Biological and Anthropogenic Markers in Taylor Valley...

Biological and Anthropogenic Markers in Taylor Valley, Antarctica

BY

CAROLINE MARIE B. JARAULA B.S., University of the Philippines M.S., University of the Philippines

THESIS Submitted as partial fulfillment of the requirements for the degree of Doctor of Philosophy in Earth and Environmental Sciences in the Graduate College of the

University of Illinois at Chicago, 2008

Chicago, Illinois

UMI Number: 3345646

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy

submitted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleed-through, substandard margins, and improper

alignment can adversely affect reproduction.

In the unlikely event that the author did not send a complete manuscript

and there are missing pages, these will be noted. Also, if unauthorized

copyright material had to be removed, a note will indicate the deletion.

®

UMI UMI Microform 3345646

Copyright 2009 by ProQuest LLC.

All rights reserved. This microform edition is protected against

unauthorized copying under Title 17, United States Code.

ProQuest LLC 789 E. Eisenhower Parkway

PO Box 1346 Ann Arbor, Ml 48106-1346

BIOLOGICAL AND ANTHROPOGENIC MARKERS IN TAYLOR VALLEY, ANTARCTICA

Caroline Marie Brinas Jaraula Department of Earth and Environmental Sciences

University of Illinois at Chicago Chicago, Illinois (2008)

Dissertation Chairperson: Fabien Kenig, Ph.D.

Establishing proxies are important to fill the information gaps for environmental studies

in Taylor Valley, Antarctica, where the area is sensitive to climate change and instrumental

records begun only a few decades ago. This study uses molecular markers, which are compounds

that preserve compositional and structural information so the sources of materials or organisms

can be traced. The compounds may also be a vital membrane component that distinguishes a

cellular function.

Temperature proxies based on the alkenone unsaturation indices and glycerol dialkyl

glycerol tetraethers (GDGTs) are tested if these are applicable to the study area. The alkenone

unsaturation indices and temperature estimates for Lake Fryxell were consistently low compared

to global records and show a good potential to use as a paleotemperature proxy. Mass spectral

analysis of alkenone moieties revealed the following compounds that have not been reported in

the literature: octatriaconta-pentaen-2-one (C38 5Me), nonatriaconta-pentaen-2 and 3-one

(C39;5Me and C395Et), as well as tetradec-pentaen-2-one (C^sMe).

Based on isoprenoidal and branched GDGTs, average temperatures for the water columns

of the lakes were within the error and deviations of the data, except for Lake Fryxell, where the

temperatures were overestimated. The GDGT distribution patterns in present-day deposits

classifies according to the different physico-chemical properties of each lake that govern oxic,

suboxic or anoxic conditions in the bottom waters. This classification was successful in

identifying the known paleoenvironments of the lakes.

Molecular markers were also used to trace the aviation diesel fuel and synthetic oils that

were spilled when a helicopter crashed on the Lake Fryxell ice cover. Naphthalenes and n-

alkanes of the fuel and C5 to C6 fatty acids of the synthetic oils were biodegraded in ice

associated with sediments. Volatile compounds from the diesel fuel were evaporated and water

washing was a significant process that fractionated the fuel composition. Potentially, only the

high molecular weight components of the synthetic oils PETs, the antioxidant tricresyl

phosphates are persistent. These compounds also tend to sorb onto sediments, and may comprise

the contaminants that eventually cross the ice cover and reach the pristine lake waters.

ACKNOWLEDGMENTS

I would like to thank my dissertation committee for their support and guidance: Fabien Kenig, Simon Brassell (Indiana University), Peter Doran, Kathy Nagy, and Neil Sturchio. Special thanks also to Kelvin Rodolfo for being a constant mentor through the years. I'm equally grateful for the collaboration of Stefan Schouten and Ellen Hopmans and for accommodating me to process samples at the Royal Netherlands Institute for Sea Research. John Priscu (Montana State University) and Kathy Welch (Ohio State University) provided invaluable input on the oil spill research; Rachael Morgan-Kiss (Miami University) readily supplied the DNA data for haptophytes; Yongsong Huang (Brown University) and William D'Andrea (Brown University) shared the DNA sequences from Greenland haptophytes; Dan McElheny (Department of Chemistry) conducted NMR analysis for the alkenone project.

Financial assistance were extended by a fellowship from the Institute of Environmental Science and Policy, award from the UIC Provost, LTER-MCM, NSF Antarctic Biology and Medicine SGER 0346316 (F. Kenig and P. Doran). Travel grants were provided by Institute for Environmental Science and Policy, Women in Science and Engineering, Earth and Environmental Sciences, Graduate College, Graduate Student Council and Gordon Research Conference Chair.

I would like to acknowledge the following for their dedication, patience, time and effort in going through the numerous laboratory steps: Apostolis Sambanis, Jeffrey Fitzgibbons, Timothy Chung, Marcus Muccianti, and Alice Hillegass. Gregory Ventura, Todd Brown, Linnea Heraty and Abelardo Beloso taught me some of the tricks in the laboratory. Jennifer Lawson Knoepfle provided me with first hand information about the study site.

Special thanks to Mary Lou Schick, Minnie Jones, Marilyn Cartlidge and Maggie Jameson for always being there, ready to help in all the administrative aspects of graduate school. D'arcy Meyer-Dombard, Carol Stein and Cynthia Jameson (Department of Chemistry) are sincerely acknowledged for their encouragement and professional insight.

I am ever thankful to my family and to my husband, Zenon, for their patience, inspiration, and support throughout my stay in graduate school.. .and to God, who has made everything possible.

CMBJ

IV

TABLE OF CONTENTS

CHAPTER PAGE

CHAPTER 1.

1.0 INTRODUCTION 1

2.0 LOCATION AND PHYSICAL ENVIRONMENT 5 2.1 Streams 6 2.2 Lakes 7

3.0 LIFE IN TAYLOR VALLEY 9 3.1 Adaptation to cold and dry environments 9 3.2 Endolithic communities 10 3.3 Soils 12 3.4 Streams 12 3.5 Lakes 13

4.0 RECENT ENVIRONMENTAL HISTORY 14

5.0 ORGANIC GEOCHEMISTRY OF TAYLOR VALLEY 15

6.0 ENVIRONMENTAL IMPACT 16

7.0 CITED LITERATURE 18

CHAPTER 2. SPME-GCMS study of the natural attenuation of aviation diesel spilled on the perennial cover of Lake Fryxell, Antarctica

Abstract 23

1.0 INTRODUCTION 24

2.0 MATERIALS AND METHODS 27 2.1 Study area and ice cover conditions 27 2.2 Samples 29 2.3 Core sampling 29 2.4 Extraction of volatiles 30 2.5 Gas chromatography-mass spectrometry 30 2.6 Fuel evaporation experiment 31

v

TABLE OF CONTENTS (continued)

CHAPTER PAGE

3.0 RESULTS „„„ 31 3.1JP-5/AN8 fuel mix 31 3.2 Ice cores from "uncontaminated" sites 32 3.3 Ice cores in contaminated sites 33 3.4 Meltpool waters and sediments 42 3.5 Fuel evaporation experiment 45

4.0 DISCUSSION 46 4.1 Evaporation 46 4.2 Biodegradation 46 4.3 Evaporation vs. biodegradation 47 4.4 Water washing 48 4.5 Lateral and vertical transport of contaminants 50

5.0 CONCLUSIONS 54

6.0 ACKNOWLEDGMENTS 55

7.0 REFERENCES CITED 56

8.0 APPENDIX II 60

CHAPTER 3. Composition and biodegradation of synthetic oils and

hydraulic fluid spilled on the perennial ice cover of Lake Fryxell, Antarctica

Abstract 73

1.0 INTRODUCTION 74

2.0 STUDY AREA 76

3.0 METHODS 77

4.0 RESULTS AND DISCUSSION 78 4.1 Synthetic oil Aeroshell 500 and 555 78 4.2 Hydraulic fluid MIL 5605 80 4.3 Degradation potential of spilled contaminants 80 4.4 GC-MS analysis of field samples 84 4.5 Lateral and vertical transports in the ice cover 88

5.0 CONCLUSIONS 89

VI


CHAPTER PAGE

ACKNOWLEDGMENT 90

REFERENCES CITED 90

CHAPTER 4 Tentative identification of pentaunsaturated alkenones from Lake Fryxell, East Antarctica

Abstract 95

1.0 INTRODUCTION 96

2.0 STUDY AREA, SAMPLE AND ANALYTICAL METHODS 100

3.0 RESULTS AND DISCUSSION 103 3.1 Alkenone identification 103 3.2 Lake Fryxell source organism(s) of alkenones 104 3.3 Alkenone distribution 107 3.4 Other biomarkers associated with alkenones 112 3.5 Factors relating to alkenone unsaturation 114 3.6 Climate proxy record 115

4.0 CONCLUSIONS 116

5.0 ACKNOWLEDGEMENTS 118


CHAPTER 5. Distribution and significance of archaeal and bacterial tetraether membrane lipids in lakes of Taylor Valley, Antarctica

Abstract 125

1.0 INTRODUCTION 127 1.1 Taylor Valley 128

1.1.1 Climate 130 1.1.2 Ice cover 130

vn


CHAPTER PAGE

1.1.3 Sources of water and materials 130 1.1.4 Lakes 132

1.2 Historical evolution of Taylor Valley 134 1.3 Archaea in Taylor Valley 135 1.4 Archaeal biomarkers 137

2.0 METHODS 141

3.0 RESULTS AND DISCUSSION 142 3.1 GDGT distribution 142

3.1.1 Cryptoendolith 142 3.1.2 Microbial mats 142 3.1.3 Lake bottom sediments 145 3.1.4 Core sediments 147

Lake Fryxell core sediments 147 Lake Hoare core sediments 150

3.2 Temperature estimate 151 3.2.1 LakeChad 151 3.2.2 West Lake Bonney 151 3.2.3 Lake Fryxell (paleo)temperature estimates 154 3.2.4 Lake Hoare (paleo)temperature estimates 155

3.3 Cluster analysis for present-day materials 155 3.3.1 Terrestrial Type 156 3.3.2 Aquatic non-stratified 156 3.3.3 Aquatic and stratified 158

3.4 Cluster analysis for present-day and ancient sediments 160

4.0 CONCLUSION 162

5.0 ACKNOWLEDGEMENTS 163


CHAPTER 6. Conclusions and Recommendations

1.0 UBIQUITY OF ARCHAEA 175

2.0 TERRESTRIAL BIOMARKERS IN ANTARCTIC SEDIMENTS 176

3.0 CRYPTOENDOLITHIC COMMUNITY 178

Vlll


CHAPTER PAGE

4.0 PENTAUNSATURATE ALKENONES 178

5.0 LAKE FRYXELL HAPTOPHYTES 179

6.0 APPLICABILITY OF TEMPERATURE PROXY RECORDS

IN COLD ENVIRONMENTS 180

7.0 ALKENONE UNSATURATION INDICES 181

8.0 TETRAETHER INDEX FOR 86 CARBONS 182

9.0 METHYLATION AND CYCLIZATION INDICES 182

10.DEVELOPINGBIOMARKER PROXIES 183

11 NATURAL ATTENUATION OF OIL SPILLS 185

12 DYNAMIC ICE COVERS 186

13 MONITORING AND REMEDIATION 187

14 REFERENCES CITED 188

VITA 189

ix

LIST OF FIGURES FIGURE PAGE

Chapter 1 1. Study area, McMurdo Dry Valleys 2 2. Taylor Valley Lakes 3 3. Granite endolithic community 11

Chapter 2 1. Helicopter crash site and locations of samples 26 2. SPME-GCMS of JP-5/AN8 diesel fuel mixture 34 3. Log of ice core 12 37 4. SPME-GCMS total ion current of aviation diesel and

ice core 12 samples 39 5. Summed mass chromatogram m/z 128+141+156 41 6. SPME-GCMS total ion current of aviation diesel and

meltpool waters and sediments 44 7. Relative strengths of evaporation vs. biodegradation 49 8. Dynamics in the Lake Fryxell ice cover that affect the

natural attenuation of contaminant fluids 53

Chapter 3 1. Total ion current trace of turbine oil Aeroshell 500 and

transmission oil Aeroshell 555 79 2. Total ion current trace of Hydraulic fluid MIL-5605 81 3. Partial total ion current of pentaerythritol trimesters 83 4. Total ion current trace of meltpool waters and sediments 85

Chapter 4 1. Locations of Lake Fryxell and Taylor Valley 98 2. Total ion current of the neutral fraction 101 3. Partial total ion current trace of F3, F4 andF5 102 4. Mass spectra data 104 5.Haptophytel8SrDNAtree 108

Chapter 5 1. Location map 129 2. Isoprenoidal glycerol dialkyl glycerol tetraethers (GDGT) 138 3. Branched GDGTs 139 4. Lake bottom sediment and a microbial mat

GDGT distribution 144 5. TEXg6 and BIT ratios for lake bottom sediments and

a microbial mat 146 6. GDGT distribution in core sediments and averages in

lake bottom sediments 149

x

LIST OF FIGURES (continued)

FIGURE PAGE

7. Cluster analysis of GDGT distribution and associated types of environment 157

8. Cluster analysis of lake bottom and core sediment GDGT distribution 161

XI

LIST OF TABLES

TABLE PAGE

Chapter 2 1. Percentage of fuel components 36

Chapter 4 1. Molecular ion and major fragment ions 105 2. Alkenone distribution from lakes, representative marine

environments and culture experiments 110 3. Sterols in Lake Fryxell 113 4. Temperature estimates for Lake Fryxell 117

Chapter 5 1. Physical and chemical descriptions of Taylor Valley lakes 131 2. Percentages of individual GDGTs relative to the total

GDGTs in modern samples 139 3. Percentages of individual GDGTs relative to the total

GDGTs in core samples 148 4. Temperature estimates from TEXg6 152 5. Temperature estimates from BIT 153

xn

APPENDIX

LIST OF APPENDICES

PAGE

Chapter 2 II-l Logoficecore 1 61 II-2 Log of ice core 2 62 II-3 Comparison of SPME-GCMS total ion current for

aviation diesel and ice core 2 samples 62 II-4 Log of ice core 3 63 II-5 Log of ice core 4 64 II-6 Comparison of SPME-GCMS total ion current for

aviation diesel and ice core 4 samples 64 II-7 Log of ice core 5 65 II-8 Comparison of SPME-GCMS total ion current for

aviation diesel and ice core 5 samples 65 Discussion of Ice core 6 "contaminated site" 66

II-9 Log of ice core 6 68 II-10 Comparison of SPME-GCMS total ion current for

aviation diesel and ice core 6 samples 69 11-11 Log of ice core 7 70 11-12 Logoficecore 11 70 II-13 Evaporation experiment trends for i-Cn/n-Cu

and Tnaph/Taik 71

Chapter 4 IV-1 Lake Fryxell bottom sediment log 170 IV-2 Lake Hoare bottom sediment log 171 IV-3 Lake Hoare core A sediment log 172 IV-4 Lake Hoare core B sediment log 173 IV-5 Lake Bonney bottom sediment log. 174

xiii

LIST OF ABBREVIATIONS

BHT

BIT

BLAST

CBT

DNAPL

GC-MS

GDGT

HMW

LGM

LMW

LNAPL

LTER

MAT

MBT

MCM

OUT

PAH

PAR

PET

SPME

TEXg6

Butylated Hydroxytoluene

Branched Index of Tetraether

Basic Logical Alignment Search Tool

Cyclization for Branched Tetraethers

Dense Non-aqueous Phase Liquid

Gas Chromatography-Mass Spectrometry

Glycerol Dialkyl Glycerol Tetraether

High Molecular Weight

Last Glacial Maximum

Low Molecular Weight

Light Non-aqueous Phase Liquid

Long-Term Ecological Modeling

Mean Air Temperature

Methylation for Branched Tetraethers

McMurdo Dry Valleys

Operational taxonomical units

Polyaromatic Hydrocarbons

Photosynthetically Active Radiation

Pentaerythritol Triesters

Solid Phase Micro-extraction

Tetraether Index for 86 Carbons

xiv

LIST OF ABBREVIATIONS (continued)

UCM Unresolved Complex Mixture

U^ Alkenone Unsaturation Index for 37 Carbons

xv

CHAPTER 1

1.0 Background of the Study

The environmental role of the cold, isolated Antarctic continent first drew the attention of

the public when the ozone hole caused by anthropogenic chlorofluorocarbon use was discovered

(Farman et al, 1985). The sensitivity of Antarctica to global warming was further recognized

after the collapse of the Larsen B ice shelf (MacAyeal et al , 2003). Extreme cold and dry

conditions in continental Antarctic preclude the growth of vascular plants as well as the presence

of vertebrates (Parker et al., 1982; Priscu, et al , 1999; Roberts et al., 2000) and Antarctica is

considered as the closest analog on Earth for conditions on Mars and Jupiter's moon Europa

(Priscu, 1998; McKay et al., 2005; Doran et al., 1998). The Antarctic ecosystem was also

considered a potential analogue to that of the Precambrian, prior to the advent of metazoan

grazers (Simmons et al., 1993) and to that of Earth in the middle to late Precambrian during

times of extensive glaciations known as "Snowball Earth" (Anderson, 1983; Kastings and Toon,

1988). Sublimation exceeds precipitation in some Antarctic regions that, consequently, lack ice

or snow cover. The largest such area is the McMurdo Dry Valleys (MCM) west of Ross Sea,

which includes Wright and Victoria valleys, in addition to Taylor Valley, the focus of this study

(Fig. 1). Numerous geochemical, meteorological, sedimentological, biological, and isotopic

studies have defined the watershed and lake-specific biogeochemical characteristics of the

McMurdo Dry Valleys and its lakes (Fig. 2). Environmental monitoring by the Long-term

Ecological Research in MCM, established since 1993, has provided significant records for

aquatic and terrestrial biogeochemistry and climatic data that are discussed in the next sections.

1

2

Ross Sea

McMurdo i Dry Valleys / Ross Ice

\ 180° I /Sheet 150° _->' r~| "~T"~~\ 210C

_ \ - J « * I

120 /

h \ so

90c i I .A-

.x

•A \ 60c \ :/-~.

30°

l/Yx \ \240°

s

LS_\ L 970° 70°

>

7 / 300° \ / /

330° 0°

Figure 1.McMurdo Dry Valleys, west of Ross Sea, is the largest ice-free area in Antarctica.

3

Taylor Valiey • Giaciers

Lakes

Streams

, •utettefttort Rhone •' Glacier Glad,

•"••,{X..i§r ,1 Creecerjf (\ I J-^-''Wains J S l rvi J f A

Figure 2. Taylor Valley lakes

4

With summer temperatures close to the melting point of water, slight climatic changes

can have significant implications on the hydrology and viability of life in the valley. Various

experiments were conducted and scenarios were modeled to study the effects of climate change,

however the record of climatic parameters and physico-chemical data only covers the last few

decades. Thus, proxy records need to be established so that climatic and paleoenvironmental

information preserved in the sediments, rocks, glaciers, and lake water can be used.

Geochemical, sedimentological and isotopic proxy records have been used and have refined the

history of the valleys and its lakes (Lyons et al., 1997; Lyons et al., 1998; Gooseff et al., 2006;

Denton et al., 1989; Hall and Denton, 2000)). This study attempts to establish proxy records

based on molecular markers

Molecular markers are lipid compounds which retain enough structural information to

trace anthropogenic sources and biosynthetic pathways specific to domains or vital membrane

components that distinguishes a cellular function. Organic geochemical techniques are used to

extract and identify these compounds. The use of molecular markers has not been systematically

applied to the Dry Valley environments. The biota of the Dry Valley lakes, being exclusively

microplanktic populations and microbial benthos (Wharton et al , 1993) afford a rare opportunity

to study an exclusively microbial ecosystem.

Increasing activity from scientific research and tourism makes the pristine Antarctic

environment more susceptible to pollution. Even without occupying the isolated continent,

atmospheric circulation transports contaminants inland. To study pollutants and

5

xenobiotic compounds in Taylor Valley, anthropogenic markers extracted from the environment

using organic geochemical techniques are used. These molecular markers can be utilized to trace

the contaminant source and transformations incurred from various environmental processes.

Four papers or chapters in this dissertation report biological and anthropogenic markers

in Taylor Valley. The fate of spilled aviation diesel and synthetic oils on the ice cover are

respectively discussed in Chapters 2 and 3. Temperature proxies using alkenone unsaturation

indices are evaluated and previously unrecognized pentaunsaturated alkenones are presented in

Chapter 4. The applicability of using glycerol dialkyl glycerol tetraethers (GDGT), another

established temperature proxy, is assessed in Chapters 5. The GDGT assemblages were also

classified based on calibrations with the varied physico-chemical conditions of the present-day

lakes. This classification was applied to older sediments to test if the known paleoenvironments

of the lakes can be resolved.

2.0 LOCATION AND PHYSICAL ENVIRONMENT

A persistent low pressure cell over the topographically low Ross embayment develops

cyclonic systems that bring moisture to the Dry Valleys (Morse et al., 1998). Annual

precipitation, mainly snow, is less than 10 cm of the water equivalent (Bromley, 1985), whereas

sublimation is about 30 cm. The modern annual average air temperature in Taylor Valley are -15

to -30 °C, and values in the summer hover around 0 °C for a few days (Chinn, 1993).

CI" contents of glaciers, streams and lakes decrease with increasing distance from the

coast, and CI" is conservative or does not precipitate in streams during the melt season, which

conveniently allows its use to estimate other major ion abundances (Lyons, et al., 1998).

6

Although smectites, illites, and vermicullites in the Dry Valleys (Campbell and Claridge, 1987),

high abundances of HC03" and H4Si04 normalized to Cl" in streams indicate high rates of

chemical weathering that is comparable to values for streams in humid to arid conditions in

lower latitudes (Lyons, et al., 1998). The effect of chemical weathering is more obvious in the

ion concentrations in the streams that drain into Lake Fryxell because these are the longest in the

Taylor Valley.

2.1 Streams

Stable Hydrogen and Oxygen isotopes were used to trace the sources of stream waters.

The Lake Bonney and Lake Hoare watershed are fed mainly by glacial melts; additionally,

streams in Lake Fryxell also receive snow melt (Goosseff et al., 2006). Glacial melt waters

sustain stream flow for about six to eight weeks from late November to early February (Conovitz

et al., 1998). Stream flow dynamics are controlled by climate, direct radiation of the Sun on the

vertical glacier face, geomorphology, and sediment texture (Conovitz et al., 1998; Fountain et

al., 1998). Alluvium also stores water within the hyporheic zone. Subsurface flow is limited to

the areas adjacent to the streams; groundwater flow is restricted (Chinn, 1993) by an

impermeable permafrost layer about 0.5 m below the surface (Conovitz et al., 1998).

2.2 Lakes

Taylor Valley consists of three major basins: Bonney, Chad-Hoare complex and Fryxell

(Chinn, 1993). A unique feature of the Dry Valley lakes is their perennial ice cover (Wharton et

al., 1993), although some lakes in the Alpine and high Arctic regions occasionally retain their ice

covers year round (Adams et al., 1998). Liquid water underneath the ice is maintained by the

7

supply of glacial melt waters when the temperatures rise to 0 °C for a few days in the summer

(Wilson and Wellman, 1962; Clow, 1988). A decrease in the range of temperatures can

completely freeze the lakes. The ice cover inhibits penetration of photosynthetically active 400-

700 nm radiation (PAR), restricts wind-generated mixing in the summer moats, retards the

exchange of gases between the water column and the atmosphere, decreases sedimentation in the

lake and traps aeolian sediments, which eventually affects PAR transmission (Wharton et al.,

1993). The ice cover insulates the liquid water and allows organisms to thrive, making this one

of the few oases to survive throughout the year (Morgan-Kiss et al , 2006). The ice thickness is a

balance between the 60 cm yr-1 sublimation on the surface and 30 cm yr"1 of freezing that adds

ice to the underside, generating an average upward movement 30 cm yr"1. This upward

movement of the ice was also observed from a microbial mat in the Lake Hoare ice that was

estimated to cross the ice in 5 to 10 years (Parker et al , 1982).

An updated bathymetry of Lake Bonney, Hoare and Fryxell by Doran et al. (1996)

enabled a better comparison of the morphologic control in the climate sensitivity of the lakes.

The liquid to frozen water ratio is lowest in Lake Fryxell. During a warm phase that completely

melts the ice covers, Lake Fryxell receives more meltwater than the volume of liquid water

already present, diluting its salt content approximately 3.5 times more than would occur in Lake

Bonney (Doran and Wharton, 1996). Moreover, the much larger ratio of surface area to volume

in Lake Fryxell also enhances evaporative loss, causing the lake levels to drop faster than in

Lake Bonney and Lake Hoare. Thus, the Lake Fryxell system is therefore more susceptible to

climate change (Doran and Wharton, 1996).

8

The lakes are closed hydrologic systems. Lacking outflows makes their hydrologic

balances sensitive to even slight perturbations in climate (Chinn, 1993). Lake Hoare records are

the longest and being in the middle of the valley will represent average conditions. A

compilation of lake ice thicknesses in Lake Hoare shows that ice thinned from 5.5 m in 1978 to

3.3 m in 1988 (Wharton, et al., 1993). The average rate of thinning from 1978 to 1988 was 22.5

cm yr"1, but recently slowed down. From 1986 to 2000, ice thickened by about 2 m (Doran et al.,

2002). In the past century, lake levels have been rising because inflows exceeded sublimation

and evaporation (Chinn, 1993). Recently, Doran et al. (2002) showed that lake levels dropped

about half a meter since 1986. Concurrent thinning of ice and increase in lake levels indicate

warming up to the mid 1980's, followed by cooling since then.

Over time, the increase in 8180 of the lake waters normalized to CI" concentration is a

more conservative parameter than depth (Lyone et al., 1998). For the surface waters of West

Lake Bonney and Lake Fryxell from 1972 to 2002 and Lake Hoare from 1990 to 2002, the

oxygen isotope data indicate an increase in the levels and subsequent freshening (Goosseff et al,

2006). This isotopic shift is a remnant of the 2001-2002 high flow that compensated about 85%

of the equivalent water volume that was evaporated or sublimated as the lake levels decreased in

the previous decades.

3.0 LIFE IN TAYLOR VALLEY

Dry and cold conditions in the valley prevent vascular plants and vertebrates to flourish

(Virginia and Wall, 1999). The area is known as one of the most extreme conditions on Earth,

but ecosystems flourish in rocks, sediments, streams, lakes and soils (Vincent, 1988; Priscu, et

9

al., 1998), although simpler in structure and function compared to temperate and tropical

counterparts. Availability of liquid water is the most limiting factor for life (Kennedy, 1993;

Wynn-Williams et al., 1997; Moorhead and Priscu, 1998). With temperature close to 0 °C in the

summer, physical factors, such as albedo and angle of incoming radiation, play important roles in

making the area habitable.

3.1 Adaptation to cold and dry environments

Maintaining membrane fluidity in cold temperatures is necessary for organisms and

photosynthesizers to function (Los and Murata, 2004; Routaboul et al., 2000). Psychrophiles and

psychrotrophs, cold-loving and cold tolerant organisms, respectively, adapt by incorporating

polyunsaturated, branched, short-chain or cyclic fatty acids in their cell membranes (White et al.,

2000).The addition of a cyclohexane ring and increase in the cyclic isoprenoidal moieties of the

tetraether membranes of aquatic Crenarchaeota are adaptation of (hyper)thermophilic ancestors

to cold temperatures (Sinninghe-Damste et al., 2002). In haptophyte microalgae, the increase in

alkenone unsaturation is also linked to cold temperatures (Marlowe, 1984; Brassell et al., 1986).

The organisms also cope physiologically by producing solutes. For example, endolithic

communities biosynthesize exopolysaccharide sheaths not only to conserve water, but also to aid

in bioweathering, cement and stabilize the soil (Wynn-Williams, 1993; Wynn-Williams et al.,

1997). Selection by endolithic communities of habitats, such as fissures or pore spaces in

translucent rocks, is a common adaptation to the dessicating conditions in the valley (Friedman,

1982). Expansion of mucilaginous sheaths in rock cracks and fissures to leach out inorganic

nutrients is habitat modification through bioweathering (Wynn-Williams et al , 1997).

10

3.2 Endolithic communities

Cyanobacteria and eukaryotic algae, together referred to as microalgae, dominantly

comprise an endolithic community that photosynthetically fixes carbon (Wynn-Williams et al.,

1997). Fungal symbionts and commensals produce organic acids as by-products that locally

weather the rock and release inorganic nutrients (Johnston and Vestal, 1993). Endoliths are

layered communities that usually consist of black-pigmented lichen, white lichen with hyaline

fungi, and green microalgae (Friedmann et al, 1988; Nienow and Friedmann, 1993) (Fig. 3).

Using Fourier Transform-Raman spectroscopy, Wynn-Williams et al. (1997) was able to

configure the endolithic communities including the ones from Beacon Sandstone from East

Beacon Mountain, Taylor Valley. The layers from top to bottom and their maximum thicknesses

are: 1 mm of abiotic iron-rich crust, 3 mm of black lichen zone, 5 mm of white fungal layer, 8

mm of green microalgae, and iron leachate. The crust, although lacking microbiota (Wynn-

Williams 1993b), is built at the top of the community to fill pore spaces and prevent dessication

in this zone (Wynn-Williams et al., 1997). The black lichen layer contains calcium oxalate

dihydrate and chlorophyll, whereas the fugal layer contains calcium oxalate monohydrate. The

layer contains traces of calcium oxalate monohydrate and iron oxide. Oxalic acids produced by

fungi weather the surrounding minerals, combine with Ca2+ and release Fe2+ that oxidize to form

a reddish color (Wynn-William et al., 1997). Microclimates within the endolithic profile is

reflected in the formation of dehydrate in the upper layer, where temperature is lower and closer

to the source of moisture. Conversely, monohydrate forms in drier and more acidic conditions

(Nienow et al., 1988a). The green algal layer is not only associated with chlorophyll, but also

carotenoid accessory pigments, needed to necessary to optimally absorb low light levels

(Edwards et al., 1995). Endolithic communities, including bacteria, fungi and yeast, constitute

11

White fungi Green

microalgae

Figure 3. A granite endolithic community from Lake Fryxell watershed. The granite was broken- up to reveal the endoliths.(photo by: F. Kenig)

12

the biota that are propagated by wind transport and serve as primary sources of soil inocula

(Wynn-Williams et al., 1997). These may even occur in glacier cryoconites (Foreman et al.,

2007), in high elevations of the Dry Valleys.

3.3 Soils

Nematodes are at the top of the food chain, Scottnema being the endemic taxa, and more

so in the driest of soils in the McMurdo Dry Valleys (Courtright, et a l , 1996; Virginia and Wall,

1999). The highest fungal densities in the Dry Valleys occur in soils beneath moss colonies

(Baublis et al., 1991). Fungi are mostly filamentous and are associated with yeasts.

3.4 Streams

Algal mats colonize stone pavements in the main stream channel, or the edge of the

parafluvial zone where the substrates are less stable (McKnight and Tae, 1996a; McKnight and

Tate 1996b; Alger et al., 1997). Nitrogen-fixing bacteria, tardigrades, nematodes and rotifers are

also reported (Allnut et al., 1981). Growth of these mats fix carbon and transforms available

nutrients ultimately influencing the nutrient, water and organic matter input to the lakes

(Simmons et al , 1993; Spaulding et al , 1994).

13

3.5 Lakes

Despite the lack of higher plants and animals, a complex assemblage of autotrophic,

heterotrophic and mixotrophic levels interact (Lisle and Priscu, 2004; Lyons et al., 1997; Priscu,

1998; Priscu et al, 1999). The assemblages are stratified along with the chemical gradients in the

water column. An excellent example is the stratification of different chryptophyte and

chlorophyte flagellate species in the water column of Lake Fryxell (Vincent, 1987; Spaulding et

al., 1994). Most of the metabolic activities are along the oxic-anoxic interface, called as the deep

chlorophyll maximum (Simmons et al, 1993). Different groups of autotrophs and photosynthetic

bacteria occupy the top of the anoxic zone and the oxic waters (Vincent, 1987; Vincent and

Howard-Williams, 1987).

Lack of water currents or convection in ice-covered lakes requires swimming

mechanisms or small sizes to avoid sinking. Strong planktonic swimmers in Antarctic lakes are

Chroomonas lacustris and Chlamydomonas subcaudata (Parker et al., 1982a; Parker and

Simmons, 1985). Minute, but numerous, heterotrophic bacteria are having highest densities in

the zone of maximum photosynthesis and at the sediment-water interface (Vincent et al, 1981;

Mikell et al., 1984; McKnight et al., 1988). Cryptophyte algae in Lake Bonney produce dimethyl

sulfide and dimethyl sulfoxide. The accumulation in Lake Bonney waters are the highest

recorded concentrations in natural systems (Lee et al., 2004).

Zooplankton are absent in the lakes (Parker and Simmons, 1985). Metazoans occur

infrequently at the underside of the ice in Lakes Hoare and Fryxell (Cathey et al., 1981),

although rotifers and protozoans were reported at the oxic/anoxic boundary (Vincent and

Howard-Williams, 1985).

14

4.0 RECENT ENVIRONMENTAL HISTORY

Advance of the Ross Sea Ice during the Last Glacial Maximum (LGM) increased the

distance of the moisture source; thereby precipitation decreasing precipitation and causing

extremely arid conditions in the Dry Valleys (Bromwich, 1988; Zwally and Giovinetto, 1997).

The aridity caused Taylor Glacier and alpine glaciers in the Dry Valleys to retreat,

contemporaneous with the maximum expansion of the Ross Sea Ice (Marchant et al, 1994;

Higgins et al., 2000). Modelling by Fountain et al. (1998) reconstructed snow-free and cloud-free

conditions during the LGM that caused melting down to the blue glacier ice (Delton et al, 1989).

The melt waters accumulated against the Ross Sea Ice and created one large Glacial Lake

Washburn in Taylor Valley (Hall and Denton, 1996; Hall and Denton 2000; Hall and Denton

2000; Stuiver et al, 1981; Wagner et al., 2006). Rising as high as 350 m above sea level at its

maximum extent, the glacial lake occupied the valley until 8,300 radiocarbon years BP (Hall and

Denton, 2000).

5.0 ORGANIC GEOCHEMISTRY OF TAYLOR VALLEY

Only a few of previous studies in Taylor Valley utilized molecular markers. In the soils

surrounding Lake Bonney, a bimodal distribution of w-alkanoic acids have maximum carbon

numbers at Ci6 and C2s (Matsumoto et al., 1981). Short-chain, <C2o, n-alkanoic acids are from

recent biological sources, whereas long-chain or >C2o and a,co-dicarboxylic acids are from the

Beacon Supergroup that outcrop in the valley, compose the moraines and fluvio-glacial deposits

(Barrett and Kyle, 1975). Matsumoto et al. (1990) elaborated that the biomarkers from the

Beacon Supergroup has low abundances of alkenoic acids and also include thermally altered

15

forms of triterpanes and steranes, such that 17a(H),2ip(H)/17P(H),2ip(H)-Hopane exceeds 1.

Basaltic intrusions into the Beacon Supergroup (Funaki, 1983) and volcanic activity during 2.5

to 4.5 m.a. (Kurasawa, 1986) in the areas surrounding Lake Bonney baked the ancient

biomarkers that included triterpanes, steranes and long-chain ra-alkanes (Matsumoto et al., 1990).

Bacteria are the dominant contributors of iso- and anteiso-alkanoic acids, but bacteria, yeasts,

and microalgae are the sources of 3-hydroxy acids in the soils (Matsumoto et al., 1989). Ancient

vascular plants are also sources of 3-hydroxy acids (Eglinton et al., 1968; Boon et al., 1977;

Kawamura and Ishiwatari, 1982) and long-chain rc-alkanes predominantly odd carbon-numbered

with maximum concentrations at «-C23, n-C25, or n-C2i.

Waters from various depths in both lobes of Lake Bonney contain abundant saturated and

unsaturated C12 to Cig fatty acids. Sediments from both lobes in Lake Bonney, however, differ in

composition (Matsumoto et al., 1979). West Bonney abundantly contains C28 and C29 fatty acids,

whereas the East lobe contains saturated and unsaturated C16 to Cig fatty acids. An alkene, C29:2,

abundant in Lake Fryxell sediments is not detected in the water column, another example of the

differences in the biomarker pools (Matsumoto, 1993).

The major sterol in lake waters and sediments, 24-ethylcholest-5-en-3p-ol, is derived

from cyanobacteria and green algae (Matsumoto et al , 1982; Matsumoto et al , 1989). Mosses

are the major sources of 24-Methyl-cholest-5-en-3p-ol (Matsumoto and Kanda, 1985), whereas

24-ethyl-cholest-5-en-3|3-ol is from diatoms (Nichols et al., 1989). Matsumoto et al. (1988)

reported abundant short-chain, <C2o, normal 2-hydroxy acids in sediments of Lakes Bonney and

Fryxell. Normal 3-hydroxy acids that has predominantly even-carbon numbers, branched 3-

hydroxy acids with predominantly odd-carbon numbers, even-carbon numbered C22 to C30

16

(oo-l)-Hydroxy acids, 9,10-Dihydroxyhexadecanoic and 9,10-Dihydroxyoctadecanoic acids were

also detected in the sediments and were inferred to derive from cyanobacteria and microalgae

(Matsumoto et al., 1988).

Cryptoendoliths that thrive in the Beacon Supergroup sediments, mainly sandstones,

consist of C20 to C30 anteiso-alkanes and anteiso-alkanoic acids (Matsumoto et al., 1992).

Biomarkers in the cryptoendoliths also include «-alkanes, cyclic and acyclic isoprenoids.

6.0 ENVIRONMENTAL IMPACT

The Antarctic area, south of 60 °S, is considered "a natural reserve devoted to peace and

science" under the Antarctic Treaty that was officially enforced in 1961. This Treaty resulted in

scientific and logistical cooperation among twelve countries that were actively involved in

Antarctic research during the 1957-1958 International Geophysical Year. The original Treaty and

about 200 more agreements, collectively known as the Antarctic Treaty System, have been

ratified and now have 26 member countries. Signed in 1991, and enforced in 1998, the Protocol

on Environmental Protection to the Antarctic Treaty prohibits extraction of mineral resources

except for scientific purposes, and provides management protocols for marine pollution,

environmental impact assessments, waste management, protected areas, fauna, and flora. A sixth

provision for liability from environmental emergencies was adopted in 2005.

A well documented example in implementing the protocols of the Treaty is the

decommissioning in 1995 of the New Zealand research base, occupied since 1969, because rising

levels in Lake Vanda threatened to inundate the base, which would release contaminants into one

of the world's most pristine ecosystems (Waterhouse, 1997). Remediation work that started in

17

1992 focused on stations, tracks between stations, helicopter pads, an old incinerator site, and

Greywater Gully, a waste repository. About 15,000 kg of contaminated soil and rocks, two 209-L

drums of polluted water, and scientific equipment were removed, minimizing the spread of

material in all practicable ways. For example, the saw was used indoors as much as possible so

that sawdust were easily collected. Several building anchors, in two meters of permafrost could

not be practically removed, and were cut and left buried. It is expected that all scientific and

tourist activities follow the "Code of Conduct" implemented by individual research bases to

support the Treaty.

18

7.0 CITED LITERATURE

Adams, W., Doran, P. T., Ecclestone, M., Kingsbury, C , & Allan, C : A rare second-year lake ice cover in the Canadian high arctic. Arctic 42; 299-306: 1989.

Barrett, P.J. and Kyle, R.A.: The early Permian glacial beds of south Victoria Land and the Darwin Mountains, Antarctica. In: Gondwana Geology. Campbell K.S.W. (ed.), Canberra, Australian National University Press, 333-346: 1975.

Baroni, C , Orombelli, G.: Abandoned penguin rookeries as Holocene paleoclimatic indicators in

Antarctica. Geology 22; 23-26: 1994.

Bromley, A.: Weather observations. Wright Valley, Antarctica. Wellington, New Zealand: Meteorological Service; vol. 11: 1985.

Campbell, I.A.: Antarctica: Soils weathering processes and environment. Amsterdam, The Netherlands Elsevier, 1987.

Chinn, T.: Physical hydrology of the Dry Valley lakes. In: Physical and biogeochemical processes in Antarctic Lakes, vol. 59, pp. 1-52. Washington, D.C., American Geophysical Union, 1993.

Clow, G., McKAy, C , Simmons, J. G., Wharton, R.: Climatological observations and predicted sublimation rates at Lake Hoare, Antarctica. J. of Climate 7; 715-728: 1988.

Conovitz, P. A., McKnight, D. M., MacDonald, L. H., Fountain, A. G., House, H. R.: Hydrologic processes influencing streamflow variation in Fryxell Basin, Antarctica. In Ecosystem dynamics in a polar desert The McMurdo Dry VAlleys, Antarctica. J. C. Priscu (ed.), vol. 72, pp. 93-108. Washington, D.C., American Geophysical Union, 1998.

Courtright, E. M., Freckman, D. W., Virginina, R. A., Thomas, W. K.: McMurdo Dry Valleys LTER: Genetic diversity of soil nematodes in the McMurdo Dry Valleys of Antarctica. Antarct. J.-Rev.; 203-204: 1996.

Denton, G., Bockheim, J., Wilson, S., & Stuiver, M.: Late Wisconsin and early Holocene glacial history, inner Ross Embayment, Antarctica. Quat. Res. 31; 151-182: 1989.

Doran, P. T., & Wharton, R. A.: McMurdo Dry Valleys LTER: Geophysical determination of bathymetry and morphology of Taylor Valley lakes. Antarct. J.-Rev.; 198-200: 1996.

Doran, P. T., Wharton, R., des Marais, D., McKay, C : Antarctic paleolake sediments and the search for extinct life on Mars. J. of Geophvs. Res.-Planets 103, E12; 28481-28493: 1998.

19

Farman, J., Gardiner, B., Shanklin, J.: Large losses of total ozone in Antarctica reveal seasonal ClOx/NOx interaction. Nature 315; 201-210: 1985.

Foreman, C. M., Sattler, B., Mikucki, J. A., Porazinska, D., Priscu, J. C : Metabolic activity and diversity of cryoconites in the Taylor Valley, Antarctic. J. of Geophys. Res. 112 (doi:10.1029/2006G000358), G04S32: 2007.

Fountain, A. G., Dana, G., Lewis, K., Vaughn, B., & McKnight, D. M.: Glaciers of the McMurdo dry valleys, Southern Victoria Land, Antarctica. In: Ecosystem dynamics in a

polar desert. Priscu, J. C. (ed.), vol. 72, pp. 65-76. Washington, D.C.: American

Geophysical Union, 1998.

Funaki, M.: Paleomagnetic investigation of McMurdo Sound region, Antarctica. Antarctic Res. 78; 1-14: 1983.

Gooseff, M. N., Lyons, W., McKnight, D., Vaugh, B., Fountain, A., Dowling, C.: A stable isotopic investigation of a polar desert hydrologic system, McMurdo Dry Valleys, Antarctica. Arctic, Antarct. and Alpine Res. 38, 1; 60-71: 2006.

Grootes, P. M., Steig, E. J., Stuiver, M., Waddington, E. D., Morse, D. L., & Nadeau, M.-J.: The Taylor Dome Antarctic O record and globally synchronous changes in climate. Quat. Res. 56: 289-298: 2001.

Hall, B., Denton, G.: Extent and chronology of the Ross Sea Ice sheet and the Wilson Piedmont Glacier along the Scott coast at and since the Last Glacial Maximum. Geografiska Annaler Series A-Phvsical Geos. 82A; 337-363: 2000.

Kennedy, A.: Water as a limiting factor in the Antarctic terrestrial environment: a biogeographical synthesis. Arctic Alpine Res. 25; 308-315: 1993.

Kurasawa, H.: Volcano and volcanic rocks in West Antarctica. In Science of Antarct. 5, Geoscience, pp. 88-101. Kokin Syoin, Tokyo, National Institute of Polar Research Japan, 1986. Japanese.

Lyons, W. B., Bartek, L. R., Mayewski, P. A., Doran, P. T.: Climate history of the McMurdo Dry Valleys since the last glacial maximum: a synthesis. In: Ecosystem processes in Antarctic ice-free landscapes. Lyons, W., Howard-Williams, C , Hawes, I. (eds.), Rotterdam, Netherlands, A.A. Balkema., 1997.

Lyons, W. B., Welch, K., Neumann, K., Toxey, J., McArthur, R., Williams, C : Geochemical linkages among glaciers, streams and lakes within the Taylor Valley, Antarctica. In: Ecosystem dynamics in a polar desert: The McMurdo Dry Valleys, Antarctica. Priscu, J. C. (ed.), vol. 72, pp. 77-92. Washington, D.C., American Geophysical Union, 1998.

20

Lyons, W., Tyler, S., Wharton Jr., R., McKnight, D., Vaugh, B.: A late Holocene dessication of Lake Hoare and Lake Fryxell, McMurdo Dry Valleys, Antarctica. Antarct. Sci. 10; 247-256, 1998.

MacAyeal, D., Scambos, T., Hulbe, C , Fahnestock, M.: Catastrophic ice-shelf break-up by an

ice-shelf-fragment-capsize mechanism. J. of Glaciology 49. 164; 22-36: 2003.

Matsumoto, G.I.: Geochemical features of the McMurdo Dry Valley lakes, Antarctica. In: Physical and bio geochemical processes in Antarctic lakes. Antarcic Research Series 59, pp. 95-118, Washington, D.C., American Geophysical Union, 1993.

Matsumoto, G.I., Friedmann, E.I., Watanuki, K., Ocampo-Friedmann, R.: Novel long-chain anteiso-alkanes and cmteis o-a\kanoic acids in Antarctic rocks colonized by living and fossil cryptoendolithic microorganisms. J. of Chrom. 598; 267-276: 1992.

Matsumoto, G. Kanda, H.: Hydrocarbons, sterols, and hydroxy acids in Antarctic mosses.

Antarct. Res. 87, 23-31: 1985.

Matsumoto, G., Torii, T., Hanya, T.: Distribution of organic constituents in lake waters and sediments of the McMurdo Sound Region in the Antarctic. Memoirs of the National Inst. of Polar Res. 13: 103-120: 1979.

Matsumoto, G., Torii, T., Hanya, T.: High abundances of long-chain normal alkanoic acids in Antarctic soil. Nature 290, 688-690: 1981.

Matsumoto, G., Torii, T., Hanya, T.: High abundance of algal 24-ethylcholesterol in Antarctic

lake sediments. Nature 299, 52-54: 1982.

Matsumoto, G.I., Hirai, A., Hirota, K., Watanuki, K..: Organic Geochemistry of the McMurdo

Dry Valleys soil, Antarctica. Org. Geochem. 16, 4-6; 781-791: 1989.

Matsumoto, G.I., Watanuki, K., Torii, T.: Hydroxy acids in Antarctic lake sediments and their geochemical significance. Org. Geochem. 13, 4-6; 785-790: 1988.

Moorhead, D., Priscu, J.: The McMurdo Dry Valley ecosystem: organization, controls, and linkages. In: Ecosystem dynamics in a polar desert. Priscu, J. C. (ed.), vol. 72, pp. 351-363. Washington, D.C., American Geophysical Union, 1998.

Morse, D., Waddington, E., Steig, E.: Ice age storm trajectories inferred from radar stratigraphy at Taylor Dome, Antarctica. Geophvs. Res. Let. 25; 3383-3386: 1998.

Nichols, P.D., Palmisano, A.C., Rayner, M.S., Smith, G.A., White, D.C.: Changes in the lipid composition of Antarctic sea-ice diatom communities during a spring bloom: an indication of community physiological status. Antarct. Sci. 1; 133-140: 1989.

21

Parker, B., Simmons Jr, G., Wharton Jr, R., & Seaburg, K.: Removal of organic and inorganic matter fromAntarctic lakes by aerial escape of blue-green algal mats. J. of Phvcol. 18; 72-78: 1982.

Parker, B., Simmons, G., Seaburg, K., Cathey, D., & Allnutt, F.: Comparative ecology of plankton communities in seven Antarctic oases. J. of Plankton Ecol. 4; 271-286: 1982.

Priscu, J. C , Fritsen, C , Adams, E., Giovannoni, S., Paerl, H., McKay, C : Perennial Antarctic lake ice: an oasis for life in a polar desert. Science 280; 2095-2098: 1998.

Priscu, J. C , Wolf, C., Takacs, C., Fritsen, C., Laybourn-Parry, J., Roberts, E.: Organic carbon transformations in the water column of a perenially ice-covered Antarctic lake. BioScience 49; 997-1008: 1999.

Roberts, E., Laybourn-Parry, J., McKnight, D., & Novarinos, G.: Stratification and dynamics of microbial loop communities in Lake Fryxell, Antarctica. Freshwater Biol. 44; 649-661: 2000.

Steig, E. J., Morse, D. L., Waddington, E. D., Stuiver, M. G., Mayewski, P. A., Twickler, M. S.: Wisconsinian and Holocene climate history from an ice core at Taylor Dome, western Ross embayment, Antarctica. Geografiska Annaler 82A; 213-235: 2000.

Vincent, W. F.: Microbial ecosystems of Antarctica. New York, Cambridge University Press,

1988.

Virginia, R. A., Wall, D. A.: How soils structure communities in the Antarctica Dry Valleys.

BioScience 49; 973-983: 1999.

Waterhouse, E.: Implementing the protocol on ice-free land: The New Zealand experience at Vanda Station. In Ecosystem processes in Antarctic ice-free landscapes. Lyons, W. B. Howard-Williams, D., Hawes, I. (eds.), Rotterdam, The Netherlands, A.A. Balkema, 1997.

Wharton, R. A., McKay, C. P., Clow, G. D., Andersen, D. T.: Perennial ice covers and their influence on Antarctic lake ecosystems. In: Physical and biogeochemical processes in Antarctic lakes. Green, W. J., Friedman, E. I. (eds.), vol. 59, pp. 53-70. Washington, D.C., American GEophysical Union, 1993.

Wynn-Williams, D., Russell, N., Edwards, H.: Moisture and habitat structure as regulators for microalgal colonists in diverse Antarctic terrestrial habitats. In: Ecosystem processes in Antarctic ice-free landscapes. W. B. Lyons, Howard-Williams, & Hawes (eds.), Rotterdam, A.A. Balkema, 1997.

CHAPTER 2

SPME-GCMS study of the natural attenuation of aviation diesel spilled

on the perennial ice cover of Lake Fryxell, Antarctica

Caroline M.B. Jaraulaa, Fabien P. Keniga, Peter T. Dorana, John C. Priscub,

Kathleen A. Welchc

"Department of Earth and Environmental Sciences, University of Illinois at Chicago, 845 WTaylor St., Chicago, Illinois 60607

Department of Land Resources and Environmental Sciences, Montana State University, Bozeman, MT 59717

cByrd Polar Research Center, Ohio State University, Columbus, OH 43210-1002

accepted in Science of the Total Environment

22

23

Abstract

In January 2003, a helicopter crashed on the 5m thick perennial ice cover of Lake Fryxell

(McMurdo Dry Valleys, East Antarctica), spilling -730 liters of aviation diesel fuel (JP5-AN8

mixture). The molecular composition of the initial fuel was analyzed by solid phase

microextraction (SPME) gas chromatography-mass spectrometry (GC-MS), then compared to

the composition of the contaminated ice, water, and sediments collected a year after the spill.

Evaporation is the major agent of diesel weathering in meltpool waters and in the ice. This

process is facilitated by the light non-aqueous phase liquid properties of the aviation diesel and

by the net upward movement of the ice as a result of ablation. In contrast, in sediment-bearing

ice, biodegradation by both alkane- and aromatic-degraders was the prominent attenuation

mechanism. The composition of the diesel contaminant in the ice was also affected by the

differential solubility of its constituents, some ice containing water-washed diesel and some ice

containing exclusively relatively soluble low molecular weight aromatic hydrocarbons such as

alkylbenzene and naphthalene homologues. The extent of evaporation, water-washing and

biodegradation between sites and at different depths in the ice are evaluated on the basis of

molecular ratios and the results of JP5-AN8 diesel evaporation experiment at 4 °C. Immediate

spread of the aviation diesel was enhanced where the presence of aeolian sediments induced

formations of meltpools. However, in absence of melt pools, slow spreading of the diesel is

possible through the porous ice and the ice cover aquifer.

Keywords: jet fuel, naphthalenes, oil spill, water washing, metabolites, SPME-GCMS

24

1.0 INTRODUCTION

The McMurdo Dry Valleys (MCM), which comprise the largest ice-free region of

Antarctica, is the coldest and driest desert on our planet and is the site of the MCM long-term

ecological research (LTER) program (http://www.mcmlter.org/). This polar desert is dominated

by microorganisms, the highest forms of organisms being rotifers and nematodes (Roberts et al,

2000). Growth of these organisms occurs during the short summer melt season in the liquid

water of the lakes, soils and glacial streams (McKnight et al., 1999; Treonis et al, 1999), within

glaciers (Christner et al., 2005; Foreman et al., 2007), as well as in sediment layers trapped in the

perennial ice cover of the lakes (Priscu et al, 1998; Priscu, 1999). Vertebrates and vascular

plants are absent (Virginia and Wall, 1999). The cold desert climate, nutrient limitations, and

slow-growing biological communities make the MCM ecosystem extremely sensitive to climatic

changes and human impact (Cowan and Tow, 2004; Lyons et al. 2006).

The MCM, because of their remote location and severe climate conditions, remained

pristine for decades after their discovery in 1903. In the last few decades, increasing exploration,

scientific operations, and tourism have affected the MCM, mainly at and around research bases

(e.g. Vincent, 1996; Aislabie et al., 2004). Limited tourism started in the valleys in the early

1990's and the MCM-LTER site was established in 1993 (Wharton and Doran, 1999).

Helicopter operations related to scientific exploration alone increased by 30% from the early

1970s into the 1990s (Vincent, 1996). Fossil fuel combustion by helicopters in the MCM

exceeds that of all other utilities such as stoves, heaters, generators, and all-terrain vehicles

(Lyons et al, 2000). During the austral summer of 1995-1996, 106 or 20% of the helicopter

landings and take-offs in the McMurdo Sound area occurred at Lake Fryxell in Taylor Valley

(Fig. 1A; Wharton and Doran, 1999). Despite the frequency of landings, particulate organic

http://www.mcmlter.org/

25

carbon or elemental carbon fluxes derived from helicopters are orders of magnitude lower than

the total active carbon mass in the valleys (Lyons et al., 2000). Most hydrocarbon

contaminations in the McMurdo Sound area occurs in landfills, helipads, refueling areas, and

from leaks in fuel tanks and fuel lines within McMurdo Station (e.g. Kennicutt et al., 1992; Gore

et al., 1999). Relatively high levels of hydrocarbon contamination are also found at research

stations in other parts of Antarctica (Saul et al., 2005). A majority of this contamination occurs

primarily in soils adjacent to the stations where human activities are focused (Aislabie et al.,

2004).

On January 17, 2003, a Bell 212 helicopter crashed on the thick perennial ice cover of

Lake Fryxell (Fig. 1). Approximately 730 liters of aviation fuel (JP5-AN8) as well as small

amounts of hydraulic fluid (MIL 5606), synthetic transmission fluid (Aeroshell 555), and

synthetic engine lubricant (Aeroshell 500) were spilled. This accident resulted in the largest

documented spill in the MCM. Within four days, an emergency cleanup team from McMurdo

Station was dispatched to the crash site. When the approaching winter season forced suspension

of the clean-up efforts, it was estimated that no more than 45% of the spilled fluids had been

recovered (Alexander and Stockton, 2003). In a review of the potential impacts of research

activities in the MCM, Wharton and Doran (1999) concluded that a hazardous spill on the lake

ice will have the least impact if cleanup response is immediate. Such action, however, is possible

only early in the field season, from September to early November, while the ice is solidly frozen.

Spills late in the summer that penetrate the ice cover would be difficult to remove or contain

(Jepsen et al. 2006). The crash occurred during a warm period when the lake ice was isothermal

near the melting point, and the ice was partly covered by large meltpools, the worst possible

scenario for a successful clean-up effort.

26

4 ^ Crash site 3 8 + Meltpool water or

Sediment sample

Figure 1-(A and B) Helicopter crash site on Lake Fryxell ice cover, Taylor Valley, Antarctica. Locations of meltpool waters, sediments and ice cores collected a year after the crash (C). W=westernmost and E=easternmost samples.

27

Few studies on the impact of accidental hydrocarbon releases in the MCM have been

conducted, focusing only on the effects on soil biogeochemistry as most incidents occurred on

land (e.g. Lyons et al., 2000; Aislabie et al., 2004). Collectively, these soil-based studies led to

the conclusion that hydrocarbon contaminants in this cold desert environment can persist for

many years and have profound effects on microbial diversity and geochemistry. However,

results of those studies cannot be directly extrapolated to the ice cover, which has significantly

different physical, chemical and biological characteristics than soils.

In this report, we compare the initial chemical composition of the fuel spilled on Lake

Fryxell ice with that of fuel residue collected a year later in the ice cover. Specifically, we

evaluate and compare the effects of biodegradation, evaporation, and water washing on the

abundance and composition of the residual aviation diesel fuel. We also discuss the potential for

spilled fuel to cross the 5 m ice cover and reach the lake water.

2.0 Materials and methods

2.1 Study area and ice cover conditions

The helicopter crash site (77°36'41.098" S, 163°06'47.228" E) is located 602 m from the

Fryxell Camp benchmark Fry 1 along azimuth 200°50'39" (Fig IB; Alexander and Stockton,

2003). At this site, the ice cover was 5.3 m thick. The average annual ice-cover thickness is a

balance between 15 to 60 cm y"1 of ablation loss at the surface via sublimation, and -30 cm y"1 of

freezing of lake water at the bottom, generating a net upward movement of the ice cover.

Precipitation in the Lake Fryxell basin is minor (<10 cm y"1 water equivalent) and almost

exclusively as snow (Bromley, 1985). The lake basin typically has 30 melt days (Henderson et

28

al., 1965). Patches of aeolian sand, embedded or covering the ice has a lower heat capacity than

the surrounding ice (Adams et al., 1998), and induce the formation of meltpools up to 1.5 m wide

on the surface during the austral summer. In the subsurface, the sediments melt into the ice and

leave a trail of liquid water in their path (Adams et al., 1998). Some of the sediments collect in

pockets and in the summer accumulate at the bottom of cavities in the porous ice. Fritsen et al.

(1998) identified a layer of sediment at ~ 0.5 m depth in Lake Fryxell and referred to this

sediment layer an "aquifer" in the ice as this contains liquid water from November to February.

Liquid water volume in the ice cover can be as high as 40% at the peak of summer (Adams et al.,

1998).

Coincidentally, prior to the accident, a thermal array was deployed to monitor the

temperature of the ice cover near the crash site. During the ten days prior to the crash, maximum

daytime temperature was 2-3 °C, considerably warmer than the average summer temperature of

-3 °C (Margesin and Schinner, 1999). Above-freezing temperatures resulted in the formation of

large meltpools, which exposed subsurface chambers generally 0.2 to 0.5 m beneath the surface

with a maximum depth of 1.2 m. These meltpools are mainly located south of the crash site

where more sediment had accumulated on the ice (Alexander and Stockton, 2003). The sampling

grid (Fig. 1C) separates the northern and southern sectors (Alexander and Stockton, 2003).

Some of the spilled fuel-oil mixture was visible on meltpool surfaces and in subsurface

chambers, some of which were accessed with chisels and ice-axes during clean-up operations.

On the 18th day after the crash, the contamination was visible to 19 m radius from the crash site

(Alexander and Stockton, 2003). Daytime temperatures above freezing lasted until January 23rd,

six days after the crash. No temperature data are available from January 23rd until February 6th,

29

when a data logger was installed at the crash site (Alexander and Stockton, 2003). Surface

temperatures fluctuated around 0 °C. Temperatures dropped below 0 °C only on the 14th and 16th

of February at a depth of 0.25 and 0.50 m, respectively.

2.2 Samples

From December 2003 to January 2004, during the austral summer, approximately a year

after the accident, twelve 7.6 cm diameter and 2-4 m long ice cores were collected using a Sipre

corer fitted with a two-cycle motor head. In the northwest quadrant of the sampling grid, cores 1

to 5 were collected at sites believed to be uncontaminated, though fuel sheen covered the

surfaces of water filling the core holes (Fig. 1C). Cores 6,11 and 12 were recovered in a

summer-melt area near the impact site. All of the cores were wrapped in aluminum foil that had

been baked at 500 °C for 12 h. The cores were kept frozen at -22 °C until processed at the

University of Illinois at Chicago. Seven samples of melt waters were collected from meltpools at

the accident site. Separate sediment samples were also collected at sites 20 and 38. The

sediments and water samples were kept at 4 °C until analyzed. Raytheon Polar Services provided

a sample of the JP5-AN8 aviation fuel used in McMurdo.

2.3 Core sampling

The upper 70 cm of the cores were logged for brittleness, transparency, presence of

bubbles, laminations, and fluorescence under 365 nm ultraviolet (UV) light. Under UV light,

JP5-AN8 diesel fuel has a yellow fluorescence, whereas transparent and opaque ice is pink to

purple. Subsamples were collected where ice character and UV light fluorescence changes as

well as where sediments were present. Subsamples were cut into 1 cm-thick ice chunks by a

30

methanol-cleaned band saw in a -22 °C walk-in freezer. Half of each subsample was archived;

the other half thawed at 4 °C in a closed glass jar. All glassware used was baked overnight at

500 °C, and all caps were lined with solvent-cleaned polytetrafluoroethylene.

Fluid in an ice bubble was sampled by manually drilling into the bubble airspace, then

collecting with a syringe. Approximately 15 ul of fluid was recovered.

2.4 Extraction of volatiles

To minimize loss of low molecular weight volatiles, the vapor phases of the meltpool

waters and thawed ice water were sampled using solid phase micro-extraction (SPME;

Pawliszyn, 1999). A 2000 ul aliquot of each thawed subsample was pipetted into headspace

sampling vials. While the samples were continuously stirred, SPME headspace sampling was

performed using a 100 um polydimethylsiloxane phase fiber for 10 minutes at 40±1 °C. For

samples overloaded with fuel, a lesser volume of aliquot (10 ul) was used and diluted with

ultrapure water to total 2000 ul. For the bubble fluid, 0.1 ul was diluted in 2000 ul ultra pure

water.

2.5 Gas chromato graph-mass spectrometry

A HP-6890 gas chromatogram (GC) coupled to a HP-5973 Mass Selective Detector was

used in electron ionization mode at 70 eV with helium as carrier gas. The column was a 30 m

long HP-5MS with 0.25 mm I.D. and 0.25 um thick film. Mass scan range was 40 to 650 at a

rate of three scans per second. Analytes were desorbed from the SPME fiber for 30 seconds into

the injector, which was operated in splitless mode at 220 °C.

31

Quantification of compounds Z-C13 and n-Cu was achieved by integrating peak areas

from total ion chromatograms, from which the total abundance of alkanes was also calculated.

Total naphthalenes was quantified by calculating the sum of integrated peak areas from mass

chromatograms m/z 128, 142, 156 for naphthalene, dimethyl- and ethylnaphthalenes, and

trimethylnaphthalenes, respectively.

2.6 Fuel Evaporation Experiment

Fuel-water mixtures were left uncapped to evaporate in a 4 °C walk-in refrigerator. Six

4000 (0,1 headspace sampling vials were filled with 1900 ul ultra pure water and 100 ul of JP5-

AN8 fuel. The vials were sealed after 2, 4, 6, 8, 16 and 21 days. Composition of the remaining

fuel in each vial was analyzed using SPME-GC-MS following the method described above, but

with the front inlet operating in 99:1 split mode. Vials were weighed before and after the

experiment.

3.0 RESULTS

3.1 JP5-AN8 fuel mix

The exact ratio of JP-5 to AN8 fuel is most likely in the range 60-70% JP-5 and 40-30%

AN8. Specifications of JP-5 are found in MIL-PRF-5624S (1996). This fuel has a high flash

point (60 °C) and a freezing point of-46 °C (Edwards, 2007). Specifications for the JP-8 grade

aviation kerosene AN8 are contained in MIL-DTL-83188E (1999). This fuel has a flash point of

at least 38 °C and a freezing point of-58 °C or lower (DESC, 2005).

32

The JP5-AN8 aviation diesel consists mainly of compounds with low molecular weights

(LMW; 0.864 g/ml at 4 °C; 0.778 g/ml at 25 °C). Short-chain «-alkanes (C7-Ci6) dominate

(-32% of the fuel), whereas n-C\6 to «-Ci8 are present in trace amounts (Fig. 2A). The most

abundant «-alkane is n-Cn. Monomethylalkanes (M) as well as isoprenoids and alkylbenzenes

from Cg-Ci6 are also abundant (Fig. 2B and C). Polycyclic aromatic hydrocarbons (PAHs)

including mono-, di- and, ethylmethylated naphthalenes comprise -4% of the total diesel mix

(Fig. 2D). Trimethylnaphthalenes, phenantrene, anthracene and methylanthracene are present in

trace amounts.

3.2 Ice Cores from "uncontaminated" sites

Cores 1 to 5 were collected northwest of the crash site, in an area designated

"uncontaminated" (Fig. 1C). After collection of ice cores 1 to 4, fuel sheen was observed on the

water that filled the core holes. The cores generally consisted of opaque granular ice and do not

contain any significant layers or patches of sediments (see supplemental material in Appendix

II). Ice cores 2, 4 and 5 did not emit yellow fluorescence under UV light, suggesting absence of

diesel contamination. Blank SPME-GC-MS total ion chromatograms from core 2 at 0-2 cm

(surface) and 10-11 cm, from core 4 at depths 0-2, 13-14, 39-40, 55-56, and 80-81 cm, core 5

from depths 3-4 and 11-12 cm confirmed the absence of contaminants (see supplemental

material in Appendix II). However, SPME-GC-MS of 67-68 cm from core 4 showed traces of

JP5-AN8 hydrocarbons, though it did not fluoresce under UV, indicating that the some of spilled

diesel penetrated the uncontaminated area. Bubbles with yellowish fluid at 78 cm in ice core 1,

13 and 21 cm in ice core 3 emitted fluorescence, again indicating that spilled diesel is present in

the "uncontaminated" area.

33

It can be hypothesized that the fuel sheen on the water filling the core holes may

derive from rupture of fuel-filled bubbles during coring but may derive also from

contamination of the ice cover aquifer. Lateral transport of contaminants with water

within the ice cover, especially within granular ice layers, from more contaminated

areas is possible.

3.3 Ice cores in contaminated sites

Ice cores 6 and 12, collected in the contaminated area, were analyzed in detail. For the

sake of brevity, only results from core 12 are described, as this core demonstrate well the various

processes involved in fuel compositional changes over time. Description and SPME-GCMS data

for ice core 6 are available in the supplemental material (Appendix II-1). The upper 15 cm of ice

core 12 comprise two couplets of opaque and transparent ice (Fig. 3). Opaque and granular

layers form from rapid freezing or compacted snowfall, whereas transparent ice is formed from

slow freezing of pockets of melt waters (Adams et al., 1998). Sparse sediments are present

between 5 and 6 cm as well as between 8 and 9 cm. Just above these sediment patches, fluid-

filled spherical and inverted teardrop-shaped bubbles emit intense yellow fluorescence under UV

light, suggesting abundance of diesel contaminant. The ice between 15 and 30 cm gradually

changes from semi-transparent to transparent downward, indicating decreasing rate of freezing

from the top to the bottom of the layer. This suggests that the ice between 15 and 30 cm

corresponds to the progressive downward freezing of a single meltpool. Between 30 and 48 cm,

the core consists of granule to pebble-sized ice chunks with interspersed sediments. Ice chunks

with sediments between45 and 48 cm emit intense yellow fluorescence under UV light,

suggesting abundance of diesel contaminant. A prominent sediment layer, between 48 and

34

JP5-AN8 A Total Ion Current

m/z 91+92+105+106+119+120+133+134+147+148 C _ C, _ Alky (benzenes

x

C, > , .

6k ^ f frfr '

m/z 128+115+141+142+155+156+170+169+178+184+192 D b . Naphthalenes

„

>>

Inte

nsi

• > ...

1

n-C, •

i 1 , tan 1

•

z-C 0

10 20

Retention time (min)

Fig. 2-SPME-GCMS of JP-5 and AN8 diesel fuel mixture. (A) Total ion current, (B) ra-alkanes, monomethylalkanes, and isoprenoids, (C) alkylbenzenes, and (D) naphthalenes, aromatic hydrocarbons (PAHs) including mono-, di- and, ethylmethylated naphthalenes comprise - 4 % of the total diesel mix (Fig. 2D). Trimethylnaphthalenes, phenantrene, anthracene and methylanthracene are present in trace amounts.

35

53 cm, consisted mainly of angular, fine to granular, buff, white or transparent quartz, and fine to

medium black hornblende, other amphibole minerals, and red iron oxide coated angular granules

of rock fragments that are all aeolian in origin. The ice from 53-80 cm is transparent and

laminated with patches of sediments between 66 and 72 cm. These sediments are similar in

composition to those observed between 48 and 53 cm.

SPME-GC-MS results indicate that the surface sample (0-1.5 cm) contained diesel-

derived compounds with a mode at «-Cn (Fig. 4B, Table I). A shift in n-alkanes distribution

mode of the contaminants relative to that of the original fuel (mode at n-Cu, Fig. 4A) suggests

evaporation of low molecular weight compounds. The fluid recovered from an inverted teardrop-

shaped bubble at 4 cm has minor compositional differences relative to the JP5-AN8 diesel

mixture, suggesting minimal weathering of the diesel as a result of evaporation (Fig. 4C).

Aromatic ketones are also present although in low abundance.

The ice between 12 andl3 cm is also contaminated with diesel-derived compounds

dominated by n-Cu (Fig. 4D), indicating lesser evaporative weathering than for the surface

sample. The sample collected between 25 and 26 cm also contains compounds derived from

diesel, but dominated by n-C^ and with high amounts of naphthalenes relative to n-alkanes (Fig.

4E). In this sample, preferential loss of w-alkanes compared to branched-alkanes, isoprenoids,

and aromatics suggests that biodegradation is the main agent of contaminant weathering as

normal alkanes are preferentially degraded relative to branched alkanes, isoprenoids, and cyclic

compounds, which in turn are more easily biodegraded than aromatics (e.g. Leahy and Colwell,

1990).

36

TA

BL

E I

Per

cent

age

of f

uel

com

pone

nts

and

proc

esse

s af

fect

ing

the

com

posi

tion

of

the

fuel

res

idue

AN

8 Si

te 2

0 Si

te 3

8 C

ore

12 (d

epth

s in

cm

)

wat

er

sed

wat

er

sed

0-1.

5 5*

12

-13

25-2

6 45

-46

Tot

al a

lkan

es

29

26

33

36

16

35

34

27

20

0

Tot

al n

apht

hale

nes

4 11

13

4

15

7 2

7 12

31

Mod

e rc

-Cn

n-C

u n-

Cu

n-C

u D

M-n

aph

n-C

u n-

Cu

n-C

\ 2

n-C

u M

e-N

aph

«-C

i4

UC

M

Y

Y

Eva

pora

tion

Y

Y

Y

Y

Y

Y

Bio

degr

adat

ion

Y

Y

Y

Dis

solu

tion

Y

Y

Val

ues

are

% o

f to

tal

fuel

. *F

luid

in

ice

bubb

le;

DM

-dim

ethy

lnap

htha

lene

; M

e-m

ethy

lnap

htha

lene

; Y

-pre

sent

37

? Core 12 GENERAL DESCRIPTION

tee character

s air bubbles w/

10

20

30

40

50

60

70

luiifiiaa 15

2gU.

Sediment

= s a r 5 - 8 3EZ3C8-9

K\" Patches of sand ifce medium to coarse, angular

•30-

Brittle, white, sub-rounded granule to pebble ice chunks

_37„

break in ice record

laminated

medium to coarse sand coating the ice chunks

3tr fine to granular and*" " - _aub=angutaf ,sanct

«66«*

}/\ medium to coarse sand L_X2 — — —•

UV light absorption

and reaction

bubbles have yellow fluorescence

Samples depth (cm)

™0-1.5~"

bubble fluid

— 12-13

•25-26

48 I 45-48

^57-59

Fluorescence legend:

_ Intense

Diffused

Figure 3.Log of ice core 12. Black rectangles indicate samples used for SPME-GCMS analyses.

38

Samples collected between 45 and 46 cm and between 57 and 59 cm contain well

resolved PAHs dominated by methylated naphthalenes (Fig. 4F and G). The presence of low

molecular weight (LMW) aromatic compounds (120 amu) in sample 45-46 cm precludes

evaporation as the only cause of alkane loss. Comparison of the distribution of aromatics in

this sample with that observed in the diesel-mix (Fig. 5) shows an increase in abundance of

LMW relative to high molecular weight (HMW) naphthalenes in the ice sample. Such a

change is incompatible with evaporation and biodegradation, which should have affected

LMW before HMW compounds (Wang et al., 1998). The well resolved peaks and the

absence of an unresolved complex mixture (UCM), typical of biodegraded diesel fuel,

excludes biodegradation as a dominant weathering process. Preferential dissolution of LMW

PAHs relative to other diesel-derived compounds can account for the exclusive presence of

PAHs and their distribution. Zurcher and Thtier (1978) and Bradley and Chapelle (1995)

showed that LMW aromatics in petroleum preferentially dissolve in water, resulting in the

increased concentration of aromatics in water. The solubility of PAHs in water generally

decrease with increasing molecular weight and varies as a function of alkyl position. This

explains why 1-methylnaphthalene (peak c; solubility, 28 mg/L at 25° C; MacKay et al.,

1992) is more abundant than 2-methylnaphthalene (peak b; solubility 24.6 mg/L at 25 °C) in

samples 45-46 and 57-59 cm, but is less abundant than 2-methylnaphthalene in the original

fuel (Fig. 5). The increased abundance of naphthalene (peak a; solubility 31 mg/L at 25°C)

relative to both methylnaphthalenes in the ice samples (Fig. 5) can also be explained by

preferential dissolution. The same observation can be made for the low abundance of

dimethylnaphthalenes (peaks labeled e; least soluble) relative to methylnaphthalenes (more

soluble). Therefore, the ice at depths 45-46 cm and 57-59 cm formed from water that was in

39

• «-aikane$ o Isoprenoids

«-C, n-C.

IS

IS

IS

IS

""**^H.

2-V-t;

ff...„ k *,..« », I .MgMdhrtU.iifl^

JP5-AN8 A H-C Total !on Current **

O I o • L

rtninttm .f, —.—,—.—

A J I L M B M M I M S I ^ . F

0-1.5 cm B

iuU*iu) L4

4 cm C Ice bubble fluid

L»^MiAMiiJm^*^!immm

: 12-13 cm D

i k « u tfAllu

•..» k JJJJU^IJMU^

• 25-26 cm £

b „ - • UCM

AiatJUu*k*«jJL e 45-46 cm p

Kl

= 05

I., M ilmliM.h i Ititii

K2f \ K4 .. 48-50 cm c

' ?~rK set^ments

lw&&aJUL Jutf&U

nJwtiiliimin wij,i n |n HI i M in

57-59 cm H

20 30 40 50 Retention time (rain)

60

Figure 4. SPME-GCMS total ion current of (A) JP5-AN8 diesel mixture, (B-H) ice core 12 samples. Structures of naphthalene homologues a, b, c and d are in Fig. 5 A. Core sediments (G) contain biodegradation products of aromatic compounds: Kl. 2,3-dihydroindan-l-one, K2. 2-methyl-5-(l-methylethenyl)-2-cyclohen-l-one, K3. 2,3-dihydro[2H]-naphthalen-l-one, K4. 5,7-dimethylindan-l-one, K5. 4,7-dimethylindan-l-one, K6. 6,7-dimethylindan-l-one. Internal standard (IS) used is trichloroethylene.

40

contact with diesel fuel and in which LMW PAHs were preferentially dissolved and was

transported to a subsurface cavity at site 12 where it froze and was sampled. The presence of

low molecular weight aromatics in sample 45-46 cm indicates some, but limited, evaporative

loss and suggests limited contact between contaminated water and the atmosphere.

The sediment layer between 48 and 50 cm mostly contain aromatics, such as

naphthalenes, methyl-, dimethyl- and ethyl-naphthalenes (Fig. 4G) but aromatic ketones

(such as napthalenones, indanones, methyl-, dimethyl-, and ethyl-indanones) are also

prominent. These aromatic ketones are metabolites formed from biodegradation of aromatic

compounds (Langbehn and Steinhart, 1995) but have also been considered products of

photooxidation (Prince et al., 2003). The contribution of photooxidation to the formation of

these ketones cannot be discounted directly, however, we have observed these compounds

only in samples associated with sediment layers, where we observed also other effects of

biodegradation.

In summary, fuel contained in the surface ice, such as sample 0-1.5 cm is likely to be

more frequently exposed to the atmosphere, enhancing the loss of volatiles and resulting in

the strong evaporation signal. More frequent and short-term freeze-thaw cycles produced the

alternating layers of sediment in transparent and opaque ice between 4 and 9 cm. Gases

exsolved during freezing created bubbles where the fuel was also encapsulated. The fuel

trapped into the bubbles is not significantly weathered, though the presence of minor

amounts of aromatic ketones suggests biodegradation. It can be speculated that this fuel-

water inclusion remained fluid for a prolonged period of time and may have formed as an

oasis of life in the ice.

41

CO

C +->

b

JP5-AN8 A Ci m/z 128+142+156

00

JUL

30 ll^l»iiiiatihiBw4*i^..i.i*Jk)ftwwil<iMi|>

45-46 cm

m/z 128+142+156

d1 ©

\ flJ^JLJfJfm

40 50 Retention time (mm)

60

Figure 5.Summed mass chromatogram m/z 128+142+156 showing the relative abundance of naphthalenes in (A) JP5-AN8 diesel mixture and (B) in ice core 12, 45-46 cm. a. naphthalene, b. 2-methylnaphthalene, c. 1-methylnaphthalene, dl and d2. ethylnaphthalenes, e. dimethylnaphthalenes.

42

In contrast, ice layer between 9 and 30 cm formed from water that froze gradually

from the top. The opaque ice layer from 9 to 12 cm froze and quickly insulated the water

below and allowed the rate of freezing to decrease as seen from the gradual change in ice

character from semi-transparent to transparent between 15 and 30 cm. In samples 12-13 cm

and 25-26 cm, the presence of trace amounts of M-C9 (128 amu), a hydrophobic, insoluble,

LMW compound indicates moderate evaporative influence (Fig. 4 D and E). Specifically for

sample 12-13 cm, where the mode is at n-Ciz, evaporation was, therefore quickly quenched.

With waters from 25-26 cm freezing much slower than the rest of the unit, biodegradation

may have persisted longer, resulting in the significant alkane depletion observed (Fig. 4E).

The distribution and abundance of diesel contaminants between 25 and 26 cm as well

as between 12 and 13 cm suggest loss of LMW contaminants via biodegradation and

evaporation, without any suggestion of selective dissolution of diesel compounds. Thus, the

ice samples 25-26 cm and 12-13 cm are derived from a different pool of water than the ice

samples 45-46 cm and 57-59 cm, which contain selectively dissolved compounds.

3.4 Meltpool water and sediment

Samples of meltpool water collected from the easternmost and westernmost areas of

the crash site (E and W, respectively in Fig. IB and C) have no detectable fuel contamination

with SPME-GC-MS (Fig. 6B and C). In contrast, meltpool water and sediment samples

collected closer to the crash site (locations 20 and 38) contained abundant JP5-AN8 residues

(Table I, Fig. 6D-G).

43

The large amount of diesel fuel contaminant in the water sample from site 20

imposed dilution of 10 ul of water sample in 1990 ul of ultra pure water for SPME-GC-MS

analysis. Volatiles in meltpool water of site 20 contain w-alkanes, with a mode at w-Cn,

indicating significant loss of low molecular weight components by evaporation (Fig. 6D).

Branched alkanes and isoprenoids are also abundant, whereas naphthalenes are present in

trace amounts.

Sediments in the meltpool at site 20 also contain JP5-AN8 hydrocarbons, with rc-Cn

as the mode (Fig. 6E). As observed in the original fuel mixture, naphthalenes are in low

abundance. Branched alkanes and isoprenoids are also present, although less prominent than

in the water sample (Fig. 6D). The high isoprenoid to alkane ratio in the water sample

relative to the sediment sample suggests a stronger effect of evaporative attenuation in the

water sample.

At site 38, the meltpool water is dominated by alkanes with n-Cn and M-C14 similarly

abundant (Fig. 6F). Branched alkanes and isoprenoids are also prominent. This hydrocarbon

distribution suggests evaporative loss similar to that observed for the surface sample of

nearby core 6 (see Appendix II).

The sediment sample from site 38 (Fig. 6G) has much less total fuel hydrocarbons

than in the water at site 38. Total n-alkanes and total naphthalenes have equal abundances

(Table I), suggesting biodegradation of the diesel. Overall, dimethylnaphthalenes are the

most abundant compounds, followed by n-Cu- Evaporative loss seems equivalent in both

water and sediment samples.

44

• «-alkasies o isoprenoids

»-Cr fc"

IS

JP5-AN8 A

fl"-\*<,*

East end B

US West end C

IS

IS

; Site 20 water D

"Raised baseline

*tw«JiX«*ajyL«3&&k«^i^

Site 20 E sediments

: Site 38 water F

-M.i......a..i..

«-C,a i-C,

n „ Site 38 G • ' j r • sediments

1 ° ? **tokA»*<A» ,t.ft* ,

0 40 Retention time (tnin)

Figure 6. SPME-GCMS total ion current of (A) JP5-AN8 diesel mixture and for (B-D) meltpool waters and sediments.

45

3.5 Fuel evaporation experiment

As the aviation diesel evaporated, combined weights of the water and fuel

for each sampling day decreased. By the sixth day, the weight loss was about 80

mg, or -10% of the initial fuel weight. The rate of loss was highest during the first

day at 0.7 mg h"1, then decreased to 0.3 mg h"1 on the eighth day. After the eighth

day, weight losses were so small that they could not be monitored accurately.

The total ion chromatogram shows a change in the mode from n-Cn in the

original fuel to n-Cn and w-C^by the fourth day, to n-Cuby the sixth day, and to n-

C B by the sixteenth day. Both n-C\i and i-Cu increase in percentage of the total

fuel residue with time, as LMW volatile compounds are evaporated first. However,

the ratio of i-Cn over n-Cn in the evaporating fuel progressively decreases as /-C13

is more volatile than n-C\j,. The i-Cn/n-Cn ratio varied linearly with time during

the evaporation experiment (r2 = 0.99; n = 6; see supplementary material). The

abundance of total naphthalene (sum of naphthalene, methyl-naphthalenes and

dimethyl-naphthalenes), and total alkane was also monitored. Both groups of

compounds increase in percentage of the total fuel residue but the total naphthalene

over total alkane ratio also increases with progressive evaporation (r2 = 0.89; n = 6;

see supplementary material).

46

4.0 DISCUSSION

4.1 Evaporation

Evaporation is a major process in the weathering of the aviation diesel spilled on the

ice cover of Lake Fryxell as all samples analyzed, including the fluid trapped in ice bubble,

show evaporative losses. Evaporation may have occurred as soon as the fuel came in contact

with the atmosphere during the accident. Further evaporation of the diesel mixture continued

whenever the contaminants were exposed to the atmosphere. In ice cores, evaporation of

diesel contaminants is more intense in surface than in deeper layers, as the former is more

often subjected to freeze-thaw cycles than the latter (Fig. 4). The light non-aqueous phase

liquid (LNAPL) behavior, the dominance of LMW compounds, and the high volatility of

spilled aviation diesel favor evaporation.

4.2 Biodegradation

Biodegradation is another major attenuation process in the ice cover, but the effects

of this process were exclusively observed in ice containing sediments (Fig. 4G and Fig. 6G).

In the ice cover, an ecologically and physiologically complex microbial consortium develops

from the relatively nutrient- and carbon-enriched mixture of sediment and water (Priscu et

al., 1998). Such microbial assemblages were not observed in sediment-free ice layers. The

sediments are transported to the ice through aeolian processes from soil and stream habitats

and provide the inoculums for in-ice habitat (Fritsen et al., 1998). Viable microbial

assemblages are capable of photosynthesis, nitrogen fixation, and decomposition.

Filamentous cyanobacterial genus Phormidium, nitrogen-fixing genus Nostoc, and diatom

47

algae were identified (Priscu et al., 1998). The microbial assemblage must have shifted to

organisms more capable of degrading diesel as observed in contaminated Antarctic soils

(Saul et al, 2005).

4.3 Evaporation vs. biodegradation

Estimating the relative effect of evaporation and biodegradation on natural

attenuation using field samples can be difficult. Snape et al. (2005) developed fuel

evaporation models to quantitatively differentiate the effects of evaporation from those of

biodegradation of diesel in cold environments (4 °C). Their experiments were based on a

diesel fuel, Special Antarctic Blend, which is most often used and spilled at Casey Station

(Australian Antarctic base). This diesel, with a «-alkane mode at n-C\2, is enriched in HMW

compounds compared with the JP5-AN8 mixture studied in this report. Most of the ratios of

Snape et al. (2005) include z'-Ci6 concentration. This compound is either unresolved or absent

from our SPME-GCMS chromatograms, as JP5-AN8 is a lighter diesel than the Special

Antarctic Blend. Thus, for this study, we use the ratio i-C^/n-Cu of the sample (S) divided

by the same ratio in the original diesel (D) mixture as an indicator of evaporation (z'-Cn/w-

Ci3)s/(z'-Ci3/tt-Ci3)D.

M-CH or n-Cn could not be used as these compounds are often in low abundance in

the chromatographic traces of field samples. As Z'-CB evaporates more easily than H-CB, a

ratio value of less than 1 indicates evaporation. Concordant with qualitative observations,

quantitative estimate of evaporations using the (z'-Cn/w-CnVO'-Cn/w-C^D show that

contaminants in all samples are primarily affected by evaporation (Fig. 7). Consistent with

qualitative observations based on n-alkane mode, the contaminants in the surface sample of

48

core 12 (0-1.5 cm) were affected by evaporation more than contaminants located deeper in

the core. Diesel contaminants in meltpool waters are more affected by evaporation than their

corresponding sediments as the sediments are shielded from the atmosphere by overlaying

waters (see samples for site 20 and 38 in Fig. 7). It is noteworthy that the contaminants in the

bubble fluid at ice core 12 4 cm is the least evaporated of all samples, as observed

qualitatively.

As a proxy for the extent of biodegradation we calculated a ratio of the

concentrations of total naphthalene (Tnaph) over the concentration of total alkanes (Taik). The

larger the values of that ratio, the more intense the biodegradation is because alkanes are

more easily biodegraded than naphthalenes. However, evaporation affects abundances of

both naphthalenes and alkanes and the ratio Tnaph/ Taik as a result of their differential

response to evaporation and their difference in initial abundance. To circumvent that caveat,

an evaporation curve based on an evaporation experiment at 4 °C (see method) was plotted

on a (Z'-CI3/«-CI3)S/(J-CI3/«-CI3)D versus Tnapht/Taik diagram (Fig. 7). Field samples falling to

the right of the evaporation curve were affected by biodegradation. For example, fuel

residues in the sediments of meltpool site 38 are more intensely biodegraded than the

meltpool diesel in the ice of Core 12 between 25 and 26 cm.

4.4 Water washing

Samples located to the left of the evaporation curve have lost naphthalenes

preferentially to alkanes. The only process we have identified able to explain these data is

loss of naphthalenes by preferential dissolution of these compounds in water (water

washing). Indeed, we have analyzed ice core samples containing only aromatic com

49

1.0 0 t = 0

Total naphthalene / total alkane 0.1 0.2

0.8-

0.6-

0.4-

0.2

0.0

0.3 Tt = 2

I l t = 4

ft = 6 f t = 8

• W = -9.02x+1.53 bubble! R

2 = o.88 fluid I

on 1 12-13

1 0-1.5 20o\ • ft = 16

38A\

* t = 21

1

* evaporation experiment • core 12 samples (cm) • core 6 samples (cm)

^•meltpool sediments Aomeltpool waters

B

38 4

45-46

Fig. 7-Tnapht/Taik versus (/-CU/M-C^S/O'-CB/W-C^D showing the relative strengths of evaporation versus biodegradation in sediments, meltpool waters and ice core subsamples. The evaporation curve is a linear regression.

50

pounds that were preferentially dissolved (washed-out) from the diesel contaminants (core

12 45-46 cm and 57-59 cm; Fig. 4F and G) and could not be plotted in Fig. 7 as they do not

contain any «-alkanes. At site 20, slight variation in Tnaph/ Taik indicates higher water

washing in the meltpool sediments than in the water. It is noteworthy that the diesel trapped

in the bubble of core 12 at 4cm is partly water-washed, a process that we could not detect

qualitatively from the chromatographic trace.

It is important to note that water washing is not a natural attenuation process, as the

dissolved compounds are not eliminated from the environment. On the contrary, the

preferential transfer of aromatics in water, away from sediments where biodegradation

preferentially takes place, will probably prevent or at least slow down the natural attenuation

of these compounds.

4.5 Lateral and Vertical transport of contaminants

Wharton and Doran (1999) noted that the initial conditions of the ice cover are of

paramount importance when considering clean up efforts, as hard ice of the early summer is

easier to clean than the porous and wet ice of the melt season. Early in the season, the solidly

frozen ice cover will be mostly impermeable, preventing penetration of contaminants into the

ice cover. The contaminants will more likely spread on the ice surface, enhancing attenuation

via evaporation, and most probably sorb on windblown surface sediments. As the summer

melt season advances, the porosity, permeability, and water content of the ice cover increases

with formation of meltpools, subsurface chambers and channels (Adams et al., 1998; Fritsen

et al., 1998), as observed in the southern quadrant of the crash area. Spilled fluids are then

able to penetrate the ice cover, and transport of contaminants is more widespread both

51

laterally and vertically. However, the presence of sediments is not merely detrimental as

biodegradation was observed only in ice containing sediments. Porous ice north of the crash

site, however, allowed the transport of contaminants to previously "uncontaminated" sites

within a year after the crash, as indicated by oil films at the surface of water filling the core

hole (e.g. ice core 4, 36 m from the crash site).

Transported waters into uncontaminated sites may also have enhanced LMW PAH

content as a result of water washing (Fig. 4 F and H). Higher molecular weight PAHs such as

dimethyl-, ethyl-, and trimethyl naphthalenes were observed unaffected by biodegradation

after 1 year. Their solubility will allow future remobilization if in contact with water. As a

result, these compounds will likely affect a wider area with time. These compounds,

particularly dimethyl- and ethyl-naphthalene may be used as tracers to monitor the spread of

residual fuel contamination as they are not occurring naturally in the ice cover.

Ice cover ablation at the surface and the accretion of ice at the bottom yield a net

upward movement of the ice cover without reducing its thickness (Fig. 8 A). This upward

flux of ice will help prevent the contaminants from reaching the lake water if the

contaminants do not move downward faster than 60 cm y"1, the maximum ablation rate

calculated (Henderson et al., 1965). Jepsen et al. (2006), however, showed experimentally

that, at temperatures close to ice melting, aviation diesel JP-8 propagates along ice crystal

boundaries at velocities up to 1.6 m h"1 ("fuel tunneling" Fig. 8B). The extent of vertical fuel

tunneling in the field, however, is difficult to predict as tortuosity and branching of the fuel

tunnels depend on the concentration of impurities and temperature variations in the ice.

52

Windblown sediments trapped in the ice cover that are <1 cm in diameter are able to

melt their way into the ice (Hendy, 2000). Pockets of sediments accumulate up to 2 m in the

ice and create an aquifer in the summer (Priscu et al., in press; Fig. 8B). Compared to the

heat capacity of ice or water (0.502 cal.g"1.'^"1), black basaltic sediment (0.201 cal.g"1 "C"1)

and diesel (0.014 cal.g"1 "C"1) have much lower heat capacities. The presence of diesel

contamination will also favor melting of the surrounding ice. We have observed diesel

sorbed onto sediment grains, in ice cores and meltpools (sites 20 and 38). Thus, the diesel

residues sorbed on the sediment grains will favor the formation of water and sediment

pockets and increase the potential for biodegradation as microbial processes are triggered by

the presence of liquid water in the summer.

The diesel contaminant, if not associated with sediments, will behave as LNAPLs.

The low heat capacity of the diesel can result in the melt of the surrounding ice and the

LNAPLs can progressively melt their way to the surface of the ice cover (Fig 8C). Such a

process is commonly observed for pieces of microbial mats, which were incorporated at the

bottom of the ice and observed to travel across the ice cover (Parker et al, 1982). When

reaching the surface of the ice cover, the contaminants may be further attenuated by

evaporation or may associate with sediments and be biodegraded.

The dissolved fraction of the diesel contamination, mostly including the LMW PAHs,

will be mobile during the melt season. In all ice samples containing the dissolved portion of

water washed diesel (Fig. 4F and H), we have not observed any of the potential effects of

biodegradation. Thus, these compounds are likely to be persistent and natural attenuation

will depend on evaporation only.

53

Ice ablation: A

15-60 .cm y'

Sediments'

B

Evaporation

D

Ice cover

V Meitpool

^^> Fuel tunnels

' + - * •

0.65 ra ~~

Fuel film

LNAPL

Hydrostatic level

« | Fluid in " bubbles

5.3 m

Water column'

f lee accretion: I 30 cm y"1

Fig. 8-Dynamics in the Lake Fryxell ice cover that affect the natural attenuation of contaminating fluids. (A) Ablation loss at the surface and accretion of ice at the bottom generates a net upward movement of 30 cm y"1. (B) Light diesel components evaporate from ice surface and meltpools. Lower heat capacity of fuel and sediments induce melting and results to "tunneling deeper into the ice. (C) Pools of water separate light non-aqueous phase liquids, LNAPLs, making them susceptible for evaporation. LNAPLs may also rise in ice straws and eventually evaporate. (D) Fractions of aromatics that are dissolved are re-frozen in subsurface chambers. Some ice cracks connect connect the "aquifer" to the lake water.

54

As sediments can melt their way to a depth of approximately 2 m in the ice cover,

where the ice aquifer occurs, it can be speculated that diesel contamination will easily reach

that level over time (Fig. 8D). Remobilization of soluble aromatic compounds in the aquifer

and the existence of discrete conduits, such as ice cracks, that connect the ice aquifer to the

lake water (Hendy et al., 2000) may result in the transfer of soluble aromatic contaminants

(e.g. naphthalenes) and sediments coated with fuel into the lake water (Fig. 8D).

Fortunately, the conjugated effects of evaporation and biodegradation, as well as the

permanent upward movement of the ice cover will limit to a minimum the amount of diesel

components able to reach Lake Fryxell water body.

5. CONCLUSIONS

With the helicopter crash occurring late in the summer, the situation was highly

unfavorable as above freezing temperatures days prior to and after the crash induced the

formation of meltpools and subsurface chambers. Spreading of the contaminants from the

crash site occurred both horizontally and vertically. The extent of horizontal spread is

controlled by the volume of pore spaces and network of channels in the ice, which are well

developed in the southern quadrants. Diesel contaminated waters also spread into the

northern sector, which have porous ice.

A year after the crash, a significant amount of spilled fluid remained in the Lake

Fryxell ice cover, particularly south of the crash site. JP5-AN8 is attenuated naturally mainly

by evaporation, especially on the ice surface and in meltpool waters. Spilled fluids evaporate

at decreased rates during the long Austral winter while confined in the ice cover. Further

natural attenuation occurred through biodegradation, especially in sediment-bearing ice

55

layers in which biological activity was previously detected. Alkane-degraders in the ice play

as prominent a role in weathering hydrocarbons as they do in the soils, although the rates are

much slower.

6.0 ACKNOWLEDGMENTS

This project was supported by the Office of Polar Program of the National Science

Foundation (Antarctic Biology and Medicine SGER 0346316) and by a fellowship grant

from the University of Illinois at Chicago Institute of Environmental Science and Policy.

Kelvin Rodolfo is acknowledged for his edits on the manuscript. We thank Apostolis

Sambanis, Marcus Muccianti, Timothy Chung, and Jeffrey Fitzgibbons who helped in the

laboratory.

56

7.0 REFERENCES CITED

Adams E.E., Priscu, J.C., Fritsen C.H., Smith S.R., and Brackman S.L.: Permanent ice covers of the McMurdo dry valley lakes, Antarctica: bubble formation and metamorphism. In: Priscu J.C., editor. Ecosystem dynamics in a polar desert: the McMurdo dry valleys, Antarctica. 72. Am. Geophvs. Union, pp. 281-295, Washington DC, 1998.

Aislabie J., Balks M.R., Foght J.M., and Waterhouse E.J.: Hydrocarbon spills on Antarctic soils: Effects and management. Environ Sci & Technol. 38:1265-1274, 2004

Alexander S.P., and Stockton W.L.: Lake Fryxell 79U Crash Site: Initial Hydrocarbon Monitoring Program. Raytheon Technical Services Company. 44: 2003.

Bradley P.M., and Chapelle F.H.: Factors affecting microbial 2,4,6-Trinitrotoluene mineralization in contaminated soil. Environ Sci & Technol. 29:802-806, 1995.

Bromley A.M.: Weather observations, Wright Valley, Antarctica. New Zealand Meteorological Service. 1985.

Christner B.C., Mikucki J.A., Foreman CM., Denson J., and Priscu J.C.: Glacial ice cores: a model system for developing extraterrestrial decontamination protocols. Icarus. 74:572-584, 2005.

Cowan D.A., and Tow L.A.: Endangered Antarctic environments doi:10.1146/annurev.micro.57.030502.090811. Annu Rev of Microbiol. 58:649-690, 2004.

DESC (Defense Energy Support Center).: Turbine fuel, aviation JP8/AN8 (Deep Freeze).2005.

Edwards T.: Advancements in gas turbine fuels from 1943 to 2005. J. of Engineering for Gas Turbines and Power-Transactions of the Asme. 129:13-20. 2007.

Foreman, CM., Sattler, B., Mikucki, J.A., Porazinska, D.L., and Priscu, J.C: Metabolic activity and diversity of cryoconites in the Taylor Valley, Antarctica. J Geophvs Res. 112: G04S32, doi:10.1029/2006JG000358. 2007.

Fritsen, C.H., and Priscu, J.C: Cyanobacterial assemblages in permanent ice covers on Antarctic lakes: Distribution growth rate, and temperature response of photosynthesis. J of Phvcol 34:587-597. 1998.

Gore, D.B., Revill, A.T., and Guille, D.: Petroleum hydrocarbons ten years after spillage at a helipad in Bunger Hills, East Antarctica. Antarct Sci 11:427-429, 1999.

57

Henderson, R.A., Prebble, W.M., Hoare, R.A., Popplewell, K.B., House, D.A., and Wilson, A.T.: An ablation rate for Lake Fryxell, Victoria Land, Antarctica. J Glaciol 6:129-133, 1965.

Hendy, C.H.: The role of polar lake ice as a filter for glacial lacustrine sediments. Geogr Ann A_82A:271-274, 2000.

Hendy, C.H., Sadler, A.J., Denton, G.H., and Hall, B.L.: Proglacial lake -ice conveyors: a new mechanism for deposition of drift in polar environments. Geog Ann A 82;249-270, 2000.

Jepsen, S.M., Adams, E.E., and Priscu, J.C.: Fuel movement along grain boundaries in ice. Cold Reg Sci and Technol 45:158-165. 2006.

Kennicutt II, M.C., McDonald, T.J., Denoux, G.J., and McDonald, S.J.: Hydrocarbon contamination on the Antarctic Peninsula I. Arthur Harbor-subtidal sediments. Mar Pollut Bull, 10:499-506, 1992.

Langbehn, A., and Steinhart, F.: Biodegradation studies of hydrocarbons in soils by analyzing metabolites formed. Chemosphere 30:855-868. 1995.

Leahy, J.G., and Colwell, R.R.: Microbial-degradation of hydrocarbons in the environment. Microbiol Rev 54:305-315. 1990.

Lyons, W.B., Laybourn-Parry, J., Welch, K.A., and Priscu, J.C.: Antarctic lake systems and climate change. In: Trends in Antarctic terrestrial and limnetic ecosystems: Antarctica as a global indicator. Bergstrom, D.M., Convey, P., Huiskes, A.H.L. (eds.), Springer, Dordrecht 2006.

Lyons, W.B., Nezat, C.A., Welch, K.A., Kottmeier, S.T., and Doran, P.T.: Fossil fuel burning in Taylor Valley, southern Victoria Land, Antarctica: Estimating the role of scientific activities on carbon and nitrogen reservoirs and fluxes. Environ Sci & Technol 34:1659-1662. 2000.

Mackay, D., Shiu, W.Y., and Ma, K.C.: Handbook of physical-chemical properties and environmental fate for organic chemicals, vol. 2. polynuclear aromatic hydrocarbons, polychlorinated dioxins and debenzofurans. Lewis Publisher, London, UK, 1992.

Margesin, R., and Schinner, F.: Biological decontamination of oil spills in cold environments. J of Chem Technol and Biot 74:381-389, 1999.

Mcknight, D.M., Niyogi, D.K., Alger, A.S., Bomblies, A., and Conovitz, P.A.: Dry valley streams in Antarctica: Ecosystems waiting for water. Bioscience 49:985-995, 1999 .

MIL-DTL-83133E, Tubine fuels, aviation, kerosene wpe NATO F-34 (JP-8) NATO F-35 andJP-8+100. 1999.

58

MIL-PRF-5624S, Turbine fuel, aviation, grades JP-4, JP-5, and JP-5/JP-8 ST. 1996.

Parker, B.C., Simmons Jr., G.M., Wharton Jr., R.A., and Seaburg, K.G.: Removal of organic and inorganic matter from Antarctic lakes by aerial escape of blue-green algal mats. J. Phvcol. 18:72-78, 1982 .

Pawliszyn, J.: RSC Chromatography Monographs. In: Applications of Solid Phase Microextraction ed.R.M. Smith, pp. 655. Royal Society of Chemistry, Cambridge U.K. 1999.

Prince, R.C., Garrett, R.M., Bare, R.E., Grossman, M.J., Townsend, T., Suflita, J.M., Lee, K., Owens, E.H., Sergy, G.A., Braddock, J.F., Lindstrom, J.E., and Lessard, R.R.: The roles of photooxidation and biodegradation in long-term weathering of crude and heavy fuel oils. Spill Sci & Technol Bull 8:145-156. 2003 .

Priscu, J.C., Fritsen, C.H., Adams, E.E., Giovannoni, S.J., Paerl, H.W., McKay, C.P., Doran, P.T., Gordon, D.A., Lanoil, B.D., and Pinckney, J.L.: Perennial Antarctic lake ice: An oasis for life in a polar desert. Sci 280:2095-2098, 1998 .

Priscu JC. Life in the valley of the dead. Bioscience 49:12,9, 1999.

Roberts, E.C., Laybourn-Parry, J., McKnight, D.M., and Novarinos, G.: Stratification and dynamics of microbial loop communities in Lake Fryxell, Antarctica. Freshwater Biol 44:649-661, 2000.

Saul, D.J., Aislabie, J.M., Brown, C.E., Harris, L., and Foght, J.M.: Hydrocarbon contamination changes the bacterial diversity of soil from around Scott Base, Antarctica. FEMS Microbiol Ecol 53:141-155. 2005.

Snape, I., Harvey, P.M.A., Ferguson, S.H., Rayner, J.L., and Revill, A.T.: Investigation of evaporation and biodegradation of fuel spills in Antarctica-I. A chemical approach using GC-FID. Chemosphere 61:1485-1494. 2005.

Treonis, A.M., Wall, D.H., and Virginia, R.A.: Invertebrate biodiversity in Antarctic Dry Valley soils and sediments. Ecosystems 2:482-492, 1999.

Vincent, W.F.: Environmental management of a cold desert ecosystem: the McMurdo Dry Valleys, pp. 57. Desert Research Institute 1996.

Virginia, R.A., and Wall, D.H.: How soils structure communities in the Antarctic Dry Valleys. BioScience 49:973-983. 1999.

Wang, Z.D., Fingas, M„ Blenkinsopp, S., Sergy, G., Landriault, M., Sigouin, L., Foght, J., Semple, K., and Westlake, D.W.S.. Comparison of oil composition changes due to biodegradation and physical weathering in different oils. J of Chromatogr A 809:89-107, 1998.

59

Wharton, R.A., and Doran, P.T.: McMurdo Dry Valle Lakes: Impacts of Research Activities, pp. 54. University of Illinois at Chicago, Chicago 1999.

Zurcher, F., and Thuer, M: Rapid weathering processes of fuel oil in natural waters. Environ Sci & Technol 12:838-843, 1978.

60

Appendix II.

Supplementary Material

The appendix contain logs of ice core 1,2, 3, 4, 5, 6, 7, and 11 which are not

described in the manuscript. The core is followed by solid phase microextraction-gas

chromatography-mass spectrometry (SPME-GCMS) data when available. Core 6 is

described in detail.

Descriptions of ice cores use the terms ice straws, bubbles, and globules, which

are defined as thin vertical air pockets, hollow spherical or teardrop-shaped air pockets,

and granular ice, respectively.

ICE CORE 1

i i

y 10

120

130

|40 1 I50

y 80

|B0

Corel GENERAL DESCRIPTION

Tfftf -Ice character

|i 1 IIH'III

1o$> stra v 5 j i

|

1 l

i i

i :

1

j

j /

• i

i 1 i

'

|

i ,|

ll

1 l| J

1 '

oif bu~b cs , i p

h

!| '.1 i , !|

'

1 i i

i

.!

Sediment

^ - no oxM colored sands f A K i:c>m surface

s in."1 by?! #nd blmok s&nd on ' j ^ cere surface

hnu «jrk brown to black sand <1isoersec in ice

Mi angular granules

Dark b"OWi to black angular r mules

UV light absorption and reaction

Ice gipbulK at 78 cm fluoresced

Appendix II-1. Log of ice core 1

ICE CORE 2

If o f m a

1 J10

J 20

J 3 0

J 40

j 5 0

J BO

J 70

Core 2 GENERAL DESCRIPTION

Ice character T r c 1 » r I I f 1

Ul

I I ' l l I l l llil } litJIi

1 j ! 111.

>>

& I L l l

I I 1 ! inn HI

II |! !r mts

i

Jfraik

1

n

Sediment

Sands dispersed in ice

UV Sight absorption

and reaction Samples

dspth (cm) MM A 9

•«10-12


120,000

+ 80,000 -

40,000

4J

= 40,000

< 40,000 H

w-C„ « «-alkanes o isoprenoids

w-C, M

JP5-AN8

|-L|3

,xJLixA.lL.jiLl,....'-j?l ""'"

0-2cm

10- 11cm

Appendix II-3. Comparison of SPME-GC-MS total ion current for (A) JP5-AN8 diesel mixture and (B-C) ice core 2 samples. Internal standard used is trichloroethylene.

ICE CORE 3

I !

1 I 1 0

120

130

140 1 |so 3 I 6 0

I 7 0


Ice character

• uompaUed ice globules j

' H I |l|> llM

1111

' grades from i Tfanspaiwii

' I1

i 1 .

' i „ 1 ' 1 f , I

opaque to

1 1

1 <

i r<i II 1 1 III! '1

grades from transparent! 1 to opaque j

J j ,. 66 |[f(*fff •<•

L 1 II

' 11 lilt

Sediment

subangular, buff-colored and mafic sands

buff sands and red angular Granules

.

UV light absorption

and reaction

globules at 13 and 21 cm fluoresce


ICE CORE 4

10

20

30

40

50

60

70

80


Granular ice; Pa-wdi v opacify in. rMS«-> ijp ore

ILLLUJ 1 Granule to pebble-sized! sub-rounded ice chursKsj

Granule to pobbie-sized [ Oat ice chunks |

Coarse sand to granule I pa:ticec \mU

Sediment

Coarse sand composed of quartz and mane minerals

ine sand composed of mafics

UV light absorption

and reaction

Fluorescence legend: «™« intense

Diffused

Samples depth (cm) WM

• 13-14

• 29-30

• 38-40

• 52-53

• 87-68

• 80-81


fi'Xji,

* w-alkanes o isoprenoids

«-C7 ^'^o

I., ,:,!,:„« M.«,„i„-Jjfe,,.«*«»

a

• | M

o n

n

M n

»~Cn

i-C« o I-C14

Afa«*tihL..Lu

ANSA

n-CK e

,IS

0-2 cm B

13-14 cm C

39-40 cm D ....... I

55-56 cm E

iTS

.IS

67-68 cm F

80-81 cm G

Appendix II- 6-Comparison of SPME-GC-MS total ion current for (A) JP5-AN8 diesel mixture and (B-G) ice core 4 samples.

ICE CORE 5

Core 5 (melt area nsar melt pool)

Sediment UV light

absorption and reaction

Samples depth (cm)

10

20

30

40

50

60

Brittle and porous

fine sand to granule Oniy sand-sized sediments

• 3-4

• 11-12

«24-25

1 cm thick alternating layers of semi-opaque and transparent see

Thinly iavered

, Med to coarse anguiar sand; qte and mafscs

- Fine to med sand - Med to coarse sand FUjomscsmns

legend:

-™- intense

WM-. Diffused


120,000 -

f 80.000 -

40,000 -IS

t

• H-alkanes o isoprenoids

n-C, * * o

«• <

r M I 0 n

•C,

c 40,000 X3

< 40,000

JP5-AN8 A

11-12 cm C

10 20 30 40 Retention time (min)

50

Appendix II-8. Comparison of SPME-GC-MS total ion current for (A) JP5-AN8 diesel mixture and (B-C) ice core 5 samples. Internal standard used is trichloroethylene.

66

ICE CORE 6

Ice core 6 from "contaminated" site

The upper 67 cm of core 6 consists of at least five pairs of thin granular opaque

ice and thick transparent ice (respectively averaging 4 and 10 cm; Appendix III-9).

Within transparent ice layers, sediment patches between 7 and 9 cm and between 14 and

15 cm are surrounded by bubbles, some of which are filled with a yellowish fluid that

intensely fluoresces under UV light. The sediment grains are mostly angular, buff or

white quartz grains, similar to the aeolian sand typically observed on the ice cover and in

ice core 12. Interspersed grains of fine black, angular sand located within intergranular

ice spaces were also observed in opaque ice between 45 and 48 cm, as well as between 57

and 60 cm. Fluid-filled bubbles between 5 and 9 cm and a layer between 57 and 60 cm

emitted intense yellow fluorescence under UV light, indicating the presence of

diesel-derived contaminants.

SPME-GC-MS of samples at 0-1, 5-6.5, 6.5-8 and 10-11 cm indicate the presence

of diesel contamination (Appendix III-10B-E ). The sample from 0-1 cm only has trace

amount of diesel constituents with n-Cn and n-Cu in near equal abundance, and no

alkanes eluting below w-Cn, suggesting evaporative loss (Appendix III-10B). Sample 5-

6.5 cm has a mode at n-C^, no alkanes eluting before «-Cn;but W-CH is significantly less

abundant than n-Cu, indicating lesser evaporative losses than in the surface sample

(Appendix III-10C). In samples 6.5-8 cm and 10-11 cm, preferential loss of «-alkanes

relative to branched alkanes and the presence of UCM (Appendix III-10D and E) suggest

that biodegradation also plays a role in altering the fuel composition. In sample 45-46 cm

diesel fuel contaminants are present in very low abundance and consist mostly of PAHs

67

dominated by methyl-naphthalenes (Appendix III-1 OF). n-Cn is the most abundant

alkane, whereas branched-alkanes are present only in trace amounts. The presence of

alkanes, branched-alkanes, and isoprenoids precludes dissolution as the process

concentrating aromatics in this sample. Instead, loss of the LMW fraction by evaporation

and strong biodegradation of alkanes and some naphthalenes could result in the

distribution of compounds observed.

In core 6, all the samples exhibit relatively high level of evaporative loss of diesel

contaminants. Such a loss can only occur from contaminated waters that were exposed to

the atmosphere, most probably as a surface meltpool, as opposed to a subsurface

chamber. It is likely that ice samples 5-6.5, 6.5-8 and 10-11 cm are from the same

meltpool that froze as a single unit because the distribution of hydocarbons are similar,

consistent with these samples belonging to the same transparent layer in the core. The

intense fluorescence between 7 and 8 cm is probably due to the concentration of the fuel

around the sediment patches. Slight differences in composition of the diesel residue in

these samples are due to heterogeneous distribution of microbial activity associated with

the sediments. The preferential sorption of diesel contaminants to sediments and

associated organic phase prevents evaporation but favors biodegradation. In contrast, the

effects of evaporation on diesel weathering increase toward the surface of the core, with

the surface sample being the most affected.

ICE CORE 6

Core 6 GENERAL DESCP.IPT10N

50

60

67-

Ii4sLlilMJi2iSlMj

niiiiisiSiii

8ed;menl:

—877- f patch of sed on JbX- core surface

r W r _ patch of sed en -"•IS"1" core surface

Dispersed ^ i i ^ s mafic, fine sand

scattered mafic sand only on ice core surface

US?

UV light absorption

and reaction Sample

depth (cm)

Legend Intense fluorescence I

diffused fluorescence!


69

m

n-C, t„J,

n-alkanes

n-C9 • 1 °

ICE CORE 6

«;Ci, JP5-AN8 £ # Total Ion Current

•

^ L « J

X8 -

IS

I .

IS i

IS ,t

IS I

XR JLtm

10

...*...* U-tSiit*4^

i-Ci3 »"Cn

i-C

JJP

I

_ . . . , j

,f0 «-C,

L _ ^

M

j

Ml

! 0-1 cm B

i i ;

; 5-6.5 cm C

L_,j .

: 6.5-8 cm p

Uu '* ;

UCM:

Wiyt 10-11 cm g»

; : """* UCM

| 45-46 cm p

20 : 50 Rete

40 ntiou ti

5 me (mi

0 n)

60 70

Appendix 11-10. Comparison of SPME-GC-MS total ion current for (A) JP-5 AN8 diesel mixture and (B-F) ice core 6 samples.

ICE CORE 7

70

Sketch Ice character

Core 7

Sediment UV Sight


10

20

30

40

50

60

Layered ..;..LLLl..LUJ.iiJ.

Brittle _ _ 2 3 _ _

Laminated Brittle

Broken-UD straw;

Med to coarse angular sands Cotains qtz, mafics and iron oxide-coated grains

Patches of sediments

iMI I I I i i J ! ice chunks!

Med to coarse angular sands Cotains qtz, mafics and Iron oxide-coated grains


ICE CORE 11

(tore 11

Sediment UV light


Angular sands around the core

Mafic sands in the core

Mafic sands in the core

Fine sub-angular sands Contains mafic minerals and Iron oxide-coated grains

42 Green 44 fluorescent „ bubbles

Appendix 11-12. Log of ice core 11

71

O

y = 0.004x + 0.0677 / .

R2 = 0.89 X

>. ^ r

NX A > ^ , / ^ "s

y = -0.04x + 0.84 ^ s

R2 = 0.99 ° v v ^

5 10 15 20

- 1.0

0.8

0.6

- 0.4

0.2

0.0

CD

alka

i

_.

e/T

ota

c 0 CO

naph

th

Tot

al

15%-i

14%-

13%-

12%-

" 1 1 % -

4 10%" d 9%-

8%-

7%-

6%-

5%-0

number of days A tot naph/tot alk o i-C13/n-C13

Linear (tot naph/tot aik) Linear (i-Ci3/n-C13)

Appendix 11-13. Evaporation experiment trends for z-Ci3/«-Ci3

and Tnaph/Taik

CHAPTER 3

Composition and biodegradation of synthetic oils and hydraulic fluid spilled on the

perennial ice cover of Lake Fryxell, Antarctica

Caroline M.B. Jaraulaa, Fabien P. Keniga, Peter T. Dorana, John C. Priscub,

Kathleen A. Welch0

"Department of Earth and Environmental Sciences, University of Illinois at Chicago, 845 W Taylor St., Chicago, Illinois 60607

Department of Land Resources and Environmental Sciences, Montana State University, Bozeman, MT 59717

cByrd Polar Research Center, Ohio State University, Columbus, OH 43210-1002

72

73

Abstract

A helicopter crashed on the Lake Fryxell (Antarctica) perennial ice cover in January

2003, spilling synthetic transmission oil (Aeroshell 500), synthetic engine lubricant (Aeroshell

555), and hydraulic fluid (MIL-5605). Molecular compositions of the oils were analyzed by gas

chromatography-mass spectrometry and compared to the composition of contaminants in ice,

melt water, and sediment samples collected a year after the spill. Aeroshell 500 and 555, both

based on C20-C33 pentaerythritol triesters (PET) with C5-C10 fatty acids, differ in isomer

distribution and additive assemblage. Biodegradation of these oils in the ice cover occur when

sediments are present. PETs with short fatty acids are preferentially degraded, whereas long

chain fatty acids hinder the esters in PET from hydrolysis by esterase from the microbial

assemblage associated with sediments. It remains to be seen if the microbial ecosystem can

degrade tricresyl phosphate. These more recalcitrant PET species and tricresyl phosphates are

likely to persist and comprise the contaminants that may eventually cross the ice cover to reach

the pristine lake water. The hydraulic fluid is a petroleum cut dominated by mono- and bicyclic

terpenoids and diamondoids. Contamination by hydraulic fluid was not observed, which is likely

related to the small amount spilled.

Keywords: oil spill, natural attenuation, antioxidant additives, synthetic esters, lubricants

74

1.0 INTRODUCTION

Petroleum reserves in the Arctic, increasing tourism and scientific traffic in the Antarctic,

leaking tanks and oil transport on alpine areas and ski resorts, and spills occurring in the winter

increase the risk of hydrocarbon spills in cold environments (Margesin and Schinner, 1999). As

opposed to warmer environments, oils spilled in cold regions or during winter seasons are more

viscous, which will decrease distribution, dissolution, diffusion rates, and degradability, the latter

of which is caused by freezing of some of its components (Margesin and Schinner, 2001). In the

largest Antarctic ice-free region, the McMurdo Dry Valleys (MCM), there are increasing

concerns about hydrocarbon contamination of soils adjacent to the research bases (Vincent,

1996; Aislabie et al., 2004). The sites that were susceptible to hydrocarbon contamination

include landfills, helicopter pads, refueling areas, and along fuel lines (Kennicutt et al., 1992;

Gore et al., 1999; Saul et al., 2005).

On January 17, 2003, a Bell 212 helicopter crashed in MCM on the 5 m thick perennial

ice cover of Lake Fryxell (Taylor Valley) spilling -730 liters of aviation diesel fuel (JP5-AN8)

as well as a total of -27 liters of hydraulic fluid (MIL-5606), synthetic transmission fluid

(Aeroshell 555), and synthetic engine lubricant (Aeroshell 500). Cleanup efforts, initiated four

days after the accident, recovered no more than 45% of the spilled fluids (Alexander and

Stockton, 2003). Environmental impacts from such spills is minimized by immediate cleanup,

but this is possible only early in the field season, from September to early November, while the

ice is solidly frozen. Later in the Austral summer, spilled fluids can penetrate the ice cover via

melt pools and are difficult to remove or contain (Jepsen et al., 2006). The crash occurred during

a warm period when the lake ice was isothermal near the melting point, and the ice was partly

covered by large melt pools, the worst possible scenario for a successful clean-up.

75

The few studies of accidental hydrocarbon releases and their impacts in the MCM focus

only on the effects on soil biogeochemistry as most incidents occurred on land (Aislabie et al.,

2004; Lyons et al., 2000). Conclusions derived from these soil-based studies indicate that

hydrocarbon contaminants in cold desert environments can persist for many years and

profoundly affect microbial diversity and geochemistry. However, results of those studies cannot

be directly extrapolated to the ice cover, which has significantly different physical, chemical and

biological characteristics.

Composition and natural attenuation of the spilled aviation diesel fuel are discussed in

Jaraula et al. (Jaraula et al., in press). Here, we compare the initial chemical composition of the

hydraulic fluid, synthetic transmission fluid, and synthetic engine lubricant spilled on Lake

Fryxell ice with that of residues in the ice cover a year later. We also discuss the potential for the

spilled oils to reach the lake water by penetrating the 5 m ice cover.

Lubricants are formulated from mineral or synthetic oils. Whereas mineral oils are

readily available and affordable, synthetic oils have superior properties such as increased thermal

stability, large temperature range of application, low volatility at high temperature, and less

susceptible to oxidation, to name a few (Rudnick and Bartz, 2005). These advantages led to

widespread adoption of synthetic oils by the lubricant industry so that worldwide demand for

synthetic oils is estimated to increase 5-7% annually since the 1980's (Whitby and Williamson,

1999). Lubricants are now regulated and tested for toxicity, potential for bioaccumulation, and

biodegradability (Bartz, 1998). Laboratory and field investigations show that synthetic oils are

more extensively and more rapidly biodegraded compared to mineral oils (Bartz, 1998;

Eisentraeger et al., 2002; Haigh, 1995; Pettersson, 2007). Synthetic esters, in particular, are more

stable to thermal and oxidative degradation from the addition of stronger carbon-oxygen bond

76

compared to the carbon-carbon linkage (Randies, 1999). Among the synthetic esters, exceptional

thermal and oxidative stability of pentaerythritol esters made these the base of choice for military

jet fighter lubricants. Synthesis and production of these oils started in Germany during the World

War II shortage of mineral oils (Randies, 1999). Ester-base lubricants also worked so well under

low temperature conditions that development of these fluids after the war was closely associated

with advances in aviation gas turbine technology (Randies, 1999).

2.0 STUDY AREA

MCM being the coldest and driest desert on our planet is the site of a long-term

ecological research program (http ://www.mcmlter. org/). The climate, limited nutrients, and slow-

growing biological communities make the MCM ecosystem extremely sensitive to climatic

changes and human impact (Cowan and Tow, 2004; Lyons et al., 2006). At the site of the

helicopter crash (77°36'41.098" S, 163°06'47.228" E) the ice cover was 5.3 m thick (Alexander

and Stockton, 2003). Average annual ice-cover thickness is a balance between -30 cm y"1 of

freezing of lake water at the bottom and 15 to 60 cm y" of ablation loss at the surface via

sublimation. Together, these processes, in combination with a fully floating ice cover, generate

-30 cm y"1 of net upward movement of the ice. Aeolian sand embedded in or covering the ice

has a lower heat capacity than the surrounding ice (Adams et al., 1998), inducing the formation

of melt pools as wide as 1.5 m during the austral summer. Thermistors deployed nearby to

monitor the temperature of the ice cover indicated that ten days prior to the crash, maximum

daytime temperature was 2-3 °C warmer than the average summer temperature of -3 °C

(Margesin and Schinner, 1999), supporting the presence of large melt pools.

http://www.mcmlter

77

In the subsurface, the sediments melt into the ice and leave trails of liquid water in their

wakes (Adams et al., 1998). Some of the sediments collect in pockets and accumulate in the

summer at the bottom of cavities in the porous ice. Fritsen et al. (1998) referred to a layer of

sediment at -0.5 m depth in the Lake Fryxell ice as an "aquifer" because it contains liquid water

from November to February. At the height of summer, the volume of liquid water in the ice

cover can be as much as 40% (Adams et al., 1998).

3.0 METHODS

Gas chromatography-mass spectrometry (GC-MS) was performed with a HP-6890 GC

coupled to a HP-5973 Mass Selective Detector used in electron ionization (70 eV) with helium

as a carrier gas. The mass range (40 to 650 amu) was scanned every three seconds. The capillary

GC column used was a HP-5MS (30 m, 0.25 mm I.D., 0.25 um film).

Raytheon Polar Services provided samples of Aeroshell 500, Aeroshell 555, and MIL-

5605 that were from the same stock of oils as used on the helicopter. Aeroshell 500 and

Aeroshell 555 were diluted in cyclohexane (100 uL/mg) and 0.5 ul was injected in splitless

mode. The GC oven was programmed to increase in temperature from 60 °C (1.5 minutes) to 130

°C at 20 °C/min, and then at 4 °C/min to 300 °C (52.5 min).

An aliquot of 0.25 uL of the MIL-5606 hydraulic fluid was injected without dilution or

processing as in 98:2 split mode. The GC oven was programmed to increase in temperature from

40 °C (10 minutes) to 200 °C at 2 °C/min and then at 10 °C/min to 320 °C (20 min).

Water and sediment samples were retrieved from melt pool sites 20 (~10 m northwest of

the crash site), and 38 (~7 m northeast of the site). Non-volatile hydrocarbons were extracted

from 2000 uL samples used for solid-phase microextraction (SPME; Jaraula et al., in press). The

78

remaining fluid was then extracted with 1 ml of hexane and 1 ml dichloromethane three times for

each solvent. The extracts were dried under N2, weighed, diluted with cyclohexane (100 uL mg"

') and injected twice into the GC-MS using the two oven temperature programs listed above.

An ice core recovered ~7 m south of the crash site contained fluid-filled bubbles that

fluoresced in ultraviolet light. A syringe was used to recover 15 uL of yellowish fluid from one

of the bubbles. From this fluid, 1.0 uL was dissolved in 1999 uL ultrapure water (Jaraula et al.,

impress). After extracting a fraction of the volatiles via SPME, the combined fluid was extracted

with hexane and dichloromethane and the extract was injected in the GC-MS twice as described

above.

4.0 RESULTS AND DISCUSSION

4.1 Synthetic oil Aeroshell 500 and 555

The synthetic lubricant Aeroshell 500 (1.083 g ml"1 at 4° C), under the U.S. military

specification MIL-PRF-23699F STD, is used in gas turbine engines. This lubricant contains C20

to C33 pentarerythritol triesters (PET) of C5 to C10 fatty acids (MW 388 to 556 amu; Fig. 1A), C20

being the most abundant, and minor amounts of didentaerythritol hexaesters.

Additives include the antioxidants phenantrene (and traces of methyl- and dimethyl-

phenantrenes), N-phenyl-2-naphthylamine, tricresyl phosphate isomers (ortho, meta, and para-

cresyl), and Z)w-(4-octylphenyl)-amine. The synthetic oil Aeroshell 555 (0.994 g ml"1 at 4° C),

under the U.S. military specification MIL-PRF-23699F STD, is used for high temperature, high

load applications including Bell helicopter transmissions. To increase thermal stability, Aeroshell

555 contains less of the C20 to C24 PET than Aeroshell 500. Instead, Aeroshell 555 has a high

79

c

A Aeroshell 500

Tricresyl 0 C v \ Pentaerythritol triesters

phosphates > ' v^ -

N-Phenyl-2-Naphthylamine

H Phenantrene CCr X5

is

R,R'orR" = C8toC10)

R> /

\ / OH

C , \ Didentaerythritol I , \ hexaesters

B Aeroshell 555

H

1_| p A A / W " ^ - v i i <-« n i 1 ^ 5 r111(-'5

*bis(4-octylphenyl)-amine

Tricresy phosphates

c, /

28

c /

30

Didentaerythritol hexaesters

10 20 30 40 50 Retention time (min)

60

Figure 1. Total ion current (TIC) trace of (A) turbine oil Aeroshell 500 and (B) transmission oil Aeroshell 555. Total carbon number is indicated above the pentaerythritol triester peaks. IS: 3,3-diethylheptadecane.

80

relative abundance of C25 to C32 PET, C26 being the most abundant (Fig. IB). Didentaerythritol

esters are less abundant than in Aeroshell 500. Antioxidant additives in Aeroshell 555 have

higher boiling points than in Aeroshell 500, with only tricresyl phosphate isomers and bis-(4-

octylphenyl)-amine, which are specific for high temperature applications.

4.2 Hydraulic fluid MIL-5605

The hydraulic fluid MIL-5605 (0.897 g ml"1 at 4° C) is a petroleum distillation cut

dominated by mono- and bicyclic terpanes as well as diamondoids, including adamantanes and

the mono-, di-, and tri-methylated adamantanes (Fig. 2A and B). The most abundant compound

in MIL-5605 is the aromatic antioxidant additive butylated hydroxytoluene (BHT).

4.3 Degradation potential of spilled contaminants

Both Aeroshell synthetic lubricants consist entirely of high molecular weight (HMW)

compounds (>388 amu) and will not evaporate unless heated. Photodegradation is not expected

because pentaerythritol esters do not absorb radiation in the UV and VIS region, but laboratory

experiments have shown that synthetic tri- and tetraester-based lubricants are biodegradable ~20

°C (23-25). These studies concur in indicating that in the presence of carboxyl esterase,

pentaerythritol esters will release easily degradable fatty acids and pentaerythritol (Fig. 1A), a C5

compound with a quaternary carbon and four hydroxyl groups, which are water-soluble but

probably poorly biodegradable (25,26). Carboxyl esterase is generally involved in the primary

degradation of organic matter (Bornscheuer, 2002). Thus, esterase is a relatively common

enzyme in microorganisms that preferentially hydrolyzes triglycerides with fatty acids shorter

than C6. The longer fatty acids probably impose steric hindrance of the carboxyl group

81

TOTAL ION CURRENT

Adamantanes


Figure 2. (A) Total ion current trace of the hydraulic fluid MIL-5605 and (B) partial summed mass chromatogram showing the distribution of adamantanes. BHT: butylated hydroxytoluene.

82

(Bornscheuer, 2002). Hence, the length of the fatty acids in PET should partially control the

biodegradation potential of Aeroshell synthetic ester oils via esterase. PET bearing two or three

C5 fatty acids probably will be much more easily biodegraded than those bearing longer chain

ones, resulting in the biodegradation of the lower molecular weight PET. The carbon number of

the three fatty acid chains of PET in Aeroshell 500 and 555 was determined from mass spectral

fragmentation and listed in column above the peaks, irrespective of their position as R, R' or R"

(Fig. 3A and B).

A synthetic lubricating oil, containing aliphatic esters (up to C34), similar to Aeroshell

500 and Aeroshell 555, was still present in sediments ten years after a spill at a helipad in Bunger

Hills, East Antarctica (Gore et al., 1999). This oil is a dense non-aqueous phase liquid (DNAPL),

and so concentrations were higher in the subsurface sediments than at the surface. Even in the

surface sediments, there was no evidence for biodegradation. In absence of biological activity,

these hindered esters will persist. Thus, their degradation in the ice cover of Lake Fryxell should

be limited to sediments, ice layers, and melt pools where biological activity, with esterase

capacity, may occur during the summer melt season.

The additives phenanthrene and methylated phenantrenes are more susceptible to

photodegradation and biodegradation than evaporation (Prince et al., 2003). Tricresyl phosphates

are not volatile, are relatively insoluble in water, and have low hydrolysis potential, although

they are rapidly biodegraded or ingested by living organisms (Saeger et al., 1979; Wageman et

al., 1974). N-phenyl-1-naphthylamine, similar to N-phenyl-2-naphthylamine, is lipophilic,

therefore it can accumulate preferentially in cell membranes and disrupt photosynthetic reactions

(Altenburger et al., 2006). N-phenyl-1-naphthylamine is biodegradable in sewer and lake water

83

Aeroshell 500 A

24 25 26 27 28 29

Total cnrbon number ol'lhe Iriesters

Figure 3. Partial total ion current (TIC) of pentaerythritol triesters showing the carbon number of the fatty acid substituents of triesters in (A) Aeroshell 500, (B) Aeroshell 555, (C) Site 38 sediments and (D) Core 12 4 cm bubble fluid. The total carbon numbers of the triesters, on the x axis, correspond to the sum of carbon atoms in the acid substituents plus five carbon atoms from pentaerythritol. The carbon number of the three fatty acid chains of PETs are listed in column above the peaks. * indicates the antioxidant additive bis(4-octylphenyl)-amine.

84

although at slow rates (Sikka et al , 1981). This additive does not hydrolyze, but sorbs onto

sediment and organic materials and its ultraviolet absorption spectrum indicates that it should

degrade photochemically in air. No data are available on the biodegradation of bis(4-

octylphenyl)amine in aquatic systems (Kronimus et al., 2004), but this compound is extremely

hydrophobic and thus unlikely to be easily biodegraded.

The hydraulic fluid is a hydrophobic light non-aqueous phase liquid (LNAPL) and its

volatile components are susceptible to evaporative loss. The diamondoids have low boiling

points and can evaporate, and adamantanes are biodegradable (Wang et al., 2006), whereas

bicyclic terpenoids and diamantanes resist biodegradation more strongly. It can be hypothesized

that although the amount of hydraulic fluid spilled is minor, the potential of bicyclic terpenoids

and diamantanes to persist in Lake Fryxell ice is relatively high. The additive BHT

photodegrades easily in the presence of oxygen; thus, neither BHT nor its degradation products

are persistent (MIkami et al., 1979a; Mikami et al., 1979b).

4.4 GC-MS analysis of field samples

The extract of melt pool waters from site 38 contains C20 to C32 PET, in which C23 and

C26 triesters are the most abundant (Fig. 4A). The distribution of the triesters and their isomers

are consistent with a mixture of Aeroshell 500 and Aeroshell 555.

Melt pool sediments in site 38 contains rc-alkanes from C12 to C22 with a mode at n-Cis (Fig. 4B)

derived from aviation diesel contamination (Jaraula et al., in press). The extract also contains

PET, in which the distribution is neither similar to that of Aeroshell 500 nor that of Aeroshell

555 (Fig 1A and B). The C20-C25 triesters are significantly less abundant than the C26 triesters,

85

Site 38 (water)

O c

-a c 3

XI

<

Tert-butylphenyl diphenyl

phosphate

I.S. n-C,

M*Jti

n-C,

n-C

»lAl*Hl?J,>s,iXjl.,J'

cj

W c22

Site 38 R

(black D

sediments)

C,

c

uu \UJZ^ r c C I I I I < AJLJWWTT^

Site 20 Q (water)

n-C |

• J

«-C,

Tricresyl phosphates,

n-C, n-C20 n-C„ MdUU L u l l * ^ ^

Site 20 r (sed intents)

Pentaerythritol esters

X8

*j

20 30 40


i1 | • . . ' t | « i ' i" • •

50 60

Figure 4. Total ion current trace of melt pool waters and sediments showing the distribution of contaminants derived from Aeroshell 500, Aeroshell 555 and aviation diesel. *residue of I.S., n-C24, from a previous injection.

86

indicating preferential loss of low molecular weight PETs. This loss is not due to evaporation nor

photooxidation as these triesters are not susceptible to such processes. In detail, PET isomers

with C5 and C6 acids are preferentially depleted (Fig. 3C) relative to the original isomers in

Aeroshell 500 and 555 (Fig. 3A and B). For example, the C20 PET containing three C5 fatty acids

(C5-C5-C5) is now present only in trace amounts. Similarly, among the C24 triesters, the C5-C5-C9

and C5-C6-C8 are more depleted than C5-C7-C7. This selective loss of isomers with shorter fatty

acid chain length is likely due to degradation via esterase activity.

Among the antioxidant additives, only the insoluble tricresyl phosphates were detected

and were present only in the sediments, as expected (Fig. 4B and D). Their low abundances

relative to high molecular weight PET and the relative abundances of the three isomers are

similar to those of the Aeroshell oils, suggesting that these compounds do not readily degrade in

the ice cover. Biodegradation usually changes the relative abundances of the three tricresyl

phosphates isomers, which have differing levels of toxicity to organisms (Wong and Chou,

1984). Tert-butylphenyl diphenyl phosphate, which is used as a flame retardant and plasticizer in

lubricants (Heitkamp et al., 1985), elutes just after the tricresyl phosphates (Fig. 4B). This

compound was not initially detected in the Aeroshell oils due to its low concentration and perfect

co-elution with the C20 triester. This additive can be metabolized by fungi (Heitkamp et al., 1985;

Heitkamp et al., 1986), which are probably present in the ice cover, as these organisms are

common in cryoconite holes (Wharton et al., 1985), cylindrical water-filled melt holes on glacier

surfaces. Thus, this additive is not expected to persist in the ice cover where sediments are

present. Bis-(4-octylphenyl)amine and N-phenyl-2-naphthylamine, major antioxidants in the

87

Aeroshell oils, were not detected in the samples that contained Aeroshell residues. This suggests

that both additives were weathered, although by-products of their biodegradation were not

observed.

The extract of melt pool water from site 20 contains only trace amounts of C20 to C30 PET

(Fig. 4C), concordant with the low solubility of these compounds and further distance of site 20

from the crash site. PETs are present in a distribution similar to that observed for the water

sample collected in the melt pool at site 38.

Melt pool sediments from site 20 are dominated by «-alkanes (C12 to C23; Fig. 4D)

derived from diesel contamination (Jaraula et al., in press). PETs are present in low abundance

relative to diesel and are dominated by C24 isomers. Loss of the low molecular weight PETs in

sediment samples of site 38 and 20 suggests that esterase is active in sediments. In contrast, the

presence of C20 PET in melt pool waters suggests a low level of biodegradation in melt water.

This observation is consistent with the observed biodegradation of diesel fuel spilled on the ice

cover, which occurred exclusively in sediment bearing layers of the ice cover and melt pool

sediments (Jaraula et al., in press).

Among the antioxidant additives, only tricresyl phosphates were detected at site 20.

These additives are present only in the sediments and the three isomers have similar relative

abundances than in the Aeroshell oils, indicating that tricresyl phosphates are not easily

biodegraded in the ice cover. The high abundance of tricresyl phophates relative to low

molecular weight PETs in both sediment samples, in contrast to what is observed in the

Aeroshell oils (Fig. 1A and B), is concordant with extensive loss of PETs via biodegradation.

88

Fluids recovered from a bubble in an ice core also contain residues of the synthetic oils

(Fig. 3D). Triester isomers containing C5 and C6 are also preferentially lost, although to a lesser

extent than in the sediments of site 38 (Fig. 3C). Biodegradation in this fluid-filled bubble is also

supported by the presence of naphthalenones (Jaraula et al., in press) that are known by-products

of naphthalene biodegradation (Langbehn and Steinhart, 1995). This oil-diesel-water inclusion,

probably remained fluid for prolonged period of time and may have served as an oasis for life in

the ice cover (Priscu et al., 1998).

The hydraulic fluid MIL-5605 was not detected in any of the samples. Thus, our

prediction of its natural attenuation potential, made on the basis of molecular composition, as

well as preferential evaporation, and biodegradation of certain compound families could not be

verified in the field.

4.5 Lateral and vertical transport in the ice cover

Aeroshell 500 and Aeroshell 555 behave as DNAPLs and, as a result, occur more

abundantly at the bottom of melt pools, usually in association with sediments (Fig. 4B and D). A

small fraction of the PETs dissolve in the melt pool waters (Fig. 4A and C) and can be

transported laterally in the summer when the ice is porous and an aquifer develops in the ice

(Fritsen and Priscu, 1998). The bulk of the Aeroshell oil residue, however, and the recalcitrant

additives tricresyl phosphates preferentially sorb onto sediments. This association with sediments

favors natural attenuation because biodegradation occurs mostly in sediment layers. DNAPLs

that sorb onto sediments and are not biodegraded may melt their way into the aquifer in the ice.

The net upward movement of the ice cover will limit the descent of DNAPLs across the ice

89

cover into the water column. However, discrete conduits in the ice that connect the aquifer to the

lake water (Priscu et al., 2002), may hasten the transfer of sorbed contaminants into the pristine

lake waters.

5.0 CONCLUSION

A year after the crash of a helicopter on the perennial ice of Lake Fryxell, residues of

engine lubricant Aeroshell 555 and transmission oil Aeroshell 500 were detected in the

sediments and melt pool waters, in contrast to the hydraulic fluid, which was not detected.

Generally, behaving as DNAPLs, the synthetic oils are mostly associated with sediments,

although a fraction is dissolved in melt pool waters. In sediments, pentaerythritol triesters with

C5 and Ce acids are more susceptible to biodegradation. The more recalcitrant species of

pentaerythritol triesters and antioxidant additives tricresyl phosphates are likely to persist and

comprise the contaminants that may eventually cross the ice cover to reach the pristine lake

water. Pentaerythritol triesters will breakdown into fatty acids and pentaerythritol, which is

soluble but non-biodegradable. It remains to be seen if the microbial ecosystem in Lake Fryxell

ice cover can degrade tricresyl phosphate.

90

ACKNOWLEDGEMENT

This project was supported by the Office of Polar Programs of the National Science

Foundation (Antarctic Biology and Medicine SGER 0346316 to F. Kenig and P. Doran) and by a

fellowship grant from the University of Illinois at Chicago Institute of Environmental Science

and Policy to C. Jaraula. We thank Kelvin Rodolfo and Neil Sturchio for their suggestions on the

manuscript. We also thank Apostolis Sambanis, Marcus Muccianti, Timothy Chung, and Alice

Hilegass for their help in the laboratory.

REFERENCES CITED

Adams, E.E.; Priscu, J.C.; Fritsen, C.H.; Smith, S.R.; and Brackman, S.L.: Ecosystem dynamics in a polar desert; the McMurdo dry valleys, Antarctica. Priscu, J. C , Ed.; American Geophysical Union Vol. 72,:pp 281-295, 1998.

Aislabie, J.; Balks, M.R.; Foght, J.M.; and Waterhouse, E.J.: Hydrocarbon Spills on Antarctic Soils: Effects and Management Environmental. Science & Technology 38: 1265-1274, 2004.

Alexander, S.P.; and Stockton, W.L.: "Lake Fryxell 79U Crash Site: Initial Hydrocarbon Monitoring Program.," Raytheon Technical Services Company, Polar Services Company, 2003.

Altenburger, R., Brack,W.; Greco, W.R.; Grote,M.; Jung,K.; Ovari, A.; Riedl, J.; Schwab, K.; and Kuster, E.: On the mode of action of N-phenyl-2-naphthylamine in plants. Environmental Science & Technology 40: 6163-6169, 2006.

Amund,0.0.: Utilization and degradation of an ester-based synthetic lubricant by Acinetobacter lwoffi Biodegradation 7: 91-95, 1996.

Bartz, W. J. Lubricants and the environment. Tribology International 31, 35-47, 1998,.

Bornscheuer,U.T.: Microbial carboxyl esterases: classification, properties and application in biocatalysis. Ferns Microbiology Reviews 26, 73-81, 2002.

91

Cowan, D. A., and Tow, L. A.: Endangered Antarctic environments doi:10.1146/annurev.micro.57.030502.090811 Annual Review of Microbiology 58:649-690, 2004.

Doran, P.T., Gordon, D.A., Lanoil, B. D., and Pinckney, J. L.: Perennial Antarctic lake ice: An oasis for life in a polar desert. Science 280: 2095-2098,1998.

Eisentraeger, A., Schmidt, M., Murrenhoff, H., Dott, W., and Hahn, S.: Biodegradability testing of synthetic ester lubricants - effects of additives and usage. Chemosphere 48: 89-96, 2002.

Fritsen, C.H., and Priscu, J.C.: Cyanobacterial assemblages in permanent ice covers on Antarctic lakes: Distribution growth rate, and temperature response of photosynthesis. Journal of Phycology 34: 587-597, 1998.

Gore, D. B., Revill, A. T., and Guille, D.: Petroleum hydrocarbons ten years after spillage at a helipad in Bunger Hills, East Antarctica. Antarctic Science 11, 427-429, 1999.

Haigh, S. D.: Fate and effects of synthetic lubricants in soil: biodegradation and effect on crops in field studies. Science of the Total Environment 168: 71-83, 1995.

Heitkamp, M. A., Freeman, J. P., and Cerniglia, C. E.: Biodegradation of ter/-Butylphenyl Diphenyl Phosphate. Applied and Environmental Microbiology 51: 316-322, 1986.

Heitkamp, M. A., Freeman, J. P., McMillan, D. C, and Cerniglia, C. E.: Fungal Metabolism of fer/-Butylphenyl Diphenyl Phosphate. Applied and Environmental Microbiology 50: 265-273, 1985.

Jaraula, C. M. B., Kenig, F., Doran, P., Priscu, J. C , and Welch, K.: SPME-GCMS study of the natural attenuation of aviation diesel spilled on the perennial ice cover of Lake Fryxell. Antarctica. Sci. of the Total Environment, In press.

Jepsen, S. M.,Adams, E. E., and Priscu, J. C : Fuel movement along grain boundaries in ice. Cold Regions Science And Technology 45: 158-165,2006.

Kenig, F., Simons, D. J. H., Crich, D.,Cowen, J. P., Ventura, G. T., and Rehbein-Khalily, T.: Structure and distribution of branched aliphatic alkanes with quaternary carbon atoms in Cenomanian and Turonian black shales of Pasquia Hills (Saskatchewan, Canada). Organic Geochemistry 36:117-138. 2005.

Kennicutt, M . C , Denoux, G. J., and McDonald, S. J.: Hydrocarbon Contamination on the Antarctic Peninsula. Marine Pollution Bulletin 24: 499-506, 1992.

Kronimus, A., Schwarzbauer, J., Dsikowitzky, S. H., and Littke, R.: Anthropogenic organic contaminants in sediments of the Lippe river, Germany. Water research 38:3473-3484, 2004.

92

Langbehn, A., and Steinhart, F.: Biodegradation studies of hydrocarbons in soils by analyzing metabolites formed. Chemosphere 30: 855-868, 1995.

Lyons, W. B., Laybourn-Parry, J., Welch, K. A., and Priscu, J. C : Trends in Antarctic terrestrial and limnetic ecosystems: Antarctica as a global indicator eds.Bergstrom, D. M., Convey, P., Huiskes, A. H. L.,; Springer Dordrecht, 2006.

Lyons, W. B., Nezat, C. A., Welch, K. A., Kottmeier, S. T., and Doran, P. T.: Fossil fuel burning in Taylor Valley, southern Victoria Land, Antarctica: Estimating the role of scientific activities on carbon and nitrogen reservoirs and fluxes. Environmental Science & Technology 34: 1659-1662, 2000.

Margesin, R., and Schinner, F.: Biological decontamination of oil spills in cold environments. Journal of Chemical Technology and Biotechnology 74:381-389, 1999.

Margesin, R., and Schinner, F.: Biodegradation and biormeediation of hydrocarbons in extreme environments. Appl Microbiol Biotechnol 56: 650-663, 2001.

Mikami, N., Gomi, H., and Miyamoto, J.: Studies on degradation of 2,6-di-tert-butyl-4-methylphenol (BHT) in the environment. Part-I degradation of 14C-BHT in soil Chemosphere 5: 305-310, 1979.

Mikami, N., Gomi, H., and Miyamoto, J.: Studies on degradation of 2,6-di-tert-butyl-4-methylphenol (BHT) in the environment. Part-II degradation of 14C-BHT in water Chemosphere 5: 311-315. 1979.

Pettersson, A.: High-performance base fluids for environmentally adapted lubricants. Tribology International 40: 638-645, 2007.

Prince, R. C , Garrett, R. M., Bare, R. E., Grossman, M. J.,; Townsend, T., Suflita, J. M., Lee, K., Owens, E. H., Sergy, G. A., Braddock, J. F., Lindstrom, J. E., and Lessard, R. R.: The roles of photooxidation and biodegradation in long-term weathering of crude and heavy fuel oils. Spill Science & Technology Bulletin 8: 145-156, 2003.

Priscu, J. C , Fritsen, C. H., Adams, E. E., Giovannoni, S. J., Paerl, H. W., McKay, C. P., Priscu, J. C , Adams, E. E.,; Paerl, H. W., Fritsen, C. H.,; Dore, J. E., Lisle, J. T., Wolf, C. F., and Mikucki, J.: Perennial Antarctic lake ice: a refuge for cyanobacteria in an extreme environment. Princeton University Press, 2002.

Randies, S. J.: Esters. Marcel Dekker, Inc., pp 807-840. New York, New York, USA, 1999.

Rudnick, L. R., and Bartz, W. J.: Synthetics, mineral oils, and bio-based lubricants: chemistry and technology. Rudnick, L. R., Ed.; CRC Press: Vol. I l l : pp. 331-350, Wilmington, Delaware, USA, 2005.

Saeger, V. W., Hicks, O., Kaley, R. G., Michael, P. R., Mieure, J. P., and Tucker, E. S.: Environmental Fate of Selected Phosphate Esters. Environmental Science & Technology 13: 840-844, 1979.

93

Saul, D. J., Aislabie, J. M., Brown, C. E.;, Harris, L., and Foght, J. M.: Hydrocarbon contamination changes the bacterial diversity of soil from around Scott Base. Antarctica Ferns Microbiology Ecology 53: 141-155, 2005.

Sikka, H. C , Pack, E. J., Sugatt, R. H., Banerjee, S., and Rosenberg, A.: Environmental Fate and Effects of N-phenyl-1-naphthylamine and its disposition and Metabolism in the Rat. Syracuse Research Company 106 pp., 1981.

Sonderkamp, S., Vomberg, A., Schmitz, C , Fassbender, U., and Klinner, U.: Interactions between bacterial populations during degradation of a lubricant base oil. Ferns Microbiology Ecology 38: 97-104. 2001. .

Vincent, W. F., ed.: Environmental Management of a Cold Desert Ecosystem: the McMurdo Dry Valleys. Desert Research Institute, 1996.

Wharton, R. A., Jr., McKay, C. P., Parker, B. C , and Simmons, G. M., Jr.: Cryonite holes on glaciers. BioScience 35: 499-503, 1985.

Whitby, R. D., and Williamson, E. I.: Synthetic lubricants and high-performance functional fluids. Rudnick, L. R., Shubkin, R. L., eds.; pp 807-840, Marcel Dekker, Inc.: New York, New York, USA, 1999.

Wong, P. T. S., and Chau, Y. K.: Structure-Toxicity of Triaryl Phosphates in Freshwater. The Science of the Total Environment 32: 157-165, 1984.

Wright, M. A., Taylor, F., Randies, S. J., Brown, D. E., and Higgins, I. J.: Biodegradation of a Synthetic Lubricant by Micrococcus-Roseus. Applied and Environmental Microbiology 59: 1072-1076, 1993.

Wagemann, R., Graham, B., and Lockhart, W. L.: Studies on Chemical Degradation and Fish Toxicity of a Synthetic Tri-aryl Phosphate Lubricating Oil, IMOL S-140. Fish.Mar. Serv. Res. Dev. Tech. Rep. 480: 30pp., 1974.

Wang, Z. D., Yang, C , Hollebone, B., and Fingas, M.: Forensic fingerprinting of diamondoids for correlation and differentiation of spilled oil and petroleum products. Environmental Science & Technology 40: 5636-5646, 2006.

CHAPTER 4

Tentative identification of pentaunsaturated alkenones

from Lake Fryxell, East Antarctica

Caroline M.B. Jaraulaa'*, Simon C. Brassellb, Rachael M. Morgan-Kiss c,

Peter T. Dorana, Fabien Keniga

^Department of Earth and Environmental Sciences, University of Illinois at Chicago, 845 W Taylor St., Chicago IL 60607-7059, USA.

Department of Geological Sciences, Indiana University, 1001 East 10th Street, Bloomington, IN47405-1405, USA

cDepartment of Microbiology, Miami University, 700 E. High St., Oxford Ohio 45056

submitted to Organic Geochemistry

94

95

Abstract

Chromatographic and mass spectrometric analysis of bottom sediments from perennially

ice-covered Lake Fryxell in the Dry Valleys of East Antarctica has allowed tentative

identification of previously unrecognized pentaunsaturated alkenones, octatriaconta-pentaen-2-

one (C38:5Me), nonatriaconta-pentaen-2 and 3-one (C39:5Me and C39:5Et), as well as tetradec-

pentaen-2-one (Gio^Me). These compounds comprise 49%, 29% and 12% of the total C38, C39

and C40 alkenones, respectively, which also include di-, tri-, and tetra-unsaturated ketones.

Analysis of environmental DNA extracted from the water column produced two 18S rDNA

sequences that were closely related to Isochrysis galbana as well as haptophyte sequences

recovered from sediment samples collected from Greenland lakes and Ace Lake, Antarctica.

Based on the alkenone distribution as well as the phylogenetic results, we conclude that

Isochrysis spp. are the likely source organisms biosynthesizing the novel alkenones. The suite of

alkenones produced may be a response to extremely cold temperatures year-round (~ 1 °C

average) and oxygen supersaturation, as well as phosphate- and light-limited conditions in the

upper water column.

Keywords: U 37, Isochrysis sp., light limitation, nutrient limitation, epibrassicasterol, Dry Valleys, haptophytes

96

1.0 INTRODUCTION

The degree of unsaturation in long chain (C37 to C40) alkenones (LCAs) decreases with

growth temperature in laboratory cultures of haptophyceae (Marlowe, 1984; Harwood and

Russell, 1984; Brassell et al., 1986), sediments (Brassell et al., 1981; Brassell and Eglinton,

1984) and water column samples (Prahl and Wakeham, 1987). Brassell et al. (1986) formulated

an alkenone unsaturation index: Uk37 = (37:2-37:4)/(37:2+37:3+37:4). Prahl and Wakeham

(1987) proposed another alkenone index, Uk'37= (37:2)/(37:2+37:3+37:4), because

tetraunsaturated alkenones are less common in marine environments than their tri- and

diunsaturated counterparts. To estimate sea surface temperature, calibration equations from

cultures and field samples specific to prymnesiophyte strains or localities are available, although

the standard equation used in the marine environment is UK 37 = 0.034T+0.039 (Prahl and

Wakeham, 1987; Prahl et al., 1988). Both alkenone unsaturation indices have become powerful

tools for the reconstruction of sea surface paleotemperature of oceans and lakes (e.g. Eglinton et

al., 2001; Schouten et al., 2002).

Boon et al. (1978) first reported sedimentary long chain alkenones. De Leeuw et al.

(1980) and Volkman et al (1980) confirmed the structures as methyl (Me) and ethyl (Et) ketones,

whereas Rechka and Maxwell (1988) specified the trans stereochemistry of the double bonds

and showed that these double bonds are separated by five methylene units. More recently, less

common alkenones, such as shorter chain diunsaturated pentatriaconta-dien-2-one,

hexatriaconta-diene-2 and 3-one, were found in cultures of Emiliania huxleyi strains deprived of

nutrients and light (Prahl et al., 2006). Xu et al. (2001) also reported the presence of a rare

shorter chain alkenone with only three methylene groups between double bonds in Holocene

Black Sea sediments, and proposed that the source organisms for this compound were distinct

97

from those known to produce commonly observed C37 to C39 alkenones. Fujine et al. (2006)

reported an even shorter diunsaturated alkenone with only 34 carbons in Quaternary sediments

from the Japan Sea.

Here we report the tentative identification, based on mass spectral analysis, of

pentaunsaturated alkenones in the surface sediments of Lake Fryxell, a perennially ice-covered

lake in the McMurdo Dry Valleys of East Antarctica. The study follows the alkenone

nomenclature by Brassell (1993), where we represent 38 carbons having three double bonds with

the ketone group on the second carbon as Csg^Me or with the ketone group on the third carbon as

C38:3Et.

2.0 STUDY AREA, SAMPLE, AND ANALYTICAL METHODS

Lake Fryxell is near the mouth of Taylor Valley, adjacent to the McMurdo Sound (Fig.

1). It has a 3.3 to 4.5 m thick perennial ice cover, which prevents wind-driven mixing of the

surface water and allows only 1.3% of photosynthetically active radiation to reach the water

column (Spigel and Priscu, 1998). Below the ice, the upper 3.5 m of the water column consists of

freshwater supersaturated with oxygen that remains at ~0 °C year long (Roberts et al, 2000;

Lawson et al., 2004). At depths of 10-11 m, a chemocline separates the surface from the bottom

waters. From 11 to 19 m, the bottom water is anoxic and sulfidic, with salinity up to 6.2

(Vincent, 1988).

Lake bottom sediments were sampled with an Ekman grab in 2000, immediately wrapped

in pre-combusted foil, and kept frozen at -20 °C until processed in the laboratory. The frozen

sample was freeze-dried and extracted successively with MeOH, dichloromethane (DCM), and

DCM:MeOH (1:1, v/v) via ultrasonication until the solvent extract was colorless.

98

Fig. 1. Locations of Lake Fryxell, Taylor Valley (East Antarctica) and Ace Lake, Vestfold Hills (West Antarctica).

99

The total extract was saponified by refluxing in 0.40 N KOH/MeOH/hexane solution, from

which neutral lipids were recovered with hexane ( 3 x 1 ml). The neutral lipid fraction was rotary-

evaporated to near dryness and dried under N2. Further fractionation of the extract was

performed using column chromatography with hexane-soaked silica gel (Merck, 100-200 mesh).

Fractions were sequentially eluted using hexane (fraction Fl), mixtures of hexane:ethyl acetate

(95:5, v/v; F2), 90:10 (F3), 85:15 (F4), 80:20 (F5), 75:25 (F6), 70:30 (F7) andMeOH (F8).

Gas chromatography-mass spectrometry (GC-MS) was performed with an Hewlett-

Packard (HP)-6890 gas chromatograph coupled to a HP-5973 Mass Selective Detector used in

electron ionization mode at 70 eV, with He as carrier gas. The extracts were diluted with

cyclohexane (100 ul/mg) and introduced into the column through a split/splitless injector. The

column was a 30 m x 0.25 mm ID. with a 0.25 um thick film of HP-5MS. MS scan range was

from m/z 40 to 650 at a rate of three scans/s. The GC oven was programmed from 60 °C (1.5

min) to 130 °C at 20 °C/min and to 300 °C (held 52.5 min) at 4 °C/min. Only the F3, F4, and F5

fractions contained alkenones.

For the purposes of environmental DNA analysis, 2 L of lake water were collected from

the water column of Lake Fryxell during at a depth of 9m and filtered onto 0.2 um Pall Nylon

Membranes. Total environmental DNA was extracted using the Ultra Clean Soil DNA Isolation

Kit following the protocol of the manufacturer (Mobio, Carlsbad, CA, USA). The quality and

fragment length of the DNA extracts were determined by agarose gel electrophoresis (data not

shown). A partial 18S rDNA was amplified using the primers Ek-IF

(CATGTTGATCCTTGCCAG) and Ek-1520R (CYGCAGGTTCACCTAC) (Lefranc et al.

2005). The PCR product was gel purified, cloned into pCR2.1-TOPO vector according to the

manufacturer's instructions, and transformed into E. coli strain Top 10 to make a clone library.

Fifty clones from the library were screened by RFLP with the restriction enzyme Haelll to

determined the number of Operational Taxonomic units (OTUs), and representatives from each

OTU were selected and sent for sequencing. Sequencing was performed at the University of

Delaware Sequencing and Genotyping Center using standard primers and protocols. Preliminary

BLAST analysis identified two sequences as Prymnesiophytes. These two sequences were

chosen for further phylogenetic analysis, and alignment of the 18S DNA sequences was

performed with ClustalW using MEGA4 (Tamura et al. 2004; 2007). Comparative sequences

were selected from GenBank to represent major lineages of algae and eukaryotes. In addition,

18S rRNA sequences from prymnesiophytes isolated from Ace Lake, Antarctica (Coolen et al.

2004) and Greenland lakes (D'Andrea et al. 2006) were also included in the resultant

phylogenetic tree. A neighbor-joining analysis was performed and bootstrap consensus trees

(1000 psuedoreplicates) were generated using the Kimura 2-parameter distance model (Kimura

1980) with pairwise gap deletion.


GC-MS of the neutral fraction show the presence of long chain alkenones with 37 to 40

carbons (Fig. 2). Within the C38 group of alkenones is an unknown compound, I. Upon

separation of the neutral fraction, alkenones were observed in F3, F4 and F5, C37 alkenones

being exclusively present in F5 (Fig. 3C). C38 alkenones were enhanced in the F4 fraction,

CO c 0)

B

Phytol

••f h I""']*' •»»•*•> " •!••».•,-., j — «|» ! , • « » I f — y . . , * fllS

Epibrassicasterol

10 15 20 25' ' 3 0 ' ' '35 40 45

:4m

'55 c :3m

I Um

I 1 4m

| | : 4 e . :3e

IS A/\.2e

Alkenones «

m m

52 53 54 55 Retention time (min)

11 1 i —

56

'37 '38 '39 '40

Fig. 2. (A) Total ion current trace of neutral lipid fraction from a Lake Fryxell surface sediment sample. (B) C37 to C40 alkenones and an unknown alkenone, I.

102

A

c 40-

4m 3e ^ III A 4e A A

62

B

to

c CD

-•—» c

56

4m

J

3m

58 60

. l 4e

W./V.

3e

III 4ei

62

56 58 60 Retention time (min)

~ T —

62

Fig. 3. Partial total ion current trace of (A) fraction F3 where C39 andC4o alkenone are enhanced, (B) fraction F4 where C38 alkenones are enhanced and (C) fraction F5 where C37 alkenones exclusively elute.

whereas C39 and C40 alkenones were enhanced in F3 (Fig. 3B, A). This separation allowed better

resolution of C39 and C40 alkenone isomers and the detection of further unknown compounds

(peaks II, III, and IV).

3.1. Alkenone identification

The mass spectrum of peak I (Fig. 4) has M+D atm/z 538, two daltons less than C3g:4, and

the peak elutes prior to C38:4. A base peak at m/z 43, fragment ions m/z 520 [M-18]+D

corresponding to the loss of water and m/z 495 [M-43]+, corresponding to the loss of a C2H3O

fragment, are typical of methyl ketones (de Leeuw et al., 1979). Thus, we tentatively assigned

the compound corresponding to peak I as a C38 methyl ketone with five double bonds,

octatriaconta-pentaen-2-one (C38:sMe; Table I). The compound cannot be C^a fatty acid methyl

ester or C-i^.2 fatty acid ethyl ester based on M+.

Peak II and peak III elute prior to C39:4Me and C40:4Me, respectively (Fig. 3A, B). In the

F3 fraction, the mass spectrum of peak II has M+D at m/z 552, 2 daltons less than that of the

C39:4Me and 14 daltons more than compound I. However, the mass spectrum of peak II does not

display a fragment ion [M-43]+ typical of methyl ketones, although fragment ions at m/z 55 and

[M-55]+ are present. This cannot be a C40 alkene because the hydrocarbon fraction (Fl and F2)

does not contain a C40 alkene and alkenes would have been removed prior to elution of the F3

fraction upon silica gel chromatography. Thus, the compound corresponding to peak II is likely

to be a pentaunsaturated C39 ketone. In the F4 fraction, the mass spectrum of peak III displays

M+D at m/z 566 (2 daltons less than C40:4Me and 28 daltons more than compound I), a base peak

at m/z 43, and a fragment ion at m/z 523 or [M-43]+ (Table I). These data suggest that the

compound corresponding to peak III may be (tentatively) a C40 pentaunsaturated methyl ketone,

163 ii 189 2ig

M+-43 495

M+

[M-18]+ 538 520 ' j—.—U-

«» ^ .-"i -V 287.315 343 369 413 .....i, f. ,.<., > • > , . , , . , i ......

M+-43 M+

495 538

50 100 150 200 250 300 350 400 450 500

m/z

Fig. 4. Mass spectra of I, tentatively identified as octatriacontapentaen-2-one (C38:5Me).

TABLE I

Molecular ion and major fragment ions of penta-unstaurated alkenones

Penta- compound nomenclature M Important fragment ions unsaturated name Alkenones

I octatriaconta-pentaen-2-one II nonatriaconta-pentaen-2-one III tetradec-pentaen-2-one IV nonatriaconta-pentaen-3-one

C38:5Me 538 43, 495 [M-43]+, 520 [M-18]+

C39:5Me 552 55,497 [M-55]+

C4o;5Me 566 43, 523 [M-43]+

C39:sEt 552 57,495 [M-57]+

tetradecapentaen-2-one (C^sMe). As peaks I, II, and III all elute prior to C38:4Me, C39:4Me and

C40:4Me, respectively and, considering the robustness of the mass spectral data for peaks I and

III, it can be speculated that peak II may also represent a methyl ketone, so it is tentatively

assigned as nonatriaconta-pentaen-2-one, C39:sMe (Table I). The first peak of the C39 alkenone

group, peak IV, elutes just before C38:4Et. The mass spectrum has a M+ at m/z 552, an ion at m/z

495 [M-57]+ corresponding to loss of C3H5O and a base peak at m/z 57. This compound is

tentatively assigned as a C39 ethyl alkenone with five double bonds, nonatriacontapentaen-3-one

(C39:sEt).

Unfortunately, there was only a small amount of the sample and there were no extracts

left for other spectroscopic identification. Re-injection of the fractions yield smaller peaks for

alkenones with four double bonds and even smaller peaks for alkenones with five double bonds.

The presence of five double bonds in ketones with 38 to 40 carbons is consistent with the

alkenone biosynthetic pathway proposed by Rontani et al. (2006) for alkenones with more than

two double bonds separated by five methylene units. Rontani et al. (2006) proposed that

desaturase activity starts at A14 and A21, followed by progressive desaturation at A7 and A28 to

produce triunsaturated and tetraunsaturated ketones, respectively. If this biosynthetic path is

followed, it can be speculated that an additional desaturation at A would have resulted in the

pentaunsaturated C38:sMe, C39:5Me, C39:sEt and C4o:5Me ketones observed in Lake Fryxell

sediments, and can also explain the absence of the alkenones C37:5Me and C38:5Et, which cannot

accommodate a fifth unsaturation within their structures.

3.2 Lake Fryxell source organism!s) of alkenones

The library yielded four distinct OTUs, The dominant OTU in our clone library was

identified by sequencing as belonging to the Cryptomonads group (Cryptophyta) (data not

shown). However, molecular phylogenies of two 18S rRNA clones inferred that two of our

isolated clones were 100% specific for Prymnesiophyceae (Fig. 5). Phlylogenetic analysis

indicated that the Fryxell sequences were most closely related to eachother (99% sequence

identity). Lake Fryxell sequences were also closely associated with Ace Lake phylotypes 4 and

5 than Ace Lake phlytypes 1 and 2. Ace lake phylotype 4 was dominant when the lake became

brackish 9400 to 7700 cal y BP when marine waters invaded as the sea level rose (Coolen et al.,

2004). The Ace Lake phylotypes 1 and 2, which are adapted to more saline conditions when the

Lake was connected to the open sea, are in a different cluster (Coolen et al., 2004). Fryxell

prymnesiophytes were also a distinct cluster from I. galbana (Fig. 5). The 18S rDNA sequences

of Greenland lakes group into Greenland phlotypes and I. galbana-like haptophytes, which has

more similarity to Lake Fryxell sequences (D'Andrea and Huang, 2005). In support of our

findings, B. Lanoil (2007, personal communication) also detected haptophytes from Isochrysis

sp. plastid genes in a culture-independent 16S rDNA molecular survey at 12 m depth, below the

chemocline in Lake Fryxell.

3.3. Alkenone distribution

C37 alkenones dominate C38 and C39 alkenones, and are accompanied by trace abundances

of C40 alkenones (Fig. 2B). C38 alkenones is dominated by C38:sMe, which represent 49% of the

total C38 ketones. The C39:4Me dominates the C39 alkenones and is followed by compound II,

C39:5Me; the ethyl C39 ketones, including compound IV, C39:sEt, have low abundances (Fig. 3a).

P ry m n e siophyceaeim

Haptophyta

DQ07S203 Isochrysis sp.

T J - DQD71573 \sochrysissp.

DQ071574 Isochrysis sp.

DQ079859 Isochrysis sp.

•AJ246266.1 Isochryss gatbana

r-AM490999.1 P.pandoxa

r*t490998.1 Chymtila la.iKlhja

V H S Sediment 1

T—HS Sediment 3

™ FRX9m-1

LFRX9m-2

SB AY303355 DGG&5 Ace Lake haptophyte

S3 AY303354 DGG&4 .fee Lake haptophyte

r „ AY3D3351 DuGE-1 Ase Lake haptophyte

-AY303352 DGG&2 Ace Lake haptophyte

y**ta0997.1 Dfcratemasp.

AM490996.1 Isochtysis/totalis

721AJ24267 tmanionia tottmda

1*1491009 dwysocnmstulina of.

sit T2I

HE L AM491004 Prymnesium nemaMethecum

X7748 0 Phaeocystis antafctta

AB183665 Gephtocapsa oceanica

AF184167.1 emUania hvxteyi

AB183618 Em&ania sp.

AJ243261 Goccolithus pelagicus

AJ24S262 QzxsiplaccolitiKis moholis

ABD58360 Gephymcapsa oceanica

MMlilQ')? Sytacosphaera putebra

* W 9 0 9 8 6 Qzmmsphaeta weoVterranea

im

kf_ AM490980 Ceftrosiofaera rnmjcosa

AJ246264 Pleumchrysis elongata

BSIate-Holocene 1

h-HS Sediment 4

I- HS Sediment !

i m IAJ246263 Pteurochrysiy carte/ae

SS6 sediment 1

BSIate-HolooeneS

p LS Sediment 1

L LS Sediment 2

SS6 sediment 2

Greenland phylotype

- H S mater filtrate 2

BS Sediment 1

IP* silj—-

Pavlophyceae »

BS Sediment 2

HS uiaerti l trate 1

BS lae-Holooene 2

U40925 coccoid haptophyte

1 AF 102371 Pavlova gyans

i m l L34669 Pavlova salma

— ABD17122 Py&mitnnasolivacea

K72708 OntoreMasp.

Figure 5. Neighbour-joining tree showing 18S rDNA-inferred relationships between haptophyte sequences isolated from Lake Fryxell (bold), west Greenland lakes (underlined) (D'Andrea & Hong, unpublished) and Ace Lake. Neighbour-joining bootstrap values greater than 50% are shown. Evolutionary distance for number of changes per site is shown on a scale bar.

109

C40 alkenones are dominated by C^^Et with compound III, C^sMe, C4o:4Me, C4o;4Et being

similarly abundant (Fig. 3). The more unsaturated species, tri, tetra and pentaunsaturated forms,

dominate alkenones of the same carbon number. Alkenone chain lengths of up to 40 carbons were

only reported in sediments from the present-day meromictic saline Ace Lake in Antarctica

(Coolen et al., 2004), Erhai Lake in China (Li et al., 1996), in a hypersaline coastal microbial

mat in Spain (Lopez et al., 2005), and in cultures of C. lamellosa (Table II; Volkman et al.,

1995). In Ace Lake, the alkenone distribution comprised of chain lengths up to 40 carbons in the

saline meromictic environment today and when the lake was brackish at the onset of Holocene

transgression (Table IV; Coolen et al., 2004). When Ace Lake was a marine inlet, the alkenone

distribution was limited to C37Me and C3sEt. E. huxleyi and Gephyrocapsa oceanica are the main

producers of alkenones in marine environments, whereas Chrysotila lamellosa and Isochrysis

galbana are associated with brackish environments (e.g., Marlowe et al., 1984; Prahl and

Wakeham, 1987; Versteegh et al., 2001; Rontani and Volkman, 2005). Among cultures of these

species, only C. lamellosa contain C40 alkenones (Table IV). C39Me is uncommon and is only

produced by G. oceanica and C. lamellosa in cultures. G. oceanica is not an expected

contributor of this alkenone This suggests a possible contribution of alkenones from C.

lamellosa, aside from I. galbana.

Carbon chain length distribution, a ratio between total C37 to total C38 abundances were

proposed to differentiate haptophyte species (Rostek, 1979). Chu et al. (2005) summarized the

C37/C38 values from other lakes and marine environments and indicated that the open sea

generally has low values (<2). The coastal I. galbana and C. lamellosa, however, have a wide

range and have the highest mean values at 10.9 and 5.6, respectively. In Lake Fryxell, the 1.45

110

TA

BL

E I

I A

LK

EN

ON

E D

IST

RIB

UT

ION

FR

OM

LA

KE

S, R

EPR

ESE

NT

AT

IVE

MA

RIN

E E

NV

IRO

NM

EN

TS

AN

D C

UL

TU

RE

EX

PER

IME

NT

S *3

8:

4 3

2 3

2 ^

39

: 4

3 '4

0:

Oth

er a

lken

ones

2 5

4 3

4 3

2 5

4

Lak

es

Lak

e Fr

yxel

l, A

ntar

ctic

a1

Ace

Lak

e, A

ntar

ctic

a2

Bra

ya S

0, G

reen

land

L

ake

SS6,

Gre

enla

nd3

Lak

e St

eiss

linge

n, G

erm

any4

Tib

etan

Sal

t Lak

e, C

hina

5

Lak

e Q

ingh

ai,

Chi

na6

Erh

ai la

ke, C

hina

7

Coa

stal

to

open

oce

an

Bal

tic S

ea8

54°

N A

tlant

ic O

cean

9

58°

S So

uthe

rn O

cean

wat

er

filtr

ate10

58°

S So

uthe

rn O

cean

se

dim

ent10

Nor

weg

ian

Fjor

d11

Tri

nita

t So

lar

Salte

rn12

Cul

ture

s E

. Hux

leyi

1' E

. Hux

ley

in

E.

Hux

leyi

u

G. o

cean

ica1

C. l

amel

losa

16

C. l

amel

losa

1

I. g

alba

na11

1. g

alba

nan

Isoc

hrys

is s

p.

(mer

omic

tic s

alin

e la

ke)

(mar

ine

inle

t)

Bra

ckis

h ol

igos

alin

e O

ligos

alin

e Y

oung

er D

ryas

dep

osit

Col

d pe

riod

Bal

tic S

ea-l

ike

mar

ine-

like

Hyp

ersa

line

; m

at

10°C

25

°C

11.r

e 10

°c

20 °c

15 °

c 15

°C

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

e e e X

X

e e e e e e e e e X

X

e e e e e e e e e

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

m

X

m

m

m

e X

X

e X e X

X

e X e X

X

m

m

m

m

m

m

m

m

m

X

m

m

m

e e e e

X

X

m

m

m

m

X

X

m

m

m

m

m

Cr

C39

> iM

e, C

40

C37

:1M

e

C33

:3M

e,C

33:2

Me,

C34

:2E

t

The

num

ber

of c

arbo

ns a

re f

ollo

wed

by

a co

lon,

num

ber

of d

oubl

e bo

nds

are

unde

rlin

ed.

Met

hyl

(m)

or e

thyl

(e)

is

the

pos

itio

n of

the

car

bony

l at

the

seco

nd o

r th

ird

carb

on,

resp

ecti

vely

. If

the

pos

itio

n of

the

carb

onyl

is

not

reso

lved

, x

is a

ssig

ned

to b

oth

met

hyl

or e

thyl

ket

ones

. Alk

enon

es p

rodu

ced

from

+ac

etyl

-SC

oA a

nd

*pro

pion

yl-S

CoA

pri

mer

s 'T

his

stud

y; 2

Coo

len

et a

l., 2

004;

3D

'And

rea

et a

l., 2

006;

4Z

ink

et a

l., 2

001;

5W

ang

and

Zhe

ng (

1998

); 6

Liu

et

al. (

2008

); 7L

i et

al

(199

6);

8Sch

ulz

et a

l. (2

000)

; 9R

osel

l-M

ele

et a

l (1

995)

; 10Si

kes

et a

l. (1

997)

; nR

osel

l-M

ele

(199

8); l

2Lop

ez e

t al

. (20

05);

13Pr

ahl

and

Wak

eham

(19

87);

14V

olkm

an e

t al

(19

79);

15

Vol

kman

et

al (

1995

); 1

6Ron

tani

et

al. (

2004

); 1

7Elt

grot

h et

al

(200

5); 1

8Mar

low

e et

al

(198

4b)

I l l

ratio, calculated from the neutral fraction, is more similar to E. huxleyi and G. oceanica.

The ratio is too low for the 5.4-15.1 range for/, galbana (Marlowe et al., 1994), but is the

minimum for C. lamellosa (1.4-9.3; Rontani et al., 2004; Prahl et al., 1988). The values

overlap with the range of ratios in the Baltic Sea (Schulz et al., 2000), Lake Van (Thiel et

al., 1997), and some of the Chinese lakes (Chu et al , 2005).

The biosynthetic pathways traced by Rontani et al. (2006) based on the position of

the double bonds presents a pattern for the distribution of alkenones. These representative

alkenone distributions from cultures, sediments and water filtrates, summarized in Table II,

more commonly have the acetyl-SCoA primer than the Propionyl-SCoA. Lake Fryxell has

a wide range of alkenones in length and number of unsaturations, whereas the Baltic Sea,

Ace Lake as a marine inlet, and the Spanish saltern all have distributions specific for the

acetyl-SCoA primer. In Ace Lake the changes in the alkenone distribution in the last 20,000

years were a result of changes in the haptophye ecology. Interestingly, the changes in

ecology did not shift the alkenone distribution to the propionyl-SCoA primer. In the Baltic

Sea, a strong gradient from marine to freshwater conditions also represents changes in the

alkenone distribution, and reflect both the effect of ecology and adaptation to salinity

changes (Schultz et al., 2000).

The tetraunsaturated form is the dominant C37 alkenone, commonly observed in

lakes (Cranwell, 1985; Schouten et al., 2001). In areas with brackish salinities in the Baltic

Sea, the lack of C3gMe ketones coincides with higher Csy^Me concentrations. A linear

relationship between the C^-AVO of the total C37 alkenones and salinity in Chinese lakes was

also observed although to a lesser degree than the C37:4% to mean annual air temperature

(Chu et al., 2005). In the open ocean, however, the salinity control over the C37% is limited

112

to the North Atlantic where salinity strongly correlates with temperature and indicates that

the trend is an artifact of other environmetal factors, such as slow growth rates, low light

and nutrient supply (Sikes and Sicre, 2002; Conte et al., 1998).

In Lake Fryxell, the C37:4 abundance of 82% in F5 or 77% in the neutral fraction is

similar to 14 of the lakes from China and within the range for Lake Van and Lake

Steisslingen, but is beyond the range of values for E. Huxleyi, I. galbana, and C. lamellosa,

although the latter two produce the highest percentage (11-42% ; Marlowe et al., 1984;

Conte et al., 1994). In Ace Lake, the maximum abundance of C37:4Me in the sediments was

associated with the high concentrations of Ace Lake phylotypes 3 and 4. Since these

phylotypes are closely associated with Lake Fryxell haptophytes, the high abundance of

C37:4Me may be due to a combination of the type of haptophyes and low salinity conditions.

3.4 Other biomarkers associated with alkenones

Alkenones are known biomarkers of certain haptophyte microalgae (Volkman et al.,

1980). The presence of 24-methylcholesta-5,22-diene-3P-ol, epibrassicasterol or its epimer

(Fig. 2A), is consistent with the presence of haptophytes (Conte et al. 1994 and references

therein), and diatoms (Volkman, 1986) and Cryptomonas spp. (Goad et al., 1983), which

both occur in Lake Fryxell waters (e.g. Lizotte and Priscu, 1998; Laybourn-Parry, 2006). A

suite of sterols in the sediments are also identified and listed in Table III.

TABLE III

Sterols in Lake Fryxell sediments

Carbon Sterol number

27 cholest 5-en-3b-ol (cholesterol) 27 5a-Cholestan-3b-ol (cholestanol) 28 24-methyl-cholest-5,22-dien-3b-ol (Brassicasterol or

epibrassicasterol) 28 24-methylcholest-5-en-3b-ol (campesterol) 28 24-methyl-5a-cholest-22-en-3b-ol (brassicastanol) 28 4-methyolcholesta-3b-ol 28 24-methylcholestanol 29 24-ethyl-5a-cholest-5-en-3b-ol (sitosterol) 29 24-ethyl-cholest-5,22-dien-3b-ol (stigmasterol) 29 4,24-dimethylcholest-22-en-3b-ol 30 4a,23,24-trimethyl-5a-cholestan-3b-ol (dinostanol) 30 4a,23,24-trimethyl-5a-cholest-22en-3b-ol (dinosterol)

114

3.5. Factors relating to alkenone unsaturation

Numbers of double bonds and chain lengths of alkenones in culture experiments are

primarily controlled by the species of alkenone-synthesizing organisms (Marlowe et al.,

1984; Volkman et al., 1995), and nutrient availability (Epstein et al., 1998; Popp et al.,

1998; Prahl et al., 2006), light deprivation (Prahl et al., 2006) and salinity (Fujine et al.,

2006), as well as the water depth at which alkenones are produced (Ternois et al., 1997).

Versteegh et al. (2001) showed that phosphate and light limitation independently resulted in

an increase in alkenone unsaturation during growth experiments with /. galbana. Lake

Fryxell is phosphorus-limited (Dore and Priscu, 2001) and light-limited because of the

thick perennial ice cover (Howard-Williams et al., 1998), suggesting that nutrient and light

stress may exert control on alkenone unsaturation.. Using staining techniques, significant

accumulations of alkenones were observed in the chloroplast, where they are synthesized,

and are probably used as sinks for surplus reducing power (Yamamoto, 2000; Eltgroth et

al., 2005). In the cells of I. galbana andis. huxleyi, Liu and Lin (2001) and Eltgroth et al.

(2005) also determined that alkenones cluster as cytoplasmic lipid bodies, which increase

during phosphate- or nitrogen-limited conditions during growth experiments, (Prahl et al.,

2003; Eltgroth et al., 2005). Lipid bodies are not just for energy storage (e.g. Pond and

Harris, 1996; Epstein, 1998), but these oils can be quickly mobilized to be used in various

aspects of cell metabolism (Cohen et al., 2000; Khozin-Goldberg et al., 2005). Such a

process may be important when algae meet transitory favorable growth conditions in a

generally unfavorable environment (Khozin-Goldberg et al., 2005). Alkenones in the oil

115

bodies must remain unfrozen and, in the cold waters of Lake Fryxell, a high level of

alkenone unsaturation, including pentaunsaturated alkenones, is required to prevent

freezing.

Alkenones are more resistant to oxygen radicals and photodegradation than

carotenoids (Rontani et al., 1997) and the trans geometry of their double bonds are more

photostable than the cis geometry of alkenes (Mouzdahir et al., 2001), explaining why

alkenones may be favored over triacylglycerol for energy storage (Eltgroth et al., 2005).

The use of alkenones for metabolic storage in E. huxleyi, instead of triacylglycerols, was

also noted by Epstein et al. (2001). In Lake Fryxell, where the upper water column is

supersaturated in oxygen, photooxidative stability afforded by alkenones may be beneficial

to the algae.

The constantly cold temperatures of Lake Fryxell surface and bottom waters,

combined with the anoxic and sulfidic bottom water conditions, result in the preservation of

pentaunsaturated alkenones in the sediments, which elsewhere might experience

degradation.

3.6 Climate proxy record

The calculated UK37 value (-0.74) from the neutral fraction for Lake Fryxell

sediments is consistently to those calculated for Greenland (-0.68 to -0.60; D'Andrea et al.,

2006), and German (-0.53 to -0.29; Zink et al., 2001), Turkish (0 to -0.61; Thiel et al.,

1997) and Chinese (Li et al., 1996) lakes, to name a few. The calculated UK 37 value of 0.04

in Lake Fryxell is similar to Greenland lakes and consistently less than UK 37 of German,

Turkish and Chinese lakes. Temperature estimates from U 37 and U 37 are not realistic

116

(Table IV). Closest estimates are from calibrations of the German lakes (Zink, 2001) and

from E. huxleyi (Prahl e al., 1988). The trends for Uk37 and Uk 37 signify that the use of

alkenone-based proxies for temperature reconstructions is viable in cold regions and in

comparing data for areas from different temperature belts, considering

that numerous non-thermal factors that affect the number of double bonds in alkenones.

There is, however, a need to calibrate alkenone-based paleotemperature reconstructions for

lakes in cold regions.

4.0 CONCLUSIONS

GC-MS analysis allowed tentative identification of pentaunsaturated C38 to C40

alkenones in Lake Fryxell. These alkenones are susceptible to photooxidative and

autooxidative degradation and are difficult to preserve. The likely organisms that

biosynthesize the alkenones are Isochrysis sp. based on its 18S rDNA sequences detected at

9 m depth, in accord with the dominance of epibrassicasterol among the sterols. Based on

the distribution of alkenones in cultures of haptophytes, only C. lamellosa biosynthesize up

to C40 alkenones, which suggest the possible contribution of C. lamellosa to the alkenones

in the sediments. The distribution of alkenones in Lake Fryxell sediments does not match

previous investigation of alkenones. It can be speculated that (i) other unidentified

organisms in Lake Fryxell contribute to the alkenones observed, or that (ii) Isochrysis spp.

produce the observed alkenones in the cold, nutrient- and light-limited conditions of Lake

Fryxell.

117

TABLE IV

Temperature estimates for Lake Fryxell based on Uk37 and Uk 37 calibration

Calibration equation Reference Temperature -3.65 7.37 6.05 -3.65 -0.07 6.57 4.20 -4.08

Uk37 = 0.0377T - 0.5992 Uk37 = 0.052T-1.12 Uk37 = 0.02T +0.121 Uk37 = 0.0377T-0.5992 Uk'37 = 0.034T + 0.039 Uk'37= 0.022T + 0.017 Uk'37 = 0.21 I T - 0.725 Uk'37 = 0.0257T-0.2608

Brasselletal. (1986) /. galbana (Marlowe, 1984) German Lakes (Zink et al., 2001) C. Lamellosa 10-22 °C (Sun et al., 2007) E. huxleyi (Prahl et al., 1988) Southern Ocean (Sikes and Volkman, 1993) German Lake (Zink et al., 2001) C. Lamellosa 14-22 °C (Sun et al., 2007)

118

5.0 ACKNOWLEDGEMENTS

The project was supported by the Office of Polar Programs of the National Science

Foundation (Antarctic Biology and Medicine SGER 0346316, LTER 0423595). Support

was also granted from the University of Illinois at Chicago Institute of Environmental

Science and Policy and the Provost office. We thank J. Lawson for collecting and initially

processing the sample, as well as T. Chung and M. Muccianti for help in the laboratory. K.

Rodolfo and J. Priscu are acknowledged for editing the manuscript. We appreciate J-F

Rontani and J. Volkman whose the constructive comments and suggestions greatly

improved this manuscript.


Boon, J.J., van der Meer, F.W., Schuyl, P.J.W., de Leeuw, J.W., Schenk, P.A., and Burlingame, A.L.: Organic geochemical analyses of core samples from Site 362, Walvis Ridge, DSDP Leg 40. In: Initial Reports of the Deep Sea Drilling Project, pp. 627-637, U.S. Government Printing Office, Washington, DC, 1987.

Brassell, S.C., Maxwell, J.R., and Eglinton, G.: Preliminary lipid analyses of two Quaternary sediments from the Middle America Trench, southern Mexico transect. In: Initial Reports of the Deep Sea Drilling Project 66. pp. 557-580. U.S. Government Printing Office, Washington, DC, 1981.

Brassell, S.C, Eglinton, G.: Lipid indicators of microbial activity in marine sediments. In: Heterotrophic Activity in the Sea. Plenum Press, eds. J.E. Hobbie, and PJ. Williams, leB., pp. 481-503. New York, 1984.

Brassell, S.C, Eglinton, G., Marlowe, I.T., Pflaumann, U., and Sarnthein, M.: Molecular stratigraphy: a new tool for climatic assessment. Nature 320: 129-133, 1986.

Brassell, S.C: Applications of biomarkers for delineating marine paleoclimatic fluctuations during the Pleistocene. In: Org. Geochem. M.H. Engel and S.A. Macko (eds.), pp. 699-737. Plenum Press, New York, 1993.

119

Cohen, Z., Khozin-Goldberg, I., Alderstein, D., and Bigogno, C : The role of triacylglycerol as a reservoir of polyunsaturated fatty acids for the rapid production of chloroplastic lipids in certain microalgae. Biochem. Soc. Trans. 26: 740-743, 2000.

Conte, M.H., Volkman, J.K., and Eglinton, G.: Lipid Biomarkers of the Haptophyta. Clarendon Press, Oxford, 1994.

Coolen, M.J.L., Muyzer, G., Rijpstra, W.I.C., Schouten, S., Volkman, J.K., and Damste, J.S.S.: Combined DNA and lipid analyses of sediments reveal changes in Holocene haptophyte and diatom populations in an Antarctic lake. Earth and Planet. Sci. Lett. 223:225-239,2004.

Cranwell, P.A.: Long-chain unsaturated ketones in recent lacustrine sediments. Geochim. et Cosmochim. Acta 49:1545-1551, 1985.

D'Andrea, W.J., and Huang, Y.: Long chain alkenones in Greenland lake sediments: Low 5 C values and exceptional abundance. Org. Geochem. 36: 1234-1241, 2005.

D'Andrea, W.J., Lage, M., Martiny, J.B.H., Laatsch, A.D., Amaral-Zettler, L.A., Sogin, M.L., and Huang, Y.S.: Alkenone producers inferred from well-preserved 18S rDNA in Greenland lake sediments. J. of Geophvs. Res.-Biogeosciences 111; G03013, doi:10.1029/2005JG000121, 2006.

de Leeuw, J.W., van der Meer, F.W., Rijpstra, W.I.C., and Schenck, P.A.: On the occurrence and structural identification of long chain unsaturated ketones and hydrocarbons in sediments. In: Advances in Org. Geochem. Douglas, A.G. and Maxwell, J.R. (eds.) pp. 21 l-217.Pergamon Press, Oxford, 1980.

Dore, J.E., and Priscu, J.C.: Phytoplankton phosphorus deficiency and alkaline phosphatase activity in the McMurdo Dry Valley lakes, Antarctica. Limnol. and Oceanography 46: 1331-1346,2001.

Eglinton, T.I., Conte, M.H., Eglinton, G., and Hayes, J.M.: Proceedings of a workshop on alkenone-based paleoceanographic indicators. Geochem., Geophvs. Geosvs. 2, art. no.-2000GC000122,2001.

Eltgroth, M.L., Watwood, R.L., and Wolfe, G.V.: Production and cellular localization of neutral long-chain lipids in the haptophyte algae Isochrysis galbana and Emiliania huxlevi. J.ofPhvcol.41: 1000-1009, 2005.

Epstein, B.L., D'Hondt, S., and Hargraves, P.E.: The possible metabolic role of C37

alkenones in Emiliania huxleyi. Org. Geochem. 32, 867-875, 2001.

120

Epstein, B.L., D'Hondt, S., Quinn, J.G., Zhang, J.P., and Hargraves, P.E.: An effect of dissolved nutrient concentrations on alkenone-based temperature estimates. Paleoceanographv 13: 122-126, 1998.

Fujine, K., Yamamoto, M., Tada, R., and Kido, Y.: A salinity-related occurrence of a novel alkenone and alkenoate in Late Pleistocene sediments from the Japan Sea. Org. Geochem. 37: 1074-1084,2006.

Goad, L.J., Holz, J.G.G., and Beach, D.H.: Identification of (24S)-24-methylcholesta-5,22-dien-3p-ol as the major sterol of a marine cryptophyte and a marine prymnesiophyte. Phytochemistrv22: 475-476, 1983.

Harwood, J.L., and Russell, N.L.: Lipids in Plants and Microbes. George Allen & Unwin, London, 1984.

Howard-Williams C , Schwarz A. M., Hawes I., and Priscu J. C. Optical properties of the McMurdo Dry Valley Lakes, Antarctica. In: Ecosystem Dynamics in a Polar Desert. The McMurdo Dry Valleys, Antarctica, ed. J.C. Priscu, pp. 189-203, American Geophysical Union, Washington, D.C, 1998.

Khozin-Goldberg, I., Cohen, Z., Pimenta-Leibowitz, M., Nechev, J., andZilberg, D.: Feeding with arachidonic acid-rich triacylglycerols from the microalga Parietochloris incisa improved recovery of guppies from infection with Tetrahymena sp. Aquaculture 255: 142-150,2006.

Khozin-Goldberg, I., Shrestha, P., and Cohen, Z.: Mobilization of arachidonyl moieties from triacylglycerols into chloroplastic lipids following recovery from nitrogen starvation of the microalga Parietochloris incisa. Biochim. et Biophys. Acta 1738: 63-71,2005.

Lawson, J., Doran, P.T., Kenig, F., de Marais, J.D., and Priscu, J.C: Stable carbon and nitrogen isotopic composition of benthic and pelagic organic matter in lakes of the McMurdo Dry Valleys, Antarctica. Aquatic Geochem. 10: 269-301, 2004.

Laybourn-Parry, J., and Pearce, D.A.: The biodiversity and ecology of Antarctic Lakes: models for evolution. Phil. Trans, of the Royal Society B-Biol. Sci. 362: 2273-2289, 2006.

Li, X-F, Cullen, W. R., Reimer, K. J., and Le, X-C: Microbial degradation of pyrene and characterization of a metabolite. The Sci. of the Tot. Environ. 77: 17-29, 1996.

Liu, C.P., and Lin, L.P.: Ultrastructural study and lipid formation of Isochrysis sp. CCMP1324. Botanical Bull, of Academia Sinica42: 207-214, 2001.

121

Lizotte, M.P., and Priscu, J.C.: Pigment analysis of the distribution, succession and fate of phytoplankton in the McMurdo Dry Valley lakes of Antarctica. In: Ecosystem Dynamics in a Polar Desert: The McMurdo Dry Valleys, ed. J.C. Priscu, pp. 229-239, American Geophysical Union, Washington, DC, 1998.

Marlowe, I.T.: Lipids as Paleoclimatic Indicators, University of Bristol, Bristol, UK, 1984.

Marlowe, I. T., Brassell, S. C , Eglinton, G., and Green, J. C : Long-chain unsaturated ketones and esters in living algae and marine sediments. Org. Geochem. 6: 135-141,1984a.

Mouzdahir, A., Grossi, V., Bakkas, S., and Rontani, J.F.: Visible light-dependent degradation of long-chain alkenes in killed cells of Emiliania huxleyi and Nannochloropsis salina. Phytochemistry 56: 677-684, 2001.

Pond, D.W., and Harris, R.P.: The lipid composition of the coccolithophore Emiliania huxleyi and its possible ecophysiological significance. J. of the Marine Biol. Assoc, of the United Kingdom 76: 579-594, 1996.

Popp, B.N., Kenig, F., Wakeham, S.G., Laws, E.A., and Bidigare, R.R.: Does growth rate affect ketone unsaturation and intracellular carbon isotopic variability in Emiliania huxleyi? Paleoceanography 13: 35-41, 1998.

Prahl, F.G., Mix, A.C., and Sparrow, M.A.: Alkenone paleothermometry: Biological lessons from marine sediment records off western South America. Geochim. et Cosmochim. Acta 70: 2856-2867, 2006.

Prahl, F.G., and Wakeham, S.G.: Calibration of unsaturation patterns in long-chain ketone compositions for paleotemperature assessment. Nature 330: 67-369, 1987.

Prahl, F.G., Muehlhausen, L.A., and Zahnle, D.L.: Further evidence of long-chain alkenones as indicators of paleoceanographic conditions. Geochim. et Cosmochim. Acta 52:2303-2310, 1988.

Prahl, F.G., Wolfe, G.V., and Sparrow, M.A.: Physiological impacts on alkenone paleothermometry. Paleoceanography 18: 1025. doi:10.1029/2002PA000803, 2003.

Rechka, J. A., and Maxwell, J.R.: Characterization of alkenone temperature indicators in sediments and organisms. Org. Geochem. 13: 727-734, 1988.

Roberts, E.C., Laybourn-Parry, J., McKnight, D.M. and Novarinos, G.: Stratification and dynamics of microbial loop communities in Lake Fryxell, Antarctica. Freshwater Biol. 44:649-661,2000.

122

Rontani, J.F., Prahl, F.G., and Volkman, J.K.: Re-examination of the double bond positions in alkenones and derivatives: biosynthetic implications. J. of the Phvcol. Soc. of Am. 42:800-813,2006.

Rontani, J.F., Cuny, P., Grossi, V., and Beker, B.: Stability of long-chain alkenones in senescing cells of Emiliania huxleyi: effect of photochemical and aerobic microbial degradation on the alkenone unsaturation ratio (UK37). Org. Geochem. 26: 503-509, 1997.

Rontani, J.F., and Volkman, J.K.: Lipid characterization of coastal hypersaline cyanobacterial mats from the Camargue (France). Org. Geochem. 36: 251-272, 2005.

Schouten, S., Rijpstra, W. I. C , Kok, M., Hopmans, E. C , Summons, R. E., Volkman, J. K., and Sinninghe Damste, J. S.: Molecular organic tracers of biogeochemical processes in a saline meromictic lake (Ace Lake). Geochim. et Cosmochim. Acta 65: 1629-1640, 2001.

Schouten, S., Hopmans, E. C , SchefUp, E., and Sinninghe Damste, J. S.: Distributional variations in marine crenarchaeotal membrane lipids: a new tool for reconstructing ancient sea water temperatures? Earth and Planet. Sci. Lett. 204: 265-274, 2002.

Simoneit, B. R., and Burlingame, A.L.: Further preliminary results on the higher weight hydrocarbons and fatty acids in the DSDP cores, Leg 9. In: Initial reports of the Deep Sea Drilling Project 9: pp. 859-901U.S. Government Printing Office, Washington, D.C., 1972.

Simoneit, B. R., and Burlingame, A.L.: Preliminary organic analyses of DSDP cores, Leg 12 and 13. In: Initial Reports of the Deep Sea Drilling Project 17: pp. 561-590. U.S. Government Printing Office, Washington, DC, 1973.

Simoneit, B. R., Scott, E.S., and Burlingame, A.L:. Preliminary organic analyses of DSDP cores, Leg 14, Atlantic Ocean. In: Initial reports of the Deep Sea Drilling Project 16: U.S. Government Printing Office, Washington, DC, pp. 575-600, 1973.

Spigel, R.H., and Priscu, J.C.: Physical limnology of the McMurdo Dry Valley lakes. In: Ecosystem Dynamics in a Polar Desert: The McMurdo Dry Valleys, Antarctica 72: ed. J.C. Priscu, pp. 153-187.. American Geophysical Union , Washington, DC, 1998.

Sun, Q., Chu, G.Q., Liu, G.X., Li, S., and Wang, X.H.: Calibration of alkenone unsaturation index with growth temperature for a lacustrine species, Chrysotila lamellosa (Haptophyceae). Org. Geochem. 38: 1226-1234, 2007.

Ternois, Y., Sicre, M.A., Boireau, A., Conte, M.H., and Eglinton, G.: Evaluation of long-chain alkenones as paleo-temperature indicators in the Mediterranean Sea. Deep-Sea Research Part I-Oceanographic Res. Papers 44: 271-286, 1997.

123

Thiel, V., Jenisch, A., Landmann, G., Reimer, A., and Michaelis, W.: Unusual distributions of long-chain alkenones and tetrahyamanol from the high alkaline Lake Van, Turkey. Geochim. et Cosmochim. Acta 61: 2053-2064.

Versteegh, G.J.M., Riegman, R., de Leeuw, J.W., and Jansen, J.H.F., 2001. U 37 values for Isochrysis galbana as a function of culture temperature, light intensity and nutrient concentrations. Org. Geochem. 32: 785-794, 1997.

Vincent, W.F.: Microbial ecosystems of Antarctica. Cambridge University Press, Cambridge, 1988.

Volkman, J.K. A review of sterol markers for marine and terrigenous organic matter. Org. Geochem. 9: 83-99, 1986.

Volkman, J.K., Burton, H. R., Everitt, D.A., and Allen D. I.: Pigment and lipid compositions of algal and bacterial communities in Ace Lake, Vestfold Hills, Antarctica. Hydrobiologia 165: 41-57, 1988.

Volkman, J.K., Barrett, S.M., Blackburn, S.I., and Sikes, E.L.: Alkenones in Gephyrocapsa oceanica - implications for studies of paleoclimate. Geochim. et Cosmochim. Acta 59: 513-520, 1995.

Volkman, J.K., Eglinton, G., Corner, E.D.S., and Forsberg, T.E.V.: Long-chain alkenes and alkenones in the marine coccolithophorid Emiliania huxleyi. Phvtochem. 19: 2619, 1980.

Xu, L., Reddy, CM., Farrington, J.W., Frysinger, G.S., Gaines, R.B., Johnson, C.G., Nelson, R.K., and Eglinton, T.I.: Identification of a novel alkenone in Black Sea sediments. Org. Geochem. 32: 633-645, 2001.

Yamamoto, M., Shiraiwa, Y., and Inouye, I.: Physiological responses of lipids in Emiliania huxleyi and Gephyrocapsa oceanica (Haptophyceae) to growth status and their implications for alkenone paleothermometry. Org. Geochem. 31: 799-811, 2000.

Zink, K-G, Leythaeuser, D., Melkonian, M., and Schwark, L.: Temperature dependency of long-chain alkenone distributions in Recent to fossil limnic sediments and in lake waters. Geochim. et Cosmochim. Acta 65: 253-265, 2001.

CHAPTER 5

Distribution and significance of archaeal and bacterial tetraether membrane lipids

in lakes of Taylor Valley, Antarctica

Caroline M.B. Jaraula3, Stefan Schoutenb, Ellen C. Hopmansb, Simon C. Brassell0, Fabien Keniga

^Department of Earth and Environmental Sciences, University of Illinois at Chicago,

845 WTaylor St., Chicago IL 60607-7059, USA

Marine Organic Biogeochemistry, Royal Netherlands Institute for Sea Research, the Netherlands

'Department of Geological Sciences, Indiana University, 1001 East 10th Street, Bloomington, IN47405-1405, USA

124

125

Abstract

Taylor Valley, Antarctica, one of the coldest and driest deserts on Earth cannot support

vascular plants or vertebrates. The lake fauna and flora are exclusively microbial and

microplanktonic, which makes it an excellent environment to study biomarkers and represent a

cold dataset with stable temperatures year round due to the perennial ice covers. Sediments and

microbial mats collected from Lakes Fryxell, Hoare, Chad and Bonney, and from a granite

endolith in the valley were analyzed for the presence of isoprenoidal and branched glycerol

dialkyl glycerol tetraethers (GDGTs) and explore the veracity of the Tetraether Index (TEX86) in

estimating temperatures for these environments.

The TEX86, based on isoprenoidal GDGTs, is higher for Lake Fryxell than other lakes in

Taylor Valley, and overestimate the calculated temperature. Methanogens in which activities are

four orders of magnitude greater in Lake Fryxell than in the other lakes may have caused

temperature overestimates. Average temperature estimate in Lake Hoare is close to instrumental

records for the water column, which is almost isothermal. In Lake Bonney, the temperature

estimate is 2.4 °C, consistent with average instrumental records of 2.2 °C for the water column.

In Lake Chad, microbial mats contain up to 95% branched GDGTs that are markers for terrestrial

bacteria. Average mean annual air temperature estimate of-18 °C based on the MBT proxy of

branched GDGTs is within the range reported for mean annual air temperature.

The strong influence of methanogens in the pool of GDGT present in the lake sediments

affects the GDGT distribution, and can be used to track environmental conditions, such as

stratified water columns, which support methanogenesis. The modern GDGT distribution is

calibrated with known environmental conditions in the valley and compared with the distribution

in older sediment deposits to infer the paleoenvironment. Older archaeal and bacterial lipid

distributions are consistent with paleoenvironmental interpretations from bulk geochemical and

sedimentological analyses that indicate a large freshwater lake inundated the valley in the

Pleistocene. After the last glacial maximum, lake levels decreased and formed small hypersaline

ponds. From the mid-Holocene onward, the basins filled up to form the modern Taylor Valley

lakes.

Keywords: paleotemperature proxy, microbial mats, methanogens, ether-lipids, biomarkers

127

1.0 INTRODUCTION

The McMurdo Dry Valleys comprise the largest ice-free area in Antarctica

because annual sublimation exceeds precipitation (Fountain et al , 1999). In Taylor

Valley, one of the southern valleys, mean annual air temperatures are consistently

subzero, which makes this setting one of the coldest and driest deserts on Earth (Fritsen et

al., 2000). These harsh environmental conditions cannot support vascular plants or

vertebrates (Parker et al., 1982; Priscu et al., 1999; Roberts et al., 2000). Hence, the lakes

in the valley consist exclusively of planktic populations of microorganisms and benthic

microbial mats (Wharton et al., 1993). Perennial ice cover on the lakes leads to stable

water column structures and temperatures. The simple biological communities and stable

water column conditions make this region an excellent site to examine the characteristics

of biomarker proxy records that represents a cold environment.

In Taylor Valley, bacteria and eukarya are have been extensively studied and are

prevalent in the lakes (e.g. Laybourn-Parry et al., 1997; Priscu, 1997; Priscu et al , 1999),

whereas Archaea have been only reported from Lake Fryxell (Brambilla et al, 2001;

Karr et al, 2006; Singh et al, 2005). Archaea are resilient to extreme environmental

conditions and live independently or in consortia with bacteria. Hence, it is hypothesized

that archaea are present in Taylor Valley lakes and that there is high potential for archaeal

biomarkers to be preserved in the sediments because of the cold temperatures and anoxic

bottom waters of some lakes. For this study, lipid biomarkers are used to seek evidence

of Archaea in the lakes and a cryptoendolith of Taylor Valley. If present, the distributions

of archaeal biomarkers will be determined and calibrated with the present-day physico-

chemical conditions in the lakes. The trends will be used to correlate modern distribution

128

patterns to those in core sediments to infer paleoenvironments of the lakes and the valley.

Existing calibrations for reconstructing water column and mean annual air temperatures

will be tested if these are applicable to temperatures in Taylor Valley.

1.1 Taylor Valley

The 33 km-long Taylor Valley is situated west of McMurdo Sound at the foot of

the East Antarctic Ice Sheet (Fig. 1). In the valley bottom, annual air temperatures

average -14.8 to -30 °C (Doran et al, 2002), precipitation as snow is less than 10 mm of

water equivalent (Bromley, 1985), and potential evaporation exceeds 300 mm (Chinn,

1993). Slight changes in the environmental parameters such as temperature and ground

albedo can drastically change the hydrologic budget of different parts of the valley

(Fountain et al., 1999; Lyons et al., 2000). At mid-valley, Nussbaum Riegel, a bedrock

ridge, rises 700 m high above sea level (Spigel and Priscu, 1998), and stands as a barrier

against airflow and moisture from the McMurdo Sound. Consequently, the valley

microclimates change from relatively cold and dry near Taylor Glacier to warmer and

wetter toward the sound (Lyons et al., 2000; Fountain et al., 1999). The valley also

changes from narrow and steep at the head to wide and gentle gradient at its mouth.

129

sff i---" 7

270*1

24Cr

2icr~

Taylor Valley • Giaewrs

• lakes

Figure 1. Location map

£\ -' Gl

130

1.1.1. Climate

Climate in the valley is seasonal, winds and temperature being governed by the

polar cycle of four months of sunlight, four months of twilight and four months of

darkness (Clow et al., 1988; Fountain et al , 1999). During winter, dry katabatic winds

from the ice sheet descend into the valley contributing strongly to ice sublimation. In the

summer, moist coastal winds from the Ross Sea advect into the valley.

1.1.2 Ice cover

The perennial ice cover restricts the exchange of gases between the lakes and the

atmosphere (Wharton et al., 1993; Priscu et al., 1997), but also traps oxygen from

photosynthesis and carbon dioxide from respiration (Neumann et al., 2001),

dimethylsulfoxide, and dimethylsulfoniopropionate (Lee et al., 2004) in distinct layers of

the water column (Table 1). The ice cover also has a significant role in protecting the

water column from wind-driven mixing, thus helping to keep the lakes stratified

(Vincent, 1988).

1.1.3 Sources of water and materials

Materials are brought into the lakes by wind and water and by in situ production

of organic material by plankton and microbial mats in the moat and deeper waters.

Aeolian sediments are blocked by the ice cover, but eventually pass into the lake through

cracks and fissures in the ice. Glacial meltwaters feed streams that flow for the six to nine

131

TABLE 1

Physical and chemical descriptions of Taylor Valley lakes

Altitude, m

distance to sea, km

Latitude

Longitude

Length, km

Maximum width, km

Surface area, km2

Volume, 1

maximum depth, m

Ice thickness, m

(PAR), transmission thru

Number of streams

C02(aQ) equilibrium depth,

ice,

ma

%

Data from Spigel and Pricu (1998);

Lake Bonney

(west)

57

28

77°43'S

162°26'E

2.6

0.9

0.99

6 .0x l0 6

40

2.8-4.5

2.73

5

10 Lawson et al.

(east)

57

25

77°43'S

162°17'E

4.8

0.9

3.32

1.1 xlO 6

37

3-4.5

2.7

4

10 (2004); www.

Chad

74

22

77°38'S

162°45'E

1

0.2

0.15

No data

5.5

3.5-4.5

No data

3

No data ltemet.edu/sites

Hoare

73

15

77°38'S

162°55'E

4.2

1

1.94

2.6 xlO 6

34

3.1-5.5

1.59

1

10 /mem/; McKnif

Fryxell

18

9

77°37'S

163°09'E

5.8

2.1

7.08

4.3 x 106

20

3.3-4.5

1.34

13

0b

>ht et al. (1993). a below this depth, water is saturated in C02 and below chemoclines are supersaturated with CO2 (Neumann et al, 2001). b all depths in Lake Fryxell are saturated with respect to C02

http://ltemet.edu/sites

132

weeks of austral summer (Chinn, 1993). The Commonwealth and Canada Glaciers, for

example, feed some of the streams that drain into Lake Fryxell and the annual total

discharge into the lake is about 4% of the volume of the lake (Hood et al, 1998).

1.1.4. Lakes

Four ice-covered lakes occupy Taylor Valley. The ice sheet extends into the head

of the valley as Taylor Glacier and drains into Lake Bonney, which is divided by a sill

into a west and an east lobes (Fig. 1). At mid-valley the Nussbaum Riegel and Suess

Glacier are followed by Lake Chad and Lake Hoare, which is blocked downvalley by

Canada Glacier. When the amount of water produced through glacial melt is unusually

large, Lake Chad can overflow into Lake Hoare through a 5 m-long spillway across a

moraine left by Suess Glacier (McKnight and Andrews, 1993). Canada Glacier drains

into Lake Fryxell. During the summer, ice along the lakeshore melts, forming an area of

open water around the lake, so called moat. The width of the moat increase with

temperature, input of glacial melt waters and gradient of the lake shore.

The lakes of Taylor Valley have different properties as a result of their different

histories, paleoenvironmental history and current biogeochemical processes. Lake

properties are listed in Table 1 (http://www.lternet.edu/sites/iTicm/). Lake Bonney is

divided into the West and East lobes by Bonney Riegel, a bedrock sill at 12 m depth. The

lake is highly stratified with the chemocline and oxycline at a depth of-15 m in the West

lobe and 20 m in the East lobe. Among the Taylor Valley lakes, bottom waters for both

lobes are the coldest, reaching as low as -5 and -2 °C, respectively (Table 1). The bottom

waters are also the saltiest, with as much as three times the salinity of seawater (Lee et

http://www.lternet.edu/sites/iTicm/

133

al., 2004). The last marine incursion into the valley was during the warm Miocene-

Pliocene epoch, 3 to 5 Ma (Porter, 1981). Fjord-like conditions extended up the valley as

far as Lake Bonney, but sea level have receded since then and Taylor Glacier advanced,

trapping remnants of the marine waters beneath it (Mickuki et al., 2004).

A few data from Lake Chad are currently available. The pH values at 4 and 5.5 m

depths are 7.24 and 7.28, respectively (Lyons et. al., www.lternet.edu/sites/mcm/). The lake

is shallow and loses its entire ice cover during the summer, which means that it is the

only Taylor Valley lake that is not perennially ice-covered (Mcknight and Andrews,

1993). The moraine spillway between Lake Chad Lake Hoare hosts abundant microbial

mats.

Lake Hoare, a fresh water lake, is supersaturated with oxygen to the 25-m depth

(Spigel and Priscu, 1998); this is the most oligotrophic of the Taylor Valley lakes

(Clocksin et al., 2007). Waters in some depressions, below 25 m are anoxic and sulfidic

(Craig etal., 1992).

Of all the lakes, Fryxell receives the most streamwater, which accumulate salts

and nutrients. (Tyler et al., 1998). Lake Fryxell has the highest productivity of the Taylor

Valley lakes (Lizotte and Priscu, 1994), which result in the anoxic and sulfidic bottom

waters (Roberts, 2000).

1.2 Historical evolution of Taylor Valley

Geomorphological and sedimentological studies of drift sheets and moraines, as

well as drill cores on the valley floor by Hall and Denton (2000a) and Hall and Denton

(2000b) has shown that during the last glacial period, the Ross Ice Sheet extended inland,

http://www.lternet.edu/sites/mcm/

damming melt waters from Taylor Glacier in the valley. Radiocarbon dating of lacustrine

deposits and microbial mats in perched deltas yielded ages of 42 to 28 ky BP, which

indicate that the valley was formerly inundated as high as 350 m above sea level

indicated by the delta fronts that formed the Glacial Lake Washburn (Hall & Denton,

2000a; Hall & Denton, 2000b; Stuiver et al, 1981; Wagner et al, 2006). After the last

glacial maximum at 8,300 y BP, lake levels fluctuated greatly as the ice sheet retreated

out of Taylor Valley to its present position (Hall and Denton, 2000b; Hendy, 2000).

Geochemical proxy records for lake level, such as total organic carbon/total sulfur from

Wagner et al. (2006) and halogen geochemistry from Lyons et al. (1998) is concordant

with stratigraphic and geomorphological analyses by Chinn (1993), Hendy (2000),

Stuiver et al. (1981) that indicated intense evaporation from 14,000 to 8,000 y BP and

2,500 to 1,000 y BP. These dessicating events severely reduced lake levels, forming

separate hypersaline ponds where the modern lakes are located (Wagner et al., 2006;

Chinn, 1993). In contrast, Lake Hoare was completely dried up and its salts were blown

away by the wind (Fountain et al., 1999). Re-filling with melt water formed the present

stratified Lake Fryxell and freshwater Lake Hoare (Doran et al, 1994; Hall & Denton,

2000a; Hall & Denton, 2000b). Thus, the current Lake Fryxell is about 10,000 years old,

whereas Lake Hoare is only 1,000 to 3,000 years old (Doran et al, 1994; Lyons et al,

1998; Wagner et al, 2006).

1.2 Archaea in Taylor Valley

16S rDNA clones from a cyanobacterial mat collected from the moat were

amplified and sequenced (Brambilla et al., 2001). Comparison to clone libraries revealed

that the archaeal sequences derive from two methanogenic Euryarchaeota that are

phylogenetically close to Ant 12 Antarctic clone 371 and Methanoculleuspalmolei.

However, staining and microscopic fluorescence specific for Archaea indicated that these

microbes are not abundant. Karr et al. (2006) conducted a molecular diversity study using

lake water and sediments and also revealed archaeal DNA sequences. The DNA cluster

analysis of the sediments also revealed Euryarchaeota. One of the groups identified is

similar to Methanosarcina (Karr et al., 2006). A methylotrophic strain of

Methanosarcina, FRX-1, from Lake Fryxell has recently been isolated by Singh et al.

(2005) although the 16s rRNA of this archaeon is distantly related from the sequences

detected by Karr et al (2006). Another group of Euryarchaeota is related to

Methanoculleus palmolei, which may be similar to the sequences identified by Brambilla

et al. (2001) from a moat mat (Karr et al., 2006). The Euryarchaeota in the deep and

anoxic waters between 14 and 17 m are related to marine Euryarchaeota and several

orders of methanogens. Karr et al. (2006) proposed that the Lake Fryxell water column

Euryarchaeota mediates the anoxic methylotrophy and uses sulfate as electron acceptors.

In contrast to their marine Euryarchaeotal relatives that thrive in oxic waters, the Lake

Fryxell phylotypes flourish in strictly anoxic and highly sulfidic freshwater (Karr et al.,

2006). No DNA from Archaea was detected in lichen- and cyanobacterial-dominated

cryptoendolithic communities (de la Torre et al., 2003).

Methanogenesis and methylotrophy were detected in all the lakes, but these

processes are more intense in Lake Fryxell by at least four orders of magnitude (Smith et

al., 1993; Lawson et al., submitted). In Lake Fryxell, the highest methane concentration

of 9,000 ug L"1 and the highest methane oxidation rates of about 30 ug L"1 day"1 were

observed at the sediment-water interface or in the sediments 15 cm deep. Smith et al.

(1993) measured -76%o for the isotopic composition of biogenic methane. Both Lawson

et al. (submitted) and Smith et al. (1993) report that at 11 m water depths in Lake Fryxell,

methane is below detection limits, which suggests efficient anaerobic methane oxidation.

1.3 Archaeal biomarkers

Core membrane lipids of Archaea are uniquely based on isoprenoid bound on 2,3-

di-0-s«-glycerol (De Rosa and Gambacorta, 1988; Kates, 1978; Kushwaha et al., 1981).

Isoprenoidal glycerol dialkyl glycerol tetraethers (GDGTs) from marine crenarchaeota,

one of the kingdoms of Archaea, have caldarchaeol as the basic structure (Fig. 2, GDGT

[I]) and can contain up to eight cyclopentane rings [II-IV]. In culture experiments of

(hyper)thermophilic Crenarchaeota, the number of rings increases with growth

temperature (De Rosa et al., 1980; Gliozi et al., 1983). In contrast, the addition of a

cyclohexane ring in crenarchaeol [V] and its regio-isomer [VI] is a biosynthetic

adaptation of (hyper)thermophiles to low temperatures as it prevents dense packing of the

membrane (Sinninghe Damste et al., 2002). Biosynthesis of cyclic isoprenoidal GDGTs

[II-IV] and the crenarchaeol isomer is a response to make the membrane more fluid in

low temperature environments (Schouten et al., 2002). The increase in the number of

rings in the GDGT structure is expressed as a ratio, the TEX86 (Schouten et al., 2002):

137

TEXU= i»wwm (1) [//]+[///]+[/n+[w]

This biomarker proxy record correlates linearly with temperature and provides a tool for

reconstructing paleotemperature:

r = (r£X86-0.28)/0.15 (2)

where T is the annual mean sea surface temperature (SST) in °C (Schouten et al., 2002).

Powers et al., (2004) calibrated this index for lakes:

T = (TEXS6-0.29)/0M5 (3)

Recently, branched GDGTs [VII-IX], were observed in terrestrial environments

such as peat bogs, lake and coastal marine sediments (Weijers et al., 2006). The branched

alkyl chains and 1,2-di-O-alkyl-sn-glycerol configuration indicate that the GDGTs are

derived from bacteria rather than from archaea and that this represent an evidence of

lateral gene transfer between Archaea and Bacteria living in consortia (Weijers et al.,

2006). The relative abundance of branched compared to isoprenoidal GDGTs (BIT) is

considered a proxy for terrestrial input into marine environments (Hopmans et al., 2004):

u r _ [VU + VM + IX]

[VII + VIII + IX] + [VI]

This index is important for refining marine SST estimated from TEX86 because terrestrial

organic matter input will also contribute cyclic GDGTs that are accounted in TEXg6 and

increase the calculated temperatures (Weijers et al., 2006).

138

no rv] ^ ^ ' ^ ^ a

13D2

•-Oil

1300

LOB

1298

L-OH

12%

OH

" O i

1292

L OH

[VI] Creaarchacol regio-isomer 1292

Figure 2. Isoprenoidal GDGT

139

HO [VUJ %x--0 .

m/z

1022

•OH

020

018

I0JS6

1034

1032

050

048

1046

Figure 3. Branched GDGTs

140

Weijers et al. (2007) analyzed soils from 90 globally distributed sites for their GDGT

content. The branched GDGTs vary in the number of cyclopentyl moieties [Vllb, VIIc,

VHIb, VIIIc, IXb, IXc] that are expressed as the cyclization ratio of branched tetraethers

(CBT):

CBT = - log f[VIIb ] + [VHIb ] ^

(5) v [VII ] + [VIII ] j

Similarly, an increase in the branching of the carbon chain enhances the

membrane permeability (Weijers et al., 2007). The ratio of branching at position 5 and 5'

of the tetraethers (MBT) was formulated by Weijers et al. (2007), a high value indicating

a low degree of branching.

MBT= yw + vm + rnc]

[VII + Vllb + VIIc] + [VIII + VHIb + VIIIc] + [IX + IXb + IXc] y '

A three dimensional scatter plot of MBT, pH and mean annual temperature (MAT) yields

a calibration equation (Weijers et al., 2007):

MBT = 0.867 - 0.096 * pH + 0.021 * MAT (8)

Incorporating methylation and cyclisatioin into a three-dimensional plot also yields MAT

(Weijers et al, 2007):

MBT = 0.122 + 0 .187* CBT + 0 .020* MAT (9)

141

2.0 METHODS

A total of 36 samples (Table 1), including modern moat microbial mats, lake

bottom sediments, a granite endolith, as well as sediment cores, and a laboratory blank

were analyzed for intact GDGT using the methods of Hopmans et al. (2004). Briefly, the

sediments and microbial mats were sonicated with dichloromethane (DCM), DCM-

methanol (MeOH; 1:1, v/v), and MeOH. The extraction was performed at least three

times for each solvent or until the supernatant turned colorless. All extracts were

combined, dried, weighed, then separated into polar and apolar fractions. For samples

where the original mat, sediments or total extracts are no longer available, archives of

neutral fractions and fractions containing alcohols, ketones and cholesterol were used and

also purified by separating the polar and apolar fractions. The purified polar fractions

were injected into a high-performance liquid chromatography-atmospheric pressure

chemical ionization mass spectrometer

Quantitation for all GDGT species was performed by integrating peak areas in

their total ion chromatograms. Additional HPLC/APCI-MS runs of samples were specific

for identifying cyclic branched GDGTs. Peak areas of crenarchaeol and branched GDGTs

from this run were integrated in selected ion chromatograms m/z 1292, 1050, 1048, 1046,

1036, 1034, 1032, 1022, 1020, 1018. Statistical analyses for the GDGT abundances were

performed in SYSTAT 12. Average linkage and Euclidean distance were used for the

cluster analysis.

142


3.1 GDGT distribution

Archaeal biomarkers are ubiquitous in both modern and ancient sediments (Table

2). GDGT-0 and Crenarchaeol are the major constituents in the sediments of Lakes

Fryxell, Hoare and West Bonney. Lake Chad microbial mats mostly contain branched

GDGTs. Below are detailed descriptions of the GDGT-0, Crenachaeol, and total

branched GDGT distributions.

3.1.1 Cryptoendolith

A community of microorganisms in a granite boulder collected at the top of a

paleodelta above Lake Fryxell contains 51.4% GDGT-0, 5.4% GDGT-l and 43.2%

crenarchaeol and does not contain any branched GDGTs (Table 2). In contrast, culture-

independent DNA surveys of cryptoendolithic communities from sandstone outcrops on

the Battleship Promontory in McMurdo Dry Valleys did not detect archaea (de la Torre et

al., 2003). Only lichen-dominated eukaryal and mostly cyanobacterial-dominated

bacterial phylotypes were identified.

3.1.2 Microbial Mats

Moat microbial mats from Lake Fryxell and Lake Hoare do not contain GDGTs

(Table 2), whereas the two moat mats from Lake Chad contain -94% branched GDGTs,

primarily composed of alkanes with the least branching [VII], and yielded an average of

0.99 BIT units.

143

TABLE 2

Percentages of individual GDGTs relative to the total GDGT in modern samples

Cryptoendolith moat microbial mats Fryxell-29 Hoare-08 Chad-06

Chad-08 Chad ave C h a d s d lake bottom sediments Lake Fryxell

0-10 cm

0-10 0-0.5

2-2.5 3-3.5 5-6

7-7.5

7.5-8 ave sd

greenish mud

quartz sands black S rich

laminated

laminated sand

laminated sand

Lake Ho are 0-0.5 0-1 1-2

2-3 ave

sd

black

Wes t Bonney

0-2.5

3.5-7 9-13

ave sd

sand

[I] GDGT

0 51.4

0.0 0.0 5.1 4.3 4.7 0.6

51.5

53.5 52.0

55.5

67.6 52.7 55.9 52.5 55.1 5.3

38.3 36.1 31.4 41.3 39.1 7.2

0.0

39.0 40.2 39.6 0.8

[II] GDGT

1 5.4

0.0 0.0 0.1 0.0 0.0 0.0

1.9 2.3 0.3

2.5

2.3 2.5

2.3 2.1 2.0 0.7

6.5 5.9 5.7 6.4 6.3 0.4

0.0

5.5 5.9

5.7 0.3

[III] GDGT

2

0.0

0.0 0.0 0.0 0.0 0.0 0.0

0.7

1.2 1.0

1.2

1.0 1.3 1.1 1.0 1.1 0.2

2.3 2.0 2.4 2.4 2.2 0.2

0.0 2.6

2.5 2.5 0.1

[IV] GDGT

3 0.0

0.0 0.0 0.0 0.0 0.0 0.0

0.2

0.5 0.4

0.4

0.0 0.4 0.4

0.4

0.3 0.2

0.6 0.1 0.5 0.5 0.5 0.2

0.0 0.5

0.0 0.2 0.3

[V] Cren 43.2

0.0 0.0 0.0 1.5 0.8 1.0

43.0

41.9 45.8

39.7

29.1 42.5

39.9 43.3 40.7 5.1

49.2 52.9 59.2 49.3 49.3

8.0

0.0 41.0

39.6

40.3 1.0

[VI] Cren

isomer 0.0

0.0 0.0 0.0 0.0 0.0 0.0

1.0

0.6 0.4

0.1

0.0 0.6 0.4

0.2 0.4 0.3

0.2 0.7 0.8 0.1 0.3

0.4

0.0

0.0

0.0 0.0 0.0

[VII] BT

0.0

0.0 0.0 57.9 55.8 56.9 1.5

0.8

0.0 0.0

0.3

0.0 0.0

0.0 0.2

0.2 0.3

1.9 1.6 0.0 0.0 1.4

1.3

0.0 7.0

7.2

7.1 0.1

[VIII] BT 0.0

0.0 0.0

31.4 31.9 31.7 0.4

0.3

0.0 0.0

0.1

0.0 0.0

0.0 0.1

0.1 0.1

0.7 0.5 0.0 0.0 0.5 0.5

0.0 2.4

2.3 2.4 0.1

[IX]

BT 0.0

0.0 0.0 5.4 6.5 5.9 0.8

0.5

0.0 0.0

0.2

0.0 0.0

0.0 0.2

0.1 0.2

0.1 0.1 0.0 0.0 0.1

0.1

0.0 1.9

2.3 2.1 0.3

Total

branched

GDGT 0.0

0.0 0.0

94.7 94.3 94.5

1.6

0.0 0.0

0.5

0.0 0.0

0.0 0.5 0.3 0.6

2.8 2.2 0.0 0.0 2.4

2.0

0.0 11.4

11.8 11.6 0.3

144

Only one stream drains into Lake Chad, which receives its water directly from

Suess Glacier. If streams transport branched GDGTs, then Lake Fryxell, which drains the

most number and longest streams should have abundant branched GDGTs. If glacial

microbial communities, or sediments eroded by glaciers, or aeolian input significantly

contribute to the branched GDGT pool in the lakes, then all the lakes should have high

branched GDGTs. None of the other lakes, however, have branched GDGT abundances

as high as Lake Chad. Therefore, branched GDGTs are produced in situ in Lake Chad

moat microbial mats.

3.1.3 Lake bottom sediments

GDGT-0 relative abundances in Lake Fryxell bottom sediments, are the highest in

this study, averaging 55.1% (Table 2 and Fig. 4A). The highest percentages are

associated with laminated deposits, whereas sandy sediments generally have lower

abundances. Lake Hoare bottom sediments, including the core top sediments, have an

average GDGT-0 of 39.1%. Sediments from the top of core site A, an anoxic depression

below 21 m, yield 51%, the highest value for the lake (Table 3). West Lake Bonney

sediments between 0 to 2.5 cm does not contain GDGT, but samples between 3.5 to 7 cm

and between 9 to 13 cm have similar values and average 39.6%.

For crenarchaeol [V], the mean relative abundances in Lake Fryxell and West

Lake Bonney are 40.7% and 40.3%, respectively (Table 2, Fig. 4B), and lower than those

from Lake Hoare that yield an average of 49.3%. The crenarchaeol isomer has low

abundances and does not correlate with any of the GDGT moieties.

145

% of Total GDGT

GDGT-0 Ave «

20 40 60 80 100 0 -

4

9

]?.

- • — —

• on

GDGT-0

•

* # HH

4 .

th (

cm)

•

ft Q

12.

0 — » - wkggpa-j-* » — —

Crenarchaeol

«

•

Cren Ave

i_t"pi

o«#

12 Branched

GDGT AVE I

LEGEND Q Fryxell A Hoare 0 Chad • West Bouncy ^ Hoarc core tops

Figure 4. GDGT distribution for lake bottom sediments and a Lake Chad microbial mat

146

TABLE 3

Percentages of individual GDGTs relative to total GDGT in core samples

Lake Fryxell 46

61.5

62.5

ave

s.d.

331

471

631

653

ave

s.d.

mid-Holocene

early Holocene

early Holocene

mid-Holocene

mid-Holocene

late Pleistocene

late Pleistocene

late Pleistocene

late Pleistocene

late Pleistocene

late Pleistocene

Lake Hoare

core A

0-0.5

0-1

8-9

12-13

core B

0-1

1-2

3-4

5-7

BLANK

laminated

oxic

[I] GDGT

0

. 48.2

43.5

44.4

45.4

2.5

28.1

29.9

34.7

30.2

30.7

2.8

30.9

50.6

23.9

32.6

45.4

31.5

n.d.

n.d.

n.d.

[II] GDGT

1

5.5

5.2

6.3

5.7

0.6

5.0

5.8

6.4

5.8

5.7

0.6

5.6

6.9

5.2

5.6

6.4

4.9

n.d.

n.d.

n.d.

[Ill] GDGT

2

2.3

2.0

2.0

2.1

0.2

1.8

2.8

2.3

2.2

2.3

0.4

2.2

2.0

2.4

n.d.

2.0

2.1

n.d.

n.d.

n.d.

[IV] GDGT

3

0.4

0.3

0.4

0.4

n.d.

0.6

0.8

0.7

0.6

0.7

0.1

0.5

0.5

2.7

n.d.

0.5

1.1

n.d.

n.d.

n.d.

[V] Cren

42.0

44.6

43.8

43.5

1.4

47.0

49.9

52.4

51.8

50.2

2.4

55.6

35.1

31.5

34.1

44.0

52.4

n.d.

n.d.

n.d.

[VI] Cren

isomer

0.7

0.8

0.7

0.7

n.d.

1.1

0.4

1.0

1.4

1.0

0.4

0.2

0.0

0.7

n.d.

n.d.

0.6

n.d.

n.d.

n.d.

[VII] bGDGT

0.3

2.2

0.7

1.0

1.0

11.1

7.8

1.6

5.9

6.6

4.0

3.4

3.4

22.4

20.6

1.1

5.0

n.d.

n.d.

n.d.

[VIII] bGDGT

0.2

0.5

0.5

0.4

0.2

3.1

1.5

0.5

1.4

1.6

1.1

1.4

1.3

9.1

7.1

0.5

1.9

n.d.

n.d.

n.d.

[IX] bGDGT

0.5

0.8

1.2

0.8

0.4

2.2

1.0

0.5

0.9

1.1

0.7

0.2

0.1

2.2

n.d.

0.1

0.5

n.d.

n.d.

n.d.

Total branched

GDGT

0.9

3.5

2.3

2.2

1.3

16.4

10.3

2.6

8.1

5.3

5.7

5.0

4.9

33.6

27.7

1.7

7.4

n.d.

n.d.

n.d.

n.d not detected

Branched GDGTs only comprise as high as 12% and is highest in Lake Bonney

sediments (Fig. 6C) that amount to 0.2 BIT units (Table 2, Fig. 5A). These biomarkers

preserved in the Lake Bonney sediments are likely derived from the watershed, in

contrast to in situ production in Lake Chad. Lake Hoare and Lake Fryxell contain minor

amounts of branched GDGTs and only yield less than 0.05 BIT units, considering that

Lake Fryxell receives the most number of streams that drain a relatively large area.. The

abundances of the three cyclic branched GDGTs exhibit an inverse correlation with

crenarchaeol (R2[Vii]=-0.88, R2

[Vm]=-0.89, R2[IX]=-0.87), whereas these branched GDGTs

co-vary with GDGT-0 (R2[Viif0.81, R2

[Viii]=0.82, R2[IX]=0.81). This indicates that the

shallow water environments where bacteria that produce branched GDGTs thrive are not

optimum conditions for aquatic Crenarchaeota.

3.1.4 Core Sediments

Lake Fryxell core sediments

Core sediments from Lake Fryxell have ages in the Holocene for depths 46, 61,

and 62 cm, and late Pleistocene for depths 331, 471, 631, and 653 cm (Wagner et al.,

2006). In the core, GDGT-0 have abundances less than the average of modern deposits

and are distinct for the Holocene and late Pleistocene sediments. Holocene deposits

average 45.4%, whereas late Pleistocene deposits average 30.7%, about 15% less than

modern GDGT-0 abundances (Table 3, Fig. 6A). This magnitude of decrease in GDGT-0

in a purely microbial community will require a significant decrease in contributions from

methanogens. Oxic conditions are inhospitable to methanogens, and therefore Lake

Fryxell might have been more oxygenated compared to the present conditions.

BIT/MBT 0.2 0.4 0 -B**A > L

0.6 -a j&A—

2 J

S

a O

H D

6 H

10 A

12 •

AVE BIT

CBT/T

1.0

" O T T

a 0.05 A 0.07

••*0.22

A 1.52/-17° 00.42/3.3° n i 15/-3 7"

01.65/-18 — T — — | — ' -\

o-99f M B T

CBT CBT 0.5 1.0 1.5

r— 2.0

0.6 0.8

2.5

1.0

0

2

a

10

12

AVE TEXJT

— ^ f r - ^ a - a - B - J — — . ——- * - a—-—

O DD

0

TEX 86

0.3/1.2\ 00.33/2.8° 0.33/2.4

SO.5/13.90

Temperature Q 1Q 2Q 3Q 40

—i 1—

50 60

LEGEND D Frvxell A Hoare O Chad 4> West Bonney ~ Hoarc core tops

gure 5. TEXg6 and BIT ratios for lake bottom sediments and Lake Chad microbial mat

149

Crenarchaeol abundances in the core are similar to modern. In contrast, branched

GDGTs are more abundant in the sediment core, particularly for the late Pleistocene

deposits, which have as much as 16.4% (Table 3). Nevertheless, the highest abundance in

the core only translates to 0.26 BIT units.

Total organic carbon/total sulfur (TOC/TS), proxy record for lake levels, from

bulk geochemical analyses by Wagner et al. (2006) indicated that in the Pleistocene when

the Ross Ice Sheet dammed the melt waters in the valley and formed Glacial Lake

Washburn, lakes levels were much higher than present and the waters were fresh. This

condition is more similar to present-day unstratified Lake Hoare. High abundances of

branch GDGT in sediments 331, 471, and 653 cm deep also contain low TOC/TS, which

indicate lowstands during Glacial Lake Washburn stage. Low lake levels also mean that

there is dessication in the watershed, which will not be hospitable to soil organisms that

would be in a state of anhydrobiosis. In contrast, a relatively shallow, fresh and

oxygenated Lake Washburn can be a suitable habitat for organisms that biosynthesize

branched GDGTs.

Lake Hoare core sediments

Sediments in core B between 3 to 4 and between 5 to 7 cm does not contain any GDGT.

However, sediments in core A between 8 to 9 cm and 12 to 13 cm have a total of-30%

branched GDGTs, abundances that translate to about 0.5 BIT units (Table 3). This

abundance is much higher compared to present-day Lake Hoare sediments. The

proximity of Canada Glacier and drainage of Lake Chad waters, as well as soils and

perched deltas in the watershed could introduce branched GDGTs to Lake Hoare.

150

20 30

A GDGT 0 % 40 50 60 70

Modern AVE

a o

28

Modern AVE

-A- 1 i — E 3 — '

0 -f—

2 J _

5 +—

...„,-,..., .......r...,„m...... „ ,., 48. .,.,„-, ...rr r„,T...„.m..„r..„.,..m.„,rvr„

*

— i pr~~^^ B Crenarchaeol %

35 42 49

C Branched GDGT %

7 14 21 28 35

Modern fkaJJL. AVE B , , A I

JBw_A-A- — ———T0r——^

——?* — ' — •

- _ _ ^ 1 _ _ _ _

LEGEND • Fryxell core A Hoare ^ West Bonney D Fryxell • Hoare core tops

Figure 6. GDGT distribution in core sediments and averages in lake bottom sediments

151

Nevertheless, we cannot rule out in situ production of these biomarkers if lake wide

biogeochemical changes occurred to allow the organisms that biosynthesize branched

GDGTs to thrive in Lake Hoare.

3.2 Temperature estimate

3.2.1 Lake Chad

Of the moat mats, only Lake Chad 06 was amenable for TEXs6 calculation, which

yields 0.33 units (eq. 1; Table 4). Using TEX86and the existing linear relationships, the

temperature calibration for lakes by Powers et al. (2004) converts the ratio to 2.8 °C (eq.

3) and calibration for oceans by Schouten et al. (2002; eq. 2) yield 3.5 °C, close to 4.6 °C

from measured temperature during sampling (Table 4). This is the only available data for

temperature, but we can infer that during the melt season, the

water temperature would have to be at least 0 °C until the ice cover is lost, and thereafter

increase to 4.6 °C. The TEX86 estimates can represent the spring to summer temperatures,

when conditions are viable for growth.

Using the calibration by Weijers et al. (2007), the BIT value 0.99 converts to

-17.9 °C (eq. 3). This temperature is within the range reported for annual MAT for

Taylor Valley. Moats form and metabolic activity start early in the spring and into the

summer. Biosynthesis of branched GDGTs in the moat mats likely represent spring to

summer air temperatures. We do not know, however, the influence of the water

temperature and the interaction between air and water temperatures on the estimates from

branched GDGTs.

152

TABLE 4

Temperature estimates from TEXg

Chad-06 moat mat

TEX8 6

0.33 Lake Fryxell bottom sediments (depths green silts 0-10 quartz sands 0-10 black mud 0- 0.5 sandy silts 2-2.5 laminated 3-3.5

sandy 5-6 sandy silts 7-7.5

sandy 7.5-8

ave

0.52 0.50

0.89 0.40 0.31 0.49 0.45

0.44

0.50

Lake Hoare bottom sediments (depths black mud 0-0.5

0-1 1-2

2-3 3

ave

0.29 0.31 0.37 0.29

0.31 West Lake Bonney bottom sediments 1

3.5-7 9-13

ave Lake Fryxell core sediment:

46

61.5 62.5

mid Holocene ave 331 471 631 653

Pleistocene ave Lake Hoare core sediments core A (local depression)

0-0.5 0-1 8-9

core B (21 m water depth) 0-1 1-2

0.36 0.29

0.33

T ( ° C ) Lake

calibration

2.8 i in cm)

15.3 14.0

40.2 7.0 1.3

13.2 10.4

10.0

13.9 in cm)

0.1 1.2 5.0 0.1

1.2 ^depths in cm)

4.6 0.2

2.4 s (depths in cm)

0.38

0.40 0.40 0.39 0.42

0.37 0.37 0.41 0.35

6.1

7.5 7.2

6.9 8.4

5.4

5.6 8.3 6.9

(depths in cm)

0.31 0.26 0.51

0.27 0.37

1.6 -2.2 14.5

-1.4 5.3

T (°C)

Ocean

calibration2

3.5

15.9 14.6

40.9 7.7 1.9

13.9 11.1

10.6

14.6

0.8 1.9 5.7 0.8

1.9

5.3 0.9

3.1

6.8

8.1 7.8

7.6

9.1 6.0 6.3 9.0 7.6

2.3 -1.6 15.1

-0.7 6.0

T ( ° C )

instrumental

record

4.5*

2.2§

0.36§

2.1+ , 1.5*

1 T = (TEX86 - 0.29)/0.01 (Powers et al, 2004) 2 T = (TEX86 - 0.28)/0.01 (Schouten et al, 2002)

average includes Lake Hoare core A 0-0.5 and 0-1 and core B 0-1 and 1-2 cm * water temperature at 1 m depth during sampling § average measured water column temperatures for 1993-1999 t average measured water column temperature for 1993-1998 { average measured water column temperatures up to 20 m, depth of oxycline, for 1993-1998

TABLE 5

Temperature estimates from branched GDGTs

BIT1 BIT2

Lake Chad moat microbial mats Chad-06 1.00 1.00 Chad-08 0.98 0.99

ave 0.99 0.99 Lake Fryxell bottom sediments

0-10 0.04 2-2.5 0.01 7-7.5 0.05 7.5-8 0.01

ave 0.02

MBT

0.06 0.07

0.24

CBT

1.65 na

1.15

pH

4.4 na

5.7

MAT 3

( °Q

-18.6 na

-4.9

MAT4

(°C)

-18.3 na

-3.7

Lake Hoare bottom sediments 0-0.5 0.05 0.05 0.05 0-1 0.04 2-3 0.03 0.04

ave5 0.09 0.07 West Lake Bonney bottom sediments

3.5-7 0.22 9-13 0.23 0.27 0.20

Lake Fryxell core sediments

1.30

1.08

0.42

5.3

5.9

7.7

46 61.5 62.5 ave 331 471 631 653

ave

0.02 0.07 0.05 0.05 0.26 0.17 0.05 0.14 0.2

Lake Ho re core sed A 0-0.5 A 0-1 A 8-9 A 12-13 B l - 2

0.08 0.12 0.52 0.45 0.12

0.08

0.52

0.12

0.04

0.07

0.07

2.13 3.17

1.56 4.65

-15.7

-14.1

-17.8

0.1

-14.5

-12.2

-17.1

3.3

-24.09 -25.01

-17.14 -16.63 Branch Index of Tetraethers from rapid scan of GDGTs Branch Index of Tetraethers from SIM mode scan for branched GDGTs

3 annual MAT=[MBT-0.867+(0.096* pH)]/0.021 (Weijers et a l , 2007) 4 annual MAT=[MBT~ 0.122-(0.187* CBT)]/0.020 (Weijers et al., 2007)

Lake Hoare average temperature estimate includes cores A 0-0.5, A 0-1 and B 1-2 cm

154

3.2.2 West Lake Bonney

The average TEX86 value from the sediments in Lake Bonney is 0.33, similar to

average values from Lake Chad and Lake Hoare, but less than those in Lake Fryxell (Fig.

6A). Temperature estimates from the calibration equation by Powers et al (2004) has an

average of 2.4 °C. similar to the average water column temperatures of 2.1 °C that was

measured from 1993 to 1998. From the branched GDGTs, the BIT value is 0.27. The

cannot be ascertained whether these are from the watershed, produced in situ, or

associated with Taylor Glacier. The similarity of TEX86-derived temperature with that of

the instrumental record indicate that even with a 0.27 BIT value, the isoprenoidal GDGTs

that may be associated with the branched GDGTs are minimal and do not affect the water

temperature estimate. This observation is consistent with the inverse correlation between

the branched GDGTs and crenarchaeol, and will indicate that the sources of these

GDGTs are not the same. The BIT value translates to 3.3 °C, which is an overestimate

and may indicate conditions, such as pH, that have greater influence on the cyclization or

branching during biosynthesis of the GDGTs.

3.2.3 Lake Fryxell (paleo)temperature estimates

The range of TEX86 values in the upper 8 cm of the lake bottom sediments is

0.31-0.89 units (Table 4, Fig. 5A). TEX86 values average 0.5, yielding 11 °C (eq. 2), an

overestimate of measured temperatures in the water column. The highest TEX86 value is

from the black surface sediments that also necessitated removal of elemental sulfur.

Methanogens probably contributed significant GDGTs to offset the TEX86 to high values

and biased temperature estimates to higher values for Lake Fryxell.

155

In the core, TEX86 averages 0.39 for both Holocene and late Pleistocene

sediments, less than the average for the surface sediments. The calculated

paleotemperatures average 7.6 °C. Using BIT values, and considering that the branched

GDGTs are allocthonous biomarkers, the mean air annual temperature is -4.3 °C, which

is close to the average summer temperature of -3 °C for Taylor Valley (Margesin and

Schinner, 1999). Most metabolic activities in the valley occur in the summer, therefore

the MAT represent average summer temperatures.

3.2.4 Lake Hoare (paleo) temperature estimates

Average TEX^ value is 0.30 in the lake bottom sediments, including core tops (0-

0.5 and 0-1 cm) from sites A and B, translates to 1.1 °C from calibrations by Powers et al

(2004; Table 2). This temperature estimate is close to the Lake Hoare average of 0.3 °C

and the temperature estimate is not overprinted by bacterial tetraethers because branched

GDGTs have low abundances.

In core A, however, sediments between 8 to 9 cm a contain abundant branched

GDGT that is equivalent to 0.51 BIT units and translates to 15.1 °C, an overestimate that

might be due to contributions from biosynthesizers of branched GDGTs and not due to

contributions from methanogens because GDGT-0 decreased in relative abundance.

Mean annual air temperatures calculated from branched GDGTs (eq. 7 and 8) has an

average of-14.1 °C and is within the range for Taylor Valley.

156

3.3 Cluster analysis of present-day materials

Cluster analysis based on relative GDGT distribution was performed to determine

if the groupings have significance to environmental conditions. Only percentages from

the isoprenoid GDGTs were used because the branched GDGTs are scarce in other

samples. Indeed the cluster tree shows major groupings according to the lakes. This is

consistent with their different geochemical environments today that would also promote

different archaeal groups to flourish. Repeated cluster analyses using different linkage

and distance measurements showed that the groupings are significant. The clusters are

terrestrial Lake Chad-type versus the aquatic non-stratified Lake Hoare-type, aquatic

stratified hypersaline bottom water Lake Bonney-type, and aquatic stratified brackish

bottom water Lake Fryxell-type (Fig. 5).

3.3.1 Terrestrial-type

The Lake Chad-type consists mostly of branched GDGTs, characteristic of the

terrestrial distribution, with BIT indices close to 1.0. Other lakes with high BIT indices

are 0.93 for Lake Paloma in Chile, and 0.91 for Siso Lake in Spain (Hopmans et al.,

2004), probably due to autochthonous production. A 0.99 to 1.0 BIT value is common for

peats, soils and wetlands (Hopmans et al., 2004; Weijers et al., 2006). A low BIT value of

0.4 in the marine environment is considered by Weijers et al. (2006) as indicating a

significant input of terrestrial GDGTs because soils have much lower concentrations.

157

Typ

e of

di

stri

buti

on

C-m

at-0

6 C

-mat

-08

H-b

otto

m-1

-2 c

m

H-c

oreA

-0-0

.5cm

H

-cor

eB-l

-2cm

H

-bot

tom

-0-1

cm

. H

-bot

tom

-0-0

.5cm

H

-bot

tom

-2-3

cm

B-b

otto

m-9

-13c

m

B-b

otto

m-3

.5-7

cm

H-c

oreB

-O-l

cm

F-b

otto

m-0

-0.5

cm

F-b

otto

m 1

8-0-

10c

m

F-b

otto

ml

8.25

-0-1

0cm

F-

gra

nite

end

olit

h F

-bot

tom

-2-2

.5cm

H

-cor

eA-0

-1 c

m

F-b

otto

m-3

-3.5

cm

Cha

d

Hoa

re

GD

GT

-0

31

Frv

xell

10

20

Dis

tanc

es

30

39

J B

onne

y 39

52

56

68

%

%

Cre

n B

ranc

hed

1

57

50

41

43

40

29

94 4 2 11

0 0

Env

iron

men

t an

d se

dim

ent

type

Ter

rest

rial

51

£ s

Subo

xic

t\a Oxi

c

I A

nox

ic,

hype

rsal

ine

bott

om

wat

er

A n

oxi

c,

brac

kish

bo

ttom

w

ater

(s

andy

se

dim

ents

)

(sil

ty t

o m

uddy

)

met

hano

gen-

rich

Fig

ure

7. C

lust

er a

nal

ysis

of

GD

GT

dis

trib

utio

n an

d th

e as

soci

ated

typ

es o

f en

viro

nm

ent

158

3.3.2 Aquatic non-stratified

The Lake Hoare archaeal distribution type has minimal branched GDGTs, but

more of crenarchaeol than GDGT-0. The -50% crenarchaeol typifies marine

environments (Weijers, 2006), specifically, mesopelagic/upwelling areas (Turich et al.,

2007). In the fresh, only slightly stratified and oxygenated waters of Lake Hoare, this

archaeal distribution is produced by Crenarchaeota, which thrive in both oxic and anoxic

environments and actively remineralize organic matter throughout the water column

(Weijers et al., 2006). Moreover, lacustrine and terrestrial crenarchaeota are

phylogenetically related to marine Crenarchaeota (Jurgens et al., 1997; Ochsenreiter et

al., 2003) and share a preference for alkaline conditions (Pearson et al., 2004; Weijers et

al., 2006). Lake Hoare is the most alkaline of the Taylor Valley lakes, and its oxic

conditions support the success of Crenarchaeota over that of Euryarchaeota.

The two subgroups in Lake Hoare that are not obviously different, but differ

subtly in the abundances of GDGT-0 (Fig. 7). One subgroup has a low average GDGT-0

of 31.3% and the highest crenarchaeol abundances in Taylor Valley. This distribution

indicates more abundant Crenarchaeota than methanogenic or methylotrophic

Euryarchaeota, thus the local environment is oxic. The other subgroup is suboxic with a

slightly higher average GDGT-0 of 39%, suggests higher contributions from

Euryarchaeota. The tops of cores, 0-1 cm, A and B cluster with Lake Fryxell because the

site is in a local depression that is suboxic to anoxic.

159

3.3.3 Aquatic and stratified-types

Stratified lakes, like Lake Fryxell and Lake Bonney, which are neighboring

clusters, allow for anoxic bottom waters even when surface waters are supersaturated in

oxygen. GDGT-0 is the dominant lipid of the Lake Fryxell-type, which has negligible

branched GDGTs. The fresh upper and brackish lower waters, as well as abundance of

methanogens support the similarity of Lake Fryxell to the GDGT distribution in

fresh/estuarine group demarcated by cluster analysis for different marine environments of

Turich et al. (2007).

Two sediment subgroups can be differentiated according to GDGT-0 abundance.

The first consists of sandy sediments, with an average GDGT-0 abundance of 52%, close

to typical estimates of 50% of Crenarchaeota in marine environments (Weijers et al.,

2006). Karr et al. (2006) recovered from waters close to the oxycline 9 m deep

Crenarchaeote sequences similar to those from the Crenarchaeota marine benthic group

C and speculated that the Lake Fryxell Crenarchaeota uses sulfur as electron donor or

acceptor.

The second subgroup has 58% GDGT-0 average abundance, the highest in Taylor

Valley. Moreover, their association with varved lake deposits and possibly benthic

microbial mats indicate that sediments of this group receive substantial contributions

from the Group II Euryarchaeota methanogens that biosynthesize GDGT-0 (Schouten et

al., 2000; Weijers et al., 2006). The association of GDGT-0 with Euryarchaeota is

consistent with the results of Karr et al. (2006) who amplified Euryarchaeota DNA from

Lake Fryxell waters 11, 14, and 17 m deep, but none in shallower depths. The

Methanosarcina and Methanoculleus phylotypes are associated with methanogens. These

160

researchers concluded that anoxic methanotrophy also occurs in Lake Fryxell with sulfate

as the likely oxidant and is probably mediated by the water column Euryarcheaota.

Stable isotope analyses from Smith et al. (1993) and Lawson et al. (unpublished data)

indicate methanogenesis occurring in Lake Fryxell. Methane concentration and methane

oxidation rates are up to four orders of magnitude higher compared to other lakes in

Taylor Valley (Lawson et a l , unpublished data).

The Lake Bonney-type cluster will represent a stratified with hypersaline and

anoxic bottom waters. Similar amounts of GDGT-0 and crenarchaeol, indicates nearly

equal contributions from Crenarchaeota and Euryarchaeota. This distribution is similar

to the epipelagic zone group by Turich et al. (2007). The proximity of the Taylor Glacier

and the supply of meltwater create water saturated sediments and shallow freshwater

pools at the head of the valley and in the surrounding areas to support organisms that

biosynthesize the branched GDGT that eventually become a significant part of the

sediment GDGT pool.

3.4 Cluster Analysis of present-day and ancient sediments

The late Pleistocene deposits from Lake Fryxell cluster in the Lake Hoare-type of

GDGT distribution (Fig. 8) is consistent with freshwater conditions deciphered from bulk

geochemical and sedimentological analyses by Wagner et al. (2006), who also suggested

that changes in the TOC/TS ratio are due to fluctuating lake level and salinity. The

sediments from these deposits were specifically chosen to test if the biomarker data can

document these environmental changes. The LZ1021 5i-331, 5ii-471, and 6ii-622.5 cm

deposits cluster with Lake Hoare oxic distribution, have relatively higher BIT values and

161

O-nwi-M Cr-njiU-QS

SlippiSSpll ;^^^^^^S IM^^^^^^w^^K t'"'' it-corcU l-2cm H-CPU'V0-0.5un ll-surf l-2cm « ^ ^ ^ ^ * ^ P T H-surf-O-lcm H-smf-0-0„'iiT,t H-&urf-2-3cm B-suri'-l>-13iiii

•K^SftS H-cureB-0 H-cure&O

F-emlplilh l--surl-0-0.5cii)

•F-surf!8-CM0em f-sutf-7.5-8cin F-surf-5-6cm •F-suifUS.25-0-10cm F-surt"-2-2.5cm F-surf-7-7.5cm

•F-surf-3-3.5cm

E«v ironnient and sediment

type

Terrestrial

'5

1 •?

-

-3 '£: 3 5 5 5

--

1 1 3 o

Oxic

Suboxic

" 1 Anoxic, hypemiline bottom water

' Anoxic, brackish bottom water

• | f.tilty to muddy)

-1 metlianogen-rich

10 20 Distances

30

Figure 8. Cluster analysis of lake bottom and core sediment GDGT distribution

162

are also the same deposits that record low TOC/TS reported by Wagner et al. (2006)

suggesting late Pleistocene lowstands of Glacial Lake Washburn. Conversely, the deposit

6i-601 cm clusters with the suboxic Lake Hoare subgroup, have a low BIT value, and

also records a highstand. Even with the water column supersaturated with oxygen down

to 25 m today, much higher lake levels can easily get suboxic conditions in the deeper

parts of the water column.

The Holocene deposits 2ii-62.5 and 3ii-61.5 cm cluster with the Lake Bonney-

type, suggesting a highly stratified lake with hypersaline and anoxic bottom waters. This

deposit also contains high total sulfur that started to accumulate 10,000 y cal BP and

reached maximum abundances around 7,500 cal y BP (Wagner et al., 2006). Both

biomarker and geochemical analyses are consistent with the intense evaporation event

around 8,000 y BP (Stuiver et al., 1981; Hendy, 2000a). The mid-Holocene deposits, 2i-

46 cm from Lake Fryxell cluster with the Fryxell-type, thus suggesting a brackish, yet

stratified lake, similar to the current Lake Fryxell conditions today.

4.0 CONCLUSION

Archaea are a ubiquitous component of the microbial assemblage in Taylor

Valley lakes. In environments, such as Taylor Valley, where primary production is

limited to a few months of the year, the presence of chemolithotrophs can significantly

influence the carbon cycle. The lake water temperatures calculated using TEX86, an index

based on isoprenoidal GDGT ratios, overestimate measured temperatures in Lake Fryxell

because GDGTs contributions by methanogens become prominent as this is exclusively a

163

microbial system. Average water column temperatures from Lakes Hoare, Chad and

West Bonney are similar to mean instrumental records from 1993 to 1999. Annual mean

air temperatures estimated from branched GDGTs closely approximate the measured

values, particularly for Lake Chad where the lake loses ice cover during summer and

where branched GDGTs are produced in situ.

The granite endolithic community, which only contains isoprenoidal GDGTs, also

indicate that the lichen-dominated and cyanobacterial-dominated communities contain

archaea. The present-day archaeal distributions as stratified hypersaline or brackish, oxic

and suboxic freshwater and terrestrial types correlate well with bulk geochemical proxy

records.

The distribution of GDGTs in Taylor Valley lakes seems to be controlled by

reduction-oxdation conditions in the bottom sediments. Anoxic conditions in the bottom

water are induced by salinity and stratification and presence or absence of ice cover. For

archaea, the contribution of Euryarchaeota is prominent in the total GDGT pool. Thus

the archaeal biomarkers are sensitive proxy records for a lake system responding to

climatic and environmental changes.

5.0 ACKNOWLEDGEMENTS

Support to C. Jaraula was granted by the Office of the Provost at the University of

Illinois at Chicago, and by a fellowship from the Institute of Environmental Science and

Policy. We thank T. Chung, A. Hilegass and Z. Mateo for help in the laboratory. K.

Rodolfo is acknowledged for editing the manuscript.


Booth, I.R.: Regulation of cytoplasmic pH in bacteria. Microbiol. Rev. 49: 359-378, 1985.

Brambilla, E., Hippe, H., Hagelstein, A., Tindall, B.J., and Stackebrandt, E.: 16S rDNA diversity of cultured and uncultured prokaryotes of a mat sample from Lake Fryxell, McMurdo Dry Valleys, Antarctica. Extremophiles 5: 23-33, 2001.

Bromley, A.M.: Weather observations, Wright Valley, Antarctica. New Zealand Meteorological Service, 1985.

Chinn, T.J.: Physical hydrology of the dry valley lakes. In: Physical and Biogeochemical Processes in Antarctic Lakes, 59: eds. W.J. Green, and E.I. Friedmann, pp. 1-51. Antarctic Research Series, American Geophysical Union. 1993.

Clocksin, K.M., Jung, D.O., and Madigan, M.T.: Cold-active chemoorganotrophic bacteria from permanently ice-covered Lake Hoare, McMurdo Dry Valleys, Antarctica. Appl. and Environ. Microbiol. 73: 3077-3083, 2007.

Clow, G.D., McKay, C.P., Simmons, G.M.J., and Wharton, R.A.: Climatological Observations and Predicted Sublimation Rates at Lake Hoare, Antartica. J. of Climate 1:715-727, 1988.

Craig, FL, Wharton, R.A., and Mckay, C.P.: Oxygen Supersaturation in Ice-Covered Antarctic Lakes - Biological Versus Physical Contributions. Science 255: 318-321, 1992.

Damste, J.S.S., Schouten, S., Hopmans, E.C., van Duin, A.C.T., and Geenevasen, J.A.J.: Crenarchaeol: the characteristic core glycerol dibiphytanyl glycerol tetraether membrane lipid of cosmopolitan pelagic crenarchaeota. J. of Lipid Res. 43: 1641-1651,2002.

de la Torre, J.R., Goebel, B.M., Friedmann, E.I., and Pace, N.R.: Microbial diversity of cryptoendolithic communities from the McMurdo Dry Valleys, Antarctica. Applied and environmental microbiology 69: 3858-3867, 2003.

de Rosa M., and Gambacorta, A.: The lipids of archaebacteria. Progress in Lipid Res. 27: 253-175, 1988.

Doran, P.T., Berger, G.W., Lyons, W.B., Wharton, R.A., Davisson, M.L., Southon, J., and Dibb, J.E.: Dating Quaternary lacustrine sediments in the McMurdo Dry Valleys, Antarctica. Palaeogeog. Palaeoclimatol. Palaeoecol. 147: 223-239, 1999.

165

Doran, P.T., McKay, C.P., Clow, G.D., Dana, G.L., Fountain, A., Nylen, T., and Lyons, W.B.: Valley floor climate observations from the McMurdo Dry Valleys, Antarctica, 1986-2000. J. of Geophys. Res. 107: (D24), 4772, doi:10.1029/2001JD002045, 2002.

Doran, P.T., Wharton, R.A., Jr., and Lyons, W.B.: Paleolimnology of the McMurdo Dry Valleys, Antarctica. J. of Paleolimnol. 10: 85-114, 1994.

Foreman, CM., B, S., JA, M., DL, P., and Priscu, J.C.: Metabolic activity and diversity of cryoconites in the Taylor Valley, Antarctica. Geophys. Res. 112: doi: 10.1029/2006JG000358, 2007.

Fountain, A., Lyons, W.B., Burkins, M.B., Dana, G.L., Doran, P.T., Lewis, K.J., McKnight, D., Moorhead, D.L., Parsons, A.N., Priscu, J.C., Wall, D.H., Wharton, R.A.J., and Virginia, R.A.: Physical controls on the Taylor Valley Ecosystem, Antarctica. Bioscience 49: 961-972, 1999.

Fritsen, C.H., Grue, A.M., and Priscu, J.C.: Distribution of organic carbon and nitrogen in surface soils in the McMurdo Dry Valleys, Antarctica. Polar Biol. 23: 121-128, 2000.

Garrity, G.: Bergev's manual of systematic bacteriology, New York, 2001.

Gliozzi, A., Paoli, G., DeRosa, M., and Gambacorta, A.: Effect of isoprenoid cyclisation on the transition-temperature of lipids in thermophilic archaeabacteria. Biochim. Biophvs. Acta 735: 234-242, 1983.

Hall, B.L., and Denton, G.H.: Extent and chronology of the Ross Sea ice sheet and the Wilson Piedmont Glacier along the Scott Coast at and since the last glacial maximum. Geografiska Annaler Series a-Phvsical Geog. 82A: 337-363, 2000a.

Hall, B.L., and Denton, G.H.: Radiocarbon chronology of Ross Sea drift, eastern Taylor Valley, Antarctica: Evidence for a grounded ice sheet in the Ross Sea at the last glacial maximum. Geografiska Annaler 82A: 305-336, 2000b.

Heijs, S.K., Haese, R.R., van der Wielen, P.W.J.J., Forney, L.J., and van Elsas, J.D.: Use of 16S rRNA gene based clone libraries to assess microbial communities potentially involved in anaerobic methane oxidation in a Mediterranean cold seep. Microbial Ecol. 53: 384-398, 2007.

Hendy, C.H.: Late Quaternary lakes in the McMurdo Sound region of Antarctica. Geografiska Annaler Series a-Phvsical Geog. 82A: 411-432, 2000.

Hopmans, E.C., Weijers, J.W.H., Schefuss, E., Herfort, L., Damste, J.S.S., and Schouten, S.: A novel proxy for terrestrial organic matter in sediments based on branched and isoprenoid tetraether lipids. Earth and Planet. Sci. Lett. 224: 107-116, 2004.

166

Hood, E.M., Howes, B.L., and Jenkins, W.J.: Limnol. and Oceanography 43: 265-272, 1998.

Jurgens, G., Lindstrom, K., and Saano, A.: Novel group within the kingdom Crenarchaeota from boreal forest soil. Appl. and Environ. Microbiol. 63: 803-805, 1997.

Karr, E.A., Ng, J.M., Belchik, S.M., Sattley, W.M., Madigan, M.T., and Achenbach, L.A.: Biodiversity of methanogenic and other Archaea in the permanently frozen Lake Fryxell, Antarctica. Appl. and Environ. Microbiol. 72: 1663-1666, 2006.

Kates, M.: The phytanyl ether-linked polar lipids and isoprenoid neutral lipids of extremely halophilic bacteria. Prog, in Chem. Fats and Other Lipids 15: 301-342, 1978.

Kennedy, M., Mrofka, D., and von der Borch, C : Snowball Earth termination by destabilization of equatorial permafrost methane clathrate. Nature 453: 642-645, 2008.

Kling, G.W., Kipphut, G.W., and Miller, M.C.: The Flux of C0 2 and CH4 from Lakes and Rivers in Arctic Alaska. Hvdrobiologia 240: 23-36, 1992.

Kushwaha, S.C., Kates, M., Sprott, G., and Smith, I.C.P.: Novel polar lipids from the methanogen Methanospirillum hungatei Gpl. Biochim. Biophvs. Acta 664: 156-173,1981.

Laybourn-Parry, J., James, M.R., McKnight, D.M., Priscu, J.C., Spaulding, S.A., and Shiel, R.: The microbial plankton of Lake Fyxell, Southern Victoria Land, Antarctica during the summers of 1992 and 1994. Polar Biol. 17: 54-61, 1997.

Lawson, J., Doran, P.T., Kenig, F., Des Marais, D.J., and Priscu, J.C.: Stable carbon and nitrogen isotopic composition of benthic and pelagic organic matter in lakes of the McMurdo Dry Valleys, Antarctica. Aquat. Geochem. 10,:269-301, 2004.

Lee, P.A., Priscu, J.C., DiTullio, G.R., Riseman, S.F., Tursich, N., and de Mora, S.J.: Elevated levels of dimethylated-sulfur compounds in Lake Bonney, a poorly ventilated Antarctic lake. Limnol. and Oceanography 49: 1044-1055, 2004.

Lizotte, M.P., and Priscu, J.C.: Natural Fluorescence and Quantum Yields in Vertically Stationary Phytoplankton from Perennially Ice-Covered Lakes. Limnol. and Oceanography 39: 1399-1410, 1994.

Lyons, W.B., Fountain, A., Doran, P., Priscu, J.C., Neumann, K., and Welch, K.A.: Importance of landscape position and legacy: the evolution of the lakes in Taylor Valley, Antarctica. Freshwater Biol. 43: 355-367, 2000.

Lyons, W.B., Welch, K.A., and Sharma, P.: Chlorine-36 in the waters of the McMurdo Dry Valley lakes, southern Victoria Land, Antarctica: Revisited. Geochim. Et Cosmochim. Acta 62: 185-191, 1998.

Margesin, R., and Schinner, F.: Biological decontamination of oil spills in cold environments. J. of Technol. and Biotechnol. 74: 381-389, 1999.

McKnight, D.M., and Andrews, E.D.: Potential hydrologic and geochemical consequences of the 1992 merging of Lake Chad with Lake Hoare in Taylor Valley. Antarct. J.-Rev. 249-251, 1993.

Mikucki, J.A., Foreman, CM., Sattler, B., Lyons, W.B., and Priscu, J.C.: Geomicrobiology of Blood Falls: An iron-rich saline discharge at the terminus of the Taylor Glacier, Antarctica. Aquat. Geochem. 10: 199-220, 2004.

Neumann, K., Lyons, W.B., Priscu, J.C., and Donahoe, R.J.: C02 concentrations in perennially ice-covered lakes of Taylor Valley, Antarctica. Biogeochem. 56: 27-50,2001.

Ochsenreiter, T., Selezi, D., Quaiser, A., Bonch-Osmolovskaya, L., and Schleper, C : Diversity and abundance of Crenarchaeota in terrestrial habitats studied by 16S RNA surveys and real time PCR. Environ. Microbiol. 5: 787-797, 2003.

Orphan, V.J., House, C.H., Hinrichs, K.U., McKeegan, K.D., and DeLong, E.F.: Direct phylogenetic and isotopic evidence for multiple groups of archaea involved in the anaerobic oxidation of methane. Geochim. Et Cosmochim. Acta 66: A571-a571, 2002.

Parker, B.C., Simmons, G.M.J., Seaburg, K.G., Cathey D.D., and Allnutt, F.C.T.: Comparative ecology of plankton communities in seven Antarctic oases. J. of Plankton Ecol. 4: 271-286, 1982.

Pearson, A., Huang, Z., Ingalls, A.E., Romanek, C.S., Wiegel, J., Freeman, K.H., Smittenberg, R.H., and Zhang, C.L.: Nonmarine crenarchaeol in Nevada hot springs. Appl. and Environ. Microbiol. 70: 5229-5237, 2004.

Porter, S.C., and Beget, J. E.: Provenance and depositional environments of late Cenozoic sediments in permafrost cores from lower Taylor Valley, Antarctica. In: Dry Valley Drilling Project ed. L.D. McGinnis, pp. 351-364 Antarctic Research Series, American Geophysical Union, Washington D. C , 1981.

Powers, L.A, Werne, J.P., Johnson, T.C., Hopmans, E.C., Sinninghe Damste, J.S., and Schouten, S.: Crenarchaeotal membrane lipids in lake sediments: A new paleotemperature proxy for continental paleoclimate reconstruction. Geol. 32: 613-616,2004.

168

Priscu, J.C.: The biogeochemistry of nitrous oxide in permanently ice-covered lakes of the McMurdo Dry Valleys, Antarctica. Global Change Biol. 3:301-315,1997.

Priscu, J.C., Wolf, C.F., Takacs, CD., Fritsen, C.H., Laybourn-Parry, J., Roberts E.C., and Lyons, B.: Organic carbon transformations in the water column of a perennially ice-covered Antarctic lake. BioScience 49:997-1008, 1999.

Roberts, E.C., Laybourn-Parry, J., McKnight, D.M. and Novarinos, G.: Stratification and dynamics of microbial loop communities in Lake Fryxell, Antarctica. Freshwater Biol. 44: 649-661, 2000.

Roberts , E.C., Priscu, J.C., Wolf, C , Lyons, W.B., and Laybourn-Parry, J.: The distribution of microplankton in the McMurdo Dry Valley Lakes, Antarctica: response to ecosystem legacy or present-day climatic controls? Polar Biol. 27: 238-249, 2004.

Schouten, S., Hopmans, E.C., Pancost, R.D., and Damste, J.S.S.: Widespread occurrence of structurally diverse tetraether membrane lipids: Evidence for the ubiquitous presence of low-temperature relatives of hyperthermophiles. Proceedings of the National Academy of Sciences of the United States of America 97: 14421-14426, 2000.

Schouten, S., Hopmans, E.C., Schefuss, E., and Damste, J.S.S.: Distributional variations in marine crenarchaeotal membrane lipids: a new tool for reconstructing ancient sea water temperatures? Earth and Planet. Sci. Lett. 204: 265-274, 2002.

Singh, N., Kendall, M.M., Liu, Y.T., and Boone, D.R.: Isolation and characterization of methylotrophic methanogens from anoxic marine sediments in Skan Bay, Alaska: description of Methanococcoides alaskense sp nov., and emended description of Methanosarcina baltica. International J. of Systematic and Evolutionary Microbiol. 55: 2531-253, 2005..

Smith, R.L., Miller, L.G., and Howes, B.L.: The geochemistry of methane in Lake Fryxell, an amictic, permanently ice-covered, Antarctic lake. Biogeochem. 21: 95-115,1993.

Spigel, R.H., and Priscu, J.C.: Physical limnology of the McMurdo Dry Valley lakes. In: Ecosystem Dynamics in a Polar Desert: The McMurdo Dry Valleys. Antarctica, 72: ed. J.C. Priscu, pp. 153-187. American Geophysical Union, Washington, D.C.1998.

Stuiver, M., Denton, G.H., Hughes, T.J., and Fastbrook, J.L.: History of the marine ice sheet in west Antarctica during the last glaciation: a working hypothesis. In: The Last Great Ice Sheets, eds. G.H. Denton, T.H. Hughes, pp. 319-436. Wiley-interscience, New York, NY., 1981.

169

Treude, T., Orphan, V., Knittel, K., Gieseke, A., House, C.H., and Boetius, A.: Consumption of methane and C02 by methanotrophic microbial mats from gas seeps of the anoxic black sea. Applied and Environ. Microbiol. 73: 3770-3770, 2007.

Turich, C , Freeman, K.H., Bruns, M.A., Conte, M., Jones, A.D., and Wakeham, S.G.: Lipids of marine Archaea: Patterns and provenance in the water-column and sediments. Geochim. Et Cosmochim. Acta 71: 3272-3291, 2007.

Tyler, S.W., Cook, P.G., Butt, A.Z., Thomas, J.M., Doran, P.T., and Lyons, W.B.: Evidence of deep circulation in two perennially ice-covered Antarctic lakes. Limnol. and Oceanography 43: 625-635, 1998.

Van de Vossenberg, J.L.C.M., Driessen, A.J.M., Da Costa, M.S., and Konings, W.N.: Homeostasis of the membrane proton permeability in Bacillus subtilis grown at different temperatures. Biochim. Biophys. Acta-membranes 1419: 97-104, 1999.

Vincent, W.F.: Microbial ecosystems of Antarctica. Cambridge University Press, 1988.

Wagner, B., Melles, M., Doran, P.T., Kenig, F., Forman, S.L., Pierau, R., and Allen, P.: Glacial and postglacial sedimentation in the Fryxell Basin, Taylor Valley, southern Victoria Land, Antarctica. Palaeogeog. Palaeoclim. Palaeoecol. 241: 320-337, 2006.

Weijers, J.W.H., Schouten, S., Spaargaren, O.C., and Damste, J.S.S.: Occurrence and distribution of tetraether membrane lipids in soils: Implications for the use of the TEX86 proxy and the BIT index. Org. Geochem. 37:1680-1693, 2006.

Weijers, J.W.H., Schouten, S., van den Donker, J.C., Hopmans, E.C., and Damste, J.S.S.: Environmental controls on bacterial tetraether membrane lipid distribution in soils. Geochim. Et Cosmochim. Acta 71: 703-713, 2007.

Weitemeyer, K.A., and Buffett, B.A.: Accumulation and release of methane from clathrates below the Laurentide and Cordilleran ice sheets. Global and Planetary Changes 53: 176-187,2006.

Wharton, R.A., Jr., Lyons, W.B., and Des Marais, D.J.: Stable isotopic biogeochemistry of carbon and nitrogen in a perennially ice-covered Antarctic lake. Chem. Geol. 107: 159-172, 1993a.

Wharton, R.A., McKay, C.P., Clow, G.D., and Andersen, D.T.: Perennial ice covers and their influence on Antarctic lake ecosystems. In: Physical and Biogeochemical Processes in Antarctic Lakes 59: eds. W. Green, E.I. Friedmann, pp. 53-70 Antarctic Res. Sen, American Geophysical Union, Washington, D.C., 1993b.

170

APPENDIX IV

ESTIMATED MEAN GRAIN SIZE

Sedimentary Structure

GENERAt DESCRIPTION

OraiH Description Organic®

OLEY 2.4/108 park bluish gray

GLEY 2. 3/1C-8 Very •&«* bhsisfi gray]

GLEY 2. 2,»5Pi Biuisb black

GLEY 2, 5?10B Bluish gray

GLEY Z 23?SPq Bluish Biack

Giey 2, S/10B Bluish gray

G18Y2, 2.5/5PB •Bluish.black

GtEY 2. 4/10B Dark blussh gray]

More pronounced laminarion

Lamination me vent Brownish line? lamination

m siit, w/tn to c sand grains

Transition layer

Poss*bi8 largo organic, particles

F lo c Saria; arigliFaf'grams n io«!y"' black, i*/ some white Qtz.

Sandy layer Fine sift wi scattered fine sand

Sandy layer

Fine lamination

Sandy layer

Clay -on outside of sample, coarse sand beneath Fine lighter colored layer below

F to m silt

Quarts sands

Mostly black, less b e than 2.5-3.6 cm

Prominent layer Well- sorted fine to med angular %&n4

Qtz. Present more inan any other layer

Appendix V-l. Lake Fryxell bottom sediment log.

171

% % | S S t-l~ w * ft -

l[ J f 1

1 if 3

1 i * Fi

1

COLOR

BUI , M-J>,Bill? til

BU.U.n.k

K-Ki i i i r - i r '

l i - . - j r l i i i -W

y-SirW'-ft-i

1! s n l W ' *

- 1 - f l 1 ill, 1' ,


CUM SiL" W » E ' f

. . . . . . I

*•* * ;J

. ,

* * *

. » # *

Visual Observations

t H ! ! <,;>, • > l . ! l « U I..1 1 * <-f 't<

UI'^I *- J r^1 (**(! iss) sO"* "»e '>

V < K Jts j a f W M * * ' A l l

1 i*i = < <-<< 1.1 l a i>3 i

^ tJ t i \ i i-. ff<. f f - ^ i> " u

i* ( I J 4 ,« 1 <*, 1 t . ' Pi V K i ^ . i t ' d i ' i i ^ n J i

M » '»i •*, Wum Mi >tli It iiihH [ i* iv

^ .3 cm

Appendix V-2. Lake Hoare bottom sediment log.

Lake HoareCoreA i

: .

P * 1 * | , o '

:

j :

COLOR

H-koreO -u.

5V; |

2.5Y5 ; :

ogge<J by: Caroline iaraula May 2b> 2004

ESTIMATED MEAN 8RAIN SHE

CLAY StLT SAND

, * • ' * • . • ' » •

' • • > • .

. . . . . . ;

• '

.

OENERAL DESCRIPTION

Visual OhservsrtktH* < «i ing Sjmpltng

Alternating thin CIS- 1 ,5 cm) layers of mud and sand

mud and sill ImeHammatian;*

Wdl-sorted sands

Mudd) sand

"Mitt*.

dw n tin in t 1 » nl f i n ihn i k e u i i ' . I' t h k k u . l

Me MMlllll- U t !aku Inirr ilu i »h.r

"the layers are less hesrt possibly <iae to lower mud cemlem and a higher Jiand fraction.

When ihe tore ^ .J - euu ihe ruhbc melk-J am! 'et'l <t hlaei imrar

c<*u'<-sr 7 em from Tbeboiusm, Uurine eoie !oe.!}U!£. the Uafk -,me<u wa^-,trujvil sideua\s uith a Wade.

Field description: anoxic

• -?-..- i

* • .

"" ' *

Appendix V-3. Lake Hoare core A sediment log.

Lake Hoare Core B I <>i'j.'Cil H. < 'jrulim; jaraula May 26, 2004 Field description: oxic; waa'rdepfii 21 in

1,1 F j * | S

| J 1 , I 1 4 I i |

f 6

I f % H>

I f 10

f 12 R

| I ' 4

I II f 16

^ 18

J 20

COLOR

5Y4/1

5Y4/2

5Y4/1

SYS/1

2 . 5 Y 5 ?

ESTIMATED MEAN SRAIN SIZE

s 'A ' . . . « r J*?"1, •

,

' * . 1 • * ". 1 ' • : :pr

• « • {«_* -,.*,

* ***** « . « ,—".—~

" ' • " . * • * • * • " . * * •

• • « «

•' '

VHoal Observations

ildit-colnK-ii organic-ricS) laniifytins

"sand l,i..i<.ii<iii upwdiJ am! the »:«' k\ji»t)iiintn iipw-atd

Sami ia u m\hkh: matrix f.'uatv sand inteilansmaJad wilh smv-<imU lines represent line vind i

MwJlts VIMl

GENERAL DESCRIPTION

Caring/Sampling Notes

Ihe Isnenaiv <lijjhtlv bmx possihk <hw i<» vorini;. MMM uf ihe ivBi lasers arc muJ-nca

'jg.!

- , • 'SI&JS i C ^ . ^ ^ 1

'4*ih3P<«fc

"tf»M.* - .* K fW|

* ••"• r,*w

AjW** ' ^ 1 * * %

• » * . . • • • *

"• 1*11'** ,.r-;W^4 ••'• ". * " •*,

t .

»

Appendix V-4. Lake Hoare core B sediment log

174

L&k& Botmsy fogetetf by: Caroline JarauSa

u \ ,

if 20

If 30

r 40

jf 50

if 60

f 70

f 80

f 90

f 100

COLOR

3@rk. (clay) ^ta^nish §?ay

OtEV 1 m iCiy

1 r


CLA <*l 1 S /ND

1

1

1

— -

" "-.-

% * 3

Legend

' V

• 1 *

GENERAL DESCRIPTION

SedfeiMMtaiy St ruc ture a m i Grain Descr ip t ion

0-3. s em ~SD% Quaite sand angular q^stg S»^ Jo ^m: ml (granutes), -15% dajf f'nwf p w y $estesd c&y&y sarifl sod maSic miOiM&is

• •»*5%qy*rtetf.69ndi -10% mafe

4 •;S% Sim; s ifKi) -:!:svs>'

7 cm ' -•£$% ftoe. sne^im 3. cost's* ssmti;

tO cm

sandy ft'ay n ? * f t » sand)

Samples

H o.

1 • 3.S .

I •

1 1 5-13

1

Appendix V-5. Lake Bonney bottom sediment log.

CHAPTER 6

CONCLUSIONS AND RECOMMENDATIONS

Extremely cold and dry conditions in Taylor Valley result in slow-growing biological

communities, simple trophic structure, and exclusively microbial lakes. This unique environment

provides a perspective of biogeochemical processes at slower rates. Biomarkers are better

preserved in the meromictic and cold lake waters that have minimal grazers and sometimes

anoxic conditions. From this study, the use of biological markers was instrumental in concluding

that Archaea is ubiquitous in Taylor Valley, average temperatures estimated from TEXg6 for the

Lakes Hoare, Chad and West Bonney water columns are close to instrumental records, alkenone

unsaturation indices are consistently in the low end.

Differences in the compositions of the spilled oils and the residues in ice are due to

evaporation, biodegradation, and water washing. Relative strengths of these processes were

assessed by plotting the ratios (/-CI3/«-CI3)S/(/-CI3/«-CI3)D and total «-alkanes/total naphthalenes

for different sites and depths in the ice. To isolate the effects of evaporation on these ratios, a

fuel evaporation experiment was conducted.

1.0 Ubiquity of Archaea

Eukarya and bacteria are well studied in the region, however, Archaea has only been

recently recognized owing to the difficulties in biological and microbiological techniques in

acquiring archaeal cells. Analysis of lipids is an efficient way to recognize and study archaea,

moreso that the lakes are purely microbial and do not harbor higher plants and animals. This

175

176

study showed that Archaea is ubiquitous in Taylor Valley, which was only reported in Lake

Fryxell. One of the key implications for archaeal activity is their influence to carbon and element

cycling in the lakes. Some Archaea are known chemolithotrophs, which use inorganic substrates

to produce energy. Most of the documented primary producers in Taylor Valley are

photoautotrophs, which have efficiently adapted to prolonged darkness and low levels of

photosynthetically active radiation during the summer. It is also possible that chemolithotrophic

archaea, even with slower metabolic rates, may have comparable annual contributions as some

photoautotrophs because of their longer or continuous activity year long compared to light-

dependent autotrophs. The influence of archaea in the carbon-fixing and cycling of nutrients are

likely more than previously thought and need to be reconsidered in the carbon and nutrient

budgets.

2.0 Terrestrial biomarkers in Antarctic sediments

Even though the continental Antarctic does not have mature soil formation, bacteria

associated with soils are present. Branched tetraethers, biomarkers for terrestrial bacteria, were

detected in Lake Chad moat microbial mats and sediments in other Taylor Valley lakes. This is

the first report of branched tetraethers in microbial mats. The abundance relative to the total

GDGT in the microbial mat is 94%, an abundance commonly reported for soils and rice paddies.

This high percentage cannot be explained by low microbial activity in the sediments in the

watershed, or by the lone stream that drains into the lake. Moreover, Lake Fryxell drains the

most number and the longest streams, but does not have significant branched GDGTs. Hence,

streams and watershed sediments cannot be a major source of branched tetraethers for Lake

Chad. If glaciers or sediments from glaciers are considered as a source, glacial-dammed Lakes

Hoaire, Fryxell and Bonney only have as much as 12% of branched GDGTs. To yield abundances

of more than 90%, it is likely that the branched tetraethers are produced in situ. There are only a

handful of lakes reported so far that have more than 90% of branched tetraethers (Hopmanns,

2004).

Consequently, branched tetraethers preserved in any lake sediments first need to be

evaluated whether these are autochthonous or allochthonous deposits before they are used for

environmental interpretations. This problem can be resolved by a survey of branched tetraethers

from different parts of the watershed and the lakes. For Taylor Valley, the Lake Fryxell

watershed has the most number and longest streams that stream transects from glaciers to the

lake may provide useful information about the source of the branched tetraethers. The database

generated will be useful in interpreting the increase of branched tetratethers in the late

Pleistocene deposits of the lake. If Archaea is present, the lipid information about the microbial

consortia will also aid in explaining the loss of nutrients from tracer experiments performed

along the stream (McKnight, pers. comm.). Sediment and microbial mat samples along a transect

in Lake Chad, channel connecting Lake Chad and Lake Hoare, from Taylor Glacier to West

Lake Bonney, and at the interface of the hyporheic zone and the permafrost are some of the

critical sites that may have preserved biomarkers and harbor impotant information about the

microbial consortia.

3.0 Cryptoendolithic community

This study confirms the presence of Archaea in the granite cryptoendolithic community.

Matsumoto et al. (1992) reported isoprenoids, archaeal biomarkers, in cryptoendoliths of the

Beacon Supergroup although Archaea was not elucidated as a source. Interestingly, the

cryptoendoliths, assumed to represent the "terrestrial" environments, does not contain branched

tetraethers, but only crenarchaeol and caldarchaeol. There is still a need to refine specific

GDGTs to groups of archaea or bacteria and to have a diversity of "terrestrial" samples to

understand the provenance and source organisms of branched tetraethers. Current developments

in GDGT analyses use intact polar lipids, where the polar groups are still attached, and this has

been shown to be specific for living prokaryotes and in some cases species-specific (Sturt et al.,

2004). Study of biomarkers and the consortia of organisms in cryptoendolithic communities also

have significant applications for astrobiological studies.

4.0 Pentaunsaturated alkenones

Previously unrecognized pentaunsaturated alkenones were identified as octatriaconta-

pentaen-2-one (C38:5Me), nonatriaconta-pentaen-2 and 3-one (C39:5Me and Ca^Et), as well as

tetradec-pentaen-2-one (C .-sMe) from mass spectral data. These compounds comprise 49%,

29% and 12% of the total C38, C39 and C40 alkenones, respectively. The alkenones also consist

mostly of tetra- and triunsaturated moieties, whereas the diunsaturated alkenones are minimal.

The alkenone distribution is the most diverse compared to those reported from marine, lake and

culture distributions. Given this diversity of alkenones, the absence of C37:sMe and C38:sEt in

contrast to the presence of longer chain pentaunsaturated moieties does not express an

179

energetically favorable biosynthesis because alkenones are first elongated before desaturation is

carried out. Based on reported alkenones, Rontani et al. (2006) noticed that for C37 to C40 moities

have five methylene units in between double bonds and also proposed that this biosynthetic

pathway is specific to this group of haptophytes. The proposed mechanism shows that

desaturation occurs in the order A14C (14th Carbon after the carbonyl group), A21C, A7C, then

A28C. This study supports that the grouping of alkenones based on their specific biosynthetic

pathway by Rontani et al. (2006) wherein C37Me and C3gEt cannot accommodate a fifth double

bond due to the five methylene unit distance desaturase enzymes require during the alkenone

synthesis.

The unusual diverse distribution of alkenones is probably also influenced by light- and

nutrient-limited conditions in the upper water column of Lake Fryxell. Cold and anoxic

conditions in the bottom water enhance the preservation of these alkenones. The combination of

the physic-chemical properties of the upper and bottom water columns provided excellent

conditions in Lake Fryxell that are not found in other lakes.

5.0 Lake Fryxell Haptophytes

Fragments of 18S rDNA of Haptophytes were amplified from isolated environmental

DNA samples. Molecular phylogenies of two 18S rRNA clones are 100% specific for

Prymnesiophyceae. The two Lake Fryxell sequences are most closely related to each other,

having 99% sequence identity, associated with brackish Ace Lake phylotypes 4 and 5, and have

similarity to Haptophytes from Greenland lakes, but are a distinct cluster from I. galbana.

The Lake Fryxell alkenone distribution have C^^Et and C40:2Et, only Chrysotila

lamellosa is reported to biosynthesize these compounds from culture experiments. In addition,

C. lamellosa is associated with /. galbana and both commonly thrive in brackish environments.

This indicates that there is high potential for Chrysotila lamellosa thriving in Lake Fryxell,

although this Haptophyte is not reported nor fragments of its 18S rDNA sequence have been

detected.

6.0 Applicability of temperature proxy records in cold environments

Due to low light conditions in the lakes, most metabolic activities occur in the summer;

therefore, any temperature estimate will represent the average for this season. Water column

temperatures were estimated from existing calibrations of alkenone unsaturation indices (JJ"37

and JJ ) and the Tetraether Index for 86 Carbons (TEXg6). Annual mean air temperatures

(MAT) were estimated from the extent of methylation and cyclization of branched tetraethers.

All the existing calibration equations were based on water temperatures that were not as

low as those of the Taylor Valley lakes. The calculated water temperatures are, therefore,

extrapolated from the linear regression lines. Slight overestimates from most of the calculated

temperatures attest to the bias towards warm temperatures. Each temperature proxy is evaluated

below.

7.0 Alkenone unsaturation Indices

Alkenones, proven important in climate studies for use as a paleotemperature proxy, are

present in Lake Fryxell. In studies that calibrated the alkenone unsaturation indices and sea

surface temperatures, a common problem with data from cold environments is a decrease in

correlation with temperature. It has been proposed that non-thermal factors, such as light- and

nutrient-limited conditions increase the unsaturated moieties and bias the estimates to low

temperatures. From this study, temperatures calculated from existing calibration equations yield

3.6 to 7.3 °C, slight overestimates from the 2.5 °C water column average even with low light and

nutrient-depleted conditions. The JJ and JJ~31 indices, however, are consistently in the low

end of reported values for lakes and ocean sediments. The indices will be a better tool when

comparing Taylor Valley records to worldwide data and as a qualitative comparison for

paleotemperatures in the lakes, rather than the calculated temperatures. Lake calibration

equations used in calculating (paleo)temperatures are biased to middle and low latitude areas.

Alkenone unsaturation indices use abundances of two, three and four unsaturations in C37

alkenones. In Lake Fryxell, however, C37:2 abundances are low. The loss in the faithfulness of

temperature estimates in the cold regions may be traced, in part, to the biosynthesis of these

compounds. Rontani et al. (2006) proposed that alkenones are first lengthened, then desaturated,

and have five methylene units in between double bonds. This also means that with colder

environments, much of the effects of temperature will be recorded in longer alkenones, C38 and

C39. At a certain low temperature threshold or C37 abundance, more C38 to C40 pentaunsaturated

moieties are biosynthesized, which can explain the decrease in correlation coefficients with C37

alkenones in cold regions.

182

8.0 Tetraether Index for 86 Carbons(TEXgf,)

Another globally calibrated temperature proxy record is based on archaeal biomarkers.

TEX86, which converts to temperature, were applicable to Lakes Hoare, Chad and Bonney, but

not for Lake Fryxell. Temperatures of 1.2and 2.4 °C are close to 0.4 and 2.2 °C averages for the

water column instrumental records from 1986 to 1998 for Lakes Hoare and Bonney,

respectively. The index is formulated with the assumption that aquatic Crenarchaeota adapt to

low temperatures by synthesizing crenarchaeol and, similar to their (hyper)thermophilic

ancestors, increase the number of cyclic moieties as temperatures increase. Cyclic GDGTs

biosynthesized by Euryarchaeota will be quantitated along with those from Crenarchaeota and

will increase the temperature estimates. Methanogens, which belong to kingdom Euryarchaeota,

were inferred present in all the lakes based on stable carbon isotopic surveys of the water column

and sediments (Lawson, unpublished data). For a purely microbial habitat, the influence of

GDGTs from methanogens can easily offset those contributed by Crenarchaeota and thus

overestimate the temperature, such as the average for Lake Fryxell. There are no long-term

temperature record for Lake Chad, so the 2.8 °C estimate from TEXs6 cannot be evaluated.

9.0 Methvlation and Cvclization indices for Branched Tetraethers (MBT and CBT)

Temperature estimates based on the number of branching and cyclization of the branched

tetraethers were correlated with annual mean air temperatures (MAT) from previous studies.

Although temperature estimates can still be calculated, a major limitation to any interpretation

for the study area is the effect of the ice cover and the temperature of the water column for

branched GDGTs that were biosynthesized in situ. Only in Lake Chad was there strong evidence

for in situ production. Incidentally the lake is in the middle of the valley, loses its ice cover and

is well mixed because it is shallow. Consequently, the temperatures in the water column can

equilibrate with air temperatures for the summer, therefore the temperature changes can robustly

represent annual MAT changes for the valley.

MAT in dry environments elsewhere tend to be underestimated, so experiments are

underway to test possible reasons (Schouten/pers. comm.). One consensus was to modify the

MBT index by dropping the term in the denominator that consisted of moieties having a methyl

group at position 5 or GDGT [IX], [IXb], and [IXc] because abundances are commonly low,

which makes quantitation difficult. For Taylor Valley, the sources of branched GDGTs for the

lakes need to be identified whether these are allocthonous or autochthonous. To improve the

reliability of the MBT temperature estimates from Lake Chad as a representative of the valley,

sediment and benthic microbial mats need to be collected from the depocenter and, for at least a

year, summer air and water column temperatures need to be monitored.

10. Developing biomarker proxies for paleoclimate and paleoenvironmental reconstructions

Summer temperatures are poised at the melting point of ice. Supply of liquid water is

crucial to the survival of organisms and is the limiting factor to life in Taylor Valley. Slight

changes in temperature and humidity can have a large effect on the liquid water budget in the

soils, streams, and lakes. During colder and drier conditions, the lakes become saline ponds, if

not desiccated. Slight and prolonged cooling can cause some of the shallow lakes and ponds to

freeze throughout its water column. During relatively warm conditions, lake ice thin or its loss

triggers lakewide biogeochemical changes as the surface water dissolved oxygen content and

184

nutrients from the bottom waters are mixed. The Taylor Valley system is sensitive to climatic

changes and the simplified trophic structure allows for biomarker proxies to record

biogeochemical changes in systems containing liquid waters.

As shown in this study and from Turich et al (2007), the modern archaeal lipidic

assemblages were calibrated with the modern physico-chemical conditions. For Taylor Valley,

archaeal assemblages are specific to each lake and are controlled, to a large extent, by the oxygen

supply of the bottom waters. Presence of the perennial ice cover, salinity, and temperature are

factors that directly affect the oxygenation of the bottom waters. Anoxia will favor methanogens

and enhance the lipid contributions from Euryarchaeota compared to Crenarchaeota. For Taylor

Valley, glacial-interglacial cooling and drying, lake level highstands and lowstands during the

existence of Glacial Lake Washburn, and a possible recent dessication event were resolved from

the archaeal assemblages.

Compared to the abundances of GDGTs in the archaeal assemblages by Turich et al.

(2007), Lake Hoare-type is equivalent to upwelling regions, Lake Fryxell-type are similar to

coastal areas, and Lake Chad is consistently terrestrial. Common to both studies, these three

types of assemblages have a predominance of crenarchaeol, GDGT-0, and branched GDGTs,

respectively. As expected, the assemblages in Taylor Valley are limited compared to the diverse

types of environments represented from in the study of Turich et al. (2007).

Establishing molecular proxy records using assemblages will have to encompass a more

diverse selection of environments that include streams, glaciers, soils and cryptoendoliths. For

Taylor Valley, additional samples from ice covers, lake water columns, lake sediments, and

paleodelta deposits. From streams, longitudinal and cross sectional transects along streams and at

185

the interface of the permafrost and hyporheic zones will have enough spatial scale to derive

environmental trends from the melt water sources to sinks. Lake transects following the sampling

scheme from isotopic studies by Lawson et al. (2004) supplemented with data from biomarkers,

compound-specific stable isotope and radiocarbon analyses are powerful tools to establish

current and paleoenvironmental trends in Taylor Valley.

It is also recommended that future lipidic studies should include quantitative analysis of

archaeal biomass, as well as microscopic and staining techniques to identify the percentage of

viable cells. Archaeal lipids in conjunction with compound-specific isotope analysis will help in

identifying the sources of carbon. Though the microbial community has a seemingly simple

trophic system, the similarity in lipidic components can pose problems in identifying the function

of the archaeal species, most of which are uncultivable. To resolve this problem, new techniques

using isotopic analysis of phylogenetically specific rRNA from prokaryotes may eventually be

worth exploring if these are applicable to Taylor Valley lakes.

11. Natural attenuation of oils spilled on the ice cover

Light non-aqueous phase liquids (LNAPLs) aviation jet fuel and hydraulic fluid, as well

as dense non-aqueous phase liquids (DNAPLs) synthetic transmission and engine oils were

spilled on the Lake Fryxell ice cover during a helicopter crash. No residues of the hydraulic fluid

were detected, so no conclusions can be made for this fluid.

Fractions of the fluids were naturally attenuated through evaporation and biodegradation

wherein their physico-chemical properties determined their persistence in the ice cover. Dry

conditions in Taylor Valley causes sublimation rates of 30 cm y'1 of the ice cover and also

186

evaporated a major fraction of the volatile aviation diesel. The synthetic fluids, however, are not

susceptible to evaporation. Another process that naturally attenuated the spilled oils was

biodegradation, but only in ice associated with sediments. Compounds that were readily

biodegraded from the diesel were «-alkanes and naphthalenes, and from the synthetic oils were

pentaerythritol triesters that have C5 to C6 fatty acids. From the biodegradation trends observed,

this study recommends that formulation for fatty acids used for pentaerythritol triesters include

more of the short chain moieties to increase the biodegradability of synthetic oils.

12. Dynamic ice covers

This study also shows evidence of the ice cover as dynamic and porous rather than static

and solid. As early as a year after the spill, oil slicks were already observed at least 40 m from

the crash site. This is evidence of the porous nature of the ice cover that can contain up to 40%

liquid water in the summer and can have discontinuous aquifers when sediments are present

(Fritsen and Priscu, 1998). Water washing was a significant process that altered and laterally

transported the fuel instantly. Jepsen et al. (2006) has shown that fuel dissolves the intergranular

boundaries between ice crystals and called this as "fuel tunneling." Rates of fuel movement were

as high as 1.6 m hr"1, however, tortuosity of the tunnels can significantly affect the vertical

distance travelled. Most contaminant dispersal studies consider transport of pollutants in the ice

along with water, but the interaction of LNAPLs or ice containing aromatic compounds and the

surrounding solid ice at the onset of the melt season may have higher vertical transport rates than

during the height of the melt season. Dynamics between the structure of the ice that contain

sediments, bubbles, ice straws, for example, and its interaction with fluids that have low heat

187

capacities are not well understood. In areas where ice is present, it might serve well to study the

interaction between them as increasing temperatures from global warming can affect the stability

and porosity of ice and the dynamics and mechanism of ice melting is not well understood to

explain the higher rates of decline compared to predicted models for marine ice shelves, glaciers

and lake ice covers.

13. Monitoring and remediation

Based on the estimated volume of oils recovered during clean-up, susceptibility of the

majority of the spilled fluids to evaporation and biodegradation, following the protocols on

environmental protection in the Antarctic Treaty, it is more practical not to disrupt the lake ice to

recover any more of the spilled fluids. Recovering the sediments that sorb the contaminating

fluids may be more reasonable. A feasibility study will have to be done to weigh the advantages,

disadvantages and practicalities of this clean-up. Monitoring the lateral and vertical extent of the

spill can be possible by analyzing the dimethyl- and trimethyl naphthalene content of the ice as

these compounds are from the fuel, more soluble than the other aromatic constituents, and less

biodegradable than rc-alkanes and naphthalene.

188

REFERENCES CITED

Fritsen, C.H., Priscu, J.C.: Cyanobacterial assemblages in permanent ice covers on Antarctic lakes: Distribution, growth rate, and temperature response of photosynthesis. Journ. of Phvcol. 34; 587-597: 1998.

Jepsen, S.M., Adams, E.E., Priscu, J.C.: Fuel movement along grain boundaries in ice. Cold Regions Sci. and Tech. 45: 158-165: 2006

Hopmans, E.C., Weijers, J.W.H., Schefuss, E., Herfort, L., Damste, J.S.S., Schouten, S.: A novel proxy for terrestrial organic matter in sediments based on branched and isoprenoid tetraether lipids. Earth and Planet. Sci. Lett. 224; 107-116: 2004.

Lawson, J., Doran, P.T., Kenig, F., Des Marais, D.J., Priscu, J.C.: Stable carbon and nitrogen isotopic composition of benthic and pelagic organic matter in lakes of the McMurdo Dry Valleys, Antarctica. Aquatic Geochem. 10; 269-301: 2004

Matsumoto, G.I.: Geochemical features of the McMurdo Dry Valley lakes, Antarctica. In: Physical and biogeochemical processes in Antarctic lakes. Antarctic Research Series 59; pp. 95-118, Washignton, D.C., American Geophysical Union, 1993.

Rontani, J.F., Prahl, F.G., Volkman, J.K..: Re-examination of the double bond positions in alkenones and derivatives: biosynthetic implications. Journ. of the Phvcol. Soc. of Am. 42;800-813: 2006.

Sturt, H.F.; Summons, R.E., Smith, K., Elvert, M., Hinrichs, K-U.: Intact polar membrane lipids in prokaryotes and sediments deciphered by high-performance liquid chromatography/electrospray ionization multistage mass spectrometry-new biomarkers for biogeochemistry and microbial ecology. Rapid Comm. in Mass Spec. 18, 6; 617-628: 2004

Turich, C, Freeman, K.H., Brans, M.A., Conte, M., Jones, A.D., Wakeham, S.G.: Lipids of marine Archaea: Patterns and provenance in the water-column and sediments. Geochimica Et Cosmochimica Acta 71, 3272-3291: 2007.

VITA

Department of Earth and Environmental Sciences 845 West Taylor St., Chicago IL 60607-7059

e-mail: [email protected] Telephone: (312) 996-7207

Education 2008 2001 1997

Positions 2008 Fall 2008 summer 2008 spring 2007 Fall 2007 summer 2004-2007 2003-2004 2002-2003 1997-2002 1997-1998 1996-1997

PhD Earth and Environmental Sciences - University of Illinois at Chicago (UIC) M.S. Geology University of the Philippines (U.P.) B.Sc. Geology University of the Philippines

Visiting Lecturer Research Assistant Graduate Fellow Teaching Assistant Lecturer Teaching Assistant Research Assistant Teaching Assistant Teaching Associate Research Associate Undergrad Assistant

UIC Earth and Environmental Sciences UIC Environmental Stable Isotope Laboratory UIC Institute of Environmental Science and Policy UIC Earth and Environmental Sciences UIC Earth and Environmental Sciences UIC Earth and Environmental Sciences UIC Organic Geochemistry Laboratory UIC Earth and Environmental Sciences University of the Philippines U.P. Marine Geology Laboratory University of the Philippines

Awards and Honors 2007 UIC Graduate Research Forum Honorable mention 2006 UlC-Office of International Services International Student Service Award 2006 UIC-EAES Best Graduate Student Award 2001 Outstanding Master's Thesis (Philippines-Dept. of Science & Technology) 2001 Best Student Oral Presentation (Philippine Assoc, of Marine Scientists)

Research Interests organic geochemistry, stable isotope geochemistry, forensics geochemistry, environmental biogeochemistry, and renewable energy

Research and Field Experience 2008 summer Gas flux survey of wetlands in Illinois 2008 summer Gas flux survey of a Chicago sewage treatment plant 2007-2003 Organic Geochemistry of Taylor Valley Lakes, Antarctica 2002-1997 Sedimentology, Geochemistry, and Malacology of Laguna de Bay: Implications for

Late Holocene Paleoenvironmental Changes (M.S. Thesis) 2002-2001 Coastal Erosion along Bauang-Aringay Coast (La Union): Rates, Causes and

Mitigation Measures 2000 Sedimentation and Tectonics of Subic Bay and Vicinity 1999 Tidal effects on groundwater fluctuations on a small oceanic island: Pag-asa Island,

Philippines (Spratlys Islands Group Research Program)

189

mailto:[email protected]

190

1998 Sedimentation Patterns and Sediment Dispersal along the Coast of Macajalar Bay: Establishing Environmental Baselines and Predictive Tools for Coastal Resource Management Using the Sediment Record

1997-1998 Sediment Quality, Sedimentation and Bathymetry of Laguna de Bay: Baseline Data for Lake Management

List of publications Ong, J., Aguda, N., Jaraula, C.M.B., Mateo, Z., Pascua, C. and Foronda, J.. Tidal Effects on

Groundwater in a Very Small Tropical Island. Science Diliman 12, 2; 33-44: 2000.

Accepted paper Jaraula, C.M.B., Kenig, F., Doran, P.T., Priscu, J.C., Welch, K.A. Aviation diesel fuel: molecular

composition and natural attenuation in the perennial ice cover of Lake Fryxell, Antarctica. Sci. of the Tot. Environ.

Papers to submit Jaraula, C.M.B., Brassell S.C., Doran, P.T. and Kenig F. Novel penta-unsaturated alkenones in

Lake Fryxell, Antarctica. (Revised) Jaraula, C.M.B., Kenig, F.P., Doran P.T., Welch, K.A. composition and natural attenuation of

synthetic oils and hydraulic fluid spilled on the perennial ice cover of Lake Fryxell, Antarctica.

Papers in preparation Jaraula, C.M.B., Schouten, S., Hopmans, E., and Kenig, F. Glycerol dialkyl glycerol tetraethers in

Taylor Valley Lakes, Antarctica. Jaraula, C.M.B., Kenig, F., Doran, P. Paleolimnology of Taylor Valley Lakes, Antarctica. Jaraula, C.M.B., Siringan, F.P., Klingel, R., Sato, H., and Yokoyama, R. Records and causes of

Holocene salinity shifts in Laguna de Bay, Philippines.

Extended abstracts/conference proceedings Siringan, F. P., Berdin R. D., J. Duruelo, Jaraula C. M. B., Mateo Z. R. P., J. Ong, Y. Yacat.

Sediment Transport Patterns Along Macajalar Bay and Cagayan de Oro River from Geomorphic and Sedimentological Parameters. Symposium on Tectonics of Southern Philippines (Proceedings) Cagayan de Oro City, Philippines: 1999.

Siringan, F. P., Berdin R. D., Jaraula C. M. B., Mateo Z. R. P., Mamaril-Villanoy M. J., Manzano J. and Juanico-Manzano L. Lithofacies and Stacking Patterns of the Cagayan Terrace Deposits: Implications to Base Level Changes. Symposium on Tectonics of Southern Philippines (Proceedings) Cagayan de Oro City, Philippines: 1999.

Invited presentations: Jun 2008 Long Term Ecological Research-McMurdo Dry Valleys Conference - The

biomarker record speak up for Haptophytes and Archaea in Taylor Valley lakes, Antarctica

Apr 2008 North Central College - Earth Science in our everyday lives Jan 2008 Geosyntec Consultants, Inc., Chicago - Using Anthropogenic and biological markers

in environmental forensics and in reconstructing paleoenvironments in Taylor Valley, Antarctica

191

Oral Presentations: Jaraula, C.M.B., Schouten, S., Hopmans, E., Doran, P. and Kenig, F. Distribution of archaeal and

bacterial tetraether membrane lipids in the lakes of Taylor Valley, Antarctica. Geological Society of America Annual Meeting, Houston, TX: 2008.

Jaraula, C.M.B., Siringan, F. Geochemistry, Sedimentology, and Malacology of Laguna Lake: Implications on Paleo-Environmental Changes. North American Lake Management Society, Lake Mendota, WI: 2002.

Jaraula, C.M.B., Siringan, F. Laguna Lake Was a Marine Arm of Manila Bay. Philippine Association of Marine Scientists, Dumaguete, Philippines, 2002.

Jaraula, C.M.B., Siringan, F., Mateo, Z., Berdin, R., Ong, J., Yacat, Y. Sediment Dispersal and Its Implications along a Segment of the Cagayan De Oro River and Macajalar Bay Coast. 7th International Interdisciplinary Conference on the Environment, San Francisco, CA: 2001.

Poster Presentations: Jaraula, C.M.B., Brassell, S.C., Schouten, S., Hopmans, E., Kenig, F.. Applicability of biomarker

temperature proxies in cold end-member lakes in Taylor Valley, Antarctica. Gordon Research Conference, Holderness N.H.: Oct. 2008)

Jaraula, C.M.B., Brassell, S. C, Doran, P.T., Kenig, F. Novel penta-unsaturated alkenones from Lake Fryxell, East Antarctica, UIC Graduate Research Forum, Chicago IL: Jun. 2008)

Chung, T.*, Jaraula, C.M.B., Kenig, F.K., Doran, Priscu, J.C., Welch. Aviation diesel fuel: molecular composition and natural attenuation in the perennial ice cover of Lake Fryxell, Antarctica. Long-Term Ecological Research-McMurdo Dry Valleys Conference, Chicago, IL: Jun. 2008.

Hillegas, A.*, Jaraula, C.M.B., Kenig, F.P., Doran P.T., Welch, K.A. composition and natural attenuation of synthetic oils and hydraulic fluid spilled on the perennial ice cover of Lake Fryxell, Antarctica. Long-Term Ecological Research-McMurdo Dry Valleys Conference, Chicago, IL: Jun. 2008.

Jaraula, C.M.B., Brassell, S. C, Doran, P.T., Kenig, F. Novel penta-unsaturated alkenones from Lake Fryxell, East Antarctica, EOS Trans. AGU, 88(52), Fall meet. Suppl., Abstract B51C-0612, San Francisco CA: Dec. 2007.

Jaraula, C.M.B., Kenig, F., Doran, P.T. Spilled helicopter fluids: molecular composition and natural attenuation in the perennial ice cover of Lake Fryxell, Antarctica, UIC Graduate Research Forum, Chicago IL: Apr. 2007.

Jaraula, C.M.B., Kenig, F., Doran, P.T. Paleolimnology and paleoecology of the Lake Fryxell basin using molecular fossils: Preliminary results, EOS Trans. AGU, 88(52), Fall meet. Suppl., Abstract B13C-1119, San Francisco, CA: Dec. 2006.

Jaraula, C.M.B., Kenig, F., Doran, P.T. Paleolimnology and paleoecology of the Lake Fryxell basin using molecular fossils: Preliminary results, LTER-ASM, Abstract 153, Estes Park, CO: Oct. 2006.

Jaraula, C.M.B., Kenig, F., Doran, P.T. Organic geochemical analyses of Taylor Valley lakes: Preliminary results, LTER-MCM, Portland, OR: Jun. 2005.

Jaraula, C.M.B., Kenig, F., Doran, P.T. Helicopter fluids: Molecular composition and natural attenuation in the Lake Fryxell ice cover, LTER-MCM Portland, OR: Jun. 2005.

* undergraduate student

192

Activity and Research Grants 2008 $ 700 Travel grants to present a poster at AGU Fall meeting 2007 from UIC Graduate

Student Council, Graduate College, and Women in Science and Engineering 2008 $14,000 Fellowship from UIC Institute of Environmental Science and Policy 2007 $ 550 Travel grants to present a poster at AGU Fall meeting 2006 from UIC-EAES and

Women in Science and Engineering 2007 $ 600 Travel grants to present a poster at LTER-ASM 2006 from UIC Graduate Student

Council, Graduate College, and Women in Science and Engineering 2006 $ 1,880 Research grant to process samples in the Royal Netherlands Institute for Sea

Research from the UIC Provost 2006 $ 5,500 Activity grant for UIC Earth Month festivities with Dr. Raymond Bradley as the

guest speaker from the UIC Student Activities Funding Committee 2006 $ 400 Fieldtrip grant for cave exploration from UIC Chicago Organization Fund 2000 $ 300 Research grant from the Univ. of the Phil. Office of the V-Chancellor for

Research & Development

Workshop/Training 2008 Apr-Dec UIC Postdoc Institute - Women in Science & Engineering Systems

Transformation Program 2007 Aug Preparing for an Academic Career in the Geosciences: Workshop

Leadership Positions, Committees, membership 2007-2009 Associate member Sigma Xi 2007-2008 student member Association for Women Geoscientists 2007-2008 student member American Association of Petroleum Geologists 2005-2008 student member American Geophysical Union 2005-2008 student member Geological Society of America 2005-2006 President Terra Society (UIC-EAES student organization) 2004-2007 Laboratory manager Organic Geochemistry Laboratory 2003-2008 member Terra Society

Biological and Anthropogenic Markers in Taylor Valley...

Documents

Transcript of Biological and Anthropogenic Markers in Taylor Valley...