Louisiana State University LSU Digital Commons
LSU Master's Theses Graduate School
2010 The nflueI nces of Microbial Diversity on Carbonate Geochemistry Across a Transition from Fresh to Saline Water in the Edwards Aquifer, Texas Cassie Jo Gray Louisiana State University and Agricultural and Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_theses Part of the Earth Sciences Commons
Recommended Citation Gray, Cassie Jo, "The nflueI nces of Microbial Diversity on Carbonate Geochemistry Across a Transition from Fresh to Saline Water in the Edwards Aquifer, Texas" (2010). LSU Master's Theses. 4268. https://digitalcommons.lsu.edu/gradschool_theses/4268
This Thesis is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Master's Theses by an authorized graduate school editor of LSU Digital Commons. For more information, please contact [email protected].
THE INFLUENCES OF MICROBIAL DIVERSITY ON CARBONATE GEOCHEMISTRY ACROSS A TRANSITION FROM FRESH TO SALINE WATER IN THE EDWARDS AQUIFER, TEXAS
A Thesis
Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Master of Science
in
The Department of Geology and Geophysics
by Cassie Jo Gray B.S., Indiana State University, 2008 May 2010
ACKNOWLEDGEMENTS
First and foremost I would like to extend my utmost gratitude to my advisor, Dr.
Annette Engel, for sharing her passion and knowledge of geomicrobiology, and for her patience and tough love throughout this learning and growing experience. Thanks also to my other committee members, Dr. Carol Wicks and Dr. Huiming Bao, for their insightful input into my research.
Thanks to Gary Schindel and the Edwards Aquifer Authority for their permission to access the study wells. I would like to acknowledge Brendan Headd and Scott Engel for their tremendous assistance in the field, especially for help stringing the microcosms and tying knots. Thanks also to the laboratory assistants, Brendan Donnelly and Skylar White, for all their time spent on specs, gels, and mini-preps. Personal thanks are due to Issaku Kohl for his time spent helping with the IC, and to Dr. Xiaogang Xie for his assistance with the SEM.
Many thanks to those family members and friends who have supported me throughout the years and encouraged me to pursue my education, even if hundreds of miles away.
Thanks to the geologists at Indiana State University for providing all the opportunities that allowed me to get to where I am, and those at Louisiana State University for this wonderful two-year experience. Finally, I am forever indebted to Tara Ewer, a true friend whom encouraged me to be something in life and reach for something better before I realized I needed it. Don’t let your…sparkle fade away.
Funding for this research was provided by the Louisiana Board of Regents, American
Association of Petroleum Geologists (AAPG) Edward B. Picou Grant, Karst Waters Institute
(KWI) William L. Wilson Scholarship, and the LSU Department of Geology and Geophysics.
ii TABLE OF CONTENTS
ACKNOWLEDGEMENTS…………………………………………………………….……...ii
LIST OF TABLES……………………………………………………………………….….…v
LIST OF FIGURES…………………………………………………………………………...vi
ABSTRACT…………………………………………………………………………………viii
CHAPTER 1 INTRODUCTION……………………………………………….………….…….1 The Edwards Aquifer………………………………………….………….……1 Karstification Processes…………………….………………………………….6 Subsurface Microorganisms………………….…………………………….…..9 Research Objectives…………………………….…………………………….11 Study Area………………………………………….…………………………12 Significance of Research……………………………….……………………..12 Thesis Organization……………………………………….,…………………14
2 MICROBIOLOGY ACROSS THE TRANSITION FROM FRESH TO SALINE WATER IN THE EDWARDS AQUIFER………………….…….16 Introduction…………………………………………………………….……..16 Materials and Methods………………………………………………….…….16 Results………………………………………………………………………...18 Discussion…………………………………………………………………….25
3 GEOCHEMISTRY ACROSS THE TRANSITION FROM FRESH TO SALINE WATER IN THE EDWARDS AQUIFER………………………...37 Introduction…………………………………………………………………...37 Materials and Methods………………………………………………………..37 Results………………………………………………………………………...39 Discussion……………………………………………………………...... ….43
4 GEOCHEMICAL AND MICROBIAL INFLUENCES ON EDWARDS AQUIFER CARBONATES………………………………………..50 Introduction…………………………………………………………………...50 Materials and Methods………………………………………………………..50 Results………………………………………………………………………...53 Discussion…………………………………………………………………….59
5 SUMMARY AND FUTURE DIRECTIONS……………………………………69
REFERENCES……………………………………………………………………………….74
iii APPENDIX A OPERATIONAL TAXONOMIC UNITS (OTUs)……………………………...85
B SEM IMAGES………………………………………………………………….92
VITA………………………………………………………………………………………...118
iv LIST OF TABLES
2.1 Distribution of major class or phylum-level taxonomy of 16S rRNA gene bacterial sequences from clones retrieved from transect wells in New Braunfels, Texas ………………………………………………………………19
3.1 Geochemical results from the first sampling of Comal Springs and transect study wells, New Braunfels, Texas...... 40
3.2 Geochemical results from the second sampling of Comal Springs and transect study wells, New Braunfels, Texas……………………………...... 41
4.1 Well depths, diameters, screened intervals, and microcosm suspension depths….……51
4.2 Average mass loss from calcite and dolomite microcosms and associated significance ……………………………………………………………54
v LIST OF FIGURES
1.1 The Edwards Aquifer in central Texas………………………………………………..…2
1.2 Generalized stratigraphic column of the Edwards Aquifer and associated units in the San Marcos Platform………………………………………………….…….3
1.3 (A) Map of the New Braunfels transect study wells and (B) the corresponding geologic cross section…..……………………………………………………………...13
2.1 Distribution of major classes and phyla in transect wells, in order of increasing salinity (from left to right)………………………………………………….20
2.2 Rarefaction curves for all 16S rRNA gene sequences retrieved from Edwards Aquifer study wells…………………………………………………………..21
2.3 Trees relating environmental microbial samples using hierarchical clustering analysis in UniFrac………………………………………………………….24
2.4 Principal Coordinate Analysis (PCA) obtained from UniFrac………………………...26
3.1 Piper diagrams showing the hydrochemical facies of transect study wells from (A) the first sampling and (B) second sampling…………………………………42
3.2 Chemical parameters…………………………………………………………………...44
- 3.3 Concentrations of HCO3 versus (A) [Ca], (B) [Ca+Mg], and (C) [Ca+Mg]-[SO4]…………………………………………………………………….45
3.4 Saturation indices (log IAP/Ksp) of (A) calcite, (B) dolomite, and (C) gypsum versus sulfate concentrations……………………………………………..46
3.5 Speciation of (A) [Ca] and (B) [Mg] in transect wells based on complexes modeled in PHREEQC…………………………………………………….47
4.1 Mass lost for reactive and sterile (A) calcite microcosms and (B) dolomite microcosms……...….……………………………………………….55
4.2 Average mass loss (mg) versus calcite saturation index values for each well…………56
4.3 Surficial features of mineral control chips……………………………………………..56
4.4 Cellular morphology and behavior…………………………………………………….58
4.5 Cell-surface associations………………………………………………………………58
vi
4.6 Dissolution features…………………………………………………………………….60
4.7 Secondary precipitates on experimental mineral surfaces……………………………..61
4.8 EDAX spectra for (A) iron sulfide mineral and (B) silica-bearing mineral…………...62
4.9 Number of putative cells observed on reactive calcite and dolomite chips per examined chip area (0.0625 mm2)…………………………………62
4.10 Average amount of mass loss (mg) versus the total average number of cells per microcosm………………………………………………………...63
vii ABSTRACT
Microbially-mediated karstification through the production of metabolic byproducts has been well-documented in cave environments, but less is known about deep karstic settings. This research aimed to distinguish between microbial and geochemical influences on carbonate dissolution in the Edwards Aquifer, a prolific karst aquifer in central Texas, specifically from a transect of six wells across a transition from fresh to saline water in New
Braunfels. For the first time, a portion of the aquifer’s bacterial diversity was examined from molecular 16S rRNA gene sequence analyses, which revealed that Alphaproteobacteria,
Gammaproteobacteria, and Betaproteobacteria dominated the aquifer, with rare
Bacteroidetes, Nitrospirae, and Firmicutes taxonomic groups. Local geochemical conditions for each well (H2S levels, TDS, and depth) accounted for ~83% of the genetic variability among wells. In general, putative chemoorganotrophic microorganisms were prevalent in all wells, but chemolithoautotrophs associated with sulfur oxidation were prevalent only in sulfidic wells. Sterile and reactive in situ microcosms containing experimental calcite and dolomite fragments were deployed in the transect wells for ~1 month. Most of the mineral fragments in the microcosms (~64%) had statistically significant mass loss, although fluids were saturated with respect to calcite (SI= +0.6 to +0.14) and dolomite (SI= +0.18 to +0.43).
There was greater mass loss in reactive microcosms having higher surface cell densities on colonized mineral surfaces. These results suggest that microbial colonization establishes geochemical disequilibrium between mineral surfaces and bulk aquifer fluids, regardless of the metabolic potential of the microorganisms. Microbially-mediated carbonate dissolution is possible in both fresh and saline water deep karstic zones of the aquifer.
viii CHAPTER 1: INTRODUCTION
The Edwards Aquifer in central Texas is a large carbonate aquifer that provides a natural study site to investigate karstification. Because the aquifer consists of distinct fresh and saline water zones, differences in carbonate dissolution and microbial diversity between the two water zones, and across geochemical gradients, can be assessed. Karstification is a process classically thought to be created by the percolation of CO2-rich groundwater through pores in carbonate rocks (Ford and Williams, 2007). However, the potential role of microorganisms in dissolving carbonate rock from corrosive metabolic byproducts has also been recognized (e.g., Sarbu et al., 2000; Bennett and Engel, 2005). In the following chapter, topics related to the geology and hydrology of the Edwards Aquifer, karstification processes, and the current understanding of microorganisms inhabiting the subsurface are reviewed to frame the current investigation of aquifer microbial diversity, and geochemical and microbial influences on carbonate dissolution in the Edwards Aquifer.
THE EDWARDS AQUIFER
The Edwards Aquifer is a karstified carbonate aquifer that provides the sole source of freshwater for the city of San Antonio, as well as numerous industries and municipalities in the region (Sharp and Banner, 1997). The Balcones Fault Zone (BFZ) portion of the aquifer is the geographic area referred to as the Edwards Aquifer (Sharp, 1990), which extends for nearly 320 kilometers and varies in width from 8 to 65 kilometers (Maclay and Small, 1986;
Deike, 1989) in Uvalde, Medina, Bexar, Comal, and Hays counties (Figure 1.1). The aquifer lies approximately 60 to 215 meters below land surface with an average thickness of about
150 meters (Deike, 1989) and thickens to the south and southeast (Sharp, 1990; Sharp and
Banner, 1997).
1
Figure 1.1. The Edwards Aquifer in central Texas. Modified from Groschen and Buszka (1997).
The rocks comprising the Edwards Aquifer (Edwards Group Limestone) were deposited on a wide, shallow marine platform in subtidal to tidal-flat depositional settings during a major marine transgression in the early Cretaceous (Albian) (Deike, 1989; Sharp,
1990; Sharp and Banner, 1997). The formations within the Edwards Aquifer are dominated by limestone or dolostone (Groschen and Buszka, 1997). For the present study, the field site is associated with the San Marcos Platform, which spans across Uvalde, Medina, Bexar, and
Comal counties. The platform consists of the (lower) Kainer Formation and (upper) Person
Formation comprised of wackestones, grainstones, packstones, and mudstones (Maclay and
Small, 1986; Sharp, 1990) (Figure 1.2). Significant permeability formed in several of the
2
Period Group Stratigraphic Unit
Anacacho Anacacho Limestone
Austin Austin Group
Eagle Ford Eagle Ford Group
Buda Limestone
Upper Cretaceous Washita Del Rio Clay
Georgetown Formation
Cyclic and marine members
Leached and collapsed members
Regional dense member Person Formation
Edwards Grainstone member
Kirschberg evaporite member Lower Cretaceous
Dolomitic member
Basal nodular member Kainer Formation Trinity Glen Rose Limestone
Figure 1.2. Generalized stratigraphic column of the Edwards Aquifer and associated units in the San Marcos Platform. Modified from Groschen and Buszka (1997) and Randall (2006).
3 formation members due to collapse breccias and burrowed wackestones (Maclay and Small,
1986), and reeflike buildups of rudists also provide primary porosity within the aquifer (Sharp and Banner, 1997). The Edwards Group is bound by the underlying Glen Rose Formation of the Upper Trinity Group and the overlying Del Rio Clay of the Washita Group (Maclay and
Small, 1986) (Figure 1.2). The Glen Rose Formation consists mostly of alternating limestone and marl (Sharp and Banner, 1997; Groschen and Buszka, 1997), whereas the Del Rio Clay is a selenitic, calcareous, pyritic clay (Sharp, 1990). Both units serve as effective, low- permeability confining layers for the Edwards Aquifer (Oetting et al., 1996). Igneous intrusions penetrate the aquifer units in Travis and Uvalde counties (Oetting et al., 1996).
Since deposition, many geologic events have occurred to alter the Edwards Aquifer.
The most important processes affecting the aquifer may be considered karstification and faulting (Oetting et al., 1996; Clark and Small, 1997). Karstification began shortly after deposition in the Cretaceous when the carbonate platform was exposed (Clement and Sharp,
1988). Subaerial exposure was widespread, particularly along the San Marcos Arch (along the crest of the San Marcos Platform), and was most intense in the upper Person Formation
(Sharp, 1990). A significant erosional hiatus occurred and likely led to the initial formation of cavernous porosity (Maclay and Small, 1986). However, the most intense karstification of these units has occurred in the late Cenozoic, further developing conduits and cave systems in the aquifer (Sharp, 1990).
Tectonic stresses in the Miocene, probably due to forebulging in the Gulf of Mexico
Basin (Maclay and Small, 1986; Sharp, 1990), led to extensive faulting of the Edwards
Group. The normal, en echelon faults are steep-angled and dip toward the Gulf of Mexico
(Sharp and Banner, 1997). The displacement on some individual fault blocks within the BFZ
4 exceeds 150 meters (Maclay and Small, 1986). These faults may act as either hydrologic barriers, when less permeable units are juxtaposed against permeable units, or as conduits by providing preferential flow paths (Sharp, 1990; Oetting et al., 1996; Clark and Small, 1997;
Groschen and Buszka, 1997; Sharp and Banner, 1997).
The aquifer consists of two distinct geochemical zones, fresh and saline (Figure 1.1), as a result of geologic, hydrologic, and diagenetic differences between the two zones. The transition between the fresh to saline water zones roughly coincides with the location of major faulting (BFZ) in the aquifer. Up-dip of the transition zone, the aquifer is unconfined, which has allowed for freshwater recharge to circulate through the rocks and enhance porosity and permeability. As a result, the unconfined portion of the aquifer has higher values of transmissivity (~190,000 m2/d) and waters have relatively low total dissolved solids (TDS) concentrations (Groschen and Buszka, 1997). Most of the recharge to the freshwater portion of the Edwards Aquifer (about 80%) is from losing streams flowing over the aquifer and into fractures, caves, and sinkholes (Sharp, 1990; Clark and Small, 1997; Sharp and Banner,
1997). After entering into the deeper artesian zone of the aquifer, water generally flows from west to east/northeast to discharge through freshwater springs (Clark and Small, 1997).
However, localized flow systems are quite complex due to faults, fractures, and karstification that create extremely heterogeneous and anisotropic conditions in the aquifer (Clark and
Small, 1997; Sharp and Banner, 1997).
On the other hand, freshwater cannot extensively flush through the down-dip, confined part of the aquifer. Therefore, the residence time of water and rock-water interaction periods increase (Groschen and Buszka, 1997), such that this portion of the aquifer has much lower transmissivity (estimated to be 1,000 m2/d) and waters have TDS values greater than 1,000
5 mg/L. Concentrations above 230,000 mg/L have also been reported (Groschen and Buszka,
1997). The degree of mixing between the two zones has been debated, but it is generally thought that less than 10% of water from the freshwater zone moves south/southeast across the saline water line (Rye et al., 1981). However, potential saline water encroachment into the freshwater zone may be possible if population growth and urbanization pressures increase pumping from freshwater wells within the transition zone (Sharp and Banner, 1997;
Cederberg et al., 1998).
Although aquifer geochemistry typically reflects a composite of diverse rock-water interactions (Oetting et al., 1996), the geochemical features of the saline water zone cannot be explained by limestone dissolution alone. Land and Prezbindowski (1981) provide geochemical and isotopic evidence for brine migration from oil fields located down-dip toward the Gulf of Mexico. Furthermore, Clement and Sharp (1988) determine that some areas of the saline water zone have been influenced by cross-formational flow from the underlying Hosston and Glen Rose Formations. Oetting et al. (1996) further constrain the origin and evolution of the saline water with isotopic analysis and geochemical modeling to define six hydrochemical facies, with geochemical differences explained within a structural and stratigraphic context. The saline water portion of the aquifer examined in the current study is thought to have evolved from complex processes, including water-rock interactions and mixing with down-dip oil field brines (Oetting et al., 1996).
KARSTIFICATION PROCESSES
The classical mechanism for initiating karst formation in carbonate-hosted rock is the percolation of CO2-rich groundwater through pore spaces (Poulson and White, 1969; White,
1988; Palmer, 1991; Ford and Williams, 2007). Carbon dioxide dissolves in water:
6 CO2 (g) ↔ CO2 (aq) (Eq. 1.1) and the CO2 (aq) reacts with water to form carbonic acid:
* CO2 (aq) + H2O ↔ H2CO3 (Eq. 1.2)
* H2CO3 is considered a weak acid, and two dissociation reactions are relevant to the geochemistry of natural waters (Kehew, 2001):
* + - H2CO3 ↔ H + HCO3 (pKa1=6.35 at 25°C; Eq. 1.3)
- + 2- HCO3 ↔ H + CO3 (pKa2=10.33 at 25°C; Eq. 1.4)
Natural waters open to the atmosphere can be slightly acidic due to the presence of carbonic acid, and so the solubility of carbonate minerals increases when in contact with meteoric waters (Fetter, 2001; Kehew, 2001):
2+ 2- CaCO3 + H2O + CO2 ↔ Ca + 2HCO3 (pK=8.48 at 25°C; Eq. 1.5)
2+ 2+ 2- CaMg(CO3)2 + H2O + CO2 ↔ Ca + Mg + 4HCO3 (pK=17.09 at 25°C; Eq. 1.6)
Continued recharge and groundwater circulation through carbonate rocks can promote karstification, and conduits or caves may develop (White, 1988; White, 2002; Ford and
Williams, 2007). Other mechanisms of karstification have also been identified. For example, dissolution may occur along fresh-saline water mixing zones, because for carbonate groundwaters, if two solutions that are both saturated with respect to calcite are mixed, the resulting solution will be undersaturated with respect to calcite (e.g., Bögli, 1964; Plummer,
1975; Wigley and Plummer, 1976; Mylroie and Carew, 1995).
Microbiological mechanisms have also been recognized in karst formation and the influence of microbes on carbonate dissolution may be as effective as purely chemical reactions (Mylroie and Balcerzak, 1992). For example, the excretion of organic acids, such as acetic or lactic acid, from microbes can promote rock weathering (Ehrlich, 1998; Barton and
7 Northup, 2007). When microbes form colonies or biofilms on rock surfaces, the organic acids can buildup and create a microenvironment of mineral undersaturation at the rock surfaces
(White, 1997; Ehrlich, 1998). For carbonates, the metabolic production of CO2 due to respiration of organic matter may be especially effective in karstification because of the strong influence of CO2 on the carbonate system (Viles, 1988; Bottrell et al., 1991; Mylroie and Balcerzak, 1992; White, 1997).
Sulfuric acid speleogenesis (SAS) is another karstification process proposed from work in Lower Kane Cave, Wyoming (Egemeier, 1981), and applied to other locations, including Carlsbad Cavern and Lechuguilla Cave (USA), Cueva de Villa Luz (Mexico), and
Frasassi Caves (Italy) (e.g., Hill, 1990; Jagnow et al., 2000; Hose et al., 2000; Sarbu et al.,
2000). SAS is initiated from H2S oxidation, such as when the dissolved gas comes into contact with oxygenated waters, to sulfuric acid (Hill, 1990; Bennett and Engel, 2005):
H2S + 2O2 ↔ H2SO4 (Eq. 1.7)
Sulfuric acid reacts with the limestone and replaces it with gypsum:
H2SO4 + CaCO3 + H2O ↔ CaSO4·2H2O + CO2 (Eq. 1.8)
Gypsum is highly soluble in most groundwater, and dissolution of it increases overall porosity
(Hose et al., 2000; Bennett and Engel, 2005). SAS can be enhanced by microbial activity, specifically by the presence of sulfur-oxidizing bacteria (Bennett and Engel, 2005). The initial reaction proceeds in the same manner as abiotic SAS (equation 1.8), but because microorganisms gain considerable energy from this metabolism (equation 1.9), the reaction can generate significant quantities of acidity:
2- + H2S + 2O2 ↔ SO4 + 2H (ΔGr°= -798 kJ/mol) (Eq. 1.9)
8 Sulfur-oxidizers, including members of the class Epsilonproteobacteria (Engel et al., 2004) and genera such as Thiothrix and Thiobacillus (Angert et al., 1998; Hose et al., 2000), have been identified from several SAS locales, thereby providing support for microbial SAS.
SUBSURFACE MICROORGANISMS
Microorganisms have been documented in the most inhospitable environments on
Earth, including the subsurface environment that may be one of the most oligotrophic environments yet explored. Microbes live in pore spaces in sedimentary rocks or fractures in igneous and metamorphic rocks (Pedersen, 2000; Amend and Teske, 2005), and microbial species have been found from terrestrial drill holes thousands of meters deep (e.g., Szewzyk et al., 1994; Liu et al., 1997; Pedersen, 2000; Kotelnikova, 2002; Kieft et al., 2005) and beneath the deep oceanic seafloor (Edwards et al., 2005; Batzke et al., 2007; Engelen et al., 2008).
The estimated total number of cells in the terrestrial subsurface ranges from 2.5 to 25 x 1029 cells (Whitman et al., 1998). When combined with values for the oceanic subsurface, the estimated total number of cells in the subsurface (3.8 - 6.0 x 1030) outnumbers the estimated abundances in any other environment (Whitman et al., 1998).
For subsurface environments, and karst aquifers in particular, it is generally thought that microbes obtain energy by living off of dissolved organic matter or organic-rich solid material deposited in sedimentary rocks (Petsch et al., 2003; Griebler and Lueders, 2009).
These organic materials could support communities of heterotrophs and chemoorganotrophs that utilize organic compounds for energy. However, evidence against this view has come from identifying and characterizing chemosynthesis in subsurface environments, including from karstic and basaltic aquifers (Stevens and McKinley, 1995; Sarbu et al., 1996; Kinkle and Kane, 2000; Chapelle et al., 2002; Birdwell and Engel, 2009). Chemolithoautotrophic
9 microbes obtain energy from the oxidation or reduction of inorganic compounds, such as nitrogen, iron, sulfur, and manganese (Northup and Lavoie, 2001), and these microorganisms in the subsurface could provide the energy base to support vast microbial and metazoan communities (Sarbu et al., 1996). In fact, recent evidence suggests the autochthonous primary production by chemolithoautotrophs may be the dominant source of organic matter in the
Edwards Aquifer (Birdwell and Engel, 2009). The potential for multiple energy sources, from both chemoorganotrophy and chemolithoautotrophy, may increase biodiversity in the subsurface (Kinkle and Kane, 2000).
Therefore, a better understanding of subsurface microorganisms is important because microbes mediate nutrient cycles, precipitate minerals, and accelerate rock weathering
(Ehrlich, 1996; Northup and Lavoie, 2001; Newman and Banfield, 2002). The role of microbes in the formation or dissolution of minerals in the subsurface environment is perhaps the most important. Several studies have been conducted that demonstrate the ability of microbes to precipitate and dissolve a wide range of minerals, including carbonates, silicates, and sulfates (e.g., Ehrlich, 1996; Warthmann et al., 2000; Frederickson and Onstott, 2001;
Northup and Lavoie, 2001; Hammes and Verstraete, 2002; Newman and Banfield, 2002;
Moore et al., 2004; Roberts et al., 2004; Pedersen, 2005; Kremer et al., 2008). Some minerals, like dolomite, are unable to precipitate at low temperatures in the absence of microorganisms (e.g., Land, 1998; Roberts et al., 2004). Microbes affect mineral dissolution processes especially if the microbes use elements in the minerals as an energy source, an electron acceptor, or for trace element requirements (Ehrlich, 1996). Microbially-influenced dissolution can be indirect, too, such as from the oxidation of sulfur, iron, or manganese in solution and production of acids as byproducts that promote mineral and rock weathering
10 (Northup and Lavoie, 2001; Bennett and Engel, 2005). Evidence for microbially-induced dissolution has come from in situ microcosm experiments (Hiebert and Bennett, 1992;
Bennett et al., 1996; Rogers and Bennett, 2004; Bennett and Engel, 2005; Engel and Randall,
2008), whereby microorganisms accelerate weathering of minerals (compared to changes in abiotic experimental controls and theoretical calculations) due to targeted colonization of the surfaces.
RESEARCH OBJECTIVES
Previous research that focused on the saline water zone of the Edwards Aquifer suggested that H2S-utilizing microbes produce localized acidity that enhanced carbonate dissolution (Engel and Randall, 2008). In that study, only wells from the saline water zone were examined over three months. Despite clear evidence that microbes contributed to dissolution, it was unclear if microbes also affected carbonate dissolution in the freshwater zone, or if biotic and abiotic dissolution rates differed between the saline and freshwater zones. Furthermore, limited assessment of the microbial communities in each well was done
(Engel and Randall, 2008). There was a need to establish a more detailed understanding of the microbial diversity from the different geochemical zones, in conjunction with experiments to characterize carbonate dissolution.
The overall goal of the current study is to evaluate microorganisms colonizing mineral surfaces and their metabolic byproducts that could enhance localized carbonate dissolution in both fresh and saline aquifer zones. The two main objectives were to (1) examine the microbial diversity along the fresh to saline water geochemical gradient, and (2) evaluate geochemical and microbial effects on carbonate dissolution along the same gradient. Six wells were used that transect the transition zone from fresh to saline water in Comal County,
11 Texas. The specific research hypotheses tested were that (1) metabolically distinct microbial groups inhabit different geochemical zones due to the availability of different energy sources
(e.g., photosynthetic products versus inorganic substrates) and (2) carbonate dissolution is greater in the saline water zone compared to the freshwater zone based on greater potential proton produced per mole of substrate (e.g., organic carbon versus H2S) from microbial metabolism.
STUDY AREA
The following six observation wells maintained by the Edwards Aquifer Authority in
Comal County near New Braunfels, Texas, were used to assess aquifer geochemistry and geomicrobiology by using in situ microcosm experiments (Figure 1.3): state wells #DX-68-
23-616A (Paradise Alley Shallow) and -616B (Paradise Alley Deep), #DX-68-23-619A (Girl
Scout Deep) and -619B (Girl Scout Shallow), and #DX-68-23-617 (LCRA Deep) and -618
(LCRA Shallow). Two wells penetrate different depths at each site (Figure 1.3B). These wells were chosen because of their close proximity to each other (less than 1 km total distance apart), and because the transect crosses the transition from fresh to saline water. Depending on recharge conditions, the wells are typically artesian, but were under non-artesian conditions during the study period because the region was experiencing a severe drought
(Birdwell and Engel, 2009). The cased wells were standard PVC four-inch diameter to ~23 to
30 meters, then transitioned to 2-inch diameter to total depth. Two-inch PVC screened intervals began at depths of ~158 to 228 meters, depending on total well depth (Table 4.1).
SIGNIFICANCE OF RESEARCH
A study of the Edwards Aquifer can be applied to carbonate aquifers worldwide to aid in understanding the factors controlling secondary porosity and karst formation, and results
12 A
B
Figure 1.3. (A) Map of the New Braunfels transect study wells and (B) the corresponding geologic cross section. Modified from Guyton & Associates, 1979.
13 will also enhance our knowledge of microbial processes in the subsurface environments.
Identification of subsurface microbes is important for global energy resources, because microbes may affect oil reservoir productivity and development both positively and negatively (Pedersen, 2000). Also, microbes in the subsurface have the potential to affect human health, and because they are capable of degrading toxic materials into less less-toxic compounds, microbes are vitally important in the natural remediation of contaminated aquifers (Wolicka and Borkowski, 2007).
The Edwards Aquifer is economically and ecologically important to central Texas. It provides freshwater to nearly 2 million people, including the entire city of San Antonio (Sharp and Banner, 1997). Possible dissolution and/or precipitation reactions have the potential to directly affect the volume of water contained within the aquifer for human usage (Randall,
2006). In addition, this is the first research to examine the microbial diversity of a portion of the Edwards Aquifer using molecular methods. The aquifer hosts a diverse ecosystem, including many threatened and endangered species (Longley, 1981; Jones et al., 2003), and the extent to which microbes provide the energy base for the macroscopic species is unknown.
THESIS ORGANIZATION
The following chapters describe the geochemistry and microbiology of the Edwards
Aquifer across the fresh to saline water transition zone in Comal County. Chapter 2 assesses the microbial diversity within five of the transect study wells in the 16S rRNA gene framework in an attempt to understand potential microbial metabolic processes in the aquifer.
Chapter 3 examines the aqueous chemistry and mineral saturation states of the six study wells from field and laboratory analysis of water samples. Chapter 4 evaluates the effects of aqueous chemistry and microorganisms on carbonate minerals in the fresh and saline zones of
14 the aquifer using an experimental in situ microcosm approach. Results and conclusions regarding the potential influence of microorganisms on karstification processes in the
Edwards Aquifer are summarized in Chapter 5.
15 CHAPTER 2: MICROBIOLOGY ACROSS THE TRANSITION FROM FRESH TO SALINE WATER IN THE EDWARDS AQUIFER
INTRODUCTION
Because most microorganisms are unculturable (e.g., Madigan et al., 2006), molecular methods have greatly increased our knowledge of microbial diversity in almost every environment. By comparing the genetic relationships among microorganisms, conservative inferences can be made about metabolism and ecophysiology (i.e., Pace, 1997; Barton et al.,
2004). The microbial diversity from five wells in New Braunfels, Texas, was assessed within the 16S rRNA gene framework. Results provide the first analysis on these wells, as well as one of the first phylogenetic studies in the Edwards Aquifer in either fresh or saline water zones.
MATERIALS AND METHODS
As part of the study done by Birdwell and Engel (2009), water was filtered onto glass fiber filters and 0.2µm PVDF (Polyvinylidene fluoride) filters. Total DNA was extracted using procedures described by Fuhrman et al. (1988). Purified nucleic acids were stored at
-20°C until use. Extracted DNA was PCR-amplified for 16S rRNA genes using the universal bacterial primer pair 8F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1510R (5′-
GGTTACCTTGTTACGACTT-3′). Amplification was executed using a MJ Research Dyad
Disciple thermal cycler for 30 cycles under the following conditions: denaturation at 94°C for
60 seconds, primer annealing at 47°C for 60 seconds, and chain extension at 72°C for 90 seconds. Amplified PCR products were purified using 0.7% TAE low-melt agarose gel with a
Wizard PCR prep DNA purification kit (Promega Corp., USA), following the manufacturer’s instructions. Concentrations of the PCR product were determined by spectrophotometry
16 (NanoDrop ND1000). The purified PCR products were then ligated into the PCR2.1-TOPO vector with the TOPO Cloning Kit (Invitrogen Corp., USA), according to manufacturer’s instructions. Clones were lysed for 10 minutes at 96°C and screened by PCR using the primer pair M13F (5'-GTAAAACGACGGCCAG-3') and M13R (5'-CAGGAAACAGCTATGAC-3') for 26 cycles under the following conditions: denaturation at 94°C for 60 seconds, primer annealing at 55°C for 60 seconds, and chain extension at 72°C for 60 seconds.
Clones containing the correct insert were diluted for Sanger sequencing, which was done by the University of Washington’s High-Throughput Genomics Unit (Seattle, WA,
USA) using capillary sequencers from both M13 primer ends and with internal primers 907R
(5'-CCGTCAATTCCTTTRAGTTT-3') and 704F (5'-GTAGCGGTGAAATGCGTAGA-3').
Resulting sequences were assembled using Contig Express, a component of Vector
NTI Advance 10.3.0 (Invitrogen Corp., USA), then submitted to Basic Local Alignment
Search Tool (BLAST) using GenBank (http://blast.ncbi.nlm.nih.gov) to determine similarity to cultured and uncultured organisms in the database. Sequences were submitted to
Greengenes (DeSantis et al., 2006), and 54 sequences were identified as chimera and removed. Gene sequences from this study can be found in GenBank under accession numbers
HM066203 to HM066729.
Nucleotide sequences were aligned with ClustalW using the Pasteur Institute Mobyle web portal (Neron et al., 2009; available online
A distance matrix was created from nucleotide sequences in BioEdit 7.0.4.1 (Hall, 1999).
Operational taxonomic units (OTUs) were determined using DOTUR (Distance-Based
Operational Taxonomic Units and Richness Determination) (Schloss and Handelsman, 2005).
17 OTUs were defined by the furthest neighbor distance at 98% sequence similarity, which corresponds conservatively to genus-level relatedness.
The phylogenetic tree obtained from Mobyle and environmental files were used as input into UniFrac (Lozupone and Knight, 2005;
RESULTS
A total of 456 nearly full-length (>1400bp) sequences and 71 partial (<1000 but
>700bp) sequences, comprising seven major taxonomic groups, were retrieved from the transect wells. Dominant phyla were most closely related to Proteobacteria, accounting for
~88% of all total sequences (Table 2.1). Gamma- and Alphaproteobacteria were the equally dominant classes (both representing ~29% of total sequences), followed by
Betaproteobacteria (28%), and Deltaproteobacteria (1.7%). Other phyla included
Bacteroidetes (5.1%), Nitrospirae (2.7%), Firmicutes (1.5%), Planctomycetes (0.8%),
Chloroflexi (0.4%), and Caldithrix (0.4%).
Distributions of the major taxonomic divisions differed among wells (Figure 2.1).
Gammaproteobacteria dominated in both LCRA Shallow (48%) and Girl Scout Deep (56%),
Alphaproteobacteria dominated in Paradise Alley Shallow (44%) and Girl Scout Shallow
(30%), and Betaproteobacteria dominated in LCRA Deep (54%). LCRA Deep was the only well to have <85% of the sequences represented by Proteobacteria (69%); this was due to an increase in the number of Bacteroidetes relative to other wells.
18 Table 2.1. Distribution of major class or phylum-level taxonomy of 16S rRNA gene bacterial sequences from clones retrieved from transect wells in New Braunfels, Texas. Girl Scout Girl Scout LCRA LCRA Paradise Alley Total # Major Class or Phylum Deep Shallow Deep Shallow Shallow Clones Gammaproteobacteria 50 30 3 50 23 156
Alphaproteobacteria 27 32 12 27 56 154
Betaproteobacteria 10 28 52 22 36 148
Deltaproteobacteria 0 3 0 1 5 9
Bacteroidetes 1 2 18 3 3 27
Nitrospirae 1 7 4 1 1 14
Firmicutes 0 1 5 0 2 8
Planctomycetes 0 2 2 0 0 4
Chloroflexi 0 1 0 1 0 2
Caldithrix 0 0 0 0 2 2
Unclassified Bacteria 0 2 1 0 0 3
Total Full-length Sequences 73 91 87 90 115 456
Total Partial Sequences 16 17 10 15 13 71
Total Sequences 89 108 97 105 128 527
19 Increasing Salinity
Girl Scout Girl Scout LCRA LCRA Paradise Alley Deep Shallow Deep Shallow Shallow
Alphaproteobacteria Gammaproteobacteria
Betaproteobacteria Deltaproteobacteria
Bacteriodetes Nitrospirae
Firmicutes Planctomycetes
Chloroflexi Caldithrix
Unclassified Bacteria
Figure 2.1. Distribution of major classes and phyla in transect wells, in order of increasing salinity (from left to right).
20 At 98% identity, conservatively corresponding to genus-level relatedness, all sequences grouped into 143 OTUs using DOTUR. All wells had between 34 and 52 OTUs, with an average of 41 OTUs for each site. Many of the OTUs were identified across several wells, but 106 individual OTUs were found in only one site. A complete list of all OTUs at
98% sequence identity, representative clones, and closest relatives can be found in Appendix
A.
Rarefaction curves obtained from DOTUR estimated how fully the bacterial diversity was represented from the clone libraries (Figure 2.2). If diversity was fully represented by the clone libraries, the curves would plateau into a nearly horizontal line, signifying that the diversity saturation had been reached. At 99%, 98%, and 97% sequence identity (blue,
250 99%
200 98% 97% 150
100
Number of OTUs of Number observed 50
0 0 100 200 300 400 500 600 Number of sequences sampled
Figure 2.2. Rarefaction curves for all 16S rRNA gene sequences retrieved from Edwards Aquifer study wells. Based on 99% sequence identity (blue curve), 98% sequence identity (orange curve), and 97% sequence identity (green curve).
21 orange, and green curves, respectively), the curves had not reached a horizontal line, indicating that diversity of these Edwards Aquifer wells was probably not fully sampled from the clone libraries. More samples may be required to achieve diversity saturation.
Two OTUs were common to all five wells (Appendix A). The first OTU, represented by 25 clones, was most closely related to Brevundimonas vesicularis [FM955876] (99% sequence identity), an Alphaproteobacteria. The second OTU had 23 clones and was also most closely related to B. vesicularis [FM955876] (99% sequence identity).
Six OTUs were common to four wells (Appendix A). Three of the OTUs contained
29, 19, and 4 clones, and were most closely related Pseudomonas poae [DQ536513] (99% sequence identity), Pseudomonas rhodesiae [FJ462694] (99% sequence identity), and
Pseudomonas migulae [EU111725] (98% sequence identity), respectively. All are
Gammaproteobacteria. Two of the OTUs had sequences most closely related to
Betaproteobacteria, with one OTU containing 18 sequences represented by an uncultured betaproteobacterium [AB194096] (99% sequence identity) identified from a study of simazine
(herbicide)-degrading bacteria. The other OTU consisted of 13 clones that were most closely related to Janthinobacterium spp. [AJ551147] (99% sequence identity). The final OTU identified in four of the five wells contained 10 clones represented by the Caulobacter crescentus [CP001340] (99% sequence identity), an alphaproteobacterium.
Four OTUs containing gammaproteobacterial sequences were found across three wells. Two of the OTUs, consisting of 27 and 4 clones each, were represented by
Pseduomonas spp. [AY247063] (99% sequence identity). Ten clones were grouped into an
OTU most closely related to Pseudomonas argentinensis [EU723817] (99% sequence
22 identity). The other OTU consisted of 27 clones represented by Pseudomonas fluorescens
[EU854430] (99% sequence identity).
Alphaproteobacterial sequences common to three of the five study wells grouped into three separate OTUs, each with rare clones. Two of the OTUs, consisting of 6 and 4 clones each, were represented by Sphingomonas aerolata [AJ429240] (99% sequence identity).
The remaining OTU had 11 clones most closely related to Agrobacterium larrymoorei
[EU373312] (99% sequence identity).
Some OTUs were not shared by several wells, but did contain many clones from one or two wells. For example, LCRA Deep and LCRA Shallow contained 5 and 11 sequences, respectively, that formed an OTU most closely related to uncultured Azospira spp.
[FJ393134] (99% sequence identity). Twenty-nine clones from LCRA Deep and 2 clones from Girl Scout Shallow were grouped into one OTU, with sequences most closely related to uncultured Betaproteobacteria [EF467590] (99% sequence identity) isolated from a sulfidic cave stream biofilm at Frasassi Caves, Italy (e.g., Macalady et al., 2008). Eleven sequences from Paradise Alley Shallow formed one OTU that were closely related to a betaproteobacterium in the Hydrogenophilaceae family [EU266809] (99% sequence similarity) isolated from a tar-oil contaminated aquifer (Winderl et al., 2008). Another OTU consisted of 14 clones from Paradise Alley Shallow and 4 clones from Girl Scout Deep most closely related to uncultured Alphaproteobacteria [EU038058] (99% sequence similarity) isolated from Lower Kane Cave, Wyoming (e.g., Engel et al., 2004b; Engel et al., 2010).
Analysis of sequences using UniFrac allows for the examination of the genetic relationships among microbial communities as a function of the environment from which they were obtained (Lozupone and Knight, 2005). This method was used to examine microbial
23 communities across a geochemical gradient, and well depth, TDS, and H2S levels were
distinguished based on the geochemical measurements from the first sampling (Chapter 3).
Values were corrected for multiple comparisons using the Bonferroni correction (Lozupone
and Knight, 2005). The three different environments were found to be suggestively different
from each other as a function of phylogeny. The P significance test (Lozupone and Knight,
2005) was also completed for comparison, and identical results were obtained.
An environmental distance matrix was created using the UniFrac metric to obtain the
raw distances between each pair of environments based on shared sequence groups. Larger
values indicate that environments are different, while smaller values indicate similarity.
A Well Environment 63 Paradise Alley H 100 LCRA M Girl Scout N
0.05
B Paradise Alley Shallow H S Sh 91 100 LCRA Shallow M F Sh 83 Girl Scout Deep N F D 100 Girl Scout Shallow N F Sh LCRA Deep M F D 0.05
Figure 2.3. Trees relating environmental microbial samples using hierarchical clustering analysis in UniFrac. Tree with (A) only H2S levels input, and (B) with H2S levels, TDS, and depth input. Numbers on nodes represent jackknifing results, 100 permutations in each simulation. Environmental parameters defined by H2S levels [H=high (>1.0mg/L), M=low/medium (0-1.0mg/L), N=none detected (0mg/L)], TDS (S=saline, F=fresh), and depth (Sh=shallow, D=deep).
24 Values ranged from ~0.55 to 0.74, suggesting that any pair of environments shared between
~45 to 25% of the total phylogenetic diversity contained by both environments, represented by hierarchical clustering analysis (UPGMA) to relate the samples with similar environments based on the microbial communities (Figure 2.3).
When the H2S level was the only parameter considered, wells with no H2S (i.e., Girl
Scout wells) clustered separately on the tree, whereas wells with varying concentrations of
H2S formed a distinct clade (Figure 2.3A). However, when depth and TDS were incorporated, the wells clustered together in an entirely different manner (Figure 2.3B). The resulting tree indicated that LCRA Shallow and Girl Scout Deep clustered together, whereas the LCRA Deep clustered separately from all other wells (Figure 2.3B). There was no apparent clustering pattern based on H2S levels, TDS, or depth. Principal coordinate analysis
(PCA) was utilized to illustrate the influence of each individual factor on phylogenetic similarities among the wells (Figure 2.4). H2S levels (P1), TDS (P2), and depth (P3) accounted for 31.94%, 27.19%, and 24.14% of the genetic variation, respectively.
DISCUSSION
Comparatively little information is available on the microbiology of pristine aquifers, including the Edwards Aquifer (e.g., Bates et al., 2006; Engel and Randall, 2008). In uncontaminated aquifers, microorganisms associated with a heterotrophic lifestyle have been characterized previously, but chemolithoautotrophs may be common members of subsurface communities as well, especially when allochthonous organic carbon input is limited (e.g.,
Birdwell and Engel, 2009; Griebler and Leuders, 2009). Several of the most abundant phyla identified from this study have been found repeatedly in other aquifer settings, including
Proteobacteria, Bacteroidetes, Firmicutes, and Nitrospirae from shallow karstic aquifers
25
A B
Girl Scout Girl Scout Shallow Shallow LCRA Deep LCRA Deep
Paradise Alley Paradise Alley Shallow Girl Scout Shallow Girl Scout Deep Deep
LCRA LCRA Shallow Shallow
C Paradise Alley Shallow
Girl Scout LCRA Shallow Shallow Girl Scout LCRA Deep Deep
Figure 2.4. Principal Coordinate Analysis (PCA) obtained from UniFrac. (A) H2S vs. TDS (P1 vs. P2); (B) Depth vs. TDS (P3 vs. P2); (C) H2S vs. Depth (P1 vs. P3).
26 (e.g., Griebler and Leuders, 2009; Pronk et al., 2009; Shabarova and Pernthaler, 2009; Pasic et al., 2010) and caves (Angert et al., 1998; Barton et al., 2004; Griebler and Lueders, 2009), suggesting that these groups are common microbial community members in karst environments.
The working hypothesis of this study was that distinct microbial groups would inhabit different wells as a function of geochemistry. Spatial heterogeneity is thought to exert a major influence on subsurface microbial diversity (Greibler and Leuders, 2009). Although physical dispersion mechanisms, such as localized groundwater flow patterns, may affect the distribution of planktonic microbial groups, the influence is not as significant as geochemical controls (e.g., Lozupone and Knight, 2005; Schauer et al., 2010). Based on the results from this study, geochemistry influences the distribution of microorganisms in the Edwards
Aquifer transect wells. The observation that many OTUs were identified only in certain wells suggests that there are environmental factors dictating their presence (or absence). Although some OTUs were common to all wells, represented by groups such as Brevundimonas and
Pseudomonas spp., these organisms have metabolic versatility and simple nutritional requirements (Madigan et al., 2006), and may not be as sensitive to subtle geochemical changes.
Taxonomic Descriptions
The following taxonomic descriptions briefly summarize the known metabolic pathways for microbial groups identified in Edwards Aquifer study wells (for later assessment of metabolic impact on carbonate geochemistry; Chapter 4). When possible, the potential geochemical influences on community distribution are discussed.
27 Alphaproteobacteria
OTUs containing alphaproteobacterial sequences were found across all wells.
Members of Alphaproteobacteria are metabolically diverse, but of special interest in sulfidic settings is their ability for complete sulfur oxidation (Sorokin, 2003; Friedrich et al., 2005).
Three major taxonomic orders were identified: Caulobacterales, Sphingomondales, and
Rhizobiales. Caulobacterales are thought to be aerobic, oligotrophic organisms (Barton et al.,
2007), and this group includes the genera Brevundimonas and Caulobacter. The two OTUs found in all five wells were most closely related to Brevundimonas spp., heterotrophic organisms capable of utilizing many different organic compounds for carbon and energy
(Madigan et al., 2006). Their metabolic versatility likely explains their presence in all study wells. Caulobacter spp. represented an OTU common to all wells except Paradise Alley
Shallow. These bacteria are well-studied chemoorganotrophs due to their ability to form cytoplasm-filled stalks for surface attachment (Madigan et al., 2006). In fact, cell attachment to surfaces has been frequently documented in subsurface environments, as this may be advantageous in nutrient-deprived environments (Griebler and Lueders, 2009), which could explain why no sequences related to Caulobacter spp. were retrieved from the Paradise Alley well, as the higher TDS levels represent greater potential availability of nutrients relative to other wells (see Chapter 3). A total of 28 sequences related to Sphingomondales, including the genus Sphingomonas, represented several OTUs across all wells. These organisms have been identified from nutrient-poor subsurface environments where they have wide metabolic capabilities, including the degradation of many different aromatic compounds, such as toluene and napthalene (Frederickson et al., 1995; Barton et al., 2004). Twenty-four total sequences related to Rhizobiales were identified in every well except LCRA Deep, and included the
28 OTU represented by A. larrymoorei found in three wells. Rhizobiales are generally found in plant nodules fixing nitrogen; however, free-living forms exist, as well (Madigan et al., 2006).
Recent findings suggest this group is ecologically important in the formation of biofilms under oxic and anoxic conditions (Pang and Liu, 2007). Furthermore, several genera within the order Rhizobiales are capable of oxidizing sulfur compounds (e.g., Ghosh and Roy, 2006;
Masuda et al., 2010). Fourteen sequences from Paradise Alley Shallow and 4 sequences from
Girl Scout Shallow formed an OTU represented by uncultured Alphaproteobacteria isolated from Lower Kane Cave, Wyoming [EU038058], a site well-studied for bacterial sulfur oxidation diversity and speleogenesis (e.g., Engel et al., 2004b). If these sequences represent sulfur-oxidizing organisms, it seems logical that they would be more abundant in sulfidic wells (Paradise Alley) versus non-sulfidic wells (Girl Scout).
Gammaproteobacteria
Almost every gammaproteobacterial sequence retrieved (150 out of 156 clones) formed OTUs represented by Pseudomonas spp., with at least one of these OTUs identified in every well. Members of Pseudomonas are classic examples of respiring, chemoorganotrophic organisms with simple nutritional requirements (Madigan et al., 2006), which may explain their ubiquity in all study wells. Some strains of the obligately heterotrophic Pseudomonas have also demonstrated the ability to oxidize thiosulfate (Mason and Kelly, 1988; Sorokin,
2003). The other 6 gammaproteobacterial sequences formed OTUs represented by
Acinetobacter calcoaceticus [AY346313] found only in Girl Scout wells. Members of this group are common organisms in soil and water environments (Barton et al., 2004), and it is not clear why these sequences were identified only in Girl Scout wells.
29 Betaproteobacteria
Sequences related to Betaproteobacteria were identified in all wells. Three major taxonomic orders comprised important OTUs: Burkholderiales, Rhodocyclales, and
Hydrogenophilales. Burkholderiales, similar to Pseudomonas in the Gammaproteobacteria, are another group of classic respiratory chemoorganotrophs (Madigan et al., 2006). Members of this order have a wide range in metabolism. For example, some species are metabolically versatile and able to use over 100 different compounds, whereas others are restricted to utilizing less than 20 (Madigan et al., 2006). Burkholderiales have also been identified as nitrogen-fixing organisms (Barton et al., 2007; Saito et al., 2008; Ishii et al., 2009). The dominant OTU from this order was represented by Janthinobacterium spp., containing 13 total clones identified from all wells except Girl Scout Deep. Most sequences related to the order Rhodocyclales (60%) formed a single OTU in the LCRA wells represented by Azospira spp. [FJ393134], organisms currently classified as free-living, aerobic, nitrogen-fixing bacteria (Madigan et al., 2006; Reinhold-Hurek and Hurek, 2006). Rhodocyclales genera were recently identified as the dominant species in a study examining denitrifiers in soil environments (Saito et al., 2008; Ishii et al., 2009), so their relative abundance in LCRA wells is reasonable because these wells had the greatest amount of nitrate available for consumption. Eleven sequences related to Hydrogenophilales formed an OTU with clones most closely related to an uncultured representative of this order [EU266809]. The clones were also found to have close similarity to uncultured Thiobacillus spp. [FJ536917] (99% sequence identity), a member of the Hydrogenophilales order. Thiobacillus are perhaps the
0 2- best-studied sulfur-oxidizing organisms, using H2S, S , or S2O3 as sources of energy
(Madigan et al., 2006). This may explain why clones related to this group were only
30 identified in the sulfidic Paradise Alley Shallow well. Similarly, 94% of the clones in an
OTU represented by uncultured Betaproteobacteria from a study of sulfur-oxidizing biofilms
(Frasassi Caves) were retrieved from the LCRA Deep well, which had H2S initially.
Thiobacillus spp. and other sulfur-oxidizing species have been identified from many cave environments where SAS is thought to occur (e.g., Angert et al., 1998; Hose et al., 2000;
Northup and Lavoie, 2001).
Deltaproteobacteria
A total of only 9 clones were closely related to Deltaproteobacteria, which was surprising given the anaerobic nature and sulfate availability in the wells (i.e., Engel et al.,
2003). All sequences were closely related to uncultured representatives of this class and formed individual OTUs. In Paradise Alley Shallow, OTUs were represented by clones isolated from a deep phreatic sinkhole [FJ484425], marine sediments [AM746084], uncultured Smithella spp. [EU888819], and an uncultured Desulfobacteraceae bacterium
[FJ517129]. The latter participate in the reduction of oxidized sulfur species, such as sulfate and elemental sulfur (Madigan et al., 2006). The presence of sulfate reducers in this well would not be unreasonable, considering the amount of sulfate available as an energy source
(Chapter 3). The other deltaproteobacterial OTUs were identified in LCRA Shallow
[FJ485074] and Girl Scout Shallow [AY792293, FJ205327, DQ811832].
Bacteroidetes
At least one clone from every well, to total 27 clones, was found to be most closely related to members of the Bacteroidetes. Two OTUs (total of 8 clones) found only in LCRA
Deep contained sequences most closely related to Flavobacterium spp. [AM888191,
GU295966]. Flavobacteria are common in aquatic environments, and are generally restricted
31 to using glucose or few other carbon compounds for energy and a carbon source (Madigan et al., 2006). Perhaps their strict requirements for particular substrates are why Flavobacteria were retrieved from only one well. Four distinct OTUs (total of 6 clones) were represented by
Pedobacter spp., including Pedobacter agri [EF660751], a species found to utilize various carbon compounds, such as glycogen, glucose, and maltose (Roh et al., 2008). Cytophaga spp. represent sequences belonging to one OTU in LCRA Deep, consisting of two clones
[AF414444], and one clone in Girl Scout Shallow [AB189358]. Cytophaga spp. are able to move by gliding, and are common in soil and water environments (Madigan et al., 2006). It has been suggested that Cytophaga spp. account for most of the cellulose degradation in natural, oxic environments (Madigan et al., 2006).
Nitrospirae
Nine OTUs (total of 14 clones) across the transect wells were represented by
Nitrospirae, chemolithoautotrophic nitrogen-oxidizing organisms that are also capable of chemoorganotrophic sulfate reduction (Barton and Luiszer, 2005). All clones were most closely related to uncultured representatives of this phylum [i.e., FJ484503, EU266868,
FJ484454]. They are thought to commonly reside in marine and soil habitats (Madigan et al.,
2006). Sequences related to Nitrospirae were present in all transect study wells, providing support that these organisms may be autochthonous microbial community members in karstic aquifers (Pronk et al., 2009).
Firmicutes
Firmicutes accounted for 1.5% of the total clone sequences from all transect wells (8 total clones). Firmicutes are nonsporulating, gram-positive bacteria that include the lactic acid bacteria (Madigan et al., 2006). Four clones formed one OTU represented by uncultured
32 Firmicutes spp. [DQ234641] isolated from groundwater. Two clones formed 2 individual
OTUs, both within LCRA Deep, that were most closely related to Vulcanibacillus modesticaldus [AM050346], an anaerobic, nitrate-reducing bacterium isolated from deep-sea hydrothermal vents (L’Haridon et al., 2006). Like the members of the betaproteobacterial order Rhodocyclales, perhaps the distribution of V. modesticaldus is influenced by nitrate levels, which are highest in the LCRA wells.
Planctomycetes
Four sequences related to uncultured Planctomycetes spp. were identified in LCRA
Deep [FJ484765, FJ517014] and Girl Scout Shallow [DQ837238, DQ837238] wells.
Planctomycetes are aerobic chemoorganotrophs found in fresh and marine water environments (Madigan et al., 2006). However, some members of this phylum have been identified as anoxic ammonia oxidizers (anammox) (Madigan et al., 2006). Planctomycetes are unique within the Bacteria domain because their cell walls lack peptidoglycan, and they are stalked, which presumably aids in attachment.
Chloroflexi
One clone from LCRA Shallow and 1 clone from Girl Scout Shallow were found to be closely related to the green nonsulfur bacteria Chloroflexi, each forming their own OTU.
Most Chloroflexi species are thermophilic, anoxygenic phototrophs (Madigan et al., 2006), so their isolation from a dark, mesophilic aquifer is unusual. However, strains have been identified from non-thermophilic microbial mats and are capable of aerobic chemoorganotrophic growth, as well (Madigan et al., 2006). It is possible that Chloroflexi are one of the microbial groups influenced by well depth, because clones were only retrieved from shallow wells.
33 Caldithrix
From the Paradise Alley Shallow well, 1.6% of the sequences (2 clones) formed an
OTU most closely related to uncultured species of Caldithrix isolated from phreatic limestone sinkholes [FJ901586]. Culture studies of Caldithrix spp. from hydrothermal vents have shown that this group consists of anaerobic thermophiles capable of reducing nitrate
(Miroshnichenko et al., 2003; Fedosov et al., 2006).
UniFrac Analysis
UniFrac results support the influence of local well geochemistry on microbial distribution across the transition zone. Significance testing in UniFrac revealed that the environments were suggestively different, so perhaps only minor changes in any geochemical parameter affects the structure of microbial communities, and this was why there was no obvious clustering pattern based on H2S, TDS, or depth. However, inspection of each individual parameter through PCA (Figure 2.3) showed that these three factors explain ~83% of the phylogenetic variation among the wells. The strong influence of H2S levels is reasonable, considering (1) its toxicity to many organisms, perhaps dismissing the possible presence of particular groups, and (2) its energy source to sulfur-oxidizing organisms, which would allow for the presence of an entirely different suite of metabolically distinct organisms.
At least the latter case appears to apply to the Edwards Aquifer transect wells, because where there were higher H2S levels, there were more putative sulfur-oxidizing organisms. While
TDS was the second most important geochemical control based on PCA (Figure 2.3), many different geochemical parameters are embedded with the TDS measurements; for example,
TDS correlated with pH and Cl (see Figure 3.2). Upon visual inspection of the PCA results with geochemical data, differences in alkalinity best explained the patterns observed along the
34 TDS axis; for example, Girl Scout Deep and LCRA Shallow had the lowest values for alkalinity (246 and 240 mg/L, respectively), and these two wells clustered tightly on the
UniFrac tree, while the other wells had alkalinity > 264 mg/L (Chapter 3). Finally, depth most evidently influenced the distribution of microbial communities in Paradise Alley
Shallow, which had a total depth of ~173 m, while all other wells had total depths of 198 m and greater.
Summary of Edwards Aquifer Microbiology
The genetic analysis based on 16S rRNA gene sequences of the Edwards Aquifer reveals that major groups present in the transect study wells include Alpha-, Gamma-, and
Betaproteobacteria, Bacteroidetes, Firmicutes, and Nitrospirae. These results are consistent with findings from other karstic aquifer and cave settings (e.g., Angert et al., 1998; Barton et al., 2004; Griebler and Leuders, 2009; Pronk et al., 2009; Shabarova and Pernthaler, 2009;
Pasic et al., 2010).
Chemoorganotrophs that respire CO2, such as Pseudomonas and Brevundimonas, comprised major OTUs identified from all wells and dominated both Girl Scout wells and the
LCRA Shallow well. It is tempting to speculate that the LCRA Deep well may have a thriving community of chemolithoautotrophs, as the most dominant OTU identified in that well contained sequences most closely related to organisms isolated from a sulfidic cave stream biofilm (Frasassi Caves) from a study of sulfur-oxidizing biofilms. Active microbial sulfur oxidation in this well would also provide a reason why no H2S was detected within the well casing water after ~1 month (Chapter 3). Similarly, the Paradise Alley Shallow well also probably contains chemolithoautotrophs because the most dominant OTU consisted of organisms closely related to putative alphaproteobacterial sulfur-oxidizing organisms.
35 Furthermore, the potential for sulfur oxidation is present within all classes of the
Proteobacteria, a phylum that accounted for 88% of the total sequences retrieved (Table 2.1).
Coupled with the sulfide availability in the wells, sulfur oxidation may represent an important metabolic pathway in the Edwards Aquifer.
The occurrence of other possible chemolithoautotrophic organisms, such as sulfate reducers, suggests that inorganic compounds may be a vital source of energy for Edwards
Aquifer microbes. Indeed, primary production by chemolithoautotrophs may account for most of the organic matter in the Edwards Aquifer (Birdwell and Engel, 2009). Additionally, the observation of several groups related to nitrogen compound-utilizing organisms, such as
Nitrospirae (nitrifying) and Rhodocyclales (denitrifying), may indicate active nitrogen cycling within the Edwards Aquifer.
Differences in geochemistry, specifically H2S, TDS, and depth, account for ~83% of the genetic variability in the study wells. Other geochemical factors, physical dispersion controls, or even changes in substrate mineralogical composition (Boyd et al., 2007), must account for the remaining ~17% of the genetic variability. Thus, geochemistry controls the microbial composition, resulting in different metabolic pathways and byproducts, which in turn exert varying effects on aquifer processes, such as carbonate geochemistry (Chapter 4).
36 CHAPTER 3: GEOCHEMISTRY ACROSS THE TRANSITION FROM FRESH TO SALINE WATER IN THE EDWARDS AQUIFER
INTRODUCTION
The assessment of the Edwards Aquifer water chemistry provides a way to understand the potential geochemical interactions occurring within the aquifer. Aqueous geochemistry from the study wells was analyzed using both field and laboratory techniques to model mineral saturation states. These results provide a characterization of well geochemistry that will be compared to experimental results (Chapter 4).
MATERIALS AND METHODS
Water samples were taken twice from the six transect wells and the freshwater Comal
Springs in New Braunfels, Texas (Comal County). Sampling was done approximately one month apart in the fall of 2009 through the duration of the in situ microcosm experiments
(Chapter 4). During the first sampling period, wells were purged using a Grundfos Redi-Flo pump for approximately 1.5 hours (equivalent to ~3 well volumes) to obtain representative aquifer water samples. During the second sampling, wells were not purged; instead, water within the well casing was collected with a peristaltic pump to obtain a representative sample of the final water chemistry to which the microcosms were exposed (i.e., these waters had the potential to have reduced input from the aquifer due to clogged screened intervals and may not have been representative of aquifer conditions).
Measurements for dissolved oxygen, temperature, conductivity, pH, and salinity were taken in the field using standard electrode methods from purged and sampled water (APHA,
1998). If a sulfidic odor was detected, spectrophotometric measurements were taken for dissolved hydrogen sulfide using the methylene blue method (CHEMetrics Inc., VA). Other
37 + 2+ unstable dissolved constituents, such as NH4 or Fe , were also measured using colorimetric
CHEMetrics methods.
Water samples to be analyzed were filtered with Whatman glass microfiber filters and
0.2 µm PVDF filters, and collected in separate sterile HDPE bottles for anion, cation, and
- 2- - alkalinity measurements in the laboratory. Concentrations of major anions (Cl , SO4 , F ,
2- - 2- NO3 , NO2 , and PO4 ) were determined on unpreserved filtered samples and major cations
(Ca2+, Na+, K+, and Mg2+) were determined on acidified filtered samples, both analyzed using a Dionex ICS-3000 Reagent-free ion chromatograph (IC). Standards checks were accurate within ±2 standard deviations (σ), except fluoride, sodium, and calcium (±3 σ). In the
- laboratory, alkalinity (as HCO3 ) was determined by end point titration using 0.1 N sulfuric acid to pH 4.3.
The geochemical data were modeled using PHREEQC 2.15.0 (Parkhurst and Appelo,
1999; http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqci) to determine aqueous speciation and mineral and complex saturation indices. The program corrects for temperature and ionic strength, and uses the Debye-Hückel equation to calculate activity coefficients.
This equation is valid for waters with ≤ 0.1 ionic strengths (Appelo and Postma, 2005), and all samples in this study had ionic strengths ≤ 0.042. The saturation indices of calcite and dolomite were of particular interest to evaluate dissolution or precipitation reactions.
Saturation index (SI) of a mineral is defined as
SI = log (IAP/Ksp) (Eq. 3.1) where IAP is the ion activity product and Ksp is a mineral thermodynamic equilibrium constant (Kehew, 2001). A positive SI value indicates that precipitation with respect to a mineral is thermodynamically possible, whereas dissolution is expected with negative SI
38 values. If the SI is 0 ± 0.5, then the mineral was assumed to be at equilibrium within the system (Appelo and Postma, 2005).
RESULTS
Four of the six transect wells (both Girl Scout and LCRA wells) were considered freshwater, with TDS < 1,000 mg/L (Table 3.1). Comal Springs and the freshwater wells had
- 2- Ca-Mg-HCO3 - type waters, and the saline water wells were Ca-Cl-SO4 -type waters (Figure
3.1). Comal Springs had much higher levels of dissolved oxygen, measuring above 5 mg/L, than any of the wells, and the most saline wells (Paradise Alley) had the highest levels of measured H2S during both sampling periods (Tables 3.1 and 3.2). From both sampling periods, there were only minor changes in the geochemical parameters, with the exception of
H2S and alkalinity. These changes reflected differences in aquifer fluids versus evolved casing fluid geochemistries during the month. During the first sampling, the Girl Scout wells had no detectable H2S, but upon microcosm retrieval, concentrations were elevated.
Conversely, the LCRA wells had measurable H2S during the first sampling, but none was detected one month later.
TDS did not strongly correlate to temperature (Figure 3.2A). All wells had near neutral pH, as measured in the field, and pH and TDS were negatively correlated (saline waters had the lowest pH) (Figure 3.2B). TDS positively correlated (r2=0.99) to chloride concentrations (Figure 3.2C). There were varying correlations between dissolved bicarbonate and various cation concentrations (Appelo and Postma, 2005). When compared to calcium concentrations, the freshwater wells had approximately a 3:1 relationship between bicarbonate and calcium (Figure 3.3A). When magnesium was added to the calcium concentrations and compared to bicarbonate concentrations, this ratio dropped to <2:1 (Figure 3.3B). However,
39 Table 3.1. Geochemical results from the first sampling of Comal Springs and transect study wells, New Braunfels, Texas. All measurements are in mg/L, except temperature (°C) and pH (standard units). BDL= below detection level; <0.1 mg/L for H2S.
Comal Girl Scout Girl Scout LCRA LCRA Paradise Alley Paradise Alley Parameters Springs Shallow Deep Shallow Deep Shallow Deep Temperature (°C)1 23.3 26.4 24.8 25.3 26.2 26.3 25.7 pH1 7.14 7.32 7.38 7.41 7.30 7.17 7.00
Alkalinity 272.79 265.47 245.95 240.83 264.25 265.47 325.01
Cl- 19.1 19.5 25.9 65.0 15.8 272.8 576.1
F- 1.2 4.8 3.2 3.7 3.9 6.4 4.0
2- SO4 31.9 54.3 52.4 92.4 44.8 328.6 621.5
2- NO3 10.6 0.1 0.2 1.3 1.9 0.3 0.5
Na+ 36.0 57.2 18.1 45.2 57.3 192.4 316.9
K+ 1.1 0.6 1.8 3.7 1.0 10.1 19.6
Mg2+ 18.8 24.1 32.8 38.4 26.7 68.0 112.6
Ca2+ 93.3 56.4 55.0 60.9 53.8 107.8 182.2
1 H2S 0 BDL BDL 0.38 0.9 2.01 7.86
1 O2 5.34 0.06 0.07 0.12 0.1 0.08 0.11 Total Dissolved Solids 421 384 367 462 381 1005 1686 1 Data from field measurements.
40 Table 3.2. Geochemical results from the second sampling of Comal Springs and transect study wells, New Braunfels, Texas. All measurements are in mg/L, except temperature (°C) and pH (standard units). BDL= below detection level; <0.1 mg/L for H2S.
Comal Girl Scout Girl Scout LCRA LCRA Paradise Alley Paradise Alley Parameters Springs Shallow Deep Shallow Deep Shallow Deep Temperature (°C)1 23.3 26.5 27.9 24.6 24.2 26.3 26.9 pH1 7.12 7.39 7.42 7.40 7.33 7.17 7.02
Alkalinity 284.50 270.35 244.00 243.51 263.52 270.84 307.44
Cl- 20.6 19.2 21.8 65.1 14.0 283.3 557.8
F- 2.0 3.9 2.3 5.9 3.5 6.9 5.3
2- SO4 33.3 41.3 41.9 93.4 41.0 346.6 612.5
2- NO3 18.1 4.3 0.0 2.3 2.5 0.0 0.4
Na+ 58.8 50.2 52.4 82.8 49.4 193.4 328.7
K+ 1.4 0.6 0.9 3.2 1.0 10.1 19.3
Mg2+ 18.7 24.7 29.0 39.7 27.2 68.2 105.0
Ca2+ 94.0 55.4 47.1 61.9 55.0 107.9 166.9
1 H2S 0 1.0 0.2 BDL BDL 4.68 2.64
1 O2 5.06 0.93 1.2 0.83 0.76 0.47 0.84 Total Dissolved Solids 474 377 361 534 375 1024 1632 1 Data from field measurements.
41
A B
Figure 3.1. Piper diagrams showing the hydrochemical facies of transect study wells from (A) the first sampling and (B) second sampling. Open symbols correspond to shallow wells; closed symbols correspond to the deeper wells.
42 when sulfate was subtracted from calcium and magnesium concentrations, there was a 2:1 trend for all wells (Figure 3.3C).
All study well waters were modeled as being saturated with respect to calcite (SI=
+0.6 to +0.14) and dolomite (SI= +0.18 to +0.43), but undersaturated with respect to nearly every other mineral for both water samplings. Comal Springs was saturated with respect to calcite (SI=+0.10), but undersaturated with respect to dolomite (SI= -0.18). There was no correlation between saturation indices of either mineral and TDS or sulfate concentrations
(Figure 3.4A and B). However, there was a strong correlation between sulfate concentration and gypsum saturation indices (r2=0.99); as TDS and sulfate concentration increased, the gypsum saturation index approached equilibrium (Figure 3.4C).
Calcium and magnesium complexes could affect calcite and dolomite solubility by lowering the activity of the free ion, and therefore increasing the solubility of minerals that contain that element (Kehew, 2001; Appelo and Postma, 2005). Minor amounts of calcium and magnesium complexes were present in the study wells (Figure 3.5), with most of the calcium and magnesium being present as free ions in freshwater wells (>90% as free ions) and saline water wells (>75% as free ions).
DISCUSSION
Clement and Sharp (1988) suggest that Miocene faulting is most intense in the present study area, resulting in minimal interchange between the two geochemical zones. Analysis of water chemistry from the transect study wells demonstrates the distinct shift from fresh to saline water in the Edwards Aquifer and the major geochemical variations between the two zones. Changes in TDS and dissolved ion concentrations within two kilometers (Figure 1.3) support previous estimates that the transition zone is fairly distinct. Restricted flow between
43 1800 A 1600
1400
1200
1000
800 TDS (mg/L) 600
400
200
0 23 24 25 26 27 28 29 Temperature (C) 1800
B 1600
1400
1200
1000
800 TDS (mg/L) 600 Girl Scout Deep Girl Scout Shallow 400 LCRA Deep 200 LCRA Shallow
0 Paradise Alley Deep 6.9 7 7.1 7.2 7.3 7.4 7.5 Paradise Alley Shallow pH Comal Springs C 700 600
500
400
300 Chloride (mg/L) Chloride (mg/L) 200
100
0 0 500 1000 1500 2000 TDS (mg/L) Figure 3.2. Chemical parameters. (A) TDS versus temperature (r2=0.08); (B) TDS versus pH (r2=0.64); (C) Chloride versus TDS (r2=0.99). Data points represent all wells and freshwater springs from both field samplings.
44 A 10
8
L 6
[Ca] mmol/ 4
2 - ]:[Ca] CO 3 2:1 [H
0 0246
- [HCO3 ] mmol/L
B 10
8 l/L 6
4
[Ca+Mg] mmo Girl Scout Deep Girl Scout Shallow Mg] 2 - ]:[Ca+ CO 3 LCRA Deep 2:1 [H LCRA Shallow 0 Paradise Alley Deep 0246Paradise Alley Shallow - [HCO3 ] mmol/L Comal Springs
10 C
8 L
] mmol/ 6 4
4 [Ca+Mg][SO - 2 Mg] - ]:[Ca+ CO 3 2:1 [H
0 0246
- [HCO3 ] mmol/L
- Figure 3.3. Concentrations of HCO3 versus (A) [Ca], (B) [Ca+Mg], and (C) [Ca+Mg]-[SO4]. Data points represent all wells and freshwater springs from both field samplings.
45 0.18 A
0.14
0.10
0.06
Calcite Saturation Index 0.02 equilibrium
-0.02 0 100 200 300 400 500 600 700
2- [SO4 ] (mg/L)
B 0.50 0.40 ex d 0.30 on In i 0.20
0.10 te Saturat i equilibrium Girl Scout Deep
om 0.00 l Girl Scout Shallow Do -0.10 LCRA Deep LCRA Shallow -0.20 Paradise Alley Deep 0 100 200 300 400 500 600 700 Paradise Alley Shallow 2- [SO4 ] (mg/L) Comal Springs
0 C equilibrium
-0.5
-1
-1.5
Gypsum Saturation Index Index Gypsum Saturation -2
-2.5 0 100 200 300 400 500 600 700
2- [SO4 ] mg/L
Figure 3.4. Saturation indices (log IAP/Ksp) of (A) calcite, (B) dolomite, and (C) gypsum versus sulfate concentrations. Note scale changes between graphs. Data points represent all wells and freshwater springs from both field samplings.
46
A 4 Ca2+ CaHCO3+ CaSO4 3 l 2 mMo
1
0
4 B Mg2+ MgSO4 MgHCO3+ 3
2 mMol
1
0 Girl Scout Deep Girl Scout Shallow LCRA Dee LCRA Shallow D Paradise Alley S Paradise Alley h e a e l p l o w
p
Figure 3.5. Speciation of (A) [Ca] and (B) [Mg] in transect wells based on complexes modeled in PHREEQC. Data points represent well chemistry from the first sampling.
47 the fresh and saline water can affect diagenetic processes (Deike, 1989; Groschen and Buszka,
1997) and indicate that the two water zones evolved through different processes. The water from the four freshwater wells (Girl Scout and LCRA wells) was Ca-Mg-HCO3 type, suggesting that limestone and dolomite dissolution influenced water chemistry (Oetting et al.,
1996). Excess cations were likely contributed through dissolution of other minerals with Ca
- or Mg but not HCO3 , such as from gypsum (Figure 3.3). Gypsum has been encountered in the Kainer Formation (Kirschberg evaporite) of the Edwards Group (Maclay and Small, 1986;
Sharp, 1990) (see Figure 1.2). Gypsum dissolution can cause calcite supersaturation due to the common ion effect (Plummer et al., 1990; Drever, 1997; Appelo and Postma, 2005).
Conversely, the Ca-Cl-SO4 hydrochemical facies of the saline wells (Paradise Alley wells) indicate that the saline waters evolved from additional processes based on evidence from thorough investigations on the source and evolution of the saline water zone, incorporating isotopic analysis, modeling, and geochemical data from surrounding aquifers and down-dip oil fields and brines (e.g., Land and Prezbindowski, 1981; Clement and Sharp,
1988; Oetting et al., 1996). Of the possible mechanisms affecting the composition of saline water in the study wells, cross-formational flow from the underlying Glen Rose Formation is most likely (Clement and Sharp, 1988; Oetting et al., 1996).
H2S in the Edwards Aquifer is derived from sulfate-reducing microbes based on previous isotopic studies (e.g., Rye et al., 1981). The notable changes in H2S concentrations from the transect wells over one month may indicate a dynamic microbial sulfur cycle within the aquifer. For example, both Girl Scout wells initially had no detectable H2S, but contained measurable amounts (<1ppm) one month later. Given that there were minor changes in TDS and ion concentrations in these wells over that month, which suggests minimal fluid mixing
48 across fault blocks, active microbial sulfate reduction produced H2S in the wells. Similarly, changes in the LCRA wells from low amounts of H2S to no measurable amounts one month later may indicate that microbial sulfur oxidation consumed the H2S more quickly than microbial sulfate reduction replenished it (see Chapter 2). The different changes in H2S levels between the freshwater wells may also suggest that there is a hydrologic disconnect between the nearby freshwater wells, not just between fresh and saline water, due to faulting, but perhaps also due to the heterogeneous and anisotropic nature of the karst aquifer (Fetter,
2000; White, 2002).
Although geochemical modeling indicated that the wells are saturated with respect to calcite and dolomite, a considerable concentration of Ca (~93 mg/L) discharges from Comal
Springs (Table 3.1 and 3.2), suggesting that Ca-bearing minerals, like calcite and dolomite, are likely still undergoing active dissolution. The modeling results only indicate what is thermodynamically possible and does not take kinetics into account (Parkhurst and Appelo,
1999). For carbonate systems, biological activity is an especially important kinetic aspect to consider, as microorganisms can shift the carbonate equilibria through production or consumption of CO2 (White, 1997; Appelo and Postma, 2005).
49 CHAPTER 4: GEOCHEMICAL AND MICROBIAL INFLUENCES ON EDWARDS AQUIFER CARBONATES
INTRODUCTION
The in situ microcosm approach was used to assess potential carbonate mineral dissolution and/or precipitation reactions along the transition zone from fresh to saline water in the Edwards Aquifer. The experimental microcosms were designed for sterile and reactive conditions to test abiotic and biotic effects on carbonate reactions, respectively, and to provide insight into how microbes, chemistry, or both, act to influence present-day aquifer processes.
MATERIALS AND METHODS
Two sets of microcosms, reactive and sterile, were made. Reactive microcosms consisted of small (125 mL), perforated HDPE containers containing fragments (chips) of
Iceland spar calcite (Wards Scientific) or dolomite (Wards Scientific). Sterile microcosms were perforated 50 mL containers, and the calcite and dolomite chips were placed into dialysis tubing (Spectra/Por 3) with a molecular weight cutoff of 3,500 Daltons. The tubing was placed into filter-sterilized (to 0.1µm) CaCO3-equilibrated solutions until microcosm deployment. The total chip mass in each microcosm was weighed three times and chips were sterilized in 100% ethanol prior to insertion in microcosms or dialysis tubing. Approximately
6 grams were used in reactive microcosms (average of ~11 chips) and 3 grams were used in each sterile dialysis bag (average of ~6 chips).
In the field, it was not possible to place microcosms at the screened interval due to sedimentation; therefore, microcosms were suspended within the larger diameter casing at
~20 to 30 m depth (Table 4.1) for 26 days. Possible changes in aqueous chemistry in the well during the experiment were accounted for by sampling purged wells (aquifer water) before the
50 experiment and from non-purged wells (casing water) when microcosms were retrieved
(Chapter 3).
Microcosms were retrieved from the wells, and microbial cells on some of the chip surfaces were preserved using the hexamethyldisilazane (HMDS) preservation technique, whereby chips were rapidly submerged in cacodylic acid, then submerged in 2.5% glutaraldehyde and washed in an ethanol series (10 – 100%) for critical point drying
(Vandevivere and Bevaye, 1992). Some chips were preserved in 100% ethanol and stored at
~20°C for later analyses. Preserved chips were weighed three times and then stored in sterile bags. Average changes in microcosm chip weight from pre- deployment to post-retrieval were determined for each microcosm, and P values were determined from student t-tests.
Several microcosm control sets were made to determine possible changes in chip mass as a consequence of the experimental design. Approximately ten dolomite and ten calcite chips were weighed and sterilized, but left in sterile bags and maintained as experimental controls. Half of these were also HMDS preserved. Six sets of both calcite and dolomite chips (~10 chips per set) were weighed, sterilized, and placed into dialysis tubing as controls
Table 4.1. Well depths, diameters, screened intervals, and microcosm suspension depths.
Total 4" 2" Screened Microcosm Well Depth1 diameter diameter Interval Depth1 Depth1 Girl Scout Deep 240 0-24 24-228 228-240 23 Girl Scout Shallow 198 0-24 24-185 185-198 31 LCRA Deep 226 0-30 30-213 213-226 18 LCRA Shallow 201 0-302 30-1862 186-2012 18 Paradise Alley Deep 223 0-23 23-207 207-223 23 Paradise Alley Shallow 173 0-23 23-158 158-173 23 1 all measurements in meters 2 approximated due to absence of driller’s logs
51 for the experimental sterile microcosms to ensure no changes in weight occurred within the tubing before microcosms were deployed. These chips were also preserved to assess possible weight changes as a result of the preservation process. Balance precision and accuracy were tested with certified analytical weights. Scale variability was ±0.008 mg.
One chip was used for scanning electron microscopy (SEM) to represent each microcosm experiment (i.e., sterile calcite, reactive calcite, sterile dolomite, and reactive dolomite) for each well. Twelve calcite chips and twelve dolomite chips were examined using the JEOL JSM – 840A Scanning Microscope in the Department of Geology and
Geophysics, Louisiana State University. Chips were mounted and gold-coated for 45 to 60 seconds. Calcite and dolomite control chips, preserved and non-preserved, were also mounted and gold-coated to observe background surficial textures (non-preserved) and potential preservation effects (preserved). Putative microbial cells on mineral surfaces were documented quantitatively by systematically viewing 100 frames at the same magnification
(5000x) for each chip. The number of putative cells was assessed over an area of 0.0625 mm2 per chip. Chips were also viewed qualitatively by scanning the surface for notable dissolution pits, precipitates, cell morphologies, colonization patterns, or other mineral surface features.
The SEM was typically operated at an optical aperture of 3, accelerating voltage of 20 kV, and 6 x 10-10 Amp probe current. Photomicrographs were digitally acquired using NIH Image
1.62 software (Rasband, 1996) and brightness or contrast adjustment was the only manual manipulation. When precipitates were observed, energy dispersive x-ray analysis (EDAX) was used to determine elemental composition.
52 RESULTS
Reactive microcosm chips lost more mass compared to sterile microcosm chips for both calcite and dolomite (Table 4.2 and Figure 4.1). Mass loss for sterile microcosm chips was generally insignificant (P > 0.05; Figure 4.1). Average mass loss for all experimental microcosms was 12.5 mg (n = 40). All control chips had < 2 mg weight change (n = 14)
(Table 4.2), which was within the standard deviation for most microcosms.
For calcite, the reactive microcosms had between 0.08 and 1.0% mass loss (average
0.33% ± 0.26, n = 12) and sterile microcosms had between 0.05 and 0.39% mass loss (average
0.21% ± 0.12, n = 8). One reactive microcosm from the Girl Scout Shallow well showed a loss of 450 mg; this was determined to be due to mechanical damage during transport and was not considered in further analysis. For calcite microcosms, only one sterile microcosm, observed in Girl Scout Deep, lost more mass than a reactive microcosm, but the difference was minor (2.8 mg). Four sterile microcosm dolomite chips had statistically insignificant weight gain. Average percent loss for reactive calcite microcosms was greatest in Girl Scout
Deep (0.56% ± 0.61, n = 2), followed by LCRA Deep (0.50% ± 0.10, n = 2) and Paradise
Alley Shallow (0.31% ± 0.02, n = 2). Average mass loss in calcite microcosms correlated with calcite saturation indices (Chapter 3), with a stronger correlation for sterile microcosms
(r2=0.77) compared to reactive microcosms (r2=0.55) (Figure 4.2). There was no correlation between average mass loss in dolomite microcosms and dolomite saturation indices.
The calcite and dolomite control chips were examined under the SEM for a background comparison. Control chips that were unpreserved had smooth, unaltered surface textures (Figure 4.3A). No cells, secondary precipitates, or dissolution features were observed. Small, sharp pits (averaging < 2µm2) were observed along cleavage planes, but
53 Table 4.2. Average mass loss from calcite and dolomite microcosms and associated significance. * denotes significant mass loss.
Calcite Dolomite Average Average Pre- and Post- Average Average Pre- and Post- Well Mass Lost Percent Weight P Mass Lost Percent Weight P (mg) Loss values2 (mg) Loss values2 Girl Scout Deep Sterile 8.93 0.28 0.01* 46.13 1.62 0.01* Reactive 6.13 0.13 0.01* 12.63 0.30 0.01* Reactive 66.77 0.99 0.01* 6.50 0.13 0.02 Girl Scout
Shallow Sterile 4.40 0.20 0.05 -0.501 0.02 0.89 Reactive 450.67 4.93 0.01* 50.90 1.12 0.01* Reactive 13.53 0.27 0.01* 15.73 0.32 0.01* LCRA Deep Sterile 9.13 0.33 0.06 1.27 0.04 0.59 Reactive 49.23 0.57 0.01* 4.73 0.12 0.04 Reactive 30.90 0.43 0.01* 4.83 0.10 0.33 LCRA Shallow Sterile 1.13 0.05 0.24 -1.871 0.07 0.62 Reactive 8.13 0.10 0.04 3.80 0.09 0.26 Reactive 17.97 0.21 0.03 7.23 0.13 0.01* Paradise Alley
Deep Sterile 4.73 0.18 0.01* -1.801 0.07 0.06 Sterile 1.53 0.08 0.21 -0.931 0.04 0.45 Reactive 6.43 0.08 0.01* 1.93 0.11 0.29 Reactive 12.20 0.20 0.01* 9.87 0.44 0.16 Paradise Alley
Shallow Sterile 5.10 0.23 0.08 7.80 0.37 0.18 Sterile 10.70 0.39 0.02 9.13 0.46 0.02 Reactive 23.03 0.29 0.02 0.00 0.00 1.00 Reactive 24.33 0.32 0.08 5.50 0.25 0.02 Controls Sterile -1.133 0.013 0.263 1.723 0.013 0.353 Reactive -0.011 0.01 0.23 -0.011 0.01 0.57
1 Negative values indicate a gain in mass (although changes were statistically insignificant) 2 P values: ≤ 0.001 Highly significant 0.001 – 0.01 Significant 0.01 – 0.05 Marginally significant > 0.05 Not significant 3 Values averaged from six control sets.
54 A 70 * Reactive Chips 60 Sterile Chips
50
) *
40
30 *
Mass Lost (mg Lost Mass ‡ 20 ‡ * ‡ † 10 * ‡ ‡ † † † 0 70 B Reactive Chips 60 Sterile Chips 50 ‡ * ) 40
30
20 Mass Lost (mg Lost Mass † * ‡ 10 ‡ ‡ † † 0
-10 Deep Girl Scout Shallow Girl Scout Shallow Paradise Alley Deep Paradise Alley Shallow LCRA Deep LCRA
* = pre- and post- weight P values ≤ 0.001 (highly significant) † = pre- and post- weight P values = 0.001 - 0.01 (significant) ‡ = pre- and post- weight P values = 0.01 – 0.05 (marginally significant)
Figure 4.1. Mass lost for reactive and sterile (A) calcite microcosms and (B) dolomite microcosms. Error bars represent standard deviation for each microcosm. Wells are represented in order of increasing salinity (Chapter 3).
55 0.16 Reactive Chips Sterile Chips 0.14
0.12
0.10
0.08
0.06
0.04 Calcite Saturation Index 0.02
0.00 0 1020304050 Average mass loss (mg)
Figure 4.2. Average mass loss (mg) versus calcite saturation index values for each well. Sterile calcite microcosms (r2=0.77) and reactive calcite microcosms (r2=0.55) are shown. Error bars represent the standard deviation for mass loss.
A B
C D
Figure 4.3. Surficial features of mineral control chips. (A) Relatively unaltered surface of unpreserved calcite control chip. (B) Small, deep pits considered to be not associated with dissolution processes because of observation on unpreserved control dolomite chip. (C) Mottling texture on preserved calcite control chip. (D) Preservation effects on dolomite chips sometimes resulting in “popcorn”-like features.
56 they were easily distinguishable as surficial defects (Figure 4.3B). Control samples that were
HMDS preserved had a “mottling” texture on the mineral surfaces (Figure 4.3C and D).
Similar textures were observed in experimental calcite microcosms from Girl Scout Deep and
Paradise Alley wells (Appendix B).
A total of 24 experimental chips were examined by using SEM. Some cells were found on sterile microcosm chips, although it is unclear how microbes could infiltrate the dialysis tubing. Cells on sterile chip surfaces may have resulted from manipulation in the lab prior or during ethanol washing, or were introduced from the dialysis tubing or solution
(although manufacturer details ensured sterility). Nevertheless, significantly more and diverse features were observed on the experimental chip surfaces. Putative cells of varying morphologies were found on all chips, including baciliform, coccoid, and filamentous; baciliform morphotypes were observed most frequently. A unique “corkscrew” morphology was found only on chip surfaces from the LCRA Deep well (Figure 4.4A and B). Cell size varied from less than 5 µm to over 10 µm in length for some filaments, but average cell size was ~1 to 2 µm (Figure 4.4C). Colonies or clusters of cells were a regular observation
(Figure 4.4D and E). Some cells showed signs of desiccation (Figure 4.4F), which may be due to the preservation process (Esteban et al., 1998). Biofilm or exopolysaccharides (EPS) was also observed on several chip surfaces (Figure 4.5A). The material seemed to be attached to the mineral surface, and in one case appeared to be “ripped” (Figure 4.5B), possibly from the preservation process. Cell attachment features were also observed (Figure 4.5A and C).
Putative microbial cells were also associated with dissolution and/or precipitation features, but dissolution features were found on all chips. When microbes were found in association with these features, however, the dissolution pits were smaller and deeper (Figure
57 A B C
D E F
Figure 4.4. Cellular morphology and behavior. Unique “corkscrew” morphology from LCRA Deep (A) reactive calcite and (B) reactive dolomite chips with rods and cocci. (C) Long filamentous cell from Paradise Alley Shallow reactive dolomite chip. Colonies observed on reactive dolomite chips from (D) Girl Scout Deep and (E) LCRA Shallow. (F) Possible desiccation feature in cells from Paradise Alley Deep reactive dolomite chip.
A. B. C.
Figure 4.5. Cell-surface associations. (A) Biofilm material and attachment structures from Girl Scout Shallow non-reactive dolomite chip. (B) Possible micro-scale “rips” in EPS attachment to dolomite surface in LCRA Deep (non-reactive chip). (C) Attachment features on cells from LCRA Shallow, non-reactive dolomite chip.
58 4.6A through D). Larger, broader areas of dissolution and dissolution pits were found with and without microbial cells in the nearby vicinity (Figure 4.6E through I). Secondary precipitates were common, as well, and observed on at least one chip in every well with the exception of Paradise Alley Deep. Precipitates included iron sulfides (average size ~1µm2) consistent with pyrite (Figure 4.7A and B), and silicate-clay-like minerals with needle-like crystals containing calcium, magnesium, silica, and oxygen (Figure 4.7C through F) confirmed with EDAX spectra (Figure 4.8). Occasionally, microbes were found on top of or attached to the secondary precipitates (Figure 4.7G, H, and I).
Because viewing the entire mineral surface under the SEM was not feasible, examining a predefined surface area for each chip allowed for quantitative comparisons to be made between chips and among the transect wells. More cells per surface area were observed on dolomite chips compared to calcite chips for almost every well (Figure 4.9). The LCRA
Deep well had the most cells observed on reactive calcite plus dolomite chips, with the Girl
Scout wells having the fewest cells. There was a weak correlation between the documented number of putative cells and average mass loss for calcite chips (Figure 4.10A), but essentially no correlation for dolomite chips (Figure 4.10B).
DISCUSSION
The overall goal of this study was to assess carbonate dissolution from geochemical and microbial processes along a geochemical gradient in the Edwards Aquifer. Geochemical results indicated that the waters from study wells were saturated with respect to calcite and dolomite. However, experimental in situ calcite and dolomite chips showed a mass loss over a one month period. I propose that microbial activity contributed to carbonate dissolution,
59
A B C
D E F
G H I
Figure 4.6. Dissolution features. Smaller, deeper pits associated with microbial cells on reactive calcite chips from (A) LCRA Deep, (B and C) Paradise Alley Shallow, and (D) LCRA Shallow. Larger, broader dissolution pits within the vicinity of cells on reactive dolomite chips from (E) Girl Scout Shallow and (F) LCRA Shallow, and on reactive calcite chip from (G) LCRA Deep. Larger areas of dissolution without microbial cells nearby on (H and I) non-reactive calcite chips from Girl Scout Shallow.
60 A B C
D E F
G H I
Figure 4.7. Secondary precipitates on experimental mineral surfaces. Iron sulfide precipitate (Figure 4.8A), on (A) a reactive dolomite chip from Girl Scout Shallow and (B) a reactive calcite chip from Girl Scout Deep. Silicate clay-like mineral (Figure 4.8B) with no observable cells from (C) a non-reactive dolomite chip in Girl Scout Shallow. Silicate clay- like mineral, with cells in the vicinity, on (D and E) reactive dolomite chips from LCRA Shallow and (F) a non-reactive calcite chip from Paradise Alley Shallow. Cells were observed in direct contact with secondary precipitates on reactive dolomite chips in (G) LCRA Deep, (H) Paradise Alley Shallow, and (I) LCRA Shallow.
61 A S
Ca OMg Fe Au
0 200 400 600 800 1000 1200 1400
B Mg Ca Au Si
O
0 200 400 600 800 1000 1200 1400
Figure 4.8. EDAX spectra for (A) iron sulfide mineral and (B) silica-bearing mineral.
Sterile Calcite Chips 600 Reactive Calcite Chips 500 Sterile Dolomite Chips Reactive Dolomite Chips
a 400
300
Cells/unit are 200
100 Deep Girl Scout Shallow Girl Scout Shallow LCRA Shallow Paradise Alley Deep Paradise Alley 0 Deep LCRA
Figure 4.9. Number of putative cells observed on reactive calcite and dolomite chips per examined chip area (0.0625 mm2).
62 A 70
60
50
40 * * 30 * 20
Average mass loss (mg) (mg) loss Average mass * * 10 *
0 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 5.E+07 Average number of cells/microcosm
B 70
60
50
40 * 30
20 Average (mg) loss mass 10 * * * * * 0 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 5.E+07 Average number of cells/microcosm
Girl Scout Deep Girl Scout Shallow LCRA Deep LCRA Shallow Paradise Alley Deep Paradise Alley Shallow
Figure 4.10. Average amount of mass loss (mg) versus the total average number of cells per microcosm. Total average number of cells was extrapolated from number of putative cells observed on representative chip surfaces for (A) all calcite microcosms (r2=0.50) and (B) all dolomite microcosms (r2=0.001). * denotes reactive microcosms. Error bars represent the standard deviation for mass loss.
63 and that differences between the fresh and saline zone results were dependent upon local geochemical conditions, microbial community structure, and mineral surface cell density.
Microbial activity enhanced the dissolution of calcite and dolomite chip surfaces, based on the observation that reactive microcosm chips usually lost more mass than non-reactive microcosm chips. Examination of mineral chip surfaces under SEM confirmed the colonization of microbial cells. Molecular analysis of the bacterial diversity also provided evidence of microbially-mediated dissolution because microbes capable of producing corrosive agents (i.e., CO2) through their metabolism were prevalent in all study wells
(Chapter 2). Some biological molecules, such as EPS in biofilms observed on several chip surfaces (Figure 4.5), can also promote carbonate dissolution (Barton and Northup, 2007).
Additionally, cells associated with dissolution zones and secondary precipitates suggest that active processes were occurring on the mineral surfaces during the experiment.
Local well geochemistry influenced carbonate dissolution both directly and indirectly.
Geochemical conditions directly contributed to the abiotic calcite dissolution, as evidenced by the correlation between calcite saturation indices and mass loss in sterile microcosms (Figure
4.2). Sterile microcosm chips that were placed in waters closer to equilibrium with respect to calcite resulted in greater mass loss, as these fluids were closer to reaching undersaturation.
Because the correlation between calcite saturation indices and mass loss in reactive microcosms was weaker, the geochemistry of the bulk solutions may not be as important to calcite dissolution when microorganisms colonize the surfaces and create localized geochemical microenvironments on mineral surfaces.
Aqueous geochemistry also affected carbonate dissolution indirectly because, as determined previously (Chapter 2), geochemistry is responsible for most of the bacterial
64 genetic variability among wells (Figures 2.2 and 2.3). As a result, the composition of the microbial communities changed from well to well, namely in the relative abundances of putative chemoorganotrophs and chemolithoautotrophs; for example, sulfidic wells harbored greater amounts of bacteria related to putative sulfur-oxidizers. The different microbial communities then produce materials of varying corrosiveness, depending on their metabolism, that may contribute to carbonate dissolution (e.g., Ehrlich, 1996; Northup and Lavoie, 2001).
Surface-attached respiratory microbes increase CO2 levels on mineral surfaces, and metabolically active sulfur-oxidizing bacteria contribute additional acidity through H2SO4 production, both of which induce carbonate dissolution. The greatest percent of mass loss in reactive calcite microcosms was Girl Scout Deep, followed by LCRA Deep and Paradise
Alley Shallow (Table 4.2). These latter two wells were sulfidic, so it is not surprising that they had greater percent mass loss because they had greater potential for microbially- mediated dissolution through metabolic production of both CO2 and H2SO4. The potential for a mix of microbially-mediated dissolution through both carbonic acid and sulfuric acid has been noted previously from karst settings (e.g., Hose and Pisarowicz, 1999), resulting in increased speleogenesis.
The correlation between average calcite mass loss and chip surface cell density suggests that the amount of biomass colonizing surfaces directly impacted the amount of dissolution. Presumably, more microbial colonization contributed to more corrosive metabolic byproducts onto chip surfaces based on the metabolic pathways inferred from phylogenetic analysis (i.e., CO2 production from chemoorganotrophy). Although the Girl
Scout Deep reactive microcosm did not follow this trend, the average value had a much higher standard deviation than the other values, revealing possible inconsistencies in the data.
65 Additionally, one important factor to consider in this correlation is time. Microcosm experiments were deployed for ~1 month, so high biomass density was not achieved.
However, biomass accumulation with time can actually inhibit calcite dissolution by effectively coating the mineral surface (i.e., Lüttge and Conrad, 2004; Randall, 2006).
The observation that dolomite mass loss did not correlate with saturation indices or surface cell density may be reflective of the relatively poor understanding of this mineral in the aquifer (Maclay and Small, 1986; Deike, 1990). However, the observation of more cells on dolomite chips compared to calcite chips suggests that microbes may be preferentially colonizing dolomite over calcite, perhaps in order to exploit trace Fe in dolomite. This strategy has been previously documented in dolomite and calcite (Jones, 1989), as well as in the preferential colonization of silicates to obtain limiting nutrients like phosphorus and iron
(Rogers and Bennett, 2004).
Because all well waters modeled as being saturated with respect to calcite and dolomite, it is unclear why sterile microcosms also showed a loss in chip mass, albeit typically insignificant (Table 4.2). Several possible factors may explain what occurred. First, most geochemical models only account for thermodynamics and do not consider kinetic effects (Kehew, 2001). For example, dolomite precipitation is rare even though solutions may be oversaturated (e.g., Land, 1998; Warren, 2000). Moreover, in some instances, mineral surface conditions may not be in equilibrium with the bulk water composition, and the sampling of water some distance away from a mineral surface may not be representative of the mineral surface environment (Sjoberg and Rickard, 1984; Compton and Pritchard, 1990;
MacInnis and Brantley, 1992). Simpler explanations are also possible, such as mechanical damage to the mineral chips. Nevertheless, the possible in situ contamination of sterile
66 microcosms cannot be ignored. Even minor amounts of microbial colonization may have had an effect on dissolution, as microbial influences are most intense for calcite when solutions are near-equilibrium to the mineral (White, 1997).
It is tempting to suggest that microbial activity is responsible for the secondary precipitates documented on experimental chips, especially when cells were observed next to or on top of the minerals. Ehrlich (1996) states that microbes will precipitate minerals (1) in the oxidation or reduction of a dissolved inorganic species for energy metabolism, (2) in the detoxification of toxic inorganic species, (3) in the uptake of one or more dissolved inorganic species followed by conversion into a cellular support or protective structure, and/or (4) in enhancing their competitiveness in a microbial community. Bacterial precipitation of clay minerals has been previously documented (e.g., Ehrlich, 1998; Konhauser and Urrutia, 1999), including in cave and karst environments (Northup and Lavoie, 2001). Formation of iron-rich precipitates has also been previously associated with microbial activity (Ehrlich, 1996;
Ehrlich, 1998). Specifically, iron sulfides may form as a result of sulfate-reducing bacteria
(Konhauser, 1998; Gramp et al., 2009). The direct precipitation of pyrite is thermodynamically unfavorable (Konhauser, 1998), so microbes are more likely to precipitate iron sulfides like mackinawite (FeS) or greigite (Fe3S4) (Gramp et al., 2009). Active microbial sulfate reduction is probably occurring in the aquifer based on the observation of secondary iron sulfide precipitates on chip surfaces, identification of microorganisms closely related to sulfate-reducing bacteria (Chapter 2), and the production of H2S within the Girl
Scout well casings over a one-month period (Chapter 3). Even though sulfate-reducing organisms were not identified in every well from genetic sequencing, the diversity was not
67 fully sampled (Chapter 2), so it is possible that this group is not abundant enough in the aquifer to be represented with the 16S rRNA gene sequencing methods used in this study.
In summary, dissolution of fresh carbonate surfaces occurred in the Edwards Aquifer.
Microbes enhanced the amount of dissolution based on the greater mass loss observed in reactive microcosms. The aqueous geochemistry (saturation indices) directly influenced the amount of abiotic dissolution in calcite. Additionally, local geochemical parameters (H2S levels, TDS, and depth) controlled the structure of microbial groups. Different microbial groups (i.e., chemoorganotrophs and chemolithoautotrophs) utilized different metabolic pathways that contributed byproducts of varying corrosiveness (i.e., CO2 and H2SO4) for carbonate dissolution. The quantity of cells attached to chip surfaces resulted in enhanced calcite dissolution, presumably because of the increase in corrosive metabolic byproducts in contact with the mineral surface.
68 CHAPTER 5: SUMMARY AND FUTURE DIRECTIONS
The subsurface environment is exploited by microorganisms in terrestrial aquifers, the basaltic subseafloor, and deep petroleum reservoirs (e.g., Amend and Teske, 2005). Microbes gain energy from their surroundings, which in this environment may be organic, photosynthetic products or inorganic chemicals. Microorganisms actively influence their environment by mediating nutrient cycling, precipitating minerals, and enhancing rock weathering through their metabolism (Ehrlich, 1996; Northup and Lavoie, 2001). Karst and cave (limestone) environments may be particularly sensitive to the microbial enhancement of mineral dissolution due to biotic production of CO2, resulting in carbonic acid dissolution that commonly forms karst (e.g., White, 1997; Kehew, 2001). Additionally, production of sulfuric acid from sulfur-oxidizing bacteria may also contribute to karst and cave formation (e.g.,
Sarbu et al., 2000; Engel et al., 2004).
The Edwards Aquifer provides a natural study site to examine these processes across a transition from fresh to saline water in the aquifer. The aquifer is locally important to people in central Texas and the endemic cave-inhabiting fauna. Examination of the aquifer microbial communities is important in understanding their potential for bioremediation or dispersion of toxic elements (Wolicka and Borkowski, 2007). Investigating the microbial processes in the
Edwards Aquifer offers insight into the factors influencing subsurface secondary porosity formation and karstification.
The prevailing goal of this study was to examine the effects of geochemistry and microbial activity on carbonate dissolution along a geochemical gradient. To accomplish this,
16S rRNA gene sequencing was performed to establish the bacterial phylogeny of both fresh and saline water zones in the Edwards Aquifer, and to examine possible metabolic pathways
69 being utilized by aquifer microbes and the subsequent metabolic byproducts that could influence aquifer carbonates (Chapter 2). Field and laboratory analyses were conducted on water samples to determine geochemical parameters, carbonate mineral saturation states, and the role of aqueous geochemistry in carbonate dissolution (Chapter 3). Finally, an in situ microcosm approach, simulating both sterile and reactive conditions, was used to experimentally determine if calcite and/or dolomite is dissolving or precipitating in the aquifer today (Chapter 4).
The first specific objective in this research was to establish the microbial diversity in the fresh and saline zones of the aquifer. The working hypothesis was that metabolically distinct microbial groups will inhabit the different geochemical zones (i.e., fresh versus saline water) due to the different energy sources available. Several microbial groups, most notably chemoorganotrophic members of Proteobacteria, were ubiquitous in all wells, regardless of local geochemical conditions. However, the distribution of some groups varied from well to well; for example, potential chemolithoautotrophic organisms related to previously described putative sulfur oxidizers were more abundant in wells with an initial presence of H2S.
Analysis through the program UniFrac indicated that H2S levels, TDS concentrations
(alkalinity), and total well depth accounted for ~80% of the genetic variability among wells, demonstrating the importance of local geochemistry on microbial communities. Thus, the hypothesis was supported but should be slightly modified to state that the distribution of metabolically distinct microbial groups will vary depending upon local geochemical parameters in different parts of the aquifer.
The second specific objective was to examine carbonate dissolution in regards to: (1) biological versus abiotic impacts, and (2) differences between fresh and saline water zones.
70 The working hypothesis was that carbonate dissolution will be enhanced by microbial activity and be more pronounced in the saline water zone compared to the freshwater zone due to the potential for more highly corrosive metabolic byproducts from microbes. Examination of mineral chip surfaces under SEM confirmed the colonization of microbial cells. Microcosm experiments were created to differentiate between interactions with calcite and dolomite chips compared to sterile (or nearly sterile) controls. Reactive chips lost more statistically significant mass, lending support to microbially-enhanced dissolution of carbonates in the
Edwards Aquifer. The amount of dissolution varied depending on geochemistry, microbial communities, and surface cell density. Wells that were initially sulfidic had high percent mass loss in experimental calcite microcosms, had organisms potentially capable of producing sulfuric acid as a metabolic byproduct, and showed more cells on reactive chip surfaces when viewed microscopically. Based on these results, I suggest that the potential for microbially- enhanced carbonate dissolution is greater in the saline (sulfidic) water zone, provided that a sufficient amount of cells are colonizing mineral surfaces.
Future studies of the Edwards Aquifer may benefit from additional microbiological analysis or microcosm experiments. Phylogenetic analysis of freshwater Comal Springs is underway to provide a comparison between spring water and subsurface water microbial populations (e.g., Shabarova and Pernthaler, 2009). Targeting functional bacterial genes may provide a more thorough understanding of microbial function at a community or ecosystem level, and may assist in examining microbially-mediated nutrient cycling within the aquifer.
The archaeal and eukaryotic diversity of the aquifer also remains unexplored. Fully characterizing the microbial diversity in aquifers may allow for potential groundwater biomonitoring of contamination or climate change (e.g., Pronk et al., 2009). Furthermore,
71 because diverse microbial communities are thought to be more resilient to environmental stresses (Griebler and Lueders, 2009), a complete representation of the microbial diversity may be important to assess aquifer vulnerability to contamination.
Possible microcosm studies in the future could be improved by incorporating microcosms for a longer duration in situ, to inspect microbial processes that occur over an extended time period. Prolonged emplacement would also allow for a sufficient amount of biomass to accumulate on mineral surfaces, a prerequisite for full-cycle techniques like FISH to compare planktonic and attached microbial communities (e.g., Reardon et al., 2004).
However, deploying microcosms for extended time periods may result in reduced dissolution due to biomass accumulation covering chip surfaces (Lüttge and Conrad, 2004), but these observed differences over various time scales are important to assess. Creating a more sturdy apparatus for sterile microcosms, while still allowing for the flow of water, is a challenge that would offer improved confidence in results concerning abiotic processes in the aquifer.
Examining wells on an extended transect from fresh to saline water would also prove interesting. Results from this study indicate that organic carbon is an essential component for aquifer communities, based on the dominance of chemoorganotrophic organisms. However, it is unclear to what extent chemoorganotrophs persist in the saline zone. It is possible that, at an increasing distance from the transition zone and into the saline water zone, chemolithoautotrophic organisms are the dominant community members.
Aquifer geochemistry and microbial activity influence carbonates in the Edwards
Aquifer. Geochemistry directly affects abiotic carbonate dissolution depending upon saturation indices. Local geochemical parameters, specifically H2S, TDS, and depth, are major factors influencing the distribution of microbial groups in the fresh and saline water
72 zones. These different microbial groups, in turn, exert corrosive materials of varying degrees through their metabolism, and if attached to mineral surfaces in adequate amounts, affect the solubility of the fresh carbonate surfaces. Thus, microbially-mediated dissolution of fresh carbonate surfaces occurs in both fresh and saline water zones of the Edwards Aquifer.
73 REFERENCES
Amend, J.P., and Teske, A., 2005. Expanding frontiers in deep subsurface microbiology. Palaeogeography, Palaeoclimatology, Palaeoecology, v. 219, pp. 131-155.
American Public Health Association (APHA), 1998, in Clesceri, L.S., Greenberg, A.E., Eaton, A. D. (eds.), Standard Methods for the Examination of Water and Wastewater, 20th ed., US Environmental Protection Agency, American Public Health Association, the American Water Works Association, and the Water Environmental Federation, p. 1220.
Angert, E.R., Northup, D.E., Reysenbach, A., Peek, A.S., Goebel, B.M., and Pace, N.R., 1998. Molecular phylogenetic analysis of a bacterial community in Sulphur River, Parker Cave, Kentucky. American Mineralogist, v. 83, pp. 1583-1592.
Appelo, C.A.J., and Postma, D., 2005. Geochemistry, groundwater, and pollution, 2nd edition. A.A. Balkema Publishers, Leiden, The Netherlands.
Barton, H.A., Taylor, M., and Pace, N., 2004. Molecular phylogenetic analysis of a bacterial community in an oligotrophic cave environment. Geomicrobiology Journal, v. 21, pp. 11-20.
Barton, H.A., and Luiszer, F., 2005. Microbial metabolic structure in a sulfidic cave hot spring: potential mechanisms of biospeleogenesis. Journal of Cave and Karst Studies, v. 67, pp. 28-38.
Barton, H.A., and Northup, D.E., 2007. Geomicrobiology in cave environments: Past, current, and future perspectives. Journal of Cave and Karst Studies, v. 69, pp. 163- 178.
Barton, H.A., Taylor, N.M., Kreate, M.P., Springer, A.C., Oehrle, S.A., and Bertog, J.L., 2007. The impact of host rock geochemistry on bacterial community structure in oligotrophic cave environments. Int. J. Speleology, v. 36, pp. 93-104.
Bates, C.L., Forstner, M.R.J., Barnes, M.B., Whiteley, M., McLean, R.J., 2006. Heterotrophic limestone-adherent biofilm isolates from the Edwards Aquifer, Texas. The Southwestern Naturalist, v. 51, pp. 299-309.
Batzke, A., Engelen, B., Sass, H., and Cypionka, H., 2007. Phylogenetic and physiological diversity of cultured deep-biosphere bacteria from equatorial Pacific Ocean and Peru margin sediments. Geomicrobiology, v. 24, pp. 261 – 273.
Bennett, P.C., and Engel, A.S., 2005. Role of micro-organisms in karstificiation. In: Gadd, G.M., Semple, K.T., Lappin-Scott, H.M. (eds.), Micro-organisms and Earth
74 Systems—Advances in Geomicrobiology. Cambridge University Press, Cambridge, U.K., pp. 345-363.
Bennett, P.C., Hiebert, F.K., and Choi, W.J., 1996. Microbial colonization and weathering of silicates in a petroleum-contaminated groundwater. Chemical Geology, v. 132, pp. 45-53.
Birdwell, J.E., and Engel, A.S., 2009. Variability in terrestrial and microbial contributions to dissolved organic matter fluorescence in the Edwards Aquifer, central Texas. Journal of Cave and Karst Studies, v. 71(2), pp. 144-156.
Bögli, A., 1965. The role of corrosion by mixed water in cave forming. In Stelcl, O. (ed.), Problems of the Speleological Research, Academia, Prague, p. 125-131.
Bottrell, S.H., Smart, P.L., Whitaker, F., and Raiswell, R., 1991. Geochemistry and isotope systematics of sulfur in the mixing zone of Bahamian blue holes. Applied Geochemistry, v. 6, pp. 97-103.
Boyd, E.S., Cummings, D.E., Geesey, G.G., 2007. Mineralogy influences structure and diversity of bacterial communities associated with geological substrata in a pristine aquifer. Microbial Ecology, v. 54, pp. 170-182.
Cederberg, J.R., Ging, P.B., and Ourso, R.T., 1998. Monitoring of selected water-quality constituents near the freshwater/saline water interface of the Edwards Aquifer, July 1996 – December 1997: USGS Fact Sheet FS-103-98.
Chapelle, F.H., O’Neill, K., Bradley, P.M, Methe, B.A., Clufo, S.A., Knobel, L.L., and Lovley, D.R., 2002. A hydrogen-based subsurface microbial community dominated by methanogens. Nature, v. 415, pp. 312-315.
Clark, A.K., and Small, T.A., 1997. Geologic framework of the Edwards Aquifer and upper confining unit, and hydrogeologic characteristics of the Edwards Aquifer, south- central Uvalde County, Texas. U.S. Geological Survey Water-Resources Investigations Report 97—4094, Austin, TX.
Clement, T.J., and Sharp, J.M. Jr., 1988. Hydrochemical facies in the bad-water zone of the Edwards Aquifer, Central Texas. Proceedings of the Groundwater Geochemistry Conference, National Water Well Association, pp. 124-129.
Compton R.G., and Pritchard, K.L., 1990. The dissolution of calcite at pH > 7: Kinetics and mechanisms. Phil. Trans. R. Soc. Lond. A., v. 330, pp. 47-70.
Deike, R.G., 1990. Dolomite dissolution rates and possible Holocene dedolomitization of water-bearing units in the Edwards Aquifer, south-central Texas. Journal of Hydrology, v. 112, pp. 335-373.
75 DeSantis, T. Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E. L., Keller, K., Huber, T., Dalevi, D., Hu, P., and Andersen, G. L., 2006. Greengenes, a Chimera-Checked 16S rRNA Gene Database and Workbench Compatible with ARB. Appl. Environ. Microbiol., v. 72, pp. 5069-72.
Drever, J.I., 1997. The geochemistry of natural waters: Surface and groundwater environments, 3rd edition. Prentice Hall, p. 436.
Edwards, K.J., Bach, W., and McCollom, T.M., 2005. Geomicrobiology in oceanography: microbe–mineral interactions at and below the seafloor. Trends in Microbiology, v. 13, pp. 449-456.
Egemeier, 1981. Cavern development by thermal waters. National Speleological Society Bulletin, v. 43, pp. 31-51.
Ehrlich, H.L., 1996. How microbes influence mineral growth and dissolution. Chemical Geology, v. 132, pp. 5-9.
Ehrlich, H.L., 1998. Geomicrobiology: its significance for geology. Earth Science Reviews, v. 45, pp. 45-60.
Engel, A.S., Meisinger, D.B., Porter, M.L., Payn, R.A., Schmid, M., Stern, L.A., Schleifer, K.H., and Lee, N.M., 2010. Linking phylogenetic and functional diversity to nutrient spiraling in microbial mats from Lower Kane Cave (USA). The ISME Journal, v. 4, pp. 98–110.
Engel, A.S., and Randall, K.W., 2008. Microbially enhanced carbonate dissolution in karst aquifers. In Sasowsky, I.D., Feazel, C.T., Mylroie, J.E., Palmer, A.N., and Palmer, M.V. (eds.), Karst from Recent to Reservoirs, Karst Waters Institute Special Publication 14, pp. 52-56.
Engel, A.S., Stern, L.A., and Bennett, P.C., 2004. Microbial contributions to cave formation; new insights into sulfuric acid speleogenesis. Geology Boulder, v. 32, no. 5, p. 369- 372.
Engel, A.S., Porter, M.L., Stern, L.A., Quinlan, S., and Bennett, P.C., 2004b. Bacterial diversity and ecosystem function of filamentous microbial mats from aphotic (cave) sulfidic springs dominated by chemolithoautotrophic “Epsilonproteobacteria”. FEMS Microbiology Ecology, v. 51, pp. 31-53.
Engel, A.S., Lee, N., Porter, M.L., Stern, L.A., Bennett, P.C., Wagner, M., 2003. Filamentous “Epsilonproteobacteria” dominate microbial mats from sulfidic cave springs. Applied and Environmental Microbiology, v. 69, pp. 5503-5511.
76 Engelen, B., Ziegelmüller, K., Wolf, L., Köpke, B., Gittel, A., Cypionka, H., Treude, T., Nakagawa, S., Inagaki, F., Lever, M.A., and Steinsbu, B.O., 2008. Fluids from the oceanic crust support microbial activities within the deep biosphere. Geomicrobiology, v. 25, pp. 56-66.
Esteban, G.F., Clarke, K.J., and Finlay, B.J., 1998. Rapid techniques for the identification of free-living Protozoa. In Burlage, R.S, Atlas, R., Stahl, R., Geesey, G., and Sayler, G., (eds.), Techniques in microbial ecology, p. 468.
Fedosov, D.V, Podkopaeva, D.A., Miroshnichenko, M.L., Bonch-Osmolovskaya, E.A., Lebedinsky, A.V., and Grabovich, M.Yu, 2006. Investigation of the catabolism of acetate and peptides in the new anaerobic thermophilic bacterium Caldithrix abyssi. Microbiology, v. 75, pp. 119-124.
Fetter, C.W., 2000. Applied Hydrogeology, 4th edition. Prentice Hall, Upper Saddle River, N.J.
Ford, D.C., and Williams, P.W., 2007. Karst Geomorphology and Hydrology, 2nd edition. John Wiley & Sons, London, p. 576.
Frederickson, J.K., Balkwill, D.L, Drake, G.R., Romine, M.F., Ringelberg, D.B., and White, D.C., 1995. Aromatic-degrading Sphingomonas isolates from the deep subsurface. Appl. Environ. Microbiol., v. 61, pp. 1917-1922.
Fredrickson, J.K., and Onstott, T.C., 2001. Biogeochemical and geological significance of subsurface microbiology. In: Fredrickson, J.K., Fletcher, M. (eds.), Subsurface microbiology and biogeochemistry. Wiley-Liss, New York, pp. 3-37.
Friedrich, C.G., Bardischewsky, F., Rother, D., Quentmeier, A., and Fischer, J., 2005. Prokaryotic sulfur oxidation. Current Opinion in Microbiology, v. 8, pp. 253-259.
Fuhrman, J.A., Comeau, D.E., Hagstrom, A., and Chan, A.M., 1988. Extraction from natural planktonic microorganisms of DNA suitable for molecular biological studies. Applied and Environmental Microbiology, v. 54(6), pp. 1426-1429.
Ghosh, W., and Roy, P., 2006. Mesorhizobium thiogangeticum sp. nov., a novel sulfur- oxidizing chemolithoautotroph from rhizosphere soil of an Indian tropical leguminous plant. Int. J. Syst. and Evol. Microbiol., v. 56, pp. 91-97.
Gramp, J.P., Bigham, J.M, Jones, F.S., and Tuovinen, O.H., 2009 (in press). Formation of Fe-sulfides in cultures of sulfate-reducing bacteria. Journal of Hazardous Materials, doi:10.1016/j.jhazmat.2009.10.119.
Griebler, C. and Lueders, T., 2009. Microbial diversity in groundwater ecosystems. Freshwater Biology, v. 54, pp. 649-677.
77 Groschen, G.E., and Buszka, P.M., 1997. Hydrogeologic framework and geochemistry of the Edwards aquifer saline-water zone, south-central Texas. U.S. Geological Survey Water-Resources Investigations Report 974133, 47 pp.
Guyton, W.F., and Associates, 1979. Geohydrology of Comal, San Marcos, and Hueco Spring. Texas Department of Water Resources, Report 234.
Hall, T., 1999. A user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT: Nucleic Acids Symposium Ser., v. 41, pp. 95-98.
Hammes, F., and Verstraete, W., 2002. Key roles of pH and calcium metabolism in microbial carbonate precipitation. Reviews in Environmental Science and Biotechnology, v. 1, pp. 3-7.
Hiebert, F.K., and Bennett, P.C., 1992. Microbial control of silicate weathering in organic- rich ground water. Science, v. 258, pp. 278-281.
Hill, C.A., 1990. Sulfuric acid speleogenesis of Carlsbad Cavern and its relationship to hydrocarbons, Delaware Basin, New Mexico and Texas. AAPG Bulletin, v. 74, pp. 1685-1694.
Hose, L. D., Palmer, A. N., Palmer, M. V., Northup, D. E., Boston, P. J., and DuChene, H. R., 2000. Microbiology and geochemistry in a hydrogen-sulphide-rich karst environment. Chem. Geol., v. 169, pp. 399–423.
Hose, L.D., and Pisarowicz, J.A., 1999. Cueva de Villa Luz, Tabasco, Mexico: Reconnaisance study of an active sulfur spring cave and ecosystem. Journal of Cave and Karst Studies, v. 61, pp. 13-21.
Ishii, S., Yamamoto, M., Kikuchi, M., Oshima, K., Hattori, M., Otsuka, S., and Senoo, K., 2009 (in press). Microbial populations responding to denitrification-inductive conditions in rice paddy soil, as revealed by comparative 16S rRNA gene analysis. Appl. Environ. Microbiol., doi:10.1128/AEM.01481-09.
Jagnow, D.H., Hill, C.A., Davis, D.G., DuChene, H.R., Cunningham, K.I., Northup, D.E., and Queen, M.J., 2000. History of sulfuric acid theory of speleogenesis in the Guadalupe Mountains, New Mexico. Journal of Cave and Karst Studies, v. 62, pp. 54-59.
Jones, B., 1989. The role of microorganisms in phytokarst development on dolostones and limestones, Grand Cayman, British West Indies. Canadian J. of Earth Sci., v. 26, pp. 2204-2213.
Jones, W.K., Horton, H.H., Wicks, C.M., Currie, R.R., Hose, L.D., Kerbo, R.C., Goodbar, J.R., and Trout, J., 2003. Recommendations and guidelines for managing caves on protected lands: Karst Waters Institute, special publication 8.
78 Kehew, A., 2001. Applied Chemical Hydrology. Prentice Hall, Upper Saddle River, N.J.
Kieft, T.L., McCuddy, S.M., Onstott, T.C., Davidson, M., Lin, L., Mislowack, B., Pratt, L., Boice, E., Lollar, B.S., Lippmann-Pipke, J., Pfiffner, S.M., Phelps, T.J., Gihring, T., Moser, D., and van Heerden, A., 2005. Geochemically generated, energy-rick substrates and indigenous microorganisms in deep, ancient groundwater. Geomicrobiology Journal, v. 22, pp. 325-335.
Kinkle, B., and Kane, T.C., 2000. Chemolithoautotrophic microorganisms and their potential role in subsurface environments, In: Wilkens, H., Culver, D.C., and Humphreys, W.F. (eds.), Ecosystems of the World 30, Elsevier, Amsterdam, pp. 309–318.
Konhauser, K.O, and Urrutia, M.M., 1999. Bacterial clay authigenesis: a common biogeochemical process. Chemical Geology, v. 161, pp. 399-413.
Konhauser, K.O., 1998. Diversity of microbial iron mineralization. Earth Science Reviews, v. 43, pp. 91-121.
Kotelnikova, S., 2002. Microbial production and oxidation of methane in deep subsurface. Earth-Science Reviews, v. 58, pp. 367-395.
Kremer, B., Kazmierczak, J., and Stal, L.J., 2008. Calcium carbonate precipitation in cyanobacterial mats from sandy tidal flats of the North Sea. Geobiology, v. 6, pp. 46- 56.
Land, L., 1998. Failure to precipitate dolomite at 25°C from dilute solution despite 1000-fold oversaturation after 32 years. Aquatic Geochemistry, v. 4, pp. 361-368.
Land, L.S., and Prezbindowski, D.R., 1981. The origin and evolution of saline formation water, lower Cretaceous carbonates, south-central Texas, USA. Journal of Hydrology, v. 54, p. 51-74.
L'Haridon, S., Miroshnichenko, M.L., Kostrikina, N.A., Tindall, B.J., Spring, S., Schumann, P., Stackebrandt, E., Bonch-Osmolovskaya, E.A., and Jeanthon, C., 2006. Vulcanibacillus modesticaldus gen. nov., sp. nov., a strictly anaerobic, nitrate- reducing bacterium from deep-sea hydrothermal vents. Int. J. Syst. Evol. Microbiol., v. 56, pp. 1047-1053.
Liu, S.V., Zhou, J., Zhang, C., Cole, D.R., Gajdarziska-Josifovska, M., and Phelps, T.J., 1997. Thermophilic Fe(III)-reducing bacteria from the deep subsurface: the evolutionary implications. Science, v. 277, pp. 1106-1109.
Longley, G., 1981. The Edwards Aquifer: Earth’s most diverse groundwater ecosystem? International Journal of Speleology, v. 11, pp. 123-128.
79 Lozupone, C., and Knight, R., 2005. UniFrac: a new phylogenetic method for comparing microbial communities. Appl. Environ. Microbiol., v. 71, pp. 8228-8235.
Lüttge, A., and Conrad, P.G., 2004. Direct observation of microbial inhibition of calcite dissolution. Appl. Environ. Microbiol., v. 70, no. 3, p. 1627-1632.
Macalady, J.L., Dattagupta, S., Schaperdoth, I., Jones, D.S., Druschel, G.K., and Eastman, D., 2008. Niche differentiation among sulfur-oxidizing bacterial populations in cave waters. The ISME Journal, advance online publication, doi:10.1038/ismej.2008.25.
MacInnis, I.N., and Brantley, S.L., 1992. The role of dislocations and surface morphology in calcite dissolution. Geochimica et Cosmochimica Acta, v. 56, pp. 1113-1126.
Maclay, R.W., and Small, T.A., 1986. Carbonate geology and hydrology of the Edwards Aquifer in the San Antonio area, Texas. Texas Water Development Board Report 296, 90 pp.
Madigan, M.T., Martinko, J.M., Dunlap, P.V., and Clark, D.P., 2006. Brock biology of microorganisms, 11th edition. Pearson Education, Inc., 1061 pp.
Mason, J., and Kelly, D.P., 1988. Thiosulfate oxidation by obligately heterotrophic bacteria. Microbial Ecology, v. 15, pp. 123-134.
Masuda, S., Eda, S., Ikeda, S., Mitsui, H., and Minamisawa, K., 2010. Thiosulfate-dependent chemolithoautotrophic growth of Bradyrhizobium japonicum. Appl. Environ. Microbiol., doi:10.1128/AEM.02783-09.
Miroshnichenko, M.L., Kostrikina, N.A., Chernyh, N.A., Pimenov, N.V., Tourova, T.P., Antipov, A.N., Spring, S., Stackebrandt, E., and Bonch-Osmolovskaya, E.A., 2003. Caldithrix abyssi gen. nov., sp. nov., a nitrate-reducing, thermophilic, anaerobic bacterium isolated from a Mid-Atlantic Ridge hydrothermal vent, represents a novel bacterial lineage. Int. J. Syst. Evol. Microbio., v. 53, pp. 323-329.
Moore, T.S., Murray, R.W., Kurtz, A.C., and Schrag, D.P., 2004. Anaerobic methane oxidation and the formation of dolomite. Earth and Planetary Science Letters, v. 229, pp. 141– 154.
Mylroie, J.E., and Balcerzak, W.J., 1992. Interaction of microbiology and karst processes in Quaternary carbonate island aquifers. In: Standford, J.A., and Simons, J.J. (eds.), Proceedings of the First International Conference on Ground Water Ecology: Tampa, Florida, pp. 37-46.
Mylroie, J.E., and Carew, J.L., 1995. Karst development on carbonate islands. In: Budd, D.A., Saller, A.H., and Harris, P.M. (eds.), Unconformities in carbonate strata – their
80 recognition and the significance of associated porosity. AAPG Memoir v. 63, pp. 55- 76.
Neron, B., Menager, H., Maufrais, C, Joly, N., Maupetit, J., Letort, S., Carrere, S., Tuffery, P, and Letondal, C., 2009. Mobyle: a new full web bioinformatics framework. Bioinformatics, v. 25, pp. 3005-3011.
Newman, D.K, and Banfield, J.F., 2002. Geomicrobiology: how molecular-scale interactions underpin biogeochemical systems. Science, v. 296, pp. 1071-1077.
Northup, D.E. and Lavoie, K.H., 2001. Geomicrobiology of caves: a review. Geomicrobiology Journal, v. 18, p. 199-222.
Oetting, G.C., Banner, J.L., and Sharp, J.M. Jr., 1996. Regional controls on the geochemical evolution of saline groundwaters in the Edwards aquifer, central Texas. Journal of Hydrology, v. 181, pp. 251-283.
Pace, N.R., 1997. A molecular view of microbial diversity and the biosphere. Science, v. 276, pp. 734-740.
Pasic, L., Kovce, B., Sket, B., and Herzog-Velikonja, B., 2010. Diversity of microbial communities colonizing the walls of a karstic cave in Slovenia. FEMS Microbiology Ecology, v. 71, pp. 50-60.
Palmer, A.N., 1991. Origin and morphology of limestone caves. Geological Society of America Bulletin, v. 103, pp. 1-21.
Pang, C.M., and Liu, W., 2007. Community structure analysis of reverse osmosis membrane biofilms and the significance of Rhizobiales bacteria in biofouling. Environ. Sci. Technol., v. 41, pp. 4728-4734.
Parkhurst, D.L., and Appelo, C.A.J., 1999. User’s guide to PHREEQC (version 2) – a computer grogram for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. U.S.G.S. Water Resour. Inv. Rep. 99-4259, 312 pp.
Pedersen, K., 2000. Exploration of deep intraterrestrial microbial life: current perspectives. FEMS Microbiology Letters, v. 185, pp. 9-16.
Pedersen, K., 2005. The deep intraterrestrial biosphere. In: Gadd, G.M., Semple, K.T., Lappin-Scott, H.M. (eds.), Micro-organisms and earth systems—advances in geomicrobiology. Cambridge University Press, Cambridge, U.K., pp. 233-245.
Petsch, S.T., Edwards, K.J., and Eglinton, T.I., 2003. Abundance, distribution and δ13C analysis of microbial phospholipid-derived fatty acids in a black shale weathering profile. Organic Geochemistry, v. 34, pp. 731–743.
81 Plummer, L.N., 1975. Mixing of seawater with calcium carbonate groundwater. In: Whitten, E.H.T. (ed.), Quantitative studies in geological sciences. Geological Society of America Memoir, v. 142, pp. 219-236.
Plummer, L.N., Busby, J.F., Lee, R.W., and Hanshaw, B.B., 1990. Geochemical modeling of the Madison aquifer in parts of Montana, Wyoming, and South Dakota. Water Resources Research, v. 26, pp. 1981-2014.
Poulson, T.L., and White, W.B., 1969. The cave environment. Science, v. 165, pp. 971-981.
Pronk, M.N., Goldscheider, N., and Zopfi, J., 2009. Microbial communities in karst groundwater and their potential use for biomonitoring. Hydrogeology Journal, v. 17, pp. 36-48.
Randall, K., 2006. Assessing the potential impact of microbes in the Edwards and Trinity aquifers of central Texas. MS Thesis, Louisiana State University, 108 pp.
Rasband, W., 1996. NIH Image 1.62 computer software: Maryland, National Institutes of Health.
Reardon, C.L., Cummings, D.E., Petzke, L.M., Kinsall, B.L., Watson, D.B., Peyton, B.M., and Geesey, G.G., 2004. Composition and diversity of microbial communities recovered from surrogate minerals incubated in an acidic uranium-contaminated aquifer. Applied and Environmental Microbiology, v. 70, pp. 6037-6046.
Reinhold-Hurek and Hurek, 2006. The genera Azoarcus, Azovibrio, Azospira, and Azonexus. Prokaryotes, v. 5, pp. 873-891.
Roberts, J.A., Bennett, P.C., Gonzalez, L.A., Macpherson, G.L., and Milliken, K.L., 2004. Microbial precipitation of dolomite in methanogenic groundwater. Geology, v. 32, pp. 277-280.
Rogers, J. R., and Bennett, P.C., 2004. Mineral stimulation of subsurface microorganisms: release of limiting nutrients from silicates. Chemical Geology, v. 203, pp. 91–108.
Roh, S.W., Quan, Z., Nam, Y., Chang, H., Kim, K., Kim, M., Im, W., Jin, L, Kim, S., Lee, S., and Bae, J., 2008. Pedobacter agri sp. nov., from soil. Int. J. Syst. Evol. Microbiol., v. 58, pp. 1640-1643.
Rye, R.O., Back, W., Hanshaw, B.B., Rightmire, C.T., and Pearson, F.J. Jr., 1981. The origin and isotopic composition of dissolved sulfide in groundwater from carbonate aquifers in Florida and Texas. Geochimica et Cosmochimica Acta, v. 45, pp. 1941-1950.
82 Saito, T., Ishii, S., Otsuka, S., Nishiyama, M., and Senoo, K., 2008. Identification of novel Betaproteobacteria in a succinate-assimilating population in denitrifying rice paddy soil by using stable isotope probing. Microbes Environ., v. 23, pp. 192-200.
Sarbu, S. M., Galdenzi, S., Menichetti, M., and Gentile, G., 2000. Geology and biology of Grotte di Frasassi (Frasassi Caves) in central Italy, an ecological multi-disciplinary study of a hypogenic underground karst system. In Wilkens, H., Culver, D.C., and Humphreys, W.S. (eds.), Ecosystems of the World, v. 30, pp. 361– 381.
Sarbu, S.M., Kane, T.C., and Kinkle, B.K., 1996. A chemoautotrophically based cave ecosystem: Science, v. 272, p. 1953-1955.
Schauer, R., Bienhold, C., Ramette, A., and Harder, J., 2010. Bacterial diversity and biogeography in deep-sea surface sediments of the South Atlantic Ocean. The ISME Journal, v. 4, pp. 159-170.
Schloss, P.D., and Handelsman, J., 2005. Introducing DOTUR, a computer program for defining operational taxonomic units and estimating species richness. Applied and Environmental Microbiology, v. 71(3), pp. 1501-1506.
Shabarova, T., and Pernthaler, J., 2009. Investigation of bacterioplankton communities in aquatic karst pools in Barenschacht Cave of Bernese Oberland. In: White, W.B. (ed.), Proceedings of the 15th International Congress of Speleology, 2009, International Union of Speleology, USA, v. 1, pp. 416-421.
Sharp, J.M. Jr., 1990. Stratigraphic, geomorphic, and structural controls of the Edwards Aquifer, Texas, U.S.A. In: Simpson, E.S., and Sharp, J.M. Jr. (ed.), Selected papers on hydrogeology. Hannover, Germany, International Association of Hydrogeologists, v. 1, pp. 67-82.
Sharp, J.M. Jr., and Banner, J.L., 1997. The Edwards Aquifer: A resource in conflict. GSA Today, v. 7, no. 8, pp. 1-9.
Sjoberg, E.L., and Rickard, D.T., 1984. Calcite dissolution kinetics: surface speciation and the origin of the variable pH dependence. Chemical Geology, v. 42, pp. 119-136.
Sorokin, D.Y., 2003. Oxidation of inorganic sulfur compounds by obligately organotrophic bacteria. Microbiology, v. 72, pp. 725-739.
Stevens, T.O., and McKinley, J.P., 1995. Lithoautotrophic microbial ecosystems in deep basalt aquifers. Science, v. 270, pp. 450-454.
Szewzyk, U., Szewzyk, R., and Stenstrom, T.A., 1994. Thermophilic, anaerobic bacteria isolated from a deep borehole in granite in Sweden. Proceedings of the National Academy of Sciences, v. 91, pp. 1810-1813.
83 Vandevivere, P., and Bevaye, P., 1992. Sampling method for the observation of microorganisms in unconsolidated porous media via scanning electron microscopy. Soil Science, v. 153, pp. 482-485.
Viles, H.A., 1988. Organisms and karst geomorphology. In: Viles, H.A. (ed.), Biogeomorphology: New York, Basin Blackwell, p. 319-350.
Warren, J., 2000. Dolomite: occurrence, evolution and economically important associations. Earth Science Reviews, v. 52, pp. 1-81.
Warthmann, R., van Lith, Y., Vasconcelos, C., McKenzie, J.A., and Karpoff, A.M., 2000. Bacterially induced dolomite precipitation in anoxic culture experiments. Geology, v. 28, pp. 1091-1094.
White, W.B., 1988. Geomorphology and hydrology of karst terrains. New York, Oxford University Press, p. 464.
White, W.B., 1997. Thermodynamic equilibrium, kinetics, activation barriers, and reaction mechanisms for chemical reactions in karst terrains. Environmental Geology, v. 30, pp. 46-58.
White, W.B., 2002. Karst hydrology: recent developments and open questions. Engineering Geology, v. 65, pp. 85-105.
Whitman, W.B., Coleman, D.C., and Wiebe, W.J., 1998. Prokaryotes: The unseen majority. Proceedings of the National Academy of Sciences, v. 95, pp. 6578-6583.
Winderl, C., Anneser, B., Griebler, C., Meckenstock, R.U., and Lueders, T., 2008. Depth- resolved quantification of anaerobic toluene degraders and aquifer microbial community patterns in distinct redox zones of a tar oil contaminant plume. Applied and Environmental Microbiology, v. 74, pp. 792-801.
Wigley, T.M.L., and Plummer, L.N., 1976. Mixing of carbonate waters. Geochimica et Cosmochimica Acta, v. 40, pp. 989-995.
Wolicka, D., and Borkowski, A. 2007. The geomicrobiological role of sulphate-reducing bacteria in environments contaminated by petroleum products. Geomicrobiology Journal, v. 24, pp. 599-607.
84 APPENDIX A
OPERATIONAL TAXONOMIC UNITS (OTUs)
85 TABLE OF CONTENTS FOR APPENDIX A
Alphaproteobacteria………………………………………………………………………….87
Gammaproteobacteria………………………………………………………………………..88
Betaproteobacteria…………………………………………………………………………....89
Deltaproteobacteria…………………………………………………………………………..90
Bacteroidetes………………………………………………………………………………….90
Nitrospirae……………………………………………………………………………………91
Firmicutes…………………………………………………………………………………….91
Planctomycetes……………………………………………………………………………….91
Chloroflexi……………………………………………………………………………………91
Caldithrix……………………………………………………………………………………..91
Unclassified Bacteria……………………………………………………………………...... 91
86 Appendix A. Operational Taxonomic Units (OTUs) of all clone libraries defined at 98% sequence identity. Occurrence of clones in OTU per site Major Taxonomic Representative Accession % Girl Scout Girl Scout LCRA LCRA Paradise Alley Closest Relative Class or Phylum Clone # Similarity Deep Shallow Deep Shallow Shallow Uncultured alpha proteobacterium Alphaproteobacteria EDW07B001_84 EU038058 99 4 0 0 0 14 clone LKC_Acid_196 EDW07B006_69 Rhizobium sp. XJ-L72 EU817493 98 1 0 0 0 0 EDW07B005_130 Agrobacterium sp. M11 GU086439 99 5 3 0 2 0 EDW07B003_49 Rhizobium huautlense AM237359 100 0 0 0 1 0 EDW07B003_68 Mesorhizobium ciceri strain C-2/2 AY206686 96 0 0 0 1 0 EDW07B003_109 Uncultured Xanthobacter sp. FJ572674 99 0 0 0 2 0 EDW07b003_53 Bosea massiliensis strain 63287 NR025118 99 0 0 0 1 0 EDW07B001_32 Dechlorospirillum sp. DB AY530551 97 0 0 0 0 1 Uncultured alpha proteobacterium EDW07B001_85 FJ535117 99 0 2 1 0 3 clone ATB-LH-6119 Uncultured alpha proteobacterium EDW07B003_91 DQ463742 97 0 0 0 1 0 clone TK-NH3 Uncultured alpha proteobacterium EDW07B003_114 DQ463742 99 0 0 0 1 0 clone TK-NH3 EDW07B003_40 Sphingomonas sp. BBCT69 DQ337553 98 0 0 0 1 0 EDW07B003_29 Novosphingobium aromaticivorans AB025012 97 0 0 0 2 0 EDW07B001_118 Sphingomonas sp. ZnH-1 EF061133 99 0 2 0 0 1 EDW07B001_92 Sphingomonas sp. DhA-95 AF177917 99 0 0 0 0 1 Sphingomonadaceae bacterium EDW07B001_55 EF066484 99 0 0 0 0 2 CBFR-1 EDW07B002_20 Sphingomonas sp. BBCT20 DQ337548 96 0 0 1 0 0 EDW07B003_75 Alpha proteobacterium KC-IT-W5 FJ711211 99 0 0 0 1 0 Candidatus Reyranella massiliensis EDW07B006_43 EF394922 95 1 0 0 0 0 strain URTM1 EDW07B001_58 Brevundimonas vesicularis FM955876 100 6 8 3 1 7 Brevundimonas kwangchunensis EDW07B001_8 AY971369 98 0 0 0 2 2 strain KSL-110 EDW07B001_26 Phenylobacterium falsum AJ717391 98 0 0 0 0 1 EDW07B002_38 Uncultured Caulobacter sp. AM936876 99 0 0 1 0 0 EDW07B003_20 Caulobacter crescentus NA1000 CP001340 99 2 5 1 2 0 Uncultured alpha proteobacterium EDW07B001_1 GQ984314 94 0 0 0 0 1 clone DBS1a18 Uncultured alpha proteobacterium EDW07B001_5 EU038058 99 3 0 0 0 6 clone LKC_Acid_196 Uncultured alpha proteobacterium EDW07B006_119 EU038058 99 1 0 0 0 0 clone LKC_Acid_196 EDW07B006_39 Rhizobium sp. W3 EU781656 99 1 0 0 0 0 EDW07B005_145 Agrobacterium larrymoorei EU373312 99 0 1 0 0 0
87 Major Taxonomic Representative Accession % Girl Scout Girl Scout LCRA LCRA Paradise Alley Closest Relative Class or Phylum Clone # Similarity Deep Shallow Deep Shallow Shallow EDW07B003_90 Uncultured Xanthobacter sp. FJ572674 99 0 0 0 1 0 EDW07B001_11 Novosphingobium subterraneum AB025014 97 0 0 0 0 5 EDW07B005_105 Novosphingobium aromaticivorans AB025012 99 0 1 0 1 0 Uncultured alpha proteobacterium EDW07B003_65 FJ535117 99 1 2 0 1 0 clone ATB-LH-6119 EDW07B001_109 Sphingomonas sp. PA219 AM900782 96 0 0 0 0 1 Sphingomonadaceae bacterium EDW07B001_80 EF066484 99 0 0 0 0 1 CBFR-1 EDW07B001_9 Brevundimonas vesicularis FM955876 99 2 3 4 4 10 EDW07B003_18 Brevundimonas lenta strain DS-18 EF363713 99 0 0 0 1 0 Brevundimonas kwangchunensis EDW07B005_176 AY971369 97 0 1 0 0 0 strain KSL-110 EDW07B003_85 Caulobacter crescentus NA1000 CP001340 99 0 4 0 1 0 EDW07B002_87 Alpha proteobacterium A0902 AF236003 95 0 0 1 0 0 Pseudomonas fluorescens strain Ps 7- Gammaproteobacteria EDW07B003_104 EU854430 99 19 3 0 5 0 12 EDW07B001_128 Pseudomonas poae DQ536513 99 2 9 0 5 13 Uncultured gamma proteobacterium EDW07B005_125 EF612432 91 0 1 0 0 0 clone Fe-K6-C35 EDW07B003_26 Pseudomonas poae strain BIHB 730 DQ536513 98 0 0 0 1 0 EDW07B003_28 Pseudomonas migulae strain CT14 EU111725 98 1 0 1 1 1 EDW07B005_113 Pseudomonas sp. 130(2zx) AM409194 99 1 1 0 0 0
EDW07B001_86 Pseudomonas sp. AEBL3 AY247063 99 2 0 0 21 4 EDW07B003_4 Pseudomonas sp. ACP14 AY464463 99 0 1 0 1 0 Pseudomonas argentinensis strain EDW07B003_77 EU723817 99 6 3 0 1 0 HDDMG01 EDW07B005_129 Pseudomonas fluorescens strain Mc07 EF672049 98 1 1 0 0 0 EDW07B001_87 Pseudomonas sp. HhSoUsc AY089990 99 0 0 0 11 2 Uncultured Pseudomonas sp. clone 1- EDW07B006_73 EU305565 99 1 0 0 0 0 C Acinetobacter calcoaceticus strain EDW07B005_81 AY346313 99 2 3 0 0 0 HPC253 EDW07B002_102 Pseudomonas sp. SPS-2 AM689946 96 0 0 1 0 0 EDW07B002_98 Uncultured Gamma proteobacterium AM935272 97 0 0 1 0 0 EDW07B005_142 Pseudomonas rhodesiae strain NO5 FJ462694 99 10 6 0 2 1 EDW07B003_56 Pseudomonas sp. AEBL3 AY247063 99 0 1 0 2 1 Uncultured Pseudomonas sp. clone EDW07B006_95 DQ279339 95 1 0 0 0 0 TM14_39
88 Major Taxonomic Representative Accession % Girl Scout Girl Scout LCRA LCRA Paradise Alley Closest Relative Class or Phylum Clone # Similarity Deep Shallow Deep Shallow Shallow Pseudomonas argentinensis strain EDW07B005_157 EU723817 99 2 1 0 0 0 HDDMG01 EDW07B006_62 Pseudomonas sp. 28R14a AF456224 97 1 0 0 0 0 Acinetobacter calcoaceticus strain EDW07B006_29 AY346313 97 1 0 0 0 0 HPC253 EDW07B001_101 Stenotrophomonas sp. An27 AJ551165 99 0 0 0 0 1 uncultured Hydrogenophilaceae Betaproteobacteria EDW07B003_5 EU266782 93 0 0 0 2 0 bacterium Uncultured beta proteobacterium EDW07B002_72 EU700151 99 0 0 7 0 0 clone STU9 Uncultured beta proteobacterium EDW07B002_57 EU700151 96 0 0 1 0 0 clone STU9 Uncultured beta proteobacterium EDW07B002_30 EF467590 98 0 2 29 0 0 clone lka50b EDW07B001_4 Janthinobacterium sp. An8 AJ551147 99 0 1 2 4 6 Uncultured Massilia sp. clone GI8-sp- EDW07B002_79 GQ129883 99 0 1 1 0 0 C19 Uncultured Oxalobacteraceae EDW07B005_101 EU362132 98 1 4 0 0 0 bacterium clone 3 EDW07B005_74 Massilia timonae AJ871463 98 0 1 0 0 0 EDW07B002_61 Uncultured beta proteobacterium AF204252 95 0 0 3 0 0
EDW07B005_18 Georgfuchsia toluolica strain G5G6 EF219370 94 0 3 0 0 0 Uncultured Azospira sp. clone MFC- EDW07B002_120 FJ393125 93 0 0 1 0 0 B162-F12 EDW07B001_42 Beta proteobacterium CDB21 AB194096 99 4 4 0 2 8 EDW07B003_101 Beta proteobacterium CDB21 AB194096 99 0 0 0 2 0 Burkholderia sordidicola isolate EDW07B005_122 DQ256491 99 0 1 0 0 0 Jm120 Uncultured Hydrogenophilaceae EDW07B001_18 EU266809 97 0 0 0 0 1 bacterium Uncultured Hydrogenophilaceae EDW07B001_70 EU266809 99 0 0 0 0 11 bacterium clone D10_45 Uncultured beta proteobacterium EDW07B005_43 FJ517027 96 0 2 0 0 0 clone TDNP_Wbc97_166_1_58 Uncultured beta proteobacterium EDW07B002_118 FJ517027 95 0 0 1 0 0 clone TDNP_Wbc97_166_1_58 EDW07B002_41 Comamonas sp. L11 EF426453 99 0 3 1 0 0 EDW07B006_13 Delftia sp. LFJ11-1 DQ140182 98 1 0 0 0 0 EDW07B006_108 Acidovorax sp. KSP2 AB076843 98 1 0 0 0 0 EDW07B005_144 Uncultured bacterium SJA-62 AJ009470 99 0 1 0 0 0 EDW07B001_68 Thiobacillus Q AJ289884 97 0 0 0 0 5
89 Major Taxonomic Representative Accession % Girl Scout Girl Scout LCRA LCRA Paradise Alley Closest Relative Class or Phylum Clone # Similarity Deep Shallow Deep Shallow Shallow Uncultured Gallionellaceae bacterium EDW07B002_116 EU266836 94 0 0 1 0 0 clone D12_32 Uncultured Azospira sp. clone MFC- EDW07B002_52 FJ393134 99 0 0 5 11 0 B162-G10 EDW07B001_98 Beta proteobacterium IMCC1716 DQ664239 95 0 0 0 0 1 Thiobacillus denitrificans ATCC EDW07B001_10 CP000116 99 0 0 0 0 2 25259 EDW07B001_37 Beta proteobacterium CDB21 AB194096 99 1 2 0 0 2 EDW07B003_88 Uncultured Azospira sp. FJ393134 98 0 0 0 1 0 EDW07B005_165 Beta proteobacterium G5G6 EF219370 96 0 1 0 0 0 Uncultured Oxalobacteraceae EDW07B006_42 EU362132 98 1 0 0 0 0 bacterium clone 3 EDW07B006_127 Janthinobacterium sp. Lc30-2 GU244361 98 1 0 0 0 0 EDW07B005_93 Comamonas sp. L11 EF426453 99 0 2 0 0 0 Uncultured delta proteobacterium Deltaproteobacteria EDW07B001_39 AM746084 92 0 0 0 0 1 clone GoM140_Bac40 EDW07B003_98 Uncultured delta proteobacterium FJ485074 97 0 0 0 1 0 Uncultured Desulfobacteraceae EDW07B001_121 FJ517129 96 0 0 0 0 1 bacterium clone Uncultured delta proteobacterium EDW07B005_14 AY792293 90 0 1 0 0 0 clone EDW07B001_47 Uncultured Smithella sp. EU888819 93 0 0 0 0 1 EDW07B005_72 Uncultured delta proteobacterium FJ205327 84 0 1 0 0 0 Uncultured delta proteobacterium EDW07B001_28 FJ484425 87 0 0 0 0 2 clone Z17M21B Uncultured delta proteobacterium EDW07B005_97 DQ811832 80 0 1 0 0 0 clone MSB-5D3 Bacteroidetes EDW07B001_46 Pedobacter sp. VA-17 DQ984208 97 0 0 0 0 1 EDW07B003_51 Pedobacter terrae strain DS-57 DQ889723 99 0 0 0 1 0 EDW07B002_60 Pedobacter agri strain PB92 EF660751 100 0 0 3 0 0 Sphingobacteriaceae bacterium KVD- EDW07B001_88 DQ490464 97 0 0 0 0 1 1969-11 EDW07B002_75 Flavobacterium sp. ARSA-45 GU295966 98 0 0 7 0 0 EDW07B002_34 Cytophaga sp. SA1 AF414444 95 0 0 2 0 0 Uncultured Bacteroidetes bacterium EDW07B003_39 AF529321 98 1 0 0 2 0 clone CLi112 Uncultured Bacteroidetes bacterium EDW07B002_47 EF562555 99 0 0 1 0 0 clone CC_04 Uncultured Bacteroidetes bacterium EDW07B002_39 AY887013 98 0 0 1 0 0 clone 1.9.7.2 EDW07B002_103 Flavobacterium sp. THWCSN34 AM888191 96 0 0 1 0 0 EDW07B002_94 Pedobacter agri strain PB92 EF660751 99 0 0 1 0 0
90 Major Taxonomic Representative Accession % Girl Scout Girl Scout LCRA LCRA Paradise Alley Closest Relative Class or Phylum Clone # Similarity Deep Shallow Deep Shallow Shallow EDW07B001_110 uncultured Bacteroidetes/Chlorobi EU266920 91 0 0 1 0 1 EDW07B002_36 uncultured Bacteroidetes bacterium FN429794 89 0 0 1 0 0 uncultured Bacteroidetes/Chlorobi EDW07B005_121 EF562135 95 0 1 0 0 0 bacterium EDW07B005_23 Uncultured Cytophaga sp. AB189358 87 0 1 0 0 0 Uncultured Nitrospiraceae bacterium Nitrospirae EDW07B005_118 EU266868 98 0 3 0 0 1 clone D15_30 Uncultured Nitrospirae bacterium EDW07B005_73 FJ484454 93 0 1 0 0 0 clone Z17M50B EDW07B002_112 uncultured Nitrospirae bacterium FJ516996 95 0 0 3 0 0 EDW07B003_16 uncultured Nitrospirae bacterium FJ484418 94 0 0 0 1 0 EDW07B005_16 uncultured Nitrospirae bacterium FJ484418 94 0 1 0 0 0 Uncultured Nitrospirae bacterium EDW07B006_84 FJ484503 97 1 0 0 0 0 clone Z53M16B Uncultured Nitrospirae bacterium EDW07B002_121 AJ534688 87 0 0 1 0 0 partial Uncultured Nitrospirae bacterium EDW07B005_127 FJ516912 95 0 1 0 0 0 clone TDNP_USbc97_185_1_45 EDW07B005_79 Uncultured Nitrospira sp. clone I11F FJ205367 94 0 1 0 0 0 Uncultured low G+C Gram-positive Firmicutes EDW07B001_27 DQ234641 95 0 0 3 0 1 bacterium EDW07B002_19 Vulcanibacillus modesticaldus AM050346 95 0 0 1 0 0 EDW07B002_44 Vulcanibacillus modesticaldus AM050346 95 0 0 1 0 0 Uncultured Firmicutes clone MSB- EDW07B005_49 DQ811928 86 0 1 0 0 0 5C10 Uncultured low G+C Gram-positive EDW07B001_141 DQ234641 95 0 0 0 0 1 bacterium Planctomycetes EDW07B005_7 Uncultured Planctomycete DQ837238 89 0 1 0 0 0 Uncultured Planctomycete clone EDW07B005_24 DQ837238 97 0 1 0 0 0 49S1_2B_12 Uncultured Planctomycete clone EDW07B002_12 FJ484765 91 0 0 1 0 0 Z195MB5 Uncultured Planctomycetales EDW07B002_28 FJ517014 90 0 0 1 0 0 bacterium Uncultured Chloroflexi bacterium Chloroflexi EDW07B005_86 EU266859 94 0 1 0 0 0 clone D15_19 EDW07B003_87 Uncultured Chloroflexi bacterium DQ463718 88 0 0 0 1 0 Caldithrix EDW07B001_41 Uncultured Caldithrix sp. FJ901586 83 0 0 0 0 2 Uncultured candidate division SPAM Unclassified Bacteria EDW07B002_65 AY921949 92 0 0 1 0 0 bacterium clone AKYG1047 Uncultured candidate division OD1 EDW07B005_131 DQ676467 86 0 1 0 0 0 bacterium clone Uncultured candidate division TM7 EDW07B005_52 GQ287613 92 0 1 0 0 0 bacterium clone
91 APPENDIX B
SEM IMAGES
92 TABLE OF CONTENTS FOR APPENDIX B
Girl Scout Deep Chip Number 1……………………………………………………….………………………………………94 13………………………………………………….…………………………………………..95 25……………………………………………………………………………………………...96 37……………………………………………………………………………………………...97
Girl Scout Shallow Chip Number 4…………………………………………………………………………………………….....98 14……………………………………………………………………………………………...99 28…………………………………………………………………………………………….100 38…………………………………………………………………………………………….101
LCRA Deep Chip Number 6……………………………………………………………………………………………...102 15…………………………………………………………………………………………….103 29…………………………………………………………………………………………….104 39…………………………………………………………………………………………….105
LCRA Shallow Chip Number 7……………………………………………………………………………………………...106 16…………………………………………………………………………………………….107 32…………………………………………………………………………………………….108 40…………………………………………………………………………………………….109
Paradise Alley Deep Chip Number 10…………………………………………………………………………………………….110 18…………………………………………………………………………………………….111 46…………………………………………………………………………………………….112 42…………………………………………………………………………………………….113
Paradise Alley Shallow Chip Number 11…………………………………………………………………………………………….114 20…………………………………………………………………………………………….115 47…………………………………………………………………………………………….116 44…………………………………………………………………………………………….117
93
Chip Number 1 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 155
Well Girl Scout deep Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Iron-sulfide precipitates, filamentous cells present.
94
Chip Number 13 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 7
Well Girl Scout deep Colonies present no
Date emplaced 9/5/2009 Dissolution features no
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Precipitates of calcite.
95
Chip Number 25 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 151
Well Girl Scout deep Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Filamentous cells, precipitates of dolomite present.
96
Chip Number 37 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 83
Well Girl Scout deep Colonies present no
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Precipitates of calcite.
97
Chip Number 4 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 104
Well Girl Scout shallow Colonies present no
Date emplaced 9/5/2009 Dissolution features no
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Possible biofilm material observed. Long filamentous cells present.
98
Chip Number 14 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 18
Well Girl Scout shallow Colonies present no
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Broad areas of dissolution without microbes present.
99
Chip Number 28 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 339
Well Girl Scout shallow Colonies present no
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Precipitates of iron sulfide and silicate-bearing clay-like mineral.
100
Chip Number 38 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 250
Well Girl Scout shallow Colonies present yes
Date emplaced 9/5/2009 Dissolution features no
Date retrieved 10/1/2009 Secondary precipitates yes
Notes: Silicate-bearing precipitate. Biofilm material, attachment structures observed.
101 Chip Number 6 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 304
Well LCRA deep Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Cells often clustered, biofilm material observed, precipitates of calcite, “corkscrew” morphology present.
102 Chip Number 15 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 172
Well LCRA deep Colonies present no
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Precipitates of calcite, “corkscrew” morphology present.
103
Chip Number 29 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 505
Well LCRA deep Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Cells often clustered together, radiating pattern observed. Iron sulfide precipitates, filamentous cells, “corkscrew” morphology present.
104
Chip Number 39 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 282
Well LCRA deep Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: “Corkscrew” morphology present. Apparent micro-scale “rips” in biofilm-like material observed.
105
Chip Number 7 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 238
Well LCRA shallow Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Calcite precipitates. Areas of dissolution observed with and without microbes nearby.
106
Chip Number 16 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 196
Well LCRA shallow Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Possible iron-sulfide precipitates.
107
Chip Number 32 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 308
Well LCRA shallow Colonies present yes
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Precipitates silicate-bearing clay-like mineral. Cells found in clusters and chains. Possible biofilm/attachment structures observed.
108
Chip Number 40 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 61
Well LCRA shallow Colonies present no
Date emplaced 9/5/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Precipitates of dolomite. Very few zones of dissolution observed.
109
Chip Number 10 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 167
Well Paradise Alley deep Colonies present yes
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Precipitates of calcite.
110
Chip Number 18 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 19
Well Paradise Alley deep Colonies present no
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Areas of dissolution with and without microbes present; precipitates of calcite.
111
Chip Number 46 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 323
Well Paradise Alley deep Colonies present yes
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Possible biofilm material observed.
112
Chip Number 42 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 4
Well Paradise Alley deep Colonies present no
Date emplaced 9/6/2009 Dissolution features no
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Large precipitates of dolomite observed.
113
Chip Number 11 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Reactive Number cells/unit area 244
Well Paradise Alley shallow Colonies present yes
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Precipitates of calcite. Cells often aligned along cleavage planes.
114
Chip Number 20 Preservation Method HMDS
Mineral Calcite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 74
Well Paradise Alley shallow Colonies present no
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Filamentous cells present. Silicate-bearing clay-like mineral observed.
115
Chip Number 47 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Reactive Number cells/unit area 297
Well Paradise Alley shallow Colonies present yes
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Possible biofilm material observed. Iron sulfide precipitate, (long) filamentous cells present.
116
Chip Number 44 Preservation Method HMDS
Mineral Dolomite Microbes Present yes
Microcosm type Non-reactive Number cells/unit area 111
Well Paradise Alley shallow Colonies present no
Date emplaced 9/6/2009 Dissolution features yes
Date retrieved 10/2/2009 Secondary precipitates yes
Notes: Filamentous cells present. Precipitates of dolomite.
117 VITA
Cassie Jo Gray was born in 1985 to Jeannie and James Gray in Sullivan, Indiana. She grew up amidst the corn fields in the rural farming town of Coalmont, Indiana, where she graduated from Shakamak High School as valedictorian. She continued her education at Indiana State
University (ISU) in Terre Haute, Indiana, to obtain a bachelor’s degree in geology. While attending ISU, Cassie worked as an undergraduate researcher in Dr. Anthony Rathburn’s oceanography laboratory, investigating the foraminiferal ecology of the Venice Lagoon. She also participated in internships with NASA’s Jet Propulsion Laboratory and the National
Science Foundation Research Experiences for Undergraduates (REU), acted as a field assistant to Scripps Institution of Oceanography in the Gulf of California, and presented research at several Geological Society of America conferences. She graduated summa cum laude from ISU, having obtained many awards and honors including McNair Scholar, ISU
Lilly Endowment Undergraduate Fellow, Outstanding Graduating Senior in Geology, and
Outstanding Undergraduate Research in Geology. Cassie began the pursuit for a master’s degree at Louisiana State University (LSU) in 2008, where she started as a teaching assistant and later became a research assistant. She applied for and received two grants to help fund the thesis research project, from the American Association of Petroleum Geologists (AAPG) and the Karst Waters Institute. She presented her thesis research at the American Geophysical
Union’s Fall Meeting in 2009 in San Francisco, California, and defended her thesis at LSU on
March 29, 2010.
118