to noise ratio [54] resulting in total of 6,694 OTUs and encompassing 95.3% of all sequence data.   2.3.4 Data analyses  Statistical analyses were performed in R and in-house perl scripts. Diversity estimates were calculated using un-normalized data. The OTU table was normalized using both proportions and VST (where reference DNA, reference RNA, contaminated DNA, and 16 contaminated RNA were used as the conditions on which to estimate size factors). In order to calculate rRNA:rDNA ratios in the proportion-normalized dataset, it was necessary to impute rDNA values for OTUs wherein rRNA was recovered, using non-parametric multiplicative replacement [68, 69]. Imputation was unnecessary for the VST transformed data, as it does not contain zeros. To avoid calculating log(0) and producing negative values which make rRNA:rDNA ratios more difficult to interpret, a constant was added prior to VST normalization resulting in an overall scaling (multiplication) of the data on the untransformed scale [54]. Both imputation (proportions) and the addition of a constant serve to eliminate dividing by zero and negative numbers respectively, but keep values small such that rRNA:rDNA ratios can be calculated as accurately as possible. Non-metric multi-dimensional scaling (NMDS) was completed using the Bray-Curtis dissimilarity index able to handle both count and abundance data [70]. rDNA was used as a proxy for microbial abundance while rRNA:rDNA was used to estimate microbial activity.   2.1 Results 2.1.1 Microbial community diversity  To compare data products resulting from conventional proportion and VST normalization, we investigated soil microbial community structure and potential activity from the pristine reference core (Core A) and the ethanol-contaminated core (Core B). Specifically, we performed 454 pyrosequencing of rDNA and rRNA sequences with three-domain resolution and following clustering at 97% identity, the removal of singletons, and operational taxonomic units (OTUs) present in <5% of the 36 samples, 6,694 operational taxonomic units (OTUs) were used for downstream analysis. Using rarefaction curves to determine sample coverage, we found 17 samples exhibited similar slopes at the 97% identity threshold approaching 4,000 unique OTUs suggesting that at this level of genetic distance only rare OTUs remain unrecovered (Figure 2.2). Of the total OTUs, 5,834 (87.2%) were affiliated with Bacteria, 519 (7.8%) were affiliated with Archaea, 243 (3.6%) were affiliated with Eukarya, and 98 (1.5%) were unclassified (representing sequencing errors or rare taxa).   Figure 2.2 Rarefaction curves for sequences generated from Core A and Core B from the Cambria denatured fuel grade ethanol spill site.  First, we compared the number of OTUs in common between the two cores and the rDNA and rRNA in both cores to determine community level response to perturbation (Figure Core A 10cm DNACore A 10cm RNACore A 30 DNACore A 30cm RNACore A 40cm DNACore A 40cm RNACore A 50 cmCore A 50cm RNACore A 50cm RNACore A 70cm DNACore A 70cm RNACore A 80 cmCore A 80cm RNACore A 100cm DNACore B 10cm DNACore B 10cm RNACore B 20cm DNACore B 20 cm RNACore B 30cm DNACore B30 cm RNACore B 40cm DNACore B 40cm RNACore B 50cm DNACore B 50cm RNACore B 60cm DNACore B 60cm RNACore B 70cm DNACore B 70cm RNACore B 80 cm DNACore B 80cm RNACore B 90 cm DNACore B 90cm RNACore B 100cm DNACore B 100cm RNACore B 110cm RNA02000400060000 2000 4000 6000sequenceschoa 118 2.3). More OTUs were shared between rDNA and rRNA within each core than between Core A and Core B (Figure 2.3). Diversity was estimated on un-normalized data using both Shannon’s and Simpson’s diversity indices [71]. While Simpsons diversity was close to 1 for all samples, in general; Shannon’s diversity was higher in Core A compared to Core B. Indeed, using a two-tailed t-test, Core A had significantly higher diversity than Core B (p=0.029). However, within each core no significant diversity between rDNA and rRNA were found p>0.05). Next, OTU counts were normalized using both proportions and VST. Variance estimates were fit using both Poisson (φ = 0) and negative binomial models (Figure 2.3). While variances were larger than estimated under the Poisson model, because the data are overdispersed, the negative binomial model fit the variances well, highlighting VST as an appropriate normalization method for the data (Figure 2.3) [54].     Figure 2.3 A) Variance estimates using the negative binomial model when each point represents an OTU B) Venn-diagram of the OTUs shared between rDNA and rRNA fractions within and between Cambria DFE soil Core A and Core B 1e+001e+021e+041e+0610 100MeanVariance136842147092022331312795283216559681098 1325853302343Core A Core B DNA RNA A B19  2.1.2 Soil microbial community structure  To evaluate differences in microbial community structure among and between Core A and Core B, we employed principal component analysis (PCA) (Figure 2.4) and non-metric multidimensional scaling (NMDS) (Figure 2.5), using both proportion and VST-normalized data. Large-scale trends in PCA produced by both normalization approaches were similar, grouping samples primarily by depth and subsequently by core (Figure 2.4). However, using proportion-normalized data, no strong trends were identified within the NMDS (Figure 2.5). Indeed, samples from both cores and across depths grouped together suggest both soil depth and disturbance had little influence on microbial abundance and activity. In contrast, using the VST-normalized data, NMDS revealed samples grouped along a depth gradient. Taken together the data suggest microbial abundance and activity are affected by depth and thus the NMDS using proportion should be interpreted with caution (Figure 2.4 and Figure 2.5).  20   Figure 2.4 Principle component analysis (PCA) plot for Core A and Core B Cambria DFE spill site rDNA and rRNA samples using proportion-normalized data (in pink) and VST-normalized data (in violet) -200204060-50 0 50PC1 (16.5% explained var.)PC2 (7.8% explained var.)% VST10 cm20 cm30 cm40 cm50 cm60 cm70 cm80 cm90 cm100 cm110 cmCore A Core B DNARNA -40-2002040-60 -40 -20 0 20 40PC1 (11.4% explained var.)PC2 (5.6% explained var.)21  Figure 2.5 NMDS plot from Core A and Core B Cambria DFE spill site rDNA and rRNA samples using proportion-normalized data (in pink) and VST-normalized data (in violet). -1.0 -0.5 0.0 0.5 1.0-1.0-0.50.00.51.0NMDS 1NMDS 2% VST10 cm20 cm30 cm40 cm50 cm60 cm70 cm80 cm90 cm100 cm110 cmCore A Core B DNARNA -2 -1 0 1 2 3-2-10123NMDS1NMDS222  2.1.3 Microbial community abundance and potential activity Next, I evaluated microbial community abundance and potential activity with depth and across natural and contaminated conditions based on phylum-level taxonomic affiliation of OTUs. At the phylum level, trends in rDNA and rRNA across Core A and Core B were similar using both proportion (Figures 2.6 – 2.7) and VST-normalized datasets (Figures 2.8 – 2.9). Cores were dominated by bacterial affiliated with Proteobacteria (Alpha, Beta, Delta and Gamma), Actinobacteria, Chloroflexi, Firmicutes, Bacteroidetes and Acidobacteria as well as archaea affiliated with Euryarchaeota and Thaumarchaeota. Additionally, 18 candidate phyla, including OP11 and WS3, were identified in the dataset. Within Core A, Alpha-Proteobacteria were most abundant (rDNA) from 10-50 cm compared to deeper depths (60 -100 cm) while Beta-, Delta-, and Gamma-Proteobacteria either remained constant with depth or showed an increase in rDNA below 60 cm (Figure 2.6 and Figure 2.8). Similarly, the abundance of Actinobacteria, Chloroflexi, and Acidobacteria and to a lesser extent Firmicutes and Bacteroidetes varied little throughout Core A using both normalization techniques (Figure 2.6 and Figure 2.8). In addition, Euryarchaeota and Thaumarchaeota were generally most abundant from 10-30 cm above the ethanol-contaminated interval (Figure 2.6 and Figure 2.8).  Next, I compared the abundance of microbial phyla between to the two cores in order to identify potential taxon-specific responses to ethanol contamination. The abundance of Bacteriodetes remained consistent between Core A and B using both normalization techniques suggesting there is little impact of ethanol contamination at the phylum level. The abundance of Deltaproteobacteria and Chlorflexi was lowest in the ethanol-contaminated interval (50-70 cm) using both normalization techniques suggesting that these phyla are sensitive to ethanol 23 contamination (Figures 2.6 – 2.9). In addition, Actinobacteria and Acidobacteria decreased in abundance below 40 cm (where ethanol concentration is higher) from Core A to Core B using both normalization techniques although the magnitude of this change differs between the two normalization techniques. Indeed, within the proportion-normalized data Actinobacteria and Acidobacteria exhibited a 2-fold and 2 - 17-fold decrease respectively while within the VST- normalized data these taxa decreased by 1-fold and 2-fold respectively. Conversely, Firmicutes displayed a 1.4 - 8.1-fold (proportions) / 1.6 - 2.4-fold increase in rDNA abundance from 50 -70 cm from Core A to Core B suggesting a biological response to ethanol contamination. Although the two normalization techniques produced similar trends in microbial abundance, this consistency was not universal within phyletic groups. For example, within the proportion-normalized data Alpha-, Beta-, and Gamma-Proteobacteria decreased in abundance below 40 cm 1.1-2.3-fold, 2.0 - 8.6-fold, and 1.8 - 2.7-fold respectively again suggesting these taxa are sensitive to ethanol as peak contamination occurred at 60 cm. However, within the VST-normalized data Alphaproteobacteria did not appear to decrease while Beta-, and Gamma-Proteobacteria decreased by only 1.0 - 1.6 fold and 1.0 -1.2-fold respectively suggesting that these groups are not as sensitive to ethanol but rather are able to tolerate the disturbance (Figures 2.6 – 2.9).  This represents a discrepancy between the normalization techniques present at the phylum level that could affect downstream interpretation of the data.  24  Figure 2.6 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core A using proportion-normalized data 25  Figure 2.7 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core B using proportion-normalized data 26  Figure 2.8 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core A using VST-normalized data 27  Figure 2.9 Phylum level rDNA and rRNA taxonomic summary for Cambria DFE soil Core B using VST-normalized data 28 2.1.4 Potential activity  As both rDNA and rRNA were sequenced, the ratio of rRNA to rDNA was used to evaluate the potential activity of the microbial community throughout both cores. Within Core A and Core B the ratio of rRNA:rDNA for most phyla produced similar trends using both normalization approaches though on different scales (Figures 2.10 – 2.11). For example, candidate phyla WS3, Euryarchaeota and Thaumarchaeota were more active between 30-60 cm in Core B compared to Core A suggesting these groups may be stimulated by ethanol contamination. 29  Figure 2.10 Phylum level rRNA:rDNA ratio taxonomic summary for Cambria DFE soil Core A using VST-normalized data 30  Figure 2.11 Phylum level rRNA:rDNA ratio taxonomic summary for Cambria DFE soil Core B using VST-normalized data  To further explore relationships between depth and potential activity of the microbial 31 community with increased taxonomic resolution, we identified the taxonomic orders whose abundance and activity varied most throughout both cores using both normalization techniques. To determine which orders were highly variable, the standard deviation of the abundance of rDNA and rRNA of each taxonomic order across Core A and Core B was calculated. The 30 orders (top 10%) with the highest standard deviation were selected from both the proportion and VST-normalized data and the potential activity (rRNA:rDNA) of these groups was explored. 18 out of 30 orders selected were common to both proportion and VST-normalized datasets (Figure 2.5 and Figure A.8). From these 18 orders, 11 were bacterial, 5 were archaeal and the remaining 2 were eukaryotic (Figures 2.12 – 2.13).  Within Core A, using both the proportion and VST-normalized data trends were similar wherein archaeal orders Methanobacteriales, Methanomicrobiales, and Methanosarcinales all displayed a peak in potential activity at 70 cm (Figure 2.12). However, within the proportion- normalized data Methanomicrobiales and Methanosarcinales were also active at 30 cm. Potential activity of Coriobacteriales, Bacteroidales and vadinHA17 also peaked at 30 cm and 70 cm in both normalized datasets. Additionally, Charophyta, a ubiquitous soil alga [72], had the highest potential activity from 50-80 cm using both normalization techniques. In groups that were identified as being highly variable within the proportion-normalized data but not the VST-normalized data, Anaerolineales, Nitrospirales, Syntrophobacterales were most active at surface depths 10 cm and 30 cm. In contrast, within groups identified as highly variable in the VST-normalized data but not the proportion-normalized data, Thaumarchaeota Group C3, Halobacteriales and Mollicutes showed elevated activity at 30 cm and within the saturated zone (>70 cm), indicating the activity of these groups in both regions of the column.  32  Figure 2.12 Core A rRNA:rDNA heatmap for the most variable orders identified using proportion-normalized (in pink) and VST-normalized (in violet) data. Panel A) depicts taxonomic orders identified as being the most variable (top 10%) in both proportion-normalized and VST-normalized datasets. Panel B) depicts orders identified in the proportion-normalized dataset only, and Panel C) depicts orders identified in the VST-normalized dataset only.   33 Within Core B, most trends in potential activity were consistent across both normalization techniques (Figure 2.13). For example, the peak in potential activity for Enterobacteriales occurred at 60 cm while Methanosarcinales and Charophyta were most active in the surface samples (10-30 cm) although these may have taxa likely occupied anaerobic and aerobic niches respectively. In addition, Methylococcales were most active below 60 cm, again using both the proportion and VST- normalized data. Within groups that were identified as being highly variable within the proportion- normalized data but not the VST-normalized data, Sphingomonadales and Syntrophobacterales were most active below 60 cm in the saturated zone suggesting these groups are most active under anaerobic conditions. Further, Burkholderiales were most active in the interval where ethanol concentration peaked, and, within groups that were identified as being highly variable within the VST-normalized data but not the proportion–normalized data; Thaumarchaeota Group C3, Halobacteriales and Mollicutes were most active in the surface samples (10-30 cm) consistent with the potential activity of these groups within Core A. Finally, OP11 was most active in the saturated zone (>70 cm).  34  Figure 2.13 Core B rRNA:rDNA heatmap for the most variable orders identified using proportion-normalized (in pink) and VST-normalized (in violet) data. Panel A) depicts taxonomic orders identified as being the most variable (top 10%) in both proportion-normalized and VST-normalized datasets. Panel B) depicts orders identified in the proportion-normalized dataset only, and Panel C) depicts orders identified in the VST-normalized dataset only.     35 2.2 Discussion  In the present study I compared the suitability of standard 16S rRNA exploratory analyses using both proportion-normalized and VST normalized data on small subunit ribosomal rRNA and rRNA gene profiles from a field site affected by the accidental release of denatured fuel-grade ethanol (DFE). I compared microbial abundance (rDNA) and activity based on OTU abundance and rRNA:rDNA ratios on samples extracted from soil cores from a pristine reference location (Core A) and a location affected by DFE (Core B). My comparison included high-level multivariate community wide exploratory techniques such as non-metric multi-dimensional scaling (NMDS) and principle component analysis (PCA) as well as phylum and order level assessments to identify potential discrepancies in taxonomic patterns between the two normalization techniques that could influence downstream interpretation.   2.2.1 Similar trends in community and diversity and structure with varying normalization technique  Diversity was estimated to be highest in the samples from Core A [71] suggesting ethanol exposure changes community composition and decreases overall taxonomic diversity. Within both cores no significant differences in diversity were identified between the rDNA and rRNA suggesting that most species present were likely active. As diversity estimates are best performed on un-normalized data [73] no comparison was made between proportion and VST data for this analysis. However, it should be noted that Shannon’s and Simpson’s diversity estimates are not affected by proportion normalization as both methods calculate the relative fraction of each taxa [71]. With respect to community structure, large-scale trends were similar regardless of which normalization approach was used. Using PCA, samples grouped first by depth and subsequently 36 by core. Community structure within Core A and Core B were most similar in samples between 10-60 cm and differed most in saturated samples (>70 cm) . However, while the NMDS generated with VST-normalized data produced trends similar to the PCA, this was not true of the proportion-normalized data. Instead, we did not observe any trend in the data consistent with depth. This discrepancy is likely due to the compositional data bias present within proportion-normalized datasets which is known to impact ordination results [56, 74, 75]. As such, I suggest VST normalization prior to ordination as this may increase the consistency of trends produced using PCA and NMDS.   2.2.2 Discrepancies in abundance and potential activity between normalization techniques  Both normalized datasets revealed an increase in the abundance of Firmicutes, a phylum known to contain taxa that can survive extreme conditions [76], and Thaumarchaeota, an archaeal phylum comprised of ammonia oxidizers [77], within saturated (>70 cm) samples from Core A and Core B (Figures 2.6 – 2.9). However, the magnitude of these increases changed with normalization approach. Such discrepancies in the magnitude of change may influence interpretation of the biological role of certain taxa. For example, the ability of Firmicutes to endure in ethanol contamination soil may be overestimated in the proportion-normalized data as the VST-normalized data suggests the increase is less extreme (Figures 2.6 – 2.9). Indeed, the proportion-normalized data suggests the increase in Firmicutes is 2-fold greater than the increase observed based on VST normalization. Conversely, the increase in Thaumarchaeota within the proportion-normalized data may be overlooked as the increase was 4-fold greater based on VST normalization, a difference that could reflect relevant nitrogen or methane transformations within 37 the soil profile [78]. I posit that discrepancies in the magnitude of change can influence the biological interpretation of the results as the influence of environmental factors on given taxa, or the functional role of a given taxa, may be exaggerated or overlooked depending on how the data were normalized. Indeed, in addition to trends in the abundance of OTUs or phyla, the magnitude of changes in abundance must also be considered. As such I propose it is best practice to employ VST, as this model better fits over-dispersed OTU data.  Next, I examined potential activity (rRNA:rDNA) throughout Core A and Core B using both  proportion and VST-normalized data (Figures 2.10 – 2.11). Several taxa including candidate phyla WS3 and the Archaeal phylum Euryarchaeota were more active between 30-60 cm in Core B compared to Core A using both normalization techniques suggesting these groups may be stimulated by the ethanol spill. Indeed, Euryarchaeota are known to emcompass the majority of archaeal methanogens [79] and candidate phyla WS3 has been previously found to degrade hydrocarbons [80]. In similar biofuel contaminated environments ethanol concentrations up to 10g/L (1.3% v/v) have been known to stimulate microbial growth while concentrations above 40-100g/L exhibit toxic, sterilizing effects [81]. In general, bacterial and archaeal microorganisms are unable to tolerate ethanol concentrations above 9% (v/v) [82]. Here, ethanol concentrations between 40-60 cm decreased from 4-6g ethanol/ kg soil (0.8% to 1.2%) thus providing a favorable growth substrate [81] across this depth interval. Indeed, I did not observe any phyla active in Core A that were no longer present or active in Core B, consistent with the idea that the ethanol concentration at this site is neither toxic nor sterilizing. Still, care must be taken when interpreting rRNA:rDNA ratios, as elevated levels of rRNA have been attributed to sources other than increased activity (e.g., multiple ribosome copies [83], dormant cell function [84, 85]. Explicitly, elevated rRNA:rDNA ratios can be indicative of cells entering dormancy or 38 be reflective of past environmental conditions prompting high ribosomal levels [84].  Finally, to explore discrepancies in the relationships between depth and potential microbial activity between proportion and VST-normalized data, I identified the 10% most variable taxonomic orders and examined the rRNA:rDNA ratio across Core A and Core B using both normalization techniques (Figures 2.12 – 2.13). 12 of the 30 orders selected (40%) were different between the proportion and VST-normalized data, indicating the magnitude of potential activity varied across the profile differently using the two normalization techniques. For example, using the proportion-normalized data the activity of Burkholderiales, known to grow on ethanol [79], varied 3-6 fold from 50-110 cm, while using the VST-normalized data, potential activity of Burkholderiales varied only 1.5-2 fold across this same depth interval. Differences in magnitude of change across depth intervals and between uncontaminated and contaminated soils contributes to the high false positive rate in calculations of differential abundance when data are normalized using proportions [54] as such, fold-changes within proportion-normalized data should be interpreted with caution.   2.2.3 Considerations for VST normalization  Although normalizing OTU count data with VST is not currently a standard of practice in microbial ecology outside the context of differential abundance testing [57-59], VST offers multiple statistical advantages that can lend to more informative data analysis [54]. However, despite the statistical rigor of VST, there are several considerations that need to be taken into account when generating exploratory summary graphics with VST. The findings from this study reveal several practical considerations that can help microbial ecologists select the most suitable normalization techniques for a given analyses and effectively use VST normalization to analyze 39 rDNA and rRNA data.   2.2.3.1 Diversity indices  As discussed previously, diversity estimates are best performed on un-normalized data [73].   2.2.3.2 Community structure   NMDS and PCA ordination both make many assumptions about the properties of the data [61, 62, 70, 74, 86-90], including homoscedasticity, a property of the data not respected within proportion-normalized data. However, VST-normalized data are homoscedastic and thus do not violate the assumptions of many exploratory statistical techniques. Indeed, compositional data bias is known to affect ordination results [74, 75]. However, the magnitude of the bias decreases with an increase in the number of variables [74]. As such, a higher number of OTUs may reduce the effect of the bias introduced via normalization technique and thus the choice of normalization technique is more likely to impact the interpretation of low diversity environments. 
  2.2.3.3 Community composition  Phylum and order-level exploratory graphics such as bubble plots and heatmaps do not have statistical assumptions and as such both proportion and VST normalization can be used. However, it is important to note that  VST normalization approaches a log base 2 transformation [60, 91]. As a result, it is important take logarithmic rules and identities into consideration when working with VST-normalized data. For example, as adding log-transformed values is equivalent to multiplying the original count data (log(a) + log(b) = log(a*b)), normalized OTU count data must first be summed at the taxonomic level at which the analysis or comparison will be 40 completed (e.g.; Phylum or Order) prior to VST normalization (Figure 2.4). Indeed, here un-normalized count data was summed at both the phylum and order level prior to VST normalization and figure generation. In contrast, proportion-normalized data scale linearly and thus the data can be normalized only once at the OTU level and proportions can subsequently be summed directly to create phylum and/ or order level tables and figures.    2.2.3.4 Indicator species analyses  Given that VST normalization approaches a log base 2 transformation at high values [60, 91] indicator species analyses should be approached with caution when using VST- normalized data to ensure results are not due to a misappropriation of logarithmic identities.   2.2.3.5 Correlations and co-occurrence networks  Correlations and co-occurrence networks generated from small subunit ribosomal RNA (SSU rRNA) genes (OTUs) are known to suffer from compositional data (relative proportions) effects that result in false correlations and/ or convert legitimate positive correlations to negative ones [63]). As such, it is advisable to use VST-normalized data to calculate correlations [54] or employ alternative methods for co-occurrence network analyses such as SPARCC [63] or the ‘ensemble’ method [92, 93] both of which have been reviewed in detail [94].   2.3 Conclusions  Using 16S rRNA and rDNA gene surveys it is possible to capture the abundance and potential activity of taxonomic groups present within microbial communities. When accurately analyzed and interpreted, this can lead to a deeper understanding of microbial responses to 41 perturbations and assist in the development of conceptual models for monitoring microbial activity in natural and engineered environments. Canonical normalization techniques such as proportions applied to discrete rDNA and rRNA sequence read count data may be problematic when analyzing and interpreting 16S rRNA sequence information. We found that normalization technique had the largest impact on exploratory graphics with statistical assumptions (e.g.; NMDS) as well as the magnitude of change in microbial abundance and activity. Indeed, trends in microbial composition with depth were only observable within the NMDS using VST normalization. Furthermore, given that differential abundance analyses using proportion-normalized data is known to increase the likelihood of type I errors, interpretation of the magnitude of change across depth intervals or in response to ethanol contamination using proportion-normalized data are more likely to be inaccurate. Based on these results, I advocate the use of VST in summary and exploratory analyses as it is statistically appropriate and propose that normalization technique can impact our understanding of microbial abundance and potential activity. This work contributes to methodological improvements in OTU count data normalization and analyses by providing both a comparison of summary graphics and a practical guide to methodological considerations given different normalization techniques.   42 Chapter 3: Evaluating microbial community responses to denatured fuel grade ethanol at the Cambria, Minnesota spill site   Increased production and transportation of ethanol-blended fuels has raised concerns on the potentially detrimental impacts to the environment if a spill were to occur. In this chapter, I evaluate the microbial community response to fuel grade ethanol contamination that occurred at the Cambria spill site. This chapter uses data from samples collected at the Cambria spill site to characterize the response of the microbial community upon exposure.   3.1 Introduction Increased awareness of the social, economic and environmental costs associated with fossil fuel use has increase a desire to seek alternative and more sustainable sources of energy [95, 96]. For example, both the United States and the European Union have promoted the use of blended ethanol (EtOH) biofuels derived from renewable plant sources such as corn, switchgrass and algae [34, 96, 97].  However, while considerable research has examined trade-offs associated with biofuel production including CO2 emissions [95, 97] environmental impacts associated with biofuel spills such as soil and groundwater contamination and fugitive gas emissions remains an important area of research poorly constrained [37] . As biofuel production and transport occur primarily via rail cars and tanker trucks [98, 99], the majority of biofuel releases will likely result in surficial spills impacting the complete soil profile from vadose to saturated zones. Contemporary studies examining biofuel releases have focused mainly on microbial processes within the saturated zone neglecting microbial community responses within the vadose zone where the majority of DFE transformation is likely to transpire [38, 40, 100, 101].  43 Here we investigate microbial community responses to a large-scale biofuel release (denatured fuel-grade EtOH or DFE) caused by a train derailment using pyrosequencing to generate rRNA and rDNA profiles through the vadose and saturated zones. We compare microbial community structure within the source site to a reference site unaffected by the DFE, and evaluate the relationship between microbial abundance and potential activity.   3.2 Methods 3.2.1 Cambria DFE spill site characteristics The Cambria study site is located in southwestern Minnesota (lat. 44.247083, long. -94.33710598677882). Contamination resulted from a multicar train derailment in November 2006, which released 9.5 x 104 L of DFE onto land surface. Following removal of ponded product, approximately 4.9 x 104 L of the E95 fuel remained, with the majority pooling in a local depression to the west of the rail line (Figure 3.1). No soil excavation occurred at the site.  Ethanol concentrations in groundwater  within the spill zone remained elevated since the spill occurred, with acetate formation measured within 6 months of release, and dissolved CH4 measured within approximately 1 year [43]. Soils surrounding the spill zone are comprised of silty sand and loam. Following spring rainfall, ponded conditions occur. During drier periods, the water table is approximately 1m below ground surface. 44  Figure 3.1 Cambria denatured fuel grade ethanol spill site. Black dots indicate the sampling location of soil core A, B, C, D and E.  3.2.2 Sample collection, nucleic acid extraction and cDNA synthesis Sample collection took place in October 2012 using a GeoProbe. Samples were collected in at 10cm intervals from 5 soil profile cores capturing areas that had been impacted by the spill and those left unaffected (Figure 3.1 and Table 3.1). Total genomic DNA and RNA were isolated using the PowerMax® Soil DNA and PowerSoil® RNA Isolation Kits, respectively (MoBio, Carlsbad, CA) at 10cm intervals along the length of the soil core. Resultant yields were quantified using the Picogreen and Ribogreen assays (Invitrogen). RNA was purified using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA) and an on-column DNase I digestion was applied to remove DNA contamination from RNA. Total RNA was reverse transcribed to 45 complementary DNA (cDNA) using a Superscript® III first-strand synthesis kit (Invitrogen, Carlsbad, CA) with random hexamers. To ensure final cDNA was free of DNA contamination, cDNA synthesis reactions were performed without reverse transcriptase and then evaluated by PCR.   Table 3.1 Sampling summary for the Cambria DFE spill site Soil Core Location characteristics Number of samples  Core A Reference site, up gradient of the spill zone 17 Core B Source site, within the spill zone 11 Core C Source site, within the spill zone 21 Core D Source site, within the spill zone 19 Core E Source site, down gradient of the spill zone 20  3.2.3 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing The V6 – V8 region of the 16S small subunit (SSU) rRNA gene was amplified from both DNA and cDNA template using the universal primer pair 926F (5′-AAACTYAAAKGAATTGRCGG-3′) and 1392R (5′-ACGGGCGGTGTGTRC-3′) to generate molecular barcodes capturing microbial community diversity [64].  Primers were modified to include 454 adaptor sequences and barcodes were added to the reverse primers to allow multiplexing during sample sequencing. Duplicate 25µL PCR reactions were performed for each sample with each well containing 2.5µL 10X buffer, 1µL 50mM MgSO4, 0.5µL 10mM dNTPs, 0.5µL 10µM forward and reverse primers, 2µL DNA or cDNA template, 0.2µL Platinum Taq 46 high fidelity polymerase (5U / µL) and 18.3µL of nuclease free water. Negative controls were included for each sample reaction to confirm that no DNA contamination had occurred. Thermal cycler protocol started with denaturation at 95ºC for 3 minutes, followed by 30 cycles of additional denaturation at 95ºC for 30 seconds, annealing at 55ºC for 45 seconds, and extension at 72ºC for 90 seconds. 10 minutes at 72ºC for final extension completed the amplification process. PCR products were visualized on a 1% 1X TBE agarose gel run at 100V for 35 minutes. Successful PCR products were pooled, cleaned using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel), eluted in 30μL of 5mM Tris buffer pH 8.5, and quantified following the Quant-iT Picogreen dsDNA kit (ThermoFisher) protocol. Each barcoded amplicon was pooled at 100ng DNA prior to sequencing. Emulsion PCR and sequencing were performed at Genome Quebec (Montreal, Canada) on the Roche 454 GS FLX Titanium platform (454 Life Sciences, Branford, CT, USA) according to manufacturer protocol.   3.2.4 Sequence processing 687,124 SSU rDNA and 1,281,546 SSU rRNA Cambria spill site were processed using QIIME version 1.9.0 software package [46]. Sequences with less than 200 bases, ambiguous ‘N’ bases, and homopolymer runs were removed. In total, 531,131 SSU rDNA and 1,020,281 SSU rRNA sequences from the Cambria spill site went through an additional quality filtering using the usearch quality filtering pipeline developed by Robert Edgar and implemented in QIIME. De novo and reference based chimeric sequences were identified via UCHIIME [102] and removed prior to taxonomic assignment. Non-chimeric sequences were clustered at 97% into OTUs with uclust where representative sequences from each cluster were queried against the SILVA 128 ribosomal RNA database [103] using the RDP classifier to assign taxonomy. OTUs present in less than 5% 47 of the total number of samples were removed prior to any downstream analyses to reduce the over prediction of rare taxa [104]. Samples with less than 200 OTUs were also removed from the dataset. All downstream data analyses were performed in R using packages including vegan, phyloseq, pvclust, DESeq2, indicspecies and tidyverse [60, 73, 105, 106]. Data used in this chapter has been clustered and analyzed alongside other datasets used in this thesis. 3.3 Results 3.3.1 Microbial community diversity To identify the microbial community structure and potential activity present within each core sample, we performed 454 pyrotag sequencing of rDNA and rRNA with three-domain primers.  Following sample processing and quality control in QIIME, 3,424 OTUs were used for all downstream analyses. Calculating chao1 indices to generate rarefaction curves revealed the slope for each sample approaching a plateau after 2000-3000 unique OTUs have been identified (Figure 3.2). This suggests that, for these samples, the majority of the microbial diversity present in the sample has been captured with the exception of extremely rarer taxa. There were however, several samples where the slope did not appear to plateau until after this point, indicating inadequate coverage. Nevertheless, of the total 3,424 unique OTUs identified across all samples, 3273 (95.6%) were bacterial, 59 (1.7%) were archaeal, 91 (2.66%) were eukaryotic and 1 (0.03%) was unclassified using the SILVA 128 database. To determine the diversity of each core, the Shannon diversity index was used to calculate alpha diversity (Figure 3.3). Alpha diversity was consistent across the entire site, with all cores revealing a similar Shannon diversity index value (Figure 3.3).   48  3.3.2 Microbial community structure Microbial community structure was determined using non-metric multidimensional scaling (NMDS) with VST-normalized data. Most noticeably, some separation between Core A rDNA and Core A rRNA samples was observed, suggesting a potential difference in which taxa are most active within this uncontaminated core (Figure 3.9). rDNA and rRNA samples collected from the remaining cores (Core B, C, D, and E) all clustered together and revealed no distinct separation by core. This suggests that the microbial community structure within these cores are similar, likely due to them all being collected either within or in closer proximity to the contaminated spill zone (Figure 3.1, Figure 3.9).  A PERMANOVA comparing the microbial community structure at the site revealed no significant differences between each sampling core. Furthermore, indicator species analyses failed to reveal any unique taxa present within one core.   49  Figure 3.2 Rarefaction curve for sequences generated from the Cambria denatured fuel grade ethanol spill site    50  Figure 3.3 Diversity estimates for the Cambria  denatured fuel grade ethanol spill site. A) Shannon diversity index across each sampling cores B)Venn diagram comparing the number of OTUs shared between each sampling core.  51   Figure 3.4 NMDS ordination plot for rDNA and rRNA samples generated from the Cambria denatured fuel grade ethanol spill site. 52  Figure 3.5 NMDS ordination plot for samples collected from the Cambria denatured fuel grade ethanol spill site. Plots have been separated by sampling core and colour-coded by zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm.53  3.3.3 Microbial community abundance and activity To evaluate microbial community structure across the spill site, the taxonomoic composition of the community was evaluated at the phylum and order level. At the phylum level, trends in community abundance (rDNA) and potential activity (rRNA) revealed Betaproteobacteria, Euryarchaeota and Firmicutes as the most abundant and active members of the community (Figure 3.6) across the entire site. Comparing the reference site to the contamination source site again revealed a dominance in the same phyla, however there was a decrease in the potential activity of Betaproteobacteria taxa and an increase in the abundance of Euryarchaeota taxa (Figure 3.7).  Within each soil core we again observe the dominance of the same phyla however, across the entire site (Core A – Core E) trends in the distribution of these groups begins to emerge (Figure 3.8).  Indeed, evaluating Euryarchaeota and Firmicutes across the spill site reveals an increase in both the relative abundance and potential activity of these groups in the source site (Core B – E) compared to the reference site (Core A) (Figure 3.8).   54  Figure 3.6 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site 55  Figure 3.7 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site comparing the reference site to the contamination source site  56  Figure 3.8 Phylum level rDNA and rRNA taxonomic summary for the Cambria denatured fuel grade ethanol spill site within each sample depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core.  57   3.3.4 Responsive taxa To better resolve microbial community structure across depth and time, the taxonomic composition of the community was evaluated for orders within the Proteobacteria, Firmicutes, and Euryarchaeota phyla. These phyla were chosen due to their relevance in both aerobic and anaerobic degradation processes as well as due to their dominance across all soil cores (Figure 3.8)).  For the Proteobacteria phylum, Delta- and Betaproteobacteria sub-phyla were the most abundant and active. Within the Betaproteobacteria, we observed congruent trends in the abundance (rDNA) and potential activity (rRNA) indicating that taxa that were abundant were also likely active (Figure 3.10). Across all cores, phylotype SC-1-84 was dominant in both abundance and potential activity, most noticeably within the saturated zone (>100cm depth) in Cores A, C, and E.  Other groups, such as Nitrosomonadales, Burkholderiales and Hydrogenophilales, displayed variable trends in abundance and activity across each of the cores suggesting that there may be differences in the physical conditions within these soil cores that are influencing microbial community structure. This is most clearly observed with Burkholderiales which are abundant throughout Core A and B however are not observed within the remaining soil cores (Figure 3.10).  58  Figure 3.9 rDNA and rRNA taxonomic summary for taxa within the Betaproteobacteria sub-phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core.   Within Deltaproteobacteria, Bdellovibrionales and Desulfovibrionales were most abundant and active across all soil cores (Figure 3.10). In the rRNA profile, Bdellovibrionales was most active in the Cores B – E, accounting for up to 10% of the potential activity across the length of the core in some instances. A similar trend was also displayed with Desulfovibrionales which had increased potential activity in the source site compared to the reference site (Core A). More variable trends were observed with the remaining taxa with groups being abundant and/or active at specific locations within the soil core.  Myxococcales displays this trend as they were largely present within Core C and D, and most active at the water table (~100cm) or within the saturated zone (>110cm) (Figure 3.10). 59  Figure 3.10 rDNA and rRNA taxonomic summary for taxa within the Deltaproteobacteria sub-phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core.  For orders within the Firmicutes, we again observe certain taxa become more abundant and active in soil cores collected within or downstream of the spill zone. For example, both Clostridia D8A-2 and Bacillales displayed this trend in both rRNA and rDNA profiles. Interestingly, Bacillales activity was largely constrained to soil cores B, C, D, and E which are most directly impacted by the E95 spill (Figure 3.11).  60  Figure 3.11 rDNA and rRNA taxonomic summary for taxa within the Firmicutes phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core Within Euryarchaeota, we observe an increase in most taxa in the source site cores compared to the reference site. Methanomicrobiales, Methanosarcinales and Methanobacteriales were abundant across all soil cores, though most dominantly in cores B to E. The dominance of these groups within the most heavily impacted cores supports the observations of acetate and methane being detected within the groundwater [36, 43] (Figure 3.12). Specifically, the abundance and activity of Methanosarcinales suggests acetoclastic methanogenic processes occurring within the spill zone as one source of methane.  61  Figure 3.12 rDNA and rRNA taxonomic summary for taxa within the Euryarchaeota phylum across each sampling depth zone (vadose zone 10cm – 90cm; water table 100cm; and saturated zone  > 100cm) for each core 62 3.4 Discussion  In the Cambria field study, we observed the presence of taxa with traits associated with hydrocarbon degradation, C1 carbon utilization and fermentation. Additionally, we observe dominance in Euryarchaeota and methanogenic taxa indicating that inherent characteristics of the Cambria site were conducive to supporting methanogenic taxa. Here, the presence and activity of methanogens, alongside the detection of acetate and methane in the groundwater, suggests the presence of anaerobic degradation processes e.g. ethanol conversion to simpler products, such as CO2 or acetate, which can then be used to power methanogenesis [36]. Indeed, the enrichment of methanogenic taxa indicates that the full anaerobic degradation process was occurring with primary degradation and fermentation being completed by bacteria within the Betaproteobacteria, Deltaproteobacteria and Firmicutes.  Within the Betaproteobacteria, while we did not observe strong differences in community structure between the reference site and source site (Figures 3.4, 3.5, and 3.9), the presence of taxa such as Burkholderiales, suggests the initial stages of hydrocarbon degradation are taking place [107]. Indeed, Burkholderiales taxa are known hydrocarbon degraders and are frequently found in similarly impacted environments [33, 107, 108]. Similarly, with Desulfovibrionales we observe an increase in the potential activity of this group across all soil cores collected within the contamination source site (Figure 3.10). The presence of Desulfovibrionales at this site provides another mechanism for attenuation as sulphate reducing bacteria are capable of growing on environmental contaminants like ethanol-blended fuels [109]. Additionally, sulphate reducers commonly use by-products such as lactate, acetate and butyrate as a main carbon source produced by fermenter such as Bacillales and Clostridales taxa, both present at this site [28, 30, 63 109].  By evaluating the relative abundance and potential activity of groups within the Betaproteobacteria, Deltaproteobacteria and Firmicutes phyla we uncover the likely mechanism for contaminant degradation and natural attenuation at this site. Additionally, the initial presence of these groups within the reference site indicates that the soil environment supported complete anaerobic degradation. Indeed, the importance of intrinsic soil properties in shaping microbial communities which must also be considered when evaluating microbial community responses to anthropogenic perturbations and any subsequent remediation efforts [108, 110-112]. 64  3.5 Conclusion In this chapter, the microbial community response to a large scale denatured fuel grade ethanol spill was evaluated at the Cambria field site. At this site, we observed the dominance of taxa within the Proteobacteria, Euryarchaeaota, and Firmicutes phyla with traits relevant to fermentation and anaerobic hydrocarbon degradation. Comparison between the reference site and source site revealed an enrichment in methanogenic archaea, sulphate-reducing taxa, and fermenting taxa as the community responds to the contamination. The relative abundance and potential activity of these groups indicates that complete anaerobic degradation is occurring at the spill site. This is important for remediation purposes as incomplete degradation can lead to the accumulation of undesirable by-products with their own negative effects.   65 Chapter 4: Evaluating microbial community responses to ethanol-blended biofuels As production of ethanol-based biofuels has increased in recent years, so too has the likelihood of accidental spills into the environment resulting in lasting ecological implications.  In this chapter, I evaluate the microbial community response to ethanol blended fuels through laboratory experiments. This work provides an initial characterization of the response of the microbial community upon exposure to ethanol blended fuels across different ethanol concentrations.    4.1 Introduction As the world grapples with ways to provide more sustainable sources of energy while also accommodating growing energy demands, the use of biofuels has entered into the conversation as a transition solution to both [34, 35]. Indeed, in Canada biofuel consumption increased to 2,800 million litres in 2018; a trend likely to continue into the future [35]. With increasing production rates likely to continue, so too does the likelihood of accidental spills into the environment often with lasting ecological implications [36-38]. In response, numerous studies have sought to understand the fate of ethanol blended biofuels in subsurface environments by evaluating the response of the indigenous microbial communities through field and laboratory testing [37-40, 42, 100, 101]. Broadly, these studies have revealed a decline in diversity paired with shifts in the microbial community structure upon exposure. Indeed, microbial taxa with abilities to degrade complex organic material have been identified alongside known methanogens [39] to complete canonical anaerobic degradation processes in subsurface environments. However, as evaluation of these communities has often only been performed at 66 the DNA level to assess microbial abundance less is known about potential microbial activity [20, 37-40, 42, 101] using rRNA as a proxy. Additionally, a thorough understanding of the microbial response across a range of biofuel blends (E20 – 20% ethanol versus E85 – 85% ethanol) is lacking, as most studies have focused on ethanol based biofuels at one particular blend [113].  In this chapter, I surveyed microbial community response to a biofuel contamination event over a range of biofuel blends and compositions in a controlled release laboratory experiment. The abundance and potential activity of the microbial community was evaluated under different biofuel blend treatments at time points 4, 8 and 12-months post spill based on 16S SSU rRNA and rDNA amplicon sequencing. These experiments allow for rigorous comparisons on the impact of different ethanol blended fuel treatments on microbial community composition, testing the hypothesis that different biofuel blends drive unique shifts in community structure.   4.2 Methods 4.2.1 Column design and construction To monitor the natural attenuation of different ethanol fuel blend ratios in subsurface environments, a series of column experiments were designed to study these processes in a controlled, laboratory setting (Figure 4.1). Each column measured 2 metres in height and 30 centimeters in diameter, were made from acrylic and polyvinyl materials, and were designed to represent an unconfined aquifer with a 35cm saturated zone and the 165cm vadose zone [114]. Gas sensors for O2, CH4 and CO2 were installed at different vertical locations along the column alongside sampling ports for additional geochemical measurements [114].  67  Figure 4.1 Schematic diagram of the column design used in the ethanol blended fuel experiments  4.2.2 Sample collection and geochemical monitoring Soil samples were collected from each column experiment by drilling an approximately 2.5 cm diameter hole into each column at specific distances away from the base. Once the column was accessible to sampling, 5 – 10 grams of material was removed, flash frozen on dry ice and stored at -80ºC for later processing. Sampling holes were sealed with rubber stoppers and epoxy to prevent any additional air ingress to maintain the integrity of the column experiment. Samples were collected at 35cm (saturated zone), 70cm, 95cm and 145cm (vadose zone) along 68 the length of the 2 metre column at 6-month time intervals over the duration of the 18-month experiment (Figure 4.1, Table 4.1, and Table 4.2). A reference sample was also collected prior to the introduction of the ethanol blended fuel treatment. In total, 164 samples were collected at three time points (October 2015 - 16 weeks post, January 2016 – 28 weeks post, and April 2016 – 40 weeks post) for the ethanol columns experiments. Geochemical measurements were taken over the duration of the experiment. Within the vadose zone (40 cm – 145 cm from the base of column), gas measurements were collected for CO2, CH4, O2, and volatile fatty acid (acetic, propionic and butyric acid) concentrations were measured in the saturated zone (0 – 35 cm from the base of the column).   4.2.3 Nucleic acid extraction and cDNA synthesis DNA and RNA were extracted simultaneously using a modified version of a protocol developed by Lee and Hallam [115] and Hurt et al. [116]. Modifications included the use of the QIAGEN AllPrep DNA/RNA kit to separate RNA from DNA, and the addition of a DNA digestion step to ensure no contaminating DNA was present in the RNA samples. Following extraction, RNA was converted to cDNA using the Invitrogen Superscript III first-strand synthesis kit with random hexamers. A negative control, that did not contain the cDNA synthesizing enzyme, was included and amplified alongside the synthesized cDNA to confirm that the RNA did not contain any contaminating DNA.   4.2.4 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing The V6 – V8 region of the 16S small subunit (SSU) rRNA gene was amplified from both DNA and cDNA template using the universal primer pair 926F (5′-69 AAACTYAAAKGAATTGRCGG-3′) and 1392R (5′-ACGGGCGGTGTGTRC-3′) to generate molecular barcodes capturing microbial community diversity [64].  Primers were modified to include 454 adaptor sequences and barcodes were added to the reverse primers to allow multiplexing during sample sequencing. Duplicate 25µL PCR reactions were performed for each sample with each well containing 2.5µL 10X buffer, 1µL 50mM MgSO4, 0.5µL 10mM dNTPs, 0.5µL 10µM forward and reverse primers, 2µL DNA or cDNA template, 0.2µL Platinum Taq high fidelity polymerase (5U / µL) and 18.3µL of nuclease free water. Negative controls were included for each sample reaction to confirm that no DNA contamination had occurred. Thermal cycler protocol started with denaturation at 95ºC for 3 minutes, followed by 30 cycles of additional denaturation at 95ºC for 30 seconds, annealing at 55ºC for 45 seconds, and extension at 72ºC for 90 seconds. 10 minutes at 72ºC for final extension completed the amplification process. PCR products were visualized on a 1% 1X TBE agarose gel run at 100V for 35 minutes. Successful PCR products were pooled, cleaned using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel), eluted in 30μL of 5mM Tris buffer pH 8.5, and quantified following the Quant-iT Picogreen dsDNA kit (ThermoFisher) protocol. Each barcoded amplicon was pooled at 100ng DNA prior to sequencing. Emulsion PCR and sequencing were performed at Genome Quebec (Montreal, Canada) on the Roche 454 GS FLX Titanium platform (454 Life Sciences, Branford, CT, USA) according to manufacturer protocol.     Table 4.1 Sampling summary for the ethanol blended fuel column experiment Experiment Total number of samples Sampling Sampling dates 70 collected  locations (T1,T2,T3) Ethanol blended fuel 164 35cm (saturated zone)  70cm, 95cm, 145cm (vadose zone) October 2015 (16 weeks post), January 2016 (28 weeks post), April 2016 (40 weeks post)    Table 4.2 Treatment and sampling summary for the ethanol blended fuel column experiment Treatment name Treatment details Column material Number of samples collected EtOH 1 E85 (85% ethanol with 15% gasoline) Sand 21 EtOH 2 E20 (20% ethanol with 80% gasoline) Sand 21 EtOH 3 Gasoline with delayed E20 (E20 was added 8 weeks after gasoline was added to the column) Sand 21 71 EtOH 4 Gasoline Sand 20 EtOH 5 E10 (10% ethanol with 90% gasoline) Sand 20 EtOH S/S1 E20 (20% ethanol with 80% gasoline) Silty sand 20 EtOH S/S2 Gasoline with delayed E20 Silty sand 20 EtOH S/S3 Gasoline Silty sand 21   4.2.5 Sequence processing 822,738 SSU rDNA and 798,710 SSU rRNA pyrotag sequences from the ethanol column experiments site were processed using QIIME version 1.9.0 software package [46]. Sequences with less than 200 bases, ambiguous ‘N’ bases, and homopolymer runs were removed. In total, 788,818 SSU rDNA and 775,709 SSU rRNA sequences from the ethanol column experiments went through an additional quality filtering using the usearch quality filtering pipeline developed by Robert Edgar and implemented in QIIME. De novo and reference based chimeric sequences were identified via UCHIIME [102] and removed prior to taxonomic assignment. Non-chimeric sequences were clustered at 97% into OTUs with uclust where representative sequences from each cluster were queried against the SILVA 128 ribosomal RNA database [103] using the RDP classifier to assign taxonomy. OTUs present in less than 5% of the total number of samples were removed prior to any downstream analyses to reduce the over prediction of rare taxa [104]. 72 Samples with less than 200 OTUs were also removed from the dataset. All downstream data analyses were performed in R using packages including vegan, phyloseq, pvclust, DESeq2, indicspecies and tidyverse [60, 73, 105, 106]. Data used in this chapter has been clustered and analyzed alongside other datasets used in this thesis. 4.3 Results 4.3.1 Geochemistry To characterize the geochemical conditions occurring within each column experiment, time-resolved CO2, O2 and CH4 gas concentrations were measured within the vadose zone (40cm – 200cm) and volatile fatty acid production was measured within the saturated zone (0cm – 35cm). Over the duration of the experiment, we observed strong differences in CO2 and O2 gas concentrations for each column (Figures 4.2 – 4.3). Noticeably, by evaluating the CO2 concentrations within the vadose zone, there were certain columns that appeared to be relatively inactive, in that there was little to no CO2 generation. This was most evident in EtOH2 (E20) where there was no CO2 generation suggesting that most of the blended fuel was not being degraded. Likewise, EtOH1 (E85) and EtOH S/S3 (silty-sand gasoline) also showed less CO2 generation in comparison to EtOH3 or EtOH4 which appeared to be quite active over the course of the experiment (Figure 4.3). Columns EtOH1, EtOH2 and EtOH S/S3 maintain near constant oxygen levels throughout the entire column (Figure 4.2). This is congruent with the accompanying CO2 gas concentration as little degradation occurred as these columns remained oxic (Figure 4.3). CH4 gas concentrations were also measured however, most columns produced quantities that were quite low and near the detection limit of the instrument (Figure 4.4).   Acetic, propionic and butyric acid concentrations, volatile fatty acids (VFAs), were measured within the saturated zone and varied between columns. For example, EtOH1, EtOH2, 73 EtOH S/S1 and EtOH S/S2 failed to produce any concentration of VFAs over the course of the entire experiment (Figure 4.5). In contrast, the remaining columns (EtOH3, EtOH4, EtOH5 and EtOH S/S3) all produced varying amounts of acetic and butyric acid during the experiment but no propionic acid (Figure 4.5). The variation between columns can be related to the ability to anaerobically degrade the contaminant but also to the overall stress of the microbial community [101].  74  Figure 4.2 O2  gas concentration measurements for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline. 75  Figure 4.3 CO2 gas concentration measurements for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline 76  Figure 4.4 CH4  gas concentration measurement for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Gasoline,  EtOH4 – Delayed E20, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline 77  Figure 4.5 Volatile fatty acid concentrations (ppm) for each ethanol column treatment. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Delayed E20,  EtOH4 – Gasoline, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline 78  4.3.2 Microbial community diversity To identify the microbial community structure and potential activity present within each ethanol blended fuel column experiment, we performed 454 pyrotag sequencing of rDNA and rRNA using three-domain primers. Following sample processing and quality control in QIIME, 6,100 OTUs were used for all downstream analyses. Calculating chao1 indices to generate rarefaction curves reveals the slope for each sample approaching a plateau after 3000 – 4000 unique OTUs have been identified (Figure 4.6). This suggests that, for these samples, the majority of the genetic diversity present in the sample has been captured with the exception of some rarer taxa. Of the total 6,100 unique OTUs identified across all ethanol column samples, 5839 (95.7%) were bacterial, 75 (1.2%) were archaeal, 185 (3%) were eukaryotic and 1(0.02%) was unclassified using the SILVA 128 database. To determine the diversity of samples collected from the ethanol column experiments, I employed the Shannon diversity index to calculate alpha diversity. Comparing reference samples to samples that had been exposed to the ethanol blended fuel treatment reveals a significant decrease in alpha diversity (p-value=0.0005) (Figure 4.7). This trend was also identified across all ethanol column treatments individually as reference samples maintained higher diversity over the ethanol treated samples. Comparing the unique OTUs identified within each ethanol treatment revealed that most OTUs were shared across samples collected from both sand and silt columns. Comparison of OTUs recovered specifically from silt or sand material columns also revealed the same trend with the majority of OTUs being shared across both material types (Figure 4.7)79   Figure 4.6 Rarefaction curves for sequences generated from the ethanol blended fuel column experiments. EtOH1 – E85, EtOH2 –E20,  EtOH 3 – Delayed E20,  EtOH4 – Gasoline, EtOH 5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline. 80   Figure 4.7 Diversity estimates for the ethanol column experiments. A) Shannon diversity index comparing reference and contaminated samples across all ethanol column treatments. B) Shannon diversity index comparing reference and contaminated samples from within each column treatment, C) Venn diagram indicating the number of OTUs shared both within and between the sand and silt material ethanol column treatments.81  4.3.3 Microbial community structure To survey the microbial community structure within the ethanol blended fuel column experiments, we used non-metric multidimensional scaling (NMDS) with data that had been normalized using a variance stabilizing transformation. Although each column experiment differed either in ethanol treatment or sediment material (sand versus silt material), there was no distinct separation or clustering of samples based on either criterion (Figure 4.8). Additionally, samples did not appear to be clustered according to sampling depth or time point (Figures 4.9 – 4.10). These results suggest that the microbial community is structured similarly, despite differences in ethanol treatment or sediment material. This result is supported by a PERMANOVA comparing the microbial community structure within each ethanol column which revealed no significant differences between treatments (p-value > 0.05). 82  Figure 4.8 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments   83  Figure 4.9 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments. Plots have been separated by column treatment and colour-coded by sampling timepoint. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline 84  Figure 4.10 NMDS ordination plot for samples collected from the ethanol blended fuel column experiments. Plots have been separated by column treatment and colour-coded by sampling depth. EtOH1 - E85, EtOH2 - E20, EtOH3 - Delayed E20, EtOH4 - Gasoline, EtOH5 - E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline.   85 4.3.4 Microbial community abundance and potential activity To better resolve microbial community structure across depth and time, the taxonomic composition of the community was evaluated at the phylum and order level. At the phylum level, Proteobacteria (specifically Gamma- and Betaproteobacteria), Actinobacteria, Acidobacteria, Firmicutes and Chloroflexi were the dominant members of the community in both abundance (rDNA) and potential activity (rRNA) (Figure 4.11). Furthermore, comparing the microbial community structure across all ethanol column treatments identified strong taxonomic similarity at the phylum level (Figures 4.12 – 4.13). Comparing the community composition between reference and ethanol blended fuel treated samples revealed shifts in the relative abundance and potential activity of Gammaproteobacteria, Firmicutes and Chloroflexi, a trend consistent across all treatments (Figure 4.14).  86  Figure 4.11 Phylum level rDNA and rRNA taxonomic summary for the ethanol blended fuel column experiments.   87  Figure 4.12 Phylum level rDNA and rRNA taxonomic summary across all sampling depths for each ethanol blended fuel column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline 88  Figure 4.13 Phylum level rDNA and rRNA taxonomic summary across all timepoints for each ethanol blended fuel column treatment within the saturated zone (35cm) and vadose zone (75cm, 90cm, 145cm). EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline   89  Figure 4.14 Phylum level rDNA and rRNA taxonomic summary comparing reference and treatment samples for the ethanol blended fuel column experiments.    90 4.3.5 Responsive taxa To evaluate the microbial community response to the ethanol blended fuel treatments, further investigation into the taxonomic composition of certain groups were evaluated at the order level. Specifically, Gammaproteobacteria, Firmicutes and Archaeal phyla were summarized at the order level to identify the specific groups responding to the treatment.  Within the Gammaproteobacteria, Alteromonadales and Oceanospirillales were the most responsive, displaying an increase in relative abundance and potential activity when compared to reference samples (Figure 4.15). Pseudomonadales, Enterobacteriales and PYR10d3 were also among some of the most responsive groups though this varied between treatments. Alteromonadales were most abundant (rDNA) within the EtOH1 – E85 and EtOH2 – E20 treatments making up to 5 – 10% of the total community across all sampling depths (Figure 4.16). For the remaining treatments, the abundance of Alteromonadales varied from 0.1 to 1%. Evaluating the potential activity of this group however reveals that Alteromonadales were active across all treatments, though again, most active in the E85 and E20 treatments. Oceanospirillales and Enterobacteriales also followed a similar trend as both were most abundant and active within the same treatments (Figure 4.16). Alteromonadales and Oceanospirillales were active and abundant across all sampling time points, most noticeably within the vadose zone (>35cm) (Figure 4.17). In contrast, Pseudomonadales and Enterobacteriales displayed variations in activity and abundance over time. Within the saturated zone, Pseudomonadales displayed an increase in abundance and activity at T2 and T3, approximately 7 and 11-months post spill for the EtOH3 (E20 delayed), EtOH5 (E10) and EtOHSS1 (silt E20) treatments (Figure 4.17). Enterobacteriales, on the other hand, appeared to only be active at specific time points such as at T3 for EtOH1 and T2 for EtOH5 and EtOHSS2 (silt delayed E20). These variations in 91 abundance and activity across the different treatments suggest exposure may be impacting specific members of the community differently, despite being present in the same subphyla.   Figure 4.15 rDNA and rRNA taxonomic summary comparing Gammaproteobacteria orders from pristine and contaminated samples collected from the ethanol blended fuel column experiments.  92  Figure 4.16 rDNA and rRNA taxonomic summary for taxa within the Gammaproteobacteria sub-phylum  across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline.   93  Figure 4.17 rDNA and rRNA taxonomic summary for taxa within the Gammaproteobacteria sub-phylum across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Silt Gasoline  Bacillales, Lactobacillales and Clostridiales were the most responsive taxa within the Firmicutes phylum (Figure 4.18). For EtOH1 and EtOH2, the abundance and activity of these groups was maintained across the length of the column, however, this was not observed for the remaining treatments (Figure 4.19). Instead, increased abundance and activity within the saturated zone (35cm) was observed for several groups, most noticeably for Bacillales and Clostridiales, at times making up to 30% of the community (Figure 4.19). Evaluating the 94 dynamics of this group over time reveals variations in relative abundance and potential activity largely within the saturated zone. Specifically, the relative abundance of Bacillales increases over time for the EtOH3, EtOH5 and EtOHSS2 treatments (Figure 4.20). A similar trend was also identified for Clostridiales within the EtOH4 and EtOHSS3 treatments.    Figure 4.18 rDNA and rRNA taxonomic summary comparing Firmicutes orders from pristine and contaminated samples collected from the ethanol blended fuel column experiments. 95  Figure 4.19 rDNA and rRNA taxonomic summary of taxa within the Firmicutes phylum across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline.  Figure 4.20 rDNA and rRNA taxonomic summary of taxa within the Firmicutes phylum across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column 96 treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline  Finally, while archaeal taxa were not highly responsive in these experiments, minor variations in relative abundance and activity were observed across the different ethanol treatments (Figures 4.21 – 4.22). Indeed, archaeal taxa were most abundant (rDNA) within the EtOH4 (gasoline) treatment at 35cm and 70cm comprising up to 1% of the total community across all time points. Additionally, methanogenic taxa Methanocellales and Methanosarcinales were most abundant at 35cm for the EtOH1 and EtOHSS3 treatments. However, while different archaeal groups were identified as being present within the column experiments, rRNA profiles suggest they were not very active (Figure 4.23).      97  Figure 4.21 rDNA and rRNA taxonomic summary comparing Archaeal taxa from pristine and contaminated samples collected from the ethanol blended fuel column experiments.  98  Figure 4.22  rDNA and rRNA taxonomic summary of Archaeal taxa  across each sampling depth for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline.  99  Figure 4.23 rDNA and rRNA taxonomic summary of Archaeal taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each ethanol column treatment. EtOH1 – E85, EtOH2 – E20, EtOH3 – Delayed E20, EtOH4 – Gasoline, EtOH5 – E10, EtOHSS1 – Silt E20, EtOHSS2 – Silt Delayed E20, EtOHSS3 – Gasoline.       100 4.4 Discussion 4.4.1 Microbial community responses to ethanol blended fuels To evaluate the microbial community response to ethanol blended fuels, the relative abundance and potential activity of the microbial community was profiled across a range of ethanol blends in a column laboratory experiment. Through these experiments, we observed shifts in the community towards taxa within Gammaproteobacteria and Firmicutes with hydrocarbon degrading and fermenting abilities. Indeed, across all ethanol treatments, we observe an increase in the relative abundance and potential activity of Alteromonadales, Oceanospirillales and Pseudomonadales, orders with known hydrocarbon degrading pathways and have also been identified as being present and active in similarly impacted environments [20, 32, 37, 38, 42]. Within the Firmicutes phyla, we observed an increase in the abundance and potential activity of Clostridiales and Bacillales, largely within the saturated zone. These known fermenters are likely acting on the ethanol component of the blended fuel treatment, converting ethanol to acetate [117-119]. The conversion of ethanol into acetate can be observed in the EtOH1, EtOH2, EtOHSS1 and EtOHSS2 columns as there was an accumulation of acetic acid detected within the saturated zone, up to 700ppm in certain treatments.   4.4.2 Column experiments Despite differences in the ethanol blended fuels treatment applied to each column, we observe strong similarities in the structure and composition of the microbial community. Indeed, NMDS results revealed no distinct clustering of samples due to treatment or material type, a result supported by a PERMANOVA which determined no significant differences in community structure within each column (p-value > 0.05). Additionally, we did not observe any significant 101 separation between the samples due to depth or sampling time point. This similarity in microbial community structure however is not reflected within the geochemical data which revealed differences in CO2 and O2 gas concentrations and volatile fatty acid production across each column wherein certain columns appeared to be actively degrading the ethanol blended fuel treatment, while others remained dormant and less active over the course of the experiment. This was evident when comparing EtOH1 (E85) to EtOH4 (gasoline) where the former remained fairly inactive and oxic while the later generated CO2 gas indicating degradation was occurring. However, comparison of the microbial community structure however, did not reveal strong difference between the two as there was strong overlap between OTUs identified within each treatment. Differences that are seen within the geochemical profiles but not as dramatically within the microbial community may speak to the resiliency of the community to combat each treatment, and/or the resolution of microbial community profiling.  With respect to resiliency, numerous studies have highlighted the resiliency of microbial communities due to high total genetic diversity [9, 14, 39, 117] . While certain members of the community may be heavily impacted by each treatment, there are also others that remain largely unaffected and contribute to stability of the community. Therefore, the differences observed within the geochemical data from each column treatment may be due to a small subset of the microbial community that is responding and likely driving these differences. This can be observed in the taxonomic composition of Gammaproteobacteria, Firmicutes and Archaea at the order level. While a high degree of similarity between the columns was observed in terms of phyletic groups, more nuanced differences at the order level were observed. For example, increased abundance and activity of Pseudomonadales and Clostridiales (both with known fermentative and hydrocarbon degrading traits) [33, 117] within the saturated zone. These 102 differences have the potential to drive some of the measured trends in CO2, O2 and volatile fatty acid generation.  With respect to the resolution of microbial community profiling, it appears that differences in microbial community structure between columns are more nuanced than previously anticipated. Sequences generated from this dataset were clustered at 97% sequence identity to generate OTUs. While this is common practice, the discrepancy between geochemical monitoring and microbial community composition suggests that this may not be as well suited for this dataset. As differences between each column treatment were subtle, grouping our sequences at a 97% sequence identity may not have been adequate to capture this subtlety. This is in line with a current push away from grouping sequences at 97% similarity and moving towards an alternative method to generate amplicon sequence variants (ASVs) [51].  which eliminates clustering altogether to capture microbial community dynamics at a finer resolution [51]. Finally, although each column experienced a different treatment, there is also a lot that each column shared in common. For example, all columns were built from the same sandy silt subsurface starting material, meaning the resident microbial community across all treatments at T0 was similar, if not near identical. This similarity is highlighted in the NMDS plot (Figures 4.8 – 4.10) which showed no distinct clustering due to treatment, depth or time point. Additionally, though the ratios differed, each column was exposed to a mixture of ethanol and gasoline, with the exception of two columns which only received the gasoline treatment. As high concentrations of ethanol and gasoline/hydrocarbons are detrimental to the growth of many microbial species [82, 120], it is understandable that a similar response was observed across all column treatments as the community initially adjusts to the contaminant. This could explain the strong similarities 103 in microbial community structure across all treatment with more nuanced responses of specific taxa driving differences in gas flux and volatile fatty acid generation as a prelude to methanogenesis.   4.5 Conclusion In this chapter, the microbial community response upon exposure to ethanol blended fuels was evaluated in a laboratory experiment. Across all treatments, we observed enrichment in microbial taxa with traits associated with hydrocarbon degradation and fermentation within the Gammaproteobacteria and Firmicutes phyla and an overall lack of methanogenic activity. Geochemical monitoring revealed differences in O2, CO2 and CH4 gas concentrations in the vadose zone as well as in the generation of volatile fatty acids in the saturated zone. However, no significant differences in microbial community structure within each column were determined using PERMANOVA suggesting that each treatment had equivalent effects on the starting microbial community. While the lack of methanogenic taxa across all column treatments suggests that the complete anaerobic degradation process may not be occurring within this experimental design, the enrichment of hydrocarbon degrading taxa reveals the initial response of the microbial community to the contamination. Indeed, the presence of taxa with traits associated with hydrocarbon degradation shows the overall impact of ethanol blended fuels on microbial community structure in a controlled setting.    104 Chapter 5: Evaluating microbial community responses to methanol-blended biofuels Though the use of methanol as a biofuel has increased in recent years, there has been a lesser focus on the ecological impact of methanol blended fuel spills on the environment. In this chapter, I evaluate microbial community responses to methanol and methanol blended fuels using microcosm and column experiments.    5.1 Introduction The provincial government has established a low carbon fuel standard for British Columbia that includes increasing ethanol content of gasoline and development of biodiesel infrastructure [121]. Indeed, in 2010 alone, British Columbians consumed more than 300 million litres of biofuel with steady increases projected [121]. This trend is reinforced by recent US legislation focused on increasing the biofuel blend ratio in gasoline with implications for future consumption practices throughout North America [122].  Despite the promise of using blended biofuels to offset climate impacts of fossil fuel consumption, recent research has demonstrated that blended biofuel production and transport are not without hazards [120] . Some fuel blends carry the risk of generating methane during degradation that can find its way into buildings and cause health, environmental and fire risks. Moreover, blended biofuels may enhance groundwater transport of, and environmental exposure to carcinogens such as benzene [95, 96, 123]. Emerging environmental regulations are demanding more stringent air and water quality standards related to biofuel production and transport. Consequently, industries as well as regulators need robust monitoring approaches for evaluating risks and assessing natural attenuation potential for blended fuels. 105 In order to address this so far unmet need, it is important to understand what changes blending will introduce to naturally evolved hydrocarbon degrading microbiological processes and how these changes might affect operational and risk management decisions in the future so tools can be developed for monitoring purposes. Our primary objective was to produce baseline information on differences in natural attenuation and vapour intrusion potential across biofuel blends, concentrations, and environmental conditions. The application of genomic technology into mainstream risk assessment and contaminated site management practices can provide an opportunity to establish robust and cost-effective solutions to evaluate and mitigate economic and environmental impact associated with British Columbia’s bioenergy strategy. Industrial partners will be able to make more informed decision based on ecological design principles that will also increase transparency and social license related to blended biofuel production, transport and use in the province, and provide a compass to guide future monitoring and remediation efforts.  While the communities of aerobic and anaerobic microorganisms that respond to traditional hydrocarbon fuel releases are well characterized and have been utilized for natural cleanup such as the Exxon Valdez oil spill, less is known about the consortia that mediate degradation of blended biofuels (with ethanol, methanol, biodiesel, etc.). Metabolic innovations associated with consortia linked to blended fuels likely include anaerobic pathways for conversion of recalcitrant compounds of concern (e.g., benzene) into biomass and energy under conditions of environmental stress (e.g. elevated alcohol content, and decreased pH), and cooperation strategies among microbial community members enabling nutrient and energy exchange (e.g., syntrophic growth). Current understanding based on analyses of small subunit ribosomal RNA (SSU or 16S rRNA) genes reveal differences between communities mediating 106 degradation of select biofuel blends in laboratory settings. However rigorous comparisons across biofuel blend scenarios have not been conducted [113].   Although ethanol is currently the predominant alcohol used in biofuel production, the use of methanol in blended biofuels is under consideration by the energy sector. Methanol, like ethanol, is an alcohol capable of replacing the hydrocarbon derived oxygenate portion of fuels required for a cleaner burn, with a more renewable and sustainable source [124]. Although methanol yields less energy than ethanol during combustion it provides a cleaner burn [125-127]. Studies investigating the impact of ethanol blended fuels on microbial community structure have shown shifts towards fermentative and methanogenic groups, as well as preferential degradation of ethanol over recalcitrant hydrocarbons [37, 38, 42, 81, 101]. Similar studies have yet to be conducted for methanol blended fuels which have a greater potential to drive hazardous methanogenic conditions as methanol can be more readily metabolized to methane over ethanol [128, 129]. Additionally, high concentrations of alcohols, like ethanol and methanol, have been known to exert a toxic effect on microorganism resulting in decreased metabolic activity and cell death; both of which impact biodegradation potential [42].  Therefore, investigations of methanol and methanol blended biofuels on microbial community structure and activity are needed to better evaluate environmental impacts.  Here, I survey microbial community responses to methanol blended biofuels under controlled laboratory conditions using column and microcosm experiments. The abundance and potential activity of the microbial community in columns was evaluated under different biofuel blend treatments at time points 4, 8 and 12-months post spill based on 16S rRNA and rDNA amplicon sequencing. These experiments allow for more rigorous comparisons on the impact of different methanol blended fuel treatments on microbial community structure and activity, 107 testing the hypothesis that different biofuel blends drive unique shifts in community structure with implications for attenuation potential in the environment.   5.2 Methods 5.2.1 Methanol blended fuel experiments 5.2.1.1 Column design and construction Column experiments were designed to monitor degradation of two methanol blend ratios, MeOH1 (M15) and MeOH2 (M85). Columns measuring 2m in height and 30cm in diameter made from acrylic and polyvinyl materials were designed to represent an unconfined aquifer with a 35cm saturated zone and the 165cm vadose zone. Gas sensors for O2, CH4 and CO2 were installed at different vertical locations along the column alongside sampling ports for additional geochemical measurements (Figure 5.1).  108  Figure 5.1 Schematic diagram of the column design used in the methanol blended fuel experiments  5.2.1.2 Sample collection and geochemical monitoring:  Soil samples were collected by drilling an approximately 2.5cm diameter hole into each column at distinct locations from the base (Figure 5.1). Once the column was accessible, between 5-10g of material was removed, flash frozen on dry ice and stored at -80°C. Sampling holes were sealed with rubber stoppers and epoxy to prevent any additional air ingress to maintain integrity of the column experiment. Samples were collected at 35cm (saturated zone), 70cm, 95cm and 145cm (vadose zone) along the length of the 2m column at 6-month time 109 intervals over the duration of the 18-month experiment (Figure 5.1, Table 5.1, and Table 5.2). A reference sample was also collected prior to the introduction of the ethanol blended fuel treatment. In total, 48 samples were collected at three time points for the MeOH2 treatment and at four time points for the MeOH1 treatment column experiments. Geochemical measurements were also taken throughout the time course of the experiment. Within the vadose zone (40-145cm from the base of column), gas measurements were collected for CO2, CH4, O2, and volatile fatty acid (acetic, propionic and butyric acid) concentrations were measured in the saturated zone (0-35cm from the base of the column).  Table 5.1 Sampling summary for methanol blended fuel column and methanol microcosm experiments Experiment Total number of samples collected Sampling locations Sampling dates (T1,T2,T3) Methanol blended fuel 44 35cm (saturated zone)  70cm, 95cm, 145cm (vadose zone) October 2015 (4 months post spill), January 2016, April 2016 Methanol microcosm 20 - Samples collected after 247 day incubation    110  Table 5.2 Sampling tally for the methanol blended fuel column and methanol microcosm experiments Treatment name Treatment details Column material DNA & RNA sample count MeOH 1 M15  (15% methanol  with 85% gasoline) Sand  27 DNA / 25 RNA MeOH 2 M85  (85% methanol  with 15% gasoline) Sand  15 DNA / 17 DNA Microcosm 100ppm methanol Sand 5 DNA / 4 RNA Microcosm 1000ppm methanol Sand 5 DNA / 4 RNA Microcosm 100ppm methanol Silt 5 DNA / 5 RNA Microcosm 1000ppm methanol Silt 5 DNA / 5 RNA   5.2.2 Methanol microcosm experiments 5.2.2.1 Microcosm design Microcosms were constructed by filling a 1L glass bottle with 500g of sediment (either sand or silt) material. Next, bottles were filled with a mixture methanol and water leaving a 100mL headspace to produce final methanol concentrations of 100ppm and 1000ppm. Each bottle was then closed with a rubber stopped, wrapped in aluminum foil and placed on their sides. This approach allowed the water stopped to be covered with the methanol-water mixture thereby minimizing the risk of gas exchange between the headspace and the atmosphere. This 111 was used as a preventative measure for a leaky rubber stopper which would compromise the results of the experiment.   5.2.2.2 Sample collection and geochemical monitoring Samples were collected at the end of the experiment after the microcosms had been opened and dismantled (247-day incubation). Between 5-10g of sediment material was removed from each 1L bottle, flash frozen on dry ice and stored to -80°C until further processing. Concentrations of CO2, CH4, and O2 gases were measured from the headspace throughout the experiment through a gastight syringe and analyzed using gas chromatography. Water samples were also collected via syringe to measure volatile fatty acid, benzene, toluene and methanol concentrations (Table 5.1 and Table 5.2).   5.2.3 Nucleic acid extraction and cDNA synthesis DNA and RNA were extracted simultaneously using a modified version of a protocol developed by Lee and Hallam [115] and Hurt et al. [116]. Modifications included the use of the QIAGEN AllPrep DNA/RNA kit to separate RNA from DNA and the addition of a DNA digestion of the RNA to ensure no contaminating DNA was present in the RNA samples. Following extraction, RNA was converted to cDNA using the Invitrogen Superscript III first-strand synthesis kit with random hexamers. A negative control, that did not contain the cDNA synthesizing enzyme, was included and amplified alongside the synthesized cDNA to confirm that the RNA did not contain any contaminating DNA.  112 5.2.4 PCR amplification of 16S SSU rDNA and rRNA for pyrotag sequencing The V6-V8 region of the 16S small subunit (SSU) rRNA gene was amplified from both DNA and cDNA templates using the universal primer pair 926F (5′-AAACTYAAAKGAATTGRCGG-3′) and 1392R (5′-ACGGGCGGTGTGTRC-3′) [64] to generate molecular barcodes capturing microbial community structure and activity. Primers were modified to include 454 adaptor sequences and barcodes were added to the reverse primers to allow multiplexing during sample sequencing. Duplicate 25µL PCR reactions were performed for each sample with each well containing 2.5µL 10X buffer, 1µL 50mM MgSO4, 0.5µL 10mM dNTPs, 0.5µL 10µM forward and reverse primers, 2µL DNA or cDNA template, 0.2µL Platinum Taq high fidelity polymerase (5U/µL) and 18.3µL of nuclease free water. Negative controls were included for each sample reaction to confirm that no DNA contamination had occurred. Thermal cycler protocol started with denaturation at 95°C for 3 minutes, followed by 30 cycles of additional denaturation at 95°C for 30 seconds, annealing at 55°C for 45 seconds, and extension at 72°C for 90 seconds. 10 minutes at 72°C for final extension completed the amplification process. PCR products were visualized on a 1% 1X TBE agarose gel run at 100V for approximately 35 minutes. Successful PCR products were pooled, cleaned using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel), eluted in 30μL of 5mM Tris buffer pH 8.5, and quantified following the Quant-iT Picogreen dsDNA kit (Thermo Fisher) protocol. Each barcoded amplicon was pooled at 100ng DNA prior to sequencing. Emulsion PCR and sequencing were performed at Genome Quebec (Montreal, Canada) on the Roche 454 GS FLX Titanium platform (454 Life Sciences, Branford, CT, USA) according to manufacturer protocol.   113 5.2.5 Sequence processing A total of 203,509 SSU rDNA and 151,463 SSU rRNA pyrotag sequences from the methanol column experiments and 53,795 SSU rDNA and 64,069 SSU rRNA from the microcosm experiment were processed using QIIME version 1.9.0 software package [46]. Sequences with less than 200 bases, ambiguous ‘N’ bases, and homopolymer runs were removed. In total, 195,494 SSU rDNA and 106,421 SSU rRNA sequences from the methanol column experiments and 52,236 SSU rDNA and 61,293 SSU rRNA sequences from the microcosm experiment went through an additional quality filtering using the usearch quality filtering pipeline implemented in QIIME. De novo and reference based chimeric sequences were identified via UCHIIME [102] and removed prior to taxonomic assignment. Non-chimeric sequences were clustered at 97% into OTUs with uclust where representative sequences from each cluster were queried against the SILVA 128 ribosomal RNA database [103] using the RDP classifier to assign taxonomy. OTUs present in less than 5% of the total number of samples were removed prior to any downstream analyses to reduce the over prediction of rare and likely chimeric taxa [73, 104]. Samples with less than 200 OTUs were also removed from the dataset. All downstream data analyses were performed in R using packages including vegan, phyloseq, pvclust, DESeq2, indicspecies and tidyverse [60, 73, 91, 105, 106]. Data used in this chapter has been clustered and analyzed alongside other datasets used in this thesis.  114 5.3 Results 5.3.1 Methanol column experiments 5.3.1.1 Geochemical results To characterize geochemical conditions within each column experiment, time-resolved O2, CO2, and CH4 gas concentrations were measured within the vadose zone (40-200cm) and volatile fatty acid production was measured within the saturated zone (0-35cm). Within the vadose zone, trends in O2 and CO2 concentrations varied for each column treatment. In the case of O2, MeOH1 (M15) remained oxic throughout the entire column while MeOH2 (M85) oxygen varied with depth (Figure 5.2). Consistent with this observation, CO2 was negligible for MeOH1 but varied inversely with O2 throughout the column for MeOH2 (M85) (Figure 5.3). These observations suggest that while MeOH1 was inactive during the time course of the experiment MeOH2 was relatively active in hydrocarbon degradation, particularly in the 50-130cm interval (Figures 5.2 – 5.3). Methane was also measured within the vadose zone and highlight differences between column treatments. Methane concentrations between 1-5% of the total gas volume were observed 12-37 weeks post treatment for MeOH2 while MeOH1 produced negligible amounts (Figure 5.4).  Volatile fatty acid profiles for acetic, propionic and butyric acid were measured within the saturated zone and also varied between columns. A relatively small increase in acetic acid between 10-20 weeks was observed in MeOH1. In contrast MeOH2 displayed elevated levels of acetic acid between 0-10 weeks followed by a decline between 10-20 weeks and a secondary increase between 20-25 weeks. MeOH2 also displayed increasing butyric acid between 10-15 and 20-25 weeks (Figure 5.5). Both acetic acid and butyric acid production co-occurred with increased CH4 between 10-35 weeks in MeOH2 consistent with hydrocarbon degradation and methanol conversion. 115  Figure 5.2 O2  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85.   116  Figure 5.3 CO2  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. 117  Figure 5.4 CH4  gas concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85. 118  Figure 5.5 Volatile Fatty Acid concentration measurements for the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 –M85.  5.3.1.2 Microbial community diversity To identify the microbial community structure and potential activity present within each methanol blended fuel column experiment, we performed 454 pyrotag sequencing of rDNA and rRNA with three-domain primers.  Following sample processing and quality control in QIIME, 3,424 OTUs were used for all downstream analyses. Calculating chao1 indices to generate rarefaction curves revealed the slope for each sample approaching a plateau after 2000-3000 unique OTUs have been identified (Figure 5.6). This suggests that, for these samples, the majority of the microbial diversity present in the sample has been captured with the exception of extremely rarer taxa. There were however, several samples where the slope did not appear to plateau until after this point, indicating inadequate coverage. Nevertheless, of the total 3,424 119 unique OTUs identified across all methanol column samples, 3273 (95.6%) were bacterial, 59 (1.7%) were archaeal, 91 (2.66%) were eukaryotic and 1 (0.03%) was unclassified using the SILVA 128 database. To determine the diversity of samples collected from the methanol column experiments, the Shannon diversity index was used to calculate alpha diversity. Comparing reference samples to samples that had been exposed to the ethanol blended fuel treatment reveals a decrease in alpha diversity (Figure 5.7A-B). This trend was also identified within each methanol column treatment (M15 and M85) individually as reference samples maintained higher diversity over treated samples. Comparing the unique OTUs identified within each methanol treatment revealed complete overlap, indicating strong similarities in the microbial community across both columns (Figure 5.7C).   Figure 5.6 Rarefaction curves generated for sequences from the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 – M85.  120  Figure 5.7 Diversity estimates for the methanol column experiments. A) Shannon diversity index comparing reference and contaminated samples across each methanol column treatment. B) Shannon diversity index comparing reference and contaminated samples from within each column treatment. C) Venn diagram indicating the number of OTUs shared both within and between the two column treatments.121  5.3.1.3 Microbial community structure To compare microbial community structure among and between methanol blended fuel column experiments, non-metric multidimensional scaling (NMDS) was conducted with data that had been normalized using a variance stabilizing transformation. Although the two column experiments had different treatments (i.e. M15 vs M85), a strong separation between treatments was not observed (Figure 5.8). Moreover, no distinct separation or clustering of samples based on depth, sampling time point or whether samples were collected from within the vadose zone or the saturated zone was observed (Figure 5.9). These observations were supported by PERMANOVA comparing the microbial community structure between M15 and M85 indicating no significant differences between treatments (p-value > 0.05).   122  Figure 5.8 NMDS ordination plot for samples collected from the methanol blended fuel column experiments.  123  Figure 5.9 NMDS ordination plot for samples collected from the methanol blended fuel column experiments. Plots have been separated by methanol column treatment and colour-coded by sampling depth.  MeOH1 – M15, MeOH2 – M85.   5.3.1.4 Microbial community abundance and activity To better resolve microbial community structure across depth and time, the taxonomic composition of the community was evaluated at the phylum and order level. At the phylum level, trends in community abundance (rDNA) and potential activity (rRNA) were congruent as Proteobacteria (specifically Gamma- and Betaproteobacteria), Actinobacteria, Acidobacteria, Firmicutes and Chloroflexi were the most abundant and active community members (Figure 5.10). Comparison between reference samples and methanol blended fuel treated samples revealed abundance and activity shifts by Gamma- and Betaproteobacteria, Firmicutes and 124 Chloroflexi (Figure 5.11), a trend consistent across both column treatments in depth and time (Figures 5.12 – 5.13).  Figure 5.10 Phylum level rDNA and rRNA taxonomic summary for the methanol blended fuel column experiments  125  Figure 5.11 Phylum level rDNA and rRNA taxonomic summary for the methanol blended fuel column experiments comparing reference/pristine and contaminated samples  126  Figure 5.12 Phylum level rDNA and rRNA taxonomic summary across each sampling depths for the methanol blended fuels column experiment. MeOH1-M15, MeOH2 – M85 127  Figure 5.13 Phylum level rDNA and rRNA taxonomic summary across all timepoints for each methanol blended fuel column treatment within the saturated zone (35cm) and vadose zone (75cm, 90cm, 145cm). MeOH1 – M15, MeOH2 – M85.   128 5.3.1.5 Responsive taxa Based on the abundance and activity results, abundant groups were evaluated at the order level to provide more granular information on potential trait-based responses. Emphasis was placed on Proteobacteria (Gamma- and Betaproteobacteria) and Firmicutes sub-groups responsive to each treatment.  Within the Gammaproteobacteria, Alteromonadales, Enterobacteriales, Oceanospiralles, Pseudomonadales, PYR10d3 and Xanthomonadales were most responsive, displaying shifts in both abundance and activity (Figure 5.14). When evaluating the abundance (rDNA) of these orders, for the M15 treatment, Alteromonadales, Enterobacteriales and Oceanospiralles were most abundant within the vadose zone samples (70cm, 95cm and 145cm) compared to the saturated zone (Figures 5.15 – 5.16). Pseudomonadales, PYR10d3 and Xanthomonadales however, were equally abundant in both saturated zone (35cm) and vadose zone samples. For the M85 treatment, most groups were abundant at 35cm and 70cm, with the exception of Pseudomonadales which was equally abundant at 95cm. Comparing the activity (rRNA) of these two orders across both treatments revealed strong similarities with most groups being active throughout the entire column (Figures 5.15 – 5.16). A noticeable difference between the two column treatments was found with Alteromonadales and Enterobacteriales taxa which displayed different patterns in potential activity. Alteromonadales were more active in the M85 treatment over the M15 treatment, specifically at 35cm, 70cm and 95cm. Enterobacteriales on the other hand were most active within the vadose zone of the M15 treatment over the M85 treatment. Within the M85 treatment, Enterobacteriales was most active in the saturated zone at 35cm. Evaluating the dynamics of these groups over time revealed similar abundance and activity at each time point with the exception of Xanthomonadales and Pseudomonadales. Both groups 129 were most abundant in the vadose zone at the T2 time point, approximately 9-months post M15 spill (Figures 5.15 – 5.16).  Figure 5.14 rDNA and rRNA taxonomic summary comparing Gammaproteobacteria orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. MeOH1 – M15, MeOH2 – M85. 130   Figure 5.15 rDNA and rRNA taxonomic summary of Gammaproteobacteria taxa  across each sampling depth for each methanol column treatment. MeOH1 – M15, MeOH2 – M85.  131  Figure 5.16 rDNA and rRNA taxonomic summary of Gammaproteobacteria taxa  across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, and 145cm) for each methanol column treatment. MeOH1 – M15, MeOH2 – M85  132 Within the Betaproteobacteria, Burkholderiales, Methylophilales and Nitrosomonadales were most responsive to the methanol blended fuel treatments (Figure 5.17). Comparing the relative abundance of these groups revealed similar trends, with Burkholderiales emerging and then increasing across all depths, though most noticeably at 35cm for the M15 treatment. Additionally, Methylophilales were most abundant at 35cm and 70cm and Nitrosomonadales at 70cm and 145cm within the vadose zone; trends observed in both columns (Figures 5.18 – 5.19).  Burkholderiales appeared to be most active within the saturated zone at 35cm, however for the M85 treatment, this actually represented a decrease in activity when compared to the reference sample. A similar trend was observed for Nitrosomonadales which was most active in the vadose zone, however this too represented a decrease in activity compared to the reference sample (Figures 5.18 – 5.19). Nitrosomonadales were least active in the saturated zone, while Methylophilales was most active at 35cm and 70cm for the M15 treatment and at 35cm and 90cm for the M85 treatment. Investigating the response of these groups over time revealed interesting trends for both Burkholderiales and Methylophilales (Figures 5.18 – 5.19). Within the M15 treatment, Burkholderiales were most abundant and active within the saturated zone however, the activity of this group increased approximately 20-fold after the T3 time point (approximately 12-months post spill). This is in contrast to the M85 treatment where the group was most active in the vadose zone at the T1 and T2 time points, 4 and 9 months after the spill respectively. Methylophilales were more abundant and active in the M85 treatment over the M15 treatment, in both vadose and saturated zones. In both regions of the column, Methylophilales was most activity at the T2 time point approximately 9-months after initial methanol exposure.  133  Figure 5.17 rDNA and rDNA taxonomic summary comparing Betaproteobacteria orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. 134  Figure 5.18 rDNA and rRNA taxonomic summary of Betaproteobacteria taxa across each sampling depth for each methanol column treatment. MeOH1 – M15, MeOH2 – M85. 135  Figure 5.19 rDNA and rRNA taxonomic summary of Betaproteobacteria taxa  across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, and 145cm) for each methanol column treatment. MeOH1 – M15, MeOH2 – M85  Within the Firmicutes, Bacillales, Lactobacillales and Clostridales were the most abundant and active groups when compared to the reference sample (Figure 5.20). The abundance and activity of these groups was relatively stable throughout the time course of the column experiments with the exception of Bacillales, Lactobacillales and Clostridiales associated with MeOH2 (Figures 5.21 – 5.22). These orders appeared to be more active in MeOH2 throughout the time course of the experiment than MeOH1, with the potential to contribute to 136 VFA generation including acetic acid and butyric acid via fermentation processes. These compounds can in turn be used as substrates for methanogenesis. Consistent with the production of methane in MeOH2, the abundance of acetate using methanogens affiliated with Methanosarcinales increased from undetected to approximately 1% over the time course of the experiment (Figures 5.23 – 5.25).    Figure 5.20 rDNA and rDNA taxonomic summary comparing Firmicutes taxa from pristine and contaminated samples collected from the methanol blended fuel column experiments. 137  Figure 5.21 rDNA and rDNA taxonomic summary comparing Firmicutes taxa orders across each sampling depth for each methanol column treatment. MeOH1- M15 MeOH2 –M85 138  Figure 5.22 rDNA and rDNA taxonomic summary comparing Firmicutes taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each methanol blended fuel column treatment       139 Finally, within the Euryarchaeota phyla, Thermoplamatales and Methanosarcinales were the most responsive groups across the methanol treatments as both displayed an increase in abundance and potential activity in contaminated samples (Figure 5.23). Indeed, in the MeOH2 column, the relative abundance of Methanosarcinales made up to 1% of the community in the contaminated samples compared to reference samples (Figure 5.23). Along the length of the columns, Thermoplasmatales was present and active in both vadose and saturated zones for the MeOH1 treatment while Methanosarcinales was more constrained to the vadose zone (Figure 5.24). In the MeOH2 treatment, Methanosarcinales was largely present at the 35cm and 70cm sampling depths and appeared less potentially active compared to the MeOH1 treatment. Across time, for Thermoplasmatales, we observe greatest activity and relative abundance at the T1 and T2 time points within the vadose zone of the MeOH1 column, however this group was present across all time points for both treatments. Methanosarcinales displayed more variable trends in 140 the MeOH2 treatment with the group being present across all time points in the saturated zone but only emerging at the T3 time point in the vadose zone (Figure 5.25)    Figure 5.23 rDNA and rDNA taxonomic summary comparing Euryarchaeota orders from pristine and contaminated samples collected from the methanol blended fuel column experiments. 141  Figure 5.24 rDNA and rDNA taxonomic summary comparing Euryarchaeota taxa across each sampling depth each methanol blended fuel column treatment. MeOH1 – M15, MeOH2 – M85  142  Figure 5.25 rDNA and rDNA taxonomic summary comparing Euryarchaeota taxa across each sampling time point within the saturated zone (35cm) and vadose zone (70cm, 95cm, 145cm) for each methanol blended fuel column treatment. MeOH1 – M15, MeOH2 – M85       143 5.3.2 Methanol microcosm experiments 5.3.2.1 Geochemistry To characterize the geochemical conditions occurring within batch experiments, i.e. methanol microcosms, CO2, O2, and CH4 gases, volatile fatty acid, toluene, benzene and methanol concentrations were measured alongside headspace pressure measurements (Figure 5.26) Across both the 100ppm and 1000ppm methanol treatments, methanol concentrations were undetectable at 100 days, indicating that complete methanol degradation had occurred by this point. For the 100ppm treatments, complete consumption of methanol was paired with a sharp increase in pressure from -4 PSI to 18 PSI, at 100 days. For the 1000ppm treatment, a similar trend emerged, wherein complete depletion of methanol coincided with an increase in pressure. However, this increase in pressure did not occur until almost 200 days. The trend in head space pressure measurement also differed from the other two treatments with a step-wise increase in pressure from -10 PSI at day 10 to 30 PSI at day 400 and up to 130 PSI by the end of the experiment for one replicate. The other replicate also displayed this trend though only reached a final pressure of 100 PSI. Across all treatments, complete consumption of methanol was also paired with an increase in acetic acid concentrations, one of the volatile fatty acids measured over the course of the experiment. Acetic acid concentrations did not remain elevated however, and decreased as headspace pressure increased. Compositional analysis of gases collected from the headspace indicated that the pressure increase directly correlated with CH4 production (Figure 5.26).  144  Figure 5.26 Geochemical measurements collected over the course of the methanol microcosm experiments. A) VFA, methanol, and pressure measurements for the 100ppm silt treatment; B) VFA, methanol, and pressure measurements for the 100ppm sand treatment; C) VFA, methanol, and pressure measurements for the 1000ppm silt treatment; D) VFA, methanol, and pressure measurements for the 1000ppm sand treatment; E) CH4, O2, CO2, and pressure measurements for the 100ppm silt sample; F) CH4, O2, CO2, and pressure measurements for the 100ppm sand sample.  145 5.3.2.2 Microbial community diversity To identify the microbial community structure and activity associated with each methanol microcosm experiment, we performed 454 pyrotag sequencing of rDNA and rRNA with three-domain primers.  Following sample processing and quality control in QIIME, 3,424 OTUs were used for all downstream analyses. Calculating chao1 indices to generate rarefaction curves revealed the slope for each sample approaching a plateau after approximately 900 unique OTUs for the sand microcosms and 600 unique OTUs for the silt microcosms (Figure 5.27). This suggests that, for these samples, the majority of the microbial diversity present in the sample has been captured with the exception of extremely rarer taxa. Diversity estimates determined for each sample based on the Shannon diversity index revealed a decrease in alpha diversity associated with both sand and silt microcosms, with the strongest decrease observed under silt growth conditions at 1000ppm (Figure 5.28A). Set difference analysis of the combined data sets indicated that 738 unique OTUs were shared across all samples including 674 (91.3%) bacterial OTUs, 55 (7.5%) archaeal (OTUs), and 9 (1.2%) eukaryotic OTUs based on annotations using the SILVA 128 database (Figure 5.28B – C). However, significant differences in the microbial community structure between sand and silt growth conditions were observed with approximately 25% of the OTUs shared in common (Figure 5.28D). 146  Figure 5.27 Rarefaction curves for sequences generated from the methanol microcosm experiment. 147  Figure 5.28 Diversity estimates for the methanol microcosm experiments. A) Shannon diversity index for the sand and silt methanol microcosms comparing  reference  start material, 100ppm treatment and 1000ppm treatment samples.  B) Venn diagram comparing OTUs shared between the sand microcosm samples. C) Venn diagram comparing OTUs shared between the silt microcosm samples. D) Venn diagram comparing OTUs shared between the sand and silt microcosm samples. 148  5.3.2.3 Microbial community structure and indicator species analysis To survey the microbial community structure among and between microcosm experiments non-metric multidimensional scaling (NMDS) was conducted using OTU distributions normalized using a variance stabilizing transformation. A separation between the silt and sand microcosms as well as separation away from the original reference samples for each methanol treatment was observed (Figure 5.29). Indicator species analysis with 1000 permutations was then performed to determine specific OTUs associated with each condition using a threshold indicator value > 0.7 and p-value < 0.05. A total of 48 indicator OTUs were identified for the sand microcosm and 28 for the silt microcosms. Indicator OTUs associated with the silt microcosm were primarily affiliated with Methanomicrobia (methanogenic archaea) and Acidobacteria while indicator OTUs associated with sand were affiliated with a larger variety of bacterial phyla such as Acidobacteria, Firmicutes, Chloroflexi and Proteobacteria (Table A.1)149    Figure 5.29 NMDS ordination plot for samples collected from the methanol microcosm experiment.   5.3.2.4 Microbial community abundance and potential activity To better resolve microbial community structure across depth and time, the taxonomic composition of the community was evaluated at the phylum and order level. Euryarchaeota, Proteobacteria, Acidobacteria, Actinobacteria, Firmicutes and Chloroflexi were the dominant phyla in both rDNA (abundance) and rRNA (potential activity) fractions in all methanol microcosm experiments (Figure 5.30). Evaluating the phylum level taxonomic composition across the 100ppm and 1000ppm methanol treatments revealed strong similarities in both the sand and silt microcosms as well as shifts in the community structure upon exposure. Indeed, 150 comparing reference samples to those exposed to the methanol treatments revealed shifts in the relative abundance and potential activity of the dominant taxonomic groups (Figure 5.31).   Figure 5.30 Phylum level rDNA and rRNA taxonomic summary for samples collected from the methanol microcosm experiments 151  Figure 5.31 Phylum level taxonomic summary for methanol microcosm experiment across each treatment and starting material.  5.3.2.5 Responsive taxa To evaluate microbial community responses to the ethanol blended fuel treatments, further investigation into the taxonomic composition of dominant groups were evaluated at the order level. Activity shifts within the Gammaproteobacteria were largely constrained to Pseudomonadales and Enterobacteriales orders within the sand microcosm experiments. The silt microcosm experiments did not display any strong shifts in relative abundance or activity for any orders within the Gammaproteobacteria (Figure 5.32). Pseudomonadales were the most active 152 Gammaproteobacteria within the two methanol treatments although there were less active in the 1000ppm treatment than the 100ppm treatment suggesting metabolic inhibition at higher methanol concentrations. For Enterobacteriales however, an opposite trend emerged as there was higher activity in the 1000ppm treatment and nearly 25 times less activity in the 100ppm treatment.   Figure 5.32 rDNA and rRNA taxonomic summary for Gammaproteobacteria taxa across each treatment and starting material for the methanol microcosm experiments.  For Acidobacteria, the Halophagae subgroup 7 order was most responsive in the silt 100ppm and 1000ppm methanol treatments. This group displayed a 60-80 fold increase in both potential activity and relative abundance (Figure 5.33).  153  Figure 5.33 rDNA and rRNA taxonomic summary for Acidobacteria taxa across each treatment and starting material for the methanol microcosm experiments  Within the Firmicutes phylum, increases in the relative abundance and potential activity of Clostridiales and Baciliales was observed across both silt and sand microcosms while a 6-fold increase in Selenomonadales was observed solely in the sand microcosms (Figure 5.34). Firmicutes response patterns were consistent between 100ppm and 1000ppm methanol treatments.  154  Figure 5.34 rDNA and rRNA taxonomic summary for Firmicutes taxa across each treatment and starting material for the methanol microcosm experiments.  Responsive orders within the Euryarchaeota phylum were primarily affiliated with Thermoplasmatales and methanogenic Methanocellales and Methanosarcinales (Figure 5.35). Within the reference material, the relative abundance of these groups differed between the silt and sand samples. Thermoplasmatales comprised up to 20% of the community in the sand reference material and Methanosarcinales comprised up to 15% of the community in the silt reference material. For the sand microcosm, the relative abundance of Thermosplasmatales decreased while the abundance and activity of Methanosarcinales in both the 100ppm and 1000ppm methanol treatments increased from less than 1% to 20% and 10% of the community, respectively. The abundance of Methanocellales in both the 100ppm and 1000ppm methanol treatments also increased from undetectable to 5% and 1% of the community, respectively. In the 155 silt microcosms, Methanosarcinales remained the dominant methanogen in both reference and treatment samples. However, the potential activity of this group increases 10-20 fold across both treatments, a noticeable shift as there was no detected activity in the reference samples.    Figure 5.35 rDNA and rRNA taxonomic summary for Archaeal taxa across each treatment and starting material for the methanol microcosm experiments.  5.4 Discussion 5.4.1 Column experiments To evaluate responses to methanol blended fuels, microbial community abundance and activity was profiled for two methanol blends in column experiments and compared to microcosm growth using pure methanol. Through geochemical monitoring, O2 gas concentration measurements revealed the MeOH1 – M85 treatment column remained oxic over the duration of 156 the experiment, while the MeOH2 – M15 treatment generated both oxic and anoxic conditions. Additionally, CO2 and CH4 gas concentrations reveal the MeOH1 column as relatively inactive with to aerobic or anaerobic metabolisms as no significant production of CO2 or CH4 was measured. The MeOH2 column however, displayed a different trend wherein CO2 and CH4 gas concentrations along the length of the column suggesting both aerobic and anaerobic degradation processes were occurring. Despite differences in gas concentration profiles between the two column treatments, a comparison of the microbial community structure revealed no significant differences (p-value > 0.05). Indeed, through NMDS ordination and PERMANOVA, no statistically significant differences in the microbial communities were identified within each column. This was especially apparent within the NMDS plot as there was no separation between M15 and M85 treated samples. However, although large-scale changes in microbial community structure were not observed in response to column methanol treatments, shifts in the relative abundance and activity of specific taxonomic groups were discerned. Across both column experiments, when comparing reference samples to methanol blended fuel samples, Gammaproteobacteria, Betaproteobacteria and Firmicutes exhibited significant changes in relative abundance and activity. Within the Gammaproteobacteria the responsive taxa included Alteromonadales, Enterobacteriales, Oceanospiralles, Pseudomonadales, PYR10d3 and Xanthomonadales. These orders all displayed an increase in abundance and potential activity suggesting that the presence of methanol blended fuel as a carbon source, stimulated growth and metabolic activity. Indeed, the increased growth and activity of these groups upon exposure to a hydrocarbon has been observed in other similarly impacted environments and has commonly been termed the “gamma-shift” [33]. Moreover, these same taxonomic groups are commonly identified as hydrocarbon degraders based on amplicon 157 and whole genome shotgun sequencing of contaminated sites and in some cases, have even been isolated using hydrocarbons as a carbon source. Within the Betaproteobacteria Burkholderiales exhibited a similar trend in abundance and activity. This order has been identified and enriched in numerous hydrocarbon contaminated soil environments and can even be used as a predictor of degradation processes in certain cases [33, 130]. The increased relative abundance and potential activity of Methylophilales, particularly within the higher methanol treatment is likely due to the ability of this taxa to consume C1 compounds such as methanol, which has been identified in similarly impacted environments [131]. Within the Firmicutes phylum, fermentative taxa including Bacillales, Lactobacillales and Clostridales were abundant and active. Similar to the groups identified within the Gammaproteobacteria and Betaproteobacteria, the enrichment of these groups upon hydrocarbon exposure has been well documented [20, 33, 37, 41, 42, 130]. Together, the enrichment in taxa associated with hydrocarbon degradation reflects canonical, syntrophic degradation processes that often occur in contaminated subsurface environments. In this process, fermentative taxa, such as Clostridiales, first act on the hydrocarbon substrate to produce acetate and other metabolites which can then be used by other secondary degraders and scavengers such as Oceanospiralles and Burkholderiales [33]. Additionally, the presence of Methylophilales suggests that methanol degradation is also occurring within the column, likely in tandem with hydrocarbon degradation. Overall, the increased relative abundance and, even more importantly, potential activity suggests that these groups are adequately equipped to cope with toxicity associated with exposure to methanol blended fuels, even those with relatively high methanol concentrations.    158 5.4.2 Microcosm experiments To evaluate the microbial community response to pure methanol exposure, microcosm bottle experiments were constructed at 100ppm and 1000ppm methanol concentrations. Through geochemical measurements, complete methanol degradation at the 100-day mark was observed paired with an increase in the atmospheric headspace pressure due to methane gas generation. In addition, acetic acid concentrations increased and then subsequently decreased in both 100ppm and 1000ppm treatments, though this trend was more apparent in the 1000ppm treatment. In this treatment, acetic acid concentrations approached 1000ppm, indicative of methanol degradation generating acetate and other secondary metabolites. A decrease in acetic acid concentrations was then met with an increase in atmospheric head pressure and methane gas generation at the 200-day mark. This 100-day delay in methane gas production, when compared to the 100ppm treatment, is likely due to high acetic acid concentrations inhibiting the activity of methanogenic archaea [132].   Microbial community structure within the microcosm experiments was consistent with a shift toward towards one carbon utilizing and methanogenic taxa. A comparison between reference and treated samples revealed enrichment in Gammaproteobacteria, Acidobacteria, Firmicutes and methanogenic Euryarchaeota accompanied by a decrease in Actinobacteria. Further evaluation of these groups at the order level revealed key players in aerobic and anaerobic degradation processes including Pseudomonadales, Clostridiales, Bacillales and Methanosarcinales [133]. The presence of Methanosarcinales is especially relevant in conjunction with significant levels of methane gas generated within the microcosm experiments alongside increases in acetic acid concentrations. Indeed, archaea affiliated with the Methanosarcinales are known to perform acetoclastic methanogenesis where acetate is the 159 primary substrate converted to methane [134]. The increased abundance and potential activity of this group, alongside geochemical characterization suggests acetoclastic methanogenesis as the dominant process in methane generation. This is especially true within the silt microcosm experiments as Methanosarcinales were the dominant group identified in the Euryarchaetoa phylum. For the sand microcosm, Methanocellales were also enriched in the treatment samples, up to 10-fold when compared to reference samples. The presence of Methanocellales within the sand microcosms but not in the silt suggests that there were differences in the microbial community composition of each material from the start. This is noteworthy as it highlights the importance of the physical environment in shaping microbial community composition and, ultimately, function in the context of blended biofuel remediation or attenuation processes.  Together, it is likely that the anaerobic degradation of methanol follows the well-defined route wherein methanol is first oxidized and degraded to acetate by members within the Firmicutes phylum such as Clostridiales or Bacillales [128]. Acetate can then be used by Methanosarcinales taxa in acetoclastic methanogenesis to generate the high methane gas concentrations that were observed within the headspace.   5.4.3 Comparing methanol column and microcosm experiments  When comparing the results of the column and microcosm experiments, one key difference emerges in the abundance and potential activity of archaeal taxa. Within the methanol column experiments, only the M85 treatment generated methane gas at different locations along the column, suggesting the presence of methanogenesis and methanogenic niches capable of supporting these taxa. This trend however does not appear to be applicable to methanol blended fuels broadly as the M85 treatment remained oxic and did not generate any methane gas 160 concentrations over the duration of the experiment. For these experiments, it appears that although methanol is an ideal substrate to feed into methanogenic processes we do not see enrichment in methanogenic archaea like we do within the microcosm experiments. This may be due to the presence of the recalcitrant and often toxic hydrocarbons found in gasoline inhibiting the growth and activity of certain taxa [41, 42]. Within the microcosm experiment however, methanogenic archaea are enriched in conjunction with methane generation. In these experiments, there was no inhibition from hydrocarbons present in the biofuel blends used in column experiments.   5.5 Conclusion In this chapter, microbial community response to methanol and methanol blended fuels were evaluated under controlled laboratory conditions using column and microcosm experiments. Within the column experiments, shifts in the overall microbial community structure were not observed between treatments, although increased abundance and activity of fermentative, hydrocarbon degrading and C1 carbon utilizing taxa upon exposure to the methanol blended fuel were observed. Methanogenic archaea were enriched to 1% in the MeOH2 (M85) column treatment. For the microcosm experiments, strong enrichment in methanogenic archaea was observed along with clear differences in microbial community structure between the sand and silt experiments. Moreover, complete degradation of methanol after 100 days for both 100ppm and 1000ppm concentration treatments was observed indicating that these methanol concentrations are not toxic and can support enrichment of methanogenic taxa under specific growth conditions.  161 From a discovery perspective, we were able to determine that microbial community responses to controlled biofuel release are variably time lagged, and that the generation of methane resulting from transformed biofuel can take weeks to months or even years to manifest a detectable signature. While not surprising, this has implications for risk management associated with contaminated sites and points towards the need for long-term monitoring efforts to constrain the impacts of biofuel release on the environment and human health. While we identified microbial responders to methanol blended biofuel release, we could not discriminate between different blend ratios. Extended time series analysis or more rapid biofuel transformation in the experimental systems might have provided a more granular perspective needed to identify indicator groups. The identification of taxonomic or functional indicators and the development of new diagnostic tests based on these indicators would go a long way towards augmenting the capacity to monitor and mitigate the effects of biofuel release in the environment. Future studies should include a more-in depth look at the functional genes and transcripts driving microbial metabolic responses at the individual, population and community levels of organization.          162 Chapter 6: Conclusion This thesis explores microbial community responses to ethanol and methanol blended fuels through the use of the 16S rRNA gene surveys. Through laboratory and field experiments, microbial community abundance and activity was evaluated upon exposure to a blended fuel treatment to identify taxonomic and trait-based shifts in community structure. This closing chapter highlights some of the key findings of my thesis work while also providing recommendations for future research directions.   6.1 Microbial ecology and the use of the 16S rRNA gene Microorganisms are everywhere. They are able to thrive in an onslaught of different conditions and underpin the ecological functioning of all environments. Given their ubiquitous nature, microbes are excellent first responders to anthropogenic disturbances as changes in abundance and metabolic activity can provide useful information into the overall health and functioning of the system [2, 5, 13] . In terrestrial systems, the response of microbial communities has been evaluated in natural or seasonal contexts as well as a result of a direct anthropogenic disturbance such as an oil or biofuel spill [20, 37, 38, 42]. In these studies, the resultant shift in abundance or potential activity of the microbial community has enabled a deeper ecological characterization of the disturbance which can aid in the development of effective remediation and mitigation strategies. Commonly, culture-independent profiling of microbial communities is accomplished using the 16S rRNA gene as a quantitative taxonomic marker. Indeed, the accessibility and popularity of this approach has led to the development of specific computational and statistical tools that are well-suited to handle 16S rRNA sequence information [46, 47]. In this thesis, the use of the 16S rRNA gene to profile microbial 163 community structure has allowed me to identify taxa with traits relevant to the natural attenuation and degradation of either ethanol or methanol blended fuels. Profiling the microbial community within the Cambria field site ethanol spill enabled me to observe the enrichment of methanogenic archaea alongside known hydrocarbon degraders and one carbon compound utilizers. Community profiling within the ethanol and methanol column experiments revealed abundant and active fermentative and hydrocarbon degrading taxa, suggesting the start of the overall degradation and attenuation process. Through both experimental designs, I was able to detect shifts in the microbial community upon exposure to the blended fuel treatment while also observing important differences between the two studies with regards to methanogenic activity. The discrepancy between laboratory experiments and the spill site highlights the importance of the physical environment in shaping microbial community structure which must be considered when attempting to characterize ecological responses and translating laboratory results to real-world systems with greater complexity.   6.2 Comparing field experiments and column experiments Comparisons between the Cambria spill field site and column experiments should be made with caution as there are important differences in the soil properties between the two experiments, especially in regards to available organic matter. Cambria site soils were classified as ranging from organic silty clay to silty sand while the columns were comprised of silty sand material which has a lower organic carbon content [36, 43, 114]. This is a key difference between the two experiments as carbon content and availability shapes microbial community structure and function [135, 136]. Indeed, this difference is highlighted in the ordination plot which revealed the Cambria site samples separating away from the ethanol and methanol column 164 experiments (Figure 6.1). Additionally, the E95 (95% ethanol 5% gasoline) blended fuel contains an ethanol content higher than any treatment applied to the columns which may drive different community responses [137]. Nevertheless, the most evident difference observed between the two experiments is in the abundance and potential activity of methanogenic archaea. Within the column experiments, no considerable methane gas was generated across any of the column experiments, an observation congruent with an overall lack of methanogenic archaea being identified within the microbial community. This is in contrast with the Cambria site supporting the metabolic activities of methanogenic taxa. The presence of methanogens is likely driving the difference in microbial community structure between these two experiments, as illustrated in the ordination plot with Cambria samples clustering away from the other datasets (Figure 6.1). At the Cambria site, enrichment of methanogenic taxa indicates that the full anaerobic degradation process was occurring with primary degradation and fermentation being completed by bacteria within the Betaproteobacteria, Deltaproteobacteria and Firmicutes. Next, the subsequent degradation and fermentation products, such as acetate, were used in methanogenesis to produce methane. In the column experiments, shifts in the community towards Gamaproteobacteria and Firmicutes suggest that the initial steps of anaerobic degradation are occurring however they are not being coupled to methanogenesis.  165  Figure 6.1 NMDS ordination plot for rDNA and rRNA samples collected from the Cambria denatured fuel grade ethanol spill site, ethanol blended fuel column experiments, methanol blended fuel column experiments, and methanol microcosm experiment.            166  Figure 6.2 Phylum level taxonomic summary for samples collected from the Cambria denatured fuel grade ethanol spill site, ethanol blended fuel column experiments, methanol blended fuel column experiments, and methanol microcosm experiment.167  6.3 Future directions To characterize microbial community responses to ethanol and methanol blended fuels across different blend ratios and sediment material types, I simulated spills through column experiments. In both ethanol and methanol column experiments, similar taxonomic response patterns were observed despites differences in CO2, CH4 and O2 gas fluxes and volatile fatty acid production within each column. While I did observe more nuanced differences in the relative abundance and potential activity of certain taxa between column treatments, the consistent response of the microbial community despite differences in geochemical measurements suggests that the community may need to be evaluated at even greater taxonomic resolution. While characterizing microbial community structure based on taxonomic markers such as the 16S rRNA  gene is an important step when evaluating ecological responses, it also has distinct limits given that most bacterial and archaeal species remain uncultivated [138, 139]. Furthermore, taxonomic diversity is not always the best predictor of functional diversity as the former may not capture local adaptations and natural history shaping genome architectures within evolutionarily coherent populations [138]. Therefore, profiling the functional diversity within each column treatment should be the next step taken in order to thoroughly characterize microbial community responses to ethanol and methanol blended fuels. Metagenomic and metatranscriptomic sequencing offers a way to add functional characterization to this study however simple qPCR surveys would also suffice to provide an initial assessment. By incorporating functional information into this study, one may also be able to better resolve the discrepancy between the geochemical data and community data.  168 6.4 Closing Advances in molecular biology and sequencing technologies have enabled scientists and engineers to study uncultivated microbial communities with ever increasing detail. By profiling the abundance and potential activity of microbial communities one can identify specific members while also monitoring changes in structure and activity due to perturbation. In this thesis, I employed a phylogenetic anchor gene approach focused on sequencing the 16S rRNA gene to evaluate microbial community responses to ethanol and methanol blended fuel contamination. I observed an enrichment in taxa with well-defined traits related to hydrocarbon degradation, fermentation and one carbon compound utilization providing an initial characterization of the potential ecological impacts of these fuels in the environment. While additional information into the overall functioning of these communities would be beneficial, this work identifies taxa responsive to blended fuels which can in turn help in the development of monitoring tools and potential remediation strategies needed to develop best practices and increase social license of biofuel production and transport in the environment.     169 Bibliography   1. Whitman, W.B., D.C. Coleman, and W.J. Wiebe, Prokaryotes: the unseen majority. Proc Natl Acad Sci U S A, 1998. 95(12): p. 6578-83. 2. Falkowski, P.G., T. Fenchel, and E.F. Delong, The microbial engines that drive Earth's biogeochemical cycles. Science, 2008. 320(5879): p. 1034-9. 3. Ducklow, H., Microbial services: challenges for microbial ecologists in a changing world. Aquatic Microbial Ecology, 2008. 53(1): p. 13-19. 4. Bissett, A., et al., Microbial community responses to anthropogenically induced environmental change: towards a systems approach. Ecol Lett, 2013. 16 Suppl 1: p. 128-39. 5. Blaser, M.J., et al., Toward a Predictive Understanding of Earth's Microbiomes to Address 21st Century Challenges. MBio, 2016. 7(3). 6. Tiedje, J. and T. Donohue, Microbes in the energy grid. Science, 2008. 320(5879): p. 985. 7. Hermans, S.M., et al., Bacteria as Emerging Indicators of Soil Condition. Appl Environ Microbiol, 2017. 83(1). 8. Ager, D., et al., Anthropogenic disturbance affects the structure of bacterial communities. Environ Microbiol, 2010. 12(3): p. 670-8. 9. Shade, A., et al., Fundamentals of microbial community resistance and resilience. Front Microbiol, 2012. 3: p. 417. 10. Wagg, C., et al., Soil biodiversity and soil community composition determine ecosystem multifunctionality. Proc Natl Acad Sci U S A, 2014. 111(14): p. 5266-70. 11. Cabral, L., et al., Microbial functional responses to long-term anthropogenic impact in mangrove soils. Ecotoxicol Environ Saf, 2018. 160: p. 231-239. 12. Cotta, S.R., et al., Exploring bacterial functionality in mangrove sediments and its capability to overcome anthropogenic activity. Mar Pollut Bull, 2019. 141: p. 586-594. 13. Graham, E.B., et al., Microbes as Engines of Ecosystem Function: When Does Community Structure Enhance Predictions of Ecosystem Processes? Front Microbiol, 2016. 7: p. 214. 14. Allison, S.D. and J.B. Martiny, Colloquium paper: resistance, resilience, and redundancy in microbial communities. Proc Natl Acad Sci U S A, 2008. 105 Suppl 1: p. 11512-9. 15. Finlay, B.J., Stephen C. Maberly, and J. Ian Cooper, Microbial Diversity and Ecosystem Function. Oikos, 1997. 80(2). 16. Little, A.E., et al., Rules of engagement: interspecies interactions that regulate microbial communities. Annu Rev Microbiol, 2008. 62: p. 375-401. 17. Rhind, S.M., Anthropogenic pollutants: a threat to ecosystem sustainability? Philos Trans R Soc Lond B Biol Sci, 2009. 364(1534): p. 3391-401. 18. Wasi, S., S. Tabrez, and M. Ahmad, Toxicological effects of major environmental pollutants: an overview. Environ Monit Assess, 2013. 185(3): p. 2585-93. 19. Simpson, J.M., J.W. Santo Domingo, and D.J. Reasoner, Microbial source tracking: state of the science. Environ Sci Technol, 2002. 36(24): p. 5279-88. 170 20. Nelson, D.K., T.M. Lapara, and P.J. Novak, Effects of Ethanol-Based Fuel Contamination: Microbial Community Changes, Production of Regulated Compounds, and Methane Generation. Environmental Science & Technology, 2010. 44(12): p. 4525-4530. 21. Delille, D., B. Delille, and E. Pelletier, Effectiveness of bioremediation of crude oil contaminated subantarctic intertidal sediment: the microbial response. Microb Ecol, 2002. 44(2): p. 118-26. 22. King, G.M., et al., Microbial responses to the Deepwater Horizon oil spill: from coastal wetlands to the deep sea. Ann Rev Mar Sci, 2015. 7: p. 377-401. 23. Bae, H.S., et al., Response of microbial populations regulating nutrient biogeochemical cycles to oiling of coastal saltmarshes from the Deepwater Horizon oil spill. Environ Pollut, 2018. 241: p. 136-147. 24. Mason, O.U., et al., Metagenomics reveals sediment microbial community response to Deepwater Horizon oil spill. ISME J, 2014. 8(7): p. 1464-75. 25. Militon, C., et al., Bacterial community changes during bioremediation of aliphatic hydrocarbon-contaminated soil. FEMS Microbiol Ecol, 2010. 74(3): p. 669-81. 26. Lindstrom, J.E., et al., Microbial populations and hydrocarbon biodegradation potentials in fertilized shoreline sediments affected by the T/V Exxon Valdez oil spill. Appl Environ Microbiol, 1991. 57(9): p. 2514-22. 27. Fuentes, S., et al., Bioremediation of petroleum hydrocarbons: catabolic genes, microbial communities, and applications. Appl Microbiol Biotechnol, 2014. 98(11): p. 4781-94. 28. Kostka, J.E., et al., Hydrocarbon-degrading bacteria and the bacterial community response in gulf of Mexico beach sands impacted by the deepwater horizon oil spill. Appl Environ Microbiol, 2011. 77(22): p. 7962-74. 29. Brooijmans, R.J., M.I. Pastink, and R.J. Siezen, Hydrocarbon-degrading bacteria: the oil-spill clean-up crew. Microb Biotechnol, 2009. 2(6): p. 587-94. 30. Leahy, J.G. and R.R. Colwell, Microbial degradation of hydrocarbons in the environment. Microbiol Rev, 1990. 54(3): p. 305-15. 31. Das, N. and P. Chandran, Microbial degradation of petroleum hydrocarbon contaminants: an overview. Biotechnol Res Int, 2011. 2011: p. 941810. 32. Morais, D., et al., Responses of microbial community from tropical pristine coastal soil to crude oil contamination. PeerJ, 2016. 4: p. e1733. 33. Gieg, L.M., S.J. Fowler, and C. Berdugo-Clavijo, Syntrophic biodegradation of hydrocarbon contaminants. Curr Opin Biotechnol, 2014. 27: p. 21-9. 34. Fargione, J.E., R.J. Plevin, and J.D. Hill, The Ecological Impact of Biofuels. Annual Review of Ecology, Evolution, and Systematics, Vol 41, 2010. 41: p. 351-377. 35. Michael Wolinetz, M.H., Biofuels in Canada 2018. 2018: Navius Research Inc. 36. Sihota, N.J., et al., Methane emissions and contaminant degradation rates at sites affected by accidental releases of denatured fuel-grade ethanol. J Contam Hydrol, 2013. 151: p. 1-15. 37. Feris, K., et al., 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, 2008. 42(7): p. 2289-2294. 38. Ma, J., et al., Adaptive microbial population shifts in response to a continuous ethanol blend release increases biodegradation potential. Environmental Pollution, 2013. 178: p. 419-425. 171 39. Ma, J., C.W. Nossa, and P.J.J. Alvarez, Groundwater ecosystem resilience to organic contaminations: microbial and geochemical dynamics throughout the 5-year life cycle of a surrogate ethanol blend fuel plume. Water Research, 2015. 80: p. 119-129. 40. Capiro, N.L., et al., Microbial community response to a release of neat ethanol onto residual hydrocarbons in a pilot-scale aquifer tank. Environ Microbiol, 2008. 10(9): p. 2236-44. 41. Ma, J., W.G. Rixey, and P.J. Alvarez, Microbial processes influencing the transport, fate and groundwater impacts of fuel ethanol releases. Curr Opin Biotechnol, 2013. 24(3): p. 457-66. 42. Elazhari-Ali, A., et al., Biofuel components change the ecology of bacterial volatile petroleum hydrocarbon degradation in aerobic sandy soil. Environmental Pollution, 2013. 173: p. 125-132. 43. Spalding, R.F., et al., Long-Term Groundwater Monitoring Results at Large, Sudden Denatured Ethanol Releases. Groundwater Monitoring & Remediation, 2011. 31(3): p. 69-81. 44. Woese, C.R. and G.E. Fox, Phylogenetic structure of the prokaryotic domain: the primary kingdoms. Proc Natl Acad Sci U S A, 1977. 74(11): p. 5088-90. 45. Pace, N.R., A molecular view of microbial diversity and the biosphere. Science, 1997. 276(5313): p. 734-40. 46. Caporaso, J.G., et al., QIIME allows analysis of high-throughput community sequencing data. Nat Methods, 2010. 7(5): p. 335-6. 47. Schloss, P.D., et al., Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol, 2009. 75(23): p. 7537-41. 48. Knight, R., et al., Best practices for analysing microbiomes. Nat Rev Microbiol, 2018. 16(7): p. 410-422. 49. Klindworth, A., et al., Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res, 2013. 41(1): p. e1. 50. Schloss, P.D., A high-throughput DNA sequence aligner for microbial ecology studies. PLoS One, 2009. 4(12): p. e8230. 51. Callahan, B.J., P.J. McMurdie, and S.P. Holmes, Exact sequence variants should replace operational taxonomic units in marker-gene data analysis. ISME J, 2017. 11(12): p. 2639-2643. 52. Weiss, S., et al., Normalization and microbial differential abundance strategies depend upon data characteristics. Microbiome, 2017. 5(1): p. 27. 53. Paulson, J.N., et al., Differential abundance analysis for microbial marker-gene surveys. Nat Methods, 2013. 10(12): p. 1200-2. 54. McMurdie, P.J. and S. Holmes, Waste Not, Want Not: Why Rarefying Microbiome Data Is Inadmissible. Plos Computational Biology, 2014. 10(4). 55. Coxe, S., S.G. West, and L.S. Aiken, The analysis of count data: a gentle introduction to poisson regression and its alternatives. J Pers Assess, 2009. 91(2): p. 121-36. 56. McMurdie, P.J., Normalization of Microbiome Profiling Data. Methods Mol Biol, 2018. 1849: p. 143-168. 57. Bottomly, D., et al., Evaluating Gene Expression in C57BL/6J and DBA/2J Mouse Striatum Using RNA-Seq and Microarrays. Plos One, 2011. 6(3). 172 58. Giorgi, F.M., C. Del Fabbro, and F. Licausi, Comparative study of RNA-seq- and Microarray-derived coexpression networks in Arabidopsis thaliana. Bioinformatics, 2013. 29(6): p. 717-724. 59. von Wyschetzki, K., H. Lowack, and J. Heinze, Transcriptomic response to injury sheds light on the physiological costs of reproduction in ant queens. Molecular Ecology, 2016. 25(9): p. 1972-1985. 60. Anders, S., Analysing RNA-Seq data with the DESeq package. Molecular Biology, 2010. 43: p. 1-17. 61. Laforge, R., Homoscedasticity. American Psychologist, 1963. 18(4): p. 213-214. 62. Zwiener, I., B. Frisch, and H. Binder, Transforming RNA-Seq Data to Improve the Performance of Prognostic Gene Signatures. Plos One, 2014. 9(1). 63. Friedman, J. and E.J. Alm, Inferring Correlation Networks from Genomic Survey Data. Plos Computational Biology, 2012. 8(9). 64. Engelbrektson, A., et al., Experimental factors affecting PCR-based estimates of microbial species richness and evenness. ISME J, 2010. 4(5): p. 642-7. 65. Edgar, R.C., Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 2010. 26(19): p. 2460-2461. 66. Shade, A., et al., Conditionally Rare Taxa Disproportionately Contribute to Temporal Changes in Microbial Diversity. Mbio, 2014. 5(4). 67. Gies, E.A., et al., Illuminating Microbial Dark Matter in Meromictic Sakinaw Lake. Applied and Environmental Microbiology, 2014. 80(21): p. 6807-6818. 68. Lawson, C.E., et al., Rare taxa have potential to make metabolic contributions in enhanced biological phosphorus removal ecosystems. Environmental Microbiology, 2015. 17(12): p. 4979-4993. 69. Martin-Fernandez, J.A., C. Barcelo-Vidal, and V. Pawlowsky-Glahn, Dealing with zeros and missing values in compositional data sets using nonparametric imputation. Mathematical Geology, 2003. 35(3): p. 253-278. 70. Beals, E.W., Bray-Curtis Ordination - an Effective Strategy for Analysis of Multivariate Ecological Data. Advances in Ecological Research, 1984. 14: p. 1-55. 71. Hill, M.O., Diversity and Evenness: A Unifying Notation and Its Consequences. Ecology, 1973. 54(2): p. 427-432. 72. Jahnke, J., et al., The influence of growth conditions on the cell dry weight per unit biovolume of Klebsormidium flaccidum (Charophyta), a typical ubiquitous soil alga. Journal of Applied Phycology, 2011. 23(4): p. 655-664. 73. McMurdie, P.J. and S. Holmes, phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One, 2013. 8(4): p. e61217. 74. Jackson, D.A., Compositional data in community ecology: The paradigm or peril of proportions? Ecology, 1997. 78(3): p. 929-940. 75. Gloor, G.B., et al., Microbiome Datasets Are Compositional: And This Is Not Optional. Front Microbiol, 2017. 8: p. 2224. 76. Bueche, M., et al., Quantification of Endospore-Forming Firmicutes by Quantitative PCR with the Functional Gene spo0A. Applied and Environmental Microbiology, 2013. 79(17): p. 5302-5312. 77. Hedlund, B.P., et al., An integrated study reveals diverse methanogens, Thaumarchaeota, and yet-uncultivated archaeal lineages in Armenian hot springs. Antonie Van 173 Leeuwenhoek International Journal of General and Molecular Microbiology, 2013. 104(1): p. 71-82. 78. Hu, H.W., et al., pH-dependent distribution of soil ammonia oxidizers across a large geographical scale as revealed by high-throughput pyrosequencing. Journal of Soils and Sediments, 2013. 13(8): p. 1439-1449. 79. Nicol, G.W., L.A. Glover, and J.I. Prosser, Molecular analysis of methanogenic archaeal communities in managed and natural upland pasture soils. Global Change Biology, 2003. 9(10): p. 1451-1457. 80. Chang, W., et al., Molecular characterization of polycyclic aromatic hydrocarbon (PAH)-degrading methanogenic communities. Biotechnology Progress, 2005. 21(3): p. 682-688. 81. Capiro, N.L., et al., Microbial community response to a release of neat ethanol onto residual hydrocarbons in a pilot-scale aquifer tank. Environmental Microbiology, 2008. 10(9): p. 2236-2244. 82. Ingram, L.O., Ethanol Tolerance in Bacteria. Critical Reviews in Biotechnology, 1990. 9(4): p. 305-319. 83. Wang, Y.C., et al., Discovery of multiple IGS haplotypes within genotypes of Puccinia striiformis. Fungal Biology, 2012. 116(4): p. 522-528. 84. Blazewicz, S.J., et al., Evaluating rRNA as an indicator of microbial activity in environmental communities: limitations and uses. Isme Journal, 2013. 7(11): p. 2061-2068. 85. Sukenik, A., et al., Massive multiplication of genome and ribosomes in dormant cells (akinetes) of Aphanizomenon ovalisporum (Cyanobacteria). Isme Journal, 2012. 6(3): p. 670-679. 86. Garson, G.D., Factor analysis: Statistical methods and practical issues. Social Science Computer Review, 1999. 17(1): p. 129-131. 87. Kruskal, J.B., Multidimensional-Scaling by Optimizing Goodness of Fit to a Nonmetric Hypothesis. Psychometrika, 1964. 29(1): p. 1-27. 88. Kruskal, J.B., Citation Classic - Multidimensional-Scaling by Optimizing Goodness of Fit to a Nonmetric Hypothesis. Current Contents/Social & Behavioral Sciences, 1979(39): p. S12-S12. 89. Norris, M. and L. Lecavalier, Evaluating the Use of Exploratory Factor Analysis in Developmental Disability Psychological Research. Journal of Autism and Developmental Disorders, 2010. 40(1): p. 8-20. 90. Pearce, P., Factor-Analysis - Statistical-Methods and Practical Issues - Jaeon,K, Muller,Cw. Australian Journal of Psychology, 1986. 38(1): p. 89-89. 91. Love, M.I., W. Huber, and S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol, 2014. 15(12): p. 550. 92. Faust, K. and J. Raes, Microbial interactions: from networks to models. Nature Reviews Microbiology, 2012. 10(8): p. 538-550. 93. Faust, K., et al., Microbial Co-occurrence Relationships in the Human Microbiome. Plos Computational Biology, 2012. 8(7). 94. Weiss, S., et al., Correlation detection strategies in microbial data sets vary widely in sensitivity and precision. Isme Journal, 2016. 10(7): p. 1669-1681. 95. Hill, J., Environmental Costs and Benefits of Transportation Biofuel Production from Food-and Lignocellulose-Based Energy Crops: A Review. Sustainable Agriculture, 2009. 174 96. Hill, J., et al., Climate change and health costs of air emissions from biofuels and gasoline. Proc Natl Acad Sci U S A, 2009. 106(6): p. 2077-82. 97. Hill, J., et al., Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proc Natl Acad Sci U S A, 2006. 103(30): p. 11206-10. 98. Susan E. Powers, C.S.H., Stephen E. Heermann, Henry X. Corseuil, David Rice & Pedro J. J. Alvarez The Transport and Fate of Ethanol and BTEX in Groundwater Contaminated by Gasohol. Critical Reviews in Environmental Science and Technology, 2001. 31(1): p. 79-123. 99. Powers, S.E., et al., Will ethanol-blended gasoline affect groundwater quality? Environ Sci Technol, 2001. 35(1): p. 24A, 26A-30A. 100. Capiro, N.L., et al., Fuel-grade ethanol transport and impacts to groundwater in a pilot-scale aquifer tank. Water Res, 2007. 41(3): p. 656-64. 101. Ma, J., W.G. Rixey, and P.J.J. Alvarez, Increased fermentation activity and persistent methanogenesis in a model aquifer system following source removal of an ethanol blend release. Water Research, 2015. 68: p. 479-486. 102. Edgar, R.C., et al., UCHIME improves sensitivity and speed of chimera detection. Bioinformatics, 2011. 27(16): p. 2194-200. 103. Quast, C., et al., The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res, 2013. 41(Database issue): p. D590-6. 104. Kunin, V., et al., Wrinkles in the rare biosphere: pyrosequencing errors can lead to artificial inflation of diversity estimates. Environ Microbiol, 2010. 12(1): p. 118-23. 105. Wickham, H. tidyverse. 2017 2017; Available from: https://cran.r-project.org/web/packages/tidyverse/index.html. 106. Jari Oksanen, F.G.B., Michael Friendly, Roeland Kindt, Pierre, D.M. Legendre, Peter R. Minchin, R. B. O'Hara, Gavin L. Simpson, Peter, and M.H.H.S. Solymos, Eduard Szoecs and Helene Wagner, vegan:Community Ecology Package. R package version 2.4-6. 2018. 107. Aburto-Medina, A., et al., Comparison of indigenous and exogenous microbial populations during slurry phase biodegradation of long-term hydrocarbon-contaminated soil. Biodegradation, 2012. 23(6): p. 813-22. 108. Xu, X., et al., Petroleum Hydrocarbon-Degrading Bacteria for the Remediation of Oil Pollution Under Aerobic Conditions: A Perspective Analysis. Front Microbiol, 2018. 9: p. 2885. 109. Kleikemper, J., et al., Activity and diversity of sulfate-reducing bacteria in a petroleum hydrocarbon-contaminated aquifer. Appl Environ Microbiol, 2002. 68(4): p. 1516-23. 110. S.A. Wakelin, L.M.M., S.L. Rogers, A.L. Gregg, T.P. Bolger, J.A. Baldock, Habitat selective factors influencing the structural composition and functional capacity of microbial communities in agricultural soils. Soil Biology and Biochemistry, 2008. 40(3): p. 803-813. 111. Zhou, J., et al., Spatial and resource factors influencing high microbial diversity in soil. Appl Environ Microbiol, 2002. 68(1): p. 326-34. 112. Fierer, N. and R.B. Jackson, The diversity and biogeography of soil bacterial communities. Proc Natl Acad Sci U S A, 2006. 103(3): p. 626-31. 113. Zamzow, K.L., Microbial communities utilizing biodiesel waste and ethanol in treatment of acid mine drainage. 2007. 175 114. Devries, J., Experimental investigation on the fate of ethanol-blended fuels in the subsurface, in Earth, Ocean and Atmospheric Sciences. 2017, University of British Columbia. 115. Lee, S. and S.J. Hallam, Extraction of high molecular weight genomic DNA from soils and sediments. J Vis Exp, 2009(33). 116. Hurt, R.A., et al., Simultaneous recovery of RNA and DNA from soils and sediments. Appl Environ Microbiol, 2001. 67(10): p. 4495-503. 117. Kleinsteuber, S., K.M. Schleinitz, and C. Vogt, Key players and team play: anaerobic microbial communities in hydrocarbon-contaminated aquifers. Appl Microbiol Biotechnol, 2012. 94(4): p. 851-73. 118. Chen, Y.D., J.F. Barker, and L. Gui, A strategy for aromatic hydrocarbon bioremediation under anaerobic conditions and the impacts of ethanol: A microcosm study. Journal of Contaminant Hydrology, 2008. 96(1-4): p. 17-31. 119. Zhang, Y., et al., Transport and degradation of ethanol in groundwater. J Contam Hydrol, 2006. 82(3-4): p. 183-94. 120. K. Bluhm, S.H., T. Seiler et al., Toxicological and ecotoxicological potencies of biofuels used for the transport sector—a literature review. Energy & Environmental Science, 2012. 5(6): p. 7381-7392. 121. Renewable and Low Carbon Fuel Requirements Regulation, B.C.M.o.E.a. Mines, Editor. 2014. 122. Summary for Compliance Period Energy Independence and Security Act (EISA)  , B.C.M.o.E.a. Mines, Editor. 2007. 123. Gomez, D.E. and P.J. Alvarez, Comparing the Effects of Various Fuel Alcohols on the Natural Attenuation of Benzene Plumes Using a General Substrate Interaction Model. Journal of Contaminant Hydrology 2010. 113: p. 66-76. 124. Elfasakhany, A., Investigations on the effects of ethanol–methanol–gasoline blends in a spark-ignition engine: Performance and emissions analysis. Engineering Science and Technology, 2015. 18(4): p. 713-719. 125. McKendry, P., Energy production from biomass (Part 2): Conversion technologies. Bioresour Technol, 2002. 83(1): p. 47-54. 126. McKendry, P., Energy production from biomass (Part 3): Gasification technologies. Bioresour Technol, 2002. 83(1): p. 55-63. 127. Zhu, L., et al., Emissions characteristics of a diesel engine operating on biodiesel and biodiesel blended with ethanol and methanol. Sci Total Environ, 2010. 408(4): p. 914-21. 128. Kolb, S., Aerobic methanol-oxidizing bacteria in soil. FEMS Microbiol Lett, 2009. 300(1): p. 1-10. 129. Murrell, J.C. and M.S. Jetten, The microbial methane cycle. Environ Microbiol Rep, 2009. 1(5): p. 279-84. 130. Bell, T.H., et al., Predictable bacterial composition and hydrocarbon degradation in Arctic soils following diesel and nutrient disturbance. ISME J, 2013. 7(6): p. 1200-10. 131. Redmond, M.C., D.L. Valentine, and A.L. Sessions, Identification of novel methane-, ethane-, and propane-oxidizing bacteria at marine hydrocarbon seeps by stable isotope probing. Appl Environ Microbiol, 2010. 76(19): p. 6412-22. 132. Levine, A.S. and C.R. Fellers, Inhibiting Effect of Acetic Acid upon Microorganisms in the Presence of Sodium Chloride and Sucrose. J Bacteriol, 1940. 40(2): p. 255-69. 176 133. Rajala, P. and M. Bomberg, Reactivation of Deep Subsurface Microbial Community in Response to Methane or Methanol Amendment. Front Microbiol, 2017. 8: p. 431. 134. Vanwonterghem, I., et al., Methylotrophic methanogenesis discovered in the archaeal phylum Verstraetearchaeota. Nat Microbiol, 2016. 1: p. 16170. 135. Zornoza, R., et al., Changes in soil microbial community structure following the abandonment of agricultural terraces in mountainous areas of Eastern Spain. Appl Soil Ecol, 2009. 42(3): p. 315-323. 136. Drenovsky, R.E., et al., Soil water content and organic carbon availability are major determinants of soil microbial community composition. Microb Ecol, 2004. 48(3): p. 424-30. 137. Chen, C.S., et al., Assessing soil and groundwater contamination from biofuel spills. Environ Sci Process Impacts, 2015. 17(3): p. 533-42. 138. Koskella, B., L.J. Hall, and C.J.E. Metcalf, The microbiome beyond the horizon of ecological and evolutionary theory. Nat Ecol Evol, 2017. 1(11): p. 1606-1615. 139. Steen, A.D., et al., High proportions of bacteria and archaea across most biomes remain uncultured. ISME J, 2019. 13(12): p. 3126-3130.                177 Appendix A: Chapter 5 supplemental A.1      Microcosm indicator species analysis table Table A.1 Indicator species analysis summary table for the methanol microcosm experiments MICROCOSM MATERIAL INDICATOR VALUE P-VALUE OTU_ID PHYLUM CLASS ORDER FAMILY  SILT 1.00 0.001 OTU_20597  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  1.00 0.001 OTU_4766  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  0.94 0.004 OTU_47745  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  0.94 0.008 OTU_29791  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  0.86 0.026 OTU_26084  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  0.87 0.04 OTU_5395  Euryarchaeota  Methanomicrobia  Methanosarcinales  Methanosarcinaceae  1.00 0.001 OTU_8320  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  1.00 0.001 OTU_46592  Acidobacteria  Holophagae  Subgroup 7   1.00 0.001 OTU_16566  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  1.00 0.001 OTU_436  Acidobacteria  Holophagae  Subgroup 7   1.00 0.001 OTU_27558  Acidobacteria  Holophagae  Subgroup 7  uncultured bacterium  1.00 0.001 OTU_16  Acidobacteria  Holophagae  Subgroup 7  uncultured soil bacterium  1.00 0.001 OTU_57033  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  1.00 0.001 OTU_50736  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  1.00 0.001 OTU_48544  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  0.98 0.001 OTU_798  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacteriales bacterium  0.94 0.001 OTU_7100  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp. 178  0.94 0.001 OTU_31062  Acidobacteria  Holophagae  Subgroup 7  uncultured bacterium  0.98 0.002 OTU_2108  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacteria bacterium  0.94 0.003 OTU_21134  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  0.94 0.004 OTU_26377  Acidobacteria  Holophagae  Subgroup 7   0.93 0.028 OTU_7538  Acidobacteria  Holophagae  Subgroup 7  uncultured Acidobacterium sp.  0.79 0.039 OTU_9307  Acidobacteria  Holophagae  Subgroup 7  uncultured soil bacterium  0.79 0.047 OTU_3482  Planctomycetes  Phycisphaerae  Phycisphaerales   0.98 0.004 OTU_16960  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.93 0.016 OTU_14164  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.87 0.027 OTU_27822  Proteobacteria  Betaproteobacteria  Methylophilales  Methylophilaceae  0.87 0.033 OTU_2236  Verrucomicrobia  Opitutae  Opitutales  Opitutaceae SAND 1.00 0.001 OTU_20589  Acidobacteria  Holophagae  Holophagales  Holophagaceae  1.00 0.001 OTU_29271  Acidobacteria  Holophagae  Holophagales  Holophagaceae  1.00 0.001 OTU_46758  Acidobacteria  Holophagae  Holophagales  Holophagaceae  1.00 0.001 OTU_54662  Acidobacteria  Holophagae  Holophagales  Holophagaceae  1.00 0.001 OTU_59678  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.99 0.001 OTU_45879  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.93 0.001 OTU_15775  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.91 0.011 OTU_50387  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.85 0.022 OTU_14227  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.84 0.025 OTU_47760  Acidobacteria  Holophagae  Holophagales  Holophagaceae 179  0.85 0.028 OTU_58227  Acidobacteria  Acidobacteria  Acidobacteriales  Acidobacteriaceae (Subgroup 1)  0.85 0.028 OTU_2217  Acidobacteria  Acidobacteria  Acidobacteriales  Acidobacteriaceae (Subgroup 1)  0.85 0.03 OTU_50197  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.85 0.034 OTU_15894  Acidobacteria  Holophagae  Holophagales  Holophagaceae  0.84 0.035 OTU_925  Acidobacteria  Solibacteres  Solibacterales  Solibacteraceae (Subgroup 3)  1.00 0.001 OTU_1148  Actinobacteria  Actinobacteria  Frankiales  Sporichthyaceae  0.91 0.007 OTU_3932  Actinobacteria  Actinobacteria  Micrococcales  Cellulomonadaceae  0.93 0.01 OTU_511  Actinobacteria     0.84 0.023 OTU_4111  Actinobacteria  Actinobacteria  Propionibacteriales  Propionibacteriaceae  0.81 0.038 OTU_2425  Actinobacteria  Actinobacteria  Propionibacteriales  Propionibacteriaceae  0.85 0.025 OTU_442  Armatimonadetes  Armatimonadia  Armatimonadales  uncultured bacterium  0.90 0.008 OTU_9095  Bacteroidetes  Sphingobacteriia  Sphingobacteriales  Chitinophagaceae  0.85 0.023 OTU_465  Bacteroidetes  WCHB1-32  uncultured soil bacterium    0.83 0.035 OTU_345  Bacteroidetes  WCHB1-32  uncultured bacterium    1.00 0.001 OTU_473  Chloroflexi  Anaerolineae  Anaerolineales  Anaerolineaceae  0.93 0.001 OTU_773  Chloroflexi  SBR2076  uncultured eubacterium WCHB1-50    0.99 0.002 OTU_47350  Chloroflexi  KD4-96  uncultured Chloroflexi bacterium    0.92 0.007 OTU_10654  Chloroflexi  KD4-96  uncultured Chloroflexi bacterium    0.91 0.013 OTU_6771  Chloroflexi  KD4-96  uncultured Chloroflexi bacterium    0.98 0.016 OTU_371  Chloroflexi  KD4-96  uncultured Chloroflexi bacterium   180  0.85 0.022 OTU_799  Chloroflexi  Anaerolineae  Anaerolineales  Anaerolineaceae  0.85 0.024 OTU_30615  Chloroflexi  SBR2076    0.85 0.024 OTU_33767  Chloroflexi  SBR2076  uncultured eubacterium WCHB1-50    0.85 0.027 OTU_16220  Chloroflexi  SBR2076  uncultured eubacterium WCHB1-50    0.85 0.034 OTU_1472  Chloroflexi  Anaerolineae  Anaerolineales  Anaerolineaceae  0.90 0.006 OTU_6065  Firmicutes  Clostridia  Clostridiales  Peptococcaceae  0.92 0.008 OTU_56210  Firmicutes  Clostridia  Clostridiales  Peptococcaceae  0.92 0.025 OTU_54164  Firmicutes  Clostridia  Clostridiales  Peptococcaceae  1.00 0.001 OTU_2785  Gemmatimonadetes  Gemmatimonadetes  Gemmatimonadales  Gemmatimonadaceae  0.99 0.001 OTU_332  Gemmatimonadetes  Gemmatimonadetes  Gemmatimonadales  Gemmatimonadaceae  0.93 0.004 OTU_2707  Nitrospirae  Nitrospira  Nitrospirales  FW13  1.00 0.001 OTU_980  Proteobacteria  Deltaproteobacteria  Myxococcales  Haliangiaceae  1.00 0.001 OTU_41912  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.99 0.001 OTU_1896  Proteobacteria  Deltaproteobacteria  Myxococcales   0.97 0.002 OTU_28028  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.98 0.004 OTU_30314  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.91 0.005 OTU_33279  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae  0.91 0.011 OTU_53081  Proteobacteria  Deltaproteobacteria  Desulfuromonadales  Geobacteraceae   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0392576/manifest

Comment

Related Items