<<

The Pennsylvania State University

The Graduate School

Department of Geosciences

INTERPRETING NITROGEN ISOTOPE EXCURSIONS

IN THE SEDIMENTARY RECORD

A Dissertation in

Geosciences and Biogeochemistry

by

James M. Fulton

 2010 James M. Fulton

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2010

The dissertation of James Mark Fulton was reviewed and approved* by the following:

Michael A. Arthur Professor of Geosciences Dissertation Co-Advisor Co-Chair of Committee

Katherine H. Freeman Professor of Geosciences Associate Head for Graduate Programs and Research in Geosciences Dissertation Co-Advisor Co-Chair of Committee

Lee R. Kump Professor of Geosciences

John M. Regan Associate Professor of Environmental Engineering

*Signatures are on file in the Graduate School

iii

ABSTRACT

This dissertation presents an examination of the biological and chemical processes that determine the nitrogen stable-isotope compositions of organic matter (OM)-rich sediments. Using the and Fayetteville Green (FGL), NY as natural laboratories, I evaluate processes that lead to low 15N values of and sediments in anoxic basins. I synthesize sedimentary profiles of bulk N and C stable isotope values with down-core pigment distributions and pigment-specific 15N and 13C values to infer ecological changes responsible for variability in sedimentary 15N values.

The primary goal of this research is to test the hypothesis that cyanobacterial N2 fixation is responsible for black shale 15N values that are near and below 0‰.

Oceanic anoxia has been prevalent during specific times in geological history.

These intervals often are associated with broadly distributed deposits of OM-rich black shales. Widespread anaerobic ammonium oxidation (anammox) and denitrification in near the chemocline (interfaces between oxygenated and anoxic waters) are expected to have led to fixed-N deficiencies that favor cyanobacterial N2 fixation.

Such conditions have also been identified in modern meromictic water bodies, most notably the Black Sea. I present sedimentary scytonemin data from the Black Sea that indicate that cyanobacterial growth in the Holocene Black Sea was variable and responded to climate-induced changes in . The sediment intervals

15 with scytonemin also have low  Ntot values, indicative of cyanobacteria using N2 as a substrate. Green bacterial (GSB) pigments in Black Sea sediments correlate inversely with scytonemin, leading to a phytoplankton productivity model whereby

iv strong density stratification inhibits mixing of high phosphate, low N:P waters to the sea surface. This model may be applied to ancient marine settings, such as the Paleotethys

Ocean during the Permian-Triassic oceanic .

FGL has a shallow (~20 m) chemocline that allows (PSB) to thrive. This compares with a Black Sea chemocline at 80-100 m, which allows GSB growth but inhibits PSB. A similar shallow chemocline may have been present in the

Mesoproterozoic , as PSB biomarkers have been identified in the ancient sediments.

Pigments from PSB and GSB are abundant in the chemocline and sediments of FGL.

Sedimentary intervals with high PSB pigment concentrations correlate with low bulk

15 13  N and  Corg values, supporting my assertion that an increased proportion of PSB biomass causes the low stable-isotope values. Low water-column particulate 15N and

13C values at the chemocline support this model. I also analyzed 13C and 15N values of specific pigments to examine responses of chemocline productivity to elevated mixolimnion productivity. These data demonstrate that sedimentary PSB and GSB pigments were derived from the chemocline, and they provide a ~500-year record of

13 15  CDIC surface-deep gradients and deep water ammonium  N values.

v

TABLE OF CONTENTS

LIST OF FIGURES ...... viii

LIST OF TABLES ...... x

ACKNOWLEDGEMENTS ...... xi

Chapter 1 Introduction: Modern Analogues for Nitrogen Cycling in Ancient Anoxic ...... 1

1.1 Introduction ...... 1 1.2 Nitrogen Isotope Excursions in Black Shales ...... 3 1.3 Ecosystem Responses to Redox-Stratification ...... 5 1.4 Dissertation Outline and Publication Information ...... 7 1.5 Cited References ...... 10

Chapter 2 Scytonemin Production and Nitrogen Fixation in the Holocene Black Sea ...... 17

Abstract ...... 17 2.1 Introduction ...... 18 2.2 Methods ...... 20 2.2.1 Sediments ...... 20 2.2.2 Pigments ...... 20 2.2.3 Nitrogen Isotopes ...... 21 2.3 Results...... 22 2.4 Discussion ...... 25 2.4.1 Scytonemin Sources ...... 25 2.4.2 of Scytonemin Production in the Black Sea ...... 26 2.4.3 Modern Analogue to Ancient ―Anoxic Events‖ ...... 29 2.5 Cited References ...... 31

Chapter 3 Black Sea Nitrogen Cycling and Phytoplankton 15N Signal Preservation during the Holocene ...... 38

Abstract ...... 38 3.1 Introduction ...... 39 3.2 Methods ...... 43 3.2.1 Samples ...... 43 3.2.2 N and C Stable Isotopes ...... 44 3.2.3 N and C Compositions ...... 45

vi

3.2.4 Pigments ...... 46 3.2.5 Pigment Isotope Measurements ...... 47 3.3 Results...... 48 3.3.1 Organic Carbon ...... 48 3.3.2 Sedimentary Nitrogen ...... 49 3.3.3 Pigments ...... 51 3.4 Discussion ...... 52 3.4.1 Sedimentary Nitrogen ...... 52 3.4.2 Compound-Specific N Isotopes ...... 55 3.4.3 Nitrogen Fixation ...... 60 3.4.4 Conceptual Model for N Isotope Distributions ...... 62 3.5 Summary ...... 66 3.6 Cited References ...... 67

Chapter 4 Chemocline-Induced C and N Isotope Excursions Preserved in the Sediments of a ...... 89

Abstract ...... 89 4.1 Introduction ...... 90 4.1.1 Study Site ...... 93 4.2 Methods ...... 94 4.2.1 Sampling ...... 94 4.2.2 Instrumentation ...... 96 4.2.3 Sample Analysis ...... 97 4.2.4 Accumulation Rate Calculations ...... 99 4.3 Results...... 100 4.3.1 Water-Column and Pore-Water DIN ...... 100 4.3.2 Organic Matter Sources ...... 101 4.3.3 Sediment Description ...... 104 13 15 4.3.4 Sedimentary  Corg and  Ntot Values ...... 107 4.4 Discussion ...... 108 4.4.1 C and N Cycling in FGL ...... 108 4.4.2 Deep Basin Sedimentation and Water-Column Productivity ...... 114 4.4.3 Historical Shift in Organic Matter Sources ...... 117 4.4.4 Recent Shift in Chemocline Depth ...... 119 4.4.5 Implications for Geological Samples ...... 120 4.5 Summary ...... 123 4.6 Cited References ...... 124

Chapter 5 Pigment-Specific C and N Isotopes in a Meromictic Lake ...... 148

Abstract ...... 148 5.1 Introduction ...... 149 5.2 Methods ...... 154 5.2.1 Samples ...... 154

vii

5.2.2 Pigment by HPLC ...... 155 5.2.3 Pigment Purification and Isotope Analysis ...... 157 5.3 Results...... 158 5.3.1 Water-Column Pigments ...... 158 5.3.2 Sedimentary Pigments ...... 160 5.3.3 Water-Column and Core-Top Pigment C and N Isotopes ...... 161 5.3.4 Down-Core Pigment N and C Isotopes ...... 162 5.4 Discussion ...... 163 5.4.1 Sources of Sedimentary Chl a and Phe a ...... 163 5.4.2 Chemocline Blooms of PSB, GSB, and Cyanobacteria ...... 166 5.4.3 PSB and GSB Pigment Isotopes in Surface Sediments ...... 168 5.4.4 Sedimentary Pigment 15N Values and Chemocline Productivity ...... 169 5.4.5 Sedimentary Pigment 13C Values and Deep DIC Evolution ...... 170 5.5 Summary ...... 171 5.6 Cited References ...... 172

Appendix A Chapter 3 Supplementary Material ...... 186

Appendix B Chapter 4 Supplementary Material ...... 198

Appendix C Chapter 5 Supplementary Material ...... 201

viii

LIST OF FIGURES

Figure 1-1: C and N Isotope Excursions in Black Shales...... 15

Figure 1-2: Mediterranean Sapropel C and N Isotope Distributions...... 16

Figure 2-1: Identification of Scytonemin in Black Sea Sediments...... 35

Figure 2-2: Map of Core Locations...... 36

Figure 2-3: Sedimentary Units and Chemical Profiles of Black Sea Sediments ...... 37

Figure 3-1: Map of Core Locations...... 74

Figure 3-2: Gravity Core Profiles—Weight % Corg...... 75

13 Figure 3-3: Gravity Core Profiles— Corg...... 76

Figure 3-4: Box Core Compiled Geochemical Profiles...... 77

15 Figure 3-5: Gravity Core Profiles— Ntot...... 78

Figure 3-6: Gravity Core Profiles—Corg:Ntot...... 79

Figure 3-7: Sedimentary Pigment Extract Chromatogram...... 80

Figure 3-8: GGC 71 Corg vs. Ntot Compared with Corg vs. Norg...... 81

Figure 3-9: GGC 71 C and N Bulk, Refractory, and Organic Fractions...... 82

Figure 3-10: Compound-Specific Pphe a 15N and 13C Sediment Profiles...... 83

15 13 Figure 3-11: Compiled  Ntot,  Corg, and Wt. % Corg on Time Axis...... 84

Figure 3-12: Mode-1 and Mode-2 Water-Column Profiles...... 85

Figure 4-1: Fayetteville Green Lake Map with Sampling Locations...... 130

Figure 4-2: Water-Column Properties...... 131

Figure 4-3: Water-Column Seston and Nutrient Profiles...... 132

Figure 4-4: Deep Basin C and N Isotope Records...... 133

ix

Figure 4-5: Shallow Neck C and N Isotope and Pigment Records ...... 134

Figure 4-6: C and N Isotope Values of FGL Sediment Organic Matter Sources ...... 135

Figure 4-7: Deep Basin Composite Core Stratigraphy...... 136

Figure 4-8: Deep Basin Sediment Age Model and Linear Accumulation Rates...... 137

Figure 4-9: Deep Basin Mass Accumulation Rates ...... 138

Figure 4-10: Shallow Neck Core Stratigraphy...... 139

15 Figure 4-11: Deep Basin  Ntot vs. Okenone...... 140

13 Figure 4-12: Deep Basin  Corg vs. Okenone...... 141

15 Figure 4-13: Shallow Neck  Ntot vs. Okenone ...... 142

13 Figure 4-14: Shallow Neck  Corg vs. Okenone...... 143

Figure 5-1: Water Column Pigment Concentrations...... 176

Figure 5-2: Pigment Chromatogram from Unit 1...... 177

Figure 5-3: Sedimentary Pigment Concentrations...... 178

Figure 5-4: Pigment-Specific C Isotope Profiles...... 179

Figure 5-5: Pigment-Specific N Isotope Profiles...... 180

Figure 5-6: PSB Source for Chemocline Biomass...... 181

Figure 5-7: Sedimentary Pigment Accumulation Rates...... 182

x

LIST OF TABLES

Table 3-1: Sedimentary Pigment Identificaiton ...... 86

15 15 Table 3-2:  Nbiomass and  NChl a Values of Cyanobacteria...... 87

Table 3-3: Summary of Properties Related to Mode-1 and Mode-2...... 88

Table 4-1: Gravity Separation of Chemocline Seston ...... 144

Table 4-2: Leaf Litter N and C Stable Isotope Values...... 145

Table 4-3: Comparison of Gray Homogeneous Layers...... 146

Table 4-4: Unit 1 and 2 Mass Accumulation Rates ...... 147

Table 5-1: Pigment Extinction Coefficients ...... 183

Table 5-2: Pigments Identified by HPLC-MS ...... 184

Table 5-3: Water-Column and Surface-Sediment Pigment C and N Isotopes...... 185

xi

ACKNOWLEDGEMENTS

Mike Arthur and Kate Freeman provided excellent advising and academic support throughout my time at Penn State. I am grateful to them for making available the resources and instrumentation required for my projects, and allowing me unfettered access to the mass spectrometers. Thanks also to Lee Kump, not just for organizing field trips to San Salvador Island, but also for piquing my curiosity about modern and ancient stratified anoxic environments. I would like to thank Jay Regan and the multidisciplinary faculty of the Biogeochemical Research Initiative for Education (BRIE) for helping open my eyes to the microbial world, and Brendan Keely from the University of York for hosting me in his lab to study organic pigments. I also acknowledge the Penn State geosciences faculty in general, and especially Jenn Macalady and Chris House for assistance with questions related to geomicrobiology. Finally, thank you to Chris

Junium, Burt Thomas, Tony Riccardi, Katja Meyer, Denny Walizer, Pratigya Polissar, and Genevieve Ellsworth who are all great colleagues and contributed to this work.

I would like to extend a special thank you to my parents and family for helping develop my many interests and my in-laws for their encouragement in my academic pursuits. Most importantly, I would like to thank Deirdre for her partnership, support, and friendship that have made the past 13 years so great. Her encouragement and affirmation helped me enjoy graduate school and maintain perspective during difficult times. Finally, thank you Jada. You may not remember your parents being in graduate school, but you have made our last 2 years more fulfilling than we could have imagined.

Chapter 1

Introduction: Modern Analogues for Nitrogen Cycling in Ancient Anoxic Oceans

1.1 Introduction

Geologists reconstruct Earth history by interpreting observations of rocks, sediments, oceans, ice, and the atmosphere. Due to the broad geographical distribution of thick accumulations of sediments, the record is especially fruitful for studies of changing environmental conditions over time. Through the Deep Sea Drilling

Project and subsequent oceanic sediment-coring programs, geochemists and paleontologists have generated tremendous numbers of stratigraphic records that represent changing sediment characteristics, and reflect subtle to major changes in ocean chemistry and biology over time. Data generated from geological samples require interpretation within constrained frameworks. Early geologists, most notably James

Hutton, proposed the framework of uniformitarianism, noting that observations of modern processes explain features in the rock record. This is still the basis for stratigraphic interpretations of most features from marine sections. It follows that geologists must determine which modern observations to apply to a given set of ancient samples.

The sedimentary record reveals periods in geological history when anoxic waters were widespread in the marine realm (Strauss, 2006). Oceanic anoxic events (OAEs)

2 generally resulted in the deposition of black shales, which record broad changes in phytoplankton assemblages and global carbon and nutrient cycling (Arthur et al., 1985;

Arthur and Sageman, 1994). Black shales are dark colored, containing at least 2% organic carbon (Corg) by weight, and are source rocks for fossil fuel resources (Schlanger et al., 1987). To better understand the processes stimulating black shale deposition, researchers have turned to modern environments as possible analogues. These locations include regions of the ocean with high primary productivity as well as restricted basins with anoxic bottom waters that preserve sinking and sedimentary organic matter (Arthur et al., 1998; Degens and Stoffers, 1976; Glenn and Arthur, 1985; Meyer and Kump,

2008; Piper and Calvert, 2009).

In this dissertation, I present nitrogen stable isotope data from two modern redox- stratified basins, the Black Sea and Fayetteville Green Lake (FGL), NY. The water- column and sedimentary processes are treated as analogous to processes that took place in ancient anoxic oceans and produced organic-matter-rich sediments. In particular, I evaluate mechanisms that cause secular shifts in sedimentary 15N values, generally called isotope excursions in the ancient sedimentary record. The interpretations of

Holocene sedimentary nitrogen isotope excursions presented herein rely on observed properties—biological, physical, and chemical—of the overlying waters. I use these observations to assess changes in the water column that are likely to drive changes recorded in the sediments. Finally, I use data generated from Holocene sediments to outline approaches for interpreting nitrogen isotope excursions in ancient samples.

3 1.2 Nitrogen Isotope Excursions in Black Shales

Fixed-nitrogen availability is a key variable affecting phytoplankton productivity, a major influence on the oceanic carbon cycle and the isotopic composition of sedimentary carbon. Carbon-isotope excursions associated with black shales deposited during OAEs have been studied extensively (Kump and Arthur, 1999). These data are critical to understanding changes in the global carbon budget throughout Earth history, as carbon-isotope excursions result from large-scale perturbations to the carbon cycle. Both positive and negative excursions in 15N values are associated with OAEs (Cao et al.,

2009; Jenkyns et al., 2001; Junium and Arthur, 2007; Kuypers et al., 2004; Rau et al.,

1987), suggesting that changes in nitrogen cycling are tightly connected with changes in carbon cycling (Fig. 1-1). Phytoplankton 15N values in modern water columns and biomass in sediments provide important information about the forms and concentrations of available fixed nitrogen (Altabet et al., 1999; Altabet and Francois, 1994; Liu and

Kaplan, 1989; Sachs and Repeta, 1999). Shifting sedimentary 15N values may reflect significant, relatively rapid changes in nutrient availability.

Nitrogen isotope excursions in the negative direction are commonly interpreted as the result of increased reliance on bacterial N2 fixation in the surface waters (Cao et al.,

2009; Kuypers et al., 2004; Ohkouchi et al., 2006; Rau et al., 1987). Diazotrophic cyanobacteria can convert N2 to NH3 via the nitrogenase enzyme, giving them a distinct advantage in settings lacking only fixed-N to support phytoplankton growth (Tyrrell,

1999; Zehr et al., 2006). Such settings include oligotrophic waters (Capone et al., 1997;

Zehr et al., 2001) and relatively eutrophic settings (Deutsch et al., 2007; Kangro et al.,

4 2007) with nitrogen pools diminished by anaerobic ammonium oxidation (anammox) and denitrification (Jensen et al., 2008; Jetten et al., 1999; Kuypers et al., 2005). The nitrogenase enzyme causes minimal fractionation of N isotopes, leading to biomass that

15 is ~2‰ N-depleted relative to N2 substrate (Carpenter et al., 1997; Hoering and Ford,

1960). Since the atmospheric N2 reservoir is very large relative to ocean and biomass reservoirs, it is not susceptible to spatial or temporal variations in nitrogen content

(Berner, 2006) and, consequently, its 15N value should not have varied from its present value of 0‰ over the Phanerozoic. Accordingly, biomass produced by N2-fixing cyanobacteria should have remained close to -2‰ on Phanerozoic timescales.

A similar but distinct interpretation for negative 15N excursions in sapropels and black shales is that they preserve the signal of nitrogen fixation. Bottom water anoxia limits oxidative respiration which preferentially oxidizes the 15N-depleted OM fraction

(Altabet and Francois, 1994; Van Mooy et al., 2002). Thus, sediments deposited under oxygenated waters are expected to have 15N-enriched 15N values relative to sediments under anoxic waters (which have values near 0‰) even though both may receive phytoplankton source material with 15N values near 0‰. This interpretation has been used to explain the distributions of 15N values in alternating sapropel and normal marine sediments in the Mediterranean Sea (Fig. 1-2) (Sachs and Repeta, 1999). An alternative interpretation asserts that increased cyanobacterial productivity and N2 fixation

15 stimulated Corg accumulation and low  N values in the Mediterranean sapropels

(Meyers and Bernasconi, 2005).

5 A positive 15N excursion has been recognized in the Toarcian black shale section from Yorkshire England (Jenkyns et al., 2001). This excursion is interpreted as resulting from local upwelling of -deficient waters that contained abundant 15N-enriched nitrate, the product of denitrification. Similar conditions are observed in similar modern settings (Brandes et al., 1998; Liu and Kaplan, 1989). As denitrification removes fixed-N from the upwelling water mass, the regional consequence of a Toarcian upwelling zone might be a positive nitrogen isotope excursion. However, it is expected that the global manifestation of increased oceanic denitrification would be a negative nitrogen isotope excursion resulting from N2 fixation in response to phosphate excess (Tyrrell, 1999).

1.3 Ecosystem Responses to Redox-Stratification

Oceans during OAEs were redox-stratified, with oxygenated surface waters and euxinic deep waters (Arthur and Sageman, 1994; Lyons et al., 2009; Meyer and Kump,

2008). is a common characteristic of OAEs (Brown and Kenig,

2004; Cao et al., 2009; Cao et al., 2008; Koopmans et al., 1996; Schouten et al., 2000).

When euxinic waters receive even very low light intensities, green sulfur bacteria (GSB) commonly proliferate at the redox boundary. This is the case in the Black Sea and in

FGL, as a prominent Chlorobium phaeobacteroides populations inhabit the chemoclines at both locations (Culver and Brunskill, 1969; Fry, 1986; Meyer et al., submitted; Repeta and Simpson, 1991). FGL, with its relatively shallow chemocline at 20 m water depth, also hosts a population of purple sulfur bacteria (PSB) that is not present in the Black

Sea, due to light limitation (Overmann and Garcia-Pichel, 2006). PSB appear to inhabit

6 only shallow chemoclines and shallow-water benthic environments (Brocks and

Schaeffer, 2008; Meyer et al., submitted). In both the Black Sea and FGL, the chemocline-dwelling phototrophic populations oxidize H2S from the anoxic deep waters, producing a sharp decrease in H2S concentration at the chemocline. GSB and PSB form a second discrete phytoplankton population that employs different biosynthetic pathways and draws on different nutrient pools than algae and cyanobacteria living in the oxygenated surface waters. These two separate, distinctive populations produce biomass with distinct C and N stable isotopic compositions (Fry, 1986; Fry et al., 1991; Ohkouchi et al., 2005; Velinsky and Fogel, 1999).

In redox-stratified waters such as the Black Sea, the Redfield nutrient ratio is not conserved (Codispoti et al., 1991; Fuchsman et al., 2008). The Redfield ratio derives from the recognition that ocean nutrients and biomass have N:P ratios near 16:1

(Redfield, 1958). In anoxic waters, fixed-N is consumed by anaerobic ammonium oxidation and denitrification (Jensen et al., 2008; Kuypers et al., 2005; Kuypers et al.,

+ 2003; Schubert et al., 2006). These dissimilatory processes convert dissolved NH4 and

- NO3 to gaseous forms N2 and N2O, which are generally biologically unavailable. The

3- decreased burial efficiency of PO4 under low oxygen waters further reduces N:P ratios, and also favors N2-fixing organisms (Ingall and Jahnke, 1994; Van Cappellen and Ingall,

1994). When these fixed-N-deficient waters are upwelled to the surface, diazotrophic cyanobacteria are favored among the phytoplankton (Deutsch et al., 2007). This

3- ecological response is not expected in the PO4 -deficient waters of FGL (Zerkle, 2006), but has been recognized recently as a potentially important component of the Black Sea

N cycle (Fuchsman et al., 2008; Konovalov et al., 2008; McCarthy et al., 2007).

7 1.4 Dissertation Outline and Publication Information

I focus on aspects of Black Sea N cycling in chapters 2 and 3. Chapter 2 examines the connection between redox stratification and surface-dwelling cyanobacteria during the Holocene, since 7800 BP. I introduce the organic biomarker scytonemin, a pigment produced only by cyanobacteria exposed to intense ultraviolet radiation (Garcia-

Pichel et al., 1992). Scytonemin has not been identified previously in deep-sea sediments, and its abundance in Black Sea sediments establishes that cyanobacteria were, at times, prolific at the sea surface. The GSB biomarker bacteriochlorophyll e, indicative of photic-zone euxinia, has an inverse relationship with scytonemin in Black Sea sediments. This relationship implies that growth of scytonemin-producing cyanobacteria was inhibited during times of strong density stratification in the Black Sea. I propose that

3- when density stratification weakens, PO4 -rich deep waters mix into the surface waters, supporting abundant N2 fixation by surface-dwelling cyanobacteria. The sedimentary intervals with scytonemin also have low sedimentary 15N values. This connection between cyanobacteria and low sedimentary15N values supports the typical interpretation that cyanobacterial N2 fixation causes negative nitrogen isotope excursions.

This chapter will be submitted for publication to the journal Nature with coauthors M. A.

Arthur and K. H. Freeman.

In chapter 3, I propose a model that uses variability in the strength of the salinity gradient, a function of regional climate change, to explain down-core relationships between 15N values, bacterial pigments, and scytonemin in the Black Sea. This work

15 includes an examination of the reliability of sedimentary  Ntot values as a record of the

8 primary 15N signal of phytoplankton. Compound-specific N isotope analyses provide better estimates of the N-isotopic compositions of phytoplankton biomass preserved in the sediments. I extracted pigments from Black Sea core samples and measured 15N

15 values of pyropheophytin a (Pphe a), a chlorophyll a derivative. On average,  NPphe a

15 values are 5.1‰ lower than  Ntot values in the Black Sea. This is the expected relationship between phytoplankton chlorophyll a and phytoplankton biomass (Sachs et

15 al., 1999). Some inconsistency is observed between  Ntot values and phytoplankton

15 15 biomass  N values calculated from  NPphe a. This difference is derived in part from inorganic nitrogen associated with detrital clay minerals. Accounting for this

15 15 allochthonous N source reveals that  Norg values are as much as 1‰ lower than  Ntot

15 values in this core. Despite this discrepancy, however, the down-core trends in  Ntot are

15 reasonably consistent with  Norg. This chapter will be submitted to the journal

Paleoceanography with coauthors M. A. Arthur and K. H. Freeman.

In chapters 4 and 5, I report data that describe changes in phytoplankton productivity and organic matter sedimentation in Fayetteville Green Lake (FGL). FGL has a shallow chemocline at 20 m that supports growth of both GSB and PSB. The

Mesoproterozoic ocean may have had a similar chemocline depth, as biomarkers of GSB and PSB have been detected in some marginal marine (possibly a restricted basin) strata of that age (Brocks and Schaeffer, 2008).

In chapter 4, I examine the possibility that chemocline-derived biomass induced sedimentary nitrogen isotope excursions, due to high PSB and GSB productivity. By calculating concentrations and accumulation rates of bulk parameters and GSB and PSB

9 pigments, I demonstrate that the balance between cyanobacterial productivity in the mixolimnion and PSB productivity at the chemocline shifted about 200 years ago,

15 13 resulting in distinctive changes in sedimentary  Ntot and  Corg values. Further, sedimentological evidence from a core taken from above the chemocline reveals a recent, transitory shoaling of the chemocline. When the chemocline was shallower, thin laminations were preserved and sedimentary PSB and GSB pigment concentrations increased. These data demonstrate that there were anoxic, sulfidic waters above the sediments which were previously buried under oxygenated waters. The laminated

15 13 interval records negative excursions in both  Ntot and  Corg values, the result of an increased PSB biomass component. This chapter will be submitted to the journal

Limnology and with coauthors M. A. Arthur, K. H. Freeman, and L. R.

Kump.

Chapter 5 examines the utility of compound-specific isotope studies of complex pigment assemblages preserved in sediments. This work required significant methodological development to detect, purify, and measure 15N and 13C values of chlorophyll a, pheophytin a, pyropheophytin a, bacteriochlorophyll e, bacteriopheophytin a, and okenone. The methods used in this study are presented in this chapter. These data are useful for examining PSB and GSB biomass 15N and 13C values, which reveal

13 15 chemocline  CDIC and  NNH4+ variability over time. In FGL, groundwater hydrology

13 may be the strongest influence on  CDIC. Groundwater influx introduces DIC

(dissolved organic carbon) that is 13C-enriched relative to monimolimnion water

+ 15 (Thompson et al., 1997). Monimolimnion NH4  N values are controlled largely by

10 15N values of sinking and sedimentary biomass. This chapter will be submitted for publication to the journal Organic with coauthors K. H. Freeman, B. J.

Keely, and M. A. Arthur.

1.5 Cited References

Altabet, M., Pilskaln, C., Thunell, R., Pride, C., Sigman, D., Chavez, F., and Francois, R., 1999, The nitrogen isotope biogeochemistry of sinking particles from the margin of the Eastern North Pacific: Deep-Sea Research Part I, v. 46, p. 655-679. Altabet, M.A., and Francois, R., 1994, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nitrate utilization: Global Biogeochemical Cycles, v. 8, p. 103- 116. Arthur, M.A., Dean, W.E., and Laarkamp, K., 1998, Organic carbon accumulation and preservation in surface sediments on the Peru margin: Chemical Geology, v. 152, p. 273-286. Arthur, M.A., Dean, W.E., and Schlanger, S.O., 1985, Variations in the global carbon cycle during the Cretaceous related to climate, volcanism, and changes in atmospheric CO2, in Sundquist, F.T., and Broecker, W.S., eds., The Carbon Cycle and Atmospheric CO2: Natural Variations Archean to Present: Geophysical Monographs 32: Washington, American Geophysical Union, p. 504-530. Arthur, M.A., and Sageman, B.B., 1994, Marine black shales: Depositional mechanisms and environments of ancient deposits: Annual Review of Earth and Planetary Sciences, v. 22, p. 499-551. Berner, R.A., 2006, Geological nitrogen cycle and atmospheric N2 over Phanerozoic time: Geology, v. 34, p. 413-415. Brandes, J.A., Devol, A.H., Yoshinari, T., Jayakumar, D.A., and Naqvi, S.W.A., 1998, Isotopic composition of nitrate in the central Arabian Sea and eastern tropical North Pacific: A tracer for mixing and nitrogen cycles: Limnology and Oceanography, v. 43, p. 1680-1689. Brocks, J.J., and Schaeffer, P., 2008, , a biomarker for purple sulfur bacteria (Chromatiaceae), and other new carotenoid derivatives from the 1640 Ma Barney Creek Formation: Geochimica et Cosmochimica Acta, v. 72, p. 1396-1414. Brown, T.C., and Kenig, F., 2004, Water column structure during deposition of Middle Devonian-Lower Mississippian black and green/gray shales of the Illinois and Michigan Basins: a biomarker approach: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 215, p. 59-85. Cao, C.Q., Love, G.D., Hays, L.E., Wang, W., Shen, S.Z., and Summons, R.E., 2009, Biogeochemical evidence for euxinic oceans and ecological disturbance presaging

11 the end-Permian mass : Earth and Planetary Science Letters, v. 281, p. 188-201. Cao, C.Q., Wang, W., Liu, L., Shen, S.Z., and Summons, R.E., 2008, Two episodes of 13C-depletion in organic carbon in the latest Permian: Evidence from the terrestrial sequences in northern Xinjiang, China: Earth and Planetary Science Letters, v. 270, p. 251-257. Capone, D.G., Zehr, J.P., Paerl, H.W., Bergman, B., and Carpenter, E.J., 1997, Trichodesmium, a globally significant marine cyanobacterium: Science, v. 276, p. 1221-1229. Carpenter, E.J., Harvey, H.R., Fry, B., and Capone, D.G., 1997, Biogeochemical tracers of the marine cyanobacterium Trichodesmium: Deep-Sea Research Part I, v. 44, p. 27-38. Codispoti, L.A., Friederich, G.E., Murray, J.W., and Sakamoto, C.M., 1991, Chemical variability in the Black Sea: implications of continuous vertical profiles that penetrated the oxic/anoxic interface: Deep-Sea Research Part A, v. 38, supplement 2, p. S691-S710. Culver, D.A., and Brunskill, G.J., 1969, Fayetteville Green Lake, New York. V. Sudies of primary production and zooplankton in a meromictic marl lake: Limnology and Oceanography, v. 14, p. 862-872. Degens, E.T., and Stoffers, P., 1976, Stratified waters as a key to the past: Nature, v. 263, p. 22-26. Deutsch, C., Sarmiento, J., Sigman, D.M., Gruber, N., and Dunne, J.P., 2007, Spatial coupling of nitrogen inputs and losses in the ocean: Nature, v. 445, p. 163-167. Fry, B., 1986, Sources of carbon and sulfur nutrition for consumers in three meromictic of New York State: Limnology and Oceanography, v. 31, p. 79-88. Fry, B., Jannasch, H.W., Molyneaux, S.J., Wirsen, C.O., Muramoto, J.A., and King, S., 1991, Stable isotope studies of the carbon, nitrogen, and sulfur cycles in the Black Sea and the Cariaco Trench: Deep-Sea Research Part A, v. 38, supplement 2, p. S1003-S1119. Fuchsman, C.A., Murray, J.W., and Konovalov, S.K., 2008, Concentration and natural stable isotope profiles of nitrogen species in the Black Sea: Marine Chemistry, v. 111, p. 90-105. Garcia-Pichel, F., Sherry, N.D., and Castenholz, R.W., 1992, Evidence for an ultraviolet sunscreen role of the extracellular pigment scytonemin in the terrestrial cyanobacterium Chlorogloeopsis sp.: Photochemistry and Photobiology, v. 56, p. 17-23. Glenn, C.R., and Arthur, M.A., 1985, Sedimentary and geochemical indicators of productivity and oxygen contents in modern and ancient basins: The Holocene Black Sea as the "type" anoxic basin: Chemical Geology, v. 48, p. 325-354. Hoering, T.C., and Ford, H.T., 1960, The isotope effect in the fixation of nitrogen by Azotobacter: Journal of the American Chemical Society, v. 82, p. 376-378. Ingall, E.D., and Jahnke, R.A., 1994, Evidence for enhanced phosphorus regeneration from marine sediments overlain by oxygen depleted waters: Geochimica et Cosmochimica Acta, v. 58, p. 2571-2575.

12 Jenkyns, H.C., Grocke, D.R., and Hesselbo, S.P., 2001, Nitrogen isotope evidence for water mass denitrification during the early Toarcian (Jurassic) oceanic anoxic event: Paleoceanography, v. 16, p. 593-603. Jensen, M.M., Kuypers, M.M.M., Lavik, G., and Thamdrup, B., 2008, Rates and regulation of anaerobic ammonium oxidation and denitrification in the Black Sea: Limnology and Oceanography, v. 53, p. 23-36. Jetten, M.S.M., Strous, M., van de Pas-Schoonen, K.T., Schalk, J., van Dongen, U.G.J.M., Van de Graaf, A.A., Logemann, S., Muyzer, G., van Loosdrecht, M.C.M., and Kuenen, J.G., 1999, The anaerobic oxidation of ammonium: FEMS Microbiology Reviews, v. 22, p. 421-437. Junium, C.K., and Arthur, M.A., 2007, Nitrogen cycling during the Cretaceous, Cenomanian-Turonian oceanic anoxic event II: Geochemistry Geophysics Geosystems, v. 8. Kangro, K., Olli, K., Tamminen, T., and Lignell, R., 2007, Species-specific responses of a cyanobacteria-dominated phytoplankton community to artificial nutrient limitation in the Baltic Sea: Marine Ecology Progress Series, v. 336, p. 15-27. Konovalov, S.K., Fuchsman, C.A., Belokopitov, V., and Murray, J.W., 2008, Modeling the distribution of nitrogen species and isotopes in the water column of the Black Sea: Marine Chemistry, v. 111, p. 106-124. Koopmans, M.P., Koster, J., van Kaam Peters, H.M.E., Kenig, F., Schouten, S., Hartgers, W.A., de Leeuw, J.W., and Damste, J.S.S., 1996, Diagenetic and catagenetic products of : Molecular indicators for photic zone anoxia: Geochimica et Cosmochimica Acta, v. 60, p. 4467-4496. Kump, L.R., and Arthur, M.A., 1999, Interpreting carbon-isotope excursions: carbonates and organic matter: Chemical Geology, v. 161, p. 181-198. Kuypers, M.M.M., Lavik, G., Woebken, D., Schmid, M., Fuchs, B.M., Amann, R., Jorgensen, B.B., and Jetten, M.S.M., 2005, Massive nitrogen loss from the Benguela upwelling system through anaerobic ammonium oxidation: Proceedings of the National Academy of Sciences, v. 102, p. 6478-6483. Kuypers, M.M.M., Sliekers, A.O., Lavik, G., Schmid, M., Jorgensen, B.B., Kuenen, J.G., Sinninghe Damste, J.S., Strous, M., and Jetten, M.S.M., 2003, Anaerobic ammonium oxidation by anammox bacteria in the Black Sea: Nature, v. 422, p. 608-611. Kuypers, M.M.M., van Breugel, Y., Schouten, S., Erba, E., and Damste, J.S.S., 2004, N2- fixing cyanobacteria supplied nutrient N for Cretaceous oceanic anoxic events: Geology, v. 32, p. 853-856. Liu, K.K., and Kaplan, I.R., 1989, The Eastern Tropical Pacific as a source of 15N- enriched nitrate in seawater off southern California: Limnology and Oceanography, v. 34, p. 820-830. Lyons, T.W., Anbar, A.D., Severmann, S., Scott, C., and Gill, B.C., 2009, Tracking euxinia in the ancient ocean: A multiproxy perspective and Proterozoic case study: Annual Review of Earth and Planetary Sciences, v. 37, p. 507-534. McCarthy, J.J., Yilmaz, A., Coban-Yildiz, Y., and Nevins, J.L., 2007, Nitrogen Cycling in the offshore waters of the Black Sea: Estuarine, Coastal and Shelf Sciences, v. 74, p. 493-514.

13 Meyer, K.M., and Kump, L.R., 2008, Oceanic euxinia in Earth history: Causes and consequences: Annual Review of Earth and Planetary Sciences, v. 36, p. 251-288. Meyer, K.M., Kump, L.R., Macalady, J.L., Schaperdoth, I., Fulton, J.M., and Freeman, K.H., submitted, Benthic production of the sulfur phototroph biomarker okenone: Geobiology. Meyers, P.A., and Bernasconi, S.M., 2005, Carbon and nitrogen isotope excursions in mid-Pleistocene sapropels from the Tyrrhenian Basin: Evidence for climate- induced increases in microbial primary production: Marine Geology, v. 220, p. 41-58. Ohkouchi, N., Kashiyama, Y., Kuroda, J., Ogawa, N.O., and Kitazato, H., 2006, The importance of diazotrophic cyanobacteria as primary producers during Cretaceous Oceanic Anoxic Event 2: Biogeosciences, v. 3, p. 467-478. Ohkouchi, N., Nakajima, Y., Okada, H., Ogawa, N.O., Suga, H., Oguri, K., and Kitazato, H., 2005, Biogeochemical processes on the saline meromictic Lake Kaiike, Japan: implications from molecular isotopic evidences of photosynthetic pigments: Environmental Microbiology, v. 7, p. 1009-1016. Overmann, J., and Garcia-Pichel, F., 2006, The phototrophic way of life, in Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H., and Stackebrandt, E., eds., The Prokaryotes, Volume 2: Dordrecht, Springer, p. 32-85. Piper, D.Z., and Calvert, S.E., 2009, A marine biogeochemical perspective on black shale deposition: Earth-Science Reviews, v. 95, p. 63-96. Rau, G.H., Arthur, M.A., and Dean, W.E., 1987, 15N/14N variations in Cretaceous Atlantic sedimentary sequences: implication for past changes in marine nitrogen biochemistry: Earth and Planetary Science Letters, v. 82, p. 269-279. Redfield, A.C., 1958, The biological control of chemical factors on the environment: American Scientist, v. 46, p. 205-221. Repeta, D.J., and Simpson, D.J., 1991, The distribution and recycling of chlorophyll, bacteriochlorophyll and in the Black Sea: Deep-Sea Research Part A, v. 38, supplement 2, p. S969-S984. Sachs, J.P., and Repeta, D.J., 1999, Oligotrophy and nitrogen fixation during eastern Mediterranean sapropel events: Science, v. 286, p. 2485-2488. Sachs, J.P., Repeta, D.J., and Goericke, R., 1999, Nitrogen and carbon isotopic ratios of chlorophyll from marine phytoplankton: Geochimica et Cosmochimica Acta, v. 63, p. 1431-1441. Schlanger, S.O., Arthur, M.A., Jenkyns, H.C., and Scholle, P.A., 1987, The Cenomanian- Turonian oceanic anoxic event, I. Stratigraphy and distribution of organic carbon- rich beds and the marine 13C excursion, in Brooks, J., and Fleet, A.J., eds., Marine Petroleum Source Rocks, Volume 26, Special Publications of the Geological Society of London, p. 371-399. Schouten, S., Van Kaam-Peters, H.M.E., Rijpstra, W.I.C., Schoell, M., and Damste, J.S.S., 2000, Effects of an oceanic anoxic event on the stable carbon isotopic composition of Early Toarcian carbon: American Journal of Science, v. 300, p. 1- 22.

14 Schubert, C.J., Durisch-Kaiser, E., Werhrli, B., Thamdrup, B., Lam, P., and Kuypers, M.M.M., 2006, Anaerobic ammonium oxidation in a tropical freshwater system (Lake Tanganyika): Environmental Microbiology, v. 8, p. 1857-1863. Strauss, H., 2006, Anoxia through time, in Neretin, L.N., ed., Past and Present Water Column Anoxia: Dordrecht, Springer, p. 3-19. Thompson, J.B., Schultz-Lam, S., Beveridge, T.J., and Des Marais, D.J., 1997, Whiting events: Biogenic origin due to the photosynthetic activity of cyanobacterial picoplankton: Limnology and Oceanography, v. 42, p. 133-141. Tyrrell, T., 1999, The relative influences of nitrogen and phosphorus on oceanic primary production: Nature, v. 400, p. 525-531. Van Cappellen, P., and Ingall, E.D., 1994, Benthic phosphorus regeneration, net primary production, and ocean anoxia - a model of the coupled marine biogeochemical cycles of carbon and phosphorus: Paleoceanography, v. 9, p. 677-692. Van Mooy, B., Keil, R., and Devol, A., 2002, Impact of suboxia on sinking particulate organic carbon: Enhanced carbon flux and preferential degradation of amino acids via denitrification: Geochimica et Cosmochimica Acta, v. 66, p. 457-465. Velinsky, D.J., and Fogel, M.L., 1999, Cycling of dissolved and particulate nitrogen and carbon in the Framvaren Fjord, Norway: stable isotopic variations: Marine Chemistry, v. 67, p. 161-180. Zehr, J.P., Church, M.J., and Moisander, P.H., 2006, Diversity, distribution and biogeochemical significance of nitrogen-fixing microorgansims in anoxic and suboxic ocean environments, in Neretin, L.N., ed., Past and Present Water Column Anoxia: Dordrecht, Springer. Zehr, J.P., Waterbury, J.B., Turner, P.J., Montoya, J.P., Omoregie, E., Steward, G.F., Hansen, A., and Karl, D.M., 2001, Unicellular cyanobacteria fix N2 in the subtropical North Pacific Ocean: Nature, v. 412, p. 635-638. Zerkle, A.L., 2006, Microbial Trace Metal Requirements: Limiting Nutrients and Potential [Ph. D. thesis]: University Park, PA, The Pennsylvania State University.

Figure 1-1. Four examples of carbon and nitrogen isotope excursions in black shales associated with Phanerozoic oceanic anoxic events. The shaded regions mark the excursion intervals. All data are from the publications cited in this figure.

16

Figure 1-2. Mediterranean Sapropel carbon and nitrogen isotope distributions. Data are from Meyers and Bernasconi (2005).

Chapter 2

Scytonemin Production and Nitrogen Fixation in the Holocene Black Sea

Abstract

In this chapter we present evidence for the preservation of scytonemin, a pigment produced only by cyanobacteria, in Holocene Black Sea sediments. This is the first reported occurrence of scytonemin in deep-sea sediments, providing unequivocal evidence of preserved cyanobacterial biomass in this setting. Cyanobacteria only produce scytonemin when they are exposed to ultraviolet radiation. Thus, the scytonemin in Black Sea sediments must have derived from cyanobacteria that lived on land surfaces, in shallow benthic settings, or as floating mats in surface waters.

Sedimentary scytonemin concentrations exhibit an inverse relationship with those of bacteriochlorophyll e, a biomarker for green sulfur bacteria, so the deposition of scytonemin in the Black Sea was temporally separated from the occurrence of photic- zone euxinia. The intervals with scytonemin also have low 15N values relative to values expected for phytoplankton growth on nitrate, an indication that cyanobacteria may have been fixing nitrogen. Our model, detailed in chapter 3, proposes that blooms of surface-dwelling pelagic cyanobacteria were stimulated sporadically by upwelling phosphate-rich deep waters. A similar scenario may help explain, for example, biomarker distributions and changes in water-column ecology during the Permian-

18 Triassic oceanic anoxic event. Further, the preservation of scytonemin in abundance in the Black Sea suggests that this diagnostic compound may be useful as a cyanobacterial biomarker in Archaean/Proterozoic samples, helping establish the timing of the evolution of colonial, sheath-forming cyanobacteria.

2.1 Introduction

The study of cyanobacterial evolution and ecology would benefit from the development of additional cyanobacterial biomarkers that may be preserved in sedimentary organic matter. Diazotrophic cyanobacteria fill an important niche in modern oceans, fixing N2 that compensates for nitrate and ammonium lost to denitrification and anaerobic ammonium oxidation (anammox) in oxygen-minimum zones (Capone et al., 1997; Deutsch et al., 2007; Zehr et al., 2001). Diazotrophic cyanobacteria are thought to have been especially vital for maintaining phytoplankton productivity during times in geological history when expanded oxygen-minimum zones or whole-basin anoxia dominated the oceans. High 2-methylhopane index (2-MeHI) values and low bulk sediment 15N values are used to support this model for oceanic anoxic events (OAEs) (Cao et al., 2009; Kuypers et al., 2004; Rau et al., 1987;

Summons et al., 1999). Subsequent work, however, challenged the specificity of the

2-MeHI by demonstrating that purple anoxygenic photosynthetic bacteria also can produce 2-methylbacteriohopanepolyols (2-MeBHP) (Rashby et al., 2007). Further, akinetes rather than actively photosynthesizing cyanobacterial cells may produce higher

19 levels of 2-MeBHPs (Doughty et al., 2009). Thus, high 2-MeHI values may coincide with sediments deposited in environments that did not favor cyanobacterial growth.

We have detected scytonemin, a biomarker for cyanobacteria, in Holocene

Black Sea sediments. The redox-stratified waters of the Black Sea have long been proposed as a modern analogue for oceans during Phanerozoic OAEs (Degens and

Stoffers, 1976; Glenn and Arthur, 1985). Sediments deposited in ocean basins during

OAEs commonly are OM-rich black shales and Holocene Black Sea sediments contain similar high OM concentrations. We identified scytonemin in pigment extracts from

Black Sea sediments based on (1) its maximum absorbance in the UVA range at 388 nm, (2) elution time using a published HPLC method (Airs et al., 2001), and (3) mass spectrum and resonance-induced fragmentation pattern using multistage (MSn) (Fig. 2-1) (Squier et al., 2004a). We also detected smaller amounts of reduced scytonemin and ―component A,‖ two additional forms detected in the analysis of natural samples (Squier et al., 2004b). We report scytonemin and Bchl e concentrations for two deep basin Black Sea cores, GGC 69 and BC 55, which together provide a continuous record of organic matter (OM)-rich sediments deposited since 7.8 ka (kiloannum before present) (Fig. 2-2).

This is the first report of scytonemin in deep-sea sediments. Scytonemin has not been isolated from planktonic cyanobacteria, though we argue that its distribution in

Black Sea deep-basin sediments provides strong evidence for a pelagic, floating-mat source. Scytonemin is a unique pigment found in the polysaccharide sheaths of colonial cyanobacteria, where it provides passive protection from ultraviolet (UV) radiation

(Proteau et al., 1993). Cyanobacteria only produce scytonemin when they are exposed to

20 UV radiation (Garcia-Pichel et al., 1992). It is an indole-alkaloid thought to form through the condensation of aromatic amino acids, though its biosynthetic pathway remains unknown (Soule et al., 2009). Published occurrences of scytonemin include desert soil crusts, rock incrustations, lake fringes, and mats in benthic, intertidal, or sporadically moistened locations (Garcia-Pichel and Castenholz, 1991).

2.2 Methods

2.2.1 Sediments

Sediment samples were collected during Leg 1 of the R/V Knorr 1988 Black Sea cruise. The cores used in this study were maintained frozen (box cores) or refrigerated

(gravity cores) at the Woods Hole Oceanographic Institute core repository until they were sub-sampled in 2007. Samples for N isotope analysis and Corg determination were decarbonated with buffered acetic acid (pH 4). Corg concentration data are reported on decarbonated samples to eliminate the effects of carbonate dilution, especially in Units I and IIb2.

2.2.2 Pigments

Free pigments were extracted into acetone and separated by reversed-phase high performance liquid chromatography (HPLC) using published methods (Airs et al.,

2001). The pigment methods are also described in detail in chapter 5. Separations were

21 achieved on an Agilent 1200 series quaternary gradient HPLC system and two consecutive 150 × 4.6 mm Waters Spherisorb ODS2 columns. Compounds were detected and identified by UV-visible light absorption and multistage tandem mass spectrometry (Agilent 6310 MSn) (Airs et al., 2001; Airs and Keely, 2002; Squier et al.,

2004a). Pigments were quantified by comparison with light absorption response factors calculated for pure standards at concentrations in the linear response range. Scytonemin response was measured at 388 nm, its absorbance maximum using this separation method; Bchl e response was measured at 658 nm, the maximum in the Qy band. Bchl e concentrations are reported for only the [Et,Et] homologue, as the [n-Pr,Et] and

[i-Bu,Et] forms coelute with other chlorins. In the Black Sea, the [Et,Et] form has been identified as the dominant homologue, comprising ~50‰ of all Bchl e in the water column (Repeta and Simpson, 1991).

2.2.3 Nitrogen Isotopes

Nitrogen stable isotope compositions were analyzed by continuous-flow elemental analysis-isotope ratio mass spectrometry (EA-IRMS). The system is composed of a Costech ECS 4010 elemental analyzer, ThermoFinnigan Conflo III interface, and a ThermoFinnigan Deltaplus XP mass spectrometer. 15N values are reported relative to the international air standard. We calibrate our instrument using standard ammonium sulfate salts IAEA-N1 and IAEA-N2. Internal standards are analyzed during sample runs to ensure precision of measurements. Samples analyzed in this way are accurate within ±0.2‰.

22

2.3 Results

Down-core scytonemin distributions are related to previously defined sedimentary units. Holocene Black Sea sediments are divided into three units that have been described in detail (Hay et al., 1991; Ross and Degens, 1974). Arthur and Dean (1998) determined unit boundary ages by robust geomagnetic secular variation data. They found that the ages differ somewhat from published radiocarbon ages (Jones and Gagnon,

1994), particularly for the Unit I/II boundary, suggesting that a component of reworked or recycled carbon caused older radiocarbon ages. Unit III is a bioturbated, clay-rich,

OM-poor interval deposited under fresh/brackish waters prior to 7.8 ka. We did not detect scytonemin in the upper 5 cm of Unit III in GGC69, the extent of our sampling of this unit (Fig. 2-3). Unit II, called the Black Sea Sapropel, consists of continuously laminated OM-rich sediments that were deposited from 7.8 to 2.1 ka. Unit II contains two intervals with abundant scytonemin. Unit I is also laminated and is distinguished from Unit II by increased calcite content from Emiliania huxleyi coccoliths. We only detected scytonemin in the base of Unit I, in samples from the First Invasion Period of E. huxleyi and the base of the Transition Sapropel (Hay et al., 1991). Surface sediment samples from box cores 10, 47, 55 and 78 yielded no scytonemin, indicating that scytonemin is not accumulating in modern slope and deep-basin sediments.

Scytonemin accumulation in Black Sea sediments was most significant during the deposition of Unit II. Unit II sediments are divided into three subunits: an aragonite-rich

23 interval (Unit IIb2; 7.8-6.8 ka), an interval of very high Corg and low carbonate content

(IIb1; 6.8-5.1 ka), and a thick interval characterized by diminished Corg content (IIa; 5.1-

2.1 ka). The lower half of Unit IIb2 (7.8-7.3 ka) contains scytonemin at concentrations up

-1 to 0.3 mg gOC . Unit IIb1 is devoid of scytonemin with the exception of one sample near the IIb1/IIa transition. Unit IIa contains the highest concentrations of scytonemin, with a maximum of 5.8 mg gOC-1. We detected scytonemin throughout Unit IIa and into the base of Unit I in both BC55 and GGC 69. We also detected scytonemin in all Unit IIa samples from 4 gravity cores spanning water depths from 205-1259 m on the basin slopes, but at concentrations about one order of magnitude lower (less than 0.6 mg gOC-

1) than in the deep basin (Fig. 2-3). The spatial distribution of the 6 cores demonstrates that scytonemin sedimentation was a basin-wide phenomenon during the deposition of

Unit IIa (Fig. 2-2).

Sedimentary occurrences of scytonemin and Bchl e in the Black Sea display a general inverse relationship. Bchl e sedimentary distributions are similar to those of isorenieratene from previous studies. These biomarkers record the presence of brown- pigmented green sulfur bacteria (GSB) and photic zone euxinia, at times, in the Holocene

Black Sea water column (Repeta, 1993; Sinninghe Damste et al., 1993; Wakeham et al.,

1995). The similarity between the records of Bchl e and isorenieratene is expected as they are both produced by Chlorobium phaeobacteroides, the GSB species identified in the modern water column. We detected Bchl e [Et,Et] concentrations up to 35 g gOC-1 in Unit IIb and Unit I (Fig. 2-3). We also detected Bchl e in the uppermost samples of

Unit IIa, below the First Invasion Period of E. huxleyi coccoliths. The absence of Bchl e

24 in most of Unit IIa is notable to this study, as that interval contains the highest concentrations of scytonemin.

Holocene Black Sea sediments have bulk 15N values ranging between -0.1‰ and

4.5‰ (Fig. 2-3). Unit III sediments are relatively 15N-enriched, with 15N values near

15 4‰, similar to typical marine values. Unit IIb2 records a decrease in  N values, from

3.9‰ to 1.6‰ in GGC 69, possibly in response to increased cyanobacterial N2 fixation

15 (Blumenberg et al., 2009). Unit IIb1 sediments have high  N values, up to 4.5‰. The

15 minimum  N value of 1.3‰ near the top of Unit IIb1 coincides with the only occurrence of scytonemin in that unit. Unit IIa 15N values are consistently low, between 0.5‰ and

1.7‰ in GGC 69, and -0.1‰ and 0.7‰ in BC 55. These low values are coincident with the highest concentrations of scytonemin. Unit I 15N values in BC 55 are low, between

0.1‰ and 1.0‰, with the exception of the Transition Sapropel (up to 2.2‰) and the increasing trend to 2.3‰ at the core top. In other cores, Unit I 15N values are consistently higher than Unit IIa 15N values (Chapter 3). Even though low sediment

15N values persist through most of Unit I in BC 55, scytonemin is not detected in sediments above the Transition Sapropel.

25 2.4 Discussion

2.4.1 Scytonemin Sources

The provenance of scytonemin in deep-basin Black Sea sediments is intriguing.

Scytonemin has neither been identified in planktonic cyanobacteria nor has its preservation and fluvial transport from land surfaces to offshore sediments been evaluated. Even though scytonemin has no known planktonic source, its distribution in

Black Sea sediments implies a water-column source. Deep-basin Unit I and II sediments contain primarily marine-derived organic matter with a small component of reworked terrigenous organic matter (Arthur et al., 1998). Conversely, the Transition

Sapropel in Unit I contains elevated concentrations of terrigenous material (Hay et al.,

1991) but it does not yield scytonemin. If sedimentary scytonemin did derive from land surfaces, cyanobacterial desert soil crusts in Turkey would be the most likely source.

Climate records from Ukraine indicate that persistent arid conditions did not develop during the middle to late Holocene (Cordova and Lehman, 2005; Kremenetski, 1995).

Arid conditions in Turkey expanded ~4 ka (Wick et al., 2003). This aridity may have stimulated increased development of cyanobacterial soil crusts, but starting ~1000 years after the first occurrence of Unit IIa scytonemin. The arid conditions that developed ~4 ka persist to modern times so we would expect to find scytonemin in Unit I sediments if sedimentary scytonemin derives from Turkey. The absence of scytonemin in Unit I and surface sediments further supports our assertion that scytonemin is not transported from land surfaces to shelf and deep-basin sediments in the Black Sea.

26 Scytonemin has not been detected in surface blooming marine cyanobacteria in modern settings, including diazotrophic Trichodesmium in the oceans and Nodularia in the Baltic Sea (Castenholz and Garcia-Pichel, 2000; Mohlin and Wulff, 2009). These organisms produce other UV-screening compounds classified as mycosporine-like amino acids (MAAs), and apparently are incapable of scytonemin production. In the

Black Sea, infralittoral zone surveys have identified many cyanobacteria, including

Oscillatoria and Lyngbya living in the shallow waters (Aysel et al., 2004; Stoica and

Herndl, 2007). These two genera include scytonemin-producers (Garcia-Pichel and

Castenholz, 1991), and marine planktonic diazotrophs (Paerl, 2000). The expansion of the habitat of near-shore cyanobacteria in the Black Sea to include pelagic UV-exposed surface waters could be responsible for the high concentrations of scytonemin in deep- sea sediments.

2.4.2 Ecology of Scytonemin Production in the Black Sea

The biomarker record for GSB additionally constrains past environmental conditions in the Black Sea. Sediment profiles of Bchl e record temporal variability in the presence of Chlorobiacae living in the Black Sea chemocline (Repeta and Simpson,

1991). GSB growth is limited to times when solar irradiance is great enough at the chemocline to support their photosynthetic growth. In the modern Black Sea, GSB occur at depths between 80-100 m. The presence of Bchl e in Black Sea sediments demonstrates that the chemocline was probably at a similar depth during much of the time when Unit IIb and Unit I were deposited. The absence of Bchl e and

27 isorenieratene in Unit IIa, considered within the context of additional sedimentological and geochemical indicators, shows that from 5.0-2.4 ka the chemocline was deeper, but the water column remained redox-stratified (Sinninghe Damste et al., 1993; Wilkin and

Arthur, 2001).

The Unit IIa time interval appears to have been characterized by a reduction in freshwater influx (Arthur and Dean, 1998; van der Meer et al., 2008). As a result, surface water salinity probably increased and the density gradient eroded, allowing the expansion of the Surface Mixed Layer (SML) to deeper depths. Such conditions would

3- have resulted in the upward mixing of PO4 -rich, low N:P deep waters into the SML and to the sea surface, stimulating the surface blooms of diazotrophic cyanobacteria proposed in this study. As well, greater penetration of oxygenated waters to deeper depths oxidized sulfide and inhibited GSB growth, thus explaining the inverse relationship between Bchl e and scytonemin in our sediment samples.

The down-core intervals with scytonemin coincide with low bulk sediment15N values, a feature that is commonly invoked as evidence for past bacterial N2 fixation (Cao et al., 2009; Kuypers et al., 2004; Rau et al., 1987). We propose that elevated phosphate

3- concentrations [PO4 ] in the absence of proportional increases in fixed nitrogen concentrations stimulated sporadic or seasonal surface blooms of N2-fixing cyanobacteria during times when scytonemin was deposited. This hypothesis draws on the model that

3- PO4 is the ultimate limiting nutrient for phytoplankton productivity (Tyrrell, 1999). N:P nutrient ratios are generally near 16:1 in ocean waters. This ―Redfield Ratio‖ is conserved in phytoplankton biomass, maintaining the balance between N and P

(Redfield, 1958). Where N:P nutrient ratios drop below ~16:1, diazotrophic

28 cyanobacteria have a competitive advantage among the phytoplankton. N2-fixing cyanobacteria produce biomass with distinctive 15N values near -2‰ (Carpenter et al.,

1997; Hoering and Ford, 1960), a signal preserved in sediments under anoxic bottom waters (Sachs and Repeta, 1999). For comparison, normal marine sediments have 15N values greater than 4‰ (Altabet and Francois, 1994). In Black Sea sediments, we find bulk 15N values between these end-members, probably reflecting a shifting balance between N2 fixation and nitrate assimilation (Liu and Kaplan, 1989).

In the modern Black Sea, both surface and deep waters have low N:P ratios

(Codispoti et al., 1991; Fuchsman et al., 2008), the result of water column denitrification and anammox, and decreased burial efficiency of phosphate in sediments (Arthur and

Dean, 1998; Jensen et al., 2008; Kuypers et al., 2003). N2 fixation has been detected in pelagic Black Sea waters (McCarthy et al., 2007). It may be a key component of the N cycle in deeper surface waters, generating low 15N biomass that is remineralized to

+ NH4 and oxidized by anammox bacteria (Konovalov et al., 2008). Surface sediment

15N values between 2.3‰ and 3.5‰ in our box core samples are similar to phytoplankton values and are 15N-depleted compared with nitrate (Coban-Yildiz et al.,

2006; Fuchsman et al., 2008), possibly reflecting inputs from N2-fixing biomass. N2

3- fixation is most likely limited in the modern Black Sea by low [PO4 ] in the SML. The

3- Suboxic and Anoxic Layers contain high [PO4 ] and have low N:P ratios, but they are separated from the SML by the Cold Intermediate Layer (Codispoti et al., 1991; Gregg and Yakushev, 2005). This layer inhibits deep mixing by intensifying the salinity-based density gradient. In the past, erosion of the density gradient by decreased freshwater

29 3- influx could have allowed upward mixing of high [PO4 ], low N:P waters, stimulating large-scale N2 fixation (Chapter 3).

2.4.3 Modern Analogue to Ancient “Anoxic Events”

Nitrogen cycling and stratification in the Paleotethys Ocean may have behaved similarly as in the modern Black Sea. Our results depict a scenario where scytonemin- producing cyanobacteria living at the sea surface are temporally separated from times with photic zone euxinia. Sediments deposited at Meishan, China during the end-

Permian OAE reveal a similar relationship. This OAE coincides with the largest mass extinction in the Phanerozoic record. Below the extinction horizon, 15N values are near

3‰, isorenieratane concentrations are up to ~55 ppm, and 2-methylhopane indices are between 3 and 5% (Cao et al., 2009). These data suggest an ecological structure similar to that in the Black Sea during the deposition of Unit IIb and Unit I. Above the extinction horizon, 15N values are between -1 and 1‰, isorenieratane is not detected, and 2-MeHI values increase as high as 33% (Cao et al., 2009). The high 2-MeHI values may result from akinete cell formation stimulated by decreased nutrient availability following a short-duration upwelling event and cyanobacterial bloom. In this interval the biomarker distributions reflect increased cyanobacterial nitrogen fixation and a loss of photic zone euxinia, as in Unit IIa of the Black Sea sediments.

Black Sea sediments do not contain 2-MeBHPs (Blumenberg et al., 2009), the precursors to 2-methylhopanes used to calculate 2-MeHI values (Summons et al., 1999).

High 2-MeHI values have been used to assign a cyanobacterial source to ancient organic

30 matter. However, their absence in the Black Sea is not surprising, as many species of cyanobacteria do not produce 2-MeBHPs (Summons et al., 1999). The recognition that some sulfide-oxidizing phototrophic bacteria produce 2-MeBHPs also weakens the cyanobacterial specificity of the 2-MeHI, especially in anoxic or redox-stratified environments (Rashby et al., 2007). As well, recent work suggests that a dormant cyanobacterial growth phase may increase 2-MeBHP production (Doughty et al., 2009).

Akinetes, thick-walled, non-vegetative cyanobacterial cells resistant to degradation contain elevated levels of 2-MeBHPs, suggesting that high 2-MeHI values could be an indication of poor growth conditions for cyanobacteria. The presence of abundant akinetes, however, might imply that favorable conditions for cyanobacteria preceded their formation.

Another cyanobacterial biomarker is needed to strengthen arguments related to the evolution of cyanobacteria and nitrogen fixation on the early Earth and during OAEs.

Scytonemin preserved in Archaean-Proterozoic samples could be an important component of testing molecular clock data that constrain the evolution of mat-forming cyanobacteria to after 2.32 Ga (Blank and Sanchez-Baracaldo, 2010). Scytonemin production is common among colonial cyanobacteria (Garcia-Pichel and Castenholz,

1991), including some species that do not produce 2-MeBHPs. Scytonemin is especially abundant in cyanobacterial mats and stromatolites, morphologies well known in

Precambrian rocks. In fact, cyanobacteria may have depended on scytonemin for UV protection on the early Earth (Dillon and Castenholz, 1999). Thus, scytonemin or a derivative with the characteristic scytoneman skeleton (Proteau et al., 1993) could be an important biomarker in ancient samples, an excellent complement to the 2-MeHI.

31 2.5 Cited References

Airs, R.L., Atkinson, J.E., and Keely, B.J., 2001, Development and application of a high resolution liquid chromatographic method for the analysis of complex pigment distributions: Journal of Chromatography A, v. 917, p. 167-177. Airs, R.L., and Keely, B.J., 2002, Atmospheric pressure chemical ionisation liquid chromatography/mass spectrometry of bacteriochlorophylls from Chlorobiaceae: characteristic fragmentations: Rapid Communications in Mass Spectrometry, v. 16, p. 453-461. Altabet, M.A., and Francois, R., 1994, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nitrate utilization: Global Biogeochemical Cycles, v. 8, p. 103- 116. Arthur, M., and Dean, W., 1998, Organic-matter production and preservation and evolution of anoxia in the Holocene Black Sea: Paleoceanography, v. 13, p. 395- 411. Arthur, M.A., Dean, W.E., and Laarkamp, K., 1998, Organic carbon accumulation and preservation in surface sediments on the Peru margin: Chemical Geology, v. 152, p. 273-286. Aysel, V., Erdugan, H., Dural-Tarakci, B., Okudan, E.S., Senkardesler, A., and Aysel, F., 2004, Marine flora of Sinop (Black Sea, Turkey): E.U. Journal of Fisheries and Aquatic Sciences, v. 21, p. 59-68. Blank, C.E., and Sanchez-Baracaldo, P., 2010, Timing of morphological and ecological innovations in the cyanobacteria - a key to understanding the rise in atmospheric oxygen: Geobiology, v. 8, p. 1-23. Blumenberg, M., Seifert, R., Kasten, S., Bahlmann, E., and Michaelis, W., 2009, Euphotic zone bacterioplankton sources major sedimentary bacteriohopanepolyols in the Holocene Black Sea: Geochimica et Cosmochimica Acta, v. 73, p. 750-766. Cao, C.Q., Love, G.D., Hays, L.E., Wang, W., Shen, S.Z., and Summons, R.E., 2009, Biogeochemical evidence for euxinic oceans and ecological disturbance presaging the end-Permian mass extinction event: Earth and Planetary Science Letters, v. 281, p. 188-201. Capone, D.G., Zehr, J.P., Paerl, H.W., Bergman, B., and Carpenter, E.J., 1997, Trichodesmium, a globally significant marine cyanobacterium: Science, v. 276, p. 1221-1229. Carpenter, E.J., Harvey, H.R., Fry, B., and Capone, D.G., 1997, Biogeochemical tracers of the marine cyanobacterium Trichodesmium: Deep-Sea Research Part I- Oceanographic Research Papers, v. 44, p. 27-38. Castenholz, R.W., and Garcia-Pichel, F., 2000, Cyanobacterial responses to UV- radiation, in Whitton, B.A., and Potts, M., eds., The Ecology of Cyanobacteria: Dordrecht, Kluwer Academic Publishers, p. 591-611. Coban-Yildiz, Y., Altabet, M.A., Yilmaz, A., and Tugrul, S., 2006, Carbon and nitrogen isotopic ratios of suspended particulate organic matter (SPOM) in the Black Sea water column: Deep-Sea Research Part II, v. 53, p. 1875-1892.

32 Codispoti, L.A., Friederich, G.E., Murray, J.W., and Sakamoto, C.M., 1991, Chemical variability in the Black Sea: implicaiotns of continous vertical profiles that penetrated the oxic/anoxic interface, in Murray, J.W., ed., Black Sea Oceanography: Results from the 1988 Black Sea Expedition: Deep-Sea Research 38 (Supplement 2A): Oxford, Pergamon Press, p. S691-S710. Cordova, C.E., and Lehman, P.H., 2005, Holocene environmental change in southwestern Crimea (Ukraine) in pollen and soil records: Holocene, v. 15, p. 263-277. Degens, E.T., and Stoffers, P., 1976, Stratified waters as a key to the past: Nature, v. 263, p. 22-26. Deutsch, C., Sarmiento, J., Sigman, D.M., Gruber, N., and Dunne, J.P., 2007, Spatial coupling of nitrogen inputs and losses in the ocean: Nature, v. 445, p. 163-167. Dillon, J.G., and Castenholz, R.W., 1999, Scytonemin, a cyanobacterial sheath pigment, protects against UVC radiation: Implications for early photosynthetic life: Journal of Phycology, v. 35, p. 673-681. Doughty, D.M., Hunter, R.C., Summons, R.E., and Newman, D.K., 2009, 2- Methylhopanoids are maximally produced in akinetes of Nostoc punctiforme: geobiological implications: Geobiology, v. 7, p. 524-532. Fuchsman, C.A., Murray, J.W., and Konovalov, S.K., 2008, Concentration and natural stable isotope profiles of nitrogen species in the Black Sea: Marine Chemistry, v. 111, p. 90-105. Garcia-Pichel, F., and Castenholz, R.W., 1991, Characterization and biological implications of scytonemin, a cyanobacterial sheath pigment: Journal of Phycology, v. 27, p. 395-409. Garcia-Pichel, F., Sherry, N.D., and Castenholz, R.W., 1992, Evidence for an ultraviolet sunscreen role of the extracellular pigment scytonemin in the terrestrial cyanobacterium Chlorogloeopsis sp.: Photochemistry and Photobiology, v. 56, p. 17-23. Glenn, C.R., and Arthur, M.A., 1985, Sedimentary and geochemical indicators of productivity and oxygen contents in modern and ancient basins: The Holocene Black Sea as the "type" anoxic basin: Chemical Geology, v. 48, p. 325-354. Gregg, M.C., and Yakushev, E., 2005, Surface ventilation of the Black Sea's cold intermediate layer in the middle of the western gyre: Geophysical Research Letters, v. 32, p. L03604. Hay, B., Arthur, M.A., Dean, W.E., Neff, E., and Honjo, S., 1991, Sediment deposition in the Late Holocene abyssal Black Sea with climatic and chronological implications: Deep-Sea Research Part A, v. 38, supplement 2, p. S1211-S1233. Hoering, T.C., and Ford, H.T., 1960, The isotope effect in the fixation of nitrogen by Azotobacter: Journal of the American Chemical Society, v. 82, p. 376-378. Jensen, M.M., Kuypers, M.M.M., Lavik, G., and Thamdrup, B., 2008, Rates and regulation of anaerobic ammonium oxidation and denitrification in the Black Sea: Limnology and Oceanography, v. 53, p. 23-36. Jones, G.A., and Gagnon, A.R., 1994, Radiocarbon chronology of Black Sea sediments: Deep-Sea Research Part I, v. 41, p. 531-557.

33 Konovalov, S.K., Fuchsman, C.A., Belokopitov, V., and Murray, J.W., 2008, Modeling the distribution of nitrogen species and isotopes in the water column of the Black Sea: Marine Chemistry, v. 111, p. 106-124. Kremenetski, C.V., 1995, Holocene vegetation and climate history of southwestern Ukraine: Review of Palaeobotany and Palynology, v. 85, p. 289-301. Kuypers, M.M.M., Sliekers, A.O., Lavik, G., Schmid, M., Jorgensen, B.B., Kuenen, J.G., Sinninghe Damste, J.S., Strous, M., and Jetten, M.S.M., 2003, Anaerobic ammonium oxidation by anammox bacteria in the Black Sea: Nature, v. 422, p. 608-611. Kuypers, M.M.M., van Breugel, Y., Schouten, S., Erba, E., and Damste, J.S.S., 2004, N2- fixing cyanobacteria supplied nutrient N for Cretaceous oceanic anoxic events: Geology, v. 32, p. 853-856. Liu, K.K., and Kaplan, I.R., 1989, The Eastern Tropical Pacific as a source of 15N- enriched nitrate in seawater off southern California: Limnology and Oceanography, v. 34, p. 820-830. McCarthy, J.J., Yilmaz, A., Coban-Yildiz, Y., and Nevins, J.L., 2007, Nitrogen Cycling in the offshore waters of the Black Sea: Estuarine, Coastal and Shelf Sciences, v. 74, p. 493-514. Mohlin, M., and Wulff, A., 2009, Interaction effects of ambient UV radiation and nutrient limitation on the toxic cyanobacterium Nodularia spumigena: Microbial Ecology, v. 57, p. 675-686. Paerl, H.W., 2000, Marine Plankton, in Whitton, B.A., and Potts, M., eds., The Ecology of Cyanobacteria: Dordrecht, Kluwer Academic Publishers, p. 121-148. Proteau, P.J., Gerwick, W.H., Garcia-Pichel, F., and Castenholz, R.W., 1993, The structure of scytonemin, an ultraviolet sunscreen pigment from the sheaths of cyanobacteria: Experientia, v. 49, p. 825-829. Rashby, S.E., Sessions, A.L., Summons, R.E., and Newman, D.K., 2007, Biosynthesis of 2-methylbacteriohopanepolyols by an anoxygenic phototroph: Proceedings of the National Academy of Sciences of the United States of America, v. 104, p. 15099- 15104. Rau, G.H., Arthur, M.A., and Dean, W.E., 1987, 15N/14N variations in Cretaceous Atlantic sedimentary sequences: implication for past changes in marine nitrogen biochemistry: Earth and Planetary Science Letters, v. 82, p. 269-279. Redfield, A.C., 1958, The biological control of chemical factors on the environment: American Scientist, v. 46, p. 205-221. Repeta, D.J., 1993, A high resolution historical record of Holocene anoxygenic primary production in the Black Sea: Geochimica et Cosmochimica Acta, v. 57, p. 4337- 4342. Repeta, D.J., and Simpson, D.J., 1991, The distribution and recycling of chlorophyll, bacteriochlorophyll and carotenoids in the Black Sea, in Murray, J.W., ed., Black Sea Oceanography: Results from the 1988 Black Sea Expedition: Deep-Sea Research 38 (Supplement 2A): Oxford, Pergamon Press, p. S969-S984. Ross, D.A., and Degens, E.T., 1974, Recent sediments of Black Sea, in Degens, E.T., and Ross, D.A., eds., The Black Sea--Geology, Chemistry, and Biology: Tulsa, The American Association of Petroleum Geologists, p. 183-189.

34 Sachs, J.P., and Repeta, D.J., 1999, Oligotrophy and nitrogen fixation during eastern Mediterranean sapropel events: Science, v. 286, p. 2485-2488. Sinninghe Damste, J.S., Wakeham, S.G., Kohnen, M.E.L., Hayes, J.M., and de Leeuw, J.W., 1993, A 6,000-year sedimentary molecular record of chemocline excursions in the Black Sea: Nature, v. 362, p. 827-829. Soule, T., Garcia-Pichel, F., and Stout, V., 2009, Gene expression patterns associated with the biosynthesis of the sunscreen scytonemin in Nostoc punctiforme ATCC 29133 in response to UVA radiation: Journal of Bacteriology, v. 191, p. 4639- 4646. Squier, A.H., Airs, R.L., Hodgson, D.A., and Keely, B.J., 2004a, Atmospheric pressure chemical ionisation liquid chromatography/mass spectrometry of the ultraviolet sreening pigment scytonemin: characteristic fragmentations: Rapid Communications in Mass Spectrometry, v. 18, p. 2934-2938. Squier, A.H., Hodgson, D.A., and Keely, B.J., 2004b, A critical assessment of the analysis and distributions of scytonemin and related UV screening pigments in sediments: Organic Geochemistry, v. 35, p. 1221-1228. Stoica, E., and Herndl, G.J., 2007, Bacterioplankton community composition in nearshore waters of the NW Black Sea during consecutive diatom and coccolithophorid blooms: Aquatic Sciences, v. 69, p. 413-418. Summons, R.E., Jahnke, L.L., Hope, J.M., and Logan, G.A., 1999, 2-Methylhopanoids as biomarkers for cyanobacterial oxygenic : Nature, v. 400, p. 554- 557. Tyrrell, T., 1999, The relative influences of nitrogen and phosphorus on oceanic primary production: Nature, v. 400, p. 525-531. van der Meer, M.T.J., Sangiorgi, F., Baas, M., Brinkhuis, H., Sinninghe Damste, J.S., and Schouten, S., 2008, Molecular isotopic and dinoflagellate evidence for Late Holocene freshening of the Black Sea: Earth and Planetary Science Letters, v. 267, p. 426-434. Wakeham, S.G., Sinninghe Damste, J.S., Kohnen, M.E.L., and de Leeuw, J.W., 1995, Organic sulfur compounds formed during early diagenesis in the Black Sea sediments: Geochimica et Cosmochimica Acta, v. 59, p. 521-533. Wick, L., Lemcke, G., and Sturm, M., 2003, Evidence of Lateglacial and Holocene climatic change and human impact in eastern Anatolia: high-resolution pollen, charcoal, isotopic and geochemical records from the laminated sediments of Lake Van, Turkey: Holocene, v. 13, p. 665-675. Wilkin, R.T., and Arthur, M.A., 2001, Variations in pyrite texture, sulfur isotope composition, and iron systematics in the Black Sea: Evidence for Late Pleistocene to Holocene excursions of the O-2-H2S redox transition: Geochimica Et Cosmochimica Acta, v. 65, p. 1399-1416. Zehr, J.P., Waterbury, J.B., Turner, P.J., Montoya, J.P., Omoregie, E., Steward, G.F., Hansen, A., and Karl, D.M., 2001, Unicellular cyanobacteria fix N2 in the subtropical North Pacific Ocean: Nature, v. 412, p. 635-638.

35

Figure 2-1. Indentification of scytonemin in Black Sea sediments. A: Scytonemin structure. B: Base peak mass chromatogram of a sample containing abundant scytonemin. The two peaks of similar magnitude preceding scytonemin are reduced scytonemin and ―component A,‖ two additional manifestations of scytonemin (Squier et al., 2004b). This particular sample contains higher than normal concentrations of reduced scytonemin and component A. C: Mass spectrum of the scytonemin peak.

36

Figure 2-2. Map of core locations for this study. Scytonemin was detected from Unit IIa of all giant gravity cores and BC 55. BC 10, 78, and 47 did not retrieve Unit IIa sediments. None of the four box cores yielded scytonemin in surface sediments.

37

Figure 2-3. Sedimentary units and chemical profiles of Black Sea sediments. The base of Unit I is the first invasion period of E. huxleyi, characterized by abundant coccoliths. That interval is overlain by the Transition Sapropel (T. Sap.), distinguished by low calcite content and higher concentrations of detrital minerals and terigenous organic matter. We include 15N data for GGC 01 as it has a continuous record at one coring location, accurately representing the basin-wide 15N trend from Unit IIa to Unit I (Chapter 3). Note the condensed scale used for scytonemin concentrations in deep-basin Unit IIa samples. The age scale is based on geomagnetic secular variation determinations (Arthur and Dean, 1998).

38 Chapter 3

Black Sea Nitrogen Cycling and Phytoplankton 15N Signal Preservation during the Holocene

Abstract

In this chapter, I present a comprehensive assessment of nitrogen preserved in

Black Sea sediments. I examine the composition of bulk sedimentary nitrogen, partitioned into exchangeable, clay-bound, and organic fractions. This assessment reveals that clay-bound nitrogen comprises up to 38% of bulk nitrogen in the analyzed

15 samples. This inorganic fraction influences  Ntotal values in sediments, especially in

15 samples with relatively low organic-matter content. It follows that  Ntot values may not

15 be adequate for estimating  Norg values in studies designed to characterize organic nitrogen in organic-matter-poor samples. The sediments examined in this study generally

15 15 contain sufficient organic matter that  Ntot values closely approximate  Norg values.

Additionally, I present a profile of compound-specific 15N values of sedimentary pyropheophytin a (Pphe a), which more accurately records changes in phytoplankton

15N values during the Holocene.

15 Bulk  N values in this study support the hypothesis that cyanobacterial N2 fixation has been an important component of the Black Sea nitrogen cycle at times in the

15 Holocene. However, examination of sedimentary  NPphe a values in Black Sea sediments in light of published and new isotopic difference calculations between biomass and chlorophyll a suggest that most of the Pphe a derives from algae, not cyanobacteria.

39 15 + Cyanobacterial biomass was recycled in the photic zone, and the N-depleted NH4 released during heterotrophy was assimilated by other phytoplankton. I outline a model for N2 fixation in the Black Sea that draws upon previous water-column nutrient and hydrographic data to explain the proposed temporal variability in N2 fixation. This conceptual model suggests that variability in the strength of the surface-deep density gradient determines the depth from which remineralized nutrients mix upward, and consequently the degree of fixed-N deficiency in the surface waters.

3.1 Introduction

The Black Sea is a brackish marine basin with a limited connection to the global ocean through the Bosporus Strait. Several large rivers, notably the Danube, Dniester, and Don, flow into the sea leading to a positive water balance and net outflow through the Bosporus (Murray et al., 1991). Mediterranean Sea water also flows into the Black

Sea as an undercurrent through the Bosporus via the Aegean Sea, Dardanelles Strait, and

Marmara Sea (Gunnerson and Ozturgut, 1974). This estuarine circulation traps nutrients in the deep waters, which, when upwelled, stimulate primary productivity in the surface waters (Arthur and Dean, 1998; Meyer and Kump, 2008). The surface waters, influenced by fresh river water and precipitation, are less dense than the high salinity Mediterranean waters. This causes the Mediterranean influx to the Black Sea to sink and mix into the deep water mass. The consequent surface-deep density gradient limits downward mixing of oxygen-rich surface waters, leading to stable stratification and a poorly ventilated deep water mass with a residence time of ~400 years (Murray et al., 1991). This slow

40 ventilation in combination with a relatively large sinking organic matter flux leads to complete oxygen consumption by heterotrophic organisms in the deep water mass

(Deuser, 1974; Meyer and Kump, 2008).

The upper boundary of the anoxic layer (AOL) is marked by the presence of H2S and defines the chemocline. The water mass above the chemocline is divided into three layers with distinct properties. The upper ~40 m are the Surface Mixed Layer (SML), which seasonally mixes and remains oxygenated. Below the SML, a dense water mass accumulates, the Cold Intermediated Layer (CIL), which forms the lower boundary of seasonal mixing. CIL waters have below 8ºC, and contain a steep salinity gradient resulting in strong density stratification confirmed by Brunt-Vaisala frequency calculations (Murray et al., 1991). Between the CIL and the AOL is the Suboxic Layer

(SOL), a water mass essentially devoid of oxygen and . The SOL is a zone of enhanced biological activity that fundamentally affects nitrogen cycling in the sea.

Traditionally, denitrification in the SOL was identified as the primary biological mechanism for fixed-N loss in the Black Sea water column (Murray et al., 1995; Ward and Kilpatrick, 1991). Researchers on a recent expedition were unable to detect denitrification, and therefore proposed that anaerobic ammonium oxidation (anammox), which was detected by several methods, serves as the dominant sink for fixed-N in the

Black Sea SOL (Jensen et al., 2008; Kuypers et al., 2003). The anammox process appears widespread in ocean oxygen-minimum zones (Kuypers et al., 2005), consuming both ammonium and nitrite and producing N2 gas (Jetten et al., 1999). Fixed-N losses near the chemocline diminish the pool of biologically available N in the Black Sea and

41 are reflected in low N:P ratios in Black Sea waters (Fuchsman et al., 2008). Molar N:P ratios of ocean nutrients and marine plankton are typically near 16:1 (Redfield, 1958).

Low N:P ratios in the Black Sea are caused not only by fixed-N removal near the

3- chemocline, but also by low sediment burial rates of PO4 , which accumulates in the deep water mass (Arthur and Dean, 1998; Codispoti et al., 1991). These processes combine to yield nutrient pools with N:P nutrient ratios between 5:1 and 1:1 that favor diazotrophs (Tyrrell, 1999), N2-fixing organisms that use the nitrogenase enzyme to convert N2 to NH3. N2-fixing cyanobacteria are a critical component of the marine N cycle (Capone et al., 1997; Zehr et al., 2001), maintaining levels of fixed nitrogen

3- generally proportional to PO4 (Deutsch et al., 2007). Marine N2-fixing cyanobacteria were also important during times in geological history when water-column anoxia was widespread, such as during some oceanic anoxic events (OAEs) (Kuypers et al., 2004).

15 Bulk ocean sediments typically have  Ntotal values greater than 4‰ (Altabet and

Francois, 1994), whereas sediments deposited during OAEs are commonly near and below 0‰ (Cao et al., 2009; Junium and Arthur, 2007; Kuypers et al., 2004; Rao et al.,

1987). Nitrogenase imparts a ~2‰ isotope effect on N2 assimilated by diazotrophic organisms (Carpenter et al., 1997; Hoering and Ford, 1960). Atmospheric N2 is defined as 0‰, and oceanic dissolved N2 rarely strays far from this value, such that microbial diazotrophic organisms in marine and lacustrine settings typically have biomass 15N values close to -2‰. Thus, sinking particulate organic matter (OM) has low 15N values in waters dominated by N2-fixing cyanobacteria, and when bottom waters are anoxic,

15 these low  Ntot signals are preserved (Sachs and Repeta, 1999). In support of this

42 interpretation, 2-Methylhopane Indices (2-MeHI) indicate cyanobacteria were abundant during OAEs (Cao et al., 2009; Kuypers et al., 2004). Further, in chapter 2, we present scytonemin distributions in Holocene Black Sea sediments that we interpret as indicative of abundant surface-water cyanobacteria during the deposition of two sediment intervals

15 15 with low  Ntot values. The combined biomarker and  Ntot evidence points toward cyanobacterial N2-fixation being an important process within both ancient and modern deep marine anoxic basins.

Additional research on nitrogen in black shales has focused on using compound-

15 15 specific  N values to ensure that low  Ntot values are not diagenetically induced

(Junium, 2010; Kashiyama et al., 2008; Ohkouchi et al., 2006). These studies focus on isolating and measuring 15N values of geoporphyrins and maleimides that are derived from chloropigments (Chikaraishi et al., 2008; Kashiyama et al., 2007). Chloropigments are tetrapyrroles, so they contain four N atoms per molecule. Chlorophyll a (Chl a; I in

Appendix A) is typically the most abundant chloropigment produced by marine phytoplankton. Chl a degrades to many different compounds in Holocene sediments, including pheophytin a (Phe a; II), pyropheophytin a (Pphe a; III), and chlorophyllone

(chlone; IV) which are found in Black Sea surface sediments (Sachs and Repeta, 2000).

In older sediments, these compounds further degrade to geoporphyrins, final products of the Treibs scheme for Chl a degradation (Keely, 2006). The structure of deoxophylloerythroetioporphyin (DPEP; V), a Chl a-derived geoporphyrin found in abundance in many ancient sediments and oils, is included in Appendix A, though it has not been detected in the Black Sea. 15N values of geoporphyrins and chlorins, when

43

15 interpreted in light of modern water column and culture observations, confirm that  Ntot values in organic-matter-rich sediments deposited under anoxic waters preserve a primary phytoplankton signal (Chicarelli et al., 1993; Sachs and Repeta, 1999).

In the Black Sea, deep-water anoxia developed nearly synchronously about 7.8 ka

(kiloannum before present) at water depths below ~200 m (Jones and Gagnon, 1994).

Influx of Mediterranean water began at some point prior to 7.8 ka, probably about 10 ka when global eustatic sea level increased high enough to breach the Bosporus sill. The ensuing stable density stratification and increased primary productivity allowed complete consumption of oxygen in the deep waters by 7.8 ka (Arthur and Dean, 1998). N:P

3- nutrient ratios likely decreased at that time, as denitrification, anammox, and PO4 accumulation in the water column would have commenced. In this study, we examine

Holocene Black Sea sediments for temporal variability in N isotope signals that reflect changes in the structure of nutrient-rich subsurface waters. We report N and C stable isotope data on bulk and refractory sediments, as well as compound-specific Pphe a samples. We propose a conceptual model for N cycling, based on changes in mixing depth, the depth of the chemocline, and the strength of the surface-deep density gradient.

3.2 Methods

3.2.1 Samples

The samples for this study came from cores taken during leg 1 of the R/V Knorr

1988 Black Sea expedition. We analyzed samples from 7 giant gravity cores (GGC)

44 spanning water depths from 205-2190 m. The core locations are all currently under anoxic bottom waters and represent sediments deposited on the marginal slopes and the deep basin floor. The geographical distribution spanned the sea from west to east (Fig 3-

1). Gravity cores collected up to 4 m of sediment. All of the gravity cores used in this study recovered the Unit II/III boundary, which has been dated to ~7.8 ka. Three cores also retrieved ~1 m of Unit III sediments, but the bases of these cores are undated. The gravity cores were supplemented with 4 box cores (BC) from depths between 184 m and

2164 m. These cores sampled the sediment-water interface and the upper layers of Unit

I, an interval that is poorly recovered in gravity cores. BC 55 provided particularly integral material that included all of Unit I, including the transition sapropel, and the Unit

I/II transition.

3.2.2 N and C Stable Isotopes

Samples for bulk stable isotope analysis were freeze-dried, ground to fine powder, and acidified with buffered acetic acid (pH 4) to remove carbonate minerals. After repeated rinsing, the samples were freeze-dried and powdered for bulk organic carbon

(Corg) and total nitrogen (Ntot) stable isotope analysis. Subsamples of selected bulk samples from GGC 71 were additionally oxidized with 10% hydrogen peroxide at 40ºC to remove reactive organic matter (Junium and Arthur, 2007). The residual fractions are termed refractory carbon (Cref) and refractory nitrogen (Nref). These fractions are analogous to Cresidual and NKOBr (Freudenthal et al., 2001) and Nfix (de Lange, 1992).

+ GGC 71 subsamples used for exchangeable NH4 (Nex) determination were not

45 decarbonated. 200-mg powder samples were suspended in 10 ml of 2 N KCl and shaken

+ for one hour to liberate loosely bound NH4 . The KCl solutions were passed through

+ 0.45-m membrane filters and collected for NH4 analysis. To calculate organic N (Norg) content we use the equation: Norg = Ntot – Nref – Nex, and we assume inorganic N is composed solely of Nref and Nex. We discuss this interpretation in section 3.4.1 of this chapter.

Stable N and C isotope analyses were performed using a standard coupled elemental analyzer-isotope ratio mass spectrometer (EA-IRMS) in the Penn State

Biogeochemical Stable Isotope Lab. The continuous-flow system is composed of a

Costech ECS 4010 EA, coupled to a ThermoFinnigan Deltaplus XP mass spectrometer via a ThermoFinnigan Conflo III interface. We use Isodat software (v1.42) to calculate 15N and 13C values, which based on uncertainty of the isotopic compositions of international standards (IAEA-N1, IAEA-N2, NBS 19, and ANU Sucrose) and variable analysis conditions have better than 0.2‰ accuracy. Our analysis of replicate in-house and international standards yields typical precision better than 0.1‰. 15N values are presented relative to standard AIR and 13C values relative to Vienna Pee Dee Belemnite

(VPDB).

3.2.3 N and C Compositions

Bulk and refractory fractions were analyzed for weight percent C and N content by EA, concurrently with isotope ratio determinations. We used Isodat software (v1.42) to calculate peak areas for N2 and CO2 eluted from the EA. The peak areas were

46 converted to weight percent composition by comparison with response factors for standards of known composition analyzed in the same way. All weight percent concentrations are presented for decarbonated samples, eliminating variability caused by high calcite or aragonite concentrations in some samples. Nex concentrations were determined spectrophotometrically using the Salicylate Method (HACH method 8155).

+ The resulting light absorbance values were converted to NH4 concentrations using response factors determined for NH4Cl standard solutions.

3.2.4 Pigments

Core material for pigment extraction was sampled at the Woods Hole

Oceanographic Institute core repository in December 2006. GGC samples had been continuously refrigerated, and BC samples were frozen since collection in 1988.

Sedimentary pigment distributions for three cores, BC 55, GGC 59, and GGC 69, were analyzed using published solvent extraction and high performance liquid chromatography

(HPLC) methods (Airs et al., 2001). We extracted pigments into 100% acetone via repeated (2-3 times) sonication, centrifugation (2000 g, 5 min.), and supernatant decanting. Extracts were filtered through dichloromethane/hexane-extracted cotton wool and dried under a stream of N2 gas in a TurboVap LV concentration workstation (Caliper

LifeSciences). The extracts were stored dry at -20ºC prior to analysis.

Pigment analysis was conducted using an Agilent 1200 series HPLC system, consisting of a vacuum degasser (G1322A), quaternary gradient pump (G1311A), autosampler (G1329A), and 8-channel multiple wavelength detector (MWD; G1365D).

47 The system is controlled by ChemStation Rev. B.01.03-SR 2 software. Separations were achieved in reversed phase on two consecutive Waters Spherisorb ODS2 columns (3m,

150 mm × 4.6 mm I.D.) protected by a pre-column (10 mm × 5 mm I.D.) and a

Phenomenex Security Guard C18 guard column. The mobile phase consisted of a gradient of 0.5 M ammonium acetate, methanol, acetonitrile, and ethyl acetate (Airs et al., 2001). Multistage mass spectrometry (MSn) was performed using an Agilent 6310

LC/MS in positive ion mode using an APCI ion source. Tuning settings were: 4000 nA corona current, 60 psi nebulizer pressure, 5 l min-1 drying gas flow, 350ºC drying , and 400ºC vaporizer temperature. Compound identification was achieved by comparison with published elution times, mass spectra, and MSn fragmentation patterns (Airs et al., 2001; Airs and Keely, 2002; Squier et al., 2004).

3.2.5 Pigment Isotope Measurements

We report compound-specific stable N and C isotope data for Pphe a extracted from sediment samples. Purification methods are described in detail in chapter 5. Initial pigment extractions and reversed phase HPLC separations were performed as described above with the exception that water replaced ammonium acetate in the mobile phase.

Pigment peaks were collected quantitatively based on detection by the MWD at appropriate wavelengths to determine collection intervals. The fractions were further purified using normal phase HPLC (Sachs and Repeta, 2000). After drying, the collected

Pphe a fractions were dissolved in 8% acetone in hexane. The samples were injected into an isocratic flow of the same composition and separated from contaminating compounds

48 on a 5 m, 250 mm × 4.6 mm Agilent SIL silica gel column. Sample peaks were collected again and dried under nitrogen gas. Samples purified using these methods have been shown to be suitably pure for isotope analysis (Ohkouchi et al., 2005; Sachs and

Repeta, 2000). Samples were dissolved in 40 l acetone, transferred to smooth-sided tin capsules (Costech 2.9 × 6 mm), and dried at 40ºC before nanoEA analysis (Polissar et al.,

2009).

3.3 Results

3.3.1 Organic Carbon

Gravity cores from all depths record (1) an abrupt increase in weight percent Corg at the Unit II/III boundary, (2) maximum Corg content in Unit IIb1, and (3) a gradual decrease in Corg content through Units IIa and I (Fig. 3-2). These trends have been recognized in previous studies (Arthur and Dean, 1998; Arthur et al., 1994; Calvert and

Fontugne, 1987; Calvert and Karlin, 1998). Unlike previous studies, all Corg data reported here are for decarbonated samples, in order to reveal the balance between inputs of organic matter and terrigenous detrital material. We also present the most

13 comprehensive record of  Corg values for Holocene Black Sea sediments. The sedimentary unit designations are well established, and are based on basin-wide changes in inorganic and organic carbon content (Arthur et al., 1994; Hay et al., 1991; Ross and

Degens, 1974). Unit III sediments were deposited under fresh-to-brackish conditions, prior to the onset of bottom water anoxia at 7.8 ka. Sediments at the top of Unit III

49

13 generally have higher  Corg values (-26 to -24‰) in the deep basin, and lower values

(-29 to -25‰) in the shallow water cores (Fig. 3-3). Above the Unit II/III transition, sediments were deposited under anoxic waters, and regardless of core location all of the

13 cores record nearly identical  Corg trends. The initial shift to high Corg content in Unit

13 IIb2 is accompanied by basin-wide  Corg values near -26.5‰. The Corg maximum in

13 Unit IIb1 coincides with maximum  Corg values of -23‰. In Units IIa and I, Corg

13 content stabilizes at lower concentrations, and  Corg values likewise are consistently between -25‰ and -24‰, with the exception of the Unit I transition sapropel. The transition sapropel, an interval thought to contain elevated concentrations of terrigenous

13 material, records lower  Corg values and Corg concentrations than the rest of Unit I. Unit

13 I samples above the transition sapropel are similarly C-enriched as Unit IIb1 samples

13 (Figs. 3-3 and 3-4). At the top of Unit I, box cores record a ~1.5‰ decrease in  Corg values up to the surface sediments.

3.3.2 Sedimentary Nitrogen

15  Ntot values follow consistent temporal trends in all cores in this study

15 (Fig. 3-5). Unit III sediments have  Ntot values between 3 and 6‰, typical for marine

15 or lacustrine sediments with low OM content. A trend toward lower  Ntot values in the top of Unit III approaching the Unit II/III boundary parallels moderate increases in Corg

15 content (Fig. 3-2). Above the Unit II/III boundary,  Ntot values drop rapidly to near 1‰ in Unit IIb2, correlative with the initial abrupt increase in Corg content. In Unit IIb1,

50

15  Ntot values reach a maximum of ~4‰, before dropping step-wise to minimum values

15 between 0 and 1‰ in Unit IIa. This decrease in  Ntot is coincident with decreasing Corg

15 content, and culminates in a ~3 ka interval of low sediment  Ntot values in Unit IIa. The

15 transition sapropel near the base of Unit I has relatively high  Ntot values. Above the

15 15 transition sapropel, Unit I sediments return to relatively N-depleted  Ntot values, a trend most obvious in BC 55 and GGC 01 (Figs. 3-4 and 3-5, respectively). All box

15 cores and GGC 01 record a 2-3‰ increase in  Ntot values to the core tops.

Nex comprises a negligible component of Ntot in GGC 71 and does not affect

15  Ntot values. We measured a maximum Nex concentration of 1.2 ppm in Unit IIb2, decreasing to a minimum of 0.7 ppm in the core-top sample. Nref, on the other hand, is a significant component of Ntot. Operationally defined as all N resistant to oxidation by

H2O2, Nref accounts for up to 38‰ of Ntot in samples from GGC 71, a core from relatively shallow water along the Turkish coast. We measured little down-core variability in Nref content, which makes up a larger proportion of Ntot in samples with low Norg content (Fig.

3-9).

We report Corg:Ntot data on a mole:mole basis (Figs. 2-4, 2-6). All gravity cores have maximum values in Unit IIb and record trends to lower values in Unit IIa. For the most part, shallow water cores have Corg:Ntot values that continue to drop or remain steady through Unit I. Deep cores GGC 09, GGC 20, GGC 69, and BC 55 have minimum Corg:Ntot values in Unit IIa and record an increase from Unit IIa to Unit I. We discuss this variability in detail in section 2.4.1 of this paper, illustrating a caveat regarding the interpretations of Corg:Ntot profiles of sediment samples (Calvert, 2004).

51 3.3.3 Pigments

Pigment analysis reveals a complex suite of chlorins and carotenoids (Fig. 3-7;

Table 3-1). Chapter 2 of this dissertation focuses on two pigments: scytonemin (V in

Appendix A), a diagnostic pigment produced by cyanobacteria, and bacteriochlorophyll e

(Bchl e, VI), a pigment produced only by green sulfur bacteria (GSB). In this chapter, we examine the diagenetic products of Chl a. We identified 16 different compounds that derive from Chl a, as well as 4 compounds that are degradation products of Chl b, also produced by phytoplankton in the Black Sea (Table 3-1). Most of these compounds are steryl chlorin esters which have been studied in detail as products of zooplankton grazing

(King and Repeta, 1994). Pphe a is the most abundant Chl a derivative in all Black Sea sediment samples, so we chose this compound for stable isotope analysis. We discuss the

15 implications of  NPphe a values in detail in section 3.4.2 of this chapter.

We also detected trace concentrations of bacteriopheophytin a (Bphe a), and pyrobacteriopheophytin a (Pbphe a) in all Unit I and II Black Sea sediment samples.

These compounds are generally viewed as degradation products of bacteriochlorophyll a

(Bchl a) produced by anaerobic purple sulfur bacteria (PSB) and purple nonsulfur bacteria. GSB also produce small proportions of Bchl a as part of their photosynthetic apparatus. Bphe a and Pbphe a are found in consistent, low concentrations in all samples, in contrast to Bchl e which is only in discrete intervals associated with basin- wide photic zone euxinia. The dissimilar distributions of Bphe a and Pbphe a compared with Bchl e also suggests that they are not derived from PSB living in the chemocline during the same times as GSB. PSB have been isolated from nearshore littoral zone

52 surface sediments in the Black Sea, but are absent from the water column today. Re- suspension of nearshore sediments and deposition in deep waters may be responsible for the occurrence of these compounds in deep-basin sediments (Gorlenko et al., 2005). As well, purple nonsulfur bacteria have been identified in the SML (Koblizek et al., 2006), so these Bchl a-producing organisms may be responsible for the background levels of

Bphe a and Pbphe a in the sediments.

3.4 Discussion

3.4.1 Sedimentary Nitrogen

15 Differences in sedimentary Ntot values can reflect variability in nitrogen sources (Freudenthal et al., 2001). Inorganic detrital N is a commonly ignored N source that can comprise a significant proportion of Ntot, especially in OM-lean samples (de

15 Lange, 1992). To examine the influence of inorganic N on  Ntot values, we focus on

GGC 71 from 411 m water depth along the coast of Turkey, near Sinop (Fig. 3-2). Due to the concentration of detrital clay minerals along the sea margin (Ross and Degens,

1974), this relatively shallow location has a bulk sediment mass accumulation rate that is about 2.6 times greater than in the center of the deep basin (Arthur and Dean, 1998).

Illite, the most abundant clay mineral in Black Sea sediments (Muller and Stoffers, 1974),

+ + commonly contains NH4 as a substitute for K in interlayer positions (Young and Aldag,

1982). Consequently, GGC 71 samples have relatively high inorganic N concentrations.

Nex is a minor component of inorganic N in GGC 71, and negligible in its effect on

53

15  Ntot values. This conclusion has also been reached in other similar studies (de Lange,

1992; Freudenthal et al., 2001). Nref, on the other hand, contributes significantly to Ntot, especially in samples with low OM content. Calvert (2004) illustrated this relationship using plots of weight percent Ntot vs. Corg for samples from several different marine stratigraphic sections. Data examined in this way generally reveal a linear relationship and a positive intercept on the Ntot axis. Assuming relatively stable inputs of inorganic N and variable Norg inputs, the intercept value is equivalent to the average inorganic N content of the sample set. As a result, stratigraphic variability in Corg:Ntot profiles can be a function of changes in the relative amounts of organic and inorganic N rather than reflective of changes in OM sources or diagenetic effects (Calvert, 2004).

The Corg and Ntot data for GGC 71 are strongly correlated, with a linear regression intercept of 0.057% N (Fig. 3-8), representing the predicted average amount of inorganic

N in this core. The average measured Nref content of 0.063 ± 0.010% is similar to the intercept, suggesting Nref does indeed reflect the inorganic N fraction. Linear regression of Corg vs. Norg data for the same samples yields an intercept at the origin (Fig. 3-8). By distinguishing between Ntot and Norg in our samples, we can calculate Corg:Norg ratios and compare them with Corg:Ntot (Fig. 3-9). Corg:Norg values are relatively constant in Units I and II, with an average value of 17.4±1.5 (mol:mol) and no apparent stratigraphic trend.

Corg:Ntot, on the hand, trends from maximum values near 17 in Unit IIb2 to a minimum of

11 in Unit I of GGC 71. Thus, the strong trend in Corg:Ntot in this core may be explained by variability in Norg content superimposed over nearly constant inorganic N content.

The stable isotopic composition of inorganic N is remarkably consistent in Units I

15 and IIa of GGC 71, with an average  Nref value of 3.3±0.3‰ (Fig. 3-9). This

54 consistency suggests that the 15N value of the detrital N source is consistent over time, provided Nref retains its original isotopic composition. This is especially obvious in the

15 15 N-depleted samples in Unit IIa, where  Nref values are consistently greater than 3‰

15 15 even though  Norg values are relatively N-depleted, near 1‰ (Fig. 3-9). On the other hand, it appears that exchange between Norg and Nref may have taken place in Unit IIb2.

15 We measured  Nref values as low as 1.9‰ in Unit IIb2, samples that also have low

15  Ntot values. As weight percent Cref and Nref are nearly constant down-core, we cannot

15 + + explain the low  Nref values in Unit IIb2 by additional substitution of NH4 for K in the interlayers or by the inclusion of OM that is resistant to oxidation by H2O2. The apparent exchange most likely occurs via pore waters, which are expected to contain 15N-depleted

+ + NH4 liberated from OM in that interval. While NH4 ions fixed in clay interlayers are not exchangeable (Young and Aldag, 1982), some form of N transfer is apparent. In

15 15 Cretaceous black shale samples,  Ntot and  Nref values are indistinguishable (Junium and Arthur, 2007), supporting the hypothesis of exchange between the organic and inorganic pools, at least in OM-rich sediments. Based on the depth of Unit IIb2, the

+ apparent ion exchange could be a function of burial duration or pore-water NH4 concentrations, which increase with depth in Black Sea sediments (Manheim and Chan,

1974).

15 The effects of Nref on  Ntot values are most pronounced in samples with low OM

15 15 content and  Norg values significantly higher or lower than  Nref (~3.3‰). Among the cores analyzed for this study, portions of GGC 71, GGC 01, GGC 59 and GGC 48 are most likely to exhibit this effect (Figures 2-2 and 2-5). To illustrate the expected

55

15 15 15 influence of Nref on  Ntot, we present calculated  Norg values and measured  Ntot data for GGC 71 (Fig. 3-9). In Unit IIb, where OM content is highest, the difference

15 15 between  Norg and  Ntot values is small. In Units IIa and I, however, we calculate

15 15 differences approaching 1‰ where  Norg values are significantly N-depleted relative

15 15 to  Nref. The  Norg values in Unit IIa suggest that bacterial N2 fixation was a greater

15 component of the nitrogen cycle than is implied by  Ntot values.

15 These data suggest that we should be cautious when using  Ntot values in place

15 15 of  Norg. In samples with low Norg (<0.2‰) and high illite content,  Ntot values are strongly affected by inorganic N, which in GGC 71 is relatively 15N-enriched. Oxidizing

15 OM with H2O2 has proven to be sufficient for determining the quantity and  N values

15 of inorganic N. In this study  Ntot values are useful, however, as most samples have

15 greater OM content than the upper layers of GGC 71, and  Ntot values do retain the

15 significant trends in  Norg values.

3.4.2 Compound-Specific N Isotopes

Compound-specific 15N values of Pphe a, the most abundant degradation product of Chl a in Black Sea sediments, record temporal variation in the N isotopic composition of phytoplankton. Phytoplankton 15N values are more accurately

15 15 determined from the  NPphe a record than from  Ntot because Pphe a is derived directly from Chl a produced by algae and cyanobacteria in the water column. As described

15 above,  Ntot values approximate the isotopic composition of sedimentary OM, but we

56

15 15 cannot be certain that sedimentary  Norg values correspond to phytoplankton  N values.

Variable influx of terrigenous OM over the duration of Unit I and II deposition

15 could also have a significant effect on  Norg trends. As the Black Sea is an enclosed basin with high river water influx, terrigenous OM may make up a significant component of sedimentary OM (Arthur et al., 1994; Calvert and Fontugne, 1987; Shimkus and

Trimonis, 1974). This appears to be the case for the Transition Sapropel in Unit I, which

13 has relatively low hydrogen index and  Corg values indicative of terrigenous OM

(Arthur and Dean, 1998). Compound-specific 13C analysis of hydrocarbon biomarkers,

13 however, suggests that low  Corg values may be derived from phytoplankton growth on

13 high concentrations of C-depleted CO2(aq) in the SML and CIL (Freeman et al., 1994).

Thus, there is uncertainty about the relative inputs of terrestrial and marine OM to Black

Sea sediments.

Bulk sedimentary OM may also be subjected to diagenetic processes that alter

15 15  Norg values. Sedimentary OM is often N-enriched relative to phytoplankton in the overlying waters, the result of degradation during sinking and early sedimentation

(Altabet and Francois, 1994). However, burial under anoxic waters apparently preserves the phytoplankton 15N signal in bulk OM, as suggested for the Mediterranean Sea

Sapropels (Sachs and Repeta, 1999). A similar argument could be made for the Black

Sea. However, particulate OM in the CIL is 15N-enriched relative to surface water particulate OM, possibly recording the effect of aerobic degradation on the 15N signal of slowly sinking material (Fry et al., 1991). As the thickness of the oxygenated layer has

57 varied during the Holocene (Sinninghe Damste et al., 1993; Wilkin and Arthur, 2001), the oxygen exposure time for sinking particulate matter may have varied. As a result, we might expect to see down-core variability in the degree of diagenetic 15N enrichment of bulk OM.

15 15 Down-core trends in  NPphe a values generally parallel  Ntot values throughout the OM-rich Black Sea sediments (Fig 3-10). Previous research has demonstrated that

15 15  NChl a values are ~5‰ lower than  N values of the source algal biomass. This relationship holds for cultured diatoms, haptophytes, green algae, and brown algae (Sachs et al., 1999). Natural phytoplankton assemblages and OM-rich sediments also yield

15 15 15 15  Ntot-pigment values near 5‰, where  Ntot-pigment =  Ntot –  Npigment, and the pigment is Chl a or one of its diagenetic breakdown products (Ohkouchi et al., 2006; Sachs and

Repeta, 1999). We observe the same relationship in Black Sea samples, as we calculated

15 an average  Ntot-Pphe a value of 5.1±0.9‰ for 13 samples from Units I and II (Fig. 3-10).

15 This suggests that  Ntot values in this study are generally valid indicators of average phytoplankton 15N values at the time of deposition.

15 15 The relationship between  Ntot and  NPphe a values breaks down in the

Transition Sapropel. Pphe a has a 15N value of -0.9‰ in the one Transition Sapropel

15 sample from BC55 (Fig. 3-10). Using the average  Ntot-Chl a value of 5‰, we calculated an expected phytoplankton biomass 15N value of 4.1‰ during the deposition of the Transition Sapropel. This 15N value for phytoplankton is 15N-enriched by 2.2‰

15 15 compared with  Ntot. Thus, the  Ntot value of 1.9‰ is lower than expected for the

Transion Sapropel of BC55, possibly reflecting an increased proportion of terrigenous

58 OM with a low 15N value. This interpretation supports the assertion that the Transition

Sapropel contains a larger component of terrigenous OM, the result of increased riverine influx to the sea (Hay et al., 1991).

15 As described in the next section, we hypothesize that intervals with  Ntot values below ~2‰ in Holocene Black Sea sediments reflect increased levels of cyanobacterial

N2 fixation. Interestingly, in sediments deposited at times when cyanobacterial N2 fixation was important, namely Units IIb2 and Unit IIa, we observe the same ~5‰

15 15 isotopic difference between  Ntot and  NPphe a. There are two published studies that

15 include  Ntot-chl a values of cyanobacteria, one for Synechococcus sp. and the other

15 Anabaena cylindrica (Beaumont et al., 2000; Sachs et al., 1999). Their  Ntot-chl a values are +10.1‰ and -8.5‰, respectively (Table 3-2). These widely different results demonstrate that different types of cyanobacteria partition 15N and 14N differently during

15 chlorophyll synthesis. The  Ntot-chl a values for cyanobacteria both also differ greatly from the relationship observed for algae. Further study is needed to explain this discrepancy, and below we outline our initial interpretation of what this means for the evolution of the Black Sea N cycle.

We propose in chapter 2 that cyanobacteria were abundant during Unit IIb2 and

IIa deposition, producing scytonemin and fixing N2. The production of scytonemin indicates the cyanobacteria were exposed to ultraviolet radiation (Garcia-Pichel et al.,

1992), suggesting they were living at or just below the sea surface. Based on an examination of reported distributions of cyanobacteria in Black Sea waters (Aysel et al.,

2004; Stoica and Herndl, 2007), we suggest in chapter 2 that planktonic, colonial

59

Lyngbya or Oscillatoria proliferated during the deposition of Units IIb2 and IIa. Both of these genera are filamentous and include species capable of N2 fixation and scytonemin production. We grew pure cultures of floating mat-forming Lyngbya to determine the

15 15 15 15 isotopic difference between  Ntot and  NChl a. We also measured  Ntot and  NChl a values for a Lyngbya mat sample collected from Assawoman Island on the Atlantic coast of Virginia. In both culture and field samples, Lyngbya produce Chl a that is ~10‰ 15N- enriched relative to total biomass, similar to published data for Anabaena cylindica

(Table 3-2).

15 As total biomass for N2-fixing cyanobacteria has  N values near -2‰, we predict that scytonemin-producing, N2-fixing cyanobacteria in Black Sea surface waters

15 15 would have produced Chl a with a  N value near 8‰.  NPphe a values in Units IIb2 and IIa range between -3‰ and -5‰, values expected for algae, not cyanobacteria. We take this as evidence that seasonal or sporadic blooms of diazotrophic cyanobacteria fixed

N2 at the sea surface, and most of their biomass decomposed and/or was consumed and recycled by heterotrophic organisms in the SML. A component of cyanobacterial biomass survived degradation, as scytonemin is preserved in the sediments, possibly

15 + because it is found in extracellular sheaths. The N-depleted NH4 released by this process supported the growth of dinoflagellates, haptophytes, and diatoms that comprise the complex algal assemblage in the SML. These organisms produced Chl a that was on average 15N-depleted by 5‰ compared with their total biomass. A recent study of the

Black Sea N cycle proposed a similar scenario for the modern Black Sea (Konovalov et al., 2008). Their model draws on a different source for N2-fixing biomass, one found deeper in the water column; nonetheless they explain that the biomass is rapidly recycled

60 15 + to release N-depleted NH4 . As most phytoplankton biomass is recycled in the upper water column, a short-duration, seasonal cyanobacterial bloom conceivably could influence the 15N composition of the whole phytoplankton community, even though algal biomarker signals are preferentially exported and therefore dominate in the sedimentary record.

3.4.3 Nitrogen Fixation

15 The range of sediment  Ntot values, between 4.5‰ and -0.1‰, suggests that there has been significant variability in nitrogen cycling during the deposition of OM-rich sediments in the Black Sea. Models for nutrient cycling in redox-stratified basins hold that nutrient N deficits in the water column favor N2-fixing organisms, primarily cyanobacteria (Tyrrell, 1999). N:P nutrient ratios in modern Black Sea waters are markedly lower than the Redfield Ratio, with reported values between 1:1 and 7:1, depending on depth in the water column (Fuchsman et al., 2008). As the Black Sea has been continuously stratified since 7.8 ka, it follows that fixed-N deficits probably existed throughout that time, and N2-fixing organisms have been an important component of the

15 ecosystem. Sedimentary  Ntot values uphold this assertion, as values near 0‰ are common, likely the product of bacterial N2 fixation (Blumenberg et al., 2009).

N2 fixation has been detected in the Black Sea water column (McCarthy et al.,

2007). N2-fixation in the modern Black Sea apparently is carried out by obscure populations of small free-living cyanobacteria (Zehr et al., 2001) or by other non- photosynthetic bacteria or archaea (Zehr et al., 2006) that are dispersed in the SML. This

61 is in contrast to other settings like the eutrophic Baltic Sea and oligotrophic ocean gyres, where prominent surface blooms of colonial diazotrophic cyanobacteria develop

(Castenholz and Garcia-Pichel, 2000; Mohlin and Wulff, 2009). The effect of diazotrophy on the N cycle appears to be important in the Black Sea, as it provides a source of 15N-depleted biomass needed to balance the N budget (Konovalov et al., 2008).

In the modern Black Sea, particulate matter in the SML has 15N values between 1 and

6‰ (Coban-Yildiz et al., 2006; Fry et al., 1991). These values support the assertion that

- N2-fixation contributes to phytoplankton growth, as they are lower than those of NO3

15 + 15 ( N = 8-18‰) and NH4 ( N = 6-8‰) that accumulate below the SML (Fuchsman et al., 2008; Velinsky et al., 1991). These 15N-enriched nutrients mix into the SML providing growth substrate for phytoplankton. As phytoplankton take up the nutrients completely, in the absence of another nutrient source, they are expected to have the same high 15N value as the nutrients (Altabet and Francois, 1994). As particulate OM and

15 surface sediments have significantly lower  Ntot values than water-column nutrient sources, N2-fixation likely contributes significantly to phytoplankton growth and the

Black Sea N cycle.

Organic biomarker studies of Black Sea sediments provide evidence for past cyanobacterial growth. Scytonemin is only produced by colonial cyanobacteria that are exposed to high levels of ultraviolet radiation. This requirement constrains their habitat to the upper several meters of the water column or, more likely, on the sea surface itself.

15 We only detected scytonemin in sediment intervals with especially low  Ntot values,

Units IIb2 and IIa. Evidently, when environmental conditions are suitable, a significant

62 population of diazotrophic cyanobacteria blooms at the sea surface, resulting in the production of abundant 15N-depleted biomass. Such blooms have not been observed in the modern sea, and core-top sediments do not contain scytonemin. Hopanoid

15 distributions in Holocene Black Sea sediments, in conjunction with low  Ntot values, have also been taken as evidence supporting the prominence of cyanobacteria in the

Black Sea during the deposition of Units I and II (Blumenberg et al., 2009). The hopanoid study was not of high enough resolution to reveal potential decreases in

15 hopanoid concentrations associated with shifts to higher  Ntot values in Units I and IIb1.

However, Blumenberg et al. (2009) propose that cyanobacterial markers are prominent

15 during the shift to lower  Ntot values starting at the Unit II/III boundary. We show that the same negative 15N excursion was a basin-wide phenomenon and thus that cyanobacterial N2-fixation was broadly distributed in the sea (Fig. 3-11).

3.4.4 Conceptual Model for N Isotope Distributions

In Black Sea subsurface waters, high nutrient concentrations occur in two depth ranges: (1) a ~40 m-thick interval spanning the boundary between the CIL and SOL, and

(2) throughout the anoxic layer including all depths below the chemocline (Fig. 3-12)

(Codispoti et al., 1991). Rapid cycling of both nitrogen and phosphorus in the SOL removes nutrients from this layer (Murray et al., 1995; Oguz et al., 2001), producing the distinction between the two nutrient pools. These two nutrient pools have different characteristics that cause distinct phytoplankton responses when they are upwelled. The

- 3- CIL/SOL pool contains NO3 and PO4 at a ratio near 5:1 (Fuchsman et al., 2008).

63

15  NNO3- values in the CIL/SOL pool average near 8‰ (Fuchsman et al., 2008). A simple isotope balance calculation predicts that phytoplankton growth on this nutrient pool

15 would result in  Nbiomass values near +1.1‰, assuming that the new biomass would have

N:P = 16:1 and newly fixed N with a 15N value of -2‰ accounts for the entire N deficit.

We term phytoplankton growth on this pool as ―Mode-1.‖

+ 3- The upper ~50 m of the nutrient pool in the anoxic layer contains NH4 and PO4

+ 15 at a ratio near 1:1 (Fig. 3-12) (Codispoti et al., 1991; Fuchsman et al., 2008). NH4  N values are near 7‰ (Velinsky et al., 1991), so phytoplankton growth on this nutrient pool

15 would result in  Nbiomass values near -1.4‰, using the same assumptions as above. This second type of phytoplankton productivity is ―Mode-2.‖ The 2.5‰ difference between phytoplankton 15N values produced by mode-1 and mode-2 is similar to the range of values observed in sediments from Units I and II. These simple calculations demonstrate that upwelling nutrients into the SML decreases phytoplankton 15N values even though

- + 15 the upwelled NO3 and NH4 are N-enriched. Phytoplankton growth by either mode results in biomass 15N values much lower than deep nutrient 15N values, though mode-

2 provides a stronger lever to drive total phytoplankton 15N values to near or below 0‰.

These calculations assume that the subsurface nutrient pools have had similar 15N enrichment throughout the Holocene. This is a reasonable assumption as diverse redox- stratified environments like the Cariaco Basin (Fry et al., 1991), Framvaren Fjord

(Velinsky and Fogel, 1999), and Fayetteville Green Lake (Chapter 4) all have nutrient pools that are 15N-enriched, presumably by nitrification/denitrification and anammox. A down-core examination of the C isotopic composition of diploptene, a possible biomarker

64 for nitrifying bacteria, and ladderanes produced by anammox bacteria could also confirm the presence and relative abundances of these bacteria in the Black Sea water column over time (Freeman et al., 1994; Jaeschke et al., 2009). Our calculations also do not account for phytoplankton growth on nutrients added to the SML via river runoff. We

- expect that NO3 influx would be proportional to river water influx and thus would be greater in mode-2. As nitrate is commonly 15N-enriched in rivers, its addition to the SML probably causes the 15N enrichment of the total phytoplankton assemblage compared with phytoplankton using deep nutrients in mode-1 and mode-2 (Table 3-3).

The density structure of Black Sea water column allows nutrients to accumulate below the CIL by limiting seasonal mixing and the thickness of the SML. Dense CIL waters are generated in the winter and may have two sources, one being cold surface water (near 0ºC) from the northwest shelf and the other, cold surface water (6-7ºC) in the gyres (Oguz, 2002; Oguz and Besiktepe, 1999). Both of these sources produce water with density that is equivalent to that of water in the CIL. Vertical density stratification in the CIL in very strong, as indicated by the steep salinity gradient and calculations of

Brunt-Vaisala frequency (Murray et al., 1991). Wintertime observations from 2003 showed that anomalously cold weather cooled SML waters in the gyres to the point that they mixed deeply into the CIL (Gregg and Yakushev, 2005). This mixing would have delivered nutrient-rich waters from the CIL to the SML. Thus, upwelling of CIL/SOL nutrient-rich waters and OM production by mode-1 may depend on especially cold winter temperatures. The addition of cold water to the CIL also maintains a strong density gradient, which ultimately inhibits the extension of deeper mixing below the chemocline and mode-2 production. Mode-1 most likely dominated during much of Unit IIb and I

65 deposition. High Caspian Sea levels during these time intervals suggest that a cold, moist climate dominated the region (Chepalyga, 1985). The cold temperatures allowed regular mixing of the SML and CIL and upwelling of CIL/SOL nutrients. High influx of fresh river water to the SML during these times also maintained the salinity gradient that ultimately strengthened stratification.

Mode-2 represents production with deeper mixing and upwelling of deep waters from below the chemocline. Unit IIa, deposited from 5.1 to 2.1 ka, contains evidence for mode-2. Unit IIa lacks pigments that signify GSB growth, providing evidence for a deeper chemocline during that time (Sinninghe Damste et al., 1993). Further evidence from GGC 48 sedimentation patterns and pyrite framboid size distributions point toward a deeper chemocline during Unit IIa deposition (Wilkin and Arthur, 2001). Alkenone hydrogen isotope values and dinoflagellate distributions suggest the Black Sea was characterized by relatively high surface water salinity during Unit IIa deposition (van der

Meer et al., 2008). The Caspian Sea level curve supports this assertion, as low sea levels at that time suggest the regional climate was warm and dry (Chepalyga, 1985). Taken together, these data demonstrate that though bottom waters remained anoxic, the surface- deep salinity gradient was weakened, possibly by decreased freshwater influx. The weaker density gradient allowed deeper mixing of oxygenated surface waters, eroding the chemocline, and significant upwelling of nutrient-rich waters from below the chemocline.

The model described above predicts that phytoplankton biomass produced during mode-2 (e.g. Unit IIa) deposition would have relatively low 15N values. This is

15 confirmed by sedimentary  Ntot values, which reach minimum values in Unit IIa (Fig.

3-11). Unit IIa sediments also provide evidence for sea-surface-dwelling, N2-fixing

66 cyanobacteria as a primary driver for low 15N values. Conversely, the 15N-enriched values in Units I and IIb suggest a trend toward increased mode-1 productivity. Though

N2 fixation is supported by mode-1 conditions, it was not as prevalent and, based on scytonemin data, may not have been associated with significant floating cyanobacterial mats in the Black Sea (Chapter 2). The modern Black Sea fits into mode-1, as evidenced

15 by relatively high core-top  Ntot values, relatively low levels of N2-fixation in the water column, and the modern salinity and redox structures. The first ~600 years of Unit IIb2 contain low concentrations of scytonemin and Bchl e, indicators for both mode-1 and mode-2 productivity (Chapter 2 Fig 2-3). We suspect that during that transitional time, instability in the water column may have resulted in relatively rapid shifts in upwelling

3- mode. As well, the initial release of sedimentary PO4 under anoxic bottom waters may have increased selective pressure toward N2-fixation in the surface waters even though stratification was strong.

3.5 Summary

Regional climate is the ultimate driver of variations in sedimentary 15N values in the Black Sea. A warm and dry climate regime weakens the salinity gradient in the Black

3- Sea, allowing for deeper mixing of the surface waters consequent upwelling of PO4 -rich deep waters. Diazotrophic cyanobacterial blooms took advantage of these conditions that favor N2-fixation. Such was the case during the deposition of Units IIb2 and IIa. A cool, wet climate, on the other hand, strengthens the salinity gradient, inhibiting deep mixing and limiting N2-fixation. These conditions were dominant during the deposition of Units

67

IIb1 and I. Thus, the depth of wintertime mixing determines nutrient dynamics in the surface waters, with variability in the productivity of N2-fixing cyanobacteria a primary control on phytoplankton 15N values.

Nitrogen in deep-basin Black Sea sediments derives primarily from phytoplankton biomass. Marginal sediments receive a greater influx of clay minerals than deep basin sediments, resulting in significantly higher bulk mass accumulation rates.

The greater relative proportion of clay in a marginal core from 411 m results in a total N pool containing up to 38% inorganic N. Accounting for inorganic N results in a

15 15 15 difference of up to ~1‰ between  Ntot and  Norg. The same general trends in  Ntot values are recorded in all cores, however, regardless of clay content. Temporal variability in the 15N values of phytoplankton generated the down-core trends obvious in sediments from all depths below the chemocline. We confirmed this assertion by measuring compound-specific 15N values of Pphe a, a diagenetic product of Chl a. Both

15 15  NPphe a and  Ntot values follow the same general trend, with the expected ~5‰ offset between the values in all samples except the Transition Sapropel. This sedimentary analysis demonstrates that nitrogen cycling in the Black Sea is controlled by basin-wide processes which produce the same isotopic signatures at all locations.

3.6 Cited References

Airs, R.L., Atkinson, J.E., and Keely, B.J., 2001, Development and application of a high resolution liquid chromatographic method for the analysis of complex pigment distributions: Journal of Chromatography A, v. 917, p. 167-177. Airs, R.L., and Keely, B.J., 2002, Atmospheric pressure chemical ionisation liquid chromatography/mass spectrometry of bacteriochlorophylls from Chlorobiaceae:

68 characteristic fragmentations: Rapid Communications in Mass Spectrometry, v. 16, p. 453-461. Altabet, M.A., and Francois, R., 1994, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nitrate utilization: Global Biogeochemical Cycles, v. 8, p. 103- 116. Arthur, M., and Dean, W., 1998, Organic-matter production and preservation and evolution of anoxia in the Holocene Black Sea: Paleoceanography, v. 13, p. 395- 411. Arthur, M., Dean, W., Neff, E., Hay, B., King, J., and Jones, G., 1994, Varve calibrated records of carbonate and organic carbon accumulation over the last 2000 years in the Black Sea: Global Biogeochemical Cycles, v. 8, p. 195-217. Aysel, V., Erdugan, H., Dural-Tarakci, B., Okudan, E.S., Senkardesler, A., and Aysel, F., 2004, Marine flora of Sinop (Black Sea, Turkey): E.U. Journal of Fisheries and Aquatic Sciences, v. 21, p. 59-68. Beaumont, V.I., Jahnke, L.L., and Des Marais, D.J., 2000, Nitrogen isotopic fractionation in the synthesis of photosynthetic pigments in Rhodobacter capsulatus and Anabaena cylindrica: Organic Geochemistry, v. 31, p. 1075-1085. Blumenberg, M., Seifert, R., Kasten, S., Bahlmann, E., and Michaelis, W., 2009, Euphotic zone bacterioplankton sources major sedimentary bacteriohopanepolyols in the Holocene Black Sea: Geochimica et Cosmochimica Acta, v. 73, p. 750-766. Calvert, S.E., 2004, Beware intercepts: interpreting compositional ratios in multi- component sediments and sedimentary rocks: Organic Geochemistry, v. 35, p. 981-987. Calvert, S.E., and Fontugne, M.R., 1987, Stable carbon isotopic evidence for the marine origin of the organic matter in the Holocene Black Sea sapropel: Chemical Geology (Isotope Geoscience Section), v. 66, p. 315-322. Calvert, S.E., and Karlin, R.E., 1998, Organic carbon accumulation on the Holocene sapropel of the Black Sea: Geology, v. 26, p. 107-110. Cao, C.Q., Love, G.D., Hays, L.E., Wang, W., Shen, S.Z., and Summons, R.E., 2009, Biogeochemical evidence for euxinic oceans and ecological disturbance presaging the end-Permian mass extinction event: Earth and Planetary Science Letters, v. 281, p. 188-201. Capone, D.G., Zehr, J.P., Paerl, H.W., Bergman, B., and Carpenter, E.J., 1997, Trichodesmium, a globally significant marine cyanobacterium: Science, v. 276, p. 1221-1229. Carpenter, E.J., Harvey, H.R., Fry, B., and Capone, D.G., 1997, Biogeochemical tracers of the marine cyanobacterium Trichodesmium: Deep-Sea Research Part I, v. 44, p. 27-38. Castenholz, R.W., and Garcia-Pichel, F., 2000, Cyanobacterial responses to UV- radiation, in Whitton, B.A., and Potts, M., eds., The Ecology of Cyanobacteria: Dordrecht, Kluwer Academic Publishers, p. 591-611. Chepalyga, A.L., 1985, Inland sea basins, in Velichko, A.A., ed., Late Quaternary Environments of the Soviet Union: Minneapolis, University of Minnesota Press, p. 229-247.

69 Chicarelli, M.I., Hayes, J.M., Popp, B.N., Eckardt, C.B., and Maxwell, J.R., 1993, Carbon and nitrogen isotopic compositions of alkyl porphyrins from the Triassic Serpiano oil shale: Geochimica et Cosmochimica Acta, v. 57, p. 1308-1311. Chikaraishi, Y., Kashiyama, Y., Ogawa, N.O., Kitazato, H., Satoh, M., Nomoto, S., and Ohkouchi, N., 2008, A compound-specific isotope method for measuring the stable nitrogen isotopic composition of tetrapyrroles: Organic Geochemistry, v. 39, p. 510-520. Coban-Yildiz, Y., Altabet, M.A., Yilmaz, A., and Tugrul, S., 2006, Carbon and nitrogen isotopic ratios of suspended particulate organic matter (SPOM) in the Black Sea water column: Deep-Sea Research Part II, v. 53, p. 1875-1892. Codispoti, L.A., Friederich, G.E., Murray, J.W., and Sakamoto, C.M., 1991, Chemical variability in the Black Sea: implications of continuous vertical profiles that penetrated the oxic/anoxic interface: Deep-Sea Research Part A, v. 38, supplement 2, p. S691-S710. de Lange, G.J., 1992, Distribution of exchangeable, fixed, organic, and total nitrogen in interbedded turbiditic/pelagic sediments of the Madeira , eastern North Atlantic: Marine Geology, v. 109, p. 95-114. Deuser, W.G., 1974, Evolution of anoxic conditions in the Black Sea during the Holocene, in Degens, E.T., and Ross, D.A., eds., The Black Sea--Geology, Chemistry, and Biology: Tulsa, American Association of Petroleum Geologists Memoirs, p. 133-136. Deutsch, C., Sarmiento, J., Sigman, D.M., Gruber, N., and Dunne, J.P., 2007, Spatial coupling of nitrogen inputs and losses in the ocean: Nature, v. 445, p. 163-167. Freeman, K.H., Wakeham, S.G., and Hayes, J.M., 1994, Predictive isotopic biogeochemistry: Hydrocarbons from anoxic marine basins: Organic Geochemistry, v. 21, p. 629-644. Freudenthal, T., Wagner, T., Wenzhofer, F., Zabel, M., and Wefer, G., 2001, Early diagenesis of organic matter from sediments of the eastern subtropical Atlantic: Evidence from stable nitrogen and carbon isotopes: Geochimica et Cosmochimica Acta, v. 65, p. 1795-1808. Fry, B., Jannasch, H.W., Molyneaux, S.J., Wirsen, C.O., Muramoto, J.A., and King, S., 1991, Stable isotope studies of the carbon, nitrogen, and sulfur cycles in the Black Sea and the Cariaco Trench: Deep-Sea Research Part A, v. 38, supplement 2, p. S1003-S1119. Fuchsman, C.A., Murray, J.W., and Konovalov, S.K., 2008, Concentration and natural stable isotope profiles of nitrogen species in the Black Sea: Marine Chemistry, v. 111, p. 90-105. Garcia-Pichel, F., Sherry, N.D., and Castenholz, R.W., 1992, Evidence for an ultraviolet sunscreen role of the extracellular pigment scytonemin in the terrestrial cyanobacterium Chlorogloeopsis sp.: Photochemistry and Photobiology, v. 56, p. 17-23. Gorlenko, V.M., Mikheev, P.V., Rusanov, I.I., Pimenov, N.V., and Invanov, M.V., 2005, Ecophysiological properties of photosynthetic bacteria from the Black Sea chemocline zone: Microbiology, v. 74, p. 239-247.

70 Gregg, M.C., and Yakushev, E., 2005, Surface ventilation of the Black Sea's cold intermediate layer in the middle of the western gyre: Geophysical Research Letters, v. 32, p. L03604. Gunnerson, C.G., and Ozturgut, E., 1974, The Bosporus, in Degens, E.T., and Ross, D.A., eds., The Black Sea--Geology, Chemistry, and Biology: Tulsa, The American Association of Petroleum Geologists, p. 99-114. Hay, B., Arthur, M.A., Dean, W.E., Neff, E., and Honjo, S., 1991, Sediment deposition in the Late Holocene abyssal Black Sea with climatic and chronological implications: Deep-Sea Research Part A, v. 38, supplement 2, p. S1211-S1233. Hoering, T.C., and Ford, H.T., 1960, The isotope effect in the fixation of nitrogen by Azotobacter: Journal of the American Chemical Society, v. 82, p. 376-378. Jaeschke, A., Ziegler, M., Hopmans, E.C., Reichart, G.J., Lourens, L.J., Schouten, S., and Damste, J.S.S., 2009, Molecular fossil evidence for anaerobic ammonium oxidation in the Arabian Sea over the last glacial cycle: Paleoceanography, v. 24. Jensen, M.M., Kuypers, M.M.M., Lavik, G., and Thamdrup, B., 2008, Rates and regulation of anaerobic ammonium oxidation and denitrification in the Black Sea: Limnology and Oceanography, v. 53, p. 23-36. Jetten, M.S.M., Strous, M., van de Pas-Schoonen, K.T., Schalk, J., van Dongen, U.G.J.M., Van de Graaf, A.A., Logemann, S., Muyzer, G., van Loosdrecht, M.C.M., and Kuenen, J.G., 1999, The anaerobic oxidation of ammonium: FEMS Microbiology Reviews, v. 22, p. 421-437. Jones, G.A., and Gagnon, A.R., 1994, Radiocarbon chronology of Black Sea sediments: Deep-Sea Research Part I, v. 41, p. 531-557. Junium, C.K., 2010, Nitrogen Biogeochemistry and Ancient Oceanic Anoxia [Ph. D. thesis]: University Park, PA, The Pennsylvania State University. Junium, C.K., and Arthur, M.A., 2007, Nitrogen cycling during the Cretaceous, Cenomanian-Turonian oceanic anoxic event II: Geochemistry Geophysics Geosystems, v. 8. Kashiyama, Y., Kitazato, H., and Ohkouchi, N., 2007, An improved method for isolation and purification of sedimentary porphyrins by high-performance liquid chromatography for compound-specific isotopic analysis: Journal of Chromatography A, v. 1138, p. 73-83. Kashiyama, Y., Ogawa, N.O., Shiro, M., Tada, R., Kitazato, H., and Ohkouchi, N., 2008, Reconstruction of the biogeochemistry and ecology of photoautotrophs based on the nitrogen and carbon isotopic compositions of vanadyl porphyrins from Miocene siliceous sediments: Biogeosciences, v. 5, p. 797-816. Keely, B.J., 2006, Geochemistry of chlorophylls, in Grimm, B., Porra, R.J., Rudiger, W., and Scheer, H., eds., Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications: Dordrecht, Springer, p. 535-561. King, L.L., and Repeta, D.J., 1994, Phorbin steryl esters in Black Sea sediment traps and sediments: A preliminary evaluation of their paleooceanographic potential: Geochimica et Cosmochimica Acta, v. 58, p. 4389-4399. Koblizek, M., Falkowski, P.G., and Kolber, Z.S., 2006, Diversity and distribution of photosynthetic bacteria in the Black Sea: Deep-Sea Research Part II, v. 53, p. 1934-1944.

71 Konovalov, S.K., Fuchsman, C.A., Belokopitov, V., and Murray, J.W., 2008, Modeling the distribution of nitrogen species and isotopes in the water column of the Black Sea: Marine Chemistry, v. 111, p. 106-124. Kuypers, M.M.M., Lavik, G., Woebken, D., Schmid, M., Fuchs, B.M., Amann, R., Jorgensen, B.B., and Jetten, M.S.M., 2005, Massive nitrogen loss from the Benguela upwelling system through anaerobic ammonium oxidation: Proceedings of the National Academy of Sciences, v. 102, p. 6478-6483. Kuypers, M.M.M., Sliekers, A.O., Lavik, G., Schmid, M., Jorgensen, B.B., Kuenen, J.G., Sinninghe Damste, J.S., Strous, M., and Jetten, M.S.M., 2003, Anaerobic ammonium oxidation by anammox bacteria in the Black Sea: Nature, v. 422, p. 608-611. Kuypers, M.M.M., van Breugel, Y., Schouten, S., Erba, E., and Damste, J.S.S., 2004, N2- fixing cyanobacteria supplied nutrient N for Cretaceous oceanic anoxic events: Geology, v. 32, p. 853-856. Manheim, F.E., and Chan, K.M., 1974, Interstitial waters of Black Sea sediments: New data and review, in Degens, E.T., and Ross, D.A., eds., Black Sea--Geology, Chemistry, and Biology: Tulsa, The Ammerican Association of Petroloeum Geologists, p. 155-180. McCarthy, J.J., Yilmaz, A., Coban-Yildiz, Y., and Nevins, J.L., 2007, Nitrogen Cycling in the offshore waters of the Black Sea: Estuarine, Coastal and Shelf Sciences, v. 74, p. 493-514. Meyer, K.M., and Kump, L.R., 2008, Oceanic euxinia in Earth history: Causes and consequences: Annual Review of Earth and Planetary Sciences, v. 36, p. 251-288. Mohlin, M., and Wulff, A., 2009, Interaction effects of ambient UV radiation and nutrient limitation on the toxic cyanobacterium Nodularia spumigena: Microbial Ecology, v. 57, p. 675-686. Muller, G., and Stoffers, P., 1974, Mineralogy and petrology of Black Sea basin sediments, in Degens, E.T., and Ross, D.A., eds., Black Sea--Geology, Chemistry, and Biology: Tulsa, The Americal Association of Petroleum Geologists, p. 200- 248. Murray, J.W., Codispoti, L.A., and Friederich, G.E., 1995, Oxidation-reduction environments: the suboxic zone in the Black Sea, in Huang, C.P., O'Melia, C.R., and Morgan, J.J., eds., Aquatic Chemistry: Interfacial and Interspecies Processes, Volume 224: ACS Advances in Chemistry Series, American Chemistry Society, p. 157-176. Murray, J.W., Top, Z., and Ozsoy, E., 1991, Hydrographic properties and ventilation of the Black Sea: Deep-Sea Research Part A, v. 38, supplement 2, p. S663-S689. Oguz, T., 2002, Role of physical processes controlling oxycline and suboxic layer structures in the Black Sea: Global Biogeochemical Cycles, v. 16. Oguz, T., and Besiktepe, S., 1999, Observations on the Rim Current structure, CIW formation and transport in the western Black Sea: Deep-Sea Research Part I, v. 46, p. 1733-1753. Oguz, T., Murray, J.W., and Callahan, A.E., 2001, Modeling redox cycling across the suboxic-anoxic interface zone in the Black Sea: Deep-Sea Research Part I, v. 48, p. 761-787.

72 Ohkouchi, N., Kashiyama, Y., Kuroda, J., Ogawa, N.O., and Kitazato, H., 2006, The importance of diazotrophic cyanobacteria as primary producers during Cretaceous Oceanic Anoxic Event 2: Biogeosciences, v. 3, p. 467-478. Ohkouchi, N., Nakajima, Y., Okada, H., Ogawa, N.O., Suga, H., Oguri, K., and Kitazato, H., 2005, Biogeochemical processes on the saline meromictic Lake Kaiike, Japan: implications from molecular isotopic evidences of photosynthetic pigments: Environmental Microbiology, v. 7, p. 1009-1016. Polissar, P.J., Fulton, J.M., Junium, C.K., Turich, C.C., and Freeman, K.H., 2009, Measurement of 13C and 15N isotopic composition on nanomolar quantities of C and N: Analytical Chemistry, v. 81, p. 755-763. Rao, G.H., Arthur, M.A., and Dean, W.E., 1987, 15N/14N variations in Cretaceous Atlantic sedimentary sequences: implication for past changes in marine nitrogen biochemistry: Earth and Planetary Science Letters, v. 82, p. 269-279. Redfield, A.C., 1958, The biological control of chemical factors on the environment: American Scientist, v. 46, p. 205-221. Ross, D.A., and Degens, E.T., 1974, Recent sediments of Black Sea, in Degens, E.T., and Ross, D.A., eds., The Black Sea--Geology, Chemistry, and Biology: Tulsa, The American Association of Petroleum Geologists, p. 183-189. Sachs, J.P., and Repeta, D.J., 1999, Oligotrophy and nitrogen fixation during eastern Mediterranean sapropel events: Science, v. 286, p. 2485-2488. —, 2000, The purification of chlorins from marine particles and sediments for nitrogen and carbon isotopic analysis: Organic Geochemistry, v. 31, p. 317-329. Sachs, J.P., Repeta, D.J., and Goericke, R., 1999, Nitrogen and carbon isotopic ratios of chlorophyll from marine phytoplankton: Geochimica et Cosmochimica Acta, v. 63, p. 1431-1441. Shimkus, K., and Trimonis, E.S., 1974, Modern sedimentation in Black Sea, in Degens, E.T., and Ross, D.A., eds., Black Sea--Geology, Chemistry, and Biology: Tulsa, The American Association of Petroleum Geologists, p. 249-278. Sinninghe Damste, J.S., Wakeham, S.G., Kohnen, M.E.L., Hayes, J.M., and de Leeuw, J.W., 1993, A 6,000-year sedimentary molecular record of chemocline excursions in the Black Sea: Nature, v. 362, p. 827-829. Squier, A.H., Airs, R.L., Hodgson, D.A., and Keely, B.J., 2004, Atmospheric pressure chemical ionisation liquid chromatography/mass spectrometry of the ultraviolet sreening pigment scytonemin: characteristic fragmentations: Rapid Communications in Mass Spectrometry, v. 18, p. 2934-2938. Stoica, E., and Herndl, G.J., 2007, Bacterioplankton community composition in nearshore waters of the NW Black Sea during consecutive diatom and coccolithophorid blooms: Aquatic Sciences, v. 69, p. 413-418. Tyrrell, T., 1999, The relative influences of nitrogen and phosphorus on oceanic primary production: Nature, v. 400, p. 525-531. van der Meer, M.T.J., Sangiorgi, F., Baas, M., Brinkhuis, H., Sinninghe Damste, J.S., and Schouten, S., 2008, Molecular isotopic and dinoflagellate evidence for Late Holocene freshening of the Black Sea: Earth and Planetary Science Letters, v. 267, p. 426-434.

73 Velinsky, D.J., and Fogel, M.L., 1999, Cycling of dissolved and particulate nitrogen and carbon in the Framvaren Fjord, Norway: stable isotopic variations: Marine Chemistry, v. 67, p. 161-180. Velinsky, D.J., Fogel, M.L., Todd, J.F., and Tebo, B.M., 1991, Isotopic fractionation of dissolved ammonium at the oxygen-hydrogen sulfide interface in anoxic waters: Geophysical Research Letters, v. 18, p. 649-652. Ward, B.B., and Kilpatrick, K.A., 1991, Nitrogen transformation in the oxic layer of permanent anoxic basins: the Black Sea and the Cariaco Trench, in Izdar, E., and Murray, J.W., eds., Black Sea Oceanography: Dordrecht, Kluwer Academic Publishers, p. 111-124. Wilkin, R.T., and Arthur, M.A., 2001, Variations in pyrite texture, sulfur isotope composition, and iron systematics in the Black Sea: Evidence for Late Pleistocene to Holocene excursions of the O2-H2S redox transition: Geochimica et Cosmochimica Acta, v. 65, p. 1399-1416. Young, J.L., and Aldag, R.W., 1982, Inorganic forms of nitrogen in soil, in Stevenson, F.J., ed., Nitrogen in Agricultural Soils: Madison, Soil Science Society of America, p. 43-66. Zehr, J.P., Church, M.J., and Moisander, P.H., 2006, Diversity, distribution and biogeochemical significance of nitrogen-fixing microorgansims in anoxic and suboxic ocean environments, in Neretin, L.N., ed., Past and Present Water Column Anoxia: Dordrecht, Springer. Zehr, J.P., Waterbury, J.B., Turner, P.J., Montoya, J.P., Omoregie, E., Steward, G.F., Hansen, A., and Karl, D.M., 2001, Unicellular cyanobacteria fix N2 in the subtropical North Pacific Ocean: Nature, v. 412, p. 635-638.

74

Figure 3-1. Map of core locations. The following pairs of cores were taken from nearly the same locations: GGC 09 and BC10; GGC71 and BC 78; and GGC 48 and BC 47.

75

Figure 3-2. Weight percent Corg profiles for all gravity cores in this study. All Corg content measurements were on decarbonated samples and are plotted on the x-axes. The depth scales are the same for all cores and start at 0 cm at the top of the scales. The upper and lower gray bands indicate correlated Units IIa and IIb2, respectively.

76

13 13 Figure 3-3.  Corg values for all gravity cores in this study. All  Corg values are on the x-axis in permil (‰) units relative to standard Vienna Pee Dee Belemnite (VPDB). The depth scales are the same for all cores and start at 0 cm at the top of the scales. The upper and lower gray bands indicate correlated Units IIa and IIb2, respectively.

77

Figure 3-4. Compiled box core geochemical profiles. BC 55, plotted on the left, is from a deep basin site and contains sediments including all of Unit I and the top of Unit IIa. All other cores only retrieved sediments from the top of Unit I, above the transition sapropel (T. Sap.). BC 55 data are also plotted in figure 2-11 along with gravity core data to provide a continuous profile since 7.8 ka.

78

15 15 Figure 3-5.  Ntot values for all gravity cores in this study. All  Ntot values are on the x-axis in permil (‰) units relative to standard atmospheric N2. The depth scales are the same for all cores and start at 0 cm at the top of the scales. The upper and lower gray bands indicate correlated Units IIa and IIb2, respectively.

79

Figure 3-6. Corg:Ntot ratios for all gravity cores in this study. The depth scales are the same for all cores and start at 0 cm at the top of the scales. The upper and lower gray bands indicate correlated Units IIa and IIb2, respectively.

80

Figure 3-7. Sedimentary pigment extract chromatogram. Numbered peaks are identified in Table 2-1. This chromatogram is for a sample at the top of Unit IIb1 that contains both scytonemin and Bchl e.

81

Figure 3-8. GGC 71 Corg vs. Ntot compared with Corg vs. Norg. This plot includes samples from Units I, II, and III. The positive intercept on the nitrogen axis of the Corg vs. Ntot linear regression line is equivalent to the average inorganic N content of the core samples. Each data point on the Corg vs. Norg plot is corrected for refractory N content. The regression line passes very close to the origin, indicating that the calculated Norg values are valid.

82

Figure 3-9. GGC71 C and N bulk, refractory, and organic fractions. Refractory C makes up a minor component of Corg, whereas refractory N comprises a more significant fraction of Ntot. Corg:Ntot ratios have a trend toward lower values from Unit II to Unit I. Corg:Norg 15 ratios in this core do not have a significant down-core trend. Calculated  Norg values 15 can be significantly different than  Ntot values, especially in Unit IIa where there is a 15 15 large difference between  Norg and  Nref, and Norg content is relatively low.

83

Figure 3-10. Compound-specific Pphe a C and N stable isotopes. The difference 15 15 15 between  Ntot and  NPphe a is consistent down-core, with  Ntot-Pphe a values averaging 5.1‰. This value is similar to the average value for algae (Sachs et al., 1999). Unit II samples are from GGC 59 and Unit I samples are from BC 55.

84

13 15 Figure 3-11. Compiled  Corg,  Ntot, and %Corg plotted against sediment age. This plot 13 illustrates that geochemical signals are vary synchronously basin-wide. Note that  Corg values cover a broad range in Unit III, reflecting local organic matter sources before the basin-wide signal is established. 

85

Figure 3-12. Water-column profiles for the upper 200 m of the Black Sea from the western gyre. The data are from May 26, 2001 (Station 6 of the R/V Knorr 2001 cruise) and are available online at www.ocean.washington.edu/cruises/Knorr2001. Water- column layers are: surface mixed layer (SML), cold intermediate layer (CIL), suboxic layer (SML), and anoxic layer (AOL). The nutrient peak at the CIL/SOL boundary results from remineralization of sinking organic matter (Codispoti et al., 1991). Ammonium in the AOL is derived from anaerobic OM decomposition in the water column and sediments. Phosphate accumulates in the AOL because it is released from sinking oxide and hydroxide minerals and organic matter. Mode-1 results from upward mixing of nutrients from the CIL and SOL. The mode-1 mixing depth depicted above is derived from observations by Gregg and Yakushev (2005), though mixing to the base of the CIL at times during mode-1 is likely. Mode-1 is proposed for wetter and/or cooler climates that result in stronger salinity stratification. Mode-2 mixes nutrients from the AOL into the surface waters. During mode-2, the nutrient peak at the CIL/SOL boundary is likely to be diminished or absent, as those waters would regularly mix into the SML due to a weaker salinity gradient. As well, the depth to the chemocline at the top of the AOL during mode-2 would be greater than in the modern water column. Mode-2 is proposed for drier and/or warmer climates that result in weaker salinity stratification.

86 Table 3-1. Sedimentary pigment identification. Pigment masses are identified as protonated positive ions detected by APCI-MS in positive ion mode. This table includes the first fragmentation step in MSn. The mass loss typically reflects the removal of the esterifying alcohol. Sterols are identified based on the number of carbon atoms per molecule, with the number of double bonds noted in parentheses.

87

15 15 Table 3-2.  Nbiomass and  NChl a values of cyanobacteria.

88

Table 3-3. Summary of properties related to mode-1 and mode-2

Mode-1 Mode-2 Sedimentary Units I, IIb1 IIa, IIb2

Climate cold, wet warm, dry Freshwater influx high low Salinity gradient strong weak Chemocline depth shallow deep Mixing depth shallow (CIL) deep (AOL)

Upwelled nutrient N:P 5:1 1:1 15 - + - +  N of deep NO3 or NH4 8‰ (NO3 ) 7‰ (NH4 ) Modeled N2 fixation proportion 70% 94% 15N phytoplankton using deep nutrientsa +1.1‰ -1.4‰ 15N total phytoplankton assemblageb 1.4 to 4.4‰ 0.3 to 2.1‰ a calculated as described in text b 15 15 calculated from  NPphe a, assuming  Nbiomass-Pphe a = 5‰

Chapter 4

Chemocline-Induced C and N Isotope Excursions Preserved in the Sediments of a Meromictic Lake

Abstract

Fayetteville Green Lake (FGL) sediments record relatively large temporal shifts

15 13 in bulk  N and  Corg values. Water-column nutrient and particulate N and C stable isotope compositions reveal that the phototrophic communities in the mixolimnion and chemocline produce isotopically distinct biomass, with lower 13C and 15N values in the chemocline than in the overlying mixolimnion. C and N isotope ratios of sediment gravity core samples suggest that mixolimnion biomass dominates the modern sediments.

The inclusion of a greater proportion of chemocline biomass could generate negative

15 13 isotope excursions in  N and  Corg values. Prior to ~1780 CE, biomass from purple sulfur bacteria was a larger component of bulk sedimentary organic matter and thus bulk

15 13  N and  Corganic values are lower in those sediments. A similar, more recent shift

15 13 toward lower  N and  Corganic values is observed in shallow water sediments of FGL, in response to a brief shoaling of the chemocline. The mechanisms outlined for generating isotope excursions are potentially applicable to Mesoproterozoic samples where biomarkers for purple sulfur bacteria have been identified.

90 4.1 Introduction

Marine and lacustrine sediments can have a wide range of bulk 15N values, affected by variability in phytoplankton source organisms, nutrient sources and concentrations, terrigenous organic matter content, and diagenetic alteration. In most modern ocean surface waters phytoplankton assimilate nearly all of the dissolved

15 inorganic nitrogen (DIN). In these settings sedimentary  Ntot values reflect the substrate N isotope composition, usually 5-8‰ (Altabet and Francois, 1994). In contrast,

15 organic matter (OM)-rich sediments and sedimentary rocks commonly have  Ntot values near and below 0‰ that may be explained by several different mechanisms.

Where substrate consumption is incomplete, phytoplankton 15N values are lower than substrate 15N values (Hoch et al., 1994; Montoya and McCarthy, 1995; Needoba et al.,

2003), resulting in lower sediment 15N values (Altabet et al., 1999). Phosphate excess relative to DIN, defined as N:P nutrient molar ratios lower than 16:1 (Redfield, 1958), favors growth of nitrogen fixing organisms (Tyrell, 1999), yielding biomass and

15 sediments with  Ntot values near -2‰ (Junium and Arthur, 2007; Kuypers et al., 2004;

Rau et al., 1987). Further, an examination of water column particulate matter and

15 sedimentary  Ntot values in the Mediterranean Sea suggests water column anoxia

15 preserves the N-depleted N2 fixation signal, which is otherwise lost during aerobic respiration (Sachs and Repeta, 1999). We present evidence for an additional mechanism for forming 15N-depleted sediments, resulting from the incorporation of significant proportions of biomass from sulfide-oxidizing phototrophic organisms in euxinic waters.

91 In deep meromictic lakes, such as Fayetteville Green Lake (FGL), biomass produced in the mixolimnion typically dominates autochthonous organic matter deposited in the sediments, even though primary productivity may be higher at the chemocline

(Culver and Brunskill, 1969; Fry, 1986; Thompson et al., 1990). This situation is explained by the slow sinking rates of unicellular chemocline-dwelling bacteria compared with fecal pellets and larger algae with mineral components. Slow sinking rates expose senescent bacterial cells to degradation in deep waters (Logan et al., 1995). Conversely, the inclusion of mixolimnion phytoplankton in rapidly sinking zooplankton fecal pellets preferentially preserves their biomass. Euxinia at the chemocline limits zooplankton grazing on purple sulfur bacteria (PSB) and green sulfur bacteria (GSB), reducing their incorporation into fecal pellets. Thus, algal and cyanobacterial biomass is preferentially incorporated into sediments while slowly sinking PSB and GSB biomass is more likely to degrade before reaching the basin floor.

A small proportion of chemocline-derived biomass reaches deep basin sediments via slow settling. PSB biomass may sink more rapidly than GSB biomass, allowing preferential sedimentation of PSB biomass in basins where planktonic PSB and GSB inhabit the chemocline. In particular, PSB with intracellular sulfur granules are expected to sink more rapidly than GSB (Fry, 1986; Mas et al., 1990). Daphnia with purple guts from consuming PSB have been observed in FGL and are expected to produce fecal pellets containing PSB biomass, which could increase the PSB:GSB ratio in sediments.

Pigments from both PSB and GSB are found in the sediments of recent and ancient anoxic marine settings, however, so at least a component of their biomass resists

92 complete degradation in anoxic waters (Brocks and Schaeffer, 2008; Koopmans et al.,

1996; Repeta, 1993).

Changes in the balance between chemocline and mixolimnion productivity may lead to large shifts in the carbon and nitrogen stable isotope compositions of bulk sedimentary organic matter. The C and N stable isotope compositions of phytoplankton are controlled primarily by variability in the concentrations and isotopic compositions of dissolved inorganic carbon (DIC) and fixed-N substrates, and the isotopic fractionation associated with their assimilation (House et al., 2003; Popp et al., 1998; Quandt et al.,

1977; Sirevag et al., 1977; Wada and Hattori, 1978; Waser et al., 1998). Growth rate, substrate concentration, and the fraction of substrate consumed have a primary effect on isotope fractionation. By making assumptions about substrate concentrations and isotopic compositions, sedimentary 13C and 15N values may be used to determine assimilation pathways and nutrient sources (Freeman et al., 1994; Robinson, 2001). As

13 meromictic lakes generally have strong gradients in DIC concentration [DIC],  CDIC,

- + 15 15 nitrate concentration [NO3 ], ammonium concentration [NH4 ],  NNO3-, and  NNH4+, biomass produced in the mixolimnion and at the chemocline have distinct 13C and 15N values.

Fayetteville Green Lake (FGL) contains photosynthetic bacterial populations in its shallow chemocline (Fry, 1986), and mixolimnion waters dominated by non- diazotrophic cyanobacteria (Thompson et al., 1997). We examine the C and N stable isotopic compositions of biomass produced by mixolimnion and chemocline-dwelling organisms in FGL. In particular, we focus on the 15N-depleted 15N values produced by

93 PSB and, to a lesser extent, GSB as a possible source for relatively low 15N values in the sediments. We compare our sediment stable isotope profiles with records of diagnostic of GSB and PSB photosynthetic pigments (Chapter 5). We demonstrate that relative

15 13 increases in chemocline biomass directly correlate with low sediment  Ntot and  Corg values. The mechanism we outline for producing 15N-depleted sediments may apply to ancient anoxic settings proposed to have had very shallow chemoclines, such as parts of the Permo-Triassic ocean (Kump et al., 2005) and the mid-Proterozoic ―

(Brocks and Schaeffer, 2008; Canfield, 1998).

4.1.1 Study Site

Fayetteville Green Lake is a small (0.3 km2, 52 m deep), salinity-stratified meromictic lake located near Syracuse, NY. It consists of a circular main basin and a northward trending neck containing the surface water outflow stream (Fig. 4-1). It most likely formed as a glacial plunge pool, leading to its steep banks cut into the Syracuse

Formation and Vernon Shale (Miner, 1933; Thompson et al., 1990). FGL receives groundwater input focused at several depths, with more saline groundwater entering at deeper depths, leading to density stratification (Brunskill and Ludlam, 1969; Takahashi et al., 1968; Torgersen et al., 1981). In the modern lake, the chemocline separating oxidizing and reducing conditions is at the base of a 4-m-thick that spans 16-

20 m water depth (Fig. 4-2). Previous studies have placed the chemocline as shallow as

15-17 m (Torgersen et al., 1981). The inhibition of downward mixing of oxygenated surface waters allows the deep waters to remain permanently anoxic.

94 Productivity in the surface mixolimnion waters is dominated by Synechococcus, which stimulate extracellular calcite precipitation resulting in ―whitings‖ (Thompson,

1997). Sulfate-rich groundwater entering the lake supports abundant bacterial sulfate reduction in the monimolimnion and the accumulation of hydrogen sulfide. Since the chemocline is relatively shallow, and the mixolimnion waters are clear (Secchi depths vary seasonally between 5-18 m), FGL is characterized by photic zone euxinia (Culver and Brunskill, 1969). The redox shift at the chemocline is accompanied by a turbidity maximum that indicates the presence of a bacterial plate. Previous studies have identified populations of PSB and GSB forming the plate at the upper boundary of the sulfidic monimolimnion (Fry, 1986). Indeed, water pumped from the chemocline has a distinct pink color, the result of abundant PSB. Annual primary productivity in the bacterial plate has been estimated at 239 g C·m-2·yr-1, compared with 51 g C·m-2·yr-1 in the mixolimnion (Culver and Brunskill, 1969). The mixolimnion productivity estimate may, however, be an underestimate as a subsequent study suggests they did not account for

Synechococcus productivity (Thompson et al., 1990).

4.2 Methods

4.2.1 Sampling

Water column samples were collected on November 4, 2005; May 9, 2006; and

August 29, 2006. Samples for nutrient analysis were pumped to the surface by peristaltic pump and filtered through an in-line 0.45-m groundwater filter cartridge into 1-liter

95 polypropylene bottles. We attached the tubing to a YSI data sonde to measure in situ specific conductance, redox potential, and turbidity while pumping, to establish depths of the pycnocline, chemocline, and the particulate maxima in the mixolimnion and the bacterial plate. Seston samples were collected onto pre-combusted Whatmann GF/F filters (14.2 cm) using an in-line stainless steel filter housing and a diaphragm pump

(5 l·min-1). Filter samples were kept cold on ice in the field and frozen at -20ºC in the lab prior to analysis.

Short gravity cores (< 0.5 m) were collected on June 29, 2004 and August 29,

2006 using a K-B Corer (Wildlife Supply Co., Buffalo, NY). Cores used in this study were collected from the deep central basin of the lake (53 m depth; FGL62904-2,

FGL82906-4) and from the shallow neck on the north end of the lake (17 m depth;

FGL82906-8). FGL62904-2 was transported upright in an ice chest to our laboratory at

Penn State. FGL82906-4 and FGL82906-8 were extruded from the core liners on site at

Green Lake, and divided into samples for stable isotope analysis and pigment analysis

(Chapter 5). FGL82906-4 was compared with our previous stratigraphic description from

FGL62904-2 to determine how it should be divided; FGL82906-8 was divided based on observed changes in color and texture. Subsamples were placed in 50 ml centrifuge tubes and stored on ice until they were returned to the lab.

Surface sediment samples were collected using a Ponar type grab sampler on

October 17, 2006. We report data from one grab sample, FGL101706-53m, which retrieved 10 cm of material from the center of the deep basin, including the sediment- water interface. It was divided into 10 subsamples, each 1 cm thick. All subsamples were

96 placed in 50 ml centrifuge tubes and stored in an ice chest until they were returned to the laboratory.

Benthic macrophyte algae, specifically Chara sp. which dominate the benthos over large areas of the lake bed above the chemocline, were collected by Penn State

Science Divers in October, 2006. Aquatic moss was retrieved from 17 m water depth by the Ponar sampler described above. Leaf litter samples from 6 species of near-shore trees were also collected in October, 2006. Leaves were collected from the ground surrounding multiple individual trees of each species and combined for C and N stable isotope analysis. Algae, moss, and leaf samples were freeze-dried and ground to a fine powder in a ball-bearing mixer mill.

4.2.2 Instrumentation

All stable isotope analyses were performed on a coupled elemental analyzer- isotope ratio mass spectrometer (EA-IRMS). The continuous flow system consists of a

Costech ECS 4010 elemental analyzer, ThermoFinnigan Conflo III interface, and a

ThermoFinnigan Deltaplus XP mass spectrometer. Sample carbon and nitrogen are

13 15 converted quantitatively to CO2 and N2 in the EA, and their  C and  N values are calculated by comparison to reference gases using Isodat software (v1.42). The reference gas isotope ratios are calibrated using international standards, and in-house standard materials are analyzed during sample runs to ensure consistency. Samples analyzed in this way have been shown to be accurate within ± 0.2‰. Replicate analyses of standard and sample material in our lab usually have better than 0.1‰ precision. All 13C values

97 are reported relative to Vienna Pee Dee Belemnite (VPDB) and 15N values relative to atmospheric N2 (AIR).

4.2.3 Sample Analysis

+ - Ammonium and nitrate concentrations, [NH4 ] and [NO3 ], were determined spectrophotometrically using Hach reagents and linear response factors (R2 = 0.99 for

+ - + both NH4 and NO3 ) determined for standard solutions. Calculated [NH4 ] for ammonium sulfate standard solutions have a standard deviation of 1.9 M. Calculated

- [NO3 ] for potassium nitrate standard solutions have a standard deviation of 1.6 M.

Nitrogen stable isotope values for nitrate and ammonium were determined using ammonium diffusion methods (Holmes et al., 1998; Sigman et al., 1997). Standard analyses demonstrated that 15N values have precision of better than ± 0.3‰ for nitrate and ± 0.2‰ for ammonium.

Seston filter samples were thawed to room temperature and visible zooplankton were picked from the filters using forceps prior to analyses. Depending on particle density, 8 mm, 10 mm, or 12 mm diameter plugs were removed in triplicate from each filter. The plugs were acidified with 10% HCl to dissolve CaCO3 and gently rinsed with deionized water. The plugs were freeze-dried and packed into tin sample capsules in preparation for EA-IRMS carbon and nitrogen analysis. The remainder of each filter sample was stored frozen (-20ºC) in preparation for pigment extraction (Chapter 5).

Unfiltered chemocline water was collected in May 2006 for gravity separation of

GSB and PSB biomass. A total of 8 liters was centrifuged (500 g, 5 min.) in 200 ml

98 conical bottles. This procedure yields cell pellets composed primarily of PSB biomass, the result of dense sulfur granules accumulated in PSB cells (Fry, 1986). The purple cell pellets were collected as the PSB fraction and the brown-tinted supernatant was filtered onto GF/F filters to collect the GSB fraction. The samples were analyzed for C and N stable isotopes as described below. Pigments were extracted and analyzed from both fractions to further examine the purity of the PSB and GSB fractions. The pigment extraction and analysis methods are described in depth in chapter 5.

Sediment core FGL62904-2 was split for sediment description in the lab.

Laminae couplets were counted using 10× magnification and the core was divided into samples for C and N stable isotope analysis based on sediment characteristics. The sediment description from FGL62904-2 was used to determine the subsampling criteria for FGL82906-4, which was divided in the field. Sediment samples were centrifuged and

+ 15 the supernatant decanted for analysis of pore water [NH4 ] and  NNH4+, using the nutrient methods described above. The sediment fractions were wet-sieved into coarse and fine fractions using a 125 m sieve. This treatment removed large particles of terrigenous and macrophyte algal OM. The fine-grained fractions were acidified with

10% HCl for 24 hours to dissolve carbonate minerals. After acidification the samples were rinsed 3 or 4 times by repeated centrifugation and sediment dispersal in deionized water and freeze-dried. The samples were ground to a fine powder using a ball-bearing mixer mill and packed into tin capsules in preparation for EA-IRMS C and N analysis.

To assess the detrital silicate mineral content of sediment samples, we treated carbonate-free subsamples with 10% hydrogen peroxide to remove organic matter. The reaction was carried out in 50 ml centrifuge tubes, initially at room temperature for ~4

99 hours. After the reaction had slowed, the tubes were transferred to a 50ºC oven overnight, until gas production stopped. The samples were centrifuged and the supernatant decanted and discarded. An additional treatment with 10% H2O2 at 40ºC ensured that the reaction had gone to completion. The remaining insoluble residue consists predominantly of silicate minerals incorporated into the sediments from weathering of the surrounding bedrock with a minor contribution from diatom frustules.

4.2.4 Accumulation Rate Calculations

Sediment mass accumulation rates (MAR) were calculated for laminated intervals using a varve-based chronology (Ludlam, 1969). MAR calculations were performed on

0.5 to 2-cm-thick sediment intervals that consisted entirely of laminations. Bulk MARs were calculated using freeze-dried sample weights, core cross section area, and the duration of sedimentation for each interval. Carbonate MAR was calculated by mass loss of bulk samples after acidification, rinsing, and freeze-drying. Organic carbon MAR was calculated using percent organic carbon values calculated from EA analysis. Averages of

MARs for Units 1 and 2 were calculated using duration-weighted values for all intervals analyzed.

100 4.3 Results

4.3.1 Water-Column and Pore-Water DIN

Water Column DIN profiles in FGL are similar to those of other redox-stratified basins (Codispoti et al., 1991; Velinsky and Fogel, 1999). In the oxygenated

- mixolimnion, NO3 is the most abundant form of DIN with concentrations up to 85 μM,

+ 15 whereas [NH4 ] is highest in the monimolimnion (Fig. 4-3).  NNO3- values in the main basin mixolimnion are remarkably consistent, with all measured values between 8-9‰.

+ We measured the highest water-column [NH4 ] of 493 μM at 47 m depth, near the base of

15 the monimolimnion. In the main basin,  NNH4+ values increase from the sediment-water

15 interface to the pycnocline. At 47 m depth, the  NNH4+ value was 1.7‰. The highest

15 +  NNH4+ value was 23.7‰, in the pycnocline at 18 m depth where [NH4 ] was 18 M.

In the deep basin, surface sediment pore water is 15N-enriched relative to the

15 overlying 47 m water-column sample. We measured  NNH4+ values between 3.7‰ and

+ 15 7.4‰ in this core with no clear trend with depth, [NH4 ], sediment  N value, or

+ sediment type (Fig. 4-4). Pore-water [NH4 ] from the 17-m core, on the other hand,

15 displays a diffusion profile to a sink at the sediment surface (Fig. 4-5.  NNH4+ values increase from values as low as 7.2‰ near the base of the core to 13.8‰ at the sediment- water interface. Since sediments at 17 m water depth are overlain by oxygenated waters,

+ nitrification and/or anammox probably consume NH4 at the sediment-water interface.

101 4.3.2 Organic Matter Sources

13 15 We measured  Corg and  N values of seston, benthic algae, aquatic moss, and leaf litter to evaluate the possible contributions of each to sedimentary organic matter

(Fig. 4-6). Seston isotope ratios cover a broad range, with 15N values between 8.4‰

13 and -3.6‰ and  Corg values between -26.0‰ and -42.4‰ (Fig. 4-3). There are

15 13 15 prominent trends with water depth in both  N and Corg values. The highest  Nseston

15 values are found in the mixolimnion, where  Nseston values range between 2.7‰ and

8.4‰, with the lowest value measured at 2 m in May 2006 and the highest at 10.8 m in

15 August 2006.  Nseston values decrease with depth from the base of the mixolimnion to the chemocline, where we measured a minimum value of -3.6‰ in May 2006. One meter

15 below the chemocline,  Nseston values increase and remain relatively constant throughout the monimolimnion at -1.4±0.7‰.

13 15 13 The  Corg water-column profile resembles that of  N. The highest  Cseston values are found in the mixolimnion, and the values decrease with depth to the

13 chemocline. The May 2006 depth profile contained the highest  Cseston value, -26.0‰,

13 at 2 m. At 16 m depth in May, the  Cseston value was -31.4‰, 5.4‰ lower than the 2-m sample. Mixolimnion samples were relatively 13C-depleted at 5 m and 10 m sample

13 depths in June and August. The lowest  Cseston values are found at the chemocline, where the average of four sampling dates is -41.6±1.1‰. Below the chemocline,

13  Cseston values increase relative to the chemocline values, averaging -38.8±0.9‰ in the

13 monimolimnion. The low  Cseston values at the chemocline produce a signal that should

102 be easily recognized if biomass produced at the chemocline contributes significantly to the sediments (Freeman et al., 1990).

Gravity separation of purple and brown-colored fractions from the chemocline reveals that PSB and GSB had different 15N and 13C values when growing in the same waters in May 2006 (Table 4-1). The purple fraction yielded a 15N value of -5.8‰, and the brown fraction had a value of -1.9‰. We also observed a difference between 13C values of the purple and brown fractions, -43.2‰ and -40.8‰, respectively. Fry (1986), using the same methods in June 1984, determined that the purple fraction had a 13C value of -41.1‰, and the brown fraction, -32.6‰, a difference of 8.5‰. His observation is explained by the distinction between isotopic fractionation during carbon assimilation by PSB and GSB (House et al., 2003). GSB use the reverse tricarboxilic acid pathway, which has a small carbon isotope discrimination resulting in biomass that is 13C-depleted by ~10‰ relative to DIC (Quandt et al., 1977). PSB assimilate carbon using the Calvin cycle, the more common pathway used by most plants, algae, and cyanobacteria. Carbon assimilation by PSB causes a larger fractionation, producing biomass that is ~30‰ 13C- depleted relative to DIC (Sirevag et al., 1977). Our isolated brown fraction probably contains PSB, cyanobacteria, and sinking biomass from the mixolimnion in addition to

GSB. All of the contaminants are expected to be 13C-depleted relative to GSB (Freeman et al., 1994).

Our analysis of pigments from the two fractions shows that the purple fraction contains phototrophic biomass almost exclusively from PSB (Table 4-1). The brown fraction contains primarily GSB pigments, but also contains chlorophyll a (Chl a),

103 pheophytin a (Phe a), bacteriopheophytin a (Bphe a) and okenone. The presence of okenone in this fraction confirms the inclusion of PSB biomass. Further, concentrations of Chl a relative to Bchl e are higher than expected for GSB production (Overmann and

Garcia-Pichel, 2006). Chl a and Phe a in this sample most likely come from cyanobacteria living at the chemocline (Meyer et al., submitted). Thus, the inclusion of cyanobacterial and PSB pigments confirms that biomass from these organisms

13 contaminates the GSB fraction, explaining the lower than expected  Corg values.

Benthic macrophyte algae and moss cover much of the lake floor above the chemocline. Macroscopic OM from these sources is observed in sediments of shallow and deep cores. Chara samples collected from four depths in the lake neck have 15N

13 values ranging from 3.5‰ to 7.3‰ and  Corg values between -25.7‰ and -40.6‰ (Fig.

4-3). These values are similar to the range of mixolimnion seston values. Aquatic moss collected in the lake neck at 17 m has C and N isotope values similar to the 18 m Chara sample and the seston sample collected from 10.8 m in August 2006. Thus, benthic and

13 planktonic phototrophs in the mixolimnion have lower  Corg values at deeper depths,

13 presumably reflecting lower  CDIC values with deeper depths (Freeman et al., 1994; Fry et al., 1991).

The slopes surrounding FGL are wooded, so leaf litter and soil organic matter may contribute to sedimentary OM. We observed leaf fragments as a component of the coarse sediment fractions removed from our C and N isotope analyses. We collected leaf litter from 6 common tree species along the shoreline of FGL for C and N isotope

104 analysis. This survey yielded 13C values ranging between -27.3‰ and -31.5‰, and

15N values between -1.4‰ and 3.2‰ (Table 4-2, Fig. 4-6).

4.3.3 Sediment Description

Sediments in the main basin are deposited either on the basin slopes or the basin floor, each of which encompass about half of the main basin surface area (Fig. 4-1)

(Brunskill and Ludlam, 1969). The steep basin slopes range from 0 to 45 m depth and generally do not accumulate thick layers of sediment. The basin floor is relatively flat with depths ranging from 45 to 53 m. Sediments on the basin floor are derived in nearly equal proportions from particles settling through the water column and gravity flow redeposition of material from the basin slopes (Ludlam, 1974). The lake neck has a shallower depth gradient than the main basin slopes, allowing stable sedimentary deposits at depths shallower than 30 m. In this study, we present organic geochemical data for cores from two depths, a 39-cm composite core from the center of the main basin at 53 m water depth and a 36-cm gravity core collected above the chemocline at 17 m water depth in the lake neck. The deep basin composite core is derived from two short gravity cores,

FGL 62904-2 and FGL 82906-4, and the upper 3 cm of grab sample FGL 101706-53m.

The deep basin sediments are laminated calcareous marl, composed of alternating light colored carbonate-rich and dark OM-rich laminae (Fig. 4-7). There is a distinct color change in the varved sediments at 25 cm depth, which defines the boundary between the upper Unit 1 (gray) and lower Unit 2 (red-brown). The gray color of Unit 1 has been ascribed to increased clay accumulation as well as increased calcite precipitation

105 in the water column and its accumulation in the sediments (Hilfinger et al., 2001). We observe this color change in all cores taken from the deep basin. Massive-bedded layers of variable thickness (1-20 mm) are interspersed among the laminated sediments in both

Units 1 and 2. These layers range from dark brown in Unit 2 to tan in Unit 1, and contain rock fragments, gastropod shells, quartz and carbonate sand, wood, fibrous organic matter, fine organic matter, and clay. Some massive beds display fining-upward texture, and have been identified as turbidites (Ludlam, 1969). We also find thin (1-3 mm thick) homogeneous gray clay layers interbedded with the laminated sediments. These have been interpreted as storm deposits (Hilfinger et al., 2001) or as gray turbidites (Ludlam,

1969) We isolated the thickest of these gray layers from 18.5-18.8 cm depth for C and N isotope analysis.

We counted 210 varves in Unit 1 and were able to recognize a distinctly reddish varve which has been identified previously as a marker deposited in 1963 (Ludlam, 1969,

1984). By counting from the 1963 varve, we determined that the base of Unit 1 was deposited in ~1787 and the uppermost couplet retrieved by gravity coring was deposited in 1996. The deepest varved layer we collected from Unit 2 corresponds to calendar year

1454. Implicit in our use of a varve chronology is that the massive deposits result from short-duration events that did not disturb the underlying sediments (Hilfinger et al.,

2001). This assumption is somewhat flawed as Ludlam (1974) observed that the sediments underlying turbidites often were eroded, leading to the loss of one or two varves per event. We estimate that erosion by turbidity currents may shift the age of the base of our composite core to ~20 years older, given the presence of 12 turbidites thicker than 5 mm in the composite core. In the deep basin core, Unit 1 consists of 16.4 cm of

106 laminated sediments (linear accumulation rate = 0.71 mm yr-1), 9.0 cm of tan/brown massive deposits, and 1.2 cm of gray homogenous clay. Our age model is consistent with that of Hilfinger et al. (2001) for Unit 1 (Fig. 4-8). The average bulk, carbonate, and organic carbon accumulation rates are 363 g m-2 yr-1, 268 g m-2 yr-1, and 10.6 g m-2 yr-1, respectively, for Unit 1 varved sediments (Fig. 4-9).

Gravity coring retrieved a maximum of 12.5 cm of Unit 2. We were not able to determine the total thickness of this unit as our coring device did not penetrate to its base.

Within this unit we counted 331 couplets of alternating red and brown laminae, accounting for a total thickness of 5.8 cm and a linear accumulation rate of 0.19 mm yr-1.

This accumulation rate is nearly two times greater than that calculated by Hilfinger et al.

(2001) based on one radiocarbon-dated sample (Fig. 4-8). Bulk, carbonate, and organic carbon accumulation rates for the laminated sediments are 139 g m-2 yr-1, 127 g m-2 yr-1, and 4.8 g m-2 yr-1, respectively, for Unit 2. We also found 17 massive brown layers (1-10 mm thick) accounting for 6.7 cm of total thickness. There were no homogeneous gray layers in this unit, and insoluble residues, primarily detrital silicate minerals, account for less than 5% of the Unit 2 sediments.

The shallow neck core, collected from 17 m water depth, is mostly massive gray calcareous marl with macroscopic fibrous organic matter (Fig. 4-10). It is currently overlain by oxygenated waters and subject to bioturbation. The gray clay is derived from the erosion of the local Silurian shale and dolomite bedrock (Syracuse Formation and

Vernon Shale) and the fibrous organic matter from benthic algae, aquatic moss, and terrigenous plant matter. There is one interval of laminated sediments in this core, between 7.5 and 12.5 cm depth. This interval apparently was deposited under anoxic

107 conditions, as bioturbation was inhibited allowing laminae to persist. At the base of the shallow neck core is a 5.5 cm-thick gray-colored fining upward layer that we interpret as a turbidite. We divided the turbidite into two layers for analysis. The upper 2 cm of the turbidite is 95% fine material (<125 m), and is composed of 13% carbonate minerals and 2.6% organic carbon. The lower 3.5 cm of the turbidite is coarser, composed of 74% fine material, and contains 72% carbonate minerals and 4.6% Corg. The upper layer of the turbidite has a sediment composition similar to the gray homogeneous layer isolated from

Unit 1 of the deep basin core (Table 4-3).

4.3.4 Sedimentary 13Corg and 15Ntot Values

The laminated sediments of Units 1 and 2 from the deep basin core have distinct

13 15 15 bulk  Corg and  Ntot signals (Fig. 4-4). Unit 2  Ntot values average of 2.2 ± 0.5‰.

15 15 Unit 1 sediments are N-enriched relative to Unit 2, with an average  Ntot value of 3.5

± 0.5‰. Unit 2 sediments are 13C-depleted relative to Unit 1 sediments, with average

13  Corg values of -32.2 ± 0.8‰ and -29.1 ± 1.0‰, respectively. The observed differences in C and N stable isotope distributions in Unit 1 and 2 laminated sediments suggest a change in organic matter source material or nutrient dynamics over time.

We analyzed the C and N stable isotope compositions of 4 tan turbidites from

Unit 1 and 8 brown turbidites from Unit 2. In contrast to the laminated sediments, the tan/brown turbidites appear to have consistent organic matter sources for both Units 1 and 2. Turbidite 15N values average 3.5 ± 0.4‰, with no significant difference between

13 15 Unit 1 and Unit 2 values. Turbidite  Corg values are more variable than  N values, but

108

13 Unit 1 and Unit 2 average carbon isotope compositions are indistinguishable, with  Corg values of -29.8±1.0‰ and -30.0±1.6‰, respectively. The homogeneous gray layer from

15 13 Unit 1 has a  N value of 5.6‰ and  Corg value of -26.6‰, distinct from the brown/tan turbidites.

13 Shallow neck core sediments display a wide range of  Corg values, from -25.7‰ in the upper fine-grained layer of the turbidite to -32.5‰ in the sample just above the laminated interval (Fig. 4-5). The laminated interval, along with the two non-laminated samples above it, represents an excursion from typical C and N isotope distributions in

13 this core. The excursion interval has an average  Corg value of -31.5‰, compared with

-29.5‰ for the rest of the core. Similarly, 15N values are relatively low in the excursion interval at 4.1‰, compared with 5.9‰ for the rest of the core. Both layers of the turbidite have identical 15N values of 5.7‰, but their 13C values differ. The upper

13 turbidite layer  Corg value is -25.7‰ and the lower layer is -29.6‰. The homogeneous

15 13 gray layer from the main basin core has similar  N and  Corg values as the upper layer of the turbidite in the shallow neck core.

4.4 Discussion

4.4.1 C and N Cycling in FGL

A strong 13C gradient develops annually in the mixolimnion. This results in

13 13 higher  Cseston values in surface waters and lower  Cseston values at depth (Fig. 4-3).

109 The same effect is observed in benthic algae (Fig. 4-6). Thompson et al. (1997) showed

13 that surface water  CDIC values range from -9.5‰ in the winter to summer values up to

13 -6.5‰. Summer and early fall  CDIC values of -7.0‰ and -6.0‰ from older studies confirm the seasonal 13C enrichment of surface water DIC (Deevey et al., 1963;

Takahashi et al., 1968), presumably due to the production of 13C-depleted planktonic biomass and its export to deeper depths prior to remineralization. Winter values are close to those of local well water (-10‰), which have been considered a proxy for groundwater entering the mixolimnion (Thompson et al., 1997). Since cool weather in the fall and spring leads to mixing above the pycnocline, we expect the mixolimnion is homogeneous

13 in early spring with a  CDIC value close to -9.5‰. This value is higher than older

13 published winter surface water  CDIC values of -11.3‰ and -12.0‰ (Takahashi et al.,

1968), possibly reflecting temporal differences in the balance between groundwater inputs and upwelling monimolimnion waters.

In May 2006, surface water seston was 13C-enriched by 5.4‰ relative to seston collected from 13 and 16 m depth (Fig. 4-3). Thompson (1997) showed that

Synechococcus dominates phytoplankton assemblages at all depths in the mixolimnion, so species variability is not likely the cause of the observed trend. Four other

13 mechanisms may explain this observation: (1)  CDIC values are expected to decrease with depth as sinking OM is remineralized, (2) aqueous CO2 concentrations increase with depth allowing for greater isotopic discrimination, (3) slower growth rates with decreased light flux allow for greater isotopic discrimination, (4) mixing of monimolimnion waters

13 into the deep mixolimnion can reduce  CDIC values (Freeman and Hayes, 1992). In

110 FGL, DIC at the chemocline is 13C-depleted relative to DIC in surface waters. Deevey et al. (1963) reported values of -13.7‰ at the deep turbidity maximum (20 m) and -11.2‰

13 at 18 m for June 30, 1962. Published monimolimnion  CDIC values range from -16.0‰ to -21.1‰, with lower values corresponding to deeper samples (Deevey et al., 1963;

Takahashi et al., 1968). We also observed that summer seston samples from 5 m and

10.8 m are 13C-depleted relative to samples from the same depths in May, suggesting 13C depletion of DIC continues through the summer.

Nitrate 15N values are consistently near 8‰ at all depths in the mixolimnion and

- pycnocline (Fig. 4-3). We detected a large decrease in [NO3 ] between November 2005

15 - and August 2006 without significant changes in  N of residual NO3 . If assimilation or

- 15 denitrification were driving the NO3 loss, we would expect to see N enrichment of the

- residual NO3 pool (Altabet and Francois, 1994; Mariotti et al., 1981; Montoya and

- McCarthy, 1995). Consequently, we propose a hydrologic explanation for NO3 loss in

- the mixolimnion. The NO3 pool receives inputs from surface runoff, groundwater,

+ sediment pore waters, and oxidation of NH4 from the monimolimnion, all of which combine to yield the observed high 15N values. Groundwater influx is important to the mixolimnion water budget, causing the short 1 to 2-year residence time (Takahashi et al.,

1968; Torgersen et al., 1981). In the mixolimnion and pycnocline, ground water enters predominantly at two depths, at ~10 m and ~18 m (Thompson et al., 1990). Assuming

- that groundwater contains low [NO3 ], its addition through winter, spring and summer

- 15 would dilute the high [NO3 ] concentrations observed the previous fall. Thus, N-

111 - enriched NO3 would be exported from the lake through the outlet without significant detectable isotope fractionation.

- Elevated NO3 concentrations are typical of late-fall to early-spring surface waters of other lakes of the northeastern United States (Gruendling and Malanchuk, 1974). Such

- [NO3 ] peaks can result from mixing upward of nutrient-rich deep waters or influx of nutrient-rich soil water or runoff associated with plant decay, decreased soil microbial activity, and snow melt (Gruendling and Malanchuk, 1974; Mitchell et al., 1996; Pardo et

- al., 1995). In lakes, increased [NO3 ] is not typically accompanied by significant

3- increases in [PO4 ], so surface water productivity is not likely to consume all of the

- - newly available NO3 . For FGL, the elevated mixolimnion [NO3 ] data for the November

3- 2005 sampling date compare directly with dissolved [PO4 ] date from Zerkle (2006) to yield N:P molar nutrient ratios between 238 to 314. These ratios are significantly greater

- than the 16:1 Redfield ratio, supporting our assertion that NO3 assimilation accounts for

- - a small proportion of NO3 drawdown in the mixolimnion, and dilution with NO3 -poor

- groundwater is more likely the cause of decreased [NO3 ].

+ Monimolimnion NH4 is derived from ammonification in the sediments and the

+ water column. Deep basin sediment pore waters have high [NH4 ] (Fig 4-4), driving

+ diffusion from the sediments to the water column. Pore-water NH4 and surface

15 15 + sediments have similar  N values near 3‰, about 2‰ N-enriched relative to NH4 in the overlying deep monimolimnion. This discrepancy suggests that the OM source for

+ 15 remineralized NH4 in the water column is N-depleted relative to the OM source in the sediments. Ammonification of slowly sinking PSB and GSB biomass (15N values near

15 + -1‰) is likely to generate N-depleted NH4 in the monimolimnion. Water-column and

112

+ + 15 sediment-derived NH4 mix to produce the observed deep monimolimnion NH4  N value of 1.7‰ at 47 m.

+ NH4 concentrations are much lower at the top of the monimolimnion than in

+ deeper waters (Fig. 4-3). NH4 is assimilated in this region by photosynthetic bacteria.

Isotope effects associated with assimilation produce biomass with low 15N values,

+ 15 15 leaving the residual NH4 N-enriched. We measured  NNH4+ values near 20‰ at and above the chemocline, the result of this fractionation. Export of this 15N-enriched fraction to the mixolimnion maintains low 15N values in the monimolimnion while possibly providing an additional 15N-enriched nutrient source to the mixolimnion. We also suspect that the dissimilatory processes of nitrification/denitrification and anaerobic ammonium oxidation (anammox) occur near the chemocline (Schubert et al., 2006).

15 + Taken to completion, these processes would be sinks for residual, N-enriched NH4 above the chemocline.

Seston samples display strong gradients in both 15N and 13C values from surface to deep waters, primarily in response to the observed changes in DIN and DIC stable isotope compositions. Synechococcus growth in the mixolimnion produces biomass with relatively high 15N values. This is the result of the small isotopic fractionation (enrichment factor () = 5.4±0.6‰) associated with nitrate assimilation

(Needoba et al., 2003) and growth on 15N-enriched nitrate. Mixolimnion samples also

13 have high  Cseston values relative to those from the monimolimnion, in line with the trend in DIC. In the mixolimnion, seston samples are more 15N-enriched later in the summer, with the highest 15N value of 8.4‰ measured at 10.8 m in August. The same

113

13 sample yielded the lowest  Cseston value from the mixolimnion, -37.6 ‰, reflecting the

13 13 15 incorporation of C-depleted DIC. Minimum  Cseston and  Nseston values are found at

 15 the chemocline. At this depth,  is strongly N-enriched relative to its source in deeper waters. The 15N depletion of chemocline seston is likely due to the large isotope

+ effect associated with NH4 assimilation (Hoch et al., 1992).

Biomass collected from all depths below the chemocline has 15N values below

15 0‰. We measured minimum  Nseston values at the chemocline, where the average

15  Nseston value was -2.4±1.2‰ during our study period. PSB concentrations are highest at the chemocline, suggesting these organisms may be responsible for the minimum 15N values. Our analysis of seston stable isotope compositions compared with biomarker pigment distributions of the gravity-separated samples supports the assertion that PSB produce biomass with particularly low 15N values (Table 4-1).

In chapter 5 we present pigment data that demonstrate that the PSB population drops rapidly one meter below the chemocline, whereas the GSB abundance remains

+ 15 relatively constant. The differences between NH4 and seston  N values are smaller below the chemocline (Fig. 4-3). This observation suggests that GSB do not fractionate

+ N isotopes during NH4 assimilation as effectively as PSB, though a mechanism for this

+ difference is not obvious. Researchers have suggested that despite high NH4 concentrations, GSB may fix N2 in a similar meromictic lake, (Ohkouchi et al., 2005).

We present pigment-specific 15N values in chapter 5 that counter this view, demonstrating that GSB are assimilating the same N substrate as PSB.

114 4.4.2 Deep Basin Sedimentation and Water-Column Productivity

The proportional increases in accumulation rates of Corg and CaCO3 in Unit 1 may result from increased Synechococcus productivity (Table 4-4) (Hilfinger et al., 2001).

3- Such an increase in productivity would require increased PO4 availability in the

3- mixolimnion, as PO4 appears to limit productivity in the modern mixolimnion. In

3- November 2005, dissolved PO4 concentrations in the mixolimnion were between 0.11

- M and 0.27 M (Zerkle, 2006). These compare with NO3 concentrations between 50.0

M and 85.2 M on the same sampling date (Fig. 4-3). Redox-stratified basins like FGL

3- 3- are expected to accumulate PO4 in their deep waters, and [PO4 ] increases over time

(Arthur and Dean, 1998; Meyer and Kump, 2008), so increased mixolimnion productivity

3- during Unit 1 deposition may have resulted from increased PO4 flux from the monimolimnion. Currently, PSB and GSB productivity is sufficient to assimilate most of

3- the PO4 flux from the monimolimnion to the chemocline (Zerkle, 2006). Consequently,

3- increasing the size of the deep PO4 pool or decreasing chemocline productivity would

3- be required to increase the upward PO4 flux to the mixolimnion.

3- Increased PO4 flux from the surrounding soils and rock outcrops could also have stimulated mixolimnion productivity during Unit 1 deposition. The increased detrital mineral content of Unit 1 resulted from erosion of watershed soils and subaqueous rock outcrops (Hilfinger et al., 2001; Thompson et al., 1990). Though the hypothesized

3- increase in PO4 flux to mixolimnion preceded widespread agricultural applications of inorganic fertilizer, land surfaces in the catchment would have been destabilized by deforestation and plowing during the interval of increased mixolimnion productivity.

115 3- Thus, the increase in erosion rate may have increased the PO4 flux to the surface waters.

3- The detrital mineral content of FGL sediments may be a proxy for increased PO4 flux to the surface waters.

Sedimentary calcite records increased mixolimnion productivity in Unit 1. Fine- grained calcite precipitates in association with Synechococcus cells in the mixolimnion, accounting for most of the carbonate material deposited in FGL deep basin sediments and preserving the carbon of mixolimnion DIC (Thompson et al., 1997). A

~1‰ positive shift in calcite 13C values in Unit 1 compared with Unit 2 sediments has been interpreted as resulting from increased mixolimnion phytoplankton productivity

(Hilfinger et al., 2001). Sediment traps deployed in the center of the deep basin from

1966-1967 collected primarily fine-grained calcite, presumably precipitated by

Synechococcus, with lesser amounts of detrital dolomite and calcite cortication tubules from Chara (Brunskill, 1969). The detrital and benthic algal carbonate constituents tend to concentrate in shallow, marginal sediments, as observed in the shallow neck core. In the deep basin, these components are commonly redeposited in turbidites. Turbidites in

Unit 1 contain 77% carbonate minerals, whereas those in Unit 2 are 87% carbonate. In the laminated intervals, carbonate minerals comprise 71% of Unit 1 and 89% of Unit 2 sediments. The decrease in weight percent carbonate in Unit 1 laminated sediments and turbidites results from increased dilution with detrital minerals. Carbonate mineral mass accumulation rates are actually higher in Unit 1 than in Unit 2 (Fig. 4-9), even though the weight percent composition is lower, providing further evidence for higher

Synechococcus productivity during Unit 1 deposition.

116 Silicate minerals, primarily detrital clays, comprise a small proportion of Unit 2 sediments (< 5% in all samples), but are a more significant component of the gray- colored Unit 1 sediments (Fig. 4-7). There are two sources for detrital minerals in FGL sediments. Brunskill (1969) observed near-shore plumes of terrigenous material entering the lake during rainstorms. During the 18th and early 19th century, deforestation by the timber industry and subsequent agriculture may have destabilized land surfaces and increased particulate fluxes to FGL (Hilfinger et al., 2001). Thompson et al. (1990) determined that detrital gray clay and dolomite in the lake neck originate from the erosion of subaqueous rock outcrops where groundwater flows into the lake. They identify two water depths where thick clay layers accumulate, 10 m and 18 m, the same depths where groundwater enters the mixolimnion. Thus, the increased detrital mineral content of Unit

1 could result from either land-surface destabilization in the watershed or increased groundwater flow into the mixolimnion.

Thin, gray homogeneous clay layers, up to 3 mm thick, appear exclusively in Unit

1 sediments (Fig. 4-7). These layers do not contain a coarse component typical of tan turbidites deposited in Units 1 and 2 in the deep basin. The similarity in chemical and textural properties between one of these gray homogeneous layers and the upper sediments of the gray turbidite in the shallow neck core suggests that they may have a common source (Table 4-3). This supports the interpretation that the gray layers in deep basin sediments are distal turbidites (Ludlam, 1974), composed of material that probably originated in the shallow mixolimnion waters (Thompson et al., 1990). The presence of these layers only in Unit 1 suggests that subaqueous bedrock weathering and groundwater flux to the mixolimnion were enhanced during Unit 1 deposition.

117 4.4.3 Historical Shift in Organic Matter Sources

13 15 Surface sediment  Corg and  Ntot values from the deep basin core are similar to those of mixolimnion seston samples collected in May 2006 (Fig. 4-3), suggesting mixolimnion phytoplankton biomass may dominate sedimentary OM. Terrestrial plants and benthic algae also contribute to sedimentary organic matter, but their isotopic signatures are not obvious in sedimentary organic matter (Fig. 4-4). It is evident,

13 15 however, from  Corg and  Ntot values that Unit 1 sedimentary organic matter is not significantly influenced by seston from the pycnocline, chemocline, or monimolimnion.

This observation seems contrary to previous work that demonstrated that chemocline productivity is ~5 times greater than mixolimnion productivity in FGL (Culver and

Brunskill, 1969). The sedimentary flux of Synechococcus cells may be enhanced by their association with whiting calcite (Hodell and Schelske, 1998), possibly accounting for a disproportional flux of mixolimnion biomass to the sediments. Further, Thompson et al.

(2000) suggested that Synechococcus productivity may have been overlooked in the previous study, causing underestimation of mixolimnion productivity.

Chemocline biomass appears to be an important component of Unit 2 laminated sediments. Compared with Unit 1, the N and C isotopic compositions of Unit 2 samples

13 15 are shifted toward values for deep biomass (Fig. 4-6). Higher  Corg and  Ntot values also correlate strongly with lower okenone concentrations in the laminated sediments

(Figs. 4-11, 4-12). As okenone is only produced by PSB, its elevated concentrations in

Unit 2 sediments most likely illustrates that PSB biomass is a larger component of Unit 2 organic matter, though pigment concentrations are not necessarily proportional to total

118 PSB biomass (Gobel, 1978; Overmann and Garcia-Pichel, 2006). Turbidites from Units

13 15 1 and 2 generally have low okenone concentrations and high  Corg and  Ntot values.

Turbidite sediments are composed primarily of material that was initially deposited above the chemocline, so turbidite OM deposited in Units 1 and 2 derives mostly from the mixolimnion with a small input of redeposited chemocline biomass. These data support the hypothesis that Unit 2 sediments contain a greater proportion of 15N- and 13C-depleted

PSB biomass, as the turbidite OM that was initially deposited above the chemocline

13 15 maintains similar  Corg and  N values in Units 1 and 2.

Table 4-4 summarizes the differences in the MAR of bulk sedimentary constituents in Units 1 and 2. Bulk sediment, carbonate, and organic carbon MARs are all more than 2 times greater in Unit 1. As described above, the increases in CaCO3 and

Corg MARs may be tied directly to Synechococcus productivity, illustrating that mixolimnion productivity was greater during Unit 1 deposition. Increased mixolimnion

13 15 OM flux to the sediments thus may have caused the increases in  Corg and  Ntot values observed in Unit 1. During Unit 1 deposition, PSB populations may have been inhibited at least seasonally by shading from Synechococcus, accounting for decreased concentrations of okenone in the sediments. Alternatively, a shallower chemocline during Unit 2 deposition may have allowed PSB abundance to increase to higher than modern levels. Meromictic Mahoney Lake has a shallow chemocline at 6 meters, allowing for a very dense PSB population and okenone concentrations in the water column and sediments that are significantly greater than in FGL (Coolen and Overmann,

1998). We prefer the former explanation, however, as there is evidence for increased

119 mixolimnion productivity and there is no a priori reason to assume the chemocline depth changed significantly during the Unit 1-2 transition, in the late 18th Century.

4.4.4 Recent Shift in Chemocline Depth

Previous publications reported a somewhat shallower chemocline in the 1960s and 1970s (Brunskill and Ludlam, 1969; Takahashi et al., 1968; Torgersen et al., 1981).

High sulfide flux from the monimolimnion to the chemocline could be responsible for the observed shoaling. We calculated high MAR of Corg, CaCO3, and detrital minerals in the laminated sediment interval from 1955-1968 (Fig. 4-9), evidence for higher nutrient flux and mixolimnion productivity. The increased OM flux to the monimolimnion could have stimulated higher sulfate reduction rates and increased the upward sulfide flux to the chemocline. The excess sulfide would lead to upward expansion of reducing conditions to shallower depths in the pycnocline. Torgersen et al. (1981) measured the chemocline at 15-17 m in 1975, about 4 m shallower than our current measured depth, but still within the modern pycnocline depth range. As a result, we expect that sediments deposited at

17-m would have been just below the chemocline at that time.

A chemocline shoaling event is recorded in the shallow neck core, evidenced by a laminated sediment interval containing elevated PSB and GSB pigment concentrations

(Figs. 4-5, 4-10). The shoaled chemocline coincides with ~3‰ negative excursions in

15 13 both  Ntot and  Corg values (Fig. 4-5). The laminated sediment interval apparently was deposited under anoxic waters, allowing the laminae to persist in the absence of

15 burrowing organisms. Okenone concentrations correlate inversely with  Ntot and

120

13  Corg values, illustrating that a larger relative component of PSB biomass may be responsible for the isotope shifts (Figs. 4-13; 4-14). The core-top sample has C and N isotope values and pigment concentrations similar to those below the excursion interval, showing that modern conditions are similar to those prior to the chemocline shoaling.

Based on its location near the top of the core, we suggest that the laminated interval resulted from the observed shallower chemocline in the mid to late 20th century.

4.4.5 Implications for Geological Samples

Using bacterial pigment distributions and bulk Corg and Ntot stable isotope data, we have demonstrated that high concentrations of PSB biomass in the sediments of FGL

15 13 generated low  Ntot and  Corg values. This relationship may also have produced

15 13 anomalously low  Ntot and  Corg values in some ancient sediments that record negative isotope excursions. This scenario requires a relative increase in chemocline biomass preservation in the sediments for the duration of the isotope excursion. In FGL, we document two mechanisms to increase the relative proportion of PSB biomass in the sediments. The first is relatively constant PSB productivity superimposed over a decrease in surface water productivity, illustrated in the comparison of Unit 2 with Unit 1 sediments. Alternatively, a transition from oxygenated to anoxic waters overlying a sampling site might allow a population of planktonic PSB to emerge. These organisms can significantly affect sedimentary C and N isotopic compositions, as observed in the laminated interval of shallow neck core.

121 PSB growth requires a chemocline shallower than 25 m (Overmann and Garcia-

Pichel, 2006). Such settings are rare on earth now, limited to meromictic lakes and small restricted brackish basins (Itoh et al., 2003; Overmann et al., 1991; Velinsky and Fogel,

1999). In the Black Sea, the most cited modern analog for ancient redox-stratified oceans, the 80-100 m depth of the chemocline inhibits phototrophic PSB growth (Repeta and Simpson, 1991). Instead GSB compose the entire phototrophic sulfide oxidizing community at the chemocline. Although their biomarkers are detected in Black Sea sediments (Freeman et al., 1994; Repeta, 1993), GSB biomass and 13C-depleted biomass from deep-dwelling chemoautotrophs are not thought to significantly influence

13 15 sedimentary  Corg and  Ntot values (Fry et al., 1991). Apparently, sinking PSB cells with their associated sulfur granules provide greater potential for generating chemocline-

13 15 derived negative shifts in  Corg and  Ntot values in sediments and geological samples.

Three GSB and PSB carotenoid derivatives have been identified in ancient samples; their distributions are used to determine the temporal ranges of apparently widespread photic-zone euxinia. Isorenieratane, a diagenetic product of isorenieratene produced by brown-pigmented GSB (Summons and Powell, 1987), is found in samples from many ancient settings (Grice et al., 2005; Koopmans et al., 1996). The presence of isorenieratane, however, does not indicate if the chemocline was shallow enough to support a PSB population. Chlorobactene, produced by green-pigmented GSB, degrades to in ancient sediments (Schouten et al., 2000). In modern settings, the

GSB species that produce chlorobactene require a similarly shallow chemocline as PSB.

The presence of chlorobactane in ancient samples suggests that the chemocline may have been shallow enough for PSB growth, though direct biomarker evidence is needed to

122 demonstrate the presence of PSB. Okenane, a diagenetic product of okenone, is the most promising compound indicating the presence of PSB in ancient environments. Okenane has been elusive, but was detected in rocks of the Proterozoic Barney Creek Formation

(Brocks et al., 2005; Brocks and Schaeffer, 2008). It follows then that there may be key sedimentary intervals, particularly in the Proterozoic, which could contain abundant chemocline-derived sedimentary organic matter.

GSB and PSB also inhabit benthic mats in shallow-water settings. Their biomarkers may be found in preserved mats in the geological record (Brocks and

Summons, 2004), so facies determinations are critical for diagnosing water-column euxinic conditions. Some researchers have argued that okenone is only produced by planktonic PSB and is therefore a marker for a shallow chemocline in the water column

(Brocks and Schaeffer, 2008). In FGL, however, benthic mat-forming PSB produce okenone on the sediment surface under oxygenated waters (Meyer et al., submitted).

Thus, known biomarkers of PSB and GSB do not provide definitive evidence for water column euxinia. Their interpretation as water-column derived pigments requires additional sedimentological evidence for deep water deposition.

The Mesoproterozoic Barney Creek Formation (BCF) presents a possible test case for the inclusion of chemocline-derived organic matter in sediments. The ocean during the Proterozoic Eon may have been redox-stratified, as atmospheric oxygen concentrations were too low to maintain oxygenated deep ocean waters (Canfield, 1998).

Subsequent modeling even suggests that the oxygen content of the atmosphere may not have been high enough to maintain an oxygenated surface layer, so anoxic waters may have reached the sea surface (Kump et al., 2005). The BCF interval studied by Brocks

123 and Schaeffer (2008) contains well-preserved OM that was deposited in the restricted

McArthur Basin (1640±3 Ma). This basin may not be representative of open ocean conditions in the Mesoproterozoic, and variability in sedimentary facies, including possible benthic mats, suggests that water depth varied considerably. The BCF has generally been viewed as a relatively deep basin with sub-wave-base sediments

(Rawlings, 1999). Photic zone euxinic conditions were prevalent in this basin, evidenced by their finding of isorenieratane, chlorobactane, and okenane (Brocks et al., 2005;

Brocks and Schaeffer, 2008). It is uncertain whether GSB and PSB producing these pigments were planktonic or benthic. The okenane finding is intriguing, however, as it is a biomarker for PSB. Assuming the okenone came from a water-column source, this is a good location to test the effect of PSB and other chemocline-derived biomass on bulk sediment C and N isotope distributions.

4.5 Summary

FGL sediments receive organic matter from several sources, including sinking organic matter from GSB and PSB living in the sulfidic waters at the top of the monimolimnion. Among these organisms, PSB, due to their inclusion of high-density sulfur granules and vulnerability to grazing and incorporation in fecal pellets are especially prone to sinking and inclusion in the sedimentary record. In FGL sediments we observed intervals characterized by increased proportions of PSB biomass as evidenced by elevated concentrations of okenone and Bphe a (See also Chapter 5).

15 13 These intervals also have lower  Ntot and  Corg values which are best explained as

124 resulting directly from the incorporation of PSB biomass. PSB thrive in FGL due to its relatively shallow chemocline, and we suggest that ancient sedimentary intervals with similarly shallow chemoclines may contain C and N isotope records affected by the inclusion of PSB biomass. The redox-stratified basins of the Mesoproterozoic may be the most similar to FGL waters, and therefore sedimentary rocks from this time interval should be examined for evidence of negative isotope excursions related to PSB biomass incorporation.

4.6 Cited References

Altabet, M., Pilskaln, C., Thunell, R., Pride, C., Sigman, D., Chavez, F., and Francois, R., 1999, The nitrogen isotope biogeochemistry of sinking particles from the margin of the Eastern North Pacific: Deep-Sea Research Part I, v. 46, p. 655-679. Altabet, M.A., and Francois, R., 1994, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nitrate utilization: Global Biogeochemical Cycles, v. 8, p. 103- 116. Arthur, M., and Dean, W., 1998, Organic-matter production and preservation and evolution of anoxia in the Holocene Black Sea: Paleoceanography, v. 13, p. 395- 411. Brocks, J.J., Love, G.D., Summons, R.E., Knoll, A.H., Logan, G.A., and Bowden, S.A., 2005, Biomarker evidence for green and purple sulphur bacteria in a stratified Palaeoproterozoic sea: Nature, v. 437, p. 866-870. Brocks, J.J., and Schaeffer, P., 2008, Okenane, a biomarker for purple sulfur bacteria (Chromatiaceae), and other new carotenoid derivatives from the 1640 Ma Barney Creek Formation: Geochimica et Cosmochimica Acta, v. 72, p. 1396-1414. Brocks, J.J., and Summons, R.E., 2004, Sedimentary hydrocarbons, biomarkers for early life, in Schlesinger, W.H., Holland, H.D., and Turekian, K.K., eds., Treatise on Geochemistry, Elsevier, p. 63-115. Brunskill, G.J., 1969, Fayetteville Green Lake, New York. II. Precipitation and sedimentation of calcite in a meromictic lake with laminated sediments: Limnology and Oceanography, v. 16, p. 830-847. Brunskill, G.J., and Ludlam, S.D., 1969, Fayetteville Green Lake, New York. I. Physical and chemical limnology: Limnology and Oceanography, v. 14, p. 817-829. Canfield, D.E., 1998, A new model for Proterozoic ocean chemistry: Nature, v. 396, p. 450-453.

125 Codispoti, L.A., Friederich, G.E., Murray, J.W., and Sakamoto, C.M., 1991, Chemical variability in the Black Sea: implications of continuous vertical profiles that penetrated the oxic/anoxic interface: Deep-Sea Research Part A, v. 38, supplement 2, p. S691-S710. Coolen, M.J.L., and Overmann, J., 1998, Analysis of subfossil molecular remains of purple sulfur bacteria in a lake sediment: Applied and Environmental Microbiology, v. 64, p. 4513-4521. Culver, D.A., and Brunskill, G.J., 1969, Fayetteville Green Lake, New York. V. Sudies of primary production and zooplankton in a meromictic marl lake.: Limnology and Oceanography, v. 14, p. 862-872. Deevey, E.S., Nakai, N., and Stuiver, M., 1963, Fractionation of sulfur and carbon isotopes in a meromictic lake: Science, v. 139, p. 407-408. Freeman, K.H., and Hayes, J.M., 1992, Fractionation of carbon isotopes by phytoplankton and estimates of ancient CO2 levels: Global Biogeochemical Cycles, v. 6, p. 185-198. Freeman, K.H., Hayes, J.M., Trendel, J.M., and Albrecht, P., 1990, Evidence from carbon isotope measurements for diverse origins of sedimentary hydrocarbons: Nature, v. 343, p. 254-256. Freeman, K.H., Wakeham, S.G., and Hayes, J.M., 1994, Predictive isotopic biogeochemistry: Hydrocarbons from anoxic marine basins: Organic Geochemistry, v. 21, p. 629-644. Fry, B., 1986, Sources of carbon and sulfur nutrition for consumers in three meromictic lakes of New York State: Limnology and Oceanography, v. 31, p. 79-88. Fry, B., Jannasch, H.W., Molyneaux, S.J., Wirsen, C.O., Muramoto, J.A., and King, S., 1991, Stable isotope studies of the carbon, nitrogen, and sulfur cycles in the Black Sea and the Cariaco Trench: Deep-Sea Research Part A, v. 38, supplement 2, p. S1003-S1119. Gobel, F., 1978, Quantum efficiencies of growth, in Clayton, R.K., and Sistrom, W.R., eds., The Photosynthetic Bacteria: New York, Plenum Press, p. 907-925. Grice, K., Cao, C.Q., Love, G.D., Bottcher, M.E., Twitchett, R.J., Grosjean, E., Summons, R.E., Turgeon, S.C., Dunning, W., and Jin, Y.G., 2005, Photic zone euxinia during the Permian-Triassic superanoxic event: Science, v. 307, p. 706- 709. Gruendling, G.K., and Malanchuk, J.L., 1974, Seasonal and spatial distribution of phosphates, nitrates, and silicates in Lake Champlain, U.S.A.: Hydrobiologia, v. 45, p. 405-421. Hilfinger, M.F., Mullins, H.T., Burnett, A., and Kirby, M.E., 2001, A 2500 year sediment record from Fayetteville Green Lake, New York: evidence for anthropogenic impacts and historic isotope shift: Journal of Paleolimnology, v. 26, p. 293-305. Hoch, M.P., Fogel, M.L., and Kirchman, D.L., 1992, Isotope fractionation associated with ammonium uptake by a marine bacterium: Limnology and Oceanography, v. 37, p. 1447-1459. —, 1994, Isotope fractionation during ammonium uptake by marine microbial assemblages: Geomicrobiology Journal, v. 12, p. 113-127.

126 Hodell, D.A., and Schelske, C.L., 1998, Production, sedimentation, and isotopic composition of organic matter in Lake Ontario: Limnology and Oceanography, v. 43, p. 200-214. Holmes, R.M., McClelland, J.W., Sigman, D.M., Fry, B., and Peterson, T.F., 1998, 15 + Measuring N-NH4 in marine, estuarine and fresh waters: An adaptation of the ammonia diffusion method for samples with low ammonium concentrations: Marine Chemistry, v. 60, p. 235-243. House, C.H., Schopf, J.W., and Stetter, K.O., 2003, Carbon isotopic fractionation by Archaeans and other thermophilic prokaryotes: Organic Geochemistry, v. 34, p. 345-356. Itoh, N., Tani, Y., Nagatani, T., and Soma, M., 2003, Phototrophic activity and redox condition in Lake Hamana, Japan, indicated by sedimentary photosynthetic pigments and molybdenum over the last ~250 years: Journal of Paleolimnology, v. 29, p. 403-422. Junium, C.K., and Arthur, M.A., 2007, Nitrogen cycling during the Cretaceous, Cenomanian-Turonian oceanic anoxic event II: Geochemistry Geophysics Geosystems, v. 8. Koopmans, M.P., Koster, J., van Kaam Peters, H.M.E., Kenig, F., Schouten, S., Hartgers, W.A., de Leeuw, J.W., and Damste, J.S.S., 1996, Diagenetic and catagenetic products of isorenieratene: Molecular indicators for photic zone anoxia: Geochimica et Cosmochimica Acta, v. 60, p. 4467-4496. Kump, L.R., Pavlov, A., and Arthur, M.A., 2005, Massive release of hydrogen sulfide to the surface ocean and atmosphere during intervals of oceanic anoxia: Geology, v. 33, p. 397-400. Kuypers, M.M.M., van Breugel, Y., Schouten, S., Erba, E., and Damste, J.S.S., 2004, N2- fixing cyanobacteria supplied nutrient N for Cretaceous oceanic anoxic events: Geology, v. 32, p. 853-856. Logan, G.A., Hayes, J.M., Hieshima, G.B., and Summons, R.E., 1995, Terminal Proterozoic reorganization of biogeochemical cycles: Nature, v. 376, p. 53-56. Ludlam, S.D., 1969, Fayetteville Green Lake, New York. III. The laminated sediments: Limnology and Oceanography, v. 14, p. 848-857. —, 1974, Fayetteville Green Lake, New York. VI. Role of turbidity currents in lake sedimentation: Limnology and Oceanography, v. 19, p. 656-664. —, 1984, Fayetteville Green Lake, New York, USA . VII. Varve chronology and sediment focusing: Chemical Geology, v. 44, p. 85-100. Mariotti, A., Germon, J.C., Hubert, P., Kaiser, P., Letolle, R., Tardieux, A., and Tardieux, P., 1981, Experimental determination of nitrogen kinetic isotope fractionation: Some principles; illustration for the denitrification and nitrification processes: Plant and Soil, v. 62, p. 413-430. Mas, J., Pedros-Alio, C., and Guerrero, R., 1990, In situ specific loss and growth rates of purple sulfur bacteria in Lake Ciso: FEMS Microbiology Ecology, v. 73, p. 271- 281. Meyer, K.M., and Kump, L.R., 2008, Oceanic euxinia in Earth history: Causes and consequences: Annual Review of Earth and Planetary Sciences, v. 36, p. 251-288.

127 Meyer, K.M., Kump, L.R., Macalady, J.L., Schaperdoth, I., Fulton, J.M., and Freeman, K.H., submitted, Benthic production of the sulfur phototroph biomarker okenone: Geobiology. Miner, N.A., 1933, The origin and history of Green and Round Lakes in Green Lake State Park at Fayetteville, New York. [M.A. thesis]: Syracuse, NY, Syracuse University. Mitchell, M.J., Driscoll, C.T., Kahl, J.S., Likens, G.E., Murdoch, P.S., and Pardo, L.H., 1996, Climate control of nitrate loss from forested watersheds in the northeast United States: Environmental Science & Technology, v. 30, p. 2609-2612. Montoya, J.P., and McCarthy, J.J., 1995, Isotope fractionation during nitrate uptake by phytoplankton grown in continuous culture: Journal of Plankton Research, v. 17, p. 439-464. Needoba, J.A., Waser, N.A., Harrison, P.J., and Calvert, S.E., 2003, Nitrogen isotope fractionation in 12 species of marine phytoplankton during growth on nitrate: Marine Ecology-Progress Series, v. 255, p. 81-91. Ohkouchi, N., Nakajima, Y., Okada, H., Ogawa, N.O., Suga, H., Oguri, K., and Kitazato, H., 2005, Biogeochemical processes on the saline meromictic Lake Kaiike, Japan: implications from molecular isotopic evidences of photosynthetic pigments: Environmental Microbiology, v. 7, p. 1009-1016. Overmann, J., Beatty, J.T., Hall, K.J., Pfennig, N., and Northcote, T.G., 1991, Characterization of a dense, purple sulfur bacterial layer in a meromictic salt lake: Limnology and Oceanography, v. 36, p. 846-859. Overmann, J., and Garcia-Pichel, F., 2006, The phototrophic way of life, in Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H., and Stackebrandt, E., eds., The Prokaryotes, Volume 2: Dordrecht, Springer, p. 32-85. Pardo, L.H., Driscoll, C.T., and Likens, G.E., 1995, Patterns of nitrate loss from a chronosequence of clear-cut watersheds: Water Air and Soil Pollution, v. 85, p. 1659-1664. Popp, B.N., Laws, E.A., Bidigare, R.P., Dore, J.E., Hanson, K.L., and Wakeham, S.G., 1998, Effect of phytoplankton cell geometry on carbon isotope fractionation: Geochimica et Cosmochimica Acta, v. 62, p. 69-77. Quandt, L., Gottschalk, H., Ziegler, H., and Stichler, W., 1977, Isotope discrimination by photosynthetic bacteria: FEMS Microbiology Letters, v. 1, p. 125-128. Rau, G.H., Arthur, M.A., and Dean, W.E., 1987, 15N/14N variations in Cretaceous Atlantic sedimentary sequences: implication for past changes in marine nitrogen biochemistry: Earth and Planetary Science Letters, v. 82, p. 269-279. Rawlings, D.J., 1999, Stratigraphic resolution of a mutiphase intracratonic basin system: the McArthur Basin, northern Australia: Australian Journal of Earth Science, v. 46, p. 703-723. Redfield, A.C., 1958, The biological control of chemical factors on the environment: American Scientist, v. 46, p. 205-221. Repeta, D.J., 1993, A high resolution historical record of Holocene anoxygenic primary production in the Black Sea: Geochimica et Cosmochimica Acta, v. 57, p. 4337- 4342.

128 Repeta, D.J., and Simpson, D.J., 1991, The distribution and recycling of chlorophyll, bacteriochlorophyll and carotenoids in the Black Sea: Deep-Sea Research Part A, v. 38, supplement 2, p. S969-S984. Robinson, D., 2001, 15N as an integrator of the nitrogen cycle: TRENDS in Ecology & Evolution, v. 16, p. 153-162. Sachs, J.P., and Repeta, D.J., 1999, Oligotrophy and nitrogen fixation during eastern Mediterranean sapropel events: Science, v. 286, p. 2485-2488. Schouten, S., Van Kaam-Peters, H.M.E., Rijpstra, W.I.C., Schoell, M., and Damste, J.S.S., 2000, Effects of an oceanic anoxic event on the stable carbon isotopic composition of Early Toarcian carbon: American Journal of Science, v. 300, p. 1- 22. Schubert, C.J., Durisch-Kaiser, E., Werhrli, B., Thamdrup, B., Lam, P., and Kuypers, M.M.M., 2006, Anaerobic ammonium oxidation in a tropical freshwater system (Lake Tanganyika): Environmental Microbiology, v. 8, p. 1857-1863. Sigman, D.M., Altabet, M.A., Michemer, R., McCorkle, D.C., Fry, B., and Holmes, R.M., 1997, Natural abundance-level measurements of the nitrogen isotopic composition of oceanic nitrate: an adaptation of the ammonia diffusion method: Marine Chemistry, v. 57, p. 227-242. Sirevag, R., Buchanan, B.B., Berry, J.A., and Troughton, J.H., 1977, Mechanisms of CO2 fixation in bacterial photosynthesis studied by the carbon isotope fractionation technique: Archives of Microbiology, v. 112, p. 35-38. Summons, R.E., and Powell, T.G., 1987, Identification of aryl isoprenoids in source rocks and crude oils - biological markers for the green sulfur bacteria: Geochimica et Cosmochimica Acta, v. 51, p. 557-566. Takahashi, T., Broecker, W., Li, Y.H., and Thurber, D., 1968, Chemical and isotopic balances for a meromictic lake: Limnology and Oceanography, v. 13, p. 272-292. Thompson, J.B., Ferris, F.G., and Smith, D.A., 1990, Geomicrobiology and sedimentology of the mixolimnion and chemocline in Fayetteville Green Lake, New York: Palaios, v. 5, p. 52-75. Thompson, J.B., Schultz-Lam, S., Beveridge, T.J., and Des Marais, D.J., 1997, Whiting events: Biogenic origin due to the photosynthetic activity of cyanobacterial picoplankton: Limnology and Oceanography, v. 42, p. 133-141. Torgersen, T., Hammond, D.E., Clarke, W.B., and Peng, T.-H., 1981, Fayetteville Green Lake, New York: 3H-3He water mass ages and secondary chemical structure: Limnology and Oceanography, v. 26, p. 110-122. Tyrell, T., 1999, The relative influences of nitrogen and phosphorus on oceanic primary productivity: Nature, v. 400, p. 525-531. Velinsky, D.J., and Fogel, M.L., 1999, Cycling of dissolved and particulate nitrogen and carbon in the Framvaren Fjord, Norway: stable isotopic variations: Marine Chemistry, v. 67, p. 161-180. Wada, E., and Hattori, A., 1978, Nitrogen isotope effects in the assimilation of inorganic nitrogenous compounds by marine diatoms: Geomicrobiology Journal, v. 1, p. 85- 101. Waser, N.A.D., Harrison, P.J., Nielsen, B., Calvert, S.E., and Turpin, D.H., 1998, Nitrogen isotope fractionation during the uptake and assimilation of nitrate,

129 nitrite, ammonium, and urea by a marine diatom: Limnology and Oceanography, v. 43, p. 215-224. Zerkle, A.L., 2006, Microbial Trace Metal Requirements: Limiting Nutrients and Potential Biosignatures [Ph. D. thesis]: University Park, PA, The Pennsylvania State University.

130

Figure 4-1. FGL map with sample locations. The lake sits in a topographic depression with steep slopes to the east, south, and west reaching 70-100 m above the lake surface. Deep basin cores and all water-column samples were collected at station A. The 17-m core was collected at station B, above the chemocline. This map is modified from Ludlam (1981).

131

Figure 4-2. FGL water chemistry, April 2004. The chemocline (Chem.) is defined as the depth where oxidation/reduction potential (ORP) rapidly transitions from positive to negative values. It is marked also by the turbidity (Turb.) maximum of the bacterial plate. Specific conductance (Sp. Cond.) increases rapidly through the pycnocline due to increased dissolved salt concentrations. This density transition inhibits seasonal mixing of the monimolimnion, and defines the base of the mixolimnion.

132

Figure 4-3. Water column seston and nutrient profiles for four sampling dates. The turbidity maximum (chemocline) for each sampling date was adjusted to 20 m water depth. The dashed line at 20 m represents the chemocline, and the shaded region from 17-20 m corresponds to the pycnocline. Vertical dashed lines in the seston 13C and 15N panels display the surface sediment values for reference.

133

Figure 4-4. Deep basin core C and N isotope records. The horizontal dashed line marks 13 the Unit 1-2 boundary. The vertical dashed lines represent the surface sediment  Corg 15 and  Ntot values for comparison down-core.

134

Figure 4-5. Shallow neck core isotope and pigment records. The shaded excursion interval demonstrates the association between elevated concentrations of bacterial 15 13 pigments and lower  Nbulk and  Corg values. The laminated interval probably extended up to the top of the shaded interval at 2.5 cm when the sediments were deposited, and the laminae were mixed by bioturbation after the overlying waters were oxygenated.

135

Figure 4-6. Survey of C and N isotope values of FGL sediment organic matter sources. Unit 1 15N and 13C values are similar to those of mixolimnion seston from the May sampling date. Unit 2 samples contain a larger relative proportion of material from the chemocline and monimolimnion. Data are reported only for laminated intervals in Units 1 and 2. Data labels list water depths of Chara samples and water depths and sampling months of mixolimnion seston. Isotope data are reported as standard ‰ values relative to Air (N) and VPDB (C).

136

Figure 4-7. Deep basin composite core stratigraphy and varve count timescale in calendar years. The Unit 1-2 boundary separates red/brown Unit 2 sediments from gray/tan Unit 1 sediments. The color change is primarily due to greater detrital clay content in Unit 1, illustrated in this figure by higher percent insoluble residue.

137

Figure 4-8. Sediment age model with linear accumulation rates for Unit 1 and Unit 2 laminated sediments from the deep basin composite core. The depth scale on the y-axis includes only laminated sediments, and the ages are based on varve counts. The change in linear accumulation rate is primarily due to increased mass accumulation rate in Unit 1 (Fig. 4-9), although greater compaction in Unit 2 is likely to contribute also.

138

Figure 4-9. Mass accumulation rates calculated for laminated intervals in the deep basin core. Accumulation rate maxima for all three parametersare reached near 1960 and 1860 13 15 in Unit 1. These peaks correlate with maximum  Corg and  Ntot values (Fig. 4-9) and minimum concentrations of Bchl e, Bphe a, and okenone presented in Ch. 5 (Fig. 5-3).

139

Figure 4-10. Shallow neck core stratigraphy and bulk properties.

140

15 Figure 4-11. Inverse relationship between okenone concentration and  Ntot values in the deep basin core. Unit 1 laminated sediments plot in the upper left quadrant whereas laminated Unit 2 sediments trend toward the lower right. Turbidites from Units 1 and 2 all plot in the upper left quadrant. The linear regression line is plotted for laminated sediments only. Okenone concentrations are from Chapter 5.

141

13 Figure 4-12. Deep basin core  Corg values plotted versus okenone concentrations. Unit 1 laminated samples plot on the left side of the chart and Unit 2 on the right. Okenone 13 concentrations in all turbidite samples are similar to Unit 1 samples, but turbidite  Corg values are similar to Unit 1 and Unit 2 laminated samples. Okenone concentrations are from Chapter 5.

142

15 Figure 4-13. Shallow neck core okenone vs.  Nbulk values. There is a strong inverse relationship between these two parameters, with the low 15N values and high okenone concentrations found in and above the laminated interval.

143

13 13 Figure 4-14. Shallow neck core okenone vs.  Corg values. Low  Corg values correlate with high okenone concentrations, demonstrating that the incorporation of more PSB 13 biomass in the laminated interval results in lower  Corg values.

144 Table 4-1. Gravity separation of chemocline seston collected May 09, 2006. Bphe a and okenone are produced by PSB. Bchl e is produced by brown-colored strains of GSB. Chl a may come from cyanobacteria living in the chemocline and Phe a from slowly sinking mixolimnion-derived biomass (Chapter 5).

Purple cell pellet Brown supernatant Bphe a (g l-1) 5.6 0.7 Okenone (g l-1) 40.3 17.2 Bchl e (g l-1) 0.4 33.1 Chl a (g l-1) 0.1 2.2 Phe a (g l-1) 0.1 4.1 15N -5.8 -1.9 13C -43.2 -40.8 POC (mg l-1) 2.49 3.18 PON (mg l-1) 0.43 0.57 C/N (molar) 6.7 6.5

145 Table 4-2. Leaf litter N and C stable isotope values.

15N 13C Fraxinus americana 1.3 -31.5 Thuja occidentalis -1.4 -27.3 Liriodendron tulipifera -0.4 -28.7 Betula lutea -0.8 -27.3 Platanus occidentalis 1.8 -30.5 Populus sp. 3.2 -27.9

146 Table 4-3. Comparison of gray homogeneous layers from the shallow neck and deep 15 13 basin cores. Both samples have relatively high  Ntot and  Corg values, low carbonate and high insoluble residue content, and are primarily fine-grained material.

17-m core 53-m core Interpretation Proximal turbidite Distal turbidite 15  Ntot 5.7 5.6 13  Corg -25.7 -26.6 Corg (%) 3.3 3.0 Carbonate (%) 13 23 Insoluble residue (%) 81 70 Fine fraction (%) 95 89

147 Table 4-4. Mass Accumulation Rates (deep basin core). The values presented here are weighted averages of the laminate interval MARs presented in Fig. 4-7.

Unit 1 Unit 2 Ratio 1:2 Bulk sediment g m-2 yr-1 363 139 2.6 Detrital minerals g m-2 yr-1 72 5 14.4 Carbonate minerals g m-2 yr-1 268 127 2.1 Organic carbon g m-2 yr-1 10.6 4.8 2.2

Chapter 5

Pigment-Specific C and N Isotopes in a Meromictic Lake

Abstract

Compound-specific 13C and 15N values can provide valuable information about specific organisms growing in a complex ecosystem. In this study we present methods for isolating pigments from microbial assemblages in sufficient purity to allow 15N and

13C measurements. We focus on pigments from purple sulfur bacteria (PSB), green sulfur bacteria (GSB), and cyanobacteria produced near the chemocline of a meromictic lake and extracted from water column and sediment samples. We demonstrate that chlorophyll a produced by cyanobacteria in the contributes almost 50% of the total chlorophyll a derivatives in the surface sediments. Total sedimentary organic matter, on the other hand, is dominated by cyanobacterial biomass from well-mixed surface waters. The seasonal bloom cycle for phototrophic bacteria near the chemocline includes interactions with the deep cyanobacterial population, which are critical for understanding temporal productivity patterns of GSB and PSB in the chemocline. Finally, the sediments record relationships between pigment C and N stable isotope compositions that illustrate that increased mixolimnion productivity affected chemocline-dwelling PSB and GSB. These relationships are important when considering using compound-specific stable isotope analysis of porphyrins preserved from ancient settings.

149 5.1 Introduction

Phytoplankton are a critical component of the global carbon cycle, assimilating

CO2 and transferring it to deep waters and sediments via the biological pump (Broecker and Peng, 1982). The biological pump also decreases surface water nutrient concentrations by transferring them to deep water masses. Further, phytoplankton regulate nutrient cycling by maintaining the balance between nitrogen and phosphorus in surface waters (Tyrrell, 1999). Past changes in global biogeochemical cycling can be examined in light of changes in phytoplankton phylogeny and growth conditions over time. Photosynthetic pigment derivatives preserved in sediments are particularly useful biomarkers for paleooceanographers who wish to characterize ancient phytoplankton populations. A variety of pigments are used by phytoplankton to harness light energy for cell growth (Britton et al., 2004; Keely, 2006). Pigment biomarkers include chlorins and carotenoids, some of which are produced by specific organisms that live in particular ecological niches (Ediger et al., 2006). In sediments and sedimentary rocks, pigment biomarkers are used to infer past marine ecology, which can have important implications for paleooceanography and paleoclimate research.

Pigment analysis is an especially robust tool when combined with compound- specific C and N stable isotope analysis. These data can further constrain phytoplankton phylogeny and growth conditions (Kashiyama et al., 2008). Pigment N and C isotopic compositions can be used to estimate whole biomass 15N and 13C values for the organisms that produced them (Huang et al., 2000; Ohkouchi et al., 2006; Sachs et al.,

1999). The calculated biomass values can be used in turn to evaluate nutrient substrates

150

13 and to calculate the isotopic compositions of dissolved inorganic carbon (DIC).  CDIC gradients from surface to deep waters are a function of export productivity to deep waters and deep water mass residence times (Freeman et al., 1994). Other uses of pigment 13C values include distinguishing between terrestrial and aquatic pigment sources

(Chikaraishi et al., 2007) and establishing the presence of preserved green sulfur bacterial

(GSB) biomass, an indicator for redox-stratification and photic-zone euxinia (Huang et al., 2000).

Sedimentary porphyrins (e.g. deoxophylloerythroetioporphyrin—DPEP), and chlorins (e.g. chlorophyll a) are tetrapyrroles and therefore contain four N atoms per molecule. Few biomarkers contain N, so tetrapyrrole pigments and pigment derivatives are unique in their potential application to studies of ancient N cycling. Chlorophyll a

(Chl a) is the most common pigment produced by organisms performing oxygenic photosynthesis. Chl a is not preserved well in sediments; but it degrades to pheophytin a

(Phe a) and pyropheophytin a (Pphe a), which are stable in recent sediments (Keely,

2006). The Treibs scheme links sedimentary porphyrins and chlorins to Chl a precursors

(Keely, 2006); thus, porphyrins are useful for determining both the N and C isotopic compositions of their phytoplankton sources (Ohkouchi et al., 2006; Ohkouchi et al.,

2008). As tetrapyrroles are not compatible with , alternative methods have been developed to measure their isotopic compositions. Typically, purification is carried out by high performance liquid chromatography (HPLC) and collected pigment fractions are converted to gas phases (N2 or N2O and CO2) for IRMS analysis (Bidigare et al., 1991; Higgins et al., 2009; Sachs and Repeta, 2000). Tetrapyrroles may also be converted to maleimides which can be analyzed by GC-IRMS (Chikaraishi et al., 2008).

151 Most pigment-specific C and N isotope studies have focused on Chl a and its derivatives, as it is produced by a wide range of phytoplankton in diverse habitats. In

15 deep-sea settings, sedimentary porphyrin  N values have demonstrated that bacterial N2 fixation provided fixed-N for phytoplankton growth in ancient redox-stratified oceans

(Chicarelli et al., 1993; Kashiyama et al., 2008; Ohkouchi et al., 2006; Sachs and Repeta,

1999). In modern estuaries on the Gulf of Mexico, C and N isotope compositions were used to establish that Chl a in particulate organic matter derives from in situ productivity

15 (Qian et al., 1996). In an estuary in Massachusetts,  NChl a values demonstrated that

+ phytoplankton preferentially assimilated remineralized NH4 from upstream sources

- instead of downstream NO3 influx (York et al., 2007).

Studies of Chl a degradation products commonly assume that the preserved compounds derive from the dominant phytoplankton sources (Kashiyama et al., 2008).

This approach yields biomass 15N and 13C estimates that may not reflect a true source or may be biased toward populations that produce high concentrations of Chl a. Chl a concentrations can be significantly greater at the deep chlorophyll maximum where growth rates are lower, nutrient concentrations are higher, and light limitation stimulates greater pigment production per cell (Overmann and Garcia-Pichel, 2006; Uysal, 2006).

Chl a produced by this population will likely have low 13C and 15N values, due to low

13  CDIC values and high CO2(aq) and nutrient concentrations that allow for greater isotopic fractionation (Freeman and Hayes, 1992; Waser et al., 1998). Using Chl a degradation products in sediments to characterize phytoplankton composition may also be confounded by seasonal variation in phytoplankton communities. Export production

152 of phytoplankton can be dominated by short-duration phytoplankton blooms (BS sed traps) that may skew the sedimentary record toward productivity during a particular season or by certain types of phytoplankton (Ediger et al., 2006; Oguz and Ediger, 2006).

Meromictic lakes are specialized stratified environments, as mixing of deep and surface waters is inhibited, producing two distinct water masses. The deep water mass

(monimolimnion) of a meromictic lake is prone to anoxia, and when the chemocline at the top of the anoxic layer is in the photic zone, populations of sulfide-oxidizing anoxygenic photosynthetic bacteria can develop (Overmann et al., 1991). Fayetteville

Green Lake (FGL) is a meromictic lake with an oligotrophic mixolimnion overlying a highly productive layer of anoxygenic, sulfide-oxidizing phototrophic bacteria associated with the chemocline at ~20 m water depth (Culver and Brunskill, 1969). These organisms are from two phylogenetic groups: green sulfur bacteria (GSB; Chlorobium) and purple sulfur bacteria (PSB; Chromatium, Thiocapsa, Lamprocystis) (Fry, 1986;

Meyer et al., submitted). Seasonal Synechococcus blooms dominate mixolimnion productivity and export production to the sediments (Thompson et al., 1997).

Cyanobacteria also inhabit the chemocline region (Meyer et al., submitted; Thompson et al., 1990).

Sedimentary pigment extracts from meromictic lakes contain numerous compounds produced by phototrophs living in the mixolimnion and chemocline

(Chikaraishi et al., 2007; Ohkouchi et al., 2005; Squier et al., 2002). Bacteriochlorophyll a (Bchl a) and bacteriopheophytin a (Bphe a) are produced by purple photosynthetic bacteria, including PSB. Bchl a also degrades rapidly to Bphe a which is stable in

Holocene sediments (Keely, 2006). Some PSB also produce okenone, a carotenoid that is

153 preserved well in Holocene sediments and degrades to okenane on geological time scales

(Brocks et al., 2005; Brocks and Schaeffer, 2008). Diverse species of GSB produce several different bacteriochlorophylls (Bchl), including Bchl c, Bchl d, and Bchl e. In

FGL, Bchl e produced by brown-pigmented GSB (Meyer et al., submitted) is the most abundant of these pigments in the sediments. Brown-pigmented GSB also produce the carotenoid isorenieratene. We detected isorenieratene in FGL in the water column and sediments, though at trace levels. Chl a and Bchl a are also produced in very small proportions by GSB (Madigan et al., 2003), and we assume this source is insignificant to our study. Down-core distributions of these pigments, coupled with their 13C and 15N values, can provide information about shifts in productivity of mixolimnion and chemocline populations as well as interactions between these populations via nutrient transfer.

In this paper, we describe HPLC methods for purifying photosynthetic pigments for C and N stable isotope analysis. We present down-core 15N and 13C values for

Bchl e, Bphe a, Chl a, Phe a, and okenone (13C only) extracted from FGL sediments.

Our cores recovered ~550 years of sediments and document an increase in surface water plankton productivity starting near 1780 CE (Hilfinger et al., 2001). We also present pigment concentrations determined by HPLC analysis and accumulation rates calculated using a varve chronology. To help interpret variability in sedimentary pigment distibutions, we measured water column pigment concentrations and pigment-specific C and N stable isotope compositions. As phytoplankton biomass produced in the mixolimnion strongly influences bulk sediment organic matter characteristics, it is impractical to evaluate changes in the chemocline community using bulk characteristics

154 alone. Our specific pigment data facilitate studying decadal and century-scale temporal changes in productivity and nutrient conditions below the chemocline.

5.2 Methods

5.2.1 Samples

Short gravity cores (50 cm) were collected from the center of the FGL main basin

(53 m depth) using a K-B Corer (Wildlife Supply Co., Buffalo, NY). The sediment samples examined in this study are from the deep basin composite core described in chapter 4 (Figs. 4-4, 4-7) and directly correlate with samples used for that discussion of sedimentation and bulk C and N stable isotope analysis. The sediments are divided into two units. Unit 2, deposited prior to ~1780, consists of red-brown laminated intervals interbedded with tan-brown turbidites. Unit 1 was deposited from ~1780 to the present and consists of gray-tan laminated sediments interbedded with gray-brown turbidites.

The turbidites are primarily redeposited shallow-water sediments, but they also incorporate lesser amounts of deep-water sediments. For this study, we analyzed 9 laminated intervals and 7 turbidites from Unit 2 and 11 laminated intervals and 5 turbidites from Unit 1.

Water column samples were pumped to the surface and collected on pre- combusted Whatman GF/F filters held in a stainless-steel filter housing. We collected water column particulate matter (seston) on 3 sample dates: November 4, 2005, May 9,

2006, and August 29, 2006. Seston density is greatest near the chemocline, so these

155 samples yielded more material for pigment extraction. Seston density is very low in the mixolimnion and many of the samples did not contain enough pigment for compound- specific stable-isotope analysis. The filter samples were wrapped in ashed aluminum foil, sealed in plastic bags, and transported on ice prior to returning to the lab within 12 hours of collection. All samples were stored frozen prior to pigment extraction.

We extracted pigments from sediment and filter samples using previously described methods (Airs et al., 2001a). Sediments (~5 g) or filters (cut into ~ 1.5 cm pieces) were placed in 50 ml centrifuge tubes. Pigments were extracted into acetone by:

(1) sonication for 5 minutes in ~25 ml acetone, (2) centrifugation for 5 minutes, (3) removal of supernatant by pipette, and (4) filtration through solvent-extracted cotton.

After repeating the extraction procedure 2 or 3 times, the combined extracts were dried under an N2 stream using a TurboVap LV concentration workstation (Caliper

LifeSciences). Excess water was evaporated by azeotrope with additional acetone.

During the extraction procedure, samples were shaded using aluminum foil, and the whole procedure was carried out rapidly to limit photo- and thermo-degradation. The dried extracts were stored frozen until HPLC analysis.

5.2.2 Pigment Chromatography by HPLC

Pigment extracts were resolved by reversed phase (RP)-HPLC, using a solvent gradient consisting of 0.01 M ammonium acetate, methanol, acetonitrile, and ethyl acetate (Airs et al., 2001a). During fraction collection for stable isotope analysis, we replaced ammonium acetate with water to limit C and N contamination of the isolated

156 pigment fractions. All HPLC analyses were performed on an Agilent 1200 system controlled by ChemStation Rev. B.01.03-SR 2 software. The system consists of a vacuum degasser (G1322A), quaternary gradient pump (G1311A), autosampler

(G1329A), and multiple wavelength UV-visible light detector (MWD; G1365D). RP separations were achieved on two coupled Waters Sperisorb ODS2 columns (3m, 150 mm × 4.6 mm I.D.), protected by a Waters 10 mm × 4.6 mm guard column and a

Phenomenex Security Guard C18 4 mm × 3 mm pre-column. Pigments were identified based on retention times, ionic mass/charge ratios, and fragmentation patterns determined by atmospheric pressure chemical ionization multistage mass spectrometry (APCI-MSn;

Agilent 6310 IonTrap LC/MS) in positive ion mode (Airs et al., 2001a; Airs et al.,

2001b). The corona current was 4000 nA, nebulizer pressure 60 psi, drying gas flow

5 l min-1, drying temperature 350ºC, and vaporizer temperature 400ºC.

Pigments were quantified by comparison with standard reference materials, using light absorbance characteristics. We calculated quantitative linear response factors for the MWD by injecting measured amounts of pure standards Chl a and Pphe a. We calculated response factors for Bchl e, okenone, Bphe a, and Phe a by comparison with

Chl a and Pphe a response factors, accounting for differences in the absorption properties of each compound in acetone and the eluent (Table 5-1). Because we use a RP-HPLC method with a solvent gradient, the eluent for each pigment is different. Absorbance maximum wavelengths were determined in acetone and eluent using a scanning spectrophotometer. Response ratios were determined for each compound and were used to convert published molar extinction coefficients () in acetone to the HPLC eluent

(Borrego et al., 1999; Britton, 1995; Coolen and Overmann, 1998; Watanabe et al.,

157 1984). For tetrapyrrole pigments we used abosorbance maxima in the Qy band; for okenone we used the absorbance maximum.

5.2.3 Pigment Purification and Isotope Analysis

Measurements of compound specific C and N stable isotope compositions are limited by two factors—sample size and purity. Both of these factors require careful sample-handling protocols, as working with small samples increases the likelihood of sample contamination. The sample size requirement for C and N isotopes is based on the capabilities of the method employed for isotope analysis, and helps determine which methods are practical for sample purification. We have minimized the sample size requirement to ~8 nmol N and ~100 nmol C for stable isotope analysis using recent improvements to the nanoEA-IRMS system (Polissar et al., 2009). Such small samples generally result in better than ±1‰ precision. Larger samples (up to 40 nmol N and 500 nmol C) lower this uncertainty to ~0.3‰. Stable isotope ratios are reported in delta ()

13 notation. For carbon,  C = (Rsample/Rstandard – 1) × 1000‰, where R is the ratio of the rare isotope (13C) compared to the common isotope (12C). 13C values are reported relative to standard Vienna Pee Dee Belemenite. 15N values are calculated similarly,

15 14 using ratios of N/ N and are reported relative to atmospheric N2. We calculated the uncertainty of replicate analyses of samples using the equations presented in Polissar et al. (2009). The small sample size facilitated by these methods allows us to use analytical-scale HPLC methods, affording much better chromatographic separation and

158 simpler purification methods than previous studies (Beaumont et al., 2000; Bidigare et al., 1991; Sachs and Repeta, 2000).

By using the long-duration, analytical-scale RP-HPLC method developed by Airs et al. (2001a), we are able to achieve base-line resolution for most compounds of interest.

The one exception is Bchl e, which requires an extended trapping time into adjacent peaks to ensure the entire peak is collected. After collecting peak fractions based on

MWD response at appropriate wavelengths, we further purify the fractions using normal phase (NP)-HPLC. This approach was described by Sachs and Repeta (2000); we used modifications of their approach to purify a more diverse set of compounds. These methods employ isocratic pumping of mobile phases composed of variable proportions of hexane and acetone. We use silica as the stationary phase (Agilent SIL, 5m, 250 × 4.6 mm), achieving excellent separation of Chl a, Phe a, Pphe a, Bchl e, Bphe a, and okenone from contaminating compounds that coelute with the pigments in RP. The specific NP-HPLC methods for each purified compound are outlined in Appendix C.

5.3 Results

5.3.1 Water-Column Pigments

In the FGL water column, pigment concentrations are highest near the chemocline

(Fig. 5-1). Bchl a, Bphe a, and okenone are abundant at the chemocline, produced by several genera of PSB (Fry, 1986). Of the PSB pigments, we report only okenone concentrations in the water column, as variable conversion of Bchl a to Bphe a and

159 hydroxy-Bchl a resulted in HPLC peak tailing that prevented the calculation of reliable peak areas. Okenone concentrations were always highest at the chemocline, coincident with the occurrence of maximum turbidity and organic carbon (Corg) concentrations.

Among our 3 sampling dates, the okenone concentration of 79.8 g l-1 in August 2006 was the highest, coincident with the highest Corg concentration. Lower okenone concentrations were measured in November 2005 and May 2006, 51.0 g l-1 and 34.7 g l-1, respectively. Okenone concentrations are at least one order of magnitude lower in samples collected one meter above and one meter below the chemocline.

Bchl e is the most abundant GSB pigment in the water column. Six homologs dominate Bchl e distributions, consisting of the [Et, Et], [n-Pr, Et], and [i-Bu, Et] forms esterified with farnesol or hexadecenol (Glaeser and Overmann, 2003). We detected trace concentrations of isorenieratene that could not be quantified due to coelution with

Bphe a. Both Bchl e and isorenieratene are produced exclusively by brown-colored

GSB, Chlorobium phaeobacteroides in Green Lake (Fry, 1986). The maximum measured Bchl e concentration of 50.3 g l-1 was from the chemocline in May 2006. In

August 2006, the highest concentration of Bchl e was 21.4 g l-1, also at the chemocline.

In November, 2005, the Bchl e peak was below the chemocline and significantly less concentrated at 12.0 g l-1. We also detected traces of Bchl c and Bchl d in the water column, but focused on Bchl e due to its abundance and the potential for stable isotope analysis.

Chl a and Phe a are also most abundant near the chemocline, forming a prominent deep chlorophyll maximum. Chl a concentrations were very low in the mixolimnion,

160 below 0.06 g l-1, apparently the result of generally oligotrophic conditions and relatively high light intensity. The concentration peaks for the deep maxima are generally broad, extending above and below the chemocline (Fig. 5-1). Sulfide-tolerant cyanobacteria are found near the chemocline (Thompson et al., 1990). These organisms are phylogenetically similar to cyanobacteria in the mixolimnion but appear to be actively photosynthesizing rather than sinking and senescent (Meyer et al., submitted). Chl a and

Phe a concentrations are of similar magnitude. The maximum Chl a concentration of 9.0

g l-1 was measured at the chemocline in November 2005. In May 2006, Chl a had a sharp peak (6.7 g l-1) at one meter above the chemoline. By August, the peak above the chemocline had dissipated, and the Chl a maximum of 4.0 g l-1 was at the chemocline.

Maximum concentrations of Phe a centered on the chemocline and ranged between 3.8 and 6.7 g l-1(Fig. 5-1). Phe a may derive from sinking biomass that is retained and accumulates at these depths due to the increase in water density.

5.3.2 Sedimentary Pigments

Chemocline-derived pigments dominate sedimentary pigment distributions (Fig.

5-2; Table 5-2). GSB pigments include the same 6 prominent Bchl e homologs and 2 small isorenieratene peaks observed in the water column. Additional pigments eluting between 18 and 35 minutes include Bchl c and Bchl d homologs, as well as bacteriopheophytins derived from bacteriochlorophylls. Okenone forms a distinctive peak eluting before 40 minutes. Okenone, along with Bphe a and Pbphe a record the

PSB contribution to sedimentary organic matter. Bchl a produced in the water column

161 has been converted to Bphe a and Pbphe a in the sediments. Chl a, Phe a, and Pphe a can all be quantified in Unit 1 sediment extracts. These compounds may derive from the chemocline cyanobacterial populations or from mixolimnion populations of cyanobacteria or diatoms (Thompson et al., 1997). Chl a abundance decreases down- core to trace concentrations in Unit 2, whereas Phe a and Pphe a concentrations are relatively constant (Fig. 5-3). Steryl chlorin esters, small peaks with retention times near

80 min in the 665 nm chromatogram (Fig. 5-2), are a minor component of the sedimentary pigment assemblage. Where water-column herbivory is prominent, such as in the Black Sea, steryl chlorin esters can comprise greater than 50% of Chl a derivatives in the water column and sediments (King and Repeta, 1994). For comparison, refer to figure 2-7 in Chapter 2 which depicts a Black Sea pigment chromatogram with abundant steryl chlorin esters.

5.3.3 Water-Column and Core-Top Pigment C and N Isotopes

The sediment core-top sample contains pigments deposited during the ~4 years prior to collecting water-column samples. Table 5-3 summarizes water-column and core- top pigment 13C and 15N values. All C and N isotope data collected for this study are

13 also presented in figures 5-4 and 5-5. The core-top  CChl a value of -37.2‰ is consistent with a source near the chemocline (Table 5-3, Fig. 5-4). Samples from near the chemocline in May and November have similar low 13C values. On the other hand,

13 the core-top  CPhe a value is relatively high (-31.5‰) compared with Phe a and Chl a in the chemocline. Okenone and Bphe a have similar C isotopic compositions due to their

162

13 common PSB source. The core-top  Cokenone value of -43.4‰ is similar to its chemocline values that range between -39.9‰ and -45.1‰. The range of GSB 13C

13 values is much narrower, with water-column  CBchl e values between -29.8‰ and

-30.9‰, and a surface sediments value of -30.2‰. The higher 13C values for Bchl e is expected, as GSB fix C by the reverse-TCA cycle and PSB employ the Calvin cycle

(Quandt et al., 1977; Sirevag et al., 1977).

The N isotopic compositions of Chl a and Phe a also differ in core-top and water- column samples. Chl a has a core-top 15N value of 0.5‰, similar to chemocline values of -0.6‰ and +1.3‰ (Table 5-3, Fig. 5-5). Core-top Phe a is 15N-depleted relative to Phe

15 a in the chemocline, with  NPhe a values of -2.3‰ at the sediment surface and 3.9‰ to

6.9‰ in the chemocline. The water column sample from 2 meters above the chemocline

15 in May has a  NPhe a value close to that of the core-top sample (Fig. 5-5). The core-top

Bphe a 15N value of -3.0‰ is indistinguishable from chemocline values between -3.4‰ and -3.1‰. The core-top Bchl e value of -14.0‰ is lower than those of water column samples, which ranged between -12.5‰ and -9.6‰ on our three sampling dates.

5.3.4 Down-Core Pigment N and C Isotopes

Bchl e and okenone both record relative 13C enrichment from the base of Unit 2 into Unit 1 (Fig. 5-4). The 3-4‰ increase in the 13C values of these compounds

13 13 coincides with a ~5‰ increase in  Corg.  CBphe a values do not follow the same trend as okenone, even though both are produced by PSB. However, the relatively low

163 concentrations of Bphe a in the sediments made it difficult to purify sufficient material for 13C analysis and thus the okenone 13C record is more complete and may be more reliable. Bchl e and Bphe a both record 15N enrichment from Unit 2 to Unit 1, following a similar pattern as C isotopes (Fig. 5-5). 13C and 15N values of Bchl e and Bphe a are lowest at the base of Unit 2, increase at the top of Unit 2, reach peak values in Unit 1, and decrease to the top of the core. These patterns are similar to those recorded by bulk sediments, as well. The down-core C and N isotope records of Chl a and Phe a are limited to Unit 1, as their concentrations were very low in Unit 2, especially relative to

15 the total pigment extract.  NPhe a values decrease by 2‰ from the middle of Unit 1 to

15 15 the core top, a pattern similar to those of  NBchl e and  NBphe a. Chl a, on the other hand, is 15N-enriched in the surface sediments compared with the rest of Unit 1.

5.4 Discussion

5.4.1 Sources of Sedimentary Chl a and Phe a

Surface sediment Chl a 15N and 13C values are distinct from those of Phe a, differing by +2.8‰ and -5.7‰, respectively (Table 5-3). These differences suggest that sedimentary Chl a and Phe a derive from different phytoplankton sources in FGL. Based on our examination of water column N and C isotope values, we suggest that sedimentary

Chl a is derived from cyanobacteria living near the chemocline in the deep chlorophyll maximum (Fig. 5-1, Table 5-3). Consistent with this suggestion, the surface sediment

13  CChl a value of -37.2‰ is similar to the weighted average value of -38.4‰ for five

164 water column samples collected from near and below the chemocline (Fig. 5-4). Low

13 13  CChl a values near the chemocline are a result of low  CDIC values and higher concentrations of CO2(aq), allowing for a greater isotope effect associated with C assimilation (Deevey et al., 1963; Freeman et al., 1994; Takahashi et al., 1968). These characteristics are typical for deep chlorophyll maxima, where remineralized C

15 accumulates. The surface sediment  NChl a value of 0.5‰ is also close to the average chemocline value of -0.2‰ (Fig. 5-5). Anoxia in the monimolimnion allows for better preservation of Chl a with the magnesium atom intact in the ring structure. Oxygen exposure and aerobic heterotrophy, characteristics of the mixolimnion that are absent in the monimolimnion, can facilitate the conversion of Chl a to Phe a (Keely, 2006). Thus,

Chl a produced near the chemocline is more likely to be preserved than Chl a from the mixolimnion.

Phe a is the most abundant Chl a degradation product in the water column and sediments. Chl a produced by phytoplankton in the mixolimnion appears to be degraded to Phe a in the mixolimnion, prior to sinking to the anoxic monimolimnion. In the mixolimnion, the highest concentrations of Chl a and Phe a were 0.06 g l-1 and 0.39 g l-1, suggesting Chl a does indeed degrade relatively rapidly to Phe a. Surface sedimentary Phe a 13C and 15N values are consistent with values expected for

13 mixolimnion Chl a production. The surface sediment  CPhe a value of -31.5‰ is higher

13 15 than any Phe a  C value measured in the chemocline (Fig. 5-4).  NPhe a of the surface sediment sample is also lower than values measured for seston samples. We were only able to extract sufficient quantities of Phe a for isotope analysis from seston samples

165 deeper than 10 m, so sedimentary Phe a may derive from a shallower depth in the water column.

Thompson et al. (1997) described a short-duration late spring-early summer

Synechococcus bloom in the upper mixolimnion at 4 m water depth. Synechococcus precipitate calcite, providing ballast to sink to the sediments. Calcite crystals observed in the guts of zooplankton also point toward fecal pellets as a possible mechanism for transferring biomass produced in the mixolimnion to the sediments (Brunskill, 1969).

Although we did not manage to sample the Synechococcus bloom for pigment analysis,

15 13 we did collect samples for bulk  N and  C determinations from 5 meters water depth

15 in June 2005 and May 2006. We measured seston  Nbiomass values near 5‰ and

13  Cbiomass values averaging near -30‰ for these samples. Assuming these values are representative of the Synechococcus bloom, and using published C and N isotopic differences between biomass and Chl a produced by Synechococcus (Sachs et al., 1999),

15 13 we estimate that the Synechococcus bloom would yield  NChl a and  CChl a values near

15 -5‰ and -32.5‰, respectively. These values are similar to surface sediment  NPhe a and

13  CPhe a values of -2.3‰ and -31.5‰. It appears that Chl a produced by the

Synechococcus bloom contributes significantly to Phe a preserved in the sediments.

Broadly speaking, 13C and 15N values of sedimentary porphyrins may reflect a mixture of highly productive surface-water phytoplankton and less productive, pigment- rich, deep-dwelling phototrophs. In FGL, Phe a and Chl a concentrations in surface sediments are 0.26 mg gOC-1 and 0.20 mg gOC-1, suggesting that almost half of the total

Phe a+Chl a that reaches the sediments derives from cyanobacteria living in the deep

166 chlorophyll maximum. Sedimentary Chl a ultimately is converted to Phe a, and these two compounds form one indistinguishable pool that continues to degrade to porphyrins on geological time scales. A similar relationship might hold in marine sediments, where preserved chlorophyll may derive largely from the deep chlorophyll maximum.

5.4.2 Chemocline Blooms of PSB, GSB, and Cyanobacteria

We examined temporal variability in blooms of chemocline phototrophic bacteria by examining pigment distributions in the water column. The maximum okenone concentration was measured at the chemocline in August 2006, coincident with the maximum particulate Corg concentration (Fig. 5-1). The elevated okenone concentration is apparently related to a late summer PSB bloom. The August 2006 and November 2005 okenone and particulate Corg concentrations suggest that okenone concentrations are roughly proportional to total PSB biomass in FGL, if we assume PSB comprise most of the chemocline biomass (Fig. 5-6). The minimum okenone concentration was measured in May 2006, when elevated Bchl e concentrations imply that GSB were most productive.

Mixolimnion waters are clearest from March to May (Secchi depths 14-18 m) and become increasingly turbid from June to August, when Secchi depths range from 5-8 m

(Brunskill, 1969; Thompson et al., 1997). When considered with the water clarity data, the chemocline pigment data seem to oppose the general relationship that Bchl e- producing GSB are more tolerant of low-light conditions than PSB (Overmann and

Garcia-Pichel, 2006). This contradiction can be explained by considering a third population of phototrophic bacteria near the chemocline.

167 The May 2006 seston sample from one meter above the chemocline contained high concentrations of Chl a, forming a sharp peak above the chemocline (Fig 5-1).

Seston Corg concentrations were also elevated one meter above the chemocline in May.

We interpret this peak as indicative of a bloom of deep-dwelling cyanobacteria

(Thompson et al., 1990). This third component of the chemocline phototrophic community emerges when light attenuation in the upper mixolimnion is at a minimum, in the spring when the mixolimnion is clear. Due to their tolerance of O2, the cyanobacteria are able to grow above the chemocline, where deep nutrients leak through the chemocline

(Zerkle, 2006). The deep-dwelling cyanobacteria shade the PSB that remain at the chemocline, where sulfide is available. The summer Synechococcus bloom in the mixolimnion causes mixolimnion waters to becomes turbid (Thompson et al., 1997) and the deep-dwelling cyanobacteria dissipate, as evidenced by low Chl a concentrations above the chemocline in August 2006. Bchl e-producing GSB bloom in the spring because they are well adapted to low-light conditions brought about by the deep cyanobacterial bloom. PSB bloom in the summer, when deep-dwelling cyanobacteria are inhibited and therefore, even though turbidity in the mixolimnion relatively high, light intensity reaching the chemocline is greater. A similar relationship between PSB and

GSB populations was observed in Lake Cadagno on interanual timescales, though deep- dwelling cyanobacteria were not implicated as the source of chemocline shading (Tonolla et al., 2005).

168 5.4.3 PSB and GSB Pigment Isotopes in Surface Sediments

Bphe a and okenone in the deep-basin sediments of FGL are derived from the

PSB population living at the chemocline. Based on the C and N isotopic compositions of

PSB pigments in the surface sediments, we determined that PSB biomass is exported to

15 the sediments throughout the growth season. Surface-sediment and chemocline  NBphe a

13 values are consistently near -3‰. The two chemocline  CBphe a values of -44.4‰ and

13 -40.3‰ are similar to  Cokenone values that vary between -45.1‰ and -39.9‰ (Fig. 5-4,

Table 5-3). Surface sediment Bphe a and okenone 13C values are within these ranges, presumably representing average values for chemocline production and export to the sediments. The wide range of water-column PSB pigment 13C values suggests that the turnover in this population may be relatively rapid. We have observed Daphnia with purple-colored guts, implying they consume PSB, though Fry (1986) did not find evidence for significant PSB biomass in the whole lake food web. Fecal pellets produced by Daphnia may help deliver PSB biomass to the sediments. PSB also produce intracellular sulfur granules that provide ballast to sink out of the chemocline (Overmann and Garcia-Pichel, 2006).

GSB do not have similar mechanisms to transfer biomass from the chemocline to the sediments, though sedimentary pigment distributions clearly demonstrate that GSB biomass accumulation in sediments is significant (Fig. 5-3). As opposed to the wide

13 13 range of  Cokenone values, water-column  CBchl e values are nearly constant, between

13 -29.8‰ and -30.9‰ (Table 5-3). The surface-sediment  CBchl e value of -30.2‰ is

13 15 consistent with chemocline seston  CBchl e values. The surface-sediment  NBchl e value

169 of -14.0‰, on the other hand, is 15N-depleted relative to those from all water-column samples. Among our 3 sampling dates, Bchl e concentrations were highest in May (Fig.

15 5-1). The May sample also had the lowest  NBchl e value of -11.9‰. The high

13 concentrations of Bchl e in May, combined with the lower  CBchl e values, suggest that a

GSB bloom at the chemocline in the early spring may produce most of the Bchl e that accumulates in the sediments. Alternatively, a deeper dwelling GSB population

+ 15 assimilating NH4 with a lower  N value (Chapter 4) could contribute significantly to preserved Bchl e.

5.4.4 Sedimentary Pigment 15N Values and Chemocline Productivity

This is the first study including down-core N and C stable isotope compositions of

Bphe a and Bchl e. This site was chosen because the sediments record an increase in mixolimnion productivity (Hilfinger et al., 2001) and thus we can examine the interactions between mixolimnion productivity and the monimolimnion nutrient pool.

3- Eutrophication during Unit 1 deposition, the result of increased PO4 flux to the mixolimnion, caused increased rates of organic matter accumulation in the sediments

3- (Chapter 4). We expected the transfer of PO4 to the monimolimnion to stimulate greater

3- PSB and GSB productivity, as those populations are currently limited, in part, by PO4 availability (Zerkle, 2006). Instead, accumulation rates of okenone and Bphe a decrease from Unit 2 to Unit 1 (Fig. 5-7). While pigment accumulation rates are not necessarily proportional to biomass accumulation rates, we demonstrated the general positive

170 correlation between PSB biomass and okenone in the water column (Fig. 5-6). Thus,

PSB productivity seems to have decreased from Unit 2 to Unit 1.

15 15  NBchl e and  NBphe a values increase from the base of Unit 2 to Unit 1 (Fig. 5-

5). This increase suggests a positive shift in the N isotopic composition of

+ monimolimnion NH4 . As PSB and GSB productivity decreased from Unit 2 to Unit 1,

15 15 15 in the absence of a positive shift in  NNH4+,  NBchl e and  NBphe a values would be

+ expected to decrease, as the isotope fractionation associated with NH4 assimilation would increase (Hoch et al., 1994). We hypothesize that increased productivity in the

- mixolimnion decreased isotopic fractionation between NO3 and phytoplankton biomass,

15 15 increasing both  Nbiomass and chemocline  NNH4+ values, and that the whole basin N- nutrient pool became 15N-enriched during Unit 1 deposition.

5.4.5 Sedimentary Pigment 13C Values and Deep DIC Evolution

As mixolimnion productivity increased from Unit 2 to Unit 1, we might expect the increased export of mixolimnion OM to the monimolimnion to cause decreased deep

13  CDIC values and increased concentrations of CO2(aq). Both of these changes would decrease 13C values of autotrophs growing on deep DIC (Freeman et al., 1994).

However, 13C values of sedimentary Bchl e and okenone increase to higher values in

Unit 1 compared with Unit 2 (Fig. 5-4). This finding suggests that the monimolimnion

DIC pool was more 13C-depleted during Unit 2 deposition. The relatively small

13 difference in average  Ccalcite values between Units 1 and 2, -4.7‰ and -5.0‰,

171

13 respectively attest to relatively constant mixolimnion  CDIC (Hilfinger et al., 2001). A

13 strong  CDIC gradients exists in the modern FGL water column, with values ranging from -6‰ to -9.5‰ in the mixolimnion and -13.7‰ to -21‰ in the chemocline and monimolimnion (Deevey et al., 1963; Takahashi et al., 1968; Thompson et al., 1997).

13 The gradient apparently was stronger during Unit 2 deposition, as  CDIC values at the

13 chemocline were 3-4‰ lower. It appears that in small systems, such as FGL, the  CDIC gradient may be controlled more by changes in deep water residence times rather than changes in the flux of sinking organic matter. Residence times for the FGL monimolimnion have been estimated between 4 and 35 years (Takahashi et al., 1968;

Torgersen et al., 1981). Longer residence times during Unit 2 deposition likely account

13 for the stronger  CDIC gradient proposed for that interval.

5.5 Summary

Using Fayetteville Green Lake as a model, we demonstrated that Chl a degradation products preserved in sediments may derive disproportionately from phytoplankton in the deep chlorophyll maximum. Calculations of phytoplankton 15N and 13C values based on preserved pigments may not be representative of the whole phytoplankton assemblage. We also showed that blooms of phototrophic organisms, in some cases, dominate the export flux to the sediments, and result in the preservation of a seasonal isotope signal. We also used 13C values of Bchl e and okenone in to help illustrate that increasing the sinking flux of biomass to the monimolimnion does not

172

13 necessarily increase the surface-deep  CDIC gradient. In small systems such as FGL, deep-water residence time is also an important influence on deep DIC 13C values.

5.6 Cited References

Airs, R.L., Atkinson, J.E., and Keely, B.J., 2001a, Development and application of a high resolution liquid chromatographic method for the analysis of complex pigment distributions: Journal of Chromatography A, v. 917, p. 167-177. Airs, R.L., Borrego, C.M., Garcia-Gil, J., and Keely, B.J., 2001b, Identification of the bacteriochlorophyll homologues of Chlorobium phaeobacteroides strain UdG6053 grown at low light intensity: Photosynthesis Research, v. 70, p. 221- 230. Beaumont, V.I., Jahnke, L.L., and Des Marais, D.J., 2000, Nitrogen isotopic fractionation in the synthesis of photosynthetic pigments in Rhodobacter capsulatus and Anabaena cylindrica: Organic Geochemistry, v. 31, p. 1075-1085. Bidigare, R.R., Kennicutt II, M.C., Keeney-Kennicutt, W.L., and Macko, S.A., 1991, Isolation and purification of chlorophylls a and b for the determination of stable carbon and nitrogen isotope compositions: Analytical Chemistry, v. 63, p. 130- 133. Borrego, C.M., Gerola, P.D., Miller, M., and Cox, R.P., 1999, Light intensity effects on pigment composition and organisation in the green sulfur bacterium Chlorobium tepidum: Photosynthesis Research, v. 59, p. 159-166. Britton, G., 1995, UV/visible spectroscopy, in Britton, G., Liaaen-Jensen, S., and Pfander, H., eds., Carotenoids, Volume 1B: Basel, Birkhauser Verlag, p. 13-62. Britton, G., Liaaen-Jensen, S., and Pfander, H., 2004, Carotenoids Handbook: Basel, Birkhauser Verlag. Brocks, J.J., Love, G.D., Summons, R.E., Knoll, A.H., Logan, G.A., and Bowden, S.A., 2005, Biomarker evidence for green and purple sulphur bacteria in a stratified Palaeoproterozoic sea: Nature, v. 437, p. 866-870. Brocks, J.J., and Schaeffer, P., 2008, Okenane, a biomarker for purple sulfur bacteria (Chromatiaceae), and other new carotenoid derivatives from the 1640 Ma Barney Creek Formation: Geochimica et Cosmochimica Acta, v. 72, p. 1396-1414. Broecker, W.S., and Peng, T.H., 1982, Tracers on the Sea: Palisades, NY, Lamont- Doherty Geological Observatory, p. 690. Brunskill, G.J., 1969, Fayetteville Green Lake, New York. II. Precipitation and sedimentation of calcite in a meromictic lake with laminated sediments: Limnology and Oceanography, v. 16, p. 830-847. Chicarelli, M.I., Hayes, J.M., Popp, B.N., Eckardt, C.B., and Maxwell, J.R., 1993, Carbon and nitrogen isotopic compositions of alkyl porphyrins from the Triassic Serpiano oil shale: Geochimica et Cosmochimica Acta, v. 57, p. 1308-1311.

173 Chikaraishi, Y., Kashiyama, Y., Ogawa, N.O., Kitazato, H., Satoh, M., Nomoto, S., and Ohkouchi, N., 2008, A compound-specific isotope method for measuring the stable nitrogen isotopic composition of tetrapyrroles: Organic Geochemistry, v. 39, p. 510-520. Chikaraishi, Y., Matsumoto, K., Kitazato, H., and Ohkouchi, N., 2007, Sources and transformation processes of pheopigments: Stable carbon and hydrogen isotopic evidence from Lake Haruna, Japan: Organic Geochemistry, v. 38, p. 985-1001. Coolen, M.J.L., and Overmann, J., 1998, Analysis of subfossil molecular remains of purple sulfur bacteria in a lake sediment: Applied and Environmental Microbiology, v. 64, p. 4513-4521. Culver, D.A., and Brunskill, G.J., 1969, Fayetteville Green Lake, New York. V. Sudies of primary production and zooplankton in a meromictic marl lake: Limnology and Oceanography, v. 14, p. 862-872. Deevey, E.S., Nakai, N., and Stuiver, M., 1963, Fractionation of sulfur and carbon isotopes in a meromictic lake: Science, v. 139, p. 407-408. Ediger, D., Soydemir, N., and Kideys, A.E., 2006, Estimation of phytoplankton biomass using HPLC pigment analysis in the southwestern Black Sea: Deep-Sea Research Part II, v. 53, p. 1911-1922. Freeman, K.H., and Hayes, J.M., 1992, Fractionation of carbon isotopes by phytoplankton and estimates of ancient CO2 levels: Global Biogeochemical Cycles, v. 6, p. 185-198. Freeman, K.H., Wakeham, S.G., and Hayes, J.M., 1994, Predictive isotopic biogeochemistry: Hydrocarbons from anoxic marine basins: Organic Geochemistry, v. 21, p. 629-644. Fry, B., 1986, Sources of carbon and sulfur nutrition for consumers in three meromictic lakes of New York State: Limnology and Oceanography, v. 31, p. 79-88. Glaeser, J., and Overmann, J., 2003, Characterization of in situ carbon of phototrophic consortia: Applied and Environmental Microbiology, v. 69, p. 3739- 3750. Higgins, M.B., Robinson, R.S., Casciotti, K.L., McIlvin, M.R., and Pearson, A., 2009, A method for determining the nitrogen isotopic composition of porphyrins: Analytical Chemistry, v. 81, p. 184-192. Hilfinger, M.F., Mullins, H.T., Burnett, A., and Kirby, M.E., 2001, A 2500 year sediment record from Fayetteville Green Lake, New York: evidence for anthropogenic impacts and historic isotope shift: Journal of Paleolimnology, v. 26, p. 293-305. Hoch, M.P., Fogel, M.L., and Kirchman, D.L., 1994, Isotope fractionation during ammonium uptake by marine microbial assemblages: Geomicrobiology Journal, v. 12, p. 113-127. Huang, Y., Freeman, K.H., Wilken, R.T., Arthur, M.A., and Jones, A.D., 2000, Black Sea chemocline oscillations during the Holocene: molecular and isotopic studies of marginal sediments: Organic Geochemistry, v. 31, p. 1525-1531. Kashiyama, Y., Ogawa, N.O., Shiro, M., Tada, R., Kitazato, H., and Ohkouchi, N., 2008, Reconstruction of the biogeochemistry and ecology of photoautotrophs based on the nitrogen and carbon isotopic compositions of vanadyl porphyrins from Miocene siliceous sediments: Biogeosciences, v. 5, p. 797-816.

174 Keely, B.J., 2006, Geochemistry of chlorophylls, in Grimm, B., Porra, R.J., Rudiger, W., and Scheer, H., eds., Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications: The Netherlands, Springer, p. 535-561. King, L.L., and Repeta, D.J., 1994, Phorbin steryl esters in Black Sea sediment traps and sediments: A preliminary evaluation of their paleooceanographic potential: Geochimica et Cosmochimica Acta, v. 58, p. 4389-4399. Madigan, M.T., Martinko, J.M., and Parker, J., 2003, Brock Biology of Microorganisms: Upper Saddle River, Prentice Hall, 1019 p. Meyer, K.M., Kump, L.R., Macalady, J.L., Schaperdoth, I., Fulton, J.M., and Freeman, K.H., submitted, Benthic production of the sulfur phototroph biomarker okenone: Geobiology. Oguz, T., and Ediger, D., 2006, Comparison of in situ and sattelite-derived chlorophyll pigment concentrations, and impact of phytoplankton bloom on the suboxic layer structure in the western Black Sea during May-June 2001: Deep-Sea Research Part II, v. 53, p. 1923-1933. Ohkouchi, N., Kashiyama, Y., Kuroda, J., Ogawa, N.O., and Kitazato, H., 2006, The importance of diazotrophic cyanobacteria as primary producers during Cretaceous Oceanic Anoxic Event 2: Biogeosciences, v. 3, p. 467-478. Ohkouchi, N., Nakajima, Y., Ogawa, N.O., Chikaraishi, Y., Suga, H., Sakai, S., and Kitazato, H., 2008, Carbon isotopic composition of the tetrapyrrole nucleus in chloropigments from a saline meromictic lake: A mechanistic view for interpreting the isotopic signature of alkyl porphyrins in geological samples: Organic Geochemistry, v. 39, p. 521-531. Ohkouchi, N., Nakajima, Y., Okada, H., Ogawa, N.O., Suga, H., Oguri, K., and Kitazato, H., 2005, Biogeochemical processes on the saline meromictic Lake Kaiike, Japan: implications from molecular isotopic evidences of photosynthetic pigments: Environmental Microbiology, v. 7, p. 1009-1016. Overmann, J., Beatty, J.T., Hall, K.J., Pfennig, N., and Northcote, T.G., 1991, Characterization of a dense, purple sulfur bacterial layer in a meromictic salt lake: Limnology and Oceanography, v. 36, p. 846-859. Overmann, J., and Garcia-Pichel, F., 2006, The phototrophic way of life, in Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H., and Stackebrandt, E., eds., The Prokaryotes, Volume 2: Dordrecht, Springer, p. 32-85. Polissar, P.J., Fulton, J.M., Junium, C.K., Turich, C.C., and Freeman, K.H., 2009, Measurement of 13C and 15N isotopic composition on nanomolar quantities of C and N: Analytical Chemistry, v. 81, p. 755-763. Qian, Y., Kennicutt, M.C., Svalberg, J., Macko, S.A., Bidigare, R.R., and Walker, J., 1996, Suspended particulate organic matter (SPOM) in Gulf of Mexico estuaries: Compound-specific isotope analysis and plant pigment compositions: Organic Geochemistry, v. 24, p. 875-888. Quandt, L., Gottschalk, H., Ziegler, H., and Stichler, W., 1977, Isotope discrimination by photosynthetic bacteria: FEMS Microbiology Letters, v. 1, p. 125-128. Sachs, J.P., and Repeta, D.J., 1999, Oligotrophy and nitrogen fixation during eastern Mediterranean sapropel events: Science, v. 286, p. 2485-2488.

175 —, 2000, The purification of chlorins from marine particles and sediments for nitrogen and carbon isotopic analysis: Organic Geochemistry, v. 31, p. 317-329. Sachs, J.P., Repeta, D.J., and Goericke, R., 1999, Nitrogen and carbon isotopic ratios of chlorophyll from marine phytoplankton: Geochimica et Cosmochimica Acta, v. 63, p. 1431-1441. Sirevag, R., Buchanan, B.B., Berry, J.A., and Troughton, J.H., 1977, Mechanisms of CO2 fixation in bacterial photosynthesis studied by the carbon isotope fractionation technique: Archives of Microbiology, v. 112, p. 35-38. Squier, A.H., Hodgson, D.A., and Keely, B.J., 2002, Sedimentary pigments as markers for environmental change in an Antarctic lake: Organic Geochemistry, v. 33, p. 1655-1665. Takahashi, T., Broecker, W., Li, Y.H., and Thurber, D., 1968, Chemical and isotopic balances for a meromictic lake: Limnology and Oceanography, v. 13, p. 272-292. Thompson, J.B., Ferris, F.G., and Smith, D.A., 1990, Geomicrobiology and sedimentology of the mixolimnion and chemocline in Fayetteville Green Lake, New York: Palaios, v. 5, p. 52-75. Thompson, J.B., Schultz-Lam, S., Beveridge, T.J., and Des Marais, D.J., 1997, Whiting events: Biogenic origin due to the photosynthetic activity of cyanobacterial picoplankton: Limnology and Oceanography, v. 42, p. 133-141. Tonolla, M., Peduzzi, R., and Hahn, D., 2005, Long-term population dynamics of phototrophic sulfur bacteria in the chemocline of Lake Cadagno, Switzerland: Applied and Environmental Microbiology, v. 71, p. 3544-3550. Torgersen, T., Hammond, D.E., Clarke, W.B., and Peng, T.-H., 1981, Fayetteville Green Lake, New York: 3H-3He water mass ages and secondary chemical structure: Limnology and Oceanography, v. 26, p. 110-122. Tyrrell, T., 1999, The relative influences of nitrogen and phosphorus on oceanic primary production: Nature, v. 400, p. 525-531. Uysal, Z., 2006, Vertical distribution of marine cyanobacteria Synechococcus spp. in the Black, Marmara, Aegean, and eastern Mediterranean sea: Deep-Sea Research Part II, v. 53, p. 1976-1987. Waser, N.A.D., Harrison, P.J., Nielsen, B., Calvert, S.E., and Turpin, D.H., 1998, Nitrogen isotope fractionation during the uptake and assimilation of nitrate, nitrite, ammonium, and urea by a marine diatom: Limnology and Oceanography, v. 43, p. 215-224. Watanabe, T., Hongu, A., Honda, K., Nakazato, M., Konno, M., and Saitoh, S., 1984, Preparation of chlorophylls and pheophytins by isocratic liquid-chromatography: Analytical Chemistry, v. 56, p. 251-256. York, J.K., Tomasky, G., Valiela, I., and Repeta, D.J., 2007, Stable isotopic detection of ammonium and nitrate assimilation by phytoplankton in the Waquoit Bay estuarine system: Limnology and Oceanography, v. 52, p. 144-155. Zerkle, A.L., 2006, Microbial Trace Metal Requirements: Limiting Nutrients and Potential Biosignatures [Ph. D. thesis]: University Park, PA, The Pennsylvania State University.

176

Figure 5-1. Water column pigment concentrations. The three sampling dates yielded chemocline depths between 19.8 and 20.5 m. The chemocline sample depth is set at 0 m for each sampling date, and other samples are plotted relative to the chemocline.

177

Figure 5-2. Example pigment chromatogram from Unit 1. The numbered pigments are listed in Table 5-2.

178

Figure 5-3. Sedimentary pigment concentrations. Squares connected by lines represent laminated sediment intervals. Diamonds are turbidite samples. This chapter focuses on laminated sediments; for a discussion of turbidites please refer to chapter 4. The horizontal dashed lines mark the boundary between Unit 1 (above) and Unit 2 (below).

179

Figure 5-4. Pigment-specific C isotope profiles. The upper panels are water column samples and the lower panels are sediment samples. Dashed vertical lines project surface sediment values into the water column and down-core. The horizontal dashed lines mark the depth of the chemocline (bacterial plate) and the age of the Unit 1-2 transition.

180

Figure 5-5. Pigment-specific N isotope profiles. The upper panels are water column samples and the lower panels are sediment samples. Dashed vertical lines project surface sediment values into the water column and down-core. The horizontal dashed lines mark the depth of the chemocline (bacterial plate) and the age of the Unit 1-2 transition.

181

Figure 5-6. Relationships between pigments and particulate Corg in the chemocline. The maximum Corg concentration in August 2006 coincides with the maximum okenone concentration. The May 2006 sample contains high Bchl e concentrations and low okenone concentrations, consistent with the proposed spring bloom of GSB and inhibited growth of PSB. Gravity separation of chemocline particulate matter from May suggests that PSB comprise 40-50% of total chemocline biomass (Chapter 4 Table 4-1). Using this value, we project an okenone/PSB biomass value that is plotted as an open diamond.

182

Figure 5-7. Sedimentary pigment accumulation rates. Accumulation rates were calculated for sediment intervals consisting of continuous laminations. The horizontal dashed lines mark the boundary between Unit 1 (above) and Unit 2 (below).

183 Table 5-1. Pigment extinction coefficients

extinction Biological  in coefficient source in a c tR eluant composition eluant () reference FGL -1 -1 -1 -1 (min) 1 : 2 : 3 : 4 mM cm mM cm Bchl e 15 4.7 : 76.9 : 15.0 : 3.4 34.7 48.9 (Borrego et al., 1999) GSB Chl a 35 3.4 : 61.1 : 15.0 : 3.4 75.3 81.3 (Watanabe et al., 1984) Cyano. okenone 40 3.7 : 64.2 : 15.0 : 17.1 110.0 134.1 (Britton et al., 1995) PSB Bphe a 50 2.9 : 54.7 : 15.0 : 27.4 78.0 81.2 (Coolen and Overmann, 98) PSB Phe a 58 2.5 : 49.7 : 15.0 : 32.8 43.0 46.0 (Watanabe et al., 1984) Cyano. b Pphe a 66 2.1 : 44.6 : 15.0 : 38.3 44.8 45.7 Cyano. a solvents: 1-ammonium acetatate, 2-methanol, 3-acetonitrile, 4-ethyl acetate b  for Pphe a in acetone was calculated in this study using pure standard material c GSB (green sulfur bacteria), PSB (purple sulfur bacteria), cyano (cyanobacteria)

184 Table 5-2. Pigments identified by HPLC-MSn. Peak numbers correlate with Figure 5-2.

tR Peak # (min) Assignment Esterifying alcohol [M+H]+ fragment loss 1 13.9 Bchl e [Et,Et] farnesol 799 595 204 2 15.1 Bchl e [n-Pr,Et] farnesol 813 609 204 3 16.3 Bchl e [i-Bu,Et] farnesol 827 623 204 4 22.1 Bchl e [Et,Et] hexadecenol 839 5 23.4 Bchl e [n-Pr,Et] hexadecenol 853 6 25.0 Bchl e [i-Bu,Et] hexadecenol 867 7 37.4 okenone 579 547 32 8 38.1 okenone 579 547 32 9 38.9 okenone 579 547 32 10 41.6 chlorophyll a phytol 893 615 278 11 50.0 isorenieratene 529 12 51.1 Bphe a phytol 889 611 278 13 52.3 isorenieratene 529 14 54.1 Bphe a isomer phytol 889 611 278 15 58.6 Pbphe a phytol 831 552 279 16 60.9 Phe a phytol 871 593 278 17 63.1 Phe a isomer phytol 871 593 278 18 68.2 Pphe a phytol 813 535 278

185

Table 5-3. Water-column and surface-sediment pigment-specific C and N stable isotope compositions.

Appendix A

Chapter 3 Supplementary Material

Pigment Structures

187 Carbon and nitrogen data tables for Black Sea sediments. Weight percent compositions are based on decarbonated samples.

GGC 01 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 200 1 2.61 0.29 10.4 3.3 -25.0 337 6 3.06 0.33 10.8 2.7 -24.7 475 11 3.46 0.34 11.7 2.4 -24.3 778 22 3.18 0.32 11.5 2.0 -24.5 998 30 3.88 0.37 12.3 1.7 -24.3 1219 38 3.25 0.32 12.0 2.1 -24.5 1356 43 4.00 0.36 12.9 2.0 -24.3 1549 50 2.78 0.25 12.8 2.6 -25.0 1659 54 2.65 0.26 12.1 2.7 -25.4 1797 59 2.61 0.27 11.4 3.0 1907 63 2.81 0.27 12.3 3.1 -25.4 1990 66 3.92 2.5 -24.9 2017 67 3.44 0.34 11.8 2.1 -25.1 2225 72 3.93 0.35 13.0 1.8 -24.5 2288 73 4.30 0.42 12.0 1.9 -24.1 2663 79 3.72 0.37 11.8 1.8 -24.9 2913 83 4.39 0.40 12.9 1.3 -25.2 3288 89 4.21 0.40 12.3 1.2 -25.1 3663 95 4.47 0.42 12.5 1.5 -24.6 3913 99 4.43 0.42 12.2 1.1 -24.4 4038 101 4.80 0.47 12.0 1.2 -24.8 4163 103 4.26 0.42 11.7 1.1 -24.4 4288 105 4.26 0.42 11.7 1.2 -24.6 4538 109 4.45 0.42 12.5 1.3 -24.9 4663 111 4.23 0.39 12.6 1.2 -25.3 4788 113 3.88 0.36 12.6 1.3 -25.6 4913 115 4.25 0.38 13.2 1.7 -25.4 5038 117 3.98 0.38 12.3 2.6 -25.5 5159 119 5.69 0.52 12.8 2.5 -25.2 5281 121 5.91 0.51 13.5 1.8 -25.4 5402 123 6.33 0.55 13.3 1.5 -27.2 5524 125 7.41 0.59 14.7 2.1 -26.0 5645 127 5.93 0.48 14.6 2.1 -24.2 5766 129 7.65 0.63 14.2 2.4 -23.7 5888 131 7.54 0.59 14.9 2.2 -23.4 6009 133 8.19 0.67 14.4 2.1 -24.6 6131 135 8.67 0.72 14.0 2.4 -24.1 6252 137 7.51 0.62 14.2 3.2 -23.0 6373 139 6.56 0.55 13.9 3.2 -25.8 6495 141 5.88 0.49 14.0 3.3 -25.4 6616 143 6.70 0.55 14.3 3.6 -25.4 6738 145 7.10 0.57 14.5 2.9 -25.0

188

6868 147 5.44 0.44 14.4 2.6 -24.9 7011 149 6.58 0.47 16.5 1.8 -26.5 7296 153 7.19 0.50 16.8 0.6 -26.4 7582 157 6.84 0.53 15.2 1.4 -26.0 7653 158 6.53 0.51 14.8 1.5 -28.9 7725 159 7.65 0.55 16.2 1.2 -26.4 161 2.26 0.20 12.9 3.3 -27.6 163 2.48 0.24 12.2 3.3 -27.8 165 2.67 0.25 12.5 3.3 -28.3 167 2.69 0.25 12.4 3.5 -29.2 169 2.44 0.24 11.9 3.7 -29.3 173 2.55 0.25 12.1 3.7 -28.7 177 2.19 0.20 12.6 3.3 -28.2

GGC 69 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 1850 12 9.38 0.73 14.9 1.9 -24.5 16 3.55 0.24 17.2 2.5 -22.8 21 3.80 0.29 15.1 2.2 -23.6 2000 23.5 6.78 0.52 15.1 1.2 -24.0 2200 24 16.34 1.30 14.6 1.0 -23.8 2400 25 11.59 1.00 13.6 1.7 -24.1 2750 26 9.90 0.88 13.1 1.5 -24.6 3200 27 8.41 0.75 13.1 1.2 -25.1 3600 28 8.20 0.62 15.4 1.0 -24.4 3913 29 9.40 0.81 13.5 0.5 -25.1 4053 30 7.48 0.69 12.7 1.0 -24.6 4194 31 7.10 0.64 13.0 1.2 -24.8 4334 32 9.19 0.85 12.6 0.8 -24.8 4475 33 5.02 0.47 12.6 0.9 -24.5 4616 34 9.19 0.85 12.6 0.9 -24.8 4756 35 9.16 0.82 13.1 1.2 -25.1 4897 36 10.02 0.86 13.7 1.4 -25.6 5038 37 11.75 1.01 13.5 3.0 -25.7 5147 38 11.88 1.00 13.8 2.7 -24.8 5256 39 13.09 1.09 14.0 1.9 -25.1 5366 40 15.57 1.28 14.2 1.2 -24.8 5475 41 17.03 1.31 15.1 1.9 -23.5 5584 42 17.27 1.28 15.8 2.2 -23.1 5693 43 17.50 1.22 16.7 2.2 -23.5 5803 44 16.16 1.21 15.5 2.0 -23.2 5912 45 17.51 1.26 16.2 1.8 -23.1 6021 46 18.39 1.35 15.9 1.8 -23.2 6131 47 21.30 1.60 15.6 2.1 -23.5 6240 48 17.92 1.36 15.3 2.5 -23.7 6349 49 19.15 1.43 15.7 3.1 -22.7 6458 50 11.70 0.89 15.3 3.4 -25.1 6568 51 12.23 0.92 15.4 3.6 -25.2

189

6677 52 16.48 1.18 16.4 3.1 -25.3 6786 53 15.00 1.07 16.3 4.5 -25.6 6911 54 14.90 1.09 15.9 3.3 -25.0 7039 55 16.50 1.18 16.3 2.4 -24.3 7168 56 12.70 0.90 16.5 2.3 -25.6 7296 57 12.40 0.85 17.0 1.4 -26.5 7360 58 15.10 1.09 16.2 1.8 -26.5 7423 59 16.60 1.11 17.4 1.9 -26.6 7487 60 15.90 1.02 18.1 2.1 -26.8 7550 61 14.50 0.92 18.3 2.2 -26.3 7614 62 7.12 0.52 15.9 2.5 -26.0 7677 63 5.34 0.42 14.7 2.8 -26.0 7741 64 4.32 0.34 14.8 3.0 -25.5 7804 65 3.29 0.21 17.9 3.9 -23.6 66 1.41 0.12 13.5 3.9 -23.9 67 1.01 0.11 11.2 3.7 -24.1 68 1.16 0.11 12.2 4.1 -23.8

GGC 48 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 577 2 1.61 0.13 15.0 1.5 -25.3 640 5 1.70 0.13 15.3 1.4 -25.4 767 11 1.21 0.07 18.8 2.1 -27.7 893 17 1.33 0.10 15.4 1.9 -25.2 998 22 1.99 0.15 15.9 1.0 -25.7 1082 26 1.19 0.10 14.2 2.3 -24.9 1167 30 0.82 0.06 17.3 2.1 -28.4 1272 35 1.92 0.14 16.0 1.3 -24.8 1356 39 1.67 0.13 15.5 1.3 -25.1 1421 43 1.64 0.11 17.1 1.4 -25.2 1534 50 1.62 0.10 19.2 2.2 -25.8 1598 54 1.46 0.09 18.2 2.3 -25.8 1695 60 1.60 0.10 18.1 1.9 -25.6 1759 64 1.43 0.10 16.9 2.2 -26.0 1872 71 1.43 0.09 18.2 2.0 -25.8 1937 75 1.29 0.09 16.2 2.1 -25.9 2017 80 1.50 0.10 17.5 1.4 -25.5 2163 85 1.55 0.11 16.1 1.5 -25.2 2413 90 2.01 0.15 15.6 1.2 -24.8 2663 95 1.42 0.11 15.2 1.5 -25.8 2913 100 1.75 0.13 15.8 1.1 -25.5 3163 105 2.00 0.15 15.7 1.2 -25.6 3413 110 1.97 0.15 15.6 0.6 -25.7 3663 115 1.65 0.12 16.0 1.0 -25.6 3912 120 2.49 0.19 15.6 0.3 -24.9 4035 125 2.10 0.17 14.5 0.4 -24.7 4157 130 1.90 0.15 14.9 0.6 -24.8 4279 135 2.10 0.17 14.5 0.4 -24.9

190

4377 139 2.19 0.17 15.0 -0.1 -24.9 4524 145 1.82 0.13 15.8 0.6 -23.3 4646 150 2.35 0.16 16.8 0.4 -24.3 4768 155 1.70 0.12 16.2 1.1 -24.6 4891 160 1.98 0.14 16.1 0.8 -26.0 5038 166 2.98 0.27 13.0 2.5 -25.2 5156 170 2.50 0.20 14.9 2.1 -25.0 5304 175 2.91 0.22 15.4 1.3 -25.8 5451 180 3.09 0.22 16.5 1.4 -23.0 5894 195 3.43 0.23 17.4 1.5 -23.0 6042 200 3.18 0.23 16.4 2.3 -23.3 6190 205 2.42 0.17 17.0 3.2 -24.9 6337 210 2.82 0.16 20.1 2.7 -24.8 6485 215 1.75 0.13 15.8 2.5 -25.5 6633 220 2.90 0.16 20.7 2.5 -24.3 6949 230 3.42 0.18 21.9 1.7 -24.1 7123 235 2.38 0.15 18.7 1.1 -25.7 7296 240 2.68 0.19 16.4 0.7 -25.9 7353 245 2.74 0.19 16.6 0.4 -26.1 7410 250 2.90 0.20 16.7 0.6 -26.1 7468 255 3.07 0.21 17.2 0.5 -26.4 7525 260 1.92 0.11 19.7 1.0 -26.5 7582 265 2.04 0.15 16.1 1.1 -25.8 7639 270 2.37 0.18 15.6 1.0 -26.6 7696 275 1.87 0.15 15.0 1.2 -26.9 280 0.87 0.07 14.4 1.2 -26.2 290 1.49 0.13 13.8 3.9 -24.7 300 1.48 0.14 12.6 4.0 -24.3 310 0.53 0.07 9.2 3.9 -25.2 320 0.54 0.07 8.6 4.7 -24.9 330 0.58 0.08 8.8 4.2 -25.2 350 0.42 0.07 7.3 4.1 -25.6 370 0.50 0.08 7.8 4.5 -26.1 390 0.48 0.07 7.9 4.5 -25.4 410 0.52 0.07 8.4 4.2 -25.3

GGC 59 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 1130 12 2.49 0.24 12.3 2.3 -24.4 1356 24 2.66 0.23 13.5 1.2 -24.0 1545 34 1.84 0.17 12.5 2.0 -25.1 1828 49 1.66 0.15 12.8 2.5 -23.7 2017 59 6.06 0.58 12.1 0.0 -24.4 2276 65 3.16 0.26 14.2 1.9 -24.6 2519 69 4.67 0.40 13.6 1.1 -24.4 2822 74 5.25 0.45 13.6 1.2 -24.8 3125 79 4.53 0.39 13.6 1.1 -24.9 3246 81 6.10 0.50 14.1 0.3 -24.6

191

3428 84 5.49 0.44 14.6 0.3 -24.7 3549 86 3.38 0.29 13.6 0.7 -24.8 3731 89 5.27 0.42 14.6 0.3 -25.0 3913 92 6.25 0.52 14.0 0.2 -24.8 4079 96 3.82 0.34 13.1 0.5 -24.4 4204 99 4.60 0.38 14.2 0.3 -24.2 4288 101 4.63 0.41 13.2 0.3 -24.5 4454 105 6.79 0.59 13.4 0.2 -24.4 4579 108 5.41 0.48 13.1 0.3 -24.4 4746 112 4.74 0.43 12.9 0.3 -24.6 5037 119 8.17 0.65 14.7 2.2 -25.0 5248 124 7.20 0.58 14.5 1.7 -25.2 5416 128 6.89 0.60 13.4 1.6 -24.0 5500 130 1.9 5584 132 8.43 0.69 14.3 2.0 -23.3 5752 136 12.27 0.89 16.1 1.9 -22.9 5878 139 11.32 0.95 13.9 2.0 -23.5 5962 141 11.22 0.89 14.7 2.2 -24.4 6130 145 9.24 0.71 15.2 3.3 -24.2 6256 148 6.52 0.54 14.0 3.6 -25.0 6383 151 6.62 0.52 14.8 3.2 -25.1 6509 154 9.79 0.71 16.0 2.7 -25.7 6635 157 9.29 0.69 15.6 4.3 -25.8 6761 160 6.77 0.52 15.3 3.6 -26.0 6851 162 7.92 0.61 15.2 3.0 -24.7 6999 165 8.71 0.57 17.8 2.3 -25.0 7098 167 6.71 0.48 16.4 1.7 -25.7 7197 169 5.73 0.41 16.1 1.2 -26.9 7297 171 5.47 0.39 16.6 0.7 -25.5 7452 174 5.44 0.47 13.5 0.9 -26.5 7504 175 4.48 0.34 15.6 1.2 -25.9 7556 176 6.64 0.42 18.6 1.0 -26.1 7608 177 4.04 0.30 15.7 1.3 -26.0 7660 178 5.56 0.39 16.8 1.0 -25.9 7712 179 5.04 0.33 17.7 1.9 -25.8 7764 180 6.41 0.41 18.4 1.3 -26.0 7816 181 7.98 0.54 17.4 0.7 -26.0 7868 182 5.01 0.35 16.9 1.9 -26.2 183 1.78 0.15 13.6 3.2 -27.2 189 1.62 0.16 11.7 2.9 -26.7 200 1.88 0.13 17.4 3.6 -26.6 210 1.60 0.16 11.8 3.3 -26.8 230 1.00 0.10 12.0 3.9 -26.6 240 1.08 0.10 12.2 4.2 -27.2 260 0.80 0.08 11.4 3.9 -26.1 270 1.06 0.10 12.4 4.3 -26.2 300 0.82 0.10 9.7 5.2 -27.0 320 0.73 0.08 10.0 4.5 -24.2 350 0.85 0.11 9.0 4.3 -25.3

192

370 0.93 0.09 11.7 4.5 -25.1 380 0.89 0.09 11.5 4.4 -25.8 401 0.91 0.10 11.1 4.8 -26.4 430 0.92 0.13 8.5 4.9 -25.2 450 1.31 0.09 16.3 4.8 -23.2

GGC 20 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 3069 5 12.57 1.01 14.5 1.5 -25.5 3350 6 17.52 1.45 14.1 1.0 -24.3 3631 7 17.84 1.56 13.4 0.7 -25.6 3913 8 17.64 1.50 13.7 0.2 -25.3 4194 9 16.32 1.52 12.5 0.4 -24.6 4475 10 16.38 1.49 12.8 0.2 -25.0 4756 11 17.81 1.48 14.0 0.8 -26.8 5038 12 22.48 1.83 14.3 2.1 -25.0 5256 13 22.51 1.82 14.4 1.2 -24.4 5475 14 23.64 1.78 15.5 1.8 -23.5 5693 15 26.49 1.93 16.0 1.8 -25.0 5912 16 28.57 2.13 15.6 1.9 -24.4 6131 17 23.51 1.76 15.6 2.8 -24.1 6349 18 23.68 1.65 16.7 2.8 -26.4 6568 19 25.39 1.71 17.3 4.1 -25.7 6786 20 24.82 1.70 17.0 2.5 -24.7 7039 21 20.92 1.44 17.0 1.6 -25.9 7296 22 20.59 1.42 16.9 0.5 -26.8 7378 23 21.83 1.48 17.3 1.2 -27.2 7460 24 20.57 1.35 17.8 1.7 -26.5 7541 25 10.56 0.71 17.2 2.2 -26.6 7623 26 6.75 0.51 15.5 2.5 -26.6 7704 27 6.26 0.49 14.8 2.6 -27.1 7786 28 4.84 0.40 14.2 2.8 -27.1 7868 29 3.75 0.35 12.6 3.3 -26.3 30 3.21 0.32 11.8 3.3 -26.2 31 2.78 0.29 11.2 3.4 -26.4 32 2.61 0.28 10.8 3.5 -26.3 33 2.62 0.28 11.0 3.2 -26.4 34 2.48 0.28 10.4 3.2 -26.6 35 3.28 0.32 11.8 3.2 -26.5 36 3.08 0.30 11.8 3.5 -26.9 37 2.58 0.24 12.6 3.7 -26.4 38 1.84 0.14 14.9 3.7 -19.2 39 1.75 0.11 19.0 2.4 -14.9 40 2.29 0.16 16.3 3.6 -20.0 41 2.30 0.19 14.1 3.8 -27.7 42 3.40 0.27 14.6 3.9 -27.3 43 3.91 0.33 14.0 4.1 -29.8 44 3.88 0.31 14.6 4.1 -29.0

193

49 2.98 0.23 15.1 3.8 -29.4 64 1.44 0.12 14.0 5.0 -27.3 74 1.78 0.14 15.3 5.9 -26.8 83 0.66 0.08 10.0 5.9 -31.9 89 1.07 0.10 12.7 5.4 -26.2 97 0.75 0.08 10.4 4.8 -25.2 109 0.47 0.06 8.8 5.6 -27.2 128 0.82 0.07 13.9 4.1 -24.5

GGC 09 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 1386 17 6.41 0.56 13.3 1.6 -24.6 1638 19 4.65 0.43 12.5 2.2 -25.2 1890 21 5.34 0.46 13.4 1.9 -24.7 2016 22 6.51 0.56 13.6 1.2 -23.6 2480 24 7.75 0.69 13.2 1.1 -25.2 2766 25 7.56 0.70 12.6 0.9 -24.0 3337 27 7.19 0.65 12.9 1.1 -24.7 3913 29 6.93 0.62 12.9 0.5 -24.6 4100 31 7.44 0.70 12.3 0.4 -24.3 4288 33 6.84 0.64 12.5 0.4 -24.3 4381 34 7.76 0.74 12.3 0.3 -24.2 4663 37 7.97 0.74 12.6 0.5 -24.7 4850 39 6.90 0.61 13.1 0.9 -26.1 5037 41 8.49 0.75 13.2 2.2 -24.7 5236 43 9.95 0.88 13.1 1.3 -24.7 5435 45 12.08 0.97 14.6 1.8 -23.4 5633 47 9.88 0.77 14.9 2.2 -22.7 5832 49 14.41 1.06 15.8 1.7 -22.6 6031 51 14.56 1.14 14.9 1.9 -22.7 6229 53 12.49 0.99 14.7 2.5 -24.6 6428 55 9.74 0.75 15.1 3.1 -24.4 6627 57 11.55 0.83 16.2 2.8 -25.5 6825 59 10.86 0.77 16.4 2.2 -24.4 7062 61 10.13 0.73 16.2 1.5 -26.3 7296 63 10.19 0.74 16.1 1.1 -25.6 7582 65 10.51 0.70 17.6 1.3 -28.0 67 2.56 0.22 13.3 3.5 -26.4 69 2.57 0.21 14.3 3.8 -26.5

GGC 71 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 1238 0 2.03 0.21 11.1 2.6 -24.5 1355 5.2 1.83 0.17 12.4 1.8 -24.4 1470 10.3 1.52 0.16 11.3 3.1 -24.8 1588 15.5 1.52 0.15 11.8 2.9 -25.6 1667 19 1.75 0.16 12.6 2.6 -25.2 1820 25.8 1.67 0.17 11.8 2.6 -25.2

194

1938 31 1.72 0.19 10.8 3.7 -24.8 2017 34.5 1.91 0.18 12.3 2.0 -24.9 2107 39.6 2.46 0.23 12.7 2.4 -24.3 2323 44.8 2.51 0.25 11.7 1.9 -24.2 2531 50 2.13 0.22 11.2 2.5 -24.6 2735 55.1 1.89 0.19 11.8 2.1 -25.4 2943 60.3 2.12 0.21 11.9 2.4 -25.0 3151 65.5 1.76 0.17 11.9 1.3 -25.1 3355 70.6 2.39 0.24 11.7 1.7 -24.4 3631 77.5 2.07 0.20 12.4 1.5 -24.2 3699 79.2 1.86 0.21 10.2 1.9 -25.1 3839 82.7 2.77 0.24 13.5 1.5 -24.1 3899 84.4 2.61 0.26 11.7 1.6 -24.4 3973 86.1 2.95 0.29 11.9 2.3 -24.5 4052 87.9 2.99 0.25 13.9 1.1 -24.5 4126 89.6 2.57 0.28 10.8 1.4 -24.5 4201 91.3 3.02 0.24 14.7 1.1 -24.5 4275 93 4.65 0.37 14.7 0.9 -24.5 4354 94.8 2.92 0.23 14.6 1.1 -24.3 4428 96.5 3.30 0.26 14.8 1.6 -25.0 4503 98.2 3.27 0.26 14.6 1.4 -24.5 4577 99.9 3.22 0.25 14.9 1.0 -24.1 4651 101.6 2.55 0.27 11.2 1.6 -24.9 4730 103.4 3.24 0.24 15.8 1.2 -23.9 4804 105.1 4.03 0.27 17.4 1.2 -24.6 4879 106.8 3.20 0.26 14.3 1.3 -24.3 4953 108.5 3.75 0.26 17.0 1.2 -24.2 5037 110.3 3.41 0.32 12.3 2.5 -25.6 5104 112 4.78 0.35 15.8 2.4 -25.1 5167 113.7 3.09 0.23 15.9 1.5 -24.5 5232 115.4 3.93 0.33 13.7 2.5 -24.7 5301 117.2 3.96 0.30 15.3 2.8 -25.2 5366 118.9 4.45 0.35 14.8 2.3 -24.8 5431 120.6 4.36 0.41 12.4 2.4 -24.9 5496 122.3 5.05 0.41 14.3 1.7 -25.3 5561 124 4.85 0.40 14.2 1.8 -24.8 5630 125.8 5.78 0.50 13.4 1.8 -24.1 5695 127.5 6.68 0.49 15.9 2.4 -23.2 5760 129.2 6.15 0.45 16.1 2.4 -23.6 5825 130.9 5.21 0.40 15.3 2.3 -23.6 5894 132.7 6.06 0.50 14.2 2.5 -23.2 5959 134.4 6.37 0.52 14.3 2.2 -23.1 6024 136.1 6.64 0.51 15.3 2.2 -23.2 6089 137.8 6.73 0.46 17.0 2.1 -23.3 6154 139.5 7.32 0.59 14.4 2.1 -23.3 6288 143 7.50 0.59 14.9 2.6 -23.6 6353 144.7 6.18 0.53 13.5 3.2 -23.7 6418 146.4 6.93 0.56 14.4 3.3 -22.8 6487 148.2 5.06 0.39 15.1 3.6 -24.9

195

6552 149.9 3.93 0.36 12.9 4.0 -24.9 6617 151.6 4.14 0.36 13.4 3.8 -24.8 6682 153.3 4.92 0.42 13.8 3.5 -24.9 6750 155.1 5.37 0.42 14.8 3.7 -25.3 6815 156.8 5.94 0.44 15.6 4.2 -25.1 6893 158.5 5.92 0.45 15.2 3.3 -24.4 6970 160.2 6.66 0.49 16.0 2.7 -24.0 7046 161.9 5.46 0.39 16.5 2.7 -24.7 7127 163.7 4.80 0.34 16.3 2.0 -25.9 7204 165.4 4.73 0.37 15.0 1.6 -25.9 7303 167.1 5.13 0.37 16.1 1.1 -25.7 7425 168.8 4.82 0.33 17.2 1.4 -25.8 7553 170.6 4.02 0.29 16.0 1.9 -26.6 7675 172.3 4.30 0.28 18.2 1.5 -25.7 7796 174 2.45 0.18 15.9 2.2 -25.3 175 1.79 0.17 12.0 2.7 -25.7 176 1.70 0.16 12.1 2.5 -25.7 177 1.56 0.17 10.4 2.9 -25.8 178 1.39 0.16 10.0 2.7 -26.1 179 1.50 0.18 9.9 2.8 -26.2 180 1.52 0.17 10.6 2.9 -26.2 181 1.41 0.15 10.6 2.9 -26.2 182 1.36 0.16 9.8 2.8 -26.1 183 1.34 0.16 9.7 3.2 -26.1 188 1.32 0.15 10.4 3.4 -26.5 193 1.30 0.13 11.7 3.6 -26.4 198 1.16 0.13 10.6 3.6 -27.2 202 1.36 0.15 10.6 4.5 -27.5

BC 55 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 200 5 11.27 0.93 14.1 2.3 -24.5 242 6 11.00 0.91 14.2 1.8 -24.2 284 7 12.24 0.91 15.7 1.5 -23.9 326 8 14.59 1.13 15.1 0.9 -24.0 369 9 15.97 1.32 14.1 0.4 -23.8 411 10 18.00 1.40 15.0 0.3 -23.7 453 11 17.42 1.41 14.4 0.1 -23.5 495 12 17.68 1.19 17.3 0.2 -22.6 537 13 13.55 1.00 15.8 0.7 -23.5 579 14 13.45 1.15 13.6 0.6 -24.0 621 15 14.37 1.16 14.5 0.7 -23.5 664 16 14.86 1.16 14.9 0.6 -23.5 706 17 15.02 1.16 15.1 0.6 -23.4 748 18 11.21 0.88 14.9 0.9 -23.6 790 19 10.41 0.79 15.4 1.0 -23.8 832 20 12.20 0.95 14.9 0.7 -23.4 874 21 14.65 1.20 14.3 0.4 -23.7

196

916 22 15.17 1.23 14.4 0.4 -23.7 959 23 17.38 1.56 13.0 0.1 -24.1 995 24 17.13 1.51 13.2 0.2 -23.9 1035 25 16.66 1.56 12.5 0.2 -24.1 1074 26 18.56 1.70 12.7 0.2 -24.1 1114 27 16.61 1.47 13.2 0.5 -24.4 1154 28 14.79 1.32 13.1 0.6 -24.3 1193 29 14.20 1.12 14.8 0.7 -23.9 1233 30 12.20 1.10 13.0 1.0 -24.6 1273 31 13.23 1.10 14.0 0.7 -24.2 1312 32 14.35 1.18 14.2 0.6 -24.2 1356 33 13.84 1.11 14.5 0.5 -24.4 1422 34 10.48 0.81 15.0 1.0 -24.8 1488 35 10.14 0.80 14.8 1.3 -25.2 1555 36 10.95 0.89 14.3 1.3 -25.5 1621 37 8.37 0.66 14.7 2.2 -25.5 1687 38 8.42 0.70 14.0 1.8 -25.5 1753 39 8.43 0.68 14.4 1.9 -25.3 1819 40 9.91 0.80 14.5 1.6 -25.1 1885 41 10.82 0.88 14.4 1.6 -25.0 1951 42 13.89 1.13 14.4 0.8 -24.3 2012 43 16.93 1.50 13.1 0.4 -24.8 2052 44 15.45 1.31 13.8 0.6 -24.1 2081 45 16.29 1.43 13.3 0.5 -24.1 2172 46 18.32 1.56 13.7 0.6 -23.9 2263 47 19.97 1.79 13.0 0.4 -24.3 2353 48 20.45 1.94 12.3 0.7 -24.7 2444 49 20.92 1.90 12.8 0.6 -24.6 2534 50 18.05 1.69 12.5 0.5 -24.8 2625 51 19.24 1.75 12.8 0.4 -24.8 2716 52 17.46 1.61 12.7 0.5 -25.2 2806 53 18.52 1.66 13.0 0.4 -25.1 2897 54 17.30 1.46 13.8 -0.1 -25.2

BC 10 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 0 15.49 1.46 12.4 2.3 -24.7 2 14.87 1.42 12.2 2.3 -24.8 3 8.49 0.76 13.0 3.0 -24.6 5 7.53 0.70 12.6 3.3 -24.3 7 7.36 0.63 13.6 1.8 -24.4 9 7.77 0.63 14.4 1.9 -23.9 11 4.89 0.44 13.0 2.0 -23.7 12 7.42 0.58 14.9 1.6 -23.6 13 9.99 0.85 13.7 1.3 -24.2 15 8.58 0.69 14.5 1.0 -23.2 18 9.24 0.69 15.6 0.9 -22.8 19 12.60 1.01 14.6 0.5 -23.3

197

21 10.05 0.84 14.0 1.2 -23.8 23 9.95 0.80 14.5 1.1 -23.4 25 8.78 0.73 14.0 1.1 -23.5 27 7.88 0.64 14.4 1.2 -23.2 29 10.79 0.87 14.5 1.0 -23.6 31 13.92 1.11 14.6 0.5 -23.8 32 6.35 0.51 14.5 1.3 -23.1 34 12.36 0.90 16.0 0.9 -23.3 36 12.52 1.02 14.3 1.0 -24.0 37 9.89 0.76 15.2 1.0 -23.5 39 11.27 0.84 15.7 0.9 -23.5

BC 47 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 0 2.26 0.20 13.2 2.9 -24.7 1 2.37 0.19 14.6 2.7 -24.9 6 2.10 0.15 16.3 2.7 -24.3 9 1.83 0.13 16.4 2.5 -24.2 11 2.11 0.15 16.4 2.3 -24.4 17 2.23 0.13 20.0 1.7 -23.3 24 1.61 0.12 15.7 1.6 -24.7 31 1.50 0.09 20.1 1.5 -23.1 36 1.98 0.13 17.8 1.8 -24.1 41 1.81 0.14 15.1 1.0 -24.8 46 0.93 0.06 18.7 1.1 -25.8 48 2.02 0.11 21.4 1.4 -23.4 49 1.95 0.10 23.0 2.0 -24.7 51 1.11 0.06 21.6 0.7 -26.2 56 1.64 0.11 17.4 0.8 -24.0

BC 78 15 13 Age depth (cm) %Corg %Ntot C/N (mol)  N  C 0 2.28 0.20 13.3 3.5 -24.2 1 2.32 0.21 12.9 3.5 -23.9 2 2.27 0.20 13.2 3.6 -24.0 5 1.92 0.17 13.2 3.6 -23.8 10 2.14 0.21 11.9 3.2 -24.1 15 2.01 0.21 11.2 3.0 -23.9 20 1.54 0.15 12.0 2.9 -23.8 25 2.30 0.20 13.4 2.7 -23.3 30 2.22 0.21 12.3 2.7 -23.9 35 1.89 0.18 12.3 2.6 -23.1 40 2.62 0.24 12.7 2.2 -23.8 45 2.14 0.19 13.1 2.4 -23.5 50 2.37 0.20 13.8 2.4 -22.8 55 2.32 0.20 13.5 0.5 -22.5 59 2.56 0.23 13.0 1.8 -22.9

Appendix B

Chapter 4 Supplementary Material

Deep Basin Core (location A in Fig. 4-1) Analytical Data

199

200 Shallow Neck Core (location B in Fig. 4-1) Analytical Data

Appendix C

Chapter 5 Supplementary Material

Deep Basin Core (location A in Fig. 4-1) Pigment Concentration Data

202 2-Dimensional Pigment Purification Summaries

203

204

205

206

207

VITA

James Mark Fulton

Education

Ph.D., Geosciences and Biogeochemistry, 2010 The Pennsylvania State University, University Park, PA, Department of Geosciences

M.A., Secondary Education, 2000 Ball State University, Muncie, IN, Dept. of Educational Studies

B.S., Geological Sciences, 1997 Wheaton College, Wheaton, IL, Geology Department

Awards, Honors, and Grants

P.D. Krynine Memorial Travel Grant—IMOG meeting (2009) 2nd place oral presentation award, PSU Geosciences Graduate Student Colloquium (2009) PSU Biogeochemistry Dual Title Degree Program research fellowship (2009) Shell Oil Graduate Fund Travel Grant—AGU Fall Meeting (2008) 1st place oral presentation award, PSU Geosciences Graduate Student Colloquium (2008) Worldwide University Network international collaboration grant (2005) BRIE research grant (2004) BRIE travel grant (2004) Biogeochemical Research Initiative for Education (BRIE) fellowship (2003-2006)

Professional Experience

Research Assistant, Penn State University (2003-2009 Visiting Researcher, Department on Chemistry, University of York, UK (summer 2005) Teaching Assistant, Penn State University (2002-2003, 2008) National Parks Geology (online course) TA, spring 2008 Oceanography lab instructor, spring 2008 Physical Geology lab instructor, Head TA, spring 2003 Physical Geology lab instructor, fall 2002 Teacher, Talawanda High School, Oxford, OH (2000-2002) Instructor, Department of Educational Studies, Ball State University (1999-2000)