xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001905057
007 cr mnu|||uuuuu
008 070808s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001655
Cross, Eric Charles.
Hydrogen isotopic ratios of algal and terrestrial organic matter in Lake Tulane, FL: from a modern calibration to the reconstruction of paleoclimatic and paleohydrologic conditions
h [electronic resource] /
by Eric Charles Cross.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: Recent sedimentary records have indicated that climate in low latitude, continental environments have varied significantly throughout the mid to late Holocene. In subtropical North America, major climatic phenomena such as the migration of the Intertropical Convergence Zone (ITCZ) and the Bermuda High have been shown to play a major role in this variability. Specifically, the northward migration of the ITCZ and the eastward position of the Bermuda High during summer months leads to warmer and wetter conditions over subtropical North America, and vice versa. A quantitative approach to understanding hydrologic dynamics (i.e. atmospheric circulation patterns, relative humidity) associated with these and other phenomena is necessary to accurately reconstruct the behavior of these hydrologic parameters in the past. Previous studies have shown that the hydrogen isotopic composition of algal material is a direct reflection of source waters, and that hydrogen isotopic enrichment in terrestrial material relative to aquatic biomass is a function of evaporative processes associated with the level of relative humidity in a given environment. This study utilizes a lacustrine system to provide a modern calibration that will attempt to develop a new climatic proxy for relative humidity and further examine varability in the behavior of environmental waters. This calibration was then applied to a sedimentary record to examine hydrologic variability in the geologic past.
Thesis ( M.A.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains137 pages.
Adviser: David Hollander, PhD.
x Marine Science
t USF Electronic Theses and Dissertations.
Hydrogen Isotopic Ratios of Algal and Terrestrial Organic Matter in Lake Tulane, Florida: From a Modern Calibration to the Reconstruction of Paleoclimatic and Paleohydrologic Conditions by Eric Charles Cross A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science College of Marine Science University of South Florida Major Professor: David Hollander, P h D Terrence Quinn, Ph.D. Edward VanVleet, Ph.D. Date of Approval: July 26, 2006 Keywords: geochemistry, paleoclimate, lacustrine, sedimentary, subtropical Copyright 2006 Eric Charles Cross
i Table of Contents List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Chapter One. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introductory Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Major Climatic Phenomena in the Low Latitudes . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Climate of Subtropical North America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Relative vs. Specific Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 The Hydrologic Cycle and Isotopic Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Justification for the Utilization of Hydrogen Isotopes . . . . . . . . . . . . . . . . . . . . . . 21 Hydrogen Isotopes in Organic Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Terrestrial vs. Aquatic Biomass: Isotopic Distinctions . . . . . . . . . . . . . . . . . . . . . 25 Utilizing Hydrogen Isotopic Analysis to Quantify Relative Humidity . . . . . . . . . 27 Bulk vs. Molecular Hydrogen Isotopic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 29 Lake Tulane Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Chapter Two. Research Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Preparatory and Field Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Lake Water, Particulate Organic Material, and Terrestrial Material Sampling . . . 39 Sediment Core Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Bulk Isotopic Analysis: General Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Bulk Isotopic Analysis of Lake Waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Particulate Organic Material, Terrestrial Plant Material, and sediments: Bulk Isotopic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Total Lipid Extraction (TLE), Methylation, and Column Chromatography . . . . . 44 Gas Chromatography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Compound Specific Isotopic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Chapter Three. Project Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Regional Hydrologic Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Lake Biological Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Isotopic Signature of Lake Waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Particulate Organic Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
ii Terrestrial Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Compound Specific Hydrogen Isotopic Results . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Chapter Four. Discussion of the Isotopic Relationships in Lake Tulane . . . . . . . . . . . . 69 General Isotopic Systematics Over the Annual Cycle in Lake Tulane . . . . . . . . . 69 Isotopic Behavior of Lake Tulane Waters Over the Annual Cycle . . . . . . . . . . . . 70 Relationship Between Lake Waters and Aquatic Biomass . . . . . . . . . . . . . . . . . . 75 Isotopic Variability in Terrestrial Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Terrestrial Biomass in Lake Tulane and Relative Humidity . . . . . . . . . . . . . . . . . 88 Molecular Isotopic Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Implications for Assessing Climatic Variability . . . . . . . . . . . . . . . . . . . . . . . . . 96 Summary of the Lake Tulane Modern Calibration . . . . . . . . . . . . . . . . . . . . . . . . 98 Chapter Five. Applications of Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Objectives of the Application Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Geochemical Analysis of Abrupt Climate Change in Subtropical North America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Anthropogenic Influence on Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Molecular Analysis of Recent Sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 The Little Ice Age and Medieval Warm Period . . . . . . . . . . . . . . . . . . . . . . . . . 112 Molecular Analysis of the Long-Term Sedimentary Record . . . . . . . . . . . . . . . 118 Climatic Shifts in the Holocene and Through Deglaciation . . . . . . . . . . . . . . . . 121 Climatic Shifts from the Holocene into Glacial Conditions . . . . . . . . . . . . . . . . 123 Correlations with Other Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Chapter 6: Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 List of References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
iv List of Figures Figure 1. Summer and Winter Climatic Dynamics Over Florida . . . . . . . . . . . . . . . . 11 Figure 2. Average Hydrographic Data For the Lake Tulane Region . . . . . . . . . . . . . 13 Figure 3. Average Annual Relative Humidity Over North America . . . . . . . . . . . . . 16 Figure 4. The Meteoric Water Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 5. Deuterium in Precipitation Over North America . . . . . . . . . . . . . . . . . . . . 22 Figure 6. Evapotranspiration Processes in Terrestrial Leaves . . . . . . . . . . . . . . . . . . 26 Figure 7. Schematic of Isotopic Relationships Between Terrestrial and Algal Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 8. Relationship Between Relative Humidity and Isotopic Fractionation . . . . . 28 Figure 9. Lake Tulane Location Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure 10. Geologic Morphology of Central Florida, The Lake Whales Formation . . . 34 Figure 11. US Geological Survey Topographic Map of Lake Tulane Region . . . . . . . 35 Figure 12. Potentiometric Contours of Groundwater in Florida . . . . . . . . . . . . . . . . . . 36 Figure 13. Correlating Thermal and Hydrologic Climatic Archives . . . . . . . . . . . . . . . 37 Figure 14. Bathymetry and Sampling Locations in Lake Tulane, FL . . . . . . . . . . . . . . 39 Figure 15. Seasonal Perspective of the D of Lake Waters With Respect to Depth . . . 53 Figure 16. D of All Sampled Lake Waters Over the Annual Cycle in Lake Tulane . . 55 Figure 17. Monthly Depth-Averaged Isotopic Values of Lake Waters Over an Annual Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 18. D of Particulate Organic Material in the Surface Waters and at 18.3 meters in Lake Tulane, FL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
iv Figure 19. Average Monthly Isotopic Values of Particulate Organic Material in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Figure 20. D of Each Species of Terrestrial Material Sampled Over an Annual Cycle in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Figure 21. D of All Terrestrial Material Sampled Over an Annual Cycle in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Figure 22. Molecular Isotopic Values of Terrestrial and Aquatic Material in Lake Tulane, FL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Figure 23. Overall Isotopic Relationships Determined from Calibration Study . . . . . . 69 Figure 24. Isotopic Data From USGS Florida River Sites and Lake Tulane Waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure 25. Seasonal Variability in the D of Precipitation Across North America . . . . 74 Figure 26. Isotopic Offset Between POM in Surface Waters and at 18.3 meters Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Figure 27. Average D of POM and Lake Waters Over the Annual Cycle in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure 28. Average Isotopic Offset Between Algal Material and Lake Waters Over the Annual Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure 29. Average Monthly Isotopic Values of Oak and Pine Biomass in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Figure 30. Schematic Showing the Isotopic Relationships Observed in and Around Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Figure 31. Average D of Terrestrial Material and Average Relative Humidity Over the Annual Cycle in Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Figure 32. Isotopic Offset Between Terrestrial and Algal Material Plotted with Relative Humidity Over the Annual Cycle . . . . . . . . . . . . . . . . . . . . . . . . . 92 Figure 33. Correlation Between the Isotopic Offset Between Terrestrial and Algal Material and the Average Annual Relative Humidity . . . . . . . . . . . . 93 Figure 34. Comparison Between the "! D (T-A) in Bulk and Molecular Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
v Figure 35 Bulk Hydrogen Isotopic Measurements from Lake Tulane Representing the Last 2000 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Figure 36. Age Model for the Lake Tulane Sedimentary Record . . . . . . . . . . . . . . . . 104 Figure 37. End-member schematic of Percentage of Algal Contribution to Sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Figure 38. Calculated Percentage of Algal Material Entering the Sediments of Lake Tulane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Figure 39. Correlation Between Lake Tulane D/H and Cariaco Basin Titanium Record During the Little Ice Age and Medieval Warm Period . . . . . . . . . 117 Figure 40. C 16 and C 28 Fatty Acid Isotopic Variability Over the Last 70,000 Years . . 120 Figure 41. Calculated Relative Humidity Over the Last 70,000 Years . . . . . . . . . . . . 121 Figure 42. Relationship Between GISPII Ice Core Data and Lake Tulane Relative Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
iii List of Tables Table 1. USGS Hydrogen Isotopic Data from Florida Rivers . . . . . . . . . . . . . . . . . . . 18 Table 2. Average Hydrographic Data for the Lake Tulane Region . . . . . . . . . . . . . . . 49 Table 3. D of Lake Waters in Lake Tulane (9/2002 9/2003) . . . . . . . . . . . . . . . . . 51 Table 4. D of Particulate Organic Material in Lake Tulane (9/2002-9/2003) . . . . . . 58 Table 5. D of Terrestrial Material in Lake Tulane (9/2002-9/2003) . . . . . . . . . . . . . 61 Table 6. D of Terrestrial and Aquatic Molecular Compounds From Lake Tulane, FL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Table 7. Results of the D Analysis of 140cm Sediment Core from Lake Tulane, FL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Table 8. Molecular Analysis of Fatty Acids & Percentage of Algal Material in Lake Tulane Sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
vii Hydrogen Isotopic Ratios of Algal and Terrestrial Organic Matter in Lake Tulane, Florida: From a Modern Calibration to the Reconstruction of Paleoclimatic and Paleohydrologic Conditions Eric Charles Cross ABSTRACT Recent sedimentary records have indicated that climate in low latitude, continental environments has varied significantly throughout the midto late Holocene. In subtropical North America, major climatic phenomena such as the migration of the Intertropical Convergence Zone (ITCZ) and the Bermuda High have been shown to play a major role in this variability. Specifically, the northward migration of the ITCZ and the eastward position of the Bermuda High during summer months leads to warmer and wetter conditions over subtropical North America, and vice versa. A quantitative approach to understanding how the hydrologic dynamics (i.e., atmospheric circulation patterns, relative humidity) associated with these and other phenomena is necessary to accurately reconstruct the behavior of these hydrologic parameters in the past. Previous studies have shown that the hydrogen isotopic composition of algal material is a direct reflection of source waters, and that hydrogen isotopic enrichment in terrestrial material relative to aquatic biomass is a function of evaporative processes associated with the level of relative humidity in a given environment. This study utilizes a lacustrine system to provide a modern calibration that will attempt to develop a new climatic proxy for relative humidity and further examine variability in the behavior of environmental waters.
vii The hydrogen isotopic behavior of aquatic biomass, five species of terrestrial biomass, and lake waters are examined over an annual cycle in order to calibrate their geochemical relationships. This calibration was then applied to a sedimentary record to examine hydrologic variability in the geologic past. Lake Tulane, located in Central Florida, USA, is a subtropical, groundwater-fed acidic lake with a high sedimentation rate and well-preserved organic matter. The small catchment size of the lake allows for relatively simultaneous deposition of algal and terrestrial biomass. The groundwater-fed nature of the system allows for the same pool of source waters to feed both the aquatic and terrestrial biomass. The bulk D of modern algal material over an annual cycle shows little to no variability in lake-water chemistry, coincident with the groundwater-fed nature of the lake and relatively constant D of precipitation. Average terrestrial oak and pine biomass shows a seasonal variability in D (~15) that coincides with seasonal relative humidity (~6%). Increased levels of enrichment in terrestrial biomass correlate well with decreased levels of relative humidity, further supporting the hypothesis that higher rates of evapotranspiration lead to increased isotopic enrichment. The offset between the hydrogen isotopic signature of bulk terrestrial and algal biomass ( "! D (T-A) ) is observed to provide a quantifiable parameter that can be used to examine relative humidity. The variability in "! D (T-A) over the annual cycle in Lake Tulane was used to generate a mathematical expression that can be used as a proxy for relative humidity (RH) and source water conditions in a sedimentary record (RH = -0.1917( "! D (A-T) ) + 82.088. Preliminary molecular analysis of specific fatty acids used as biomarkers for algal (C 16 ) and terrestrial (C 28 ) shows that
viii the hydrogen isotopic behavior in these molecular compounds yields the same relationships as those observed in the bulk calibration, providing a tool for resolving these parameters in a sedimentary archive and utilizing the climatic proxy developed in the calibration to examine hydrologic change in the geologic past. This calibration and new proxy were applied to two separate sedimentary archives obtained from Lake Tulane spanning a low-resolution, long-term record (70k years) and a high-resolution, short-term record (2k yrs). The long-term record was utilized to examine changes in relative humidity and environmental water over the last glacial/interglacial transition. The short term record was utilized to examine specific abrupt climatic events over the last 2,000 years (The Little Ice Age, from 1450 to 1750 c.y., the Medieval Warm Period, from 750 to 1150 c.y., and recent anthropogenic effects. Preliminary applications of these this approach incorporated both a bulk and a molecular isotopic approach. The molecular approach was used to examine the last glacial transition, showing more depleted source waters and a lower average relative humidity during glacial times. The bulk approach was applied to the high-resolution short core, and indicated higher relative humidity during the Medieval Warm Period and lower relative humidity during the Little Ice Age. These results correlate with previous research that suggests that the long-term migration of the ITCZ, and potentially the Bermuda High have led to the observed variability in hydrologic conditions in the low latitudes. Anthropogenic influence manifested itself in the most recent 70 years of the sedimentary record as an increase in the relative contribution of algal material to the sediments as a function of eutrophication. These results correlate well with previous paleoclimate research, and indicate that this
ix novel approach provides a quantifiable method with which to reconstruct relative humidity and the isotopic composition of source waters in the geologic past.
1 Chapter 1: Introduction Introductory Statements High-resolution sedimentary climatic archives from lacustrine and oceanic systems have indicated that major changes in temperature, ocean and atmospheric circulation, and hydrologic dynamics have occurred throughout the globe over a wide range of time scales (decades, centuries, millennia) (Thompson et al., 1986; Grimm et al., 1993; Stauffer et al., 1997; Bond et al., 1999; Peterson et al., 2000; Crowley & Berner, 2001; Huang et al. 2002). These studies have shown that the response to climatic variability as a function of latitude in the geologic past is observed to manifest itself in different ways. In the mid to high latitudes the response to climatic variability has been shown to be primarily thermal in nature. CO 2 measurements in ice cores and 18 Omeasurements in marine sediments at higher latitudes have been used as proxies to show dramatic changes in atmospheric and oceanic temperatures in the geologic past (Dansgaard et al., 1971; Heinrich, 1988; Andrews et al., 1998; Keigwin & Jones, 1995; Bond et al., 1993). Climatic records from Antarctic ice cores reflecting glacial/interglacial transitions show temperature fluctuations of up to 10 degrees and variability in atmospheric O 18 of up to 2 (indicative of atmospheric temperature variability) at higher latitudes throughout the last 400,000 years (Petit et al., 1999). Bond et al., (1992) has demonstrated that these temperature fluctuations correlate to other environmental records including assemblage changes in planktonic organisms and geochemical analysis of ice-rafted debris. These relationships provide linkages between temperature, the extent of ice sheets and associated ecological dynamics throughout the last several hundred thousand years. Moritz et al.. discusses the fact that temperature variability in the high latitudes can
2 potentially be directly responsible for altering the Arctic Oscillation mode of atmospheric circulation, providing a direct link between the effects of thermal variability and large-scale climatic behavior. Thus it is apparent that the thermal changes in the high latitudes are a major forcing factor in atmospheric and oceanic processes. In contrast, the response to climate change in the lower latitudes over tens to thousands of years has been shown to be primarily hydrologic in nature (Baldini et al.; Northfelt et al., 1981; Burk & Stuiver, 1981; Jasper & Hayes, 1993; Werne & Hollander, 2000; Haug et al., 2001; Huang et al., 2002; Hughen et al., 2004). Marine records of titanium percentages and isotopic fluctuations in the Cariaco Basin sediments have shown significant variability in precipitation levels and terrestrial runoff in the low-latitudes over periods of hundreds to thousands of years linked to the latitudinal migration of the Inter-Tropical Convergence Zone (ITCZ), one of the major climatic phenomena controlling moisture balance in the tropics (Haug et al., 2001; Hughen et al., 2004). Other continental archives such as tropical ice cores, lake sediments and annual banding in tree rings and speleothems complement these marine records by showing corresponding changes in moisture balance over land (Baldini et al.; Northfelt et al., 1981; Burk & Stuiver, 1981; Thompson et al., 1984; Huang et al., 2002). In a specific example, Thompson et al. (1984), has utilized oxygen isotopic measurements and analysis of dust layers in tropical ice cores to examine variability in wet and dry periods over the continents in the low-latitudes during the last 2,000 years.
3 Understanding the continental response to hydrologic variability is especially crucial in providing a comprehensive understanding of climatic dynamics. The teleconnections between continental climate and behavior over the oceans can provide insights into the cause and response of climate change across the globe. In order to examine these relationships, specific variables can be used as indicators of variability in the source and circulation patterns of atmospheric moisture over land, plant species dominance, the balance between precipitation and evaporation in a given region, and temperature (Epstein et al., 1977; Jasper & Hayes, 1993; Grimm et al., 1993; Haug et al., 2001). Understanding the behavior of these variables over the continents can answer important climatic questions, such as the location of moisture sources, how moisture levels are changing over time, the role that temperature plays in hydrologic variability, and the biologic response to climatic change over the continents in the low-latitudes. In this way, a comprehensive assessment of the continental response to abrupt climate change events can be obtained. Previous research has examined these parameters in a variety of ways in order to broaden the understanding of hydrologic dynamics over land. Specifically, Burke & Stuiver (1981) have utilized the geochemical analysis of tree-ring cellulose in an attempt to understand moisture balance and temperature change in terrestrial environments, and Grimm et al. (1999) have demonstrated that fluctuations in the relative abundance of specific species of pollen in lacustrine sediments can be used to infer changes in general moisture conditions in sub-tropical North America. Epstein et al. (1976) complements these studies by examining the oxygen and hydrogen isotopic composition of aquatic and
4 terrestrial plant cellulose and relating these parameters to the isotopic behavior of meteoric water and, in turn, hydrologic processes. Developing even more quantitative methods with which to examine specific hydrologic parameters in continental systems will help to provide a more comprehensive understanding of the importance of the lowlatitude continents in global climatic dynamics, and further establish climatic relationships between the land and oceans. Subtropical North America provides a unique perspective on making connections between climate in the tropics, subtropics, and higher latitudes, as well as between the oceans and the continents. Florida sits in a transition zone for moisture moving from the equator to the poles (Peixoto et al., 1996; Davis et al.,1997; Wells, 1998; Katz et al., 2003). Here, the intimate linkages between the hydrology associated with the Gulf of Mexico, the Caribbean, the Atlantic Ocean, and the southeastern United States can provide a wealth of information describing heat and moisture transport in present-day and geologic climatic conditions (Jasper 1989; Roulier and Quinn, 1995). This project will focus on the analysis of the hydrologic environment of Lake Tulane, located in Avon Park, Florida. The geographic position of Lake Tulane will allow for a comprehensive analysis of the various hydrologic phenomena that converge in sub-tropical North America, providing an ideal location to observe continental response to hydrologic variability in the low-latitudes. Specifically, the stable isotopic analysis of organic and inorganic materials in lake sediments has been shown to provide quantitative perspectives on temperature, the origin of source waters, and moisture conditions over the continents, and can be applied to sediments of many ages (Huang et al. 2002; McKenzie &
5 Hollander, 1993). Lake systems generally contain high-resolution sedimentary archives abundant in organic matter (due to high accumulation rates associated with high productivity), and their small catchment sizes and isolated environments provide a natural laboratory with which to reconstruct regional climate change. Due to their ubiquitous nature and sensitivity, observing isotopic relationships in a lake system is the ideal approach for expanding the understanding and potential for reconstructing the continental response to hydrologic variability. Hydrogen isotopes ( D) provide an ideal parameter with which to examine hydrologic dynamics due to their intimate linkage to the hydrologic cycle (Craig, 1961, to be discussed in detail below). D analysis of waters and organic materials can provide insights into a wide variety of the hydrologic parameters such as temperature, the origin of source waters, and potentially relative humidity conditions. The goal of this project is to develop a modern calibration in the Lake Tulane hydrologic environment that will enable the use D isotopic analysis of organic materials as a new proxy to observe the continental response to abrupt climate change over a wide range of latitudes and locales. Calibration studies have been shown to be an essential first step in understanding isotopic relationships to climatic parameters. For example, Huang et al. (2002) and Sauer et al. (2001) both expanded on previous research discussing important isotopic relationships between surface sediments and lake waters in lacustrine systems, and how these relationships can be exploited in sediments at depth to understand climate change. This study further expands on those principles by not only observing the isotopic behavior of the sediments in a lake system, but also the aquatic and surrounding terrestrial vegetation
6 associated with their formation. The hypotheses and specific goals of this project are as follows: 1) Obtain a detailed understanding of the hydrologic dynamics, including relative humidity, precipitation, and temperature, associated with subtropical North America over an annual cycle. 2) Correlate the behavior of these parameters with specific climatic phenomena affecting subtropical North America, including the seasonal migration of the Intertropical Convergence Zone (ITCZ) and the Bermuda High (BH), as well as the seasonal penetration of northern storm fronts into the region. 3) Using Lake Tulane, FL, provide a modern calibration of the bulk hydrogen isotopic behavior and relationships of the lake waters and terrestrial and algal organic materials in this system. 4) Use this calibration to test the hypotheses that: 1. The isotopic composition of source waters is directly reflected in the D of algal material. 2. Relative humidity can be calculated directly from the level of hydrogen isotopic enrichment in terrestrial material relative to algal material ( "! D (T-A) ), thus defining a new climatic proxy. 5) Further examine the hypothesis that seasonal variability in relative humidity and the isotopic composition of source waters can be correlated to variability in the climatic phenomena associated with this region, specifically the migration of the ITCZ and BH and the effects of northern storm fronts penetrating into subtropical North America.
7 6) Implement a compound-specific isotopic analysis of the terrestrial and algal organic matter in Lake Tulane to test the hypothesis that specific molecular biomarkers can be utilized to isolate the D of algal material and terrestrial material in a sedimentary archive. 7) Test the hypothesis that the modern calibration of bulk and molecular D analysis can be applied to a sedimentary record spanning the last 70,000 years to observe how source-water chemistry and relative humidity have changed over abrupt and long-term time scales, and infer the climatic forcing responsible for this variability. a) Bulk and molecular D analysis will be applied to a short (170cm) sedimentary archive to examine the correlation between excursions in the bulk hydrogen composition of the sediments during: 1) The last 100 years, 2) The Little Ice Age, and 3) The Medieval Warm Period. b) The molecular isotopic approach will be applied to a sedimentary archive representing the last glacial transition, and will be used to examine fluctuations in temperature and relative humidity across this time period.
8 Major Climatic Phenomena in the Low-latitudes Understanding the role that the tropics have played in hydrologic and climatic variability is imperative because of their importance in global hydrologic dynamics and the increasing need to quantify climate change relative to sociological issues such as global warming. The low-latitudes are the major source of heat and water vapor to the atmosphere, acting as the starting point for the global circulation of moisture and the heat engine (Broeker, et al.; 1997, Wells 1998). Higher latitudes then act as a sink for water vapor and heat, creating a general pole-ward movement of these parameters and maintaining a global balance of energy (Wells, 1998). The atmospheric and oceanic circulation associated with maintaining this balance are two of the most important processes affecting climate on both regional and global scales. Moisture balance has been shown to be the dominant factor in abrupt climate change events in the low-latitudes (spanning time scales of hundreds of years), causing variability in parameters such as the length of growing seasons, levels and sources of runoff, plant species dominance, and water levels in lakes and rivers (Jasper, 1989; Jasper and Hayes, 1993; Huang et al., 2001; Poore et al., 2003). Latitudes near the equator absorb much larger amounts of solar radiation relative to locations closer to the poles (Wells, 1998). This major heat influx leads to high levels of both evaporation and precipitation throughout the year. However, variability in the relative amounts of these two parameters does occur, and is generally associated with specific climatic phenomena that characterize the tropics and subtropics. The major climatic phenomena associated with the generation of moisture in the tropics is the position of the Intertropical Convergence Zone (Peterson et al., 2000; Haug et al.,
9 2001; Philander et al., 1995). The ITCZ, a low pressure zone defined by the convergence of easterly trade winds positioned to the north and south of the equator, is characterized by high levels of precipitation, relative humidity and warm temperatures. The latitudinal position of the ITCZ varies throughout the annual cycle, as well as over longer time scales (Haug et al., 2001, Poore et al., 2003). During the northern hemisphere summer, the ITCZ migrates north, bringing warmer temperatures and higher levels of precipitation to the northern tropics and subtropics. In contrast, during the northern hemisphere winter the ITCZ migrates south, causing cooler and drier conditions to persist in the northern tropics and subtropics. Subtropical North America sits at the northern-most end of the area directly affected by the position of the ITCZ, and thus will show a continental response to its seasonal and longitudinal migration. In addition to the ITCZ, however, there are other localized climatic phenomena that play a role in controlling the hydrology of this region, such as the relative position of the North Atlantic Subtropical Anticyclone, or the Bermuda High. Studies have shown that the position of the Bermuda High is a major factor in relative humidity and precipitation levels in the southeastern United States (Peixoto et al., 1996; Davis et al., 1997; Katz et al., 2003; Winsberg et al.). The BH is a high-pressure anticyclone positioned over the Central Atlantic. During the northern hemisphere summer, this zone migrates eastward, causing an increase in the amount of warm, moist air flowing from the tropics and the Gulf of Mexico up along the entire east coast of the United States, thus bringing increased levels of moisture into the Lake Tulane region of Florida (Davis et al., 1997). The northward migration of the ITCZ during this same time
10 period plays a significant role as a major source of moisture to the Lake Tulane region. During the northern hemisphere winter, the BH migrates westward, causing a relative decrease in the amount of warm, moist air being supplied from the lower latitudes to the southeastern United States (Davis et al., 1997). However, an increase in the migration of northern storm fronts into the southeastern United States during the winter allows for precipitation rates to continue to play a significant role in the hydrologic dynamics of the region (Davis et al., 1997; Katz et al., 2003). The meridional extent of the migration of the BH, as well as the length of time it persists, has been shown to directly correlate with moisture level conditions in the region (Davis et al., 1997; Katz et al., 2003). Additionally, the position of the mid-latitude jet stream across North Amnerica throughout the annual cycle can also lead to fluctuations in moisture levels across subtropical North America. Thus it is the combination of the latitudinal migration of the ITCZ, the east-west migration of the BH, the southward penetration of storm fronts and the position of the jet stream that can potentially lead to the seasonal hydrologic variability seen in subtropical North America. Summer months see a northward shift of the ITCZ and a eastward shift of the BH, leading to potential increases in temperature, precipitation, and relative humidity. The remaining percentage of annual moisture throughout the Lake Tulane region and sub-tropical North America is associated with the introduction of the frontal systems from the north, allowing for a relatively wet climate throughout the year. These relationships are summarized in Figure 1 (from Lutgens and Tarbuck, 2001).
11 Figure 1. Seasonal migration of the ITCZ from January (top) to July (bottom). Note the variability in wind direction and the Bermuda High as well as the ITCZ (Figure obtained from Lutgens and Tarbuck, 2001)
12 Climate of Subtropical North America Subtropical North America and the Lake Tulane region in Florida are characterized by a relatively warm, wet climate throughout most of the year. For example, data from the National Weather Service indicate that average temperatures range from ~18.5 to 29.5 C. Average monthly precipitation ranges from 6.5-8 inches during the summer months (June September) during the time when the ITCZ has migrated to its northern position (Katz et al., 2003). Rainfall is still relatively high in the winter, ranging from 2-3 inches per month as storm fronts from the northwest enter into this region. Average relative humidity shows a seasonal trend with higher values during peak summer and winter months (approximately 73-75.5%) and lower relative humidity ranging from 68.5-70% during the spring months (March-May). These annual climatic parameters are presented in Figure 2. These seasonal fluctuations in precipitation, temperature and relative humidity are associated with the combined effects of the migration of the ITCZ and the Bermuda High (Haug et al. 2001; Katz et al., 2003).
13 0 5 10 15 20 25 30 Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Temperature (celsius) Regional average temperature 0 1 2 3 4 5 6 7 8 9 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Precipitation (inches) Average monthly precipitation levels 64 66 68 70 72 74 76 Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Average Relative Humidity (%) Regional average relative humidity Figure 2. Precipitation levels, regional average temperature and relative humidity observed on a monthly basis in the Lake Tulane region (from National Weather Service). Note that although summer months are very humid, winter months still record high levels of relative humidity.
14 Relative vs. Specific Humidity There is a distinct and important difference between specific and relative humidity. In Central Florida there is a relatively small seasonal trend in the average relative humidity throughout the year, ranging from average monthly minimums of ~69%, and maximums of ~76% (see Figure 2). However, specific humidity can range up to between 40% and 50% on an annual basis. This contrast is due to the fact that relative humidity is a measurement composed of two separate parameters temperature and specific humidity (Peixoto & Oort, 1996). Specific humidity is the absolute measurement of the total moisture present in a given parcel of air. During summer months, specific humidity is considerably higher, but increases in air temperature elevate the atmospheres ability to hold moisture, thus preventing a significant increase in the relative humidity. The opposite is also true, that in winter months, specific humidity is lower but decreased air temperatures lessen the ability of the atmosphere to hold moisture. Thus it is this balance between levels of actual moisture in the air and the temperature of the air that lead to a small degree of variability in relative humidity throughout the seasons. However, there is a high degree of variability in relative humidity throughout North America (Figure 3). It can be assumed that as circulation patterns are altered over time, either seasonally or geologically, air masses with different hydrologic characteristics may move over the southeast United States. The slight seasonal trends observed in Central Florida can be related to the migration the ITCZ and the Bermuda High as discussed earlier, with higher levels of atmospheric moisture seen in summer months when the ITCZ is positioned more to the north and the BH more to the east, both phenomena allowing for increased levels of moisture to enter the southeastern US. Relative fluctuations in precipitation
15 levels, temperature, and relative humidity have the potential to cause chemical fluctuations in the hydrogen isotopic signature of waters and associated organic materials in a specific locale (Epstein et al. 1976; Deniro et al. 1981; Yapp & Epstein, 1982; Sauer et al., 2000; Huang et al. 2002). By quantifying this hydrogen isotopic behavior and associating it with atmospheric and climatic phenomena, the potential exists for developing isotopic proxies that can be used to understand hydrologic variability during the geologic past.
16 Figure 3
17 The Hydrologic Cycle and Isotopic Behavior 18 O and deuterium ( D) have both been shown to be climatic parameters that are directly linked to the hydrologic cycle, and can be used as indicators of its variability (Craig, 1961; Burk & Stuiver, 1981). The most direct link between these parameters and the global hydrologic cycle is the isotopic composition of source waters, or precipitation. The latitudinal distribution of oxygen and hydrogen isotopes in precipitation has been quantified and is represented by the meteoric water line (MWL, Figure 4, from Craig, 1961). The MWL shows isotopic variability of up to 200 in D and 20 in 18 O from the equator to the poles. Enriched D values of 0 to -20 and 18 O of to are seen in the low-latitudes, and extremely depleted values of in 18 O down to D values of have been observed near the poles. This is due to Rayleigh Distillation, a process associated with a rain-out effect seen as precipitation moves to higher latitudes (Pilson, 1998). Temperature gradients can also play a role in the isotopic composition of source waters as both a function of latitude and altitude, also manifested in the process of Rayleigh Distillation (Craig, 1961; Huang et al., 2002). Due to the large isotopic effect associated with hydrogen and deuterium, a much higher degree of isotopic variability in precipitation is observed in this parameter on a global scale.
18 Figure 4. Altered from Craig, 1961, The meteoric water line, depicting the relationship between the isotopic composition of deuterium and oxygen to latitude It would be expected that ground and surface waters in rivers and lakes of varying latitude would reflect these isotopic changes in precipitation. Specifically, in correlation with the regionally averaged data from the IAEA for isotopes in precipitation, the US Geological Survey has conducted research on the hydrogen and oxygen isotopic composition of depth-integrated stream and surface waters across the United States, including multiple sites in Florida near Lake Tulane (Coplen and Kendall, 2000). Table 1 presents selected hydrogen isotopic data from this report from river systems in central Florida surrounding the Lake Tulane region. Hydrogen isotopes recorded in terrestrial surface waters fall into the same range as predicted by the MWL, exhibiting relatively
19 enriched values of D. Isotopic variability among all sites in Florida is relatively low compared to global ranges, indicating that the D composition of precipitation and surface waters is relatively constant at these latitudes. However, each site does exhibit annual variability that can potentially be correlated to annual climatic shifts. Table 1 St. John's River Near Deland, FL Fisheating Creek at Palmdale, FL Date D (VSMOW) Date D (VSMOW) Jan-86 -7.7 Feb-86 3.5 Mar-86 -7.4 Apr-86 2.7 May-86 -3.9 Jul-86 -3.8 Jul-86 -3.9 Aug-86 -3.6 Aug-86 -4.5 Jan-87 -11.7 Jan-87 -1.9 Feb-87 -1.8 Feb-87 -4.7 Mar-87 1.1 May-87 -7.1 May-87 6.1 Jun-87 -2.5 Jun-87 11 Sep-87 -2.2 Aug-87 7.5 Alafia River at Lithia, FL Date D (VSMOW) Feb-86 -6.3 Mar-86 -3.1 Jun-86 -6.4 Aug-86 -5.2 Sep-86 -4.4 Dec-86 -6.1 Mar-87 -1.5 Jun-87 0.3 Sep-87 -3.4 Hydrogen Isotopic Data from the US Geological Survey From Surface Waters of Rivers in Central Florida (1986-1987)
20 Changes in atmospheric circulation and the temperature and altitude of precipitation provide the potential to bring air masses with isotopically distinct precipitation into a given system. In subtropical North America we would expect to see the isotopic composition of source waters vary as changes in the dominant circulation patterns such as the BH and the position of the ITCZ occur over time. While this variability may be relatively small on a seasonal scale, understanding these potentially minor fluctuations in a modern environment will help to reconstruct large-scale variability in the past. Previous research (Lin et al., 1997; Poore et al., 2003) has examined the oxygen isotopic composition of planktonic organisms in the Gulf of Mexico in an attempt to understand climate change in sub-tropical North America. However, these studies still do not provide insights into the terrestrial realm. It is imperative to obtain a quantitative method of analysis with which to observe this variability in a continental system.
21 Justification for the Utilization of Hydrogen Isotopes The meteoric water line shows that both oxygen and hydrogen provide the potential to be used as proxies for hydrologic conditions. The oxygen and hydrogen isotopic composition of organic matter has been shown to be an extremely useful tool in understanding climatic phenomena. Poore et al. (2003) specifically used 18 O in the subtropics as an indicator of the latitudinal migration of the ITCZ, one of the major climatic factors affecting the northern Gulf of Mexico, adjacent to the study site for this project. In the terrestrial environment, Cooper & Deniro (1989) and Sternerg et al. (1986) have shown that both oxygen and hydrogen isotopic ratios in plant tissue provide direct linkages to the isotopic composition of surrounding source waters. Burk and Stuiver (1981) have utilized the oxygen isotopic composition of tree-ring cellulose as a potential indicator of relative humidity conditions in the surrounding atmosphere. Oxygen isotopic analysis is also often implemented in the analysis of carbonate material found in marine sediments to reconstruct temperature. However, lacustrine systems often exhibit conditions such as acidic waters or unique digenetic processes that prevent the preservation of carbonate material that can be used for oxygen isotopic analysis. Due to the fact that lakes often contain extensive sedimentary records that have the potential to be used for hydrologic reconstructions, another isotopic parameter is necessary to properly exploit these conditions. Huang et al. (2002) discusses the fact that hydrogen is abundant in all organic matter, is resistant to varying pH levels and water chemistry, and can potentially provide a wealth of climatic information if it is utilized properly. Additionally, the large isotope effect associated with deuterium and hydrogen exhibits itself in the extremely large range in the isotopic composition of precipitation over
22 varying latitudes (Figure 5). This project hypothesizes that this wide isotopic range presents the opportunity for even minor fluctuations in atmospheric circulation to potentially manifest themselves in the hydrogen isotopic record as recognizable isotopic shifts. Figure 5. General model of the D of precipitation over North America (from www.iaea.org). Note the wide range of values seen latitudinally. The migration of air masses can cause isotopically distinct precipitation to enter a given environment, and these values can be recorded in organic matter.
23 Hydrogen Isotopes in Organic Matter Multiple studies (Epstein et al. 1976; Deniro et al. 1981; Yapp & Epstein, 1982; Sauer et al., 2000; Huang et al. 2002) have shown that the isotopic composition of source waters is recorded in the D of terrestrial and aquatic organic matter as these organisms incorporate the source waters into their own biomass. Fractionation processes associated with biosynthesis can alter the original isotopic composition of the source waters being used, but these fractionation processes can be quantified and taken into account by analyzing specific parts of the plant biomass, such as cellulose nitrate or individual lipids (Epstein 1976; Estep & Hoering,1980; Deniro 1981; Sessions et al., 1999). Northfelt et al., 1981 implemented this approach using both deuterium and carbon in the cellulose nitrate of tree rings in order to show that the hydrogen isotopic composition of organic matter in plants will provide information useful for climatic reconstruction provided the initial isotopic record has not been changed by subsequent physiological or diagenetic processes. Epstein et al. (1976) utilized the hydrogen isotopic analysis of the cellulose of both aquatic and land plants to show that organic matter from both the terrestrial and aquatic environments provides the potential to be indicators of climate change. Congruently, Sauer et al. (2001) examined the isotopic composition of lipid biomarkers from algal and terrestrial sources in both marine and freshwater environments to determine their relationship to source waters. Results of this research exhibited relatively stable direct fractionation levels of up to in algal biomarkers relative to source waters. Terrestrial material exhibited an enrichment of approximately 30 relative to algal material, for reasons to be discussed below. These studies provide evidence that the D of organic matter can be useful in climate research, and it is feasible to assume that
24 the fractionation processes associated with the biosynthesis of terrestrial and aquatic organic matter can be calibrated and directly related to the associated lake waters and the surrounding terrestrial environment. This calibration will provide a present-day geochemical understanding of isotopic relationships that can then be applied to sedimentary archives. The Lake Tulane system provides an environment where fluctuations in temperature and the isotopic composition of source water are minimal throughout the annual cycle, thereby making this an ideal system with which to analyze the isotopic behavior of algal biomass ( www.iaea.org www.nws.noaa.gov ). By observing the D of algal material in the modern Lake Tulane system, this project will test the hypothesis that the relationships between source waters and aquatic plant biomass can be quantified, and isotopic variability can be related to changes in the origin and/or temperature of those waters. In contrast, the synthesis of terrestrial organic matter undergoes unique atmospheric processes that do not occur in aquatic environments, making isotopic analysis of terrestrial material somewhat more complex (Epstein and Yapp, 1977; Epstein and Hall, 1976; Sauer et al., 2000). However, if these processes are quantified and correlated to the isotopic behavior of algal material in a specific system, isotopic relationships between the two environments can be used to provide unique inferences into hydrologic processes of a specific region.
25 Terrestrial vs. Algal Biomass: Isotopic Distinctions Terrestrial plants undergo an added step of isotopic fractionation relative to aquatic organisms (Epstein and Yapp, 1977; Epstein and Hall, 1976; Sauer et al., 2000). Evapotranspiration in terrestrial leaves causes an isotopic enrichment in leaf water and thus leaf biomass relative to algal material (Epstein et al., 1977; Allison et al., 1985; Cooper and DeNiro, 1989). This is due to the fact that during biosynthesis, small amounts of waters in leaf waxes are lost to the atmosphere through evaporation, leaving behind leaf water that contains a relatively larger amount of deuterium due to the kinetic fractionation associated with evaporative processes (Pilson, 1998, Figure 6 presents a schematic of evapotranspiration). This source water is then incorporated into the plant biomass, leading to a relative isotopic enrichment in the D of terrestrial organic matter compared to the original source waters in that environment. Congruently, this relative enrichment is also observed between the isotopic composition of algal and terrestrial material. Figure 7 presents a general schematic showing these isotopic relationships. The extent of isotopic enrichment is dependent upon the rate of evapotranspiration, which in turn is dependent upon both the temperature and moisture content of the air surrounding the leaf at the time of organic synthesis, or the relative humidity (Burk and Stuiver, 1981). This research has shown that as relative humidity increases the level of isotopic enrichment in terrestrial plant biomass decreases.
26 Figure 6. Schematic of Evapotranspiration Processes in a Terrestrial Plant Leaf (From the Food and Agricultural Organization of the United Nations online resources, www.fao.org) Figure 7. Schematic showing the general isotopic offsets between source waters, aquatic biomass, and terrestrial biomass ( # p = biological fractionation).
27 Utilizing Hydrogen Isotopic Analysis to Quantify Relative Humidity The isotopic fractionation processes unique to terrestrial plant biomass allow for a novel approach to be used in the understanding of relative humidity conditions and their relationship to the isotopic dynamics in a given system. Figure 8 presents a generalized schematic comparing relative humidity levels to isotopic fractionation in terrestrial biomass. Because this a process unique to terrestrial material, it is possible to observe the difference between the D of algal and terrestrial material ( "! D (T-A) ) in a specific environment (assuming they are incorporating the same source waters) and use the variability in this offset to potentially make inferences into variability in relative humidity. Specifically, the theory presents itself that an increase in "! D (T-A would be an indicator of a decrease in relative humidity, and vice versa. Thus, the D of algal material and terrestrial material in a given system can potentially be used to quantify 2 distinct hydrologic parameters: 1) D of algal material can be used as an indicator of the isotopic composition of source waters, and 2) The offset ( "! D (T-A) ), or relative isotopic enrichment of terrestrial material relative to algal material in a system can be used to examine levels of relative humidity. Epstein et al. (1977) has generated models using laboratory analysis of the oxygen and hydrogen isotopic composition of organic material in an attempt to conceptualize the effects of evapotranspiration (relative humidity conditions) on isotopic enrichment of plant biomass. Burk and Stuiver (1981) have made direct comparisons from the isotopic composition of terrestrial plant biomass and relative humidity, but have not examined the potential for the offset between the D of aquatic and terrestrial biomass to provide similar information on the D composition of
28 atmospheric moisture. This project attempts to expand on these studies by implementing isotopic analysis in a natural setting in order to make direct in-situ comparisons between geochemical relationships in variable plant biomass and climatic effects. Bulk isotopic analysis in a modern system can provide useful insights into these relationships and show their connection to hydrologic and climatic fluctuations. This project will calibrate the bulk hydrogen isotopic behavior of waters, aquatic and terrestrial biomass to test the hypothesis that their geochemical relationships can be used to observe the hydrologic dynamics in Lake Tulane. However, in order to isolate these parameters in a sedimentary record, a molecular approach to analysis must be implemented. Figure 8. Generalized relationship between the level of relative humidity and the degree of isotopic fractionation observed in terrestrial plant biomass
29 Bulk vs. Molecular Hydrogen Isotopic Analysis When the isotopic composition of specific source materials is necessary for a climatic reconstruction, as with terrestrial and aquatic material in this study, relying on a bulk isotopic perspective can be useful, but presents inherent problems. A molecular approach to the isotopic analysis of organic materials can complement bulk isotopic analysis by providing a consistent comparison between parameters in a modern system, and by helping to resolve these specific components in the sediments. This is because bulk isotopic analysis of lacustrine sediments generates an isotopic value that is generally the result of combined sources of organic materials. Huang et al. (2002) discusses the fact that organic substances in most lake sediments are from multiple sources including plankton, terrestrial and aquatic plants, as well as microbes. Previous research (Estep and Hoering, 1980; Deniro et al., 1981; Sessions et al. 1999; Sauer et al., 2001; Huang et al., 2002) has shown that compound specific hydrogen isotopic analysis of organic materials can provide climatic and hydrologic information, including temperature and source-water origins, as well as wet/dry conditions. Problems with multiple sources of organic matter into a sedimentary environment can thus be avoided using this approach (Deniro and Epstein, 1981; Cooper and Deniro, 1989; Sessions et al., 1999; Sauer et al., 2001). Several compounds have been shown to be useful for this type of analysis, including sterols, hydrocarbons, and fatty acids. Sessions et al. (1999) implemented compound specific analysis of lipids spanning all three of these categories, as well as alcohols, triterpenoids, and other isoprenoids in an attempt to calibrate the biosynthetic processes associated with various compounds. This research indicated that Isotopic
30 variations within compound classes (e.g., n-alkanes) are usually less than ~50, however, there can be a large degree of variability between classes, even within the same organism. This demonstrates that it is imperative to select a single compound class for analysis in order to make potential comparisons in isotopic variability as a function of a non-biological factor. Sauer et al. (2001) has demonstrated that sterols can be used as an indicator of environmental water, however, the complex series of steps and large sample sizes required for these analyses bode for a compound class that may require less sample and a more straightforward approach to analysis. Additionally, the analysis of sterols allows for the possible exchange of hydrogens on and adjacent to carbon-carbon double bonds with hydrogens in sedimentary substances, which could alter original isotope signal, (Huang et al., 2002). Huang et al. (2002) has shown that the isotopic signature of fatty acids obtained from organic material can be used as indicators of both environmental water and potentially temperature. Specifically, this study demonstrated that the long chain fatty acid C 16 (palmitic acid) can be used as an aquatic biomarker, and that its isotopic signature can be directly related to that of the surrounding environmental water, and indirectly to temperature. Huang et al. (2002) also demonstrated that longer chain fatty acids such as C 26 and C 28 can be used as terrestrial plant biomarkers. In contrast to the analysis of sterols, palmitic acid is an abundant organic compound in late Quarternary lake sediments, does not contain carbon-carbon double bonds, and does not require large samples for analyzing its D values (Huang et al., 2002).
31 This study will apply the bulk isotopic relationships between source waters, algal biomass, and terrestrial biomass as determined from a modern calibration study (see relationships presented in Figure 7) to long chain fatty acid compounds in a sedimentary record. It should be noted that C 16 or palmitic acid, is found in all plant materials, however, decay processes remove this compound from terrestrial and bacterial components before they are deposited in the sediments (Huang et al., 2002). Therefore any palmitic acid analyzed in the sedimentary record can be interpreted as having an aquatic, algal origin. This parameter will be used to infer isotopic shifts in environmental water throughout a sedimentary record. The second parameter, "! D (T-A) provides the potential to quantify relative humidity using the offset between algal and terrestrial material, as suggested by the general isotopic relationships discussed above. This relationship has been can be applied to molecular compounds by observing "! D (C28-C16) which provides a compound-specific method with which to resolve how the offset between terrestrial and algal material varies throughout a sedimentary record. The modern calibration study of bulk isotopic relationships provides a unique introduction to the potential for using hydrogen isotopic analysis as a proxy for reconstructing hydrologic conditions.
32 Lake Tulane Background Figure 9. Location Map for Study Site at Lake Tulane, FL (modified from www.wearefla.com) Lake Tulane is a groundwater-fed, subtropical lake, approximately 89 acres in size, located at 23 39N, 81 20W, in Avon Park, FL (Figure 9). The lake is positioned along the Lake Wales Ridge geologic formation along the spine of the Florida Peninsula (Figure 10). The lake was formed due to the collapse of underlying limestone causing a relatively large sinkhole that became connected to the surficial Floridian aquifer system. A topographic map of this region (Figure 11) indicates that the lake sits at 117 feet above sea level, and a contour map of the position of the Upper Floridian Aquifer in this area
33 (Figure 12) indicates that the surface of the aquifer is at a depth of approximately 80 feet above sea level. The maximum depth of the lake is observed to be approximately 21 meters, or between 65 and 70 feet, supporting the concept that groundwater is flowing throughout the lower portion of the lake. The 1989 Highlands County Soil Survey indicates that soils in this area are classified as Astatula-Urban Land complex, 0 to 8 percent slopes. Previous studies (Grimm et al., 1993; Huang et al., unpublished data) have shown that this lake contains a high resolution, climatically sensitive sedimentary record extending back at least 100,000 years. Figure 13 presents data from Grimm et al. (1993) in correlation with archives from the North Atlantic and the GISP II ice-core record, showing hydrologic variability exhibited in sub-tropical North America and its relationship to thermal parameters in other locations throughout the geologic past. In this instance, relative oak and pine pollen abundance within the sedimentary record of Lake Tulane is used to examine large-scale changes in precipitation, exhibited as Heinrich Events, over the last 70,000 years. It is apparent that the sediments in Lake Tulane contain a detailed climatic and hydrologic record that can potentially define hydrologic variability in sub-tropical North America with other climatic parameters across the globe. More extensive geochemical analysis of these sediments can potentially provide additional data with which to quantify these moisture changes. Due to the low pH associated with the waters in this system there is no carbonate preservation, preventing oxygen isotopic analysis of carbonates. However, the lake sediments are abundant in both algal and terrestrial organic matter, making hydrogen isotopic analysis an ideal approach to performing climate reconstructions. The extremely
34 small catchment area (89 acres) of this lake enables relatively simultaneous deposition of algal and terrestrial organic matter, allowing for comparisons between these two parameters in the sedimentary record. Figure 10. Graphical representation of the location of Avon Park, FL, along the Lake Wales Ridge geologic formation. Map obtained from White, 1970.
35 Figure 11. US Geological Survey Topographic Map of the Region Directly Surrounding Lake Tulane in Avon Park, FL (www.terraserver-usa.com)
36 Figure 12. Potentiometric contour map of groundwater elevations throughout the Southeast US relative to sea level and direction of groundwater movement (from capp.water.usgs.gov)
37 Figure 13. Correlation between hydrologic and thermal archives in Lake Tulane and other latitudes (from Grimm et al., 1993; Bond et al., 1999; GISPII online data)
38 Chapter 2: Research Methods Preparatory and Field Methods The study site chosen for this analysis was Lake Tulane, located in Avon Park, FL. The field study portion of the project was conducted over an annual cycle beginning in September 2002 and ending in September 2003. Regional hydrographic data consisting of precipitation levels, atmospheric temperature, relative humidity, and chemical hydrology were obtained through the National Oceanic and Atmospheric Administration online at www.noaa.gov and the International Atomic Energy Agency online at www.iaea.org Regional and site-specific geologic information was obtained from the US Geologic Survey, the 1989 Highlands County Soils Survey, and previous research involving this lake system (Grimm et al., 1993; Huang et al., unpublished data). Chemical nutrient data for Lake Tulane were obtained from Floridas LAKEWATCH program ( http://lakewatch.ifas.ufl.edu ). These data are presented in the Results section of this document. Field measurements and procedures were performed by two scientists during the first week of each month. Access to the lake was granted by the city of Avon Park. The monthly sample set consisted of lake waters, particulate organic material isolated from the lake waters, and 5 distinct groups of terrestrial plant material (Oak and pine leaves, decayed macrophytic material, live macrophytic material, and marsh grasses). Aquatic samples were collected with a ten foot john boat, which was put in using a ramp on the west side of the lake. Terrestrial plant samples were obtained by hand on the south shore
39 of the lake. Figure 14 shows the location of all sampling areas in and surrounding Lake Tulane, as well as the location of the boat ramp entry. Figure 14. Lake Tulane field sampling locations. Green regions indicate the location of terrestrial plant sampling. The black star indicates the location for water and particulate organic material sampling, as well as the location of sediment coring. Lake Water, Particulate Organic Material, and Terrestrial Material Sampling Water samples were collected at six depths (surface waters, 1.5m, 3m, 6.1m, 12.2m, and 18.3m) within the water column at the location specified on Figure 14. Total water depth at this location was approximately 20 meters. The depth selected for bottom waters (18.3m) was chosen to avoid having suspended sediments near the lake bottom contaminate the samples. Water was collected using an 8L Niskin bottle lowered by hand from the surface. After bringing the water sample back to the surface,
40 approximately 2000mL was placed in a sealed, solvent-cleaned container and brought back to the lab at the University of South Florida. 1000mL of water for each sample was filtered through 47mm GF/F filters using a pressurized nitrogen gas filtration system in order to isolate the particulate organic material from the water. Filtered water was transferred into 100mL, clean nutrient bottles, capped, sealed, and placed in a refrigerator until ready for D isotopic analysis. Particulate organic material was collected from each water sample obtained throughout the annual study during the filtering process discussed above. Filters were then dried at 60 C overnight to remove excess water. Once dry, the particulate material was scraped off of the GF/F filters and placed in solvent-cleaned glass vials, capped, and sealed until ready for isotopic analysis. Pine and oak leaves, marsh grasses, and both living and decaying macrophytic materials were sampled at the location specified on Figure 14, on the same day that water and POM samples were collected each month. The oak and pine trees were located approximately 15 meters to the south of the southern shore of the lake. Marsh grasses were located approximately 1 meter to the south of this same area. Macrophytic material was collected to the north of the marsh grasses in approximately 0.3m of water. All materials were collected by hand from the same area of each plant every month and placed in sealed containers for transport. Samples were immediately brought back to the lab and placed in a drying oven (~60 C) to remove water in the plants. Once dry, samples were homogenized for isotopic analysis.
41 Sediment Core Sampling A 140cm sediment core was extracted from Lake Tulane in November of 2002 at approximately the same sampling location as the waters and particulate organic material obtained for the modern calibration. The 3 piston core was sampled with the aid of researchers from the University of South Florida. The core was lowered from a barge using integrated sections of tubing. The tube was pushed by hand into the sediments, capped, and brought to the surface where the both ends were sealed. The core was kept refrigerated until sampling. The sediments were divided into 5mm sections, placed in sealed containers that were pre-cleaned using organic solvents, and placed in refrigeration until ready for hydrogen isotopic analysis. Prior to organic extraction and bulk and molecular analysis, the samples were dried and ground for weighing purposes. Bulk Isotopic Analysis: General Procedures All bulk hydrogen isotopic analysis was performed using thermal conversion/elemental analyzer-isotope ratio mass spectrometry (TC/EA-IRMS, instrument accuracy for D) at the University of South Florida. All samples were converted to hydrogen gas through a pyrolysis reaction in the TC/EA reactor, set at a temperature of 1450 C. The reactor uses extreme temperature to pyrolyze the sample to hydrogen gas and carbon monoxide or graphite. This process allows the sample to be converted to its gaseous state, while preventing contamination from additional hydrogen that would be generated in a combustion reaction. Helium was used as the carrier gas for all sample runs, set at a pressure of 1.5 bars. For hydrogen isotopic measurements, the mass spectrometer
42 collects deuterium (D-H gas, mass 3) and hydrogen (H-H gas, mass 2) in collector cups using a magnetic field and measures the ratio between the two isotopes. This ratio is analyzed relative to a reference hydrogen gas in order to generate isotopic values presented in standard per mil notation. Two reference hydrogen gas injections were inserted into the beginning of every sample run prior to the sample gas entering the instrument. The reference gas was measured and calibrated in the lab using Vienna Standard Mean Ocean Water (VSMOW). Measured VSMOW samples exhibited a standard deviation of <2.1, which is within instrument error. The hydrogen reference gas exhibited an D isotopic value of .59, VSMOW An H 3 factor was determined before the beginning of each run to account for any He in the carrier gas, which can corrupt hydrogen isotopic measurements. The average H 3 factor for all sample sets was 7.3. Standard on/off procedures were performed prior to the analysis of unknown samples to confirm the precision of the instrument. Bulk Isotopic Analysis of Lake Waters Waters were transferred from 100mL nutrient bottles to 2mL vials and inserted into the TC/EA auto-sampler tray. VSMOW water was used with every run as a standard with which to determine the accuracy of hydrogen isotopic measurements. An auto-sampler using a 10uL syringe was programmed to inject 1uL of water for hydrogen isotopic analysis of VSMOW and lake waters. Between five and ten VSMOW samples were analyzed at the beginning of every run, and throughout the run five VSMOW samples were inserted after every 20 unknown injections to account for any isotopic drift that occurred. Between five and ten VSMOW samples were also inserted at the end of every
43 run. Water samples for each month were injected five times each for isotopic analysis. Drift corrections were performed as necessary by linear adjustment of unknown values directly correlated with linear drift of VSMOW throughout the run. Drift-corrected data was then averaged for each month to obtain the final values. Isotopic statistics and results for the waters, particulate organic material, and terrestrial plant material analyzed in this study are presented in the Results section of this document in standard per mil notation. Particulate Organic Material, Terrestrial Plant Material, and Sediments: Bulk Isotopic Analysis Due to the relatively high POM concentrations in the lake water, approximately 2.5L of filtered water was sufficient to obtain the proper amount of particulate material needed for bulk isotopic analysis. Between 700 and 1200ug of POM for each sample and between 70 and 100ug of homogenized terrestrial material for each sample was placed in a silver capsule for insertion into the TC/EA. Approximately 1000ug of sediment was utilized for each interval taken from the 140cm sediment core for analysis. Each of these samples were analyzed in triplicate for statistical purposes and then averaged for final values. The accuracy of hydrogen isotopic measurements was determined using benzoic acid, measured and developed in the lab as a running standard. Between five to ten benzoic acid samples were inserted into the beginning of each run, and five samples were inserted between every twelve unknown samples. Five benzoic acid samples were also added to the end of every sample run. Drift corrections were performed as necessary by linear adjustment of unknown values directly correlated with linear drift of benzoic acid
44 throughout the run. Drift-corrected data were then averaged for each month to obtain the final values. Total Lipid Extraction (TLE), Methylation, and Column Chromatography Approximately 1.5g of dried pine and oak samples from January and July were weighed for extraction. In addition, filters containing particulate organic material from the month of October were used as a representative of algal material in the lake. Between 1.0 and 3.0g of sediment was utilized from select intervals of the core for organic extraction. Total lipid extractions were performed on this material following the procedures of S. Wakeham and T. Pease at the Skidaway Institute of Oceanography (unpublished manuscript). The TLE was then evaporated down to 1-2mL for methylation. Samples were transferred to ~17mL culture tubes and evaporated to dryness under N 2 1mL each of 3N methanolic HCl, methanol (HPLC grade), and toluene were added to the culture tube, which was then flushed with nitrogen and capped. Tubes were heated for 30 minutes in a 100 C water bath. After cooling, 5mL of DI water and 4mL of hexane were added and mixed. The top hexane/toluene layer was removed and placed in a 50mL pearshaped flask. The remaining samples were extracted twice more with 5mL of hexane each time. Sample was then evaporated to near-dryness and stored in ~1mL of hexane for column chromatography. 5g of activated silica gel was added to a 1cm i.d. glass column. 2.5g of deactivated alumina was then added, and a small plug of quartz wool placed on top. The column was filled and drained twice with 25mL of hexane. The sample was then transferred to the
45 column using a Pasteur pipet and allowed to absorb into the alumina. Hydrocarbons were eluted using 25mL of 1:1 toluene:hexane. Methyl esters were then eluted using 25mL of HPLC methanol. Samples were evaporated to near-dryness and then stored in ~2mL of hexane. Gas Chromatography Each organic extract was analyzed in a Hewlett Packard gas chromatograph with a flame ionization detector in order to determine the relative abundance of the fatty acid compounds utilized in this research (C 16 and C 28 for algal and terrestrial material, respectively) and to examine the full molecular compound chromatographic traces of all fatty acids in each sample. A column compensation procedure (baking the GC oven at 310 C for approximately 2 hours) was performed before sample runs to ensure a clean column and to compensate for background material that may alter the baseline signal of the chromatogram. A 10uL syringe was used to inject all samples into a fused silica capillary column 30m in length with an inner diameter of 0.32mm. Ultra-pure helium was used as a carrier gas, kept at a constant flow of 1ml/minute for all runs and column compensation procedures. Compressed air and compressed hydrogen gas were used as the gaseous mixture that supplied fuel to the flame ionization detector. C 20 fatty acid, the standard discussed previously, was injected between 3 to 5 times before running unknown samples in order to confirm background levels, retention times and instrument accuracy. Between 1 and 3uL of each sample was then injected into the column for each sample run.
46 Compound Specific Isotopic Analysis Molecular analysis was performed using a Finnigan GC/TC interface coupled to a Finnigan Delta Plus XL mass spectrometer. An HP-5 capillary column 50 meters in length with an inner diameter of 0.32mm was used to separate the molecular compounds in the GC. A 10uL syringe was used to inject samples into the GC column. Oven temperature ramp for all runs began at 50 C for one minute, then was increased to 100 C at a rate of 15 C/minute. The temperature was left at 100 C for five minutes, and then was increased to 310 C at a rate of 7 C/minute. The oven was then left at this final temperature for 20 minutes before cooling down for the next sample run. Upon exiting the gas chromatograph column each compound was converted to it gaseous state in the pyrolysis reactor, set at 1450 C under the conditions discussed previously for bulk isotopic measurements. Sample gas was then introduced to the mass spectrometer for isotopic analysis of individual compounds. C 20 fatty acid, obtained from Indiana University, Department of Geological Sciences, Biogeochemical Laboratories (delta value .8), was used as an isotopic standard to evaluate the accuracy of isotopic measurements,. Each run focused on isolating and identifying the isotopic values of C 16 and C 28 fatty acid as representative of algal and terrestrial material, respectively. 3uL injections were made for sample analysis and 1uL for standard analysis. H 3 factors were determined before each run (average value 8.7), and standard on/offs were performed to determine reference gas precision. Samples were analyzed in triplicate for statistical purposes. Five standards were analyzed at the
47 beginning of every run, and three standards were inserted after every 6-9 samples to observe accuracy throughout the run. Hydrogen isotopic values for C 16 (algal material) and C28 (terrestrial material) were calculated relative to the standard. The retention time associated with C 16 was 1435 seconds, and the time associated with C 28 was 1732 seconds. Isotopic statistics and results are presented in the Results section of this document in standard per mil notation.
48 Chapter 3: Project Results Regional Hydrologic Data Hydrologic and climatic data obtained from the National Weather Service for the region in which Lake Tulane is located show clear seasonal trends in average temperature, precipitation, and relative humidity. Although this is regionally averaged data obtained from outside resources, its relevance to the data collected in the Lake Tulane study is such that it has been included in this Results section. Table 2 lists all regional hydrologic data analyzed for this study. Average temperature has a range of 11.4 C, with a maximum in July (27.3 C) and a minimum in January (15.9 C). Average precipitation exhibits a range of 6.38 inches, with a maximum during June (8.25 in) and a minimum during December (1.87 in). More than half of the annual rainfall occurs during summer months, from June through September (~57%). It should be noted that although summer months represent a large portion of annual precipitation, winter months (December, January, and February) still play a significant role in contributing moisture to the area, representing approximately 40% of the total rainfall in the area. Average monthly relative humidity exhibits a relatively smaller range than other parameters (6.5%), with a maximum during the month of August (75.5%) and a minimum during April (69%). Unlike temperature and precipitation, relative humidity remains fairly high during winter months. As stated previously, this is due to the fact that relative humidity is a function of both the temperature and specific humidity in the atmosphere. Thus even as moisture content drops during the winter season, lower temperatures lead to a calculated relative humidity that is similar to that observed during the summer. The implications of these trends will be elaborated upon in the discussion section of this document.
49 Table 2 Month Average Temp. (Celsius) Average Precipitation (inches) Annual Percentage of Rainfall per Month Average Relative Humidity (%) Jan 15.89 2.48 5.05 73.50 Feb 16.89 2.41 4.91 71.50 Mar 19.28 3.02 6.15 70.50 Apr 21.61 2.17 4.42 69.00 May 24.67 3.63 7.39 69.50 Jun 26.67 8.25 16.81 72.00 Jul 27.33 6.81 13.87 73.50 Aug 27.28 7.18 14.63 75.50 Sept 26.50 5.98 12.18 74.00 Oct 23.39 3.02 6.15 72.00 Nov 20.17 2.27 4.62 71.00 Dec 17.00 1.87 3.81 71.50 Monthly average temperature, precipitation, and relative humidity throughout the Lake Tulane region, obtained from the National Weather Service, representing monthly averages over the last 20 years.
50 Lake Biological Data Data obtained from Floridas LAKEWATCH program ( http://lakewatch.ifas.ufl.edu/ ) indicates that average chlorophyll concentrations in Lake Tulane are approximately 3.0 ug/L, average total nitrogen concentrations are 440 ug/l, and average total phosphorous concentrations are 6.0 ug/l. Additionally, the average Secchi depth (the depth at which a secchi disk is no longer visible from the surface) measured in the Lake is 5.1 meters below the surface. According to the Trophic State Index equations developed by Carlson (1997) these conditions indicate that the TSI for Lake Tulane is approximately 41, describing it as a mesotrophic system (TSI = 9.81(ln Chl a) + 30.6). The Secchi depth suggests that light is prevented from penetrating below approximately 5 meters due to either an increase in floating algae near this depth and/or variability in the concentration of other particles or fluctuations in water color. These zonal changes have the potential to be reflected in the isotopic composition of algal material throughout the water column, to be discussed in a later section.
51 Isotopic Signature of Lake Waters Table 3 ! surface 1.5m 3m 6.1m 12.2m 18.3m StdDev Average ! September ('02) -2.62 -2.62 ! October -3.56 -3.71 -4.40 -2.28 -1.98 -2.26 0.99 -3.03 ! November -1.98 -0.01 3.73 4.25 4.96 3.02 2.19 ! December -0.01 1.10 2.38 -0.54 2.93 -0.12 1.43 0.96 ! January -0.87 0.49 3.49 -0.28 1.88 3.35 1.85 1.34 ! March -0.23 1.24 -1.46 -3.72 -1.51 0.05 1.71 -0.94 ! April -5.91 0.42 4.38 -3.73 2.10 -3.81 4.00 -1.09 ! May 4.16 -0.56 -1.50 -0.27 3.17 0.45 2.25 0.91 ! June 0.87 -2.18 6.08 -0.91 0.83 -1.99 3.06 0.45 ! July -0.58 * -0.01 -2.52 -0.10 1.17 -0.80 ! August ('03) 0.60 -0.21 -0.63 3.59 -2.66 2.53 2.26 0.53 ! Average -2.04 -1.30 -0.68 0.30 1.73 0.86 1.81 -0.63 ! StdDev 1.50 3.40 3.44 3.09 3.28 3.71 1.07 2.59 ! Annual StdDev 2.62 ! Annual Average 0.04 ! D Lake Water, September 2002 September 2003 (VSMOW) Sample contamination and/or instrument error (-) September 2002 only surface water was obtained
52 Hydrogen isotopic data was collected from the waters in Lake Tulane from September 2002 August 2003. Relatively small fluctuations in the isotopic signature of waters as a function of depth were observed in Lake Tulane throughout the annual cycle. Fluctuations were higher throughout the water column during spring and summer months, with maximum variability observed in April, exhibiting an isotopic range of 10.4. During this month the D of the water is observed to exhibit the largest gradient in the upper section of the water column, fluctuating from .0 at the surface to 4.4 at a depth of three meters. The lowest isotopic gradient was observed during the fall and winter months, with a minimum during the month of October exhibiting an isotopic range up to only 4.6. Here the gradient in the upper 3.0 meters of the water column was minimal (~0.9), with the largest isotopic shift seen from 3.0 to 6.1 meters (1.1). These gradients may be an indication of stratification of the water column isolating certain depths from isotopic homogeneity, potentially due to microbial processes affecting the geochemistry of the waters. Figure 15 presents isotopic data from four months throughout the year (January, April, August, and October) as representative of seasonal fluctuations in the D of Lake Tulane waters with respect to depth. These plots show that when instrument error is taken into account the fluctuations observed with depth in the lake are relatively small, indicating that overall the isotopic composition of waters can be said to be relatively invariant with respect to depth.
53 Figure 14 January 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 -10 -5 0 5 10 15 April 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 -10 -5 0 5 10 15 August 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 -10 -5 0 5 10 15 October 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 -10 -5 0 5 10 15 Figure 15. Seasonal perspective of isotopic variability in D of lake waters over time at varying depths in Lake Tulane, FL. Dashed lines indicate the relatively small window if isotopic fluctuations and invariant geochemical behavior seen throughout the annual cycle (within a 10 range) In addition to observing isotopic fluctuations on a monthly perspective, isotopic variability was examined for each sampling depth throughout the annual cycle. Figure 16 presents this data. Slight variability is observed at each depth throughout the year. The largest fluctuations are observed at a depth of three meters during the transition into and out of the summer season (maximum shift of 7.6). The 1.5-meter sampling depth exhibits the smallest fluctuations throughout the year, with the largest isotopic shift observed to be an enrichment of 4.81 from October to December, however, this may be
54 an exaggerated change due to the lack of data from November for this depth (see Table 2). A slight trend toward more depleted values during fall and early winter months is observed in the surface waters and the 3.0 and 12.2-meter depths, however, this trend is not supported in the remaining depths in the water column. Standard deviations for all depths range from 1.72 to 3.36, all of which fall into the boundaries of instrument error (3) except for 3.36 (three-meter depth). As stated previously, while slight trends are observed during specific months, no strong correlations exist among depth or seasonality to suggest regular isotopic shifts throughout an annual cycle. Qualitative analysis of temperature gradients within the lake during sampling indicates that a thermocline is present at approximately 12 meters below the surface. However, isotopic data does not suggest that the geochemistry of the waters is significantly affected by this stratification. Congruent to the trends seen in waters with respect to depth, by taking instrument error into account the water column can be said to be isotopically homogenized, regardless of depth or season, and thus it is possible to incorporate monthly averages of all depths in order to make further comparisons to other biological parameters.
55 -10 -8 -6 -4 -2 0 2 4 6 8 10 October November December January March April May June July August D (VSMOW) surface 1.5m 3m 6.1m 12.2m 18.3m Figure 16. Hydrogen isotopic value of lake waters at all depths over the annual cycle, from October 2002 August 2003. Note that although some random variability is observed, the range is relatively low, allowing for depth-averaged water values to be used for further comparisons Figure 17 shows the average D of all depths for each month during the course of study. This data exhibits a range of 5.2 over the annual cycle, with a maximum of 2.2 during November and a minimum of .0 during October. This total range is observed to be very close to instrument error (3). The error bars clearly show the relatively small amount of variability in the average isotopic composition of the water column throughout the year. The average D of the waters is 0.04, a value representative of averages (between 5 and ) indicated by data obtained from the IAEA (Global Network of Isotopes in Precipitation, www.iaea.org ).
56 -10 -8 -6 -4 -2 0 2 4 6 8 10 September October November December January March April May June July August D () Figure 17. Monthly Depth-Averaged Isotopic Values of Lake Waters Over an Annual Cycle Particulate Organic Material All D values for the particulate organic material over the annual cycle are shown in Table 4. D values of particulate organic material exhibited a range of up to 24 over the annual cycle. Surface waters were compared to waters at a depth of 18.3 meters in order to examine these two zones in the water column of Lake Tulane (Figure 18). The same range (24) was recorded separately in both the surface waters and those at 18.3m. Both depths exhibited the largest variability during the late fall and early winter months, fluctuating from to over oneto two-month spans from October to December. The greatest difference observed between the two depths was recorded during
57 the month of November ( "! D of 24.03). In order to make general comparisons between algal materials and other parameters analyzed in this study, a plot of the depthaveraged POM throughout the year is presented in Figure 19. The average value of the D POM over the annual cycle for all depths was The overall range for the averaged values was observed to be 18.72, with a maximum during October (-123.7) and a minimum during June (-142.4). Although this range is somewhat large, for the purposes of this study the fluctuations in POM are small. It should be noted that runoff of terrestrial material into the lake waters may also make a contribution to the bulk samples that were analyzed. The isotopic composition of terrestrial material (discussed in detail below) is observed to be significantly enriched relative to the algal material. Thus, a maximum such as that observed during October may be related to an increase in the contribution of terrestrial organic matter to the suspended particulates that were collected from Lake Tulane. More comprehensive molecular analysis of the particulate organic material within the lake is necessary to resolve the source and relative contribution of organic matter to the waters. For the purposes of this study, the assumption will be made that the particulate organic material is comprised of biomass that is 100% aquatic in origin in order to make direct comparisons between waters, algae, and terrestrial material. Assuming a relatively constant isotopic value for the POM correlates well with the concept that the D of the POM is a direct reflection of the source water it formed in (here, the lake waters), which also show little variability in their average values over the annual cycle. The average offset between the D of POM and lake waters was indicating a relatively constant level of biological fractionation throughout the year.
58 Table 4 surface 18.3m Average DD October () -120 -127.40 -123.7 7.39 November -144 -119.97 -132 24.03 December -130 -140.75 -135.6 10.32 January -129 -129.64 -129.4 0.51 March -127 -129.63 -128.3 2.71 April -139 -138.6 May -132 -133.97 -133.2 1.60 June -141 -143.83 -142.4 2.81 July -134 -130.402 -132.4 4.08 August () -127 -127.2 Annual Average -132 D Particulate Organic Material October 2002 August 2003 (VSMOW) Missing values due to contamination and/or instrument error
59 Figure 18. D of Particulate Organic Material in the Surface Waters and at 18.3 meters in Lake Tulane, FL
60 -150 -145 -140 -135 -130 -125 -120 -115 -110 Oct Nov Dec Jan March April May June July Aug dD () Figure 19. Monthly Average D of particulate organic material (POM) in the water column
61 Terrestrial Material Table 5 Oak Pine Marsh Grass Mod. Mac. Dec. Mac. Average November ('02) -94.0 -70.0 -74.9 -89.5 -77.3 -81.1 ! December -79.5 -85.8 -93.9 -87.9 -106.7 -90.8 ! January -90.8 -89.8 -89.8 -68.0 -71.3 -81.9 ! March -83.3 -77.6 -71.3 -77.0 -77.3 -77.3 ! April -84.7 -63.8 -57.0 -53.9 -82.0 -68.3 ! May -90.2 -80.5 -82.6 -71.7 -75.9 -80.2 ! June -88.1 -78.0 -81.5 -81.2 -86.5 -83.1 ! July -93.8 -69.6 -89.0 -89.7 -80.7 -84.6 ! August ('03) -87.9 -87.1 -92.2 -75.7 -88.7 -86.3 ! Annual Average -88.0 -78.0 -81.4 -77.2 -82.9 -81.5 ! Annual StdDev 4.8 8.8 12.0 11.7 10.4 6.3 D of terrestrial material surrounding Lake Tulane, September 2002 September 2003 () Isotopic analysis of terrestrial materials surrounding Lake Tulane is presented in Table 4. As discussed previously, monthly samples of pine and oak leaves, modern and decayed macrophytic biomass, and marsh grasses were obtained from November 2002 August 2003. A relatively large degree of isotopic variability was observed both in a single species during the annual cycle and between species throughout the year (Figure 20). The largest degree of isotopic variability was recorded in the biomass of the marsh grasses, which exhibited their most enriched values during the month of April (-56.95)
62 and their most depleted values during December (-93.92), recording an overall range of 36.97. The average value for marsh grasses throughout the year was .68. Modern macrophytic material exhibited the second-highest degree of variability, exhibiting its most enriched values during the month of April (-53.85) and their most depleted during November and July (-90) recording an overall range of 36.15. The average value for modern macrophytic material throughout the year was -77.17. Decayed macrophytic material exhibited a lesser degree of isotopic variability throughout the year with the exception of the month of December (-106.73), which may represent a lack of homogenization in the sample before analysis. Removing this value from the sample set provides an average of Including December in the sample set alters the average to .94 (a change that is within instrument error), suggesting that the overall isotopic signature of decayed macrophytic material is generally consistent, regardless of one potentially anomalous value. Previous research (Grimm et al. 1993) utilizing pollen analysis of sediments in Lake Tulane have implied that oak and pine trees are the dominant sources of terrestrial pollen reaching the sediments, and thus the isotopic signature of the leaf biomass from these plants throughout the year is most relevant to this study. Oak leaves exhibited the lowest degree of variability among all terrestrial species, exhibiting their most enriched values during the month of December (-79.54) and the most depleted values during November (-93.97), recording an overall range of 14.42. Pine leaves also showed a comparatively lower degree of variability, exhibiting their most enriched values during
63 April (-63.81) and the most depleted during January (-89.79), recording a range of 25.98. Figure 21 provides an overlay of all species of terrestrial material throughout the annual cycle in order to more clearly show the relationships among species. All terrestrial samples indicate a slight seasonal trend, with more enriched values generally in the late spring and early summer and more depleted values during the winter, with the exception of a few outlying points. These trends will be further analyzed in the discussion section of this paper. It should be noted that although samples were homogenized before analysis, heterogeneous analysis due to an isotopically distinct portion of the leaf entering the instrument is still a possibility.
64 D of Oak Leaves Over the Annual Cycle -100 -95 -90 -85 -80 -75 -70 -65 -60 -55 -50 Nov Dec Jan Mar Apr May Jun Jul Aug D of Oak Leaves Over the Annual Cycle D of Pine Leaves Over the Annual Cycle -100 -95 -90 -85 -80 -75 -70 -65 -60 -55 -50 Nov Dec Jan Mar Apr May Ju n Jul Aug D of Pine Leaves Over the Annual Cycle D of Marsh Grasses Over the Annual Cycle -100 -95 -90 -85 -80 -75 -70 -65 -60 -55 -50 Nov Dec Jan Mar Apr May Jun Jul Aug D of Marsh Grasses Over the Annual Cycle D of Modern Macrophytes Over the Annual Cycle -100 -95 -90 -85 -80 -75 -70 -65 -60 -55 -50 Nov Dec Jan Ma r Ap r May Ju n Jul Au g D of Modern Macrophytes Over the Annual Cycle D of Decayed Macrophytes Over the Annual Cycle -110 -100 -90 -80 -70 -60 -50 Nov Dec Jan Mar Apr May Jun Jul Aug D of Decayed Macrophytes Over the Annual Cycle Figure 20. Bulk hydrogen isotopic signature of terrestrial plants surrounding Lake Tulane
65 -120 -110 -100 -90 -80 -70 -60 -50 -40 -30 -20 Nov Dec Jan Mar Apr May Jun Jul Aug D (VSMOW) oak pine grass Mod. Mac. Dec. Mac. Figure 21. D of all terrestrial material surrounding Lake Tulane, November 2002August 2003
66 Compound Specific Hydrogen Isotopic Results A select sample set of modern terrestrial and algal material was analyzed for molecular isotopic signatures based on sample availability, size, and time of the year needed to be observed. This sample set provides an initial perspective on the molecular isotopic relationships that may exist between the various biological parameters in and surrounding Lake Tulane. Pine and oak leaves were chosen as being representative of the main source of terrestrial material around the lake that reaches the sediments. Long chain fatty acids C 26 and C 28 were analyzed for both pine and oak samples collected during the months of January and June in order to examine extreme differences that may exist from winter to summer seasons. The retention time for C 26 was (2142 sec) for all runs, and the retention time for C 28 was (2235 sec). D of pine leaves exhibited little to no variability in the C 26 between January and June, recording isotopic values of .4 in January and .4 in June, the difference between the two being within instrument error. The C 28 in pine leaves exhibited variability of approximately 17 between the two seasons, recording a value of .8 in January and .3 in June. The isotopic composition of oak leaves was somewhat more depleted, exhibiting a value of .5 for C 26 during the month of January. The remaining oak samples that were analyzed produced signals too low to be statistically valid. C 16 fatty acid in particulate organic material obtained during the month of October was analyzed for isotopic distinction due to the relatively large sample sizes retrieved during
67 that month. The retention time for C16 during all runs was (1435 sec). POM C 16 exhibited a value of -149.5, more depleted than all terrestrial samples. Figure 22 shows the relationship of all samples analyzed for compound specific isotopic signatures. The full set of data collected is presented in Table 6. All gas chromatograph traces produced from these analyses depicted clear peak separation and a steady baseline, allowing for interpretation with only a small degree of error. Figure 22 Molecular isotopic results from long chain fatty acids extracted from terrestrial and aquatic biomass. Note, the dashed lines indicate the average bulk isotopic values for terrestrial and algal biomass, as discussed in previous results.
68 Table 6 D of Terrestrial and Aquatic Molecular Compounds From Lake Tulane, FL January D VSMOW June D VSMOW Pine C28 -118.80 -95.17 -116.93 -102.21 -117.64 -103.61 average -117.79 -100.33 Pine C26 -84.93 -89.67 -86.65 -85.71 -87.61 -78.18 average -86.40 -84.44 Oak C26 -127.61 -126.23 -128.62 average -127.49 October D VSMOW POM C16 -149.23 -150.08 -149.08 average -149.46
69 Chapter 4: Discussion of the Isotopic Relationships in Lake Tulane General Isotopic Systematics Over the Annual Cycle in Lake Tulane Figure 23. Monthly average D of lake waters, POM, and dominant terrestrial sources to the lake sediments. Offsets indicate levels of biological fractionation (algal and terrestrial material), and level of evaporative enrichment (terrestrial material) The annual monthly average isotopic data obtained from the waters, aquatic organic matter, and terrestrial organic matter in and surrounding Lake Tulane from September 2002 to September 2003 is presented in Figure 23. Three clear relationships/behaviors can be established from the data presented in this figure: 1) the isotopic composition of lake waters over the annual cycle is observed to remain relatively constant (near ~0), regardless of seasonality, 2) the isotopic composition of algal material is also observed to remain relatively constant over the annual cycle, potentially providing a direct relationship with the behavior of lake waters, and 3) there is a clear enrichment in the
70 isotopic composition of terrestrial material relative to the algal material in Lake Tulane due to the evaporative processes associated with terrestrial plant biosynthesis. The relatively small degree of variability in the isotopic composition of source waters and average algal material allows for a straightforward calibration of the isotopic fractionation between waters and algae, and the offset observed in terrestrial material due to evaporative enrichment. This discussion will focus first on the geochemical processes associated with each of these relationships, and second, on how these processes can potentially be used in the development of an isotopic proxy that will provide a tool to understand hydrologic and climatic conditions in subtropical North America during the present-day as well as on geologic time scales. Isotopic Behavior of Lake Tulane Waters Over the Annual Cycle Analytical results show that the isotopic composition of waters in Lake Tulane remains relatively constant, exhibiting minimal variability over the annual cycle (annual average = 0.04). Instrument error in isotopic measurements is 3, and the largest standard deviations observed in water with respect to depth and over the annual cycle were 4.0 and 3.36, both relatively close to instrument error. The standard deviation of all sample depths over the entire annual cycle was 2.62, within instrument error. Isotopic data from multiple fresh-water locations across the state of Florida have shown a similar lack of seasonality to surface water D. Specifically, the St. Johns River sampled by the US Geological Survey (mentioned in the introduction section of this paper; Coplen and Kendall, 2000) exhibits a standard deviation of only 2.2 among the data presented from multiple months during 1986 and 1987. The Alafia River, located at a similar latitude to
71 Lake Tulane, exhibits a standard deviation of only 2.3, and Fisheating Creek records a slightly higher deviation of 6.5, both incorporating multiple months during 1986 and 1987. Figure 24 presents these isotopic data, along with the depth-averaged isotopic data from Lake Tulane from September 2002 August 2003. It is evident that the isotopic signature of near-surface freshwater across the state of Florida falls into a range between and 5, despite seasonal changes or latitudinal variability. The invariant nature of the isotopic signature of waters in Lake Tulane can also be attributed to the groundwaterfed nature of the system. Constant input of isotopically homogenized groundwater into Lake Tulane also plays a role in generating the lack of seasonal variability observed in the data. Potential isotopic variability that may occur over the annual cycle in the precipitation entering the system would have less of an effect on the groundwater across the Avon Park area, which supplies the Lake Tulane basin. Thus it is the combination of precipitation over Florida that is relatively isotopically invariant, and the constant flow of isotopically homogenized groundwater into the lake that ultimately results in the lack of seasonality observed in the isotopic measurements of lake waters. Additionally, regional data from the Global Network of Isotopes in Precipitation (GNIP), a section of the International Atomic Energy Agency (IAEA), across North America also indicates that the isotopic composition of precipitation over Florida shows little variability over the annual cycle (http://isohis.iaea.org). Figure 25 presents these data during the months of January and August (data are averages from multiple stations from 1961-1999) in an attempt to show the maximum change in the isotopic composition of precipitation from winter to summer. The range predicted by the GNIP for the Florida
72 region correlates relatively well with measured values in Lake Tulane, showing enriched environmental waters across subtropical North America relative to higher latitudes. This range remains stable, regardless of seasonality. In contrast, higher latitudes to the northwest exhibit large fluctuations in the D of precipitation from winter to summer, potentially as much as 100 over the annual cycle. It can be asserted that over longer time scales these more northern air masses may play a more significant role in the hydrology of the Lake Tulane region as climate patterns shift (Katz et al., 2003), leading to more significant variability in the isotopic composition of precipitation entering the system. However, in correlation to the lack of variability observed in Lake Tulane waters during the sampling period, average monthly data from the IAEA supports the assertion that seasonality has little affect on the isotopic composition of environmental waters across the region. Furthermore, due to the small variability in the isotopic composition of lake waters with respect to depth over the annual cycle, it is possible to incorporate a depth-averaged perspective, as shown in Figure 23, in order to make inferences into potential temperature conditions, as well as comparisons with other parameters, such as aquatic and terrestrial biomass, over the sampling time period.
73 Figure 24. Isotopic data from USGS river sampling sites and the depth-averaged isotopic data from Lake Tulane over the annual cycle.
74 Figure 25. Average D of precipitation across North America (from http://isohis.iaea.org ) during the months of January and August. Note the lack of significant variability over Florida, correlating to the invariant nature of the geochemical signature of lake waters obtained from this study. Huang et al. (2002) has demonstrated that there is a direct relationship between the hydrogen isotopic composition of source waters and temperature. Specifically, the D/H
75 ratios of lake waters from multiple sites across North America at varying latitudes were determined and compared with mean annual temperature. The resulting relationship (with a correlation of R = 0.969) was used to quantify temperatures associated with a specific D. This relationship ( D H2O = -73.6 + 4.384 T C) was used with the isotopic results from the analysis of waters in Lake Tulane in an attempt to further show that the hydrogen isotopic composition of source waters can potentially be used to infer temperature. Using this equation, the annual average D of water in Lake Tulane (-0.19) suggests a mean annual temperature of approximately 16.8 C, or ~62 F. This temperature correlates relatively well with the average temperature of the lake, providing further evidence that the hydrogen isotopic composition of source waters can be used as a proxy for surrounding temperature conditions. Relationships Between Lake Waters and Aquatic Biomass It has been recognized that the isotopic composition of water that has diffused into aquatic plants is the same as the water in which they are submerged (Yakir et al., 1989), and that the hydrogen isotopic composition of aquatic plant biomass differs from source waters due to enzymatic reactions associated with the synthesis of organic matter from the water incorporated into the aquatic plant cells (Deniro, et al., 1981). For the case of partially submerged aquatic plants or terrestrial plants this is not the case due to an isotopic enrichment that occurs during evapotranspiration (Cooper & Deniro, 1989). However, the aquatic material analyzed in this study obtained from Lake Tulane is assumed to be representative of fully submerged algal organic matter, and thus will potentially reflect the geochemistry of surrounding source waters. As reported by
76 Sessions et al. (1999), there are several processes that can affect the hydrogen isotopic composition of organic matter in plants relative to source waters, including isotope effects coupled to biosynthetic reactions, and isotopically distinct hydrogen being added during biosynthesis. However, Epstein et al. (1977) has examined the fractionation processes associated with the uptake of water into plant cellulose using 18 O/ 16 O, and has asserted that, because this process occurs in a state of equilibrium, the 18 O/ 16 O ratio of plant cellulose of aquatic plants should have an approximately one-to-one correlation with the 18 O/ 16 O ratio of the water, regardless of the oxygen isotopic fractionation that might be associated with the synthesis of cellulose. This research incorporates both oxygen and hydrogen in the examination of plant cellulose and source waters, and shows that the hydrogen isotopic composition of aquatic plant biomass also is a direct reflection of the surrounding source waters. These studies provide clear evidence that it is possible to use the D of a whole host of proxies such as bulk plant cellulose, cellulose nitrate, organic lipids and other molecular compounds as indicators of the isotopic signature of the source water it formed in, and thus make inferences into the hydrologic dynamics of a given system (Sauer et al., 2000; Deniro et al., 1981; Sessions et al., 1999; Huang et al. 2002). Analytical results from Lake Tulane indicate that, similar to the isotopic composition of the lake waters, little variability is observed in the isotopic signature of particulate organic material over the annual cycle. Some variability is observed between the isotopic composition of POM in the surface waters relative to the bottom waters during the months of October and November, however, the lack of offset during any other month
77 during the sampling period indicates that these months may contain anomalous data. The month of October may also represent a slight increase in the contribution of terrestrial material to the suspended particulates in the system, leading to the difference observed. Figure 26 presents a plot of the isotopic offset between POM in the surface and bottom waters. It is clear that, with the exception of the month of October, the majority of the data shows little to no isotopic difference between the surface of the lake and a depth of 18.3 meters once instrument error is taken into account. This lack of variability correlates well with the small D differences in the lake waters over a variety of depths, as well as throughout the annual cycle. Examining the depth-averaged D of POM during the sampling year further indicates a small degree of isotopic variability over the annual cycle. Similar to the lake waters, the standard deviation of average POM during the year was observed to be 4.8, only 1.8 higher than instrument error. The important relationship to be examined for this calibration study is the relationship between the isotopic composition of lake waters and the associated aquatic biomass throughout the year. Due to the homogenous nature of the water column in Lake Tulane, we use the depth-averaged values for both parameters in an attempt to comprehensively observe general isotopic relationships. Figure 27 presents a plot of the average D of POM and lake waters over the annual cycle. The average offset between aquatic biomass and waters is -132.77, and this offset ranges by only 5.30 throughout the year, indicating a relatively constant level of photosynthetic fractionation ( p ), regardless of seasonality. Figure 28 provides a plot of the offset between algal material and lake waters during the sampling period. More specifically, the average apparent fractionation
78 between aquatic material and environmental water ( A/W ) was [ a/b = 1000( # a/b -1) where # a/b = ( a +1000)/( b +1000)]. This approach incorporates a more rigorous mathematical difference between isotopic ratios of algal organic matter and water rather than observing the general isotopic offset. This value had a range of 5.21 throughout the year, indicating relatively constant isotopic fractionation. Figure 26. Isotopic offset between the D of the particulate organic material in the surface waters of Lake Tulane relative to waters at a depth of 18.3 meters.
79 Figure 27.Average D of POM and Lake Waters Over the Annual Cycle in Lake Tulane. Note the relatively constant level of biological fractionation ( $ p) throughout the year (133) Figure 28. Isotopic offset between algal material and lake waters during the sampling period. Note that, within instrument error, the offset remains relatively constant during the year.
80 Estep & Hoering (1981) have examined the direct relationship between the D of bulk algal biomass and the source waters in which it formed. Using source waters with an initial D of approximately Estep & Hoering (1981) generated algal biomass within the controlled source waters in order to examine how autotrophic and mixotrophic growth affects both the organic material as well as the source waters. Their study produced initial algal cells ( A. quadruplicatum ) with a D value of generating a p of This value correlates very well with the that is observed between the algal biomass and source waters within Lake Tulane. A second species ( C. sorokiniana ) yielded an p of still within the general range of isotopic offset observed in Lake Tulane. Due to the fact that the Tulane study is incorporating algal material from a natural environment it is most likely a mixture of a variety of algal species. However, the isotopic offsets generated in the Lake Tulane study further complement the research performed by Estep & Hoering (1981), and provide clear evidence of quantifiable and constant fractionation between algal biomass and source waters. In a similar study, Epstein et al. (1976) examined the direct relationship between the D of cellulose nitrate in multiple plant species and the D of the environmental waters surrounding the plants. The results of this research provide further support of a direct and constant offset between the hydrogen isotopic composition of aquatic biomass and the source water in which it formed. The absolute magnitude of isotopic offsets observed in Epstein et al. (1976) was in general less than the observed in Lake Tulane, however our study has incorporated bulk plant biomass as opposed to cellulose nitrate,
81 and thus direct comparisons may not correlate. However, the relationships established by Epstein (1976) are clearly seen in the Lake Tulane results, as well as in the research by Estep & Hoering (1981) discussed above. Epstein (1976) concluded that, Should the D/H ratio of meteoric water at a particular site change in response to a change in climate, that ratio change would be reflected by a corresponding change in the D/H ratio of cellulose non-exchangeable hydrogen largely independent of the plant involved. In the case of Lake Tulane, climatic fluctuations leading to a shift in the isotopic signature of environmental waters are evidenced as negligible during the sampling period, leading to the constant and invariant offset observed between lake waters and aquatic biomass. Congruently, the groundwater-fed nature of the lake allows for a constant cycling of homogenized water through the system, further adding to the isotopically invariant water and aquatic biomass throughout the year. This scenario provides an ideal environment with which to make further comparisons with other parameters, such as terrestrial plant material, due to the lack of climatic fluctuations that may otherwise play a role in isotopic variability. In addition to the study by Epstein (1976), Huang et al. (2002) has also demonstrated a direct relationship between the D of palmitic acid (a molecular compound shown to be representative of algal material) and environmental waters. Huang et al. (2002) shows a line of constant fractionation between the two parameters exhibiting a correlation of R = 0.894 ( D PA = -167.0 + 0.939 D H2O ), providing further evidence that the hydrogen isotopic composition of aquatic biomass can provide direct insights into the composition of environmental water. Huang et al. (2002) also suggests that, using this relationship, it
82 is possible to directly correlate the isotopic composition of aquatic biomass to temperature using the equation discussed previously in this section. Due to the fact that this calibration has incorporated bulk isotopic analysis, a direct correlation to these equations cannot be made, however, the relationships that have been shown in Huang (2002), Epstein (1976), and Estep & Hoering (1981) are clearly evident here, and this approach provides a unique analytical method with which to examine the direct reflection of the geochemistry of environmental source waters in organic matter. Specifically, the results support the hypothesis that the isotopic composition of algal material is a direct reflection of the source waters in which it formed plus an added step of biological fractionation ( D (Algal) = D (H2O) + p ). It would be expected that added variables affecting the isotopic composition of terrestrial biomass ( D (Terr) = D (H2O) + p +Evapotranspiration) will cause an offset between algal material and terrestrial material, and this is in fact observed in the results. Isotopic Variability in Terrestrial Material In addition to testing the hypothesis that isotopic composition of aquatic biomass in Lake Tulane is a direct reflection of source waters, this study has also hypothesized that the isotopic signature of terrestrial material can provide useful information regarding relative humidity conditions. Multiple studies have shown that the hydrogen and oxygen isotopic signatures of terrestrial plant biomass differs from that of algal biomass due to the added factor of evapotranspiration occurring outside of submerged environments (Epstein et al., 1977, Cooper & Deniro, 1989, Burk & Stuiver, 1981, Sternberg et al., 1986, Yapp & Epstein, 1982). Enrichment in terrestrial plants does not occur when source water is
83 incorporated into the root structure because this is simply a diffusive process (Epstein et al., 1977). Enrichment only occurs at the leaf surface, and is a function of the relative humidity in the air surrounding the leaf (Burk & Stuiver, 1981). The analysis of terrestrial material performed in this study complements this research by providing an isotopically homogenous pool of source waters that is being utilized by both algal and terrestrial material, allowing for a direct analysis of the degree of enrichment observed in terrestrial biomass. It has been suggested through the examination of pollen abundance in the geologic record that pine and oak biomass have dominated the terrestrial input to the sediments in Lake Tulane (Grimm et al., 1993). This study examines these two parameters in correlation with the waters and aquatic biomass in order to calibrate the behavior of specific terrestrial inputs and determine their relationship to the environmental waters and hydrologic conditions surrounding the Lake Tulane region over the annual cycle. The average isotopic value of oak and pine biomass throughout the annual cycle is shown in Figure 29. It should be noted that the D of oak biomass is observed to be slightly more depleted than pine biomass throughout the majority of the year. This offset is most likely due to small differences in the biosynthetic pathways that may exist between oak and pine trees leading to the isotopic signatures preserved in the plant cellulose. Cooper & Deniro (1989) implemented a study of the isotopic composition of multiple species of terrestrial plants in an attempt to understand the physiological differences that may exist that could lead to isotopically distinct biomass forming from the same pool of source waters. Multiple factors were recognized as playing a role in fluctuating D of plant
84 cellulose between species. One important factor was the residence time of water within the leaf prior to biosynthesis, which can potentially lead to extended levels of evapotranspiration and thus a higher degree of enrichment. It was recognized that residence times were significantly lower in drought-tolerant plants relative to waterstressed succulents, leading to a potential enrichment in those plants with longer residence times. Although the species of oak and pine present around Lake Tulane were not included in the study by Cooper & Deniro (1989), it has been recognized by Grimm et al. (1993) that pine trees at this location thrive during potentially wetter periods, such as Heinrich events in the geologic past as well as during more abrupt climatic events such as the Medieval Warm Period during the Holocene. In correlation with the findings of Cooper & Deniro (1989), these more water-stressed plants may be associated with leaf water having a slightly higher residence time before biosynthesis than the oak trees at this location, potentially leading to the relative enrichment observed in this study. A more detailed analysis of the biosynthetic pathways associated with these specific species is necessary to comprehensively resolve this issue. However, the average difference between the two species was observed to be approximately 8, which is a relatively small offset in hydrogen isotopic variability and for the purposes of this study. Additionally, in order to apply the relationships determined in this calibration to the sedimentary record of Lake Tulane (see Application Section following this chapter) terrestrial material is observed as the combination of oak and pine cellulose deposited in the sedimentary environment. For these reasons, the remainder of this discussion will focus on the average isotopic value between oak and pine biomass over the annual cycle to make comparisons with the waters and aquatic material.
85 There is a clear enrichment in the D of these terrestrial samples throughout the entire year in comparison to the algal biomass from Lake Tulane discussed above (see Figure 23). Due to the fact that the isotopic composition of lake waters and precipitation in this system are relatively equal and invariant over the annual cycle (as indicated by the results of the calibration of waters and data presented from the IAEA), and that the entire nature of the hydrologic system within Lake Tulane is groundwater-fed, it can be asserted that the same pool of source waters is feeding both the aquatic and land plants. It has been determined that the isotopic composition of algal material in Lake Tulane is a direct reflection of source waters, undergoing only one step of photosynthetic fractionation ( $ p ). Due to the fact that these same source waters are being incorporated into terrestrial biomass surrounding the lake, a second conclusion can be made: The isotopic composition of terrestrial plants is a function of both $ p as well as the process of evapotranspiration, leading to the relative enrichment observed in the terrestrial plant cellulose. Figure 30 provides a schematic of these isotopic and hydrologic relationships.
86 Total Vegetation -140 -120 -100 -80 -60 -40 -20 0 Nov Dec Jan Mar Apr May Jun Jul Aug oak pine Avg offset: 8 Oak ( Avg -87 ) Pine ( Avg -79 ) Figure 29. Average D of Oak and Pine Biomass Over the Annual Cycle Surrounding Lake Tulane Figure 30. Schematic detailing the relationships between the waters, aquatic material and evaporative processes associated with terrestrial material in and surrounding Lake Tulane.
87 Burk & Stuiver (1981) have examined the degree of evaporative enrichment affecting the oxygen isotopic signature of terrestrial plants and have directly related this enrichment to the relative humidity of the surrounding environment at the time of biosynthesis. Burk & Stuiver (1981) incorporated a single D measurement of source waters and associated plant cellulose at a variety of locations at different latitudes. Burk & Stuiver (1981) confirmed that the isotopic enrichment of the cellulose was directly related to the relative humidity of the environment, and that the level of enrichment could potentially be used to calculate relative humidity (results from their research generated relative humidity values within one standard deviation of measured relative humidity). The isotopic data from Lake Tulane can be used to support this hypothesis by observing how both relative humidity and the isotopic composition of terrestrial biomass vary over an annual cycle at a single location. This approach allows us to provide further evidence that the isotopic signature of terrestrial plant cellulose reflects relative humidity.
88 Terrestrial Biomass in Lake Tulane & Relative Humidity Figure 31. Average D of Terrestrial Material and Average Relative Humidity Over the Annual Cycle in Lake Tulane. The relationship between the average D of oak and pine biomass with relative humidity over the annual cycle in Lake Tulane is shown in Figure 31. Isotopic trends observed in the terrestrial plant biomass correspond well with changes in relative humidity over the sampling time period. In general, more enriched terrestrial plant biomass is recorded during times of lower relative humidity, and vice versa. Burk & Stuiver (1981), Sternberg et al. (1986), Cooper & Deniro (1989) and Allison et al. (1985) have all demonstrated that drier conditions (lower relative humidity) lead to a higher degree of evaporative enrichment in the oxygen and/or hydrogen isotopic composition of terrestrial biomass. Conversely, during times of high relative humidity terrestrial biomass would be
89 expected to undergo a smaller degree of evaporative enrichment. This is indeed the case, as shown in Figure 31. Terrestrial biomass exhibits isotopic variability of up to 24, with a maximum enrichment during the month of April (-74.3) and the most depleted material recorded during January (-90.3). A second minimum of similar magnitude is observed to occur during August (-87.5). Congruently, relative humidity levels are observed to be the lowest during April (69%), and the highest during August (75.5%), with the second-highest humidity recorded during January (73.5%). Thus the maximum change in the isotopic composition of terrestrial biomass (24) is associated with a maximum shift in the relative humidity of ~6.5%. The two maximum levels of relative humidity, one during the summer and one during the winter, correlate well with the general behavior of moisture across Florida (Katz et al., 2003). Specifically, lower temperatures during the winter months decreases the atmospheres ability to hold water, so that even with a lower specific humidity (relative to summer months), the lower temperature results in a relative humidity that is still high (Peixoto, 1996). The lowest relative humidity, observed in April, is associated with mild temperatures and a low level of specific humidity, creating a situation where the available capacity for the atmosphere to hold water is relatively high, but the specific humidity is still low, leading to an overall decrease in relative humidity. Climatic implications for these results will be discussed in more detail below. It is apparent that a relatively strong relationship exists between the isotopic composition of terrestrial material and the relative humidity of the environment. However, the bulk D signal of oak and pine leaves presented here represents a combination of the isotopic signature of source waters and the effect of evaporative processes ( D (Terr) = p + Evapotranspiration). The unique environment of Lake Tulane
90 provides a situation where the effect of source waters ( p ) can be removed from this equation in an attempt to further define the direct relationship between the isotopic composition of organic matter and the relative humidity of the surrounding atmosphere. In order to take into account variability in the D of source waters as a separate factor from relative humidity, it is necessary to develop a parameter that incorporates the isotopic signature of both aquatic and terrestrial material. The analysis of the offset between the D of terrestrial and aquatic ( %! D (T-A ) material provides the potential to generate more quantitative insights into relative humidity conditions. By observing the difference between these two parameters, the effects of environmental waters (which are isotopically invariant in the modern environment and originate from the same source for both terrestrial and algal material) are eliminated, and the effects of evapotranspiration and relative humidity can be isolated and observed. Figure 32 shows the relationship that exists between the variability in %! D (T-A) and fluctuations in relative humidity throughout the year. These trends further support the relationships seen in the isotopic composition of terrestrial material and relative humidity, as well as the conclusions of Burk & Stuiver (1981). The data show that as relative humidity decreases the offset between terrestrial and algal material ( %! D (T-A) ) increases. These data confirm the concept that lower relative humidity leads to higher rates of evapotranspiration and D enrichment in terrestrial plants. Aquatic biomass is not affected by this process (Epstein et al., 1976 & 1977, Keeley et al., 1986), and has been shown to exhibit a relatively constant isotopic signature over the annual cycle (see Figures 27,28). Thus the fluctuations observed in the offset between terrestrial and algal material can be interpreted as a response to changing
91 levels of relative humidity. This relationship can be quantified, as shown in Figure 33. Specifically, relative humidity (RH) is equal to .19x + 82.1, where x is the %! D (T-A) This relationship generates a new climatic proxy ( %! D (T-A) ) for quantifying relative humidity. The correlation between relative humidity and %! D (T-A) is relatively good (R 2 = 0.72). This relationship indicates that the %! D (T-A) reaches zero at an average relative humidity of approximately 82%, at which point the temperature and moisture levels are such that evaporative enrichment at the leaf surface is minimal to none. This behavior has been observed in other research, such as Cooper & Deniro (1989), where significant increases in the isotopic enrichment of terrestrial plants are no longer observed after relative humidity levels have reached levels above ~78-81%. Average relative humidity in Lake Tulane does not reach this 82% threshold. Further regulated laboratory testing of the geochemical response of oak and pine trees to large changes in the humidity of their growth environment may help to more clearly identify this maximum level of enrichment in the Lake Tulane system. However, these data provide a quantitative approach to utilizing isotopic relationships in order to understand relative humidity conditions in a novel way. Specifically, previous research of Epstein et al. (1976,1977) and Sternberg et al. (1986) directly correlated the enrichment in terrestrial material to the isotopic composition of the source waters in which it formed. This approach allows for the quantification of isotopic enrichment in terrestrial material (and thus relative humidity) using two sources of organic matter algal and terrestrial biomass. The sedimentary environment is a mixture of these two sources of organic matter. This mixture contains a D signal of algal and terrestrial organic matter that can, using the calibrated
92 relationships described in the Lake Tulane study, be used to reconstruct the D of environmental waters and the level of relative humidity in the past. Figure 32. The isotopic offset between terrestrial and algal material and the average relative humidity in the Lake Tulane region over the annual cycle.
93 y = -0.1917x + 82.088 R 2 = 0.7178 40 50 60 70 80 90 100 110 -20 0 20 40 60 80 100 120 Maximum average relative humidity of ~82% Isotopic Offset ( D ) Figure 33. Correlation between the isotopic offset observed in terrestrial and algal material and the average relative humidity in the Lake Tulane region. Molecular Isotopic Relationships The mathematical relationships presented in this paper obtained from bulk D measurements provide the first step in developing new comprehensive proxies for examining hydrologic behavior in the geologic past over a variety of time scales. In order to isolate terrestrial and algal organic matter in sedimentary environments and thus calculate the %! D (T-A) during abrupt climate change in the geologic record a compound specific isotopic approach must be implemented. Sessions et al. (1999) has demonstrated that the molecular hydrogen isotopic analysis of organic compounds can be used to examine the fractionation between environmental waters and biomass. Sessions et al. (1999) also illustrate the importance of making comparisons between a single compound
94 class due to the fact that fractionation processes can differ between compound classes, resulting in sometimes significant isotopic distinctions between hydrocarbons, sterols, and fatty acids. In addition, research performed by Sauer et al. (2000) have demonstrated that observing the D of specific algal sterols provide[s] a viable means of reconstructing D/H of environmental waters. Huang et al. (unpublished data) have shown that fatty acids C 16 C 26 and C 28 can be isolated in the sediments of Lake Tulane and related to hydrologic and thermal conditions in the geologic past. Huang et al. (unpublished data) state that C 16 can be used as a biomarker for aquatic material and C 28 as a biomarker for terrestrial material. However, these data do not include modern plant samples with which to provide an initial calibration of present-day geochemical relationships to a specific climatic environment. Compound specific isotopic analysis of these fatty acids in select samples of our modern terrestrial and aquatic material will help to test the hypothesis that these relationships can be exploited on a molecular level. Additionally, the ability to resolve between terrestrial and aquatic organic matter using these two compounds allows for the %! D (T-A) to be observed on a molecular level. The results of the molecular isotopic analysis of organic material from Lake Tuilane show a clear enrichment in the D of the C 26 and C 28 in terrestrial material relative to the C 16 in aquatic material. The average of all C 28 analyzed (representative of terrestrial pine biomass) exhibited a value of .06, and the average of the C 16 (representative of algal material) exhibited a value of .5. These values show an average enrichment in the D of the terrestrial compounds of approximately 40. In order to compare the overall relationships observed during the modern calibration, the offset between the
95 annual average terrestrial material (-83) and the annual average algal material (-133) was determined to be 50, correlating extremely well with the 46 observed in the molecular compounds. Figure 34 presents these comparisons to show the similarity between the %! D (T-A) of the bulk and molecular data. It is clear that the enrichment due to evapotranspiration in terrestrial material is recorded in the molecular compounds. If the %! D (T-A) obtained from this averaged molecular perspective (46) is applied to the quantified relationship established in the calibration of the bulk materials (y = -0.1917x + 82.088), a relative humidity of ~73.8% is predicted correlating relatively well with the average relative humidity of the Lake Tulane region (~72%). It should be noted that there is also a significant difference in the isotopic signature of the D C 26 in oak and pine biomass. As observed in the bulk relationships, the oak C 26 records a D value that is 42.1 more depleted than the pine. The average of the C 26 in the pine biomass analyzed was determined to be .4, and the average of the C 26 in the oak biomass analyzed was determined to be .49. This difference is much greater than the average 8 offset observed in the bulk measurements, and is most likely due to an enhanced effect of the potentially longer residence times associated with the leaf water in pine leaves in this compound (Cooper & Deniro, 1989). Additionally, the amount of material used in the C 26 analysis may have potentially been lower than that which is required for reliable samples, potentially leading to a larger error in those results. C 28 may provide a more accurate molecular perspective for comparisons between terrestrial and aquatic biomass in Lake Tulane. In general, it is evident from this initial molecular perspective that offsets between C 16 and C 28 can be quantified and then observed in a sedimentary record in order to reconstruct variability in relative humidity in the geologic
96 past. Further detailed analysis of the isotopic relationships between the molecular structure of oak and pine in the modern Lake Tulane system is the next step to providing a truly quantitative method with which to make this type of reconstruction. Figure 34. Comparison between the %! D (T-A) in the average bulk terrestrial and algal material vs. the molecular terrestrial and algal biomarkers. Note that the two offsets are very similar, providing further evidence of the validity of the approaches implemented in this study. Implications for Assessing Climatic Variability The fluctuations in the D of terrestrial biomass throughout the year can be related to some of the major climatic phenomena associated with this region. Increases in the relative humidity during summer months, as indicated by the isotopic depletion of
97 terrestrial biomass and a lower %! D (T-A) during this time period, may potentially be connected to the seasonal migration of the Inter-Tropical Convergence Zone and the meridional shift of the Bermuda High (Haug et al., 2001, Katz et al., 2003, Davis et al., 1997). Summer months in subtropical North America exhibit a northward shift in the ITCZ and a westward shift of the Bermuda High, and both systems are associated with large amounts of precipitation and humidity (See Figure 1 of Introduction section). The potential increase in relative humidity during this time of year would result in lower levels of evapotranspiration an isotopic enrichment of terrestrial material. These fluctuations are clearly evident in the hydrographic data for this study site as well as the isotopic variability of terrestrial biomass and %! D (T-A) in the sediments. Conversely, the decreases in isotopic enrichment of terrestrial biomass and increase in the %! D (T-A) correlate well with contrastingly drier conditions resulting from a southerly shift in the ITCZ, and a westward shift in the Bermuda High during late winter/early spring (Haug et al., 2001, Katz et al., 2003, Davis et al., 1997). It is clear that the techniques developed in this calibration are a direct reflection of the hydrologic processes associated with the Lake Tulane region. Studies have shown that climate patterns such as the ITCZ have undergone longer-term changes in the geologic past (Haug et al., 2001, Correge et al., 2001, discussed in detail in the Application section of this document). The approaches presented in this paper provide a new and unique proxy with which to quantify these climatic fluctuations by observing relative humidity changes. It would be expected that the degree of variability observed in the %! D (T-A) over the present-day annual cycle would exhibit much larger
98 isotopic ranges in a sedimentary record that spans time scales associated with abrupt and large-scale climatic change. Additionally, long-term alterations in the origin and average temperature of source water entering the Lake Tulane system provide the potential to be reflected in the isotopic signature of algal material (and thus the $ P) in contrast to the relatively invariant composition recorded throughout the modern calibration. The approaches defined in this study allow for these two parameters to be accurately isolated, providing the potential to examine hydrologic conditions in a variety of lacustrine systems without the need for direct measurements of the isotopic composition of source waters. The ubiquitous nature of lake environments throughout the world allows for this approach to be implemented throughout an abundance of latitudes and environments. Summary of the Lake Tulane Modern Calibration Results from the modern calibration support the hypothesis that in lake systems, the D of algal material can be used as a direct indicator of the isotopic composition of the source water in which it formed ( D (Algal) = D (H2O) + p ). Specifcally, the p was determined to be in the bulk isotopic analysis. Similarly, the molecular isotopic analysis of C 16 fatty acid (used as an algal biomarker) yielded an p of approximately By examining the isotopic signature of source waters, it is possible to make inferences into hydrologic phenomena that may be affecting a specific region, such as Lake Tulane. These results indicate that the p in the Lake Tulane region remains relatively stable over the annual cycle, regardless of seasonal changes in precipitation or temperature. The groundwater-fed nature of the Lake Tulane system plays an integral
99 role in helping to generate this isotopically homogenized water column, as well as the water supplying the terrestrial plants surrounding the lake. Terrestrial material surrounding Lake Tulane incorporates the same source waters as the aquatic biomass as well as an additional evaporative component ( D (Terr) = D (H2O) + p + Evapotranspiration). By observing the difference between the two ( %! D (A-T) ), it is possible to observe and quantify the relative humidity affecting the terrestrial environment, thereby providing a new climatic proxy for a specific hydrologic parameter. Using the %! D (A-T) recorded for the months of sampling during this study, the following relationship was established to quantify relative humidity: RH = -0.1917( %! D (A-T) ) + 82.088. The data suggests that a maximum threshold in the degree of relative humidity (~82%) exists, at which point the offset between the D of terrestrial and algal material will not increase further. This level of relative humidity was not observed during the annual cycle, however, further laboratory testing of the specific biosynthetic processes associated with terrestrial plants and controlled levels of humidity will help to further constrain the maximum levels that may affect isotopic signatures. The fluctuations in relative humidity observed over the annual cycle in Lake Tulane can be attributed to the combined effects of the north/south migration of the Inter-Tropical Convergence Zone, the east/west migration of the relative position of the Bermuda High, and the presence or absence of northern storm fronts penetrating into subtropical North America. If the D of algal and terrestrial material can be isolated in the sediments using molecular analysis, it is then possible to use these calibrated relationships to reconstruct both the
100 isotopic composition of environmental water and the degree of relative humidity in the geologic past. Our initial molecular analysis, in conjunction with Huang et al. (2002) and Huang (unpublished data) indicates that these parameters can indeed be isolated and observed in the sediments using long chain fatty acids C 16 and C 28 as representative of algal and terrestrial material, respectively. This initial research suggests that the %! D (A-T) of these molecular compounds is similar to the offset recorded in the bulk isotopic approach, further supporting the concept that the hydrogen isotopic analysis of organic matter in lake sediments provides quantitative information regarding hydrologic conditions. More rigorous molecular analysis of modern plant and sediment samples in conjunction with detailed hydrologic and climatic records and archives will help to further resolve the initial quantitative relationships established in this calibration.
101 Chapter 5: Applications of Research Modern calibration studies are essential in providing an understanding how present-day climatic fluctuations can affect the isotopic behavior of organic matter and environmental waters. Obtaining a detailed understanding of these fluctuations provides the potential to make inferences into climatic variability in the geologic record. Specifically, the relationships between the hydrogen isotopic composition of organic matter and lake waters observed in this study have provided insights into the role of relative humidity fluctuations on the chemical composition of these materials. These relationships can now be applied to a sedimentary record in order to examine hydrologic change in subtropical North America during the geologic past. Objectives of the Application Study 1. Test the hypothesis that the isotopic proxies established in the modern calibration (that the isotopic composition of algal material is a direct reflection of source waters, and that relative humidity is directly related to the !" D (T-A) ) can be applied to the bulk and molecular isotopic analysis of a 140cm core from Lake Tulane to examine abrupt climatic fluctuations (focusing on three distinct events) over the last 2,000 years a. Examine the effect of the anthropogenic eutrophication of the Lake during the 1930s and the variability in the D of sediments. b. Examine the influence of potential decreases in temperature and relative humidity during the Little Ice Age (LIA).
102 c. Compare isotopic trends observed during the LIA with potential fluctuations in D during the Medieval Warm Period (MWP) as temperatures and relative humidity undergo a relative increase. 2. Test the hypothesis that the same proxies established in the modern calibration can be applied to compound specific hydrogen isotopic analysis of long chain fatty acids C 16 and C 28 from and 18m-long sediment core from Lake Tulane representing the last ~70,000 years (data from Huang, unpublished). a. Compare and contrast the isotopic behavior associated with more rapid climatic events (LIA and MWP) with long term climatic change related to the last glacial/interglacial transition. b. Analyze the isotopic offsets and potential benefits of implementing compound specific analysis on lacustrine sediments.
103 Geochemical Analysis of Abrupt Climate Change in Subtropical North America Figure 35. Bulk hydrogen isotopic ratios of sediments from Lake Tulane representing the last ~2000 years. Note the three time intervals that this application will focus on. The results of the bulk isotopic analysis of the 140cm core extracted from Lake Tulane are presented above in Figure 35 (Data also presented below in Table 7). An age model obtained from previous sedimentary studies of Lake Tulane (Grimm et al., 1999; Swarzenski et al., unpublished data) was utilized in order to correlate isotopic fluctuations in the sedimentary system with age (Figure 36). The 140 cm core was determined to contain sediments representing approximately the last 2000 years. Specifically, sediments above a depth of 30cm were determined to have accumulation rates of 0.42 cm/yr, and sediments below this depth were determined to have
104 accumulation rates of 0.05 cm/yr. This variability in accumulation rates can be attributed to the eutrophication of the lake during the early to mid 1900s, leading to an increase in algal production and overall sediment deposition, to be discussed in detail below. The results indicate that three distinct time periods exhibit unique isotopic excursions. Figure 36. Pb 210 analysis from a sedimentary core taken from Lake Tulane (from Swarzenski, USGS unpublished data)
105 Table 7. Bulk D From the 140cm Lake Tulane Core Date (CY) D/H (VSMOW) Date (CY) D/H (VSMOW) 2003 -106.620 1468 -94.548 2002 -106.341 1438 -98.914 1991 -107.173 1428 -96.091 1986 -113.370 1418 -97.830 1982 -99.147 1398 -91.383 1977 -100.953 1378 -93.548 1972 -100.267 1358 -91.747 1967 -98.968 1318 -89.758 1963 -88.785 1278 -93.290 1958 -91.315 1218 -93.646 1955 -87.669 1198 -89.224 1953 -85.494 1178 -94.089 1951 -80.410 1158 -97.241 1948 -81.586 1138 -94.119 1946 -73.195 1128 -91.228 1943 -80.439 1118 -99.354 1941 -71.288 1108 -90.943 1939 -78.992 1098 -93.840 1936 -68.737 1088 -92.020 1923 -78.806 1078 -103.568 1883 -85.729 1068 -92.880 1843 -86.812 1058 -99.063 1803 -90.939 1048 -97.431 1783 -73.452 1038 -102.277 1763 -85.671 1028 -97.147 1738 -64.795 998 -105.256 1728 -69.665 988 -96.192 1723 -84.741 978 -95.468 1708 -77.901 968 -96.883 1683 -79.505 958 -105.402 1668 -79.868 948 -98.606 1663 -85.931 918 -108.080 1648 -80.625 878 -107.226 1643 -77.037 838 -107.548 1628 -82.753 758 -105.020 1623 -84.679 718 -101.810 1608 -82.498 678 -102.794 1598 -80.684 638 -97.239 1558 -84.145 558 -98.516 1548 -87.615 458 -93.424 1538 -89.975 363 -93.632 1528 -92.537 263 -95.459 1518 -87.716 163 -97.978 1508 -91.892 63 -104.465 1498 -94.844 -37 -100.988 1488 -95.167 -117 -105.636 1478 -92.653
106 Anthropogenic Influence on Lake Tulane The first shift occurs from approximately 1937 to the present day surface sediments, and records a depletion in deuterium of 44.3 over the last 58 years. The modern calibration provides two end-member D values associated with present-day conditions that can be used as a baseline to analyze this trend. Specifically, the average bulk isotopic measurements of particulate organic material (algal) are approximately (potentially incorporating some degree of terrestrial contribution, as stated in previous sections), while the average value of the dominant terrestrial material is approximately The general isotopic shift observed in the sedimentary record from 1937 to the present shows a change from bulk D values between and in the mid 1900s towards more depleted values below during the latter part of the century and measured at ~-114 in the present day surface sediments. This isotopic shift potentially represents an increase in the overall contribution of algal material to the lake sediments during the period of eutrophication. Isotopic signatures recorded during the early and middle 1900s indicate that terrestrial material was the dominant source of organic matter to the sediments (falling into a range near the average of recorded in the calibration study). As algal production increased, its contribution to the sedimentary record became larger, and organic matter that was more depleted (closer to ) began to enter the system. Specifically, using the two endmember values mentioned above that represent the baseline present-day signatures of terrestrial and algal material, todays sediments represent an approximate 60% contribution of algal material to the sedimentary environment (y = -1.9937x 165.98).
107 These relationships are summarized in Figure 37. The same baseline value for terrestrial material (-83) indicates that climatic conditions during the 1950s were similar to today, and that algal production was near zero. However, the data indicate that the isotopic signature of the sediments prior to the 1950s (from approximately 1936 to 1951) reached levels that were up to 15 more enriched than even todays baseline value for terrestrial material. If the assumption is made that virtually no algal material was entering the system during this time, then the fluctuations observed here must be attributed to small changes in hydrologic conditions. Specifically, decreased relative humidity during this time provides the potential for the isotopic enrichment observed. Historical records from the region indicate that the development of a railroad and an increase in residential structures around the lake beginning in the late 1930s provided the potential for increased nutrient runoff into the system leading to increased algal production and a general eutrophication of the lake (online reference www.apfla.com).
108 Figure 37. End-member schematic showing the calculated D values for terrestrial and algal material and the associated percent-contribution of algal material necessary to alter isotopic signatures Molecular Analysis of Recent Sediments In addition to bulk isotopic analysis of the sediments from Lake Tulane, an initial molecular analysis of select samples through the 140cm core was performed in order to further examine the changes in the relative contributions of terrestrial and organic matter to the sedimentary record. Specifically, gas chromatographic analysis of fatty acid methyl esters was implemented in order to examine variability in the ratio of algal to terrestrial material using C16 and C28 as biomarkers (analytical methods are the same as those discussed in the previous Research Methods section). Table 8 presents the results of this analysis.
109 The percentage of algal material in the surface (present-day) sediments correlates very well with the percentage calculated using the end-member mixing model (both at approximately 60%). Recent data within the last ~75 years indicate an increase in the percentage of algal material entering the sediments, suggesting that a peak in algal production was reached prior to present day conditions. Data from the ~100 years before present show a clear trend to lower percentages of algal material entering the system, leading to terrestrially dominated sediments. However, this molecular data suggest that there was still a relatively significant amount of algal production prior to the development around the lake and the eutrophication of the system (between 7.6 and 47.1%). Figure 38 provides a plot of the percentage of algal material in the sediments calculated from the molecular data.
110 Table 8. Results of Molecular D Analysis of Tulane Sediments Calendar Yrs. C16 area C28 area C16/C28 % Algal 2003 168758.0 115149.0 1.5 59.4 1991 69456.0 32623.0 2.1 68.0 1960 101220.0 34839.0 2.9 74.4 1955 72392.0 19534.0 3.7 78.8 1951 73743.0 22102.0 3.3 76.9 1963 16512.0 83525.0 0.2 16.5 1863 37684.0 67132.0 0.6 36.0 1803 43952.0 49287.0 0.9 47.1 1683 7882.0 47387.0 0.2 14.3 1663 9135.0 51171.0 0.2 15.1 1603 6453.0 33858.0 0.2 16.0 1543 18750.0 45445.0 0.4 29.2 1523 278443.0 404220.0 0.7 40.8 1503 23854.0 45043.0 0.5 34.6 1403 27550.0 68717.0 0.4 28.6 1243 14712.0 51778.0 0.3 22.1 1163 17505.0 29855.0 0.6 37.0 1063 64324.0 208103.0 0.3 23.6 983 31496.0 383576.0 0.1 7.6
111 Figure 38. Percent of algal material entering the sediments of Lake Tulane, calculated using molecular analysis of C16 and C28 fatty acids The isotopic data from the sedimentary environment during the last ~70 years show a potential peak in algal production during the 1950s, near 80%. This time period correlates with the enriched bulk D values observed in the sediments (~-65), in contrast with the more depleted signatures that have been associated with algal material in this study. Using the end-member relationships described above, an 80% algal contribution would have to be associate with a contribution of terrestrial material containing an isotopic signature of ~ -43, far more enriched than any of the material analyzed in the modern calibration study. This would suggest that extremely drier conditions were associated with this time period, leading to the large degree of enrichment in terrestrial material. Alternatively, hydrologic conditions during the mid
112 20 th century may have been associated with a change in the origin of source waters to the Lake Tulane system from a more isotopically depleted region to a more enriched region, leading to an enrichment in the isotopic signature of aquatic material. The modern calibration has shown that seasonal migration of the ITCZ has a potential direct effect on hydrologic conditions in subtropical North America, and provides the potential to contribute more southern, and thus more enriched (from www.iaea.gov ) waters to the system. Without a full molecular isotopic analysis of all Lake Tulane sediments across this time period a comprehensive resolution cannot be made, however, the data presented here provide an initial perspective into hydrologic and biologic conditions surrounding the lake during the 20 th century. Following the middle of the century, the data suggest that increased algal production in the lake gradually led to an overall 60% contribution of aquatic material to the sedimentary environment, as exemplified by the mixed D value of in the surface sediments. The Little Ice Age and Medieval Warm Period The second major isotopic shift recorded in the bulk hydrogen signature of the sediments occurs from approximately 1440 to 1740 calendar years, known from previous research as the Little Ice Age (Hughen et al. 1996; deMenocal et al. 2000; Correge et al. 2001; Haug et al. 2001; Lloyd Keigwin 2003). The shift observed in the geologic record during this time exhibits an enrichment in deuterium levels from to The molecular data presented above indicate that, prior to development around Lake Tulane in the mid 20 th century, terrestrial material was the dominant source of organic matter to the sediments (potentially as large as 93%). Although algal production still represents a
113 portion of the organic matter in the system, the bulk hydrogen isotopic analysis can be used in conjunction with the geochemical behavior of terrestrial material to make general assertions into the climatic forcing that has led to this isotopic shift. It has been shown that the relative enrichment in the D of terrestrial material corresponds to the degree of evapotranspiration occurring throughout the plant biomass, which is directly related to relative humidity. The fluctuations shown during this time period can be used to infer that relative humidity levels potentially decreased throughout this time window, gradually leading to increased levels of evapotranspiration and thus an increase in the hydrogen isotopic values of terrestrial organic matter. Specifically, some assumptions can be made to potentially quantify this flux in relative humidity. Due to the bulk isotopic perspective used for analysis of this core, a quantified !" D (T-A) cannot be resolved. However, if we assume that the isotopic signature of source waters during this time was similar to that observed in the modern calibration, then a value of can be associated with algal material during the Little Ice Age. Using this value, and assuming the highest contribution of terrestrial material observed in the molecular data (93%) the equations for relative humidity derived in the calibration suggest that a value would indicate a drop in relative humidity to ~68%, or approximately 5% lower than todays annual average conditions. Climatic evidence for subtropical North America also suggests that cooler and drier time periods are also associated with the penetration of more northern storm fronts into the region (Davis et al., 1997, Katz et al., 2003). These air masses provide the potential to contribute more depleted source waters to the system, leading to even more depleted algal material
114 contributing to the sediments. Thus, the offset assumed above may be an underestimate of the true conditions present during the Little Ice Age. Lower relative humidity values would be calculated as the isotopic composition of aquatic material becomes more depleted. A more comprehensive compound specific isotopic analysis of these sediments will help to clarify these issues. Studies have observed a shift to cooler temperatures and drier conditions during this time period (Hughen et al. 1996; deMenocal et al. 2000; Correge et al. 2001; Haug et al. 2001; Keigwin 2003). Correge et al. (2001) has utilized the isotopic signature of corals from the southwest tropical Pacific to reconstruct sea surface temperatures in the geologic record, suggesting that, from 1701 to 1761, surface temperatures were on average 1.4 C cooler than during the past 30 years. Peter deMonecal et al. (2000) implemented the isotopic analysis of marine sediments off West Africa to show an even greater decrease of 3-4 C in sea-surface temperatures between 1350 and 1850 A.D. Closer to the Lake Tulane region, Haug et al. (2001) has shown that decreases in titanium percentages in sediments from the Cariaco Basin during the Little Ice Age are related to decreased levels of runoff due to precipitation minima at this time. Historical records of the Lake Tulane region have shown that decreases in temperature and precipitation coincide with decreases in relative humidity. Our application demonstrates that the continental response to the Little Ice Age is recorded in the sediments of Lake Tulane as changes in relative humidity.
115 The third major isotopic shift observed in the Lake Tulane geologic record occurred from approximately 700 to 1100 calendar years, known from previous research as the Medieval Warm Period (Broeker 2001; Haug et al. 2001; Keigwin 2003). This isotopic excursion exhibited a trend opposite that of the Little Ice Age, represented by an enrichment in the bulk D of the sediments to values as high as -108. This trend indicates that relative humidity levels may have increased during the described time period, leading to lower levels of evapotranspiration and thus a smaller degree of evaporative enrichment in the isotopic signature of terrestrial plant biomass. If the assumption is again made that source waters during this time were similar to today, then a value of (assuming a 93% contribution of terrestrial material) would be associated with a relative humidity of ~76.8%, or approximately 3-4% higher than todays average conditions. However, this may also be an underestimate of relative humidity increases due to the fact that changes in source waters during this time may have been associated with more southern air masses, leading to more enriched D values, in contrast to the conditions during the Little Ice Age. Previous research discussing the Medieval Warm Period complements the Lake Tulane data by providing evidence for increases in temperatures and a trend to wetter conditions from approximately 800-1200 A.D. (Broeker 2001; Haug et al. 2001; Keigwin 2003). Keigwin (2003) has examined the isotopic signature of marine sediments from the Sargasso Sea, providing sea-surface temperature reconstructions that indicate a 1 C cooling of SST during the Medieval Warm Period (~1000 calendar years ago) relative to today. Additionally, records from the Cariaco Basin presented in Haug et al. (2001) also
116 indicate that lower levels of precipitation during the Little Ice Age led to decreased terrestrial runoff and lower percentages of titanium in the marine sediments, correlating well with the evidence suggesting drier conditions in the Lake Tulane sediments. The opposite trends that are observed during the Medieval Warm Period in the Lake Tulane sediments are also recorded in the titanium data from Haug et al. (2001). Figure 39 provides a plot overlay of the titanium data from Haug (2001) with the bulk isotopic data from Lake Tulane. It is clear that the timing of drier conditions in the Cariaco Basin coincide with the potential lower relative humidity observed in the Lake Tulane area, and that the warmer and wetter conditions indicated in the titanium data coincide with the higher relative humidity in Lake Tulane. Haug (2001) has attributed this hydrologic variability, in part, to the migration of the Inter-Tropical Convergence Zone. A southward migration of the ITCZ during the Little Ice Age would potentially lead to lower levels of precipitation and humidity, as discussed above. In contrast, a northward migration during the Medieval Warm Period would exhibit itself as higher levels of precipitation and humidity. The correlations between Lake Tulane and the Cariaco Basin suggest that changes in regional hydrologic conditions are a likely reason for the variability in both titanium percentages and geochemical signatures. Additionally, temperature reconstructions during this period (deMenocal et al. 2000; Correge et al. 2001) provide evidence of cooler conditions during the Little Ice Age on a more global scale, indicating that the trends observed in Lake Tulane can be characterized as hydrologic evidence of a rapid global climatic event. A more detailed, compound specific analysis of these Lake Tulane sediments will help to further resolve the potential role of aquatic contributions to the sedimentary record and more accurately quantify the
117 shifts in relative humidity observed over from the Medieval Warm Period, through the Little Ice Age and into present-day conditions. Figure 39. Correlation between fluctuations in the percentage of titanium in marine sediments and the D values of Lake Tulane sediments. Note that drier conditions in the LIA are evidenced by lower titanium levels and enriched D, potentially as a result of the southward migration of the ITCZ, and vice versa for the Medieval Warm Period.
118 Molecular Analysis of the Long-Term Sedimentary Record Molecular analysis of organic materials is essential when attempting to resolve variations in the isotopic composition of terrestrial and algal material in a sedimentary system. While the above application of bulk isotopic analysis and basic gas chromatography provides useful insights for this particular lacustrine system, the isotopic analysis of specific compounds that have been isolated in the sediments provides a more quantitative and accurate approach to examining these parameters. The results of the modern calibration have shown that fatty acids C 16 and C 28 can be used as biomarkers for algal and terrestrial material in sedimentary archives, and that the relationships between source waters, algal material, and the !" D (T-A) can be applied to the sediments in order to more comprehensively resolve fluctuations in relative humidity and environmental waters in the past. Molecular isotopic data describing the D values of C 16 and C 28 fatty acids in an 18-meter sediment core from Lake Tulane was obtained from Dr. Yong Song Huang at Brown University. This 18-meter core was determined to include the last glacial/interglacial transition, representing approximately the last 70,000 years. The relationships between the hydrogen isotopic behavior of algal and terrestrial material that were obtained from the modern calibration were applied to the isotopic fluctuations observed in these compounds.
119 Figure 40 shows a plot of both C 16 and C 28 extending back 70,000 years. This transitional time period from glacial conditions to todays hydrologic environment manifests itself in three specific time windows: 1) The Holocene, representing approximately the last 10,000 years, 2) The period of deglaciation, extending from approximately 10kyr to 18kyr, and 3) The last glacial period, extending from 18kyr to the end of the core data at 70kyr. Figure 41 shows the calculated relative humidity over this time period, using the !" D (T-A) and the equations derived in the modern calibration. Variability in the isotopic signature of C 16 (algal) and C 28 (terrestrial), as well as relative humidity, will be examined in an attempt to understand hydrologic change over longterm geologic time scales.
120 Figure 40. C 16 and C 28 fatty acids in Lake Tulane sediments over the last ~70,000 years, obtained from Dr. Y. Huang, Brown University. Note the three time periods of interest: The Holocene, degalaciation, and the last glacial period.
121 Figure 41. Calculated relative humidity and !" D (T-A) from the last glacial period, through deglaciation and into the Holocene. Note the wide variability during each period, and the overall transition from glacial to interglacial. Climatic Shifts in the Holocene and Through Deglaciation A major shift in the isotopic composition of both the algal and terrestrial biomarkers is recorded from the last glacial period into the Holocene (discussed below). However, less intense but still significant variability is also observed through the Holocene and during the period of deglaciation. Fluctuations of up to ~9% from todays conditions are observed during the last 10,000 years (discussed below). These data suggest that relative humidity conditions were generally higher than present day. However, it should be noted
122 that the relationships used in the modern calibration study to derive the relative humidity equation were associated with bulk D measurements. There may be some degree of difference in the relative magnitude of isotopic offset between molecular compounds in the modern system. Further compound specific analysis of modern samples will help to resolve this issue. However, the trends observed provide clear evidence of geologic fluctuations in hydrologic conditions. Some specific events, such as the period known as the Younger Dryas, associated with cooler and drier conditions, may be exhibited in the Tulane sediments. Specifically, there is a shift in the C 28 immediately preceding the Holocene to more enriched values, corresponding to the timing of the Younger Dryas approximately 12,000 years ago, correlating with the assertion that this time period was characterized by conditions that would favor lower relative humidity. Congruently, a shift to lower relative humidity at approximately 8ka corresponds with previous research (see detailed references below) that discusses an 8.2ka cooling event at this time. Conversely, a rapid jump to increased levels of relative humidity is observed during the middle of the Holocene prior to this potential cooling event, exhibiting values even higher than todays average humidity. This time period corresponds to the Mid-Holocene Thermal Maximum. The Tulane data provide clear evidence of abrupt climatic events recorded in a continental environment in the low latitudes. Multiple studies have shown that climatic conditions across the globe have undergone variability during these time periods (Poore et al., 2003; McDermott et al., 2003; Baldini et al., 2003; Haug et al., 2001; Grimm et al., 1993). The titanium records from Haug (2001) that were associated above with the LIA and MWP also extend back through the
123 entire Holocene, recording increases in precipitation and runoff over periods of thousands of years during the Holocene Thermal Maximum near 8kyr before present, and significant decreases in runoff during the Younger Dryas, near 12kyr before present. Speleothem records from Baldini et al. (2003) have utilized strontium and phosphorus concentrations to pinpoint specific cold events, such as a major cooling episode at 8200 years before present, also associated with drier conditions. D 18 O from marine sediments in the Gulf of Mexico has been analyzed by Poore et al. (2003) to show that century-scale variability throughout the Holocene has affected sea surface temperatures and can be related to the migration of the Intertropical Convergence Zone. This variability is evident in the hydrogen isotopic data from Lake Tulane as well. Further high-resolution analysis of these sediments will potentially provide more connections between hydrologic change in subtropical North America and thermal or other responses to climatic fluctuations on centennial to millennial time scales through the Holocene and deglaciation. Climatic Shifts from the Holocene into Glacial Conditions Large transitions in the isotopic signature of both C 16 and C 28 are observed from the Holocene back into the last glacial period. C 16 data record an average isotopic depletion of algal material during the last glacial period of 15 relative to values in the interglacial. This depletion correlates well with the hydrologic and climatic changes that are associated with glacial intervals, specifically lower temperatures and potentially an alteration in the origin of storm fronts and source waters entering the Lake Tulane region, potentially exhibiting amplified effects that have been observed over the annual cycle (Katz et al., 2003). It could be assumed that the isotopic shift seen in this record is
124 related solely to a drop in atmospheric temperatures during the last glacial episode. Calibration studies performed by Huang et al. have provided a direct relationship between the D of fatty acid C 16 and temperature (Huang et al., 2002). These relationships indicate that an isotopic shift of 15 in C 16 would correspond to a temperature change of approximately 3.2 C. This degree of temperature fluctuation between glacial and interglacial periods in subtropical North America may be possible, however, it is also conceivable that the isotopic shift observed in aquatic material is related to both a change in temperature combined with fluctuations in the isotopic composition of source waters (due to more northern, and thus more depleted environmental water entering the system). However, quantifying this distinction using only hydrogen isotopic measurement is not possible with current methods, as both temperature and source waters play a role in the isotopic signature preserved in aquatic organic matter. Further research and correlations between other paleoclimatic proxies will help to resolve the relative contributions of each parameter. The average isotopic shift of 55 in C 28 from the glacial to the interglacial is representative of the isotopic composition of terrestrial material in the Lake Tulane system. This behavior, in conjunction with the relationships established in the modern calibration, suggests that glacial climate was characterized by drier atmospheric conditions, and thus lower relative humidity. Evapotranspiration rates were congruently higher, leading to an increased enrichment in terrestrial organic matter relative to the associated aquatic organic matter in the Lake Tulane system. Examining only the C 28 as a climatic indicator leads to problems, because as it was stated in the modern calibration,
125 the D of terrestrial organic matter is a reflection of both the source waters it incorporated in biosynthesis and the isotopic enrichment caused by relative humidity conditions ( # p + Evapotranspiration ) Thus the record of C 28 only provides an initial perspective into the hydrologic changes associated with the last glacial transition. However, by examining the offset between C 28 and C 16 ( !" D (T-A) ) as described in the modern calibration, the effects of source water variability are removed, and inferences into relative humidity variability can be made. These data record a shift in the offset between the D of terrestrial and algal material of approximately 45 over the last glacial transition, with a significantly larger !" D (T-A) observed during glacial times. These initial observations indicate that a corresponding change in average relative humidity of up to 6% was experienced in the Lake Tulane region from todays average measurements to glacial times. A 6% shift is a potential indicator of significant climatic variability. Specifically, the data suggests that average humidity during the last glacial period was more similar to the driest conditions observed in subtropical North America during the modern annual cycle (see Discussion section). Correlations with Other Research A wide variety of climatic proxies utilized in other research complement the approach implemented in Lake Tulane by observing changes in temperature, productivity, and atmospheric dynamics over the glacial/interglacial time scale (Jasper 1989; Jasper and Hayes 1993; Crowley and Berner 2001; Huang et al. 2001). Jasper (1989) utilized the ratio (UK37) of long chain (C 37 ) unsaturated alkenones in sediments from the Pigmy Basin in the Gulf of Mexico to examine fluctuations in sea surface temperatures over the
126 last ~100,000 years, predicting temperatures that were 1 C cooler than the Holocene high SML temperature of 25.6 0.5 C. In the same ocean basin, Jasper and Hayes (1993) compared the relative contributions of terrestrial and marine organic carbon to the sedimentary record over the same time period using molecular carbon isotopic measurements, showing significant increases in the percentage of terrigenous inputs during glacial conditions as a function of decreases in marine assemblage productivity. Huang et al. (2001) has compared the relative abundance of C 3 and C 4 plants in the lower latitudes using carbon isotopic measurements of lacustrine sediments in order to make inferences into variability in dryness levels and temperature conditions during the last glacial maximum. These studies show that significant decreases in temperature, various types of productivity, and humidity were all experienced during the last glacial period. One specific comparison was made between the Lake Tulane hydrogen isotopic data and another unique climatic archive. The calculated relative humidity data from the Lake Tulane sedimentary archive were compared to 18 O data from the GISPII ice core in an effort to observe how global climate change may manifest itself in subtropical North America relative to higher latitudes (Figure 43). It is clear that the fluctuations in calculated/estimated relative humidity correspond extremely well to fluctuations in temperature proxy data from the high latitudes. This further suggests that as the high latitudes respond thermally to global climate change, the low latitudes provide a clear hydrologic response. Additionally, events such as the Mid-Holocene Thermal Maximum, the 8.2 Kyr Cooling Event, and the Younger Dryas can potentially be observed in both records with relatively small time offsets. These results indicate that the Lake Tulane
127 environment provides tools with which to make useful teleconnections between other climatic archives across the globe at a variety of latitudes. In the Tulane environment, both abrupt climatic events and long-term glacial/intergacial transitions are exhibited by hydrologic variability. It is clear that the relationships and offset observed in the modern system are also recorded in the sedimentary environment, and thus can be used as a tool to understand climate in the geologic past. Specifically, C 16 can provide information regarding fluctuations in temperature and the origin of source waters, and the !" D (C 28 C 16 ) can be used to quantify fluctuations in relative humidity over a wide variety of geologic time scales. These applications show that rigorous research of the hydrogen isotopic composition of molecular organic compounds in modern lacustrine systems is the next step in resolving the source of organic materials contributing to the sediments and providing a quantitative method to reconstruct relative humidity in the geologic record.
128 Figure 42. Oxygen isotopic data, reflecting temperature changes, correlated to the offset in terrestrial and algal hydrogen isotopic signatures in Lake Tulane. Note the high degree of similarity in the timing of large-scale changes in both proxies.
129 Chapter 6: Conclusions Bulk Hydrogen Isotopic Analysis Results from the bulk isotopic analysis of the modern calibration and the application of bulk analysis to the 140cm sediment core support the hypotheses asserted in this project that bulk hydrogen analysis of organic material can provide useful quantitative information about the regional hydrologic dynamics of subtropical North America. Specifically, in lacustrine environments, the D of algal material can be used as a direct indicator of the isotopic composition of the source water in which it formed ( D (Algal) = D (H2O) + p ). In the Lake Tulane system, the p was determined to be in the bulk isotopic analysis of modern particulate organic materials. This value was observed to remain relatively stable throughout the annual cycle, correlating with the invariant nature of the hydrogen isotopic signature of the lake waters throughout all seasons and leading to the assertion that the geochemical behavior of the bulk aquatic material is the same as that of the source waters in which it formed. This lack of variability was observed to correspond with relatively invariant hydrogen isotopic values of precipitation reported by the IAEA for the region surrounding Lake Tulane and subtropical North America, and was also supported by the fact that the groundwater-fed nature of the lake system results in a homogenous water supply feeding both the lake as well as the surrounding terrestrial material throughout the year. This stable isotopic behavior provides an ideal environment with which to make comparisons to large fluctuations that occur in the sedimentary record as a function of both abrupt and long term climatic change.
130 Bulk hydrogen analysis of terrestrial material surrounding the Lake Tulane environment confirmed the hypothesis that the degree of relative humidity in the atmosphere surrounding terrestrial plants plays a direct role in their hydrogen isotopic signature relative to the adjacent aquatic material. Specifically, decreased relative humidity leads to an increase in the level of evapotranspiration affecting the plant, resulting in an increased hydrogen isotopic enrichment in terrestrial biomass relative to algal biomass. Thus the isotopic composition of terrestrial biomass is defined as D (Terr) = D (H2O) + p + Evapotranspiration. The offset between the hydrogen isotopic signature of terrestrial biomass and algal biomass was used to define a new proxy for quantifying relative humidity. This proxy (isotopic offset) has been defined as #! D (T-A) Average relative humidity in the Lake Tulane environment was observed to vary by up to ~6% over the annual cycle, and the relative enrichment of terrestrial biomass was observed to directly correlate with this hydrologic variability. This correlation was used to quantify the relationship between relative humidity and #! D (T-A) leading to the resulting equation: RH = -0.1917( #! D (A-T) ) + 82.088. This mathematical relationship suggests that there is a maximum level of relative humidity (~82%) that will affect the isotopic enrichment of terrestrial material relative to algal biomass, beyond which levels of evapotranspiration are to a low enough degree that isotopic differences between the two sources of organic matter are negligible. The fluctuations in relative humidity and the associated #! D (T-A) were determined to be associated with the seasonal migration of three climatic phenomena associated with subtropical North America. During summer months, a northward migration of the Inter
131 Tropical Convergence Zone and an eastward migration of the Bermuda High combine to result in increased levels of precipitation and humidity over the Lake Tulane region. Conversely, during winter months the high pressure center of the Bermuda High expands westward over Florida, and the ITCZ migrates to the south, resulting in lower temperatures and somewhat lower relative humidity. Additionally, the penetration of northern storm fronts into subtropical North America during winter months allows for relative humidity conditions that are still somewhat high from December through March. The interaction of these three phenomena result in the lowest relative humidity conditions during late spring (April-May). The #! D (T-A) follows the same trend, with maximums in late spring during times where evaporative processes are high, and minimums during summer and winter when humidity conditions lead to lower rates of evapotranspiration. The results of this modern calibration clearly confirm the hypothesis that the #! D (T-A) can be used as a proxy for calculating relative humidity, and that the continental response to seasonal climate change in the low latitudes manifests itself in hydrologic parameters. Application of the proxy to bulk hydrogen isotopic analysis of a sediment core confirmed that the approach can be used to identify climatic events in a lacustrine archive. The data suggests that subtropical North America contains climatically sensitive records that can be used to examine abrupt events such as the Little Ice Age and Medieval Warm Period. This study asserts that the lower relative humidity during the Little Ice Age is a function of the long-term migration of the ITCZ to the south and potentially the Bermuda High to the east during this time period. The opposite is observed during the Medieval Warm Period. The results confirm the hypothesis that the low latitudes, and specifically the
132 continental environment of subtropical North America, exhibit a hydrologic response to abrupt climate change. The results correlate well with other climatic archives that examine hydrologic responses in the low latitudes, as well as thermal responses observed in the high latitudes, providing new teleconnections between locations across the globe. Molecular Hydrogen Isotopic Analysis Molecular hydrogen isotopic analysis confirmed that the relationships observed in both the modern system and sedimentary archives on a bulk level can be applied to a compound specific approach in order to more comprehensively resolve parameters in the sediments. The #! D (T-A) was examined using specific fatty acids that have been shown to be biomarkers for algal (C 16 ) and terrestrial (C 28 ) material. The relative enrichment in terrestrial organic matter due to evapotranspiration was also recorded on the molecular level, indicating that this proxy can indeed be used to resolve multiple sources of organic matter in a sedimentary archive. Molecular analysis of recent sediments confirmed that the relative contribution of algal material to lake Tulane decreased significantly prior to the 20 th century. Examining the #! D (C28-C-16) in a sedimentary archive spanning the last 70,000 years confirmed that Lake Tulane contains an extensive climatically sensitive collection of organic matter. The proxy indicates a significant shift to decreased relative humidity during the last glacial period. Additionally, the hydrogen isotopic signature of algal material indicates a significant decrease in temperature and/or the origin of source waters entering the system during glacial times. Wide variability in relative humidity was also observed through deglaciation and into the Holocene. Specific events such as the Younger Dryas, the 8.2ka Cooling Event, and the Mid-Holocene Thermal Maximum
133 exhibit themselves in the sediments of Lake Tulane as shifts in humidity conditions. It is apparent that this continental, low-latitude environment manifests a hydrologic response to global climate change on a wide variety of time scales. The development of this new proxy will allow for the quantification of hydrologic conditions across multiple latitudes during the geologic past.
134 List of References Allison, G.B., Gat, J.R., and Leaney, F.W.J. (1985 ) The relationship between deuterium and oxygen-18 delta values in leaf water. Chemical Geology (Isotope Geoscience section), 58, 145-156. Baldini, J.U.L., McDermott, and F., Fairchild, I.J. (2000) Structure of the 8200-Year Cold Event Revealed by a Speleothem Trace Element Record. Science, 296 (5576) 2203. Bond, G., Heinrich, H., Broeker, W., Labeyrie, L., McManus, J., Andrews, J., Huoni, S., Jantschik, R., Clasen, S., Simet, C., Tedesco, K., Klas, M., Bonani, G., and Ivy, S. (1992 ) Evidence for massive discharges of icebergs into the North Atlantic ocean during the last glacial period. Nature, 360 245-249. Bradley, R.S., Briffa, K.R., Crowley, T.J., Hughes, M.K., Jones, P.D., and Mann, M.E. (2001) The scope of Medieval warming. Science, 292, 2011-2012. Broeker, W. (2001) Was the medieval warm period global? Science, 291 1497-1499. Burk, R.L. and Stuiver, M. (1981) Oxygen Isotope Ratios in Trees Reflect Mean Annual Temperature and Humidity. Science 211 1417-1419. Chikaraishi, Y. and Naroaka, H. (2001) Organic hydrogen-carbon isotopic signatures of terrestrial higher plants during biosynthesis for distinctive photosynthetic pathways. Geochemical Journal, 35 451-458 Cooper, L.W., and DeNiro, M.J. (1989) Covariance of oxygen and hydrogen isotopic compositions in plant water: Species effects. Ecology, 70 (6), 1619-1628. Correge, T., Quinn, T., Delcroix, T., Le Cornee, F., Recy, J., and Cabioch, G. (2001) Little Ice Age sea surface temperature variability in the southwest tropical Pacific. Geophysical Research Letters 28 3477-3480. Crowley, T.J. and Berner, R.A. (2001) CO 2 and Climate Change. Science 292 870-872. Davis, R.E., Hayden, B.P., Gay, D.A., Phillips, W.L., and jones, G.V. (1997) The North Atlantic Subtropical Anticyclone. Journal of Climate, 10, 728-744. DeMenocal, P., Ortiz, J., Guilderson, and T., Sarnthein, M. (2000) Coherent Highand Low-Latitude Climate Variability During the Holocene Warm Period. Science, 288 2198-2202. DeNiro, M.J. (1981) The effects of different methods of preparing cellulose nitrate on the determination of the D/H ratios of non-exchangeable hydrogen of cellulose. Earth and Planetary Science Letters, 54 177-185.
135 Epstein, S., Thompson, P., and Yapp, C.J. (1977) Oxygen and hydrogen isotopic ratios in plant cellulose. Science, 198 (4323) 1209-1215. Epstein, S., Yapp, C.J., and Hall. J.H. (1976) The determination of the D/H ratio of nonexchangeable hydrogen in cellulose extracted from aquatic and land plants. Earth and Planetary Science, 30, 241-251. Estep, M.F., and Hoering, T.C. (1981) Stable Hydrogen Isotope Fractionations during Autotrophic and Mixotrophic Growth of Microalgae. Plant Physiology 67 474-477. Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C., and Rohl, U. (2001) Southward Migration of the Intertropical Convergence Zone Through the Holocene. Science 293 1304-1308. Huang, Y., Shuman, B., Wang, Y., and Webb III, T. (2002) Hydrogen isotope ratios of palmitic acid in lacustrine sediments record late Quaternary climate variations. Geology 30 no. 12, 1103-1106. Hughen, K.A., Overpeck, J.T., Peterson, L.C., and Trumbore, S. (1996) Rapid climate changes in the tropical Atlantic region during the last deglaciation. Nature, 380 (7) 5154. Jasper, J.P. (1989) Alkenone molecular stratigraphy in an oceanic environment affected by glacial freshwater events. Paleoceanography, 4 (6) 603-614. Jasper, J.P. and Hayes, M. (1993) Refined estimation of marine and terrigenous contributions to sedimentary organic carbon. Global Biogeochemical Cycles 7, (2) 451461. Katz, R.W., Parlange, M.B., and Tebaldi, C. (2003) Stochastic modeling of the effects of large-scale circulation on daily weather in the southeastern U.S. Climatic Change, 60, 189-216. Keeley, J.E., Sternberg, L.O., and DeNiro, M.J. (1986) The use of stable isotopes in the study of photosynthesis in freshwater plants. Aquat. Bot., 26, 213-223. Keigwin, L.D. (1996) The Little Ice Age and Medieval Warm Period in the Sargasso Sea. Science, 274 (5292), 1503. Lin, H., Peterson, L.C., Overpeck, J.T., Trumbore, S.E., and Murray, D.W. (1997 ) Late Quaternary climate change from 18 O records of multiple species of planktonic foraminifera: High-resolution records from the Cariaco Basin, Venezuela. Paleoceanography, 12 (3) 415-427.
136 Lutgens and Tarbuck (2001) The Atmosphere. 8 th edition, Figure 7.9 McDermott, F., Mattey, D.P., and Hawkesworth, C. (2001) Centennial-Scale Holocene Climate Variability Revealed by a High-Resolution Speleothem 18 O Record from SW Ireland. Science 294 (5545) 1328. McKenzie, J.A. and Hollander, D.J. (1993) Oxygen isotope record in recent carbonate sediments from Lake Greifen, Switzerland (1750-1986): Application of continental isotopic indicator for evaluation of changes in climate and atmospheric circulation patterns. Climate Change in Continental Isotopic Records. Geophysical Monograph 78 Moritz, R.E., Bitz, C.M., and Steig, E.J. (2002) Dynamics of Recent Climate Change in the Arctic. Science, 297 (5586) 1497. Northfelt, D.W., DeNiro, M.J., and Epstein, S. (1981) Hydrogen and carbon isotopic ratios of the cellulose nitrate and saponifiable lipid fractions prepared from annual redwood growth rings of a California redwood. Geochimica et Cosmochimica Acta, 45 1895-1898. Peixoto, J.P. and Oort, A.H. (1996) The Climatology of Relative Humidity in the Atmosphere. Journal of Climate, 9, 3443-3463. Peterson, L.C., Haug, G.H., Hughen, K.A., and Rohl, U. (2000) Rapid changes in the hydrologic cycle of the tropical Atlantic during the last glacial. Science, 290 (8), 19471950. Poore, R.Z., Dowsett, H.J., and Verardo, S. (2003) Millennialto century-scale variability in Gulf of Mexico Holocene climate records. Paleoceanography, 18 (2) 26.126.11. Rich, J.J., Hollander, D., and Birchfield, G.E. (1998 ) The role of regional bioproductivity in atmospheric CO 2 changes. Global Biogeochemical Cycles, 1-24. Sauer, P.E., Eglinton, T.I., Hayes, J.M., Schimmelmann, A., and Sessions, A.L. (2001) Compound-specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Cosmochimica Acta 70 213222. Schimmelmann, A. (1991) Determination of the concentration and stable isotopic composition of nonexchangeable hydrogen in organic matter. Anal. Chem. 63, 24562459. Sessions, A.L., Burgoyne, T.W., Schimmelmann, A., and Hayes, J.H. (1999) Fractionation of hydrogen isotopes in lipid biosynthesis. Organic Geochemistry, 30 1193-2000.
137 Stauffer, B., Blunier, T, Dallenbach, A., Indermuhle, A., Schwander, J., Stocker, T.F., Tschumi, J., Chappellaz, J., Raynaud, D., Hammer, C.U., and Clausen, H.B. (1997) Letters to Nature, Sternberg, l.S.L ., DeNiro, M.J., and Johnson, H.B. (1986) Oxygen and Hydrogen Isotope Ratios of Water from Photosynthetic Tissues of CAM and C 3 Plants. Plant Physiology, 82, 428-431. Terwilliger, V.J., and DeNiro, M.J. (1995) Hydrogen isotope fractionation in woodproducing avocado seedlings: Biological constraints to paleoclimatic interpretations of D values in tree ring cellulose nitrate. Geochimica et Cosmochimica Acta, 59 (24) 5199-5207. Thompson, L.G. (2001) Climatic Changes for the last 2000 Years Inferred from Ice-Core Evidence in Tropical Ice Cores. NATO ASI Series 141 281-295. Werne, J.P. and Hollander, D.J. (2000) Climate-induced variations in productivity and planktonic ecosystem structure from the Younger Dryas to Holocene in the Cariaco Basin, Venezuela. Paleoceanography 15 (1) 19-29. Yakir, D., DeNiro, M.J., and Rundel, P.W. (1989) Isotopic inhomogeneity of leaf water: Evidence and implications for the use of isotopic signals transduced by plants. Geochimica et Cosmochimica Acta, 53, 2769-2773. Online references: Avon Park, Florida Chamber of Commerce, www.apfla.com Floridas LakeWatch program, http://lakewatch.ifas.ufl.edu International Atomic Energy Association, www.iaea.org National Oceanic and Atmospheric Administration, www.noaa.gov National Weather Service, www.nws.noaa.gov Terraserver Aerial Photography, www.terraserver-usa.com