RECONSTRUCTING FLORIDA LACUSTRINE ENVIRONMENTS FROM THE LATE PLEISTOCENE TO THE PRESENT By THOMAS ELLIOTT ARNOLD A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE R EQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2017
2017 Thomas Elliott Arnold
To my committee
4 ACKNOWLEDGMENTS I would like to thank the entire Geological Sciences Departm ent: professors, staff, students, and lab technicians. I am especially indebted to Mark Brenner, Ja son chemistry. The ir assistance was invaluable, and will never be forgotten I would also like to thank Tom Bianchi for unbridled access to his organic geochemistry lab and Kate Freeman and Aaron Diefendorf for training in lab methods Funding from the University of Florida was provided by the Land Use and Environ mental Change I nstitute (LUECI) and the Water Institute. Additional s upport for this research was provided by the Inter university Training in Continental scale Ecology (ITCE Project) under Award Numbers EF 1137336 and EF 1240142 from the National Science Foundation
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 C H A P T E R 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 SOURCES OF ORGANIC MATTER TO THREE LAKES OF DIFFERENT TROPHIC STATUS ................................ ................................ ................................ 13 Introduction: Sheelar, Wauberg, and Apopka ................................ ......................... 13 Study Sites ................................ ................................ ................................ .............. 15 Organic Matter Sources of n Alkyl Lipids ................................ ................................ 16 Core Collection and Sampling ................................ ................................ ................ 18 Chronology ................................ ................................ ................................ ............. 19 Lipid Extraction, Purification, and Quantif ication ................................ ..................... 19 Bulk Geochemical Data: Lakes Sheelar, Wauberg, Apopka ................................ ... 21 Alkane Concentrations: Lakes Sheelar, Wauberg, Apopka ................................ .... 21 Fatty Acid Concentrations: Lakes Sheelar, Wauberg, Apopka ............................... 22 Interpreting Bulk Geochemical Data: Lakes Sheelar, Wauberg, Apopka ................ 23 Interpreting Alkyl Lipid Data: Lakes Sheelar, Wauberg, Apopka ............................. 26 Conclusions: Sheelar, Wauberg, and Apopka ................................ ........................ 31 3 THE BIOGEOCHEMICAL EVOLUTION OF A SUBTROPICAL LAKE .................... 41 Introduction: Lake Harris ................................ ................................ ......................... 41 Site Description: Lake Harr is ................................ ................................ .................. 44 Sediment Sampling: Lake Harris ................................ ................................ ............ 45 Lipid Extraction and Quantification: Lake Harris ................................ ..................... 46 Compound specific Isotope Measurements on n Alkanes: Lake Harris .................. 47 Bulk Geochemistry and Isotopic Compositions: Lake Harris ................................ ... 47 Concentrations and Isotopic Compositions of Hydrocarbons: L ake Harris ............. 48 Discussion Overview ................................ ................................ ............................... 52 Concentrations of n Alkanes and Shifts in Geochemical Biomarkers as Indicators of Organic Matter Source in Lake Harris ................................ ............. 52 Interpreting Carbon Isotope Variability in TOC and Hydrocarbon Biomarkers in Lake Harris ................................ ................................ ................................ .......... 56 Conclusions: Lake Harris ................................ ................................ ........................ 61
6 4 SUBTROPICAL CLIMATE RESPONSE TO HEINRICH EVENTS IN THE NORTH ATLANTIC ................................ ................................ ................................ 68 Introduction: Lake Tulane and Heinrich Events ................................ ...................... 68 Carbon Isotopes and Precipitation ................................ ................................ .......... 71 Hydrogen Isotopes a nd Precipitation ................................ ................................ ...... 72 Study Site and Sample Collection: Lake Tulane ................................ ..................... 73 Lipid Extraction, Purification, and Quantification: Lake Tulane ............................... 74 Compound Specific Isotope Measurements: Lake Tulane ................................ ...... 75 n Alkane Concentrations and Chain Length Distributions ................................ ....... 77 Carbon Isotope Results: Lake Tulane ................................ ................................ ..... 79 Hydrogen Isotope Results: Lake Tulane ................................ ................................ 79 Reconstructing Paleohydrology from Leaf Wax Carbon and Hydrology Isotopes ... 80 Implications of Rapid Climate Change Events in the Subtropics ............................ 88 Conclusions: Lake Tulane ................................ ................................ ....................... 91 5 CONCLUSIONS ................................ ................................ ................................ ... 100 LIST OF REFERENCES ................................ ................................ ............................. 102 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 114
7 LIST OF TABLES Table page 2 1 Lake Sheelar alkane and fatty acid data ................................ ............................. 38 2 2 Lake Wauberg alkane and fatty acid data. See Table 2 1 for full description. .... 39 2 3 La ke Apopka alkane and fatty acid data. See Table 2 1 for full description. ...... 40 3 1 Concentrations, maxima, minima, and ranges for the five n alkane homologues used for organic matter source inte rpretation. Zones are 13 C TOC values. ................................ ................................ ............. 66 3 2 Absolute values for carbon preference index (CPI), average chain length (ACL), and the ratio of submerged to emergent vegetation (Paq ). .................... 67 4 1 Values for the carbon preference index (CPI), average chain length (ACL), and submerged to emergent/terrestrial veget ation in the Lake Tulane core. ..... 97 4 2 leaf values for select n alkane chain lengths extracted from the Lake Tulane core. ................................ ................................ ................................ ................... 98 4 3 values for select n alkane chain lengths extracted from the Lake Tulane core. Co re phases are as follows: Tulane Pinus (TP), and Tulane Quercus (TQ). ................................ ................................ ................................ ................... 99
8 LIST OF FIGURES Figure page 2 1 Down core variability in bulk geochemical variables for lakes Sheelar, Wauberg, and Apopka. ................................ ................................ ....................... 32 2 2 Down core variability in select n alkane chain lengths f or Sheelar, Apopka, and Wauberg ................................ ................................ ................................ ..... 33 2 3 Down core variability in select fatty acid chain lengths from Lake Sheelar ........ 34 2 4 Down core variability in select fatty acid chain lengths from Lake Wauberg. .... 35 2 5 Down core variability in select fatty acid chain lengths from L ake Apopka. ........ 36 2 6 Relative abundances (wt/wt) of nitrogen and organic carbon from the Lake Sheelar core. ................................ ................................ ................................ ...... 37 3 1 Down core carbon isotope variability of TOC in the Lake Harris core. .............. 62 3 2 Bulk geochemical variability in the Lake Harris core. ................................ .......... 63 3 3 Select n alkane chain l ength abundances in the Lake Harris core. Dashed lines delineate core zones 1 3. ................................ ................................ ........... 64 3 4 Carbon isotope variability in sel ect n alkane chain lengths in the Lake Harris core. ................................ ................................ ................................ ................... 65 4 1 Concentrations of select n alkane chain lengths. ................................ .............. 93 4 2 Select n alkane chain lengths and their leaf values (dashed lines) plotted with the percent abundance of Pinus pollen (solid line) ................................ ..... 94 4 3 Select n the perce nt abundance of Pinus pollen (solid line). ................................ ........... 95 4 4 (terr aq) values (dashed line) plotted with the percent abundance of Pinus pollen (solid line). Tulane Pinus zones are represented with the orange bars. .. 96
9 Abstract of D issertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy RECON S TRUCTING FLORIDA LACUSTRINE ENVIRONMENTS FROM THE LATE PLEISTOCENE TO THE PRE SENT By Thomas Elliott Arnold May 2017 Chair: Mark Brenner Major: Geolog y I analyzed bulk geochemical and biomarker data from subtropical lakes in Florida to determine how environmental change and anthropogenic effects are recorded in lacustrine sediment s. On short (i.e. centennial) time scales, anthropogenic impacts are apparent in more recent sediments from three Florida lakes and manifest themselves as increased algal lipid biomarker concentrations, more enriched 13 C values, and low er TOC:TN ratios. Across millennial time scales in Lake Harris biomarker concentrations and their carbon isotopic values respond to regional hydrological changes in the early 13 C) and hydrogen pe ratios of lipid biomarkers extracted from a Lake Tulane sediment core serve as paleohydrology proxies across stadial/interstadial time scales during the late 13 n alkane chain lengths, demonstrate that subt ropical Florida responded out of phase with high latitude environmental changes during at least three of the late Pleistocene stadial interstadial transitions. The region remained warm and wet, when the North Atlantic was cold and dry, because of oceanic/a tmospheric patterns that limited northward export of warm, saline water. All three studies presented within, broadly agree with published literature on similar topics
10 spanning identical timeframes. Thus, we are confident in the application of lacustrine b iomarker data to measure regional change across highly variable time scales.
11 CHAPTER 1 INTRODUCTION In the following three research papers, I present the results of bio marker studies from five Florida lakes. The first paper compares up core chang es in bulk geochemical data, as well as alkyl lipid concentrations to estimate the sources of organic matter and the changes in trophic status to three Florida lakes over the past ~150 years. Bulk geochemical variables including: organic carbon percentages organic carbon to nitrogen ratios, and carbon isotope values of total organic carbon ( 13 C TOC ) can be applied as general techniques to measure productivity, vascular versus non vascular plants sources, and autoch thonous carbon assimilation, respectively Although they only represent a small fraction of bulk organic matter, alkyl lipids, including n alkanes and fatty acids, provide greater source specificity. Alkanes are hydrocarbons, a class of organic compounds that consist of the elements carbon and hy drogen (Bianchi and Canuel, 2011). Alkanes with fewer than 20 carbon atoms are typically synthesized by algae. Alkanes with intermediate numbers of carbon atoms, i.e. between 20 25 carbon atoms, are typically, but not exclusively, synthesized by aquatic ma crophytes. L ong chain alkanes, i.e. > 25 carbon atoms, are most abundant in the leaves of woody, terrestrial plants. Fatty acids are lipid biomarkers that consist of variable chain length hydrocar bon units, bound to a carboxyl functional group. Similar to alkanes, fatty acids can be subdivided into bacterial, algal, and terrestrial sources. As a group, fatty acids are less source specific than n alkanes: long chain lengths are synthesized by vascular plants, polyunsaturated varieties by algae, and branched fatty acids by bacteria. We measured concentrations of these biomarkers to assess temporal changes in organic matter source to each lake.
12 The second paper uses identical bulk geochemical variables to assess changes in organic matter sources to Lake Harris, Florida, USA but introduces the 13 C analyses of individual alkane chain length s known as compound specific isotope analyses (CSIA). With the CSIA approach I document complex changes in regional hydrology over millennial scales. I hypothesize that the sources of organic matter (i.e. all ochthonous vs. autoch th onous regional hydrologic and environmental events tha t occurred as the e stabilized in the mid Holocene. Shifts in alkane biomarkers should correspond to fluctuations in the regional water table that were caused by these hydrological changes. The final paper documents shifts in sub tropical climate during Heinrich Events (HE) in the North Atlantic. HE are recorded in the sediment record of the North Atlantic as layers of ice rafted debris (IRD), lithic fragments from rocks of continental origin, which were derived from the calving and melting of large continental ice sheets (Bond et pertur bed global heat transport via reduction in the Atlantic meridional overturning circulation (AMOC) (Lynch Stieglitz et al., 2014). CSIA of carbon and hydrogen in n alkanes extracted from the sedimentary record of Lake Tulane were analyzed to provide an appr oximation of precipitation changes, or more specifically aridity levels in the subtropics during these extreme cold periods in the North Atlantic.
13 CHAPTER 2 SOURCES OF ORGANIC MATTER TO THREE LAKES OF DIFFERENT TROPHIC STATUS Introduction : Sheelar, Wau berg, and Apopka There are nearly 8,000 lakes in the state of Florida with surface area >10 ha M ost are relatively shallow dissolution feature s (mean depth < 5 m maximum depth <20 m ). From a geological perspective, these lakes are relatively recent additi ons to the landscape. Most began to accumulate lacustrine sediment in the early Holocene (~8,000 yrs BP) as water table elevations rose and climate became wetter (Watts, 1980). Florida lakes span the entire trophic state continuum, from ultra oligotrophic in the quartz sands of the Lake Wales Ridge, to hyper eutrophic in some low lying urbanized and agricultural areas. Furthermore, m any naturally eutrophic lakes are located in regions characterized by deposits of the phosphorus rich Hawthorn Group. The tro phic status of lakes in close proximity to Hawthorn sediments is strongly correlated to regional geology, and these systems may have a long history of having been eutrophic (Brenner et al., 1999). The consequences of natural and anthropogenic eutrophicatio n of fresh water bodies have been well documented since the early 20 th century ( e.g., Hasler, 1947). More recently, studies have verified the correlation between excessive nutrient loading (primarily nitrogen and phosphorus) and harmful algal blooms (Corre ll, 1998; Anderson et al., 2002), fish kills (Landsberg et al., 2009), and reductions in biological diversity (Craft et al., 1995 ; Craft and Richardson, 1997). A recent analysis of surface water quality in Florida classified 8% of rivers and streams, 26% o f lakes, and 21% of estuaries in the state as impaired because of excessive nitrogen and/or phosphorus concentrations (Florida Department of Environmental Protection, 2008).
14 Lakes function as natural repositories for organic matter derived from aquatic (au tochthonous) and terrestrial (allochthonous) primary and secondary productivity. Thus, lake sediments serve as archives of past changes in the water column and watershed In Florida, numerous studies have tracked past changes in lacustrine productivity (e. g. Gu et al., 1996; Riedinger Whitmore et al., 2005). Organic and inorganic elemental concentrations in dated sediment cores, including N, P, C and Si, have been used to infer temporal trends in cultural eutrophication (Bianchi and Canuel, 2011). More rece ntly, 13 C and 1 5 N measurements on bulk organic matter (OM) (Torres et al., 2012) and photosynthetic pigments in sediments (Waters et al., 2015) have been employed to reconstruct past lacustrine trophic status and identify the primary producer communities Most previous work on eutrophication in Florida lakes relied on analysis of bulk sediment or bulk OM, which limited the ability to identify the source(s) of the analyzed material. In other studies, for which biomarkers were assayed, one had to be cogniz ant of the fact that such molecules may represent a small fraction of the total organic matter preserved within the lake. Determination of the bio geo chemical processes that occur in the water column and in deposited sediments remains a challenge. Combined use of bulk sediment and biomarker data, however, can help answer general questions about such processes in elemental pools, as well as specific questions about the sources of these elements to the overall pool. The primary goals of this study were to inve stigate the source ( s ) of organic carbon to the sediment in three subtropical Florida lakes evaluate recent changes in lacustrine productivity of the water bodies, and better understand recent human impacts on these aquatic ecosystems. I used a multi
15 bioma rker approach that include d n alkanes and fatty acids, as well as bulk sediment isotopic analyses to examine recent changes in autochthonous and allochthonous carbon inputs to the three lakes. Study Sites Lake Sheelar is a small (0.07 km 2 ), relatively dee p (mean depth = ~8 m, z max = ~21 m) lake in Mike Roess Goldhead Branch State Park, north of Keystone Heights, FL. It is underlain by quartz sands and is a poorly buffered system, with low pH (5.4), low conductivity (18 S/cm), and low total alkalinity (0.8 mg/L as CaCO 3 ). It is also an oligotrophic (total phosphorus = 4 g/L, total nitrogen = 86 g /L, chlorophyll a = 1.7 water (Secchi depth = 7.5 m) system ( Florida Lakewatch, 2003 ), which receives some hydro logic input from shallow groundwater seeps along its northwest shoreline. Lake Wauberg is located south of Gainesville, FL and is primarily surrounded by hardwood pine and oak forests. The lake has a surface area of 1.5 km 2 and is shallow, with a maximum depth of ~3.7 m (Florida Lakewatch, 2003). Wauberg receives water mainly from groundwater and direct rainfall, and has a shallow outflow into Sawgrass Pond. Unlike Sheelar, Wauberg is eutrophic, with mean concentrations of total phosphorus = 112 g/L, total nitrogen = 1670 g/L, chlorophyll a = 82.1 g/L, and a Secchi depth of 0.6 m ( Florida Lakewatch, 2003 ). Near surface geologic deposits around the lake are derived from the phosphate rich Hawthorn Formation, and high phosphorus concentrations ha ve been measured in groundwater entering the lake. Gu et al. (2006) reported that phytoplankton is the dominant primary producer group in the lake, with limited macrophyte abundance, possibly a consequence of light limitation.
16 The phytoplankton populations since at least the 1930s (Carr, 1934). Lake Apopka, central Florida, is the largest (124 km 2 ) and shallowest (mean depth = 1.7 m) of the three study lakes. It is a hyper eutrophic system with a mean tot al phosphorus concentration of 192 g/L, mean total nitrogen concentration of 3906 g/L, mean chlorophyll a concentration of 96 g/L, and a Secchi depth of 0.3 m. Water from the lake flows out to Beauclair Canal, which is hydrologically linked to other lak es in the Harris Chain of Lakes. Apopka shifted from a clear water, macrophyte dominated system, to a phytoplankton dominated water body in the middle of the twentieth century (Waters et al., 2015). This shift has been attributed to high phosphorus loading and inputs of dissolved color from muck farms that operated along the north shore ( Schelske et al. 2005 ). Organic Matter Sources of n Alkyl Lipids The major sources of alkanes to lacustrine sediments are algae, bacteria, submerged and emergent macrophyte s, and woody terrestrial plants. Because these sources generally synthesize unique alkane homologues, alkanes can be used to identify the source of organic matter in lake sediments. Algae and bacteria primarily synthesize short chain n alkanes, including n C 17 19 (Cranwell, 1987). Macrophytes synthesize a broad range of n alkanes, but n C 21 25 predominate among the submerged varieties (Ficken et al., 2000). Odd numbered, long chain hydrocarbons ( n C 27 33 ) are characteristic of h igher terrestrial plants, and are stable over the longest time periods (Muri and Wakeham, 2006). Aliphatic or acyclic, least susceptible to post depositional degradation (Meyers, 2003). Studies on
17 hydrocarbon preservation i n lakes demonstrated that these compounds are stable over at least decadal time scales (Wakeham et al., 2004). Other studies have reported stability of hydrocarbons over much longer, i.e. geologic timescales ( Castaeda and Schouten, 2011 ). This does not im ply that n alkanes are immune from decomposition/alteration, as biodegradation does occur in both aerobic and anaerobic environments, and preferentially targets low molecular weight molecules (Wakeham et al., 2004; Bianchi and Canuel, 2011). Unlike hydroc arbons, fatty acids are more susceptible to diagenesis and alteration. A study from a single lake demonstrated that fatty acids degraded twice as fast as alkanes in oxic sediments (Muri and Wakeham, 2006). Another study of a 0.5 m long lake sediment core s howed that fatty acids were sensitive to post depositional alteration, but hydrocarbons were not (Ho and Meyers, 1994). Nevertheless, these lipids (fatty acids) have been used successfully to track sources of organic matter in many aquatic systems (Meyers, 2003; Zhang et al., 2015). Freshwater algae synthesize C 16 and C 18 saturated fatty acids (chain lengths that are ubiquitous among higher lifeforms), as well as polyunsaturated (C 18:2 and C 18: 3 ) and monounsaturated (C 16:1 ) varieties. Odd numbered (C 15 C 17 ) and branched short chain fatty acids ( iso and anteiso C 15 ) are typically synthesized by bacteria, and even numbered, long chain fatty acids ( C 24 ) are from higher plants ( Palomo and Canuel, 2010) Ratios of long cha in length to short chain length lipids can be used to assess the relative contributions of terrestrial and aquatic organic matter to the sediments (Meyers, 2003). For fatty acids (TAR FA ) this is calculated as: TAR FA = (C 24 + C 2 6 + C 2 8 )/(C 12 + C 14 + C 16 ) ( 1 1)
18 and for hydrocarbons (TAR HC ) as: TAR HC = (C 2 7 + C 29 + C 31 )/(C 15 + C 17 + C 19 ) (1 2) Furthermore, the proportional contribution of submerged aquatic vegetation, relative to emergent and terrestrial vegetation, P aq (Ficken, 2000), is calculated as: P aq = (C 2 3 + C 25 )/(C 23 + C 25 + C 29 + C 31 ) (1 3) These equations are broadly applicable, but must be used with caution as shorter chain length n alkyl lipids are preferentially degraded. Core Collection and Sampling We collected a sediment water interface core from each of the three lakes using a piston corer designed to retrieve undisturbed sediments ( Fisher et al., 1992 ) In Lake Sheelar, 80 cm of sediment was retrieved, in Lake Wauberg, 90 cm, and in Lake Apopka, 100 cm. C ore s were kept vertical and the poorly consolidated deposits were extruded upward and sampled into labeled containers at 2 4 cm intervals. Samples were frozen, freeze dried, and ground with a mortar and pestle in preparation for geochemical analyses. Total carbon and total nitrogen (TN) were measured on a Carlo Erba NA1500 CNS elemental analyzer Inorganic carbon was measured on a UIC Coulometrics 501 2 CO 2 coulometer coupled with an AutoMate automated carbonate preparation device (AutoMateFX.com) Total organic carbon (TOC) was calculate d as the difference between total carbon and inorganic carbon TOC:TN values are expressed on a (wt/wt) basis. 13 C TOC analyses was pretreated with 1N HCl to remove inorganic carbon, and then measured on a Carlo Erba NA1500 CNS
19 elemental analyzer interfaced with a Thermo Scientific Delta V Advantage isotope ratio mass spectrometer. Isotopic compositions were normalized to the VPDB scale, and reported in standard delta notation as follows: 13 C= [(( 13 C / 12 C) sample/ ( 13 C / 12 C) standard) ) 1] x 1000 (1 1) Chronology Cores were dated by 210 Pb, using gamma counting methods described by Schelske et al. (1994). Sediment ages were calculated using the constant rate of supply model ( Appleby and Oldfield 1983). Age errors were propagated using fir st order approximations and calculated according to Binford (1990). 210 Pb dates could not be calculated for Lake Apopka, because the lake surface area was altered in the middle of the 20 th century, thus violating an assumption of the constant rate of suppl y model. A 210 Pb based age model developed by Schelske (1997) for a core from the same site, however, showed that unsupported 210 Pb activity (dpm/g) was negligible below 70 cm core depth, suggesting an age at that depth of ~110 years, i.e. five 210 Pb half lives. Lipid Extraction, Purification, and Quantification Lipids were extracted from 1 g of freeze dried sediment with an Accelerated Solvent Extractor ASE200 (Dionex), using 2:1 (v/v) dichloromethane(DCM):methanol through three extraction cycles at 10.3 MPa (1500 psi) and 100 C. Total lipid extracts (TLE) were concentrated under a gentle stream of nitrogen, and the neutral and acidic lipid fractions were separated via base saponification of the TLE with 0.5 M KOH in methanol at 70 C for 2 h. Neutral lipi ds were extracted from the aqueous solution with hexane, and after acidification with concentrated HCl, the acidic fraction was extracted
20 using 4:1 (v/v) hexane:DCM. The acidic fraction was blown to dryness and deriv a tized with BF 3 in methanol and extract ed again with DCM as fatty acid methyl esters (FAMES). Neutral lipids were further separated, based on polarity, into compound classes by column chromatography, using 5% deactivated silica gel, according to methods modified from Nichols (2011). Hydrocarbo ns were eluted from the silica gel column with 4.5 ml of 9:1 Hexane:DCM, and saturated hydrocarbons were separated from alkenes on 5% Ag impregnated silica gel (w/w) with 4 ml of hexane and ethyl acetate, respectively. Branched and cyclic saturated hydroca rbons were isolated from n alkanes with triple urea adduction. Alkane and fatty acid concentrations were measured and identified on a Thermo Scientific Trace 1310 gas chromatograph with a Supelco Equity 5 column, interfaced to a Thermo Scientific TSQ 8000 triple quadrupole mass spectrometer with electron ionization. For the hydrocarbons, the inlet was operated in splitless mode at 280 C. The column flow rate was set to 2.0 ml/min and the oven was program m ed to an initial temperature of 60 C and held for 1 minute, then ramped to 320 C at 6 C/min and held for 20 minutes. Quantification was based on the calibration curves generated from the peak areas of external standards (C 7 C 40 ) with concentrations ranging from 5 250 g/ml. Androstane was used as an interna l standard, and recovery rates for n alkanes were >80%. For fatty acids, the split/splitless inlet was operated in split mode, at 280 C with a split ratio of 10:1. The column flow rate was 1.5 ml/min and the oven started at an initial temperature of 60 C a nd was held for 1 minute, followed by a ramp to 140 C at 15 C/min, and finally to 320 C and held at that temperature for 20 min.
21 Bulk Geochemical Data : Lakes Sheelar, Wauberg Apopka 13 C TOC are shown in Figure (2 1). Highest average TOC p ercentages were found in Lake Wauberg sediments, followed by the sediments of Lake Apopka. Lowest average TOC values were measured in the Lake Sheelar core. Average TOC:TN was similar for Wauberg and Apopka, with values of 10.67 and 11.58, respectively, bu t was substantially higher in Sheelar (15.40). In Sheelar, the TOC:TN values decrease up core, from a maximum value of 18.69 at 80 cm depth, to a minimum value of 12.03 at 5 cm depth. Ranges for TOC:TN values are less extreme in Wauberg and Apopka, with st 13 C TOC values across the three lakes become more positive with increasing trophic status, from values became enri ched in 13 C up core in the Wauberg and Apopka sediments, 13 C TOC values remained relatively constant throughout the core. Alkane Concentrations : Lakes Sheelar, Wauberg, Apopka ACL, P aq and TAR HC values for each lake are displayed in Table s 2 1, 2 2 and 2 3. Average ACL values for all three lakes were similar (~28) and varied only slightly throughout each core. Average P aq values for the three lakes were also relatively similar, ranging from 0.44 to 0.49. Average TAR HC values in Lake Sheela r (39.41), however, were approximately 4 fold and 10 fold higher than TAR HC values in Wauberg (10.98) and Apopka (3.49). Alkanes were grouped by their most probable biological source: algae ( n C 17 19 ), macrophyte ( n C 23 25 ), and terrestrial ( n C 27 31 ). Re presentative homologues are shown in Figure 2 2. In Sheelar, there was no discernible stratigraphic trend in the concentrations of n alkanes within the three groups. Of note, however, are the low
22 concentrations of algal derived hydrocarbons, which never ex ceed 15 /g OC throughout the core, and high concentrations of the terrestrial alkane, n C 29 throughout the entire core. In contrast to Sheelar, n C 17 concentration in the Wauberg and Apopka cores increases by an order of magnitude at ~40 cm core depth, reaching maximum concentrations in the upper 10 cm. This is the most abundant chain length in the top 20 cm of sediment in both lakes. The increasing up core trend in n C 17 concentration in Wauberg and Apopka is accompanied by a decrease in terrestrial derived alk anes, by 119.3 /g OC and 37.8 /g OC, respectively. Fatty Acid Concentrations : Lakes Sheelar, Wauberg Apopka Fatty Acids were grouped based on their primary source, in a manner similar to the n alkane data. These data, along with TAR FA values for each lake are displ ayed in Table s 2 1 2 2, and 2 3 Terrestrial, long chain fatty acids (LCFA) represent the sum of even carbon fatty acids from C 24 to C 30 Algae derived polyunsaturated fatty acids (PFA) are the sum of C 18:2 and C 18:3 Bacteria produce a wide array of fatt y acids (BFA), and are represented in this study by the following chain lengths: branched iso and anteiso C 15 :0 monounsaturated C 16:1 7 C 18 :1 7 C 18 :1 9 and the odd carbon number saturated fatty acids, C 15:0 and C 17:0 Long chain, saturated fatty acids dominated the record in all three lakes, with C 12:0 C 16:0 C 24:0 C 26:0 C 28:0 being the most abundant homologues. Each of the lakes displays large up core concentration increases in algae and bacteria derived fatty acids, and to a lesser magnitude, decreases in LCFA concentrations (Figure 2 3 2 4 2 5 ). Of note are distinct patterns of bacteria derived fatty acid (BDFA) abundances i n the sediment records. In Sheelar, two groups of BDFA display similar concentration
23 patterns throughout the core: group 1 (saturated, branched iso and anteiso C 15 :0 C 16:1 7 and C 18 :1 9 ), group 2 (C 1 7:0 and C 18 : 1 7 ). The two groups of BDFA from Wauberg that vary correspondingly in concentrations are: group 1 (saturated, branched iso and anteiso C 15 :0 C 17 :0 ), and group 2 (C 16:1 7 C 18 :1 9 and C 18 : 2 7 ). In Apopka, all the BDFA follow similar up core patterns of abundance. Interpreting Bulk Geochemical Data: Lakes Sheelar, Wauberg Apopka The three lakes in this study can be loosely grouped into trophic state categories using the bulk geochemical data. The Sheelar core has the lowest TOC concentrations, highest TOC:TN values, and most depleted 13 C TOC signatures throughout, whereas the Lake Apopka and Wauberg cores have higher OC levels, lower TOC:TN, and less negative carbon isotope values. The values of these three variables in the Sheelar core are indicative of lower trophic status, and in Ap opka and Wauberg of higher productivity (Meyers, 1997; Brenner et al., 1999). There are, however, limitations to using only bulk sediment data to infer trophic status, which are apparent when comparing up core changes in sediment TOC:TN and 13 C TOC values. In Lake Sheelar, TOC:TN exhibits the largest up core decrease, suggestive of increasing relative algal contributions to the sediment. Across the same interval, 13 C TOC signatures remain essentially unchanged, which implies no change in primary productivit y. Furthermore, the carbon isotope values ( 30.39 to 28.44 ) are characteristic of a constant, and primarily C 3 terrestrial carbon source to the sediments (Peterson and Fry, 1987; Magill et al., 2013). TOC:TN may display an up core decrease in the Sheel ar core because of N processing in the sediments. T he plot of %N and %C (Figure 2 6 ) shows that the lower TOC:TN values near the surface are a consequence of relatively higher %N values up core, not greater
24 TOC concentrations. Nitrogen removal from sedimen ts and re deposition into the near surface sediments of this low nutrient system could be responsible for the stratigraphic distribution of TN in the sediments. In Apopka, TOC:TN values decrease slightly up 13 C TOC values increase substantially, from 13 C TOC to 13 C signatures in Apopka and Wauberg both indicate primarily algal so urces of organic carbon in the sediment (Filley et al., 2001), negligible changes in TOC:TN fail to capture the increasing algal contributions that are indicated by the up core decrease in carbon isotope values. A possible explanation for this discrepancy may lie in the historic concentrations of algal and aquatic macrophyte communities. If substantial populations of both phytoplankton and submerged macrophytes with similar TOC:TN values (see review of TOC:TN values in Bianchi and Canuel, 2011) were already well established in both lakes, a further increase in trophic state would cause an increase in 13 C TOC values, but not significantly affect TOC:TN levels. Based on 13 C TOC alone, we infer relatively stable, oligotrophic conditions in Lake Sheelar through out the time period represented by the core. The 210 Pb dates extend to a depth of only 24 cm (1885 AD), indicating relatively low mean sedimentation rate, isotope st udies on a sediment core from Lake Annie, another oligotrophic Florida lake, similarly conclude that low 13 C TOC values are the result of primarily allochthonous c ontributions to the sediments (Torres et al., 2012). The up core trend in 13 C TOC values in the Wauberg and Apopka cores suggests the trophic status of both lakes increased
25 from the bottom of the co re to the most recent sediments. The up core increase in 13 C TOC values is most likely associated with greater phytoplankton abundance that results in reduced discrimination against 13 C and/or active uptake of HCO 3 (Brenner et al., 1999). This phenomenon was documented in Lake Wauberg by Gu et al. (2006), who measured 13 C enrichment in particulate organic matter during periods of high productivity and carbon limitation. In the Apopka core, both average 13 C TOC values and the rate of 13 C enrichment were gre ater than in the Wauberg core. Some of this enrichment is almost certainly a consequence of high 13 C values of dissolved inorganic carbon (DIC). Carbon 13 enrichment of the DIC pool in Lake Apopka (average DIC = 9.0 ) was documented by Gu et al. (2004), a nd attributed to methanogenesis in the sediments. Methanogenesis produces isotopically depleted CH 4 and isotopically enriched CO 2 (Whiticar, 1999). Utilization of this enriched carbon source by primary producers elevates their 13 C signatures. Over time, as the lake becomes more productive, and incorporation of 13 C enriched CO 2 increases, the carbon isotope value of the sedimented organic matter consequently becomes more enriched. The bulk geochemical data enable us to reach the following conclusions abou t each of the three lakes: 1) Lake Sheelar is, and historically has been an oligotrophic water body, with primarily allochthonous, but some autochthonous inputs to its sediment carbon pool. TOC:TN values decrease up core, possibly indicating an increase in algal input relative to terrestrial input, but the 13 C TOC values remain well within the terrestrial range throughout the entire core. The Wauberg and Apopka cores both show evidence for recent increases in primary productivity and contributions of algal biomass, as evidenced by higher 13 C TOC and lower TOC:TN values. The up core increase in
26 13 C TOC in both cores is the result of carbon limitation and/or uptake of HCO 3 In Apopka, 13 C TOC is very high because of methanogenesis, which led to 13 C enriched D IC. Interpreting Alkyl Lipid Data: Lakes Sheelar, Wauberg Apopka Concentrations of n alkanes in the Sheelar core suggest little change in organic matter sources through time. From the bottom of the core to the top, n C 17 concentrations remain low, while n C 23 25 and n C 27 31 concentrations, although variable, are the most abundant homologues. Furthermore, we see no evidence for cultural eutrophication in the hydrocarbon record (i.e. increasing n C 17 concentrations, higher 13 C TOC values) In addition, hydrocarbon proxy ratios TAR HC and P aq both indicate primarily terrestrial and submerged/emergent macrophyte contributions to the sediment organic carbon pool. These data, however, conflict with bulk TOC:TN data, which decreas e upcore, implying relatively greater algal contributions to the sediment, and hence increased lacustrine productivity. TOC:TN values, however, can be misleading and are susceptible to in situ alteration (Meyers, 2003), and the large range in ratios result s in overlapping values for terrestrial and aquatic sources (Bianchi and Canuel, 2011). Furthermore, the TOC:TN data from sediments are at odds with modern lake water phosphorus ( 4 g/L), nitrogen ( 86 and chlorophyll a (1.7 g/L) concentrations, all of which imply very low limnetic production. From the bottom of the Lake Wauberg core (88 cm) to ~40 cm core depth, vascular plant n alkanes ( n C 23 31 ) dominate the hydrocarbon reco rd. Above 40 cm (AD 1986), algal derived n alkanes increase, and higher plant biomarkers decrease. This shift is concurrent with an up core rise in sedimentary organic carbon, a decrease in TOC:TN ratios, and higher 13 C TOC values. Interestingly, substanti al development of the
27 Lake Wauberg watershed began after 1980, and the rise in n C 17 concentration, a 9 fold increase since 1980, is concurrent with this development. Despite this increase in the algae biomarker, vascular plant biomarkers remain the most a bundant chain lengths throughout the Wauberg core, and although TAR HC values indicate a decline in macrophyte abundance up core, values never drop below the threshold value of <1, indicative of primarily algal input. Furthermore, all P aq values fall within the range reported for submerged macrophytes, meaning that for the entirety of the Wauberg record, chain lengths > n C 21 are primarily derived from aquatic sources (Ficken et al., 2000). The interpretation of the Wauberg n alkane record contrasts with in ferences based on cyanobacterial pigment data and diatom analysis of another core from the lake, which appeared to show persistent eutrophic conditions and stable primary producer populations since at least ca. AD 1894 ( Riedinger Whitmore et al., 2005 ). I propose that some of the variability in the pigment record of Riedinger Whitmore et al. (2005 ) is the result of post depositional alteration of pigment concentrations. Pigment biomarkers are unsaturated compounds, and are thus more likely to degrade than s aturated compounds such as n alkanes. The authors reported that native chlorophyll abundance declined by 50% from the base of the core (10%) to the surface (5%). Percent native chlorophyll is a metric used to quantify preservation of photosynthetic pigment s in sediments (Swain, 1985). The significant up core reduction in preservation of pigments would probably have resulted in an underestimation of cyanobacterial abundance in the most recent sediments.
28 As in the Wauberg core, the most striking trend in the Lake Apopka lipid concentration data is the rapid increase in algal derived biomarkers beginning at ~40 cm core depth (Figure 2 2.). Multiple studies (Shumate et al., 2002; Waters et al., 2005; Waters et al., 2015) have documented that Lake Apopka transit ioned from a macrophyte dominated system to an algal dominated system in the 1940s. The increase in algal biomarker concentration at 40 cm likely represents this shift. Another study on Lake Apopka presented n alkane concentration data from a core raised f rom the same location as our core (Silliman and Schelske, 2003). In that study, TAR HC values indicated a shift from macrophyte dominated organic matter at 65 cm core depth, to mixed macrophyte/phytoplankton sediments at 50 cm, and finally to phytoplankton rich sediments at 30 cm. In our core, the concentrations of macrophyte derived n C 23 dominated the n alkane record beginning at ~60 cm core depth, before their decline at 50 cm, and eventual replacement by n C 17 as the dominant chain length at 40 cm. The T AR HC values are consistent with this trend, and their decrease at 40 cm indicates a switch to organic matter dominated by phytoplankton remains. Our P aq results confirm that terrestrial leaf waxes were never a significant source of n alkanes to Lake Apopka sediment. Rather, long chain n alkanes in sediments from this lake were primarily derived from aquatic macrophytes. Reliable 210 Pb dates for the Apopka core could not be obtained, but the age of the sediments is well constrained by the disappearance of un supported 210 Pb at 70 cm and we are confident that the rapid rise in phytoplankton hydrocarbon concentrations at 40 cm represents the primary producer shift that occurred in the 1940s.
29 Several authors have reported on the long term stability of aliphatic hydrocarbons (Ho and Meyers, 1994; Meyers, 2003; Eglington and Eglington, 2008). We argue that both long chain and short chain n alkanes were stable over the time frames represented by our cores. The carboxylic acids, however, show evidence of bacterial de gradation. Previous work demonstrates that bacteria commonly produce the following fatty acids: branched iso and anteiso C 15 :0 monounsaturated C 16:1 7 C 18 :1 7 C 18 :1 9 and the odd carbon number saturated fatty acids, C 15:0 and C 17:0 (Palomo and Canuel, 2010). In our study lakes, the BDFA concentrations follow two distinct trends (with all BDFA from Apopka following a single trend), implying two bact eria communities are synthesizing distinct fatty acids. This has also been reported in studies from estuaries (Zhang et al., 2015) and marine sediments (Parkes and Taylor, 1983). Presence of multiple bacteria communities in lakes is not surprising, given t he range of redox conditions in the water column and sediments, and studies have demonstrated that under anaerobic and aerobic settings, bacteria produce different fatty acids (Parkes and Taylor, 1983). Saturated fatty acids that are probably, but not excl usively, of autochthonous origin (i.e. C 12 16 ), also follow the up core trends of the BDFA, but the LCFA do not (Figure s 2 3 2 4 2 5 ). I hypothesize that these two groups track one another because the BDFA used the short chain fatty acids SCFA (C 12 16 ) a s a carbon source, and/or both groups increased simultaneously, albeit independently, as a consequence of an external forcing such as increased nutrient loading. If the former hypothesis is correct, then SCFA are not reliable indicators of organic matter s ource, as their labile nature makes them susceptible to bacterial consumption. If the latter is correct, then each of the three lakes displays increased bacterial (likely cyanobacteria)
30 and algal concentrations that are tied to anthropogenic nutrient input s. Whatever the case, concentrations of BDFA and SCFA are linked in the cores, and from this, we conclude that: 1) two separate groups of bacteria have created and/or consumed fatty acids in our study lakes, and 2) short chain fatty acids are susceptible t o bacterial reworking, but LCFA are not. The ability to differentiate between the ultimate sources of these FA will require additional research Biomarker data from each of the three lakes confirms that organic matter in the lacustrine sediments came from t errestrial and aquatic sources. In oligotrophic Lake Sheelar, low n C 17 concentrations throughout the core imply negligible algal input to the sediments. The dominance of n C 23 31 in the sediments confirms that organic matter in the sediment of Lake Sheela r is of primarily vascular plant origin, and that it has been a low productivity system throughout historic times. The Lake Wauberg and Apopka sediment cores both display increases in n C 17 in response to cultural activities. In Wauberg this increase occur red very recently, after watershed development in the 1980s. Non vascular plants, however, never became the dominant source of organic carbon to Wauberg, and P aq values, in conjunction with TAR HC values, demonstrate that submerged aquatic vegetation contri butes the most n alkanes. In Lake Apopka, the transition from a macrophyte dominated to a phytoplankton dominated state in the 1940s is documented by the n alkane concentration data. Over the last several decades, algae have contributed the majority of hyd rocarbons to the sediment record. Fatty acid biomarkers show evidence of bacterially mediated diagenesis, which may be post depositional. We cannot use SCFA data to definitively establish the source of these compounds in sediments, but future work on fatty acid biomarkers, especially
31 compound specific isotope analyses, should provide insights into the extent of post depositional changes. Conclusions: Sheelar, Wauberg, and Apopka Bulk and biomarker data from sediment cores indicate that Lake Sheelar has rema ined a low productivity system dominated by macrophyte and terrestrial inputs since at least the early 19 th century. Increases in algal contributions, measured as n C 17 concentrations, in Lake Wauberg in the 1980s and Lake Apopka in the 1940s demonstrate the detrimental impacts of anthropogenic influences on lacustrine systems. Both systems increased in trophic status and transitioned from macrophyte to algal dominated systems after human alteration of the landscape. In all three lakes, BDFA concentrations suggest increased bacterial contributions to the modern sediments, however, this trend may be the result of post depositional degradation of fatty acid biomarkers. Bulk elemental and isotopic data, while somewhat diagnostic, are plagued by myriad issues ( see overview in Bianchi and Canuel, 2011). Overall, the hydrocarbon data represents the most source specific and stable biomarker and therefore concentrations of n alkane chain lengths can be used for reconstructions of organic matter sources and trophic status over decadal to centennial timescales
32 Figure 2 1 Down core variability in bulk geochemical variables for lakes Sheelar, Wauberg, and Apopka.
33 Figure 2 2 Down core variability in select n alkane chain lengths for Sheelar, Apopka, and Wa uberg. The chain lengths are assigned to the following biological sources: n C 17 (algae), n C 23 (macrophytes), n C 29 (woody terrestrial vegetation).
34 Figure 2 3 Down core variability in select fatty acid chain lengths from Lake Sheelar. We grouped the bacterial derived fatty acids (BDFA) together with other non BDFA chain lengths based upon similar down core concentration patterns. The first row includes saturated, branched iso and anteiso C 15 :0 C 16:1 7 and C 18 :1 9 BDFA, and C 12:0 which follows a similar pattern. The second row includes C 17:0 and C 18 : 1 7 BDFA, and C 16:0 C 18:0 and C 18:2 which follow a similar pattern. In the final row are two representative LCFA. All concentrations are in /g OC.
35 Figure 2 4 Down core variability in select fatty acid chain lengths from Lake Wauberg. We grouped the bacterial derived fatty acids (BDFA) together with other non BDFA chain lengths based upon similar down core concentration patterns. The first row inc ludes saturated, branched iso and anteiso C 15 :0 C 16:1 7 and C 18 :1 9 BDFA, and C 12:0 and C 16 :0 which follow a similar pattern. The second row includes C 16:1 7 C 18 :1 9 and C 18 : 2 7 BDFA, and C 18 : 2 which follows a similar pattern. In the final row are two representative LCFA. All concentrations are in /g OC.
36 Figure 2 5 Down core variability in select fatty acid chain lengths from Lake Apopka. All bacterial derived fatty acids (BDFA) displayed similar down core concentration patterns, shown in the first row. The second row displays representative SCFA, and the final row representative LCFA. All concentrations are in /g OC.
37 Figure 2 6 Relative abundances (wt/wt) of nitrogen and organic carbon from the Lake Sheelar core.
38 Table 2 1 Lake Shee l ar alkane and fatty acid data: a verage chain length (ACL), Submerged to emergent/terrestrial vegetation ratio ( P aq ), the ratio of long to short chain hydrocarbons (TAR HC ), the sum of even carbon fatty acid concentrations ( g/ gOC) (LCFA), the sum of a lgal derived polyunsaturated fatty acids concentrations ( g/ gOC ) (PUFA) bacterial derived fatty acids concentrations ( g/ gOC ) (BDFA), and the ratio of long to short chain fatty acids (TAR FA ) Depth (cm) ACL P aq TAR HC LCFA PUFA BDFA TAR FA 2 28.81 0.36 88.06 5261.76 7.12 2262.08 0.53 8 28.31 0.51 21.61 7685.75 5.58 1299.64 1.08 16 28.13 0.53 17.37 8339.15 2.25 492.46 2.58 2 4 28.40 0.44 18.82 8570.78 1.79 410.92 3.31 3 4 28.14 0.50 21.10 8230.01 3.54 393.69 3.07 40 28.64 0.41 136.16 7785.54 1.94 316.72 3.32 46 28.31 0.47 18.22 5021.97 2.30 245.72 3.04 52 28.35 0.46 21.91 5207.13 2.04 268.53 3.26 58 28.42 0.44 24.97 4690.01 2.10 271.44 2.53 64 28.56 0.39 29.43 4871.15 1.56 236.24 2.96 70 28.63 0.38 42.66 5964.74 2.16 221.77 3.56 78 28.57 0.41 32.66 4571.05 0.95 153.56 4.90 Average 28.44 0.44 39.41 6349.92 2.78 547.73 2.84
39 Table 2 2 Lake Wauberg alkane and f atty acid data. See Table 2 1 for full description. Depth (cm) ACL P aq TAR HC LCFA PUFA BDFA TAR FA 3 28.08 0.40 1.33 3149.72 10.53 1925.83 0.39 8 28.16 0.41 1.57 3588.62 20.09 1490.64 0.61 16 28.19 0.40 2.60 3581.50 5.67 1163.40 0.70 24 28.06 0.43 5.25 3885.90 2.54 851.43 0.91 32 28.21 0.40 7.38 3969.21 6.14 579.74 1.68 40 28.19 0.41 10.17 3929.50 0.79 257.81 2.76 48 28.17 0.44 8.24 2366.77 1.34 165.18 2.69 56 28.15 0.46 8.89 4154.41 0.94 169.19 4.31 64 28.05 0.48 13.65 5296.46 3.87 251.22 4.16 72 28.13 0.51 19.63 2395.42 0.63 100.42 5.11 80 28.21 0.41 27.10 4944.09 1.03 137.68 7.06 88 28.11 0.44 26.00 3321.36 0.80 192.89 3.64 Average 28.14 0.43 10.98 3715.25 4.53 607.12 2.84
40 Table 2 3 Lake Apopka alkane and fatty acid data. Se e Table 2 1 for full description. Depth (cm) ACL P aq TAR HC LCFA PUFA BDFA TAR FA 2 28.14 0.48 0.75 2484.96 248.81 3639.91 0.29 8 28.08 0.47 0.82 1807.38 164.72 3147.25 0.23 16 28.08 0.49 1.46 3241.15 97.91 2190.96 0.42 26 28.06 0.49 2.62 4401.39 18.32 1636.55 0.66 34 28.06 0.51 3.14 4605.20 10.57 1379.17 0.78 42 27.99 0.50 5.10 4619.93 4.23 535.43 1.78 50 28.01 0.53 5.26 7142.13 3.35 413.36 3.03 58 27.97 0.50 4.32 6735.27 2.50 275.60 5.10 66 27.89 0.48 4.72 7073.56 4.08 206.46 7.07 74 27.91 0.48 4 .60 6895.52 1.93 142.18 8.69 82 27.89 0.46 4.69 6870.40 1.51 132.78 9.40 9 0 27.89 0.46 4.41 5493.12 2.25 152.53 6.95 Average 28.00 0.49 3.49 5114.17 46.68 1154.35 3.70
41 CHAPTER 3 THE BIOGEOCHEMICAL EVOLUTION OF A SUBTROPICAL LAKE Introduction : Lake Harris Lakes are a critical component of terrestrial carbon cycling. Their sedimentary archives contain a history of changes in organic carbon sources within the lakes watershed (allochthonous) and the water column (autochtonous). They serve as a pool for sequestered organic carbon that is estimated to exceed that of the ocean by a factor of three (Dean and Gorham, 1998). A nalysis of the organic carbon preserved within a lacustrine sedimentary archive can be used to reconstruct complex environmental changes throughout the lake s existence. In Florida, paleolimnological reconstructions with organic biomarkers have been used to document ecological successions (Watts and Hansen, 1994) and hydrology changes (reviewed in Castaeda and Schouten, 2011) during the t ransition from the last glacial period to the present. These reconstructions routinely employ elemental C/N ratios and stable isotopes 13 C TOC ) to distinguish sources of organic matter within lake sediments (Meyer s 1994; Silliman et al., 1996). Differences in C/N ratios arise from the absence of carbo hydrate rich structural components in algae (Meyers and Ishowatar, 1993), and values for vascular (C/N >20) and non vascular plants (C/N between 4 and 10) have been used to gauge the relative contributions of the two end member components to the organic se diment pool. The range in C/N values between algal and woody plant species demonstrates how basic bulk elemental data can be applied a s a proxy for 13 C TOC can be interpreted as a mixture of values from various end 13 C values of plant and algal organic matter is primarily governed by the isotopic signature of the carbon substrate (a tmospheric CO 2
42 or dissolved HCO 3 ) and the combined enzymatic fractionations associated with fixing through two photosynthetic pathways (C 3 or C 4 13 C signat ures vary between these two groups (Farquhar et al., 1989). Although algae photosynthesize using the C 3 pathway, they are enriched in 13 C relative to C 3 plants in high pH systems because dissolved HCO 3 is enriched in 13 C relative to CO 2( aq) y 1988). Generalizations about carbon sources inferred from C/N ratios must be interpreted with caution. Recent studies have shown that C/N ratios vary considerably at the species level (Cloern et al., 2002), can be greater in N limited systems (Bianchi an d Canuel, 2011), and lower in systems where there is selective degradation of labile 13 C TOC signatures in relation to organic matter provenance is not unequivocal. In lake sediments, stratigraphic shifts 13 C values have been explained as reflecting greater input from algal material (Meyer, 1997), increased autotrophic carbon fixation (Brenner et al., 1999; Torres et al., 2012), or transitions from C 3 to C 4 terrestrial plant co mmunities (Bianchi et al., 2002). An alternative to analysis of bulk organic matter in lake sediments is the application of compound specific isotope measurements of molecular biomarkers as a tool for trac k ing the source and history of organic matter withi n the lacustrine system. With biomarker specific studies, reconstructions of complex hydrology changes have been carried out across millennial time scales (Filley et al., 2001). Carbon isotopic values of lipid compounds from lake sediments (Mud Lake, Flori da, USA) were used by Filley et al. (2001) to document vegetation changes associated with Holocene climate
43 shifts. They found that shifts in alkane biomarkers corresponded to fluctuations in the regional water table that were caused by hydrological changes in the early Holocene. The changes recorded in the carbon isotopes of the n alkanes mirrored similar changes from p ollen based environmental reconstructions from other Florida lakes (Watts et al., 1994). A recent study of Lake Harris, FL, USA, used multi ple geochemical proxies to identify prehistoric shifts in the primary producer community structure, from a sediment core dated to ~10,000 cal yrs BP (Kenney et al., 2016) A shift from carbonate dominated to organic dominated sediment occurred at approxima tely 5,540 yrs BP (Kenney et al., 2016). The timing of this shift may have been related to the onset of modern hydrological conditions in Florida, as inferred from pollen data (Watts el al. 1994), and core studies of basal peat layers form the Mississippi delta (Trnqvist et al., 2004). The primary objectives of this study we re two fold: 1) to assess the major sources of organic matter to the sediments of Lake Harris from its nascent stages to the present, and 2) to determine the environmental and hydrolog ical changes that drove the shifts in organic matter. We employed a suite of geochemical analyses throughout a complete Holocene sediment record from Lake Harris to evaluate these objectives quantitatively The concentrations and stable isotope composition s of total sedimentary organic carbon (TOC), total nitrogen (TN), and select biomarker hydrocarbons, were used as proxies for measuring temporal trends in the primary producer communities. Geochemical analyses, including, concentrations and accumulation ra tes of carbonate (CaCO 3 ), biogenic silica (BioSi), and total organic carbon (TOC) were used as proxies
44 for hydrological changes, such as, increases or decreases in: groundwater input, lake level, and water table elevation. I hypothesize that the sources of organic matter (i.e. allochthonous vs. autoch th onous regional hydrologic and environment al events that occurred as the e stabilized in the mid Holocene. Specifically, these climate events incl ude: the steady rise in relative sea level in the Gulf of Mexico that began in the early Holocene and abruptly decreased at 7,000 yrs BP (Trnqvist et al., 2004), wetland formation and expansion in Florida at 5,000 yrs BP (Willard and Bernhardt, 2011), and the onset of the modern El Nio cycle at 3, 000 yrs BP (Donders et al., 2008 ). The data demonstrate that sedimentary variables in Lake Harris responded to each of these events and showed a progressive trend towards eutrophication in the modern sediments. Site Description : Lake Harris Lake Harris is a relatively shallow (mean depth = 3.5m), productive lake (mean total nitrogen = 1707 ), with a mean surface area of 75km 2 (Fulton and Smith, 2008). It is part of the Lake Harris chain of lakes located in the Upper Ocklawaha River Basin, Central Florida, USA. Flow through the system begins in a spring that emanates from the southwest section of hypereutrophic Lake Apopka and continues through four additional lakes and discharges i nto the Ocklawaha River (Kenney et al., 2010). Three other lakes in the chain, including Lake Harris, receive minimal flow from Lake Apopka and are all considered to be mesotrophic to eutrophic The hydraulic retention time for Harris is relatively short ( 2.9 years) and can vary by orders of magnitude during periods of high and low flow (Fulton and Smith, 2008). Most of the lakes in the Harris chain receive surface and groundwater inputs that pass through organic and mineral rich soils, thus,
45 are hypothesi zed to be naturally productive and alkaline (Fulton and Smith, 2008). However, recent geochemical analyses of the same 5.9m sediment core from Lake Harris used in this study placed the lake on a trajectory towards increasing oligotrophication throughout th e Holocene ( K e nney et al., 2016). This trajectory was interrupted after AD 1950, when the human population around the region increased six fold and the lake veered towards the highly productive end of the trophic state spectrum (Kenney et al., 2016). Sed iment Sampling : Lake Harris We sampled sediment from the 5.9m core retrieved by K e nney et al. (2016). The upper portion of the core (first 1m) was dated using 210 Pb, and the remaining portion of the core was AMS 14 C dated using four charcoal, and three bul k sediment samples. For details of the chronology see K e nney et al. ( 201 6 ). The core was split in half, lengthwise, and one half of the core was sampled at 4 cm intervals for geochemical analyses. Wet subsamples were then frozen, freeze dried, and ground and homogenized for analyses. CaCO 3 BioSi (both diatom and sponge spicule derived), TN, and %OM sampling methods are outlined in K e nney et al. (2016). TOC was measured as the difference between total carbon and inorganic carbon on a Carlo Erba NA1500 CNS elemental analyzer equipped with an auto sampler. Dried 13 C TOC analyses was first pretreated with 1N HCl to remove inorganic carbon, and then measured on a Carlo Erba NA1500 CNS elemental analyzer interfaced with a Thermo Scientific Delta V Advantage isotope ratio mass spectrometer. Isotopic compositions were normalized to the VPDB scale, and reported in standard delta notation as follows:
46 13 13 13 C standard) 1] x 1000 (2 1) Lipid E xtraction and Q uantification : Lake Harris Lipids were extracted from 1g of freeze dried sediment with an Accelerated Solvent Extractor ASE200 (Dionex), using 2:1 (v/v) dichloromethane(DCM):methanol through three extraction cycles at 10.3 MPa (1500 psi) and 100 C. Between 1 and 2 g of sediment were used for lipid extraction. Total lipid extracts (TLE) were concentrated under a gentle stream of nitrogen, and the neutral lipid fraction was obtained after base saponification of the TLE. Neutral lipids were further separated, based on polarity, into com pound classes by column chromatography, using 5% deactivated silica gel, according to methods modified from Nichols (2011). Hydrocarbons were eluted from the silica gel column with 4.5 ml of 9:1 Hexane:DCM, and saturated hydrocarbons were separated from al kenes on 5% Ag impregnated silica gel (w/w) with 4 ml of hexane and ethyl acetate, respectively. Branched and cyclic saturated hydrocarbons were isolated from n alkanes with triple urea adduction. Alkane concentrations were measured and identified on a Th ermo Scientific Trace 1310 gas chromatograph with a Supelco Equity 5 column, interfaced to a Thermo Scientific TSQ 8000 triple quadrupole mass spectrometer with electron ionization. The inlet was operated in splitless mode at 280 C. The column flow rate wa s set to 2.0 ml/min and the oven was program m ed to an initial temperature of 60 C and held for 1 minute, then ramped to 140 C at 15 C/min, and to 320 C at 4 C/min and held for 25 minutes. Quantification was based on the calibration curves generated from th e peak areas of external standards (C 7 C 40 ) with concentrations ranging from 5 to 250 g/ml.
47 Compound specific I sot ope M easurements on n A lkanes : Lake Harris Compound specific carbon isotope values for n alkanes were measured on an Agilent 6890 GC connect ed to a Delta V Advantage IRMS interfaced with a GC C combustion system. The GC flow rate was set to 2.0 ml/min, and the oven was programmed as follows: 60 C for 1 minute, then increased at a rate of 6C/min to 320C, and held for 20 minutes. Compounds wer e combusted over a nickel/platinum/copper wire with O 2 at 960 C. The isotope ratios of carbon in CO 2 were measured and normalized to the VPDB scale using the Uncertainty Calculator (Polissar s above. Standard errors of the mean (SE) were calculated using the Uncertainty Calculator, which yielded a 1 SE value of 0.39 Bulk Geochemistry and Isotopic C ompositions : Lake Harris The core was divided into three zone s based on shifts in 13 C TO C : from the bottom of the core to 403 cm is zone 3, from 402 to 241 cm is zone 2 and from 240 cm depth to the surface is zone 1 (Figure 3 1 ). 13 C TOC values upcore, from a minimum of a maximum of isotope values remain relatively stable in zone 2 (mean sharply in zone 1 from carbon percentages remain below 15% throughout zone 3, while %CaCO 3 decreases from a maximum value of 86% at 579 cm depth, to just under 30% at 418 cm, with an average value of ~60% throughout zone 3. Values of %TOC and %CaCO 3 change suddenly at ~350 cm depth in zone 2: values of organic carbon increase fro m 2.7% to 38.1%, and carbonate drops to 1% and never increases above 25% in zone 2. In zone
48 1, %TOC remains at elevated levels and varies about its mean of 21.3%. Carbonate values never exceed 11% from 240 cm to the surface of the core and display average zone 1 values of 0.6%. In zone 3, TOC:TN values reach a maximum of 39.33 at 574 cm, before decreasing rapidly into zone 2 to a minimum value of 10.91. TOC:TN values continue to decrease slowly from zone 2 to zone 1 and reach a minimum value of 8.57 at 8 cm depth. Biogenic silica, both diatom and sponge derived, displays considerable variability throughout all zones. Of note is a dramatic increase in diatom silica at 144 cm, where concentrations increase from 11.96 to 51.96 mg/g. These values continue to ri se in zone 1, reaching a maximum value of 147.30 mg/g at 52 cm. Concentrations and Isotopic C ompositions of H ydrocarbons : Lake Harris Concentrations for select n alkane chain lengths are displayed in F igure 3 3 Mean n alkane concentrations for these chain lengths are displayed in Table 3 1 The n a lkane concentrations are dominated by long chain alkanes (i.e. > n C 25 ). The most abundant n alkane in our record is n C 27 with an average concentration of 36.01 g/g OC throughout the entire core. We calculated the average chain length (ACL) for all samples using the Eglinton and Hamilton (1967) equation: ACL(C 25 C 35 ) = (2 2) There is low variability in ACL througho 13 C TOC zones 1 and 3 (ranges = 1.2 and 1.1, respectively), however ACL values in zone 2 vary by 2.8, with a high of 29.7 and a low of 26.9 (Table 3 2 ). Carbon preference index (CPI) measures the maturity of the hydrocarbon source. We calculated CPI va lues from the equation of Bray and Evans (1961):
49 CPI = 0.5 + (2 3) The CPI values for all hydrocarbon samples averaged 3.68, with a maximum of 6.56 and a minimum of 2.14 (Table 3 2) All sample depth s exhibit an odd over even predominance, which is indicative of vascular terrestrial plant sources. There is not a clearly discern i ble trend in CPI values throughout t he three core zones. Differences in alkane concentrations are apparent across each zone in the Lake Harris core (Figure 3 3) We subdivided alkanes into groups based upon their most probable source(s): cyanobacterial (7 methylheptadecane and diploptene), a lgal ( n C 17 ), emergent/submerged macrophytes ( n C 23 ), woody terrestrial vegetation ( n C 27 and n C 29 ), and mixed woody terrestrial/C 4 grasses ( n C 35 ) and related changes in these organic matter sources to regional environmental change. In broadest terms, t here are three significant trends in alkane concentrations: 1) proliferation of algal and cyanobacterial biomarkers in the top 40 cm of the core, 2) a rapid increase in submerged macrophyte biomarker concentrations at 390 cm, which persisted until 330 cm, and 3) the dominance of terrestrial biomarkers throughout the rest of the core, with a concentration maximum at 530 cm. More specifically, zone 3 is dominated by terrestrial alkane inputs: n C 27 and n C 29 concentrations are ~5x that of n C 23 and ~20x that of n C 17 P aq values are also indicative of primarily terrestrial inputs in this zone. A sudden increase in concentrations of n C 23 n C 27 n C 29 and n C 35 occurs at 530 cm. Both terrestrial hydrocarbon biomarkers attain their maxima at this depth. In zo ne 2 the average concentration of n C 17 decreases 8 fold, 7 methylheptadecane is nonexistent, and the terrestrial alkane concentrations are
50 approximately halved, although they remain the most abundant chain lengths in this zone. Of note is an increase in n C 23 concentrations from 6.23 g/g OC at 340 cm to 32.51 g/g OC at 330 cm, which then drops back down to 6.85 g/g OC at 280 cm. A similar, albeit smaller increase in n C 17 concentrations also occurs during this in terval (Figure 3 2 ). We measured the rela tive contributions of submerged and floating aquatic macrophytes to emergent terrestrial vegetation via the P aq proxy (Ficken et al 2000), described by the formula (C 23 + C 25 )/(C 23 + C 25 + C 29 + C 31 ) P aq values from 330 cm to 290 cm range from 0.42 to 0 .81, indicative of a freshwater submerged vegetation source. This is the only portion of the Lake Harris core to have P aq values that fall in the range of submerged vegetation. The top zone of the core is characterized by a ~50 fold increase in the conce ntration of n C 17 and a ~40 fold increase in 7 methylheptadecane concentrations in the upper 40 cm of sediments. These are the hydrocarbons in the highest abundance in the top 40 cm of the core, but their concentrations drop to < 5 g/g OC below 120 cm. At this depth, and throughout the rest of zone 1, the terrestrial biomarkers n C 27 and n C 29 become the most abundant hydrocarbons with average concentrations of 52.32 g/g OC and 43.25 g/g OC, respectively. A hydrocarbon proxy, known as the terrestrial t o aquatic ratio (TAR HC ), was developed to distinguish between autochthonous and allochthonous source s of organic matter in sediments (Meyers, 1997). TAR HC results, unsurprisingly, matched our comparative analyses of single chain length alkane concentration s (Table 3 2). Values <1 reflect a dominance of aquatic hydrocarbons, and are only observed in the top 15 cm of our core. TAR HC are highly variable, but average 16.40, which indicates that
51 contributions from terrestrial plants dominate the record. The maxi mum TAR HC (74.75) is found at 530 cm, whereas values <4 occur at 510, 400 360, 330 290, and 35 cm. Isotopic compositions of selected cha in lengths are shown in Figure 3 4 Based upon alkane concentrations, and organic matter source assignments, we interpr eted only the following hydrocarbons in t his study: 7 methylheptadecane n C 17 n C 23 n C 27 and n C 29 Reliable carbon isotope ratios could not be measured on n C 17 and n C 23 below 415 cm depth. The carbon isotopes of 7 methylheptadecane have the highes t average value in 13 C values of this branched alkane average 13 C signature of n C 17 decreases progressively across all zones from a maximum value o f 13 C of n C 23 increase s from s at a mean value of s of n C 23 decrease from a maximum of 13 C values of vascular plant alkanes, n C 27 and n C 29 both increase in a pattern that follows the enrich ment of 13 C TOC Carbon isotope signatures then rapidly decrease at 362 cm for n C 27 and at 400 cm for n C 29 These chain lengths return relatively invariable 13 C values in core zone 1, with average values of 24.8 and 28.0 for n C 27 and n C 2 9 resp ectively (Figure 3 4 ).
52 Discussion Overview The shifts in geochemical variables measured in the sediments of Lake Harris coincide with major environmental events that transpired during the Holocene. Hydrocarbon biomarker data indicate a series of hydrologi cal shifts that resulted from Northern Hemisphere deglaciation. These shifts in Lake Harris are temporally linked to records from other lakes in north and central Florida (Watts and Hansen, 1994; Filley et al., 2001; Donar et al., 2009). Broadly speaking, Lake Harris evolved from a marsh like system in the early Holocene, to a shallow lake in the middle Holocene, and then deepened to its modern state after ~2,600 years BP. The carbon isotopes of the autochthonous and allochthonous organic matter pools are p roxies for these changes, and reflect multiple sources of inorganic carbon utilization by the primary producer communities. Concentrations of n Alkanes and Shifts in Geochemical Biomarkers as Indicators of Organic Matter Source in Lake Harris The biologic al sources of alkanes have been described in numerous publications (for review see Bianchi and Canuel, 2011). The n C 27/29 homologues of n alkanes represent organic matter sourced from woody terrigenous plants (Cranwell, 1982). C 4 graminoids contain n C 35 alkanes in their leaves at concentrations an order of magnitude greater than woody C 3 angiosperms (Garci n et al., 2014; Diefendorf and Freimuth 2017 ). Submerged and emergent macrophytes are characterized by chain lengths of n C 21/23 (Ficken et al., 2000), and algae synthesize alkanes with a predomina nt chain length of n C 17 (Cranwell 1982). Additionally, the branched alkanes, 7 and 8 methylheptadecane, are produced exclusively by cyanobacteria and account for 90% of their total branched alkanes (Han and C alvin, 1969). Although hydrocarbon
53 biomarkers are source specific, they represent only a fraction of the total biochemical composition of bacteria and plants. The biomarker data gleaned from this study can supplement, and be supplemented by, bulk analyses that represent larger fractions of the organic matter. The earliest portion of the core (zone 3) is dominated by n alkanes derived from terrestrial and, to a lesser extent, macrophyte sources, as well as high sponge spicule derived biogenic silica. Approx imately 800 years (at 550 cm core depth) after the lake began to fill with water in the early Holocene (10,600 yrs BP), there was a pulse of n C 23 n C 27 and n C 29 to the sediments (Figure 3 3 ). This occurs shortly after TOC:TN values jump from 24 to 39 a t 574 cm. TOC:TN values continue to increase towards maximum values in agreement with terrestrial biomarker abundances. This is the earliest stage in the lake s evolution when it was a shallow, low productivity, marsh like system that received abundant v ascular plant input from littoral communities (e.g. Taxodium spp ), and supported macrophyte growth. The temporal correlation between the TOC:TN, sponge spicule, and terrestrial hydrocarbon biomarker increases is in strong agreement with palynological (Wat ts, 1975) and diatom based paleolimnological recon structions (Quillen et al., 2009 ) that describe central Florida lakes as low nutrient, shallow water systems during the early Holocene. Zone 2 is highlighted by localized maxima of algae and ma crophyte biom arkers (Figure 3 3 ) The increase in n C 23 begins at 358 cm (~6 700 yrs BP) which corresponds to a shift from predominantly CaCO 3 sediments to organic rich sediments without measurable carbonate by 232 cm (~5,000 yrs. BP) (Fig ure 3 2 ). The carbonate colla pse is coincident with the reduction in freshwater sponge derived biogenic silica,
54 ACL values below 27, and average P aq values of 0.61 (Fig ure 3 2 ), are indicative of extensive macrophyte coverage in Lake Harris. Kenny et al. (2016) correlated these change s in geochemical proxies with the proliferation of Pinus pollen in Florida and the onset of wetter conditions in the mid dle Holocene (Watts and Hansen, 1994). During this interval, hydrologic inputs to Lake Harris shifted from a primarily groundwater sourc e to a primarily meteoric source. Lake Panasoffkee, in central FL, is a modern analogue to Lake Harris during the early to mid dle Holocene. Currently, Lake Panasoffkee receives substantial groundwater inputs and is a hard water, macrophyte dominated syste m (Brenner et al., 2006). TOC:TN values of submerged aquatic vegetation (SAV) measured in the Brenner et al. (2006) study (14 22) are only slightly greater than the average TOC:TN values in zone 2 of the Lake Harris core (12.1). From this, we can infer tha t sedimentary organic carbon in Lake Harris from ~8,000 5,000 yrs BP was a mixture of macrophytes and structurally poor, low TOC:TN, algae and periphyton. This reconstruction is further supported by vegetation studies from shallow, low nutrient lakes that have primary producer communities dominated by SAV (Schelske et al., 2005). The n alkane biomarker concentrations, with maxima of n C 17 (at ~400 and 330 cm) and n C 23 (at ~330 cm), the sponge derived silica concentrations, and the TOC:TN data measured from modern SAV, periphyton, and algae, coincide with one another in this section of the core. Together they indicate a primarily groundwater fed system with abundant submerged and emergent macrophyte populations. The largest shift in any of the hydrocarbon co ncentrations occurs in zone 1, over the top ~40 cm of the core, where n C 17 and 7 methylheptadecane increase by 50 fold
55 and 40 fold, respectively. This rapid increase in algal and cyanobacterial biomarkers corresponds with the timing of an increase in trop hic status of other Florida lakes and a six fold population increase in the counties surrounding the Harris chain of lakes after AD ~1940 (Fulton and Smith, 2008). The 7 methylheptadecane concentrations, in particular, are indicative of a highly eutrophic, possibly N limited lake that supports cyanobacteria proliferation (Dolman et al., 2012). Terrestrial ( n C 27 ) and macrophyte ( n C 23 ) biomarker concentrations fluctuate between high and low values, but display no discernible trend across zone 1. Biomarker c oncentrations from 40 cm depth to the bottom of zone 1 (240 cm) are more evenly distributed among the three major organic matter sources. Combined algal and macrophyte n alkane concentrations equate to approximately half of the terrestrial n alkane concent rations. The observed trends in the biomarker data are not clearly borne out by the bulk geochemical data. The concentrations of algal and bacterial biomarkers are an order of magnitude greater than any terrestrial biomarker throughout the top 40 cm of th e core, yet the TOC:TN values do not decrease as would be expected with an increase in non vascular sources. Instead, TOC:TN values remain relatively stable throughout zone 1, with an average value of 10.16, indicative of a primarily algal source (Meyers, 2003). Post depositional alteration of TOC:TN values occurs in sediments with oxic pore waters, which can result in artificially low values (Bianchi and Canuel, 2011). Organic matter decomposition can remove as much as 20% of the carbon that is buried on t he lake bottom (Meyers, 2003; Glman et al., 2008). Although such alteration is possible in Lake Harris, it is likely to occur during the winter when the water column loses its thermal stratification and bottom waters become oxygenated. Even if the lake is
56 thermally stratified, sediment trap studies have shown that much of the carbon in the water column is oxidized before leaving the epilimnion. In Lake Michigan, for instance, only 6% of the organic carbon formed via autochthonous production reached the sed iment surface (Eadie et al., 1984). We cannot expect sinking particulate organic carbon to have similar exposure rates to oxidation in Lake Harris due to vast differences in water depth compared with Lake Michigan but we can assume that a significant frac tion is lost before being integrated into the sediment archive. Finally, algae are protein rich but carbohydrate and lipid poor. Lipids account for variable amounts of their biochemical composition, but are generally between 5 20% of the total constituen ts and hydrocarbons account for only 3 5% of the lipid fraction (Bianchi and Canuel, 2011). The limited n alkane production from algae coupled with carbon loss is the most probable explanation for the incongruous TOC:TN and hydrocarbon concentration data Interpreting Carbon Isotope Variability in TOC and Hydrocarbon Biomarkers in Lake Harris The carbon isotope value 13 C TOC values in the 13 C TOC values in the middle section, then back to depleted values in the latest section (Figure 3 1 ). This variability reflects the changing sources of c arbon utilized by the primary producer communities. In the early Holocene, Lake Harris began to fill as the water table in the region started to rise in response to deglaciation (Watts and Stuiver, 1980). Most lakes in Florida began to accumulate sediment at this time, but remained as shallow marsh systems for millennia as precipitation and water table elevations were both lower during the early Holocene (Watts, 1971; Quillen et al., 2013). The carbon isotope values of n C 27 13 C TOC
57 13 C values of n C 27 during alkane synthesis, plot within the isotopic range of TOC values. The environmental reconstructions and carbon isotope data indicate that the major source of carbon to the lake sediments was from terrestrial plants that utilize d atmospheric CO 2 As Florida gradually became wetter during the middle Holocene (between 7,000 and 5,500 yrs BP, zone 2 in our record) the sclerophyllous oak and open prairie plant communities were replaced by modern vegetati on communities dominated by pine forests (Watts and Hansen, 1994; Donders et al., 2014). The fir st significant change in 13 C TOC values occur red at ~6,500 yrs BP (350 cm core depth ) and corresponds to this 13 C TOC that a new source of carbon wa s being utilized by the primary pro ducers. The average 13 C TOC 4 vegetation, meaning a shift from C 3 to C 4 13 C TOC enrichment. T his however, is improbable for two reasons. First, C 4 plants are adapted to low p CO 2 and arid environments (Ehleringer et al., 1997), and the mid dle Holocene was a period of increasing precipitation and relatively high p CO 2 Second, our n C 35 record, which is a proxy for graminoid abundance (Diefendorf and Freimuth 2017), and the record of C 3 /C 4 changes from nearby Lake Tulane (Huang et al., 2006), do not indicate an increase in the relative abundance of C 4 plants during this time interval. It is also possible that the 13 C enrichment wa s a consequence of decreased carbon isotope discrimi nation among phytoplankton under highly eutrophic conditions. This has been observed in multiple high productivity lacustrine systems in Florida (Gu et al., 2006; Torres et al., 2013), and has been attributed to recent cultural eutrophication in others
58 (Br 13 C TOC enrichment, then Lake Harris would have been eutrophic during the mid dle Holocene, but Kenney et al. (2016) demonstrated that nutrient accumulation was much lower at that time compared to modern nu trient accumulation. I 13 C TOC record in zone 2 reflects a shift to a more enriched carbon source for the in lake primary producer 13 C signature of n C 23 is the only possible source for this enrichment, as: 1) this chain l ength exhibits the 13 C values across the depth interval 350 250 cm, 2) these values, once isotope data, and 3) the n C 23 carbon isotope profile tracks the bulk isot ope data throughout zone 2. Lake Harris had deepened as the modern hydrosphere began to develop. Algae initially filled this newly created niche. From 430 to 400 cm there is a ~15 fold increase in n C 17 concentrations. In Lake Griffin, which is part of the Harris chain of lakes, a similar m id Holocene peak in algal abundance was recorded in cyanotoxins (Waters, 2016). I speculate that these findings indicate a relatively rapid rise in water levels that supported phytoplankton populations before the prolifer ation of macrophytes indicated by a ~30 fold increase in n C 23 contributions to the sediment at 380 to 330 cm. There are two possible explanations for the 13 C enrichment. First, the CaCO 3 levels recorded in the sediment from the beginning of the zone to 350 cm core depth indicate a well buffered system with high pH (>9). Bicarbonate ( HCO 3 ) is the major form of DIC at pH >9, and active uptake of this HCO 3 would enrich the phytoplankton and 2(aq) (Mook et al., 1974). This c ould explain the high
59 13 C values measured in n C 23 13 C values in two SAV taxa ( Vallisineria sp. and Potamogeton sp.) in nearby Lake 1 3 C range ( 13 C of the bulk organic matter in 13 C in zone 2 indicates proliferation of Vallisineria sp. and Potamogeton sp. i n the lake. From 250 cm c ore depth to the present (zone 1 ), the organic matter becomes progressively more depleted in 13 C. This depletion coincides with a second stage of increased precipitation in Florida that occurred between 5,000 and 3,000 yrs BP (Donders, 2014). Pollen based reconstructio ns of summer precipitation show persistently wetter conditions after 3,000 yrs BP which we re related to the intensification of ENSO (Donders, 2014). During the latter period, south Florida transitioned from a wet prairie to swamp forest environment (Donde rs et al., 2005). F a rther north, Quillen et al. (2013) linked changes in benthic diatom assemblages to an increase in water depth in Lake Anni e around the same time that Don ders et al. (2005) noted ENSO intensification. Our geochemical proxies indicate tha t Lake Harris also achieved its modern water depth around this time. Organic matter abundance stabilize d at levels >50% and CaCO 3 drop ped to trace leve ls after ~4,800 yrs BP (Figure 3 2 ). Diatom biogenic silica concentrations overtake sponge derived silica concentrations at ~2,800 yrs BP. Together these proxies track the transition from a primarily shallow, groundwater fed lake, to a relatively deep lake, whose depth wa s controlled by precipitation and/or surface water inputs.
60 Our n alkane data also suppor t this interpretation. The shift towards more 13 C TOC values in zone 3 results from increased algal contributions to the sediment. The algal proxy n C 17 13 C values among all n alkanes analyzed, and the transition to more neg ative values in the bulk isotope data signifies the rising contribution of algae to the organic matter pool. Interestingly, this is the opposite trend observed by Filley et al. (2001) in Mud Lake Marion County, FL In that 13 C val ues were recorded in n C 17 We argue that the photic zone to water column ratio is larger in Mud Lake, and therefore isotopic discrimination 13 C TOC signature. The dissolved inorganic ca rbon (DIC) pool in Lake Harris is larger, because it is deeper, thus primary productivity will have a less significant impact on altering the isotopic signature of the DIC pool. The cyanobacteria biomarker, 7 methylheptadecane, is the biomarker in highest 13 C values 13 C TOC values do not become more positive up 13 C TOC values would be expected as the la ke bec ame more eutrophic and cyanobacterial blooms thrive d (Brenner et al., 1999). Currently, I have no explanation for the dis crepancy between 7 methylheptadecane 13 C TOC values. The 13 C values ( are near ly identical to values measured from 7 methylheptadecane ( the top 30 cm of Mud Lake (Filley et al., 2001), implying accurate isotopic measurement. Diagenetic alteration of the cyanobacteria signal is unlikely as these compounds are
61 present in ab undant concentrations in our record, and are major components in the surface sediments of other Florida lakes (Riedinger Whitmore, 2005; Water s 2016). Conclusions: Lake Harris The evolution of Lake Harris is recorded in the geochemical proxies preserved w ithin its sediments. The lake began to fill with water ~10,000 yrs BP, when many lakes in Florida began to accumulate sediment in response to wetter conditions caused by deglaciation and eustatic sea level rise. From the bottom of the core to 403 cm depth (~7,800 yrs BP), terrestrial carbon inputs dominated the record, with limited input from macrophytes and algae, and the primary carbon source was atmospheric CO 2 Throughout the early Holocene, Lake Harris was a marsh like system in a relatively dry, open 13 C TOC values at 402 cm, and stabilization about these values at 350 cm, represents the onset of wetter conditions in Florida. Elevated CaCO 3 levels in the sediments indicate that Lake Harris was primarily f ed by groundwater, and isotopic signatures of n alkanes and bulk organic matter suggest that organic matter was primarily derived from macrophytes that utilize enriched HCO 3 as their carbon source. It was during th at time between 7,000 and 5,000 yrs BP that Florida transitioned from dry oak scrub vegetation to pine forests. Above 240 cm co 13 C TOC values begin to decline, CaCO 3 levels in the sediments decrease to below trace levels, and organic carbon concentrations increase. Around th at time (3,000 yrs BP) the effects of ENSO intensified, and many Florida lakes deepened to their c urrent limnetic state. This is observed in Lake Harris at ~2,900 yrs BP, when diatom biogenic silica concentrations increase from 10 to 120 mg/g. Concentrations of algal and cyanobacteria biomarkers increase by orders of magnitude after about AD 1940 in re sponse to human induced eutrophication.
62 Figure 3 1 Down core c arbon isotope variability of TOC in the Lake Harris core. The 13 C TOC (shown in the figure as dashed lines). Zone 3 is from 590 to 403 cm zone 2 is from 402 to 241 cm, and zone 1 extends from 240 cm to the core surface
63 Figure 3 2 Bulk geoche mical variability in the Lake Harris core.
64 Figure 3 3 Select n alkane chain length abundances in the Lake Harris core. Dashed lines delineate core zones 1 3.
65 Figure 3 4 Carbon isotope variability in select n alkane c hain lengths in the Lake Harris core. 0 100 200 300 400 500 600 35.00 25.00 15.00 Depth (cm) 13 C TOC C TOC C C17 C C23 C C27
66 Table 3 1. Concentrations, maxima, minima, and ranges for the five n alkane homologues used for organic matter source interpretation. Zones are 13 C TOC values. n alkane C 17 g/gOC C 23 g/gOC C 27 g/gOC C 29 g/gOC C 35 g/gOC Average 34.06 10.04 36.10 30.30 6.99 Max 482.57 32.51 120.00 79.70 50.45 Min 0.00 0.81 1.65 1.71 0.00 Range 482.57 31.70 118.35 77.99 50.45 Zone 1 Average 81.41 11.23 50.95 40.15 8.10 Zone 2 Average 10.63 12.50 29.22 26.28 4.59 Zone 3 Average 1.16 5.00 25.15 22.28 8.76
67 Table 3 2. Absolute values for carbon preference index (CPI ), average chain length (ACL), and the ratio of submerged to emergent vegetation (Paq). Dates determined by the 210 Pb method are in Anno Domini (AD) and carbon 14 dated layers are in years before present (yrs BP), with BP set to 1950 AD, as is the conventi on. Depth (cm) CPI ACL P aq Age Aging Method 4 3.53 28.72 0.47 2011.5 210 Pb AD 14 6.56 29.67 0.29 1998.3 210 Pb AD 36 4.01 29.27 0.30 1943.0 210 Pb AD 44 2.74 28.99 0.38 1911.8 210 Pb AD 70 3.36 29.17 0.30 900.0 C14 yrs BP 120 2.88 29.05 0.34 1916.7 C14 yrs BP 134 3.35 28.44 0.31 2398.0 C14 yrs BP 150 2.52 29.38 0.25 2775.9 C14 yrs BP 177 3.47 29.31 0.24 3565.4 C14 yrs BP 206 2.57 29.12 0.29 4770.1 C14 yrs BP 229 3.88 29.25 0.23 4843.2 C14 yrs BP 252 3.10 29.27 0.27 5199.8 C14 yrs BP 269 3.94 29.2 9 0.21 5449.5 C14 yrs BP 282 3.12 29.48 0.24 5620.0 C14 yrs BP 297 3.82 27.50 0.61 5840.5 C14 yrs BP 313 2.93 29.06 0.42 6049.6 C14 yrs BP 330 2.53 26.92 0.81 6252.4 C14 yrs BP 343 3.91 29.50 0.24 6425.4 C14 yrs BP 358 3.58 29.53 0.22 6741.6 C14 yrs BP 362 5.98 29.73 0.21 6826.3 C14 yrs BP 382 2.81 28.71 0.22 7770.0 C14 yrs BP 404 3.30 29.64 0.30 7845.3 C14 yrs BP 417 4.95 30.15 0.12 8019.3 C14 yrs BP 430 5.92 29.85 0.13 8167.3 C14 yrs BP 462 4.32 29.43 0.20 8477.8 C14 yrs BP 490 4.31 29.29 0.2 9 8827.0 C14 yrs BP 510 n/a 29.45 0.34 9300.0 C14 yrs BP 530 5.03 29.97 0.32 9472.3 C14 yrs BP 550 2.14 29.03 0.37 9832.4 C14 yrs BP 570 2.31 29.77 0.30 9900.0 C14 yrs BP
68 CHAPTER 4 SUBTROPICAL CLIMATE RESPONSE TO HEINRICH EVENTS IN THE NORTH ATLAN TIC Introduction: Lake Tulane and Heinrich Events Lake Tulane is a relatively small (~36 ha), deep (z max ~25 m) solution lake located in south central Florida, USA. Its depth and location on a structural high, the Lake Wales Ridge, enabled continuous sedi ment accumulation since before the last glacial maximum. Palynological analysis on a sediment core from Lake Tulane indicated major shifts in plant communities over the past 60,000 years (Grimm et al., 1993). Of particular note are six peaks in Pinus (pine ) pollen relative abundance, which coincide with the most intense cold phases of high latitude Dansgaard Oeschger (D O) cycles and the Heinrich Events (HE) that terminate them. Alternating with Pinus peaks are zones with high relative percentages of Quercu s (oak), Ambrosia (ragweed), Lyonia (staggerbush) and Ceratiola (rosemary) pollen, genera that today occupy the most xeric sites on the Florida landscape (Grimm et al., 2006). Additionally, the Quercus zones are replete with seeds from emergent aquatic pla nts, whereas all but one of the Pinus zones is devoid of such macrofossils. Lack of emergent macrofossils in Pinus phases suggests that the lake was too deep to support emergent vegetation close to the core site. Based on the quantitative similarity of the Pleistocene Pinus zones with modern/Holocene Florida vegetation, the Pinus peaks, and therefore the HE, were interpreted as warm and wet periods in Florida, whereas the Quercus zones were inferred to have been drier, and likely colder than the Pinus zones The assertion that the Pinus phases, and HE, represent warm and wet periods in the subtropics is contentious. HE are recorded in the sediment record of the North Atlantic as layers of ice rafted debris (IRD), lithic fragments from rocks of continental
69 or igin, which were derived from the calving and melting of large continental ice sheets episodic iceberg discharges perturbed global heat transport via reduction in the Atlantic meridional over turning circulation (AMOC) (Lynch Stieglitz et al., 2014). Although the mechanism responsible for these events is still debated (Hem ming, 2004; Marcott et al., 2011 ), teleconnections between HE and rapid climate oscillations have been recorded globally (Br oecker, 2006). Their occurrence at the end of progressively cooler Greenland D O interstadial cycles has led to the conclusion that HE were synchronous with lower temperatures and increased aridity over large regions of the Northern Hemisphere (Broecker et al., 1992; Zhao et al., 2003; Zhou et al., 2008). This hypothesis is supported by numerous lines of evidence, including: Iberian Margin sea surface temperature (SST) reconstructions derived from Mg/Ca ratios (Patton et al., 2011) and alkenone records (Paillier and Bard, 2001) the relative abundance of the polar planktonic foraminiferan Neogloboquadrina pachyderma (sinistral) in the western Mediterranean (Cacho et al., 1999), and 18 O records from marine sediment cores in the Iceland ic and Irminger Seas (van Kreveld et al., 2000). Although these studies indicated an in phase relationship between SST and climate conditions in Greenland, there appears to be an anti phase tendency in numerous other climate reconstructions, in particular, those from low to middle latitudes based temperature reconstructions from the southernmost portions of the IRD belt (40 55) indicate that surface waters were 2 4C warmer during all six Hein rich Events. Farther south in the western Atlantic, temperatures approached modern values during HE1 (Weldeab et al.,
70 2006; Carlson et al., 2008) and HE5a (Schmidt et al., 2006), implying warm surface waters prior to transitions into interstadials. These r esults are in line with late glacial circulation models, which produce warmer and more saline waters in middle latitudes of farther north (Renold et al., 2009). Reduced n orthward transport of heat just before and during HE, could be the source of the subtropical warming, hypothesized by Grimm et al. (2006) to be a prerequisite for Pinus proliferation. This idea was corroborated by coupled climate model simulations (Weaver et al., 2003 ), and pollen climate inference models specific to Florida (Donder s et al., 2011). Results from these modeling experiments demonstrate that a decrease in SST in the North Atlantic would increase the meridional temperature gradient, strengthen t he trade winds and expand the Atlantic Warm Pool (AWP), resulting in more precipitation across the Florida Peninsula during HE (Donders et al., 2011). Despite the results of modeling experiments, there are few empirical terrestrial records that document c hanges in temperature and precipitation across stadial/interstadial boundaries. In fact, the Grimm et al. (1993, 2006) studies of Lake Tulane represent the primary archive for terrestrial changes associated with HE in North America. Shortcomings of palynol ogical reconstructions, however, which include inter species differences in pollen production, dispersal, and preservation, make sedimentary pollen records potentially equivocal with respect to their ability to provide an accurate picture of plant communit y composition across the landscape (Bennet and Wi llis, 2002 ). Furthermore, pollen of different species within the genera Pinus and Quercus cannot be
71 distinguished, and today, taxa within each genus occupy different habitats across ure spectrum, and this complicates assignment of plant genera to wet or dry environments. Given the lack of terrestrial records associated with HE, and potential problems associated with palynological reconstructions, additional lines of evidence are neede d to test hypotheses regarding climate changes at low latitudes during HE. Carbon Isotopes and Precipitation Carbon isotopes of plant biomarkers in lake sediment cores have been used to infer shifts in precipitation and aridity over geologic timescales (e .g., Magill et al., 2013). Variations in the carbon isotopic composition of plant biomarkers, assuming no shifts in plant functional types or changes in the 13 C of atmospheric CO 2 are caused primarily by changes in water use efficiency (WUE = Assimilatio n/Transpiration) in plant communities as the y respond to shifts in mean annual precipitation (MAP). A positive leaf values (as defined below) was demonstrated using data from the literature for >3,000 leaf samples from nine biomes with different MAP (Diefendorf et al., 2010). In areas with higher MAP, there was greater 13 leaf ) by C 3 plants resulting in more negative plant 13 C values. The association between aridity (or decreased precipitation), and more enriched 13 C leaf values has been obser ved in many studies, with isotopic variability exceeding Danin, 2010). Measurements from single Pinus 13 C leaf values across a 400 mm grad ient of increasing precipitation in northeastern
72 (Waghorn et al., 2015). Taken together, these studies indicate that, once corrected for 13 C atmosphere and carbo leaf is primarily driven by water availability (Diefendorf et al., 2010). Hydrogen Isotopes and Precipitation weighted mean annual precipitation has been demonstrated by various investigations (Sachse et al., 2004, 2 012; Polissar and Freeman, 2010 ). Other studies have documented additional climatological (new precipitation sources and patterns), physiological (species related photosynthetic offsets), o r hydrological (precipitation ( lipid ) Chief among these effects, driven by relative humidity (Kahmen et al., 2008, 2013a, b) Studies of leaf wax hydrogen isotope variability measured across precipitation lipid values are significantly enriched by transpiration in more arid environments (Smith and Freeman, 2006; Douglas et al., 2012). A study of n alka nes from field grown barley ( Hordeum vulgare ) showed that relative humidity lipid values, suggesting that hydrogen isotopes in leaf waxes yield a record of precipitation that is strongly modified by leaf water evaporation (Sachse et al., 2010) isotopic fractionation in lipids in sediments from arid sites (Polissar and Freeman, 2010), and among n alkanes from plants across aridity gradients (Feakins and Sessions, 2010). leaf lipid values in lake sediments should provide an indicator of past precipitation and these tools can be used to evaluate changes in Florida climate during HE. For example, during the hypothesized wet Pinus leaf should be higher and
73 lipid should be relatively lower, whereas during the dry Quercus leaf should be lower lipid relatively higher. In this study we inferred temperature and precipitat ion changes across three HE (2 4 13 C and n alkanes extracted from the same sediment core from Lake Tulane that Grimm et al. (2006) used to develop their pollen record. I hypothesize that precipitation in Florida was relatively greater during HE because of low latitude warming associated wi th reduced AMOC strength. If, in fact, precipitation did change across stadial/interstadial boundaries, this change should be recorded in the carbon and hydrogen isotopes of plant leaf waxes. Study Site and Sample Collection: Lake Tulane Lake Tulane (Lati tude 27 surficial groundwater fed solution basin on the Lake Wales Ridge, an elongated, relict coastal ridge that has been above sea level since the early Pliocene (White, 1970). The sediments on the ridge are compo sed of coarse grained quartz sands and gravels, often bound together with clay s of fluvial origin (White, 1970). Lake Tulane is a relatively deep (z max ~25m), oligotrophic water body (total phosphorus = 6 g/L; Chl a = 3 g/L) that is hydrologically isolat ed from the deep, Eocene age limestone Floridan Aquifer, but receives seepage input through the shallow surficial aquifer. In spite of low aquatic productivity, the lake sediments are organic rich (>10% organic matter). Despite its location above a limesto ne platform, the lake deposits lack carbonate. A total of 24 sediment samples, four from each of Pinus zones 2 4 (TP 2 4) and Quercus zones 6 8 (TQ 6 8), were taken from the same core retrieved by Grimm et al. (2006) at 8 10 cm intervals. The core chrono logy is based on 55 AMS dates on bulk
74 sediments and, where available, terrestrial macrofossils. Our sampled interval (40,000 17,000 yr BP) fell within the range of radiocarbon dating. Grimm et al. (2006) calibrated s amples < 20,000 14 C yr BP using CALIB 5. 0.2 with the INTCAL04 calibration curve, and for samples dated to 20,000 40,000 14 C yr BP, the Fairbanks0805 calibration curve was applied (Fairbanks et al., 2005). See Grimm et al. (2006) for further details on calibration of the AMS dates. Lipid Extracti on, Purification, and Q uantification: Lake Tulane Sediment samples were freeze dried and lipids were extracted with an Accelerated Solvent Extractor ASE200 (Dionex), using 2:1 (v/v) dichloromethane(DCM):methanol through three extraction cycles at 10.3 MPa (1500 psi) and 100 C. Between 1 and 2 g of sediment were used for lipid extraction. Total lipid extracts (TLE) were concentrated under a gentle stream of nitrogen, and the neutral lipid fraction was obtained after base saponification of the TLE. Neutral li pids were further separated, based on polarity, into compound classes by column chromatography, using 5% deactivated silica gel, according to methods modified from Nichols (2011). Hydrocarbons were eluted from the silica gel column with 4.5 ml of 9:1 Hexa ne:DCM, and saturated hydrocarbons were separated from alkenes on 5% Ag impregnated silica gel (w/w) with 4 ml of hexane and ethyl acetate, respectively. A Thermo Scientific Trace 1310 gas chromatograph with a Supelco Equity 5 column, interfaced to a Ther mo Scientific TSQ 8000 triple quadrupole mass spectrometer with electron ionization, was used for compound identification. The split/splitless inlet was operated in splitless mode at 300 C with helium as the carrier gas. The column flow rate was 2.0 ml/min and the oven was held at an initial
75 temperature of 60 C for 1 minute, then ramped to 140 C at 15 C/min, and to 320 C at 4 C/min and held for 25 minutes. Alkane quantification was carried out on a Thermo Scientific Trace 1310 GC coupled with a flame ionize d detector (GC FID), using the same oven program as above. Androstane was added to sample vials prior to injection for use as an internal standard. Quantification was based on the calibration curves generated from the peak areas of external standards (C 7 C 40 ) with concentrations ranging from 5 to 250 g/ml Compound S pecific Isotope Measurements: Lake Tulane Compound specific stable carbon isotope ratios were measured at Pennsylvania State University using a gas chromatograph coupled to an isotope ratio mas s spectrometer interfaced with a GC C combustion system. The n alkanes were separated on a Varian model 3400 GC with a split/splitless injector operated in splitless mode. A fused silica capillary column (Agilent J& W DB 5; 30 m long, 0.32 mm I.D., 0.25 ml/min. The oven program began at a temperature of 60C, was held for 1 minute, then increased at a rate of 6C/min to 320C, whi ch was held for 20 minutes. Following GC separation, n alkanes were combusted over nickel platinum wire with O 2 in He (1%, v/v) at 1000C. Isotope ratios of carbon in CO 2 were measured using a Finnigan Mat 252 isotope ratio mass spectrometer. Isotopic abun dances were determined relative to a reference gas calibrated with Mix B ( n C 16 to n C 30 ; Arndt Schimmelmann, Indiana University). Carbon isotope values of samples were normalized to the VPDB scale using the Uncertainty Calculator 014) and are reported in standard delta notation as follows:
76 13 C= [(( 13 C / 12 C) sample /( 13 C / 12 C) standard ) 1] x 1000 (3 1) Standard errors of the mean (SE) were calculated using the Uncertainty Calculator, which yielded a 1 SE value of 0.35 (Polissar To eliminate possible confounding results associated with past changes in 13 C values of atmospheric CO 2 all carbon isotope data are presented as leaf values (Farquhar et al, 1989), where: leaf 13 C atm 13 C leaf 13 C leaf / 1000) (3 2) Carbon isotope values for atmospheric CO 2 were taken from Leuenberger et al. (1992) and data were interpolated using a cubic spline. Carbon isotope values of atmospheric CO 2 ranged from By accounting for leaf value primarily becomes a function of WUE, with leaf indicative of lower stomatal conductance and reduced precipitation (Diefendorf et al., 2010). Compound specific hydrogen isotopes were measured at the U niversity of Cincinnati using a Thermo Trace GC Ultra coupled to a Thermo Electron Delta V Advantage IRMS instrument with an Isolink combustion furnace. A fused silica capillary column (Agilent J&W DB used with helium as the carrier gas and a column flow rate of 1.5 ml/min using a split/splitless injector operated in splitless mode. The oven program was slightly modified from the schedule above: 80C (2 min), then to 320C (held 10 min) at 8C/min. Hyd rogen isotope abundances and standard errors were calculated as above for carbon isotopes. Reference gas values were calibrated with Mix A ( n C 16 to n C 30 ; Arndt Schimmelmann, Indiana University). The H 3 + factor had a mean value of 6.244. Calculation of is otope
77 (VSMOW) scale was done in the same way as for carbon isotopes, and the pooled 1 nota tion relative to VSMOW as follows: D /H) sample /(D/H) standard ) 1] x 1000 (3 3) To estimate the environmental controls on hydrogen isotope ratios in leaf waxes, l/w values, defined as [(( D /H) lipid /(D/H) water ) 1 ], for C 29 n l/w values for terrestrial plants in this study were taken from Magill et al. (2013), who defined three primary PFTs on the basis of photosynthetic pathway and growth habit: C 4 graminoi 3 3 l/w values were taken from Aichner et al. (2010), who calculated the biosynthetic offset between the C 23 n alkane from Potamogeton (pondweed) and lake water to be l/w evapotranspiration (and aridity) lipid values of terrestrial and aquatic plants as follows: (terr aq) lipid (C 29 lipid (C 23 ) + 1000) 1)] (3 4) n Alkane Concentrations and Chain Length Distributions Concentrations for select n alkane ch ain lengths are provided in Figure 4 1 The n a lkane chain lengths are dominated by the long chain alkanes (i.e., > n C 25 ), with n C 29 display Tulane sediment record. We calculated the average chain length (ACL) for all samples using the Eglinton and Hamilton (1967) equation:
78 ACL(C 25 C 35 ) = (3 5) There is low variability in ACL throughout the record. The highest ACL value is 30.1 and the lowest is 29.2, with a mean ACL of 29.7 for the entire record (Table 4 1). All samples exhibit a strong odd over even predominance, consistent with vascular terrestrial plant sources. This is also in agreement with carbon preference index (CPI) values calculated from the equation of Bray and Evans (1961): CPI = 0.5 + (3 6) The n alkane CPI values for all samples are >1, which indicates higher odd chain n alkane abundances. CPI values range from a maximum of 3.2 to a minimum of 2.0, with an average of 2.7 (Table 4 1). Differences in mean n alkane concentrations are apparent between Pinus and Quercus phases. Highest average abundances for carbon chain lengths n C 16 C 22 occurred during Pinus phases, whereas cha in lengths n C 23 C 35 had the highest average abundances during Quercus phases. Alkanes derived from C 3 dicots ( n C 27 and n C 29 ) decreased in average abundance by 75% ( n C 27 ) and 70% ( n C 29 ) during the Pinus intervals. Furthermore, alkane biomarkers from mi xed C 3 /C 4 sources, n C 33 and n C 35 were reduced by 64% and 62%, respectively, during the Pinus intervals. Although abundances of the submerged aquatic plant biomarker ( n C 23 ) were higher during Quercus phases by ~60%, the P aq proxy (Ficken et al, 2000), d escribed by the formula (C 23 + C 25 )/(C 23 + C 25 + C 29 + C 31 ), showed no significant change in the ratio of submerged to emergent/terrestrial vegetation across vegetation zones (Table 4 1).
79 Carbon Isotope Results: Lake Tulane leaf values for carbon chai n lengths n C 27 C 35 are shown in Figure 4 2 leaf values and Pinus pollen percentages are greater for shorter chain length alkanes n C 27 (r = 0.62, n = 24, p < 0.001) and n C 29 (r = 0.54, n = 24, p < 0.01), than for longer chain leng ths n C 31 (r = 0.44, n = 24, p < 0.05) and n C 33 (r = 0.40, n = 24, p < 0.05), however, n C 35 (r = 0.58, n = 24, p < 0.01) is also more strongly correlated to Pinus pollen than the other long chain n leaf values for all Pinus zones were an a Quercus zones and all leaf values were significantly positively correlated with Pinus pollen leaf values for all chain lengths are provided in Table 4 2. The n alkanes with 31, 33, and leaf leaf values of n alkanes with Hydrogen Isotope Results: Lake Tulane Hydrogen isotope profiles for select n alkanes are displayed in Figure 4 3 and Table 4 r during Pinus phases and lower during the Quercus zones. The greatest difference between means was recorded in the n C 23 means ( Pinus Quercus begin to decrease prior to the end of each Pinus chain length ( n C 23 ) are not significantly correlated with Pinus pollen percentages (Figure 4 3 ). As with carbon isotope results, the longer chain length alkanes ( n C 33 =
80 n C 35 ater D ranges compared to shorter chain length alkanes ( n C 27 n C 29 found in alkane n C 23 After adjusting the n C 23 l/w n C 29 l/w account for fr (terr aq) values were calculated as a proxy for evapotranspiration. Larger values are indicative of greater aridity and (terr aq) is negatively correlated with percent Pinus pollen (r = 0.79, n = 21, p (terr aq) Quercus phases (Figure 4 4 (terr aq) Reconstructing Paleohydrology from Leaf Wax Carbon and Hydrology Isotopes The late Pl eistocene was characterized by periods of rapid climate fluctuations brought about by changes in ocean circulation and ice volume. These climate events have been studied extensively in sediment records from marine sites in the high latitude north Atlantic, but rarely have they been documented in subtropical terrestrial sites. C arbon and hydrogen isotope leaf wax data from this study were used to reconstruct changes in paleohydrology at Lake Tulane, Florida (USA) and supplement the limited data on the effect s of rapid climate change at low latitudes. Both isotopes respond to changes in water availability, therefore vegetation transitions from Pinus dominated to Quercus dominated phases should have produce d measurable differences in carbon and hydrogen stable isotope ratios. If variations in carbon source (i.e. 13 C atms ) leaf value is primarily a function of WUE with low leaf indicative of lower stomatal conductance and thus low precipitation leaf values are small and their
81 correlation w ith Pinus percentages are all below r = 0.6, changes in water availability are still apparent across the studied section of the Lake Tulane sediment record. The n C 29 leaf profile exhibits the strongest positive correlation (r = 0.54) with changes in Pinu s pollen relative abundance. This n alkane, a primary component of leaf waxes in higher terrestrial angiosperms, was mainly derived from Quercus with minor contributions from Ambrosia These are the two dominant angiosperms in the Lake Tulane pollen reco rd, accounting for ~50% of all pollen recorded during interstadials, and 10 25% during stadials. Despite the high Pinus pollen during stadials, very few n alkanes were contributed by gymnosperms, as they are noted for their low alkane production (Diefendor f et al., 2011). n C 29 leaf values between Pinus and Quercus phases 13 with a single taxon, during which Pinus halepensis received different amounts of water (Ferrio et al., 2003), and with across rainfall gradients in the Mediterranean (Hartmann and Danin, 2010). I infer from leaf values during Pinus phases that WUE decreased and discrimina tion against 13 C increased. The interpretation of these data relies heavily on modern empirical studies of WUE in C 3 plants ( e.g. Hartmann and Danin, 2010). Plants respond to water stress by reducing their stomatal conductance, which lowers the CO 2 concent ration within leaves (internal leaf = c i ) relative to the concentration of CO 2 in the atmosphere (c a ), i.e. c i /c a Partial closure of the stomata reduces the flux of CO 2 and H 2 O, but the two fluxes are not reduced equally. Whereas the reduction in transpir ation is proportional to the amount of stomatal closure, photosynthesis continues at a high rate,
82 and carbon assimilation thus declines less than transpiration. Because assimilation decreases slower than transpiration, WUE increases. Related to this, when c i is low plants are less able to discriminate against 13 13 C leaf values increase and leaf values decrease (Marshall et al., 2007). From this I interpret Pinus zones (cold stadials) to have been wetter than Quercus zones (warm interstadials). Although WUE is 13 C values in plants it is not the only factor that can affect leaf wax carbon isotope signature s A small amount of variability in the carbon isotope record could be the result of mixing of n alkanes from different PFTs. Previous isotope and pollen research on the same sedi ment core from Lake Tulane showed that C 4 contributions to the organic carbon pool dropped from highs of ~50% during Quercus phases to nearly 0% during Pinus phases ( Grimm et al., 2006; Huang et al., 2006). These results are in agreement with our n alkane concentration data, which record a ~60% drop in n C 33 and n C 35 concentrations, but not in n C 27 and n C 29 concentrations in Pinus zones. The disparate behavior of n C 27/29 and n C 33/35 in the record is likely related to plant physiology: C 4 graminoids con tain n C 33 and n C 35 alkanes in their leaves at concentrations an order of magnitude greater than in woody C 3 angiosperms (Garci n et al., 2014; Diefendorf and Freimuth 2017 ). This bias towards longer chain (greater than C 31 ) n alkane synthesis in graminoi ds was also observed in a marine sediment core taken off the coast of equatorial Africa (Wang et al., 2013). The authors applied results from empirical isotope studies (e.g. Sasche et al., 2012) to their data to estimate the relative input of C 3 /C 4 vegetat ion to their site over the past 37 k a They found that n C 33 35 were mostly contributed by C 4 grasses that are depleted in deuterium and enriched in 13 C relative to the same n alkane homologues from C 3 dicots.
83 The low leaf n C 33 35 leaf record from this study suggests that these n alkanes are primarily from C 4 sources. From the n alkane concentration and isotope data we infer: 1) a negligible contribution of n alkanes from C 4 grasses during HE, and 2) a limited contribution from grasses throughout the entire n C 29 record. The n C 29 data can be further classified as a terrestrial C 3 angiosperm biomark er based upon research on chain length distributions and isotop e values among different PFTs, which show that most gymnosperms do not produce n alkanes in substantial quantities (Diefendorf et al., 2011). Mixing of C 3 /C 4 vegetation and variable input of conifer organic matter would not substantially 13 C values of the n C 27/29 record, nor would the introduction of new plant species as illustrated by Pollen counts showed that relative contributions of plant pollen changed significantly b etween stadial and interstadial periods, but relative pollen abundances were quite similar among all Pinus zones, and among all Quercus zones. Of the two major plant genera at the site, only Quercus produces n C 27/29 in significant quantities, and palynolo gical studies suggest that the most probable source of these homologues was from one or more Quercus species in Florida (Grimm et al., 2006). Therefore, changes in 13 C values, unrelated to WUE, could be from physiological differences in carbon isotope fractionation among plants of the same functional type. Isotopic differences have been detected in n alkanes sampled from angiosperms growing under identical environm ental conditions and irrigated with the same water (Pedentchouck et al., 2008). Leaves measured from two species in that study, Betula pendula and Populus tremuloides 13
84 the authors to co nclude that different rates of stomatal conductance and carbon metabolism lead to isotopic differences among angiosperms (Pedentchouck et al., 2008). Furthermore, carbon isotope ratios measured in n C 29 alkanes from three species of Quercus returned values 13 C measurements on individual Q. chrysolepis from a single study site ranged from n C 27/29 leaf record responds to both chan ges in water availability, and to a lesser extent, to mixtures of n alkanes from plants within the same genus and/or functional type. This finding does not preclude our data from being used to reconstruct water availability. Th e relatively strong correlat ion between Pinus pollen relative abundance and n C 27/29 indicate s that water availability probably change d as the climate system in the northern hemisphere responded to HE. Because n C 27/29 were least influenced by changes in relative contributions from d n C 27/29 can be used as a lipid values of these chain lengths (and all other long chain n alkanes analyzed) decrease during the interstadial periods ( Quercus zo nes ) and reac h maximum values during stadial ( Pinus zones ) (Figure 4 3 values of leaf When relative humidity is higher, both precip itation and lower. Assuming this n alkanes suggest greater aridity during the Pinus zones. The values recorded in the n alkanes of terrestrial vegetation however, does not solely track th e effects of aridity, but rather reflects both aridity and variability in the mid latitudes, hydrogen isotope ratios in precipitation are
85 primarily controlled by temperature (Gat, 1996; Rach et al., 2014). Similar to the effect values of precipitation. likely shifted toward more temperate i.e. cooler conditions As such, the is otopic composition of precipitation was controlled more by temperature and less so by the amount effect. To separate the effects of aridity from condensation temperature on leaf l/w values to long chain terrestrial and short cha in aquatic n alkanes, and expressed the effect of aridity, or more specifically, evapotranspiration, (terr aq) ). Even though we cannot exclude the possib i lity of multiple woody C 3 angio sperm sources in our n C 27 29 record, we are confident these homologues represent negligible input from gymnosperms and C 4 grasses. Because n C 27 is less abundant throughout our record, we assign ed l/w value of n C 29 using Magill e (2013) calculation for woody C 3 l/w for submerged aquatic vegetation (SAV) is more challenging because few studies have addressed isotopic offsets betwee n plants and lake water. We do know however, that submerged vegetation co ntains a greater relativ e abundance of mid chain length, primarily n C 23 alkanes (Ficken et al., 2000), and in a few studies, the offset between n (2008) measured hydrogen isotope values of n C 23 in a suite of aquatic macrophytes from lakes on the l/w values with a mean of lower relative to measurements on the submerged macrophyte Potamogeton l/w = mid
86 chain length n C 23 are higher chain n alkanes, terrestrial isotope record. The greater values measured in n C 23 can be explained by l/w value, a value which is Potamogeton measurements. This explanation is supported by Potamogeton macrofossil evidence within the Lake Tulane sediment core (Grimm et al., 2006), thus, l/w of n C 23 alkane. l/w the lake water n C 23 decrease by an average of Quercus periods. The simplest explanation for this decrease in hydrogen isotope ratios invokes decreasing air temperature in the tropics durin g Greenland by evaporative enrichment because of the lakes relatively great depth and the moderating effect of shallow groundwater inputs. This is supported by data from modern lake water studies that tracked isotopic shifts in Lake Tulane over a three year period (Escobar et al., 20 13). During that time span, the average intra 18 O is equivalent that annual evaporative enrichment of lake waters was similar during the Pleistocene and had minim al influence on the isotope values of lake water. If, however, evaporative enrichment did increase during Quercus n C 23 from SAV record during Quercus zones (Figure 4 3 ). This, of course, could be interpreted as a
87 decrease in aridity and an increase in precipitation during interstadial periods, but an interpretation of wetter conditions is at odds with: 1) our carbon isotope data, from which we infer gr eater WUE during Quercus periods, 2) the proliferation of C 4 grasslands during the Lake Tulane Quercus phases, which is suggestive of increased aridity (Huang et al., 2006), and 3) climate models and SST reconstructions that indicate ocean warming and rela tively greater precipitation in the Gulf of Mexico region during Pinus as opposed to Quercus zones (Flower et al., 2004; Schmidt et al., 2006; Donders et al., 2011). From this, we conclude that the isotopic variability in the n C 23 record reflects shifts in condensation and evaporation temperatures during cloud formation. During both evaporation of water from the source (ocean surface) and condensation of decreasing tempe n C 23, once corrected for the biological fractionation is interpreted as a signal of temperature during precipitation. values of n C 23 during Pinus periods are the result of higher temperatures, values during Quercus periods are from lower atmospheric temperatures. Hydrogen isotope values of the terrestrial n alkane ( n C 29 ) biomarker track precipitation, but are also influenced by evapotranspiration and mixing of plant communit l/w values (Smith and Freeman, 2006). When analyzed l/w corrected n C 29 record does not produce results that lead to meaningful paleo hydrological interpretation. The ~5 during Pinus phases could be attributed to either climate variables, such as increased aridity and higher temperatures, or paleo hydrological factors, such as shifts in the
88 source and timing of precipitation. When the n C 29 data are analyzed in conjunction with l/w corr ected n C 23 record, however, (terr aq) value, changes in aridity between Pinus and Quercus zones can be quantified. On average, the values for Pinus pollen concentrations. These results are in line with model simulations that produce deuterium enrichment in n alkanes by as much as 10 (terr aq) and Pinus zo nes (r = 0.79, p<0.001) (Figure 4 4 ) identifies these as zones of lower isotopic enrichment from evapotranspiration, and hence less aridity, which is defined as the ratio of annual precipitation to potential evapotranspiration (Maliva and Missimer, 2012). If bot h aridity and evapotranspiration decreased during Pinus zones, then precipitation in Florida must have increased when Pinus proliferated. Implications of Rapid Climate Change Events in the Subtropics Carbon and hydrogen isotope data from n C 29 and n C 23 in dicate that evapotranspiration decreased and atmospheric temperatures increased during Pinus phases at Lake Tulane and HE/Greenland stadials. The independent chronology developed by Grimm et al. (2006) also found an inverse correlation between Greenland te mperatures and the Lake Tulane pollen record. Multiple reconstructions of SST during HE indicate that surface waters south of the IRD belt warmed as temperatures over Greenland dropped (Flower et al., 2004; Schmidt et al., 2006; Weldeab et al., 2006 ), and temperature in the Gulf of Mexico also remained warm throughout Greenland stadials, implying persistence of the AWP around peninsular Florida (Carlson et al., 2008; Ziegler et al., 2008; Rasmussen and Thomsen, 2012). But what could have kept Florida an d the Gulf of Mexico warm and wet during the extreme cold events in the
89 North Atlantic? Results of pollen climate inference models (Donders et al., 2011) link a warm, persistent AWP with northward displacement of the ITCZ and increased precipitation across the Florida Peninsula during cold periods in the high latitude North Atlantic. At present, the source and seasonal delivery of precipitation in Florida is similar to the late Pleistocene. The northward displacement of the ITCZ during the boreal summer inc reases trade wind strength over the subtropical North Atlantic, resulting in greater rainfall on the peninsula. Other studies showed that during Greenland stadials the southward displacement of the polar front wa s associated with an expansion of the subpo lar gyre and a westward movement of the subtropical gyre (Eynaud et al., 2009). The westward shift of the subtropical gyre results in a narrower Gulf Stream and mov es it closer to the continent (Hoogakker et al., 2013). This agrees with inferred stadial/i nterstadial shifts in the position of the Gulf Stream (Hoogakker et al., 2013), and warmer mid latitude North Atlantic SST s during stadial periods (Naafs et al., 2013). A narrower Gulf Stream with reduced northward heat transport located closer to the Flo rida Peninsula could explain why Florida remain ed warm and relative ly wet during the Heinrich stadials D ata from multiple studies (e.g. Kreveld et al., 2000) indicate that the duration and magnitude of the climate system response to HE6 1 was not unifor m. Our (terr aq) record spans a range of 36 and records multiple abrupt shifts in aridity. During HE4 3 departures towards negative (terr aq) values indicate that these HE elicited the strongest response from the hydrological cycle, while HE2 produced a more muted response. It is not surprising that HE4 resulted in the largest shift toward less arid values as it is the largest of the six HE in terms of %IRD (Hemming 2004, Tierney et al., 2008). The
90 subdued response in our hydrogen isotope record during HE2, is also expected, as studies have demonstrated that there was no reduction in AMOC intensity across that interval (Lynch Stieflitz et al., 2014; Parker et al., 2015). T he pronounced response during HE3 however, does not match data that classify it as a low foraminifera interval, with limited change in AMOC strength, rather than a true ice rafting event (Hemming 2004; Lynch Stieflitz et al., 2014). Tulane Pinus zone 3, which correlates to HE3, has lower values of Pinus pollen and higher values of Ambrosia tha n Pinus zones 1, 4, 5, and 6, suggesting cooler temperatures during HE3 (Grimm et al., 2006). If HE3 was not caused by a reduction in AMOC strength associated with freshening of the North Atlantic, another yet unidentified mechanism is needed to explain th e hydrological changes during HE3. It is possible that ITCZ position and attendant atmospheric teleconnections remained similar during all six HE, even if the AMOC strength and sedimentary IRD percentages varied. We converted our values to 18 O space a nd applied the range of values to the Dansgaard (1964) temperature equation. Results showed a positive maximum assuming all isotope variability for this chain length was temperatur e driven. Evaporation is a secondary mechanism that could contribute to the enrichment in of n C 23 If the effect of modern evaporative enrichment (a value of ~5 ) were deducted from our calculation, the maximum temperature shift would reduce to a 6 during stadials. The range in calculated temperatures from the warm stadials to the cool interstadials in Florida is comparable to the estimated 6 the cold early deglacial and warmer Holocene in the lowla nds of northern Guatemala
91 (Hodell et al., 2012). A TEX 86 derived surface air temperature record from Lake Tanganyika demonstrates a similar magnitude of warming (~5 YD (Tierney et al., 2008). The fact that both temperature and preci pitation at Lake Tulane responded in concert with rapid climatic variability in the high latitude Northern Hemisphere demonstrates that glacial calving events have a strong influence on climate in the subtropics. Although these events produce cooling throu ghout much of the Northern Hemisphere, it is apparent that this was not the case in peninsular Florida. The antiphase relationship underscores the complex teleconnections linking atmospheric temperatures with oceanic SSTs and circulation patterns. The redu ction in AMOC strength limited the amount of thermal energy that could be redistributed to the poles, but resulted in the persistence of the AWP and intensification of ocean currents and trade winds in the low latitude North Atlantic that increased atmosph eric temperatures and precipitation. Atmospheric warming is seen in other terrestrial records ( e.g. Tierney et al., 2008) during HE, h owever warming never coincides with an increase in precipitation. Thus, the response to rapid climate change events is reg ional and, therefore, highly variable. The climate system s reaction depends on myriad factors, including, but not limited to: SST gradients, latent heat fluxes, and circulation patterns that influence regional atmospheric circulation patterns. Conclusions : Lake Tulane I presented a record of environmental change recorded in the sediments of Lake Tulane, Florida, w hich were deposited across HE 2 4 leaf ) measurements from leaf waxes show the combined effects of WUE and mixing of PFTs. Correlations between carbon isotopes and Pinus pollen concentrations increase with
92 decreasing carbon chain length, indicating less mixing from multip le plant groups in shorter chain length n lipid values from these chain lengths were lipid values to estimate aridity changes across HE and intervening (terr aq) values correlate strongly w ith Pinus pollen relative leaf values from n C 27 29 signify lower WUE and less aridity lipid values record an ~8.5 Florida during HE. Taken together, our results demonstrat e that the subtropics responded out of phase with high latitude environmental changes during at least three of the late Pleistocene stadial interstadial transitions. The region remained warm and wet because of oceanic/atmospheric patterns that limited nort hward export of warm, saline water.
93 Figure 4 1 Concentrations of select n alkane chain lengths. The vertical orange bars mark the extent of the Pinus zones, and both the Pinus (TP) and Quercus (TQ) zones are labeled at the top of the figure in blue, a nd red, respectively.
94 Figure 4 2 Select n alkane chain lengths and their leaf values (dashed lines) plotted with the percent abundance of Pinus pollen (solid line). Tulane Pinus zones are represented with the orange bars. All data points have a 1 SE value of 0.35
95 Figure 4 3 Select n alkane chain lengths and their values (dashed lines) plotted with the percent abundance of Pinus pollen (solid line). Tulane Pinus zones are represented with the orange bars. T he pooled 1 SE for all samp les was
96 Figure 4 4 The (terr aq) values (dashed line) plotted with the percent abundance of Pinus pollen (solid line). Tulane Pinus zones are represented with the orange bars. The (terr aq) values are essentially an approximation of ar idity, with higher values indicating more arid conditions. T he pooled 1 SE for all samples was
97 Table 4 1 Values for the carbon preference index (CPI), average chain length (ACL), and submerged to emergent/terrestrial vegetation in the Lake Tulane core. The ages are calibrated following the methods outlined in Grimm et al. (2006), and the phases are Tulane Quercus (TQ) and Tulane Pinus (TP), which are also described in the Grimm et al. (2006) paper. Age (cal. Yrs BP) CPI ACL25 35 P a q Phase 17038 3.21 29.64 0.17 TQ6 17345 2.97 29.54 0.18 TQ6 17655 0.84 29.94 0.16 TQ6 17965 2.04 29.73 0.17 TQ6 23522 2.75 29.23 0.21 TP2 23706 2.77 29.23 0.26 TP2 24447 2.81 29.69 0.19 TP2 25519 2.67 29.44 0.21 TP2 27716 2.63 29.48 0.24 TQ7 2792 5 2.90 29.63 0.22 TQ7 28030 2.89 29.64 0.21 TQ7 28134 2.76 29.68 0.22 TQ7 29693 3.09 29.55 0.18 TP3 30075 2.85 29.36 0.21 TP3 30457 2.93 29.31 0.24 TP3 31804 3.04 29.51 0.21 TP3 35511 2.27 30.00 0.19 TQ8 36068 2.35 30.04 0.15 TQ8 36207 2.33 29.93 0.15 TQ8 36347 2.71 29.84 0.14 TQ8 37180 2.38 30.13 0.11 TP4 38014 2.62 29.50 0.20 TP4 39386 2.66 29.96 0.16 TP4 39923 3.11 29.82 0.18 TP4
98 Table 4 2 leaf values for select n alkane chain lengths extracted from the Lake Tulane core. Age (cal yr BP) C 23 leaf C 27 leaf C 29 leaf C 31 leaf C 33 leaf C 35 leaf Phase 17,038 23.73 22.11 21.38 20.70 23.00 19.10 TQ6 17,345 21.62 20.69 20.02 19.20 19.86 16.64 TQ6 17,655 23.85 21.65 20.70 19.52 19.38 17.17 TQ6 17,965 23.66 21.22 20.37 19.66 18.77 17.36 TQ6 23,522 n/a 22.34 21.01 20.16 20.26 20.15 TP2 23,706 22.53 21.07 20.42 19.46 18.64 n/a TP2 24,447 22.57 22.19 21.01 20.10 19.45 18.88 TP2 25,519 22.39 22.05 20.83 20.24 19.56 18.68 TP2 27,716 21.19 21.46 20.74 20. 08 21.41 18.99 TQ7 27,925 21.94 20.85 19.93 19.51 19.70 18.48 TQ7 28,030 22.54 21.44 20.50 19.75 20.37 18.51 TQ7 28,134 21.59 21.20 20.86 20.12 20.59 17.92 TQ7 29,693 n/a 21.56 20.76 19.73 21.57 n/a TP3 30,075 24.55 22.41 21.31 21.44 21.40 n /a TP3 30,457 24.42 21.93 20.85 20.02 22.02 n/a TP3 31,804 23.01 20.87 20.21 18.77 20.61 n/a TP3 35,511 21.00 21.29 20.58 20.20 19.55 17.84 TQ8 36,068 21.27 21.42 20.96 20.33 21.39 19.57 TQ8 36,207 20.96 21.20 20.63 19.89 20.80 19.47 TQ8 36,347 21.33 21.24 20.70 19.83 20.47 19.19 TQ8 37,180 n/a 22.39 20.96 20.33 24.30 20.61 TP4 38,014 21.94 21.73 21.22 21.18 22.56 18.67 TP4 39,386 22.80 22.24 21.90 21.43 23.82 21.30 TP4 39,923 21.82 21.76 21.60 20.60 22.59 19.88 TP4
99 Table 4 3 values for select n alkane chain lengths extracted from the Lake Tulane core. Core phases are as follows : Tulane Pinus (TP), and Tulane Quercus (TQ). Age (cal yr BP) C 23 C 27 C 29 C 31 C 33 C 35 Phase 17,038 n/a 137.80 142.28 151.42 141.14 144.74 TQ6 17,345 127.91 123.30 140.89 152.81 148.98 149.05 TQ6 17,655 n/a 143.76 151.31 159.74 156.16 146.41 TQ6 17,965 148.09 138.05 149.16 157.94 154.56 156.54 TQ6 23,522 n/a n/a 144.21 n/a n/a n/a TP2 23,706 114.94 124.69 135.69 153.80 137.55 140.93 TP2 24,447 123.48 132.01 145.74 155.02 157.02 155.75 TP2 25,519 136.61 133.97 151.40 157.44 149.56 148.89 TP2 27,716 149.54 138.46 151.62 161.61 163.89 162.25 TQ7 27,925 146 .19 132.48 148.22 158.86 161.96 163.15 TQ7 28,030 140.46 134.38 148.25 158.89 158.29 160.51 TQ7 28,134 144.15 136.46 149.27 156.72 158.06 158.40 TQ7 29,693 116.02 136.45 145.11 153.21 149.52 156.03 TP3 30,075 120.97 130.33 141 .67 151.62 143.82 138.71 TP3 30,457 129.80 138.57 145.44 151.97 146.13 149.46 TP3 31,804 129.09 139.11 145.03 157.99 155.74 143.82 TP3 35,511 158.47 135.15 156.69 169.64 170.28 164.54 TQ8 36,068 149.61 132.87 155.91 170.89 169 .13 n/a TQ8 36,207 153.24 140.69 155.34 166.87 168.35 158.48 TQ8 36,347 143.48 143.49 151.53 168.93 167.26 156.18 TQ8 37,180 124.40 131.83 152.92 159.72 161.84 156.31 TP4 38,014 123.74 135.23 143.84 155.64 148.83 150.23 TP4 39, 386 131.76 137.69 147.86 162.96 165.22 160.37 TP4 39,923 129.70 135.56 144.90 161.02 156.71 150.97 TP4
100 CHAPTER 5 CONCLUSIONS Based upon analyses of bulk geochemical variables and alkyl lipid concentrations I found that both Lake s Wauberg and Apopka became increasingly more productive and derived more of their orga nic matter from algal sources immediately after anthropogenic influences disturbed their watersheds. In Lake Apopka this occurred in the mid 1940s after nearby wetlands were con verted to muck farms and in Lake Wauberg algal abundances increased in the 1980s after major residential development began around the lake Lake Sheelar has remained a low productivity system since at least the turn of the 20 th century, and algae contri bute only minor amounts of organic matter to the sediments The evolution of Lake Harris the focus of my second study, is recorded in the geochemical variables preserved within its sediments. The lake evolved from a shallow marsh like system in the early Holocene ( ~10,000 yrs BP ), when many lakes in Florida began to fill with water in response to deglaciation, to a shallow lake in the middle Holocene (7,000 5,000 yrs BP) as Florida transitioned from dry oak scrub vegetation to pine forests Around 3,000 y rs BP, the effects of ENSO intensified, and Lake Harris achieved its modern limnetic state. In the modern era (after ca. AD 1950), c oncentrations of algal and cyanobacteria biomarkers increase d by orders of magnitude in response to human induced eutrophica tion. Temporal patterns of c arbon and hydrogen isotope signatures of n alk a nes extracted from Pleistocene age sediments in Lake Tulane show correlations to Heinrich Events in the North Atlantic. During periods of cold sea surface temperatures in the North Atlantic, hydrogen isotope ratios indicate that Florida remained relatively warm.
101 Combined hydrogen and carbon isotope values from multiple n alkane chain lengths demonstrate that aridity increased during warm phases in the North Atlantic and decreased du ring cold phases, i.e. Heinrich Events. In paper 3, I conclude that Florida was likely buffered by warm sea surface temperatures that resulted from the reduced flow of the Atlantic meridional overturning circulation
102 LIST OF REFERENCES Aichner, B., Herz alkanes in Tibetan lake sediments and aquatic macrophytes A surface sediment study and application to a 16ka record from Lake Koucha. Organic Geochemistry 41 (8), 779 790. Aller, R. C. (1994). Bioturbation and remineralization of sedimentary organic matter: effects of redox oscillation. Chemical Geology 114 (3 4), 331 345. Anderson, D. M., Glibert, P. M., & Burkholder, J. M. (2002). Harmful algal blooms and eutrophication: nutrient sour ces, composition, and consequences. Estuaries 25 (4), 704 726. Appleby, P. G., & Oldfield, F. (1983). The assessment of 210Pb data from sites with varying sediment accumulation rates. In Paleolimnology (pp. 29 35). Springer Netherlands. Bennett, K. D., & Willis, K. J. (2002). Pollen. In Tracking environmental change using lake sediments (pp. 5 32). Springer Netherlands. Bianchi, T. S., & Canuel, E. A. (2011). Chemical biomarkers in aquatic ecosystems Princeton University Press. Bianchi, T. S., Mitra, S ., & McKee, B. A. (2002). Sources of terrestrially derived organic carbon in lower Mississippi River and Louisiana shelf sediments: implications for differential sedimentation and transport at the coastal margin. Marine Chemistry 77 (2), 211 223. Binford, M. W. (1990). Calculation and uncertainty analysis of 210Pb dates for PIRLA project lake sediment cores. Journal of Paleolimnology 3 (3), 253 267. Bond, G., Broecker, W., Johnsen, S., McManus, J., Labeyrie, L., Jouzel, J., & Bonani, G. (1993). Correlatio ns between climate records from North Atlantic sediments and Greenland ice. Nature 365 (6442), 143 147. Bray, E. E., & Evans, E. D. (1961). Distribution of n paraffins as a clue to recognition of source beds. Geochimica et Cosmochimica Acta 22 (1), 2 15. Brenner, M., Hodell, D. A., Leyden, B. W., Curtis, J. H., Kenney, W. F., Gu, B., & Newman, J. M. (2006). Mechanisms for organic matter and phosphorus burial in sedimentsof a shallow, subtropical, macrophyte dominated lake. Journal of Paleolimnology 35 (1) 129 148. Brenner, M., Whitmore, T. J., Curtis, J. H., Hodell, D. A., & Schelske, C. L. (1999). 13 15 N) signatures of sedimented organic matter as indicators of historic lake trophic state. Journal of Paleolimnology, 22(2), 205 221.
103 Broecker, W. S. (2006 ). Abrupt climate change revisited. Global and Planetary Change 54 (3), 211 215. Broecker, W., Bond, G., Klas, M., Clark, E., & McManus, J. (1992). Origin of the northern Atlantic's Heinrich events. Climate Dynamics 6 (3 4), 265 273. Cacho, I., Grimalt, J. O., Pelejero, C., Canals, M., Sierro, F. J., Flores, J. A., & Shackleton, N. (1999). Dansgaard Oeschger and Heinrich event imprints in Alboran Sea paleotemperatures. Paleoceanography 14 (6), 698 705. Carlson, A. E., Oppo, D. W., Came, R. E., LeGrande, A. N., Keigwin, L. D., & Curry, W. B. (2008). Subtropical Atlantic salinity variability and Atlan tic meridional circulation during the last deglaciation. Geology 36 (12), 991 994. Carr, A.F. (1934). The plankton and carbon dioxide oxygen cycle in Lake Wauberg, Castaeda, I. S., & Schouten S. (2011). A review of molecular organic proxies for examining modern and ancient lacustrine environments. Quaternary Science Reviews 30 (21), 2851 2891. Chikaraishi, Y., & Naraoka, H. (2003). Compound alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63 (3), 361 371. Cloern, J. E., Canuel, E. A., & Harris, D. (2002). Stable carbon and nitrogen isotope c omposition of aquatic and terrestrial plants of the San Francisco Bay estuarine system. Limnology and oceanography 47 (3), 713 729. Correll, D. L. (1998). The role of phosphorus in the eutrophication of receiving waters: A review. Journal of Environmental Quality 27 (2), 261 266. Craft, C. B., & Richardson, C. J. (1997). Relationships between soil nutrients and plant species composition in Everglades peatlands. Journal of Environmental Quality 26 (1), 224 232. Craft, C. B., Vymazal, J., & Richardson, C. J. (1995). Response of Everglades plant communities to nitrogen and phosphorus additions. Wetlands 15 (3), 258 271. Cranwell, P. A. (1982). Lipids of aq uatic sediments and sedimenting particulates. Progress in lipid research 21 (4), 271 308. Cranwell, P. A., Eglinton, G., & Robinson, N. (1987). Lipids of aquatic organisms as potential contributors to lacustrine sediments II. Organic Geochemistry 11 (6), 513 527.
104 Dansgaard, W. (1964). Stable isotopes in precipitation. Tellus 16 (4), 436 468. Dean, W. E., & Gorham, E. (1998). Magnitude and significance of carbon burial in lakes, reservoirs, and peatlands. Geology 26 (6), 535 538. Diefendorf, A. F., Freeman, K. H., Wing, S. L., & Graham, H. V. (2011). Production of n alkyl lipids in living plants and impli cations for the geologic past. Geochimica et Cosmochimica Acta 75 (23), 7472 7485. Diefendorf, A. F., & Freimuth, E. J. (2017). Extracting the most from terrestrial plant derived n alkyl lipids and their carbon isotopes from the sedimentary record: A revi ew. Organic Geochemistry 103 1 21. Diefendorf, A. F., Mueller, K. E., Wing, S. L., Koch, P. L., & Freeman, K. H. (2010). Global patterns in leaf 13C discrimination and implications for studies of past and future climate. Proceedings of the National Acad emy of Sciences 107 (13), 5738 5743. Dolman, A. M., Rcker, J., Pick, F. R., Fastner, J., Rohrlack, T., Mischke, U., & Wiedner, C. (2012). Cyanobacteria and cyanotoxins: the influence of nitrogen versus phosphorus. PloS one 7 (6), e38757. Donar, C., Stoe rmer, E. F., & Brenner, M. (2009). The Holocene paleolimnology of Lake Apopka, Florida. Nova Hedwigia 57 70. Donders, T. H. (2014). Middle Holocene humidity increase in Florida: Climate or sea level?. Quaternary Science Reviews 103 170 174. Donders, T H., de Boer, H. J., Finsinger, W., Grimm, E. C., Dekker, S. C., Reichart, G. J., & Wagner Cremer, F. (2011). Impact of the Atlantic Warm Pool on precipitation and temperature in Florida during North Atlantic cold spells. Climate Dynamics 36 (1 2), 109 11 8. Donders, T. H., Wagner, F., Dilcher, D. L., & Visscher, H. (2005). Mid to late Holocene El Nino Southern Oscillation dynamics reflected in the subtropical terrestrial realm. Proceedings of the National Academy of Sciences of the United States of Americ a 102 (31), 10904 10908. Donders, T. H., Wagner Cremer, F., & Visscher, H. (2008). Integration of proxy data and model scenarios for the mid Holocene onset of modern ENSO variability. Quaternary Science Reviews, 27(5), 571 579. Douglas, P. M., Pagani, M. Brenner, M., Hodell, D. A., & Curtis, J. H. (2012). Aridity and vegetation composition are important determinants of leaf southeastern Mexico and Central America. Geochimica et Cosmochimica Acta 97 24 45.
105 Eadie, B. J., Chambers, R. L. Gardner, W. S., & Bell, G. L. (1984). Sediment trap studies in Lake Michigan: Resuspension and chemical fluxes in the southern basin. Journal of Great Lakes Research 10 (3), 307 321. Eglinton, G., & Hamilton, R. J. (1967). Leaf epicuticular waxes. Scien ce 156 (3780), 1322 1335. Ehleringer, J. R., Cerling, T. E., & Helliker, B. R. (1997). C4 photosynthesis, atmospheric CO2, and climate. Oecologia 112 (3), 285 299. Ehleringer, J. R., & Cooper, T. A. (1988). Correlations between carbon isotope ratio and m icrohabitat in desert plants. Oecologia 76 (4), 562 566. Escobar, J., Buck, D. G., Brenner, M., Curtis, J. H., & Hoyos, N. (2009). Thermal stratification, mixing, and heat budgets of Florida lakes. Fundamental and Applied Limnology/Archiv fr Hydrobiologi e 174 (4), 283 293. Eynaud, F., De Abreu, L., Voelker, A., Schnfeld, J., Salgueiro, E., Turon, J. L., ... & Malaiz, B. (2009). Position of the Polar Front along the western Iberian margin during key cold episodes of the last 45 ka. Geochemistry, Geophys ics, Geosystems 10 (7). Farquhar, G. D., Ehleringer, J. R., & Hubick, K. T. (1989). Carbon isotope discrimination and photosynthesis. Annual review of plant biology 40 (1), 503 537. Fairbanks, R. G., Mortlock, R. A., Chiu, T. C., Cao, L., Kaplan, A., Gui lderson, T. P., ... & Nadeau, M. J. (2005). Radiocarbon calibration curve spanning 0 to 50,000 years BP based on paired 230 Th/234 U/238 U and 14 C dates on pristine corals. Quaternary Science Reviews 24 (16), 1781 1796. Feakins, S. J., & Sessions, A. L. (2010). Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74 (7), 2128 2141. Ferrio, J. P., Voltas, J., & Araus, J. L. (2003). Use of carbon isotope composition in monitoring environmental changes. Manage ment of Environmental Quality: An International Journal 14 (1), 82 98. Ficken, K. J., Li, B., Swain, D. L., & Eglinton, G. (2000). An n alkane proxy for the sedimentary input of submerged/floating freshwater aquatic macrophytes. Organic geochemistry 31 (7 ), 745 749. Filley, T. R., Freeman, K. H., Bianchi, T. S., Baskaran, M., Colarusso, L., & Hatcher, P. G. (2001). An isotopic biogeochemical assessment of shifts in organic matter input to Holocene sediments from Mud Lake, Florida. Organic Geochemistry 32 (9), 1153 1167.
106 Fisher, M. M., Brenner, M., & Reddy, K. R. (1992). A simple, inexpensive piston corer for collecting undisturbed sediment/water interface profiles. Journal of Paleolimnology 7 (2), 157 161. Florida Department of Environmental Protection ( FDEP). (2008). Integrated Water Quality Assessment for Florida: 2008 305(b) Report and 303(d) List Update. Florida LAKEWATCH. (2003). Florida LAKEWATCH Annual Data Summaries 2002. Department of Fisheries and Aquatic Sciences, University of Florida/Institu te of Food and Agricultural Sciences. Library, University of Florida. Gainesville, Florida. Flower, B. P., Hastings, D. W., Hill, H. W., & Quinn, T. M. (2004). Phasing of deglacial warming and Laurentide Ice Sheet meltwater in the Gulf of Mexico. Geology 32 (7), 597 600. Florea, L. J., & McGee, D. K. (2010). Stable isotopic and geochemical variability within shallow groundwater beneath a hardwood hammock and surface water in an adjoining slough (Everglades National Park, Florida, USA). Isotopes in enviro nmental and health studies 46 (2), 190 209. Fulton III, R. S., & Smith, D. (2008). Development of phosphorus load reduction goals for seven lakes in the upper Ocklawaha river basin, Florida. Lake and Reservoir Management, 24(2), 139 154. Glman, V., Rydb erg, J., de Luna, S. S., Bindler, R., & Renberg, I. (2008). Carbon and nitrogen loss rates during aging of lake sediment: Changes over 27 years studied in varved lake sediment. Limnology and Oceanography 53 (3), 1076 1082. Garcin, Y., Schefu, E., Schwab, V. F., Garreta, V., Gleixner, G., Vincens, A., ... & Sachse, D. (2014). Reconstructing C 3 and C 4 vegetation cover using n alkane carbon isotope ratios in recent lake sediments from Cameroon, Western Central Africa. Geochimica et Cosmochimica Acta 142 482 500. Gat, J. R. (1996). Oxygen and hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24 (1), 225 262. Grimm, E. C., Jacobson, G. L., Watts, W. A., Hansen, B. C., & Maasch, K. A. (1993). A 50,000 year record of c limate oscillations from Florida and its temporal correlation with the Heinrich events. SCIENCE NEW YORK THEN WASHINGTON 261 198 198. Grimm, E. C., Watts, W. A., Jacobson, G. L., Hansen, B. C., Almquist, H. R., & Dieffenbacher Krall, A. C. (2006). Evid ence for warm wet Heinrich events in Florida. Quaternary Science Reviews 25 (17), 2197 2211.
107 Gu, B., Chapman, A. D., & Schelske, C. L. (2006). Factors controlling seasonal variations in stable isotope composition of particulate organic matter in a soft water eutrophic lake. Limnology and Oceanography 51 (6), 2837 2848. Gu, B., Schelske, C. L., & Brenner, M. (1996). Relationship between sediment and lakes. Canadian Journal of Fis heries and Aquatic Sciences 53 (4), 875 883. 13 C enrichments in a shallow hypereutrophic lake: Implications for carbon cycling. Limnology and Oceanography 49 1152 1159. Han, J., & Calvin, M. (19 69). Hydrocarbon distribution of algae and bacteria, and microbiological activity in sediments. Proceedings of the National Academy of Sciences 64 (2), 436 443. Hartman, G., & Danin, A. (2010). Isotopic values of plants in relation to water availability i n the Eastern Mediterranean region. Oecologia 162 (4), 837 852. Hasler, A. D. (1947). Eutrophication of lakes by domestic drainage. Ecology 28 (4), 383 395. Hemming, S. R. (2004). Heinrich events: Massive late Pleistocene detritus layers of the North Atl antic and their global climate imprint. Reviews of Geophysics 42 (1). Ho, E. S., & Meyers, P. A. (1994). Variability of early diagenesis in lake sediments: evidence from the sedimentary geolipid record in an isolated tarn. Chemical Geology 112 (3 4), 309 324. Hodell, D. A., Turchyn, A. V., Wiseman, C. J., Escobar, J., Curtis, J. H., Brenner, M., ... & B rown, E. T. (2012). Late Glacial temperature and precipitation changes in the gypsum hydration water. Geochimica et Cosmochimica Acta 77 352 368. Hoogakker, B. A. A., Downy, F. Andersson, M. A., Chapman, M. R., Elderfield, H., McCave, I. N., ... & Gruetzner, J. (2013). Gulf Stream subtropical gyre properties across two Dansgaard Oeschger cycles. Quaternary Science Reviews 81 105 113. Huang, Y., Shuman, B., Wang, Y., Webb, T. Grimm, E. C., & Jacobson, G. L. (2006). Climatic and environmental controls on the variation of C 3 and C 4 plant abundances in central Florida for the past 62,000 years. Palaeogeography, Palaeoclimatology, Palaeoecology 237 (2), 428 435.
108 Kahmen, A., Hoffmann, B., Schefu, E., Arndt, S. K., Cernusak, L. A., West, J. B., & Sachse, D. (2013 b ). Leaf water deuterium enrichment shapes leaf wax n values of angiosperm plants II: Observational evidence and global implications. Geochimica et Cosmochi mica Acta 111 50 63. Kahmen, A., Schefu, E., & Sachse, D. (2013 a ). Leaf water deuterium enrichment shapes leaf wax n mechanistic insights. Geochimica et Cosmochimica Acta 111 39 49. Kahmen, A., Simonin, K., Tu, K. P., Merchant, A., Callister, A., Siegwolf, R., ... & Arndt, S. K. (2008). Effects of environmental parameters, leaf physiological properties and leaf ies. Plant, Cell & Environment 31 (6), 738 751. Kenney, W. F., Brenner, M., Curtis, J. H., Arnold, T. E., & Schelske, C. L. (2016). A Holocene Sediment Record of Phosphorus Accumulation in Shallow Lake Harris, Florida (USA) Offers New Perspectives on Rece nt Cultural Eutrophication. PloS one 11 (1), e0147331. Kenney, W. F., Brenner, M., Curtis, J. H., & Schelske, C. L. (2010). Identifying sources of organic matter in sediments of shallow lakes using multiple geochemical variables. Journal of Paleolimnology 44 (4), 1039 1052. Kreveld, S. V., Sarnthein, M., Erlenkeuser, H., Grootes, P., Jung, S., Nadeau, M. J., ... & Voelker, A. (2000). Potential links between surging ice sheets, circulation changes, and the Dansgaard Oeschger cycles in the Irminger Sea, 60 18 kyr. Paleoceanography 15 (4), 425 442. Landsberg, J. H., Flewelling, L. J., & Naar, J. (2009). Karenia brevis red tides, brevetoxins in the food web, and impacts on natural resources: Decadal advancements. Harmf ul Algae 8 (4), 598 607. Leuenberger, M., Siegenthaler, U., & Langway, C. (1992). Carbon isotope composition of atmospheric CO 2 during the last ice age from an Antarctic ice core. Nature 357 (6378), 488 490. Lynch Stieglitz, J., Schmidt, M. W., Henry, L G., Curry, W. B., Skinner, L. C., Mulitza, S., ... & Chang, P. (2014). Muted change in Atlantic overturning circulation over some glacial aged Heinrich events. Nature Geoscience 7 (2), 144 150. Magill, C. R., Ashley, G. M., & Freeman, K. H. (2013). Wate r, plants, and early human habitats in eastern Africa. Proceedings of the National Academy of Sciences 110 (4), 1175 1180. Maliva, R., & Missimer, T. (2012). Aridity and drought. In Arid lands water evaluation and management (pp. 21 39). Springer Berlin H eidelberg.
109 Marcott, S. A., Clark, P. U., Padman, L., Klinkhammer, G. P., Springer, S. R., Liu, Z., ... & He, F. (2011). Ice shelf collapse from subsurface warming as a trigger for Heinrich events. Proceedings of the National Academy of Sciences 108 (33), 13415 13419. Marshall, J. D., Brooks, J. R., & Lajtha, K. (2007). Sources of variation in the stable isotopic composition of plants. Stable isotopes in ecology and environmental science 2 22 60. Meyers, P. A. (1994). Preservation of source identificati on of sedimentary organic matter during and after deposition. Chemical Geology 144 (3/4), 289 302. Meyers, P. A. (1997). Organic geochemical proxies of paleoceanographic, paleolimnologic, and paleoclimatic processes. Organic geochemistry 27 (5), 213 250. Meyers, P. A. (2003). Applications of organic geochemistry to paleolimnological reconstructions: a summary of examples from the Laurentian Great Lakes. Organic geochemistry 34 (2), 261 289. Meyers, P. A., & Ishiwatari, R. (1993). Lacustrine organic geoch emistry an overview of indicators of organic matter sources and diagenesis in lake sediments. Organic geochemistry 20 (7), 867 900. Mook, W. G., Bommerson, J. C., & Staverman, W. H. (1974). Carbon isotope fractionation between dissolved bicarbonate and ga seous carbon dioxide. Earth and Planetary Science Letters 22 (2), 169 176. Mgler, I., Sachse, D., Werner, M., Xu, B., Wu, G., Yao, T., & Gleixner, G. (2008). Effect alkanes: a comparison of Nam Co (Tibetan Plateau) and Holzmaar (Germany). Organic Geochemistry 39 (6), 711 729. Muri, G., & Wakeham, S. G. (2006). Organic matter and lipids in sediments of Lake Bled (NW Slovenia): source and effect of anoxic and oxic depositional regimes. Organic Geochemistry 37 (12), 1664 1679. Naafs, B. D. A., Hefter, J., Gruetzner, J., & Stein, R. (2013). Warming of surface waters in the mid latitude North Atlantic during Heinrich events. Paleoceanography 28 (1), 153 163. Nichols, J. E. (2011). Procedures for extraction and purification of leaf wax biomarkers from peats. Mires and Peat, 7(13), 1 7. O'Leary, M. H. (1988). Carbon isotopes in photosynthesis. Bioscience 38 (5), 328 336.
110 Pailler, D., & Bard, E. (2002). High frequency palaeoceanographic changes during the past 140 000 yr recorded by the organic matter in sediments of the Iberian Margin. Palaeogeography, Palaeoclimatology, Palaeoecology 181 (4), 431 452. Palomo, L., & Canuel, E. A. (2010). Sources of fatty acids in sediments of the York River estuary: relations hips with physical and biological processes. Estuaries and coasts 33 (3), 585 599. Parker, A. O., Schmidt, M. W., & Chang, P. (2015). Tropical North Atlantic subsurface warming events as a fingerprint for AMOC variability during Marine Isotope Stage 3. Pa leoceanography 30 (11), 1425 1436. Parkes, R. J., & Taylor, J. (1983). The relationship between fatty acid distributions and bacterial respiratory types in contemporary marine sediments. Estuarine, Coastal and Shelf Science 16 (2), 173IN5175 174189. Patt on, G. M., Martin, P. A., Voelker, A., & Salgueiro, E. (2011). Multiproxy comparison of oceanographic temperature during Heinrich Events in the eastern subtropical Atlantic. Earth and Planetary Science Letters 310 (1), 45 58. compositions of n alkanes from modern angiosperms and conifers: an experimental set up in central Washington State, USA. Organic Geochemistry 39 (8), 1066 1071. Peterson, B. J., & Fry, B. (1987). Stable isotopes in ecosystem studies. Annual review of ecology and systematics 18 (1), 293 320. Geochimica et Cosmochimica Acta 129 146 156. Quillen, A. K. (2009). Diatom based paleolimnological reconstruction of quaternary environments in a Florida Sinkhole Lake. Quillen, A. K., Gaiser, E. E., & Grimm, E. C. (2013). Diatom based paleolimnological reconstruction of regio nal climate and local land use change from a protected sinkhole lake in southern Florida, USA. Journal of paleolimnology 49 (1), 15 30. Rach, O., Brauer, A., Wilkes, H., & Sachse, D. (2014). Delayed hydrological response to Greenland cooling at the onset of the Younger Dryas in western Europe. Nature Geoscience 7 (2), 109 112. Rasmussen, T. L., & Thomsen, E. (2012). Changes in planktic foraminiferal faunas, temperature and salinity in the Gulf Stream during the last 30,000 years: influence of meltwater vi a the Mississippi River. Quaternary Science Reviews 33 42 54.
111 Riedinger Whitmore, M. A., Whitmore, T. J., Smoak, J. M., Brenner, M., Moore, A., Curtis, J., & Schelske, C. L. (2005). Cyanobacterial proliferation is a recent response to eutrophication in many Florida lakes: a paleolimnological assessment. Lake and Reservoir Management 21 (4), 423 435. Renold, M., Raible, C. C., Yoshimori, M., & Stocker, T. F. (2010). Simulated resumption of the North Atlantic meridional overturning circulation Slow basin wide advection and abrupt local convection. Quaternary Science Reviews 29 (1), 101 112. Sachse, D., Billault, I., Bowen, G. J., Chikaraishi, Y., Dawson, T. E., Feakins, S. J., ... & Polissar, P. (2012). Molecular paleohydrology: interpreting the hydrogen isotopic composition of lipid biomarkers from photosynthesizing organisms. Annual Review of Earth and Planetary Sciences 40 221 249. Sachse, D., Gleixner, G., Wilkes, H., & Kahmen, A. (2010). Leaf wax n values of field grown barley reflect le formation. Geochimica et Cosmochimica Acta 74 (23), 6741 6750. Sachse, D., Radke, J., & Gleixner, G. (2004). Hydrogen isotope ratios of recent lacustrine sedimentary n alkanes record modern climate variability. Geoch imica et Cosmochimica Acta 68 (23), 4877 4889. Schelske, C. L. (1997). Sediment and phosphorus deposition in Lake Apopka. Schelske, C. L., Lowe, E. F., Battoe, L. E., Brenner, M., Coveney, M. F., & Kenney, W. F. (2005). Abrupt biological response to hydr ologic and land use changes in Lake Apopka, Florida, USA. AMBIO: A Journal of the Human Environment 34 (3), 192 198. Schelske, C. L., Peplow, A., Brenner, M., & Spencer, C. N. (1994). Low background gamma counting: applications for210Pb dating of sediment s. Journal of Paleolimnology 10 (2), 115 128. Schmidt, M. W., Vautravers, M. J., & Spero, H. J. (2006). Rapid subtropical North Atlantic salinity oscillations across Dansgaard Oeschger cycles. Nature 443 (7111), 561 564. Sharp, Z. (2007). Principles of s table isotope geochemistry (p. 344). Upper Saddle River, NJ: Pearson education. Shumate, B. C., Schelske, C. L., Crisman, T. L., & Kenney, W. F. (2002). Response of the cladoceran community to trophic state change in Lake Apopka, Florida. Journal of Paleo limnology 27 (1), 71 77. Silliman, J. E., Meyers, P. A., & Bourbonniere, R. A. (1996). Record of postglacial organic matter delivery and burial in sediments of Lake Ontario. Organic Geochemistry 24 (4), 463 472.
112 Silliman, J. E., & Schelske, C. L. (2003). Saturated hydrocarbons in the sediments of Lake Apopka, Florida. Organic Geochemistry 34 (2), 253 260. Smith F. A., & Freeman, K.H. (2006). Influence of physiology and climate on delta D of leaf wax n alkanes from C 3 and C 4 grasses. Geochimica Cosmochim ica Acta 70, 1172 1187. Stewart, G. R., Turnbull, M. H., Schmidt, S., & Erskine, P. D. (1995). 13C natural abundance in plant communities along a rainfall gradient: a biological integrator of water availability. Functional Plant Biology 22 (1), 51 55. S wain, E. B. (1985). Measurement and interpretation of sedimentary pigments. Freshwater Biology 15 (1), 53 75. Tierney, J. E., Russell, J. M., Huang, Y., Damst, J. S. S., Hopmans, E. C., & Cohen, A. S. (2008). Northern hemisphere controls on tropical sout heast African climate during the past 60,000 years. Science 322 (5899), 252 255. Trnqvist, T. E., Gonzlez, J. L., Newsom, L. A., van der Borg, K., de Jong, A. F., & Kurnik, C. W. (2004). Deciphering Holocene sea level history on the US Gulf Coast: A hig h resolution record from the Mississippi Delta. Geological Society of America Bulletin, 116(7 8), 1026 1039. Torres, I. C., Inglett, P. W., Brenner, M., Kenney, W. F., & Reddy, K. R. (2012). Stable in subtropical lakes of different trophic status. Journal of paleolimnology, 47(4), 693 706. Waghorn, M. J., Whitehead, D., Watt, M. S., Mason, E. G., & Harrington, J. J. (2015). Growth, biomass, leaf area and water use efficiency of juvenile Pinus radia ta in response to water deficits. New Zealand Journal of Forestry Science 45 (1), 3. Wakeham, S. G., Forrest, J., Masiello, C. A., Glinas, Y., Alexander, C. R., & Leavitt, P. R. (2004). Hydrocarbons in Lake Washington sediments. A 25 year retrospective i n an urban lake. Environmental science & technology 38 (2), 431 439. Wang, Y. J., Cheng, H., Edwards, R. L., An, Z. S., Wu, J. Y., Shen, C. C., & Dorale, J. A. (2001). A high resolution absolute dated late Pleistocene monsoon record from Hulu Cave, China. Science 294 (5550), 2345 2348. Wang, Y. V., Larsen, T., Leduc, G., Andersen, N., Blanz, T., & Schneider, R. R. (2013). Geochimica et Cosmochimica Acta 111 128 139. Waters, M. N. (2 016). A 4700 year history of cyanobacteria toxin production in a shallow subtropical lake. Ecosystems 19 (3), 426 436.
113 Watts, W. A. (1971). Postglacial and interglacial vegetation history of southern Georgia and central Florida. Ecology 52 (4), 676 690. Watts, W. A. (1975). A late Quaternary record of vegetation from Lake Annie, south central Florida. Geology 3 (6), 344 346. Watts, W. A., & Hansen, B. C. S. (1994). Pre Holocene and Holocene pollen records of vegetation history from the Florida peninsula and their climatic implications. Palaeogeography, Palaeoclimatology, Palaeoecology 109 (2 4), 163 176. Watts, W. A., & Stuiver, M. (1980). Late Wisconsin climate of northern Florida and the origin of species rich deciduous forest. Science 210 (4467), 325 327. Weaver, A. J., Saenko, O. A., Clark, P. U., & Mitrovica, J. X. (2003). Meltwater pulse 1A from Antarctica as a trigger of the Blling Allerd warm interval. Science 299 (5613), 1709 1713. Weldeab, S., Schneider, R. R., & Klling, M. (2006). Deglacia l sea surface temperature and salinity increase in the western tropical Atlantic in synchrony with high latitude climate instabilities. Earth and Planetary Science Letters 241 (3), 699 706. White, W. A. (1970). The geomorphology of the Florida peninsula. GSA Bulletin, 51, 1 172. Whiticar, M. J. (1999). Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology 161 (1), 291 314. Willard, D. A., & Bernhardt, C. E. (2011). Impacts of past climate and sea leve l change on Everglades wetlands: placing a century of anthropogenic change into a late Holocene context. Climatic Change, 107(1 2), 59 80. Zhang, X., Bianchi, T. S., & Allison, M. A. (2015). Sources of organic matter in sediments of the Colville River del ta, Alaska: A multi proxy approach. Organic Geochemistry 87 96 106. Zhao, J. X., Wang, Y. J., Collerson, K. D., & Gagan, M. K. (2003). Speleothem U series dating of semi synchronous climate oscillations during the last deglaciation. Earth and Planetary Science Letters 216 (1), 155 161. Zhou, H., Zhao, J., Feng, Y., Gagan, M. K., Zhou, G., & Yan, J. (2008). Distinct climate change synchronous with Heinrich event one, recorded by stable oxygen and carbon isotopic compositions in stalagmites from China. Q uaternary Research 69 (2), 306 315. Ziegler, M., Nrnberg, D., Karas, C., Tiedemann, R., & Lourens, L. J. (2008). Persistent summer expansion of the Atlantic Warm Pool during glacial abrupt cold events. Nature Geoscience 1 (9), 601 605.
114 BIOGRAPHICAL SKET CH Thomas Elliott Arnold was born in Lancaster, PA. He graduated with a B.S. degree from Pennsylvania State University and with a PhD degree f rom the University of Florida in 2017 D r. Arnold plans to continue his work in the field of organic geochemistry at the University of Pittsburgh. Although he learned a lot at the University of Florida, if he could do it all over again, he would pursue a career in theater and set design.