CALIBRATION AND APPLICATION OF LIPID H ISOTOPIC RATIOS FOR QUANTITATIVE RECONSTRUCTION OF NEW ENGLAND CLIMATE VARIABILITY OVER THE PAST 15 KYR BY JUZHI HOU B.A., CHINA UNIVERSITY OF GEOSCIENCES, 1998 PH.D., CHINESE ACADEMY OF SCIENCES, 2003 SC.M., BROWN UNIVERSITY, 2006 A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE DEPARTMENT OF GEOLOGICAL SCIENCES AT BROWN UNIVERSITY PROVIDENCE, RHODE ISLAND MAY, 2009 © Copyright 2009 by Juzhi Hou This dissertation by Juzhi Hou is accepted in its present form by the Department of Geological Sciences as satisfying the dissertation requirements of the degree of Doctor of Philosophy Date______________ Dr. Yongsong Huang (Advisor) Recommended to Graduate Council Date______________ Dr. Timothy Herbert (Reader) Date______________ Dr. James Russell (Reader) Date______________ Dr. Alberto Saal (Reader) Date______________ Dr. Julie Brigham-Grette (Reader) Approved by the Graduate School and Research Date______________ Dr. Sheila Bonde Dean of the Graduate School iii VITA JUZHI HOU was born in Xingtai, Hebei Province, China on November 29, 1975. He attended China University of Geosciences, Wuhan, from 1994-1998. Prior to his graduate studies at Brown University in 2003, Juzhi studied Quaternary Geology and Environment at Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, from 1998 to 2003. PREVIOUS EDUCATION: Sc.M. Geological Sciences, Brown University, May, 2006 Ph.D. Institute of Geology and Geophysics, Chinese Academy of Sciences, June 2003 B.S. College of Geosciences China University of Geosciences, Jun 1998 RESEARCH EXPERIENCE: Brown University, Sep 2003 – July 2008. Investigating the application of lipid H isotopic ratios climatic and environmental change and biogeochemical cycles in terrestrial ecosystems. Major specific activities include: Developing new proxies for precipitation isotopic ratios and surface air temperature, and investigating quantitative climate variability in northeastern North iv America, Tibetan Plateau, East Africa; Investigating terrestrial biogeochemical dynamics by developing a novel method to organic matter; Investigating the influence of human activities on environment using carbon isotopic ratios of terrestrial biomarkers. Chinese Academy of Sciences, Sep 1998 – June 2003. Investigated variability of East Asian monsoon during late Holocene. Specific activities included: Examining optical characteristics of speleothem micro-bands in north and south China; Constructing speleothem micro-band chronology by comparing radiometric dates and micro-band counting; Investigating temperature variability at annual-scale, precipitation and atmospheric circulation associated with East Asian monsoon during late Holocene. TEACHING EXPERIENCE: Teaching Assistant, Brown University. Environmental Geochemistry, Limnology. Planed, organized field trips; Organized, prepared, and lead laboratories for 3 groups of 7-8 students each; Lectured at beginning of each laboratory; Supervised entire lab procedures and data production; Graded homework; Met with students one-on-one for tutoring and mentoring. Rated very good to excellent as a TA. Supervisor, Brown University. Senior thesis, Luke Parsons (2005-2006); Corynn Brodsky (2007-present); Independent research by undergraduate students, Mena Ramos v (2004-2005), Sophie McCoy (2005-2006), Gideon Ng (2006-2007), Tsveta Krumova (2005-2007), Baird Langenbrunner (2006-present). Teaching Assistant, Chinese Academy of Sciences, Quaternary Environment. Rated excellent as a TA. PEER REVIEWED PAPERS: 1. Hou J., D'Andrea W.J. and Huang Y. (2008) Can sedimentary leaf waxes record D/H ratios of continental precipitation? Field, model and experimental assessments. Geochimica et Cosmochimica Acta, 72, 3503-3517, doi:10.1016/j.gca.2008.04.030 2. Hou J., Huang Y., Oswald W.W., Foster D.R. and Shuman B. (2007) Centennial-scale compound-specific hydrogen isotope record of Pleistocene - Holocene climate transition from southern New England. Geophysical Research Letters, 34, L19706, doi:10.1029/2007GL030303. 3. Hou J., D'Andrea W. J., Macdonald D. and Huang Y. (2007) Evidence for water use efficiency as an important factor in determining the !D values of tree leaf waxes. Organic Geochemistry, doi: 10.1016/j.orggeochem.2007.03.011. 4. Hou J., D'Andrea W. J., Macdonald D. and Huang Y. (2007). Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts (USA). Organic Geochemistry, doi:10.1016/j.orggeochem.2006.12.009. vi 5. Lindbladh M., Oswald W. W., Foster D. R., Faison E., Hou J. and Huang Y. (2007) A late-glacial transition from Picea glauca to Picea Mariana in southern New England. Quaternary Research, 67, 502-508. 6. Hou J., Huang Y., Wang Y., Shuman B., Oswald W.W., Faison E. and Foster D.R. (2006) Postglacial climate reconstruction based on compound-specific D/H ratios of fatty acids from Blood Pond, New England. Geochemistry Geophysics Geosystems, 7, doi:10.1029/2005GC001076. 7. Cai B., Cheng H., Hou J., Liu H., Wang G. and Liu T.S. (2005) The impact of soil erosion on the formation of Yunnan Stone Forest: Evidence from Stalagmite and field observations. Quaternary Sciences (in Chinese with English Abstract), 25, 170-176. 8. Tan M., Hou J. and Liu T.S. (2004) Sun-coupled climate connection between eastern Asia and northern Atlantic. Geophysical Research Letters , 31, doi:10.1029/2003GL019085. 9. Hou J. and Tan M. (2003) Image characteristics of annual layers in very young Chinese stalagmites. Supplementi di Geografia Fisica e Dinamica Quaternaria, 65- 69. 10. Hou J., Tan M., Liu T.S. and Cheng H. (2003) Stable isotope records of plant cover change and monsoon variation in the past 2200 years: evidence from laminated stalagmites. Boreas, 32, 304-313. 11. Tan M., Liu T.S., Hou J., Qin X., Zhang H. and Li T. (2003) Cyclic rapid warming on centennial-scale revealed by a 2650-year stalagmite record of warm season temperature. Geophysical Research Letters, 30, doi:10.1029/2003GL017352. 12. Hou J., Tan M. and Liu T.S. (2002) Counting chronology and climate records with vii about 1000 annual layers of a Holocene stalagmite from the Water Cave in Liaoning, China. Science in China, 45(5), 385-391. 13. Tan M., Hou J. and Cheng H. (2002) Methodology of quaternary reconstructing paleoclimate from annually laminated stalagmites. Quaternary Sciences (in Chinese with English Abstract), 22(3), 209-219. 14. Hou J., Tan M., Cheng H. and Liu T.S. (2001) Preliminary research on Layer counting chronology of young stalagmite from Water Cave, Benxi. Science in China (Series D, Earth Sciences, in Chinese), 31(5), 387-392 15. Hou J. (2001) Image characteristics of annual micro-bands in stalagmites. Quaternary Sciences (in Chinese), 21(3), 278 16. Tan M., Cheng H., Edwards R.L., Hou J. and Liu T.S. (2000) TIMS-230Th dating for very young annual laminated stalagmites. Quaternary Sciences (in Chinese), 20, 391. PRESENTATIONS: 1. Hou J., Russell J.M. and Huang Y. Penetration of Atlantic Walker circulation into easternmost Africa during Early-mid Holocene: Hydrogen isotopic evidence from Sacred Lake. Oral presentation at American Geophysical Union Fall Meeting, 2007. 2. Majumdar S., Feng X., Hou J., Faiia A.M. and Huang Y. Relationship between Leaf Water and Leaf Waxes in Growing Pine Needles. Poster presentation at American Geophysical Union Fall Meeting, 2007. 3. Hou J., D’Andrea W.D., Toney J. and Huang Y. Factor controlling hydrogen isotope ratios of sedimentary leaf waxes: Implications for interpreting !D records of viii paleoclimate (PP41A-1178). Poster presentation at American Geophysical Union Fall Meeting, 2006. 4. Hou J., Parsons L., Huang Y., Shuman B. and Donnelly J. High resolution hydrogen isotope records of climate change during the early Holocene from two adjacent lakes in northeastern North America (PP43A-1227). Poster presentation at American Geophysical Union Fall Meeting, 2006. 5. Toney J., Hou J. and Huang Y. H-isotope fractionation during biosynthesis of leaf waxes in C3 plants: trees vs. grasses (PP41A-1177). Poster presentation at American Geophysical Union Fall Meeting, 2006. 6. Hou J., Huang Y., Shuman B., Oswald W.W., Faison E. and Foster D.R. Postglacial climate reconstruction based on D/H ratios of fatty acids from Blood Pond, New England. Oral presentation at American Geophysical Union Fall Meeting, 2005. 7. Hou J., Huang Y., Tan M., Edwards R.L. and Smith E. Carbon isotopic ratios of lignin moieties in stalagmites as a novel indicator of past C3/C4 plant variation (PP43-0607). Poster presentation at American Geophysical Union Fall Meeting, 2004. 8. Hou J., Tan M. and Liu T.S. Layer chronology of stalagmite from northeast China. Oral presentation at The 5th International Conference on Geomorphology, Tokyo, 2001. 9. Hou J., Tan M., Liu T.S. and Cheng H. Stable isotope records of plant cover change and monsoon variation in the past 2200 years: evidence from laminated stalagmites. Oral presentation at The 5th International Conference on Geomorphology, Tokyo, 2001. 10. Hou J. and Tan M. Optical characteristics of speleothem micro-bands and their ix climatic implications. Oral presentation at Annual of High-resolution Committee in Chinese Association of Quaternary Research, Guangzhou, 2001. PROFESSIONAL SOCIETY MEMBERSHIPS: Member, American Geophysical Union Member, Geological Society of America Member, International Paleolimnology Association Member, Rhode Island Natural History Survey AWARDS AND GRANTS: National Ocean Sciences Accelerator Mass Spectrometry, Woods Hole Oceanographic Institution, Graduate Student Internship, Jan 2008 - Jul 2008. Brown University, University Fellowship, Sep 2003 – Jun 2004. Chinese Academy of Sciences, Peng Yin-Gang Scholarship, 2001. Chinese Academy of Sciences, Institute of Geology and Geophysics, 2003, Award to excellent publications. x PREFACE INTRODUCTION The late Pleistocene – early Holocene transition is characterized by abrupt climatic fluctuations around North Atlantic Ocean (e.g., Stuiver et al., 1995; Hughen et al., 1996). However, the spatial variations in the timing, amplitude, and phasing of the abrupt events are less understood on the adjacent continents. This is particularly true for the northeastern North America where the driving forces for climate were particularly complex, comprising a combination of changes in North Atlantic surface temperature, Laurentide ice sheet (LIS) extent, atmospheric composition, and insolation (e.g., Webb et al., 1993). Centennial-scale quantitative records from the northeastern North America are thus extremely important for better understanding marine-terrestrial-atmosphere- cryosphere connections and regional climatic responses. Existing paleoclimate records from the northeastern North America are mainly based on paleoecological and stratigraphic approaches, such as assemblages of pollen (e.g., Peteet et al., 1990; Webb et al., 1993; Shuman et al., 2002), Chironomidae (midge, e.g., Cwynar and Spear, 2001; Walker et al., 1997) and lake level (Shuman et al., 2001; Newby et al., 2000) from lake sediments. However, the possibility of transient vegetation responses to short-term, small-magnitude climate variations (e.g., Davis and Botkin, 1985), and the non-analogue conditions for Chironomidae (e.g., Kurek et al., 2002) could increase the difficulty of using pollen and Chironomidae data to assess abrupt xi ( 5°C cooler than before (based on the global relationship between surface air temperature and !D for mean annual precipitation, –5.6‰/°C (Dansgaard, 1964)). At Blood Pond, the temperature change at the YDC derived from hydrogen isotope records of palmitic acid and behenic acid ranges from 3 to 5 °C (the small difference between the two compounds may be due to mixing the contributions of organisms with small fractionation differences). The 15 temperature estimates are close to the temperature variation estimated from nitrogen and argon isotopes from the Greenland ice core (Severinghaus et al., 1998). This cold period was accompanied by increased abundance of boreal taxa such as Picea, Alnus and Betula, and lower abundance of Pinus and Quercus around Blood Pond (Figure 4). At the end of the YDC, the isotopic records indicate rapid warming (Figure 4), which is consistent with the isotopic records from Crooked Pond (Huang et al., 2002; Shuman et al., 2002b), and the Greenland ice core isotopic records (Stuiver et al., 1995). Moreover, the YDC in our study site was not a uniformly cold period. There were some fluctuations, including a warm (~2°C) and wet reversal occurred during the later part of the YDC. Pollen data (and to some extent, Greenland ice core data) also appear to record corresponding variations. From the end of the YDC to 9000 cal yr B.P., temperature variations derived from the hydrogen isotopic records at Blood Pond show different patterns from the Greenland ice core isotopic records. Greenland ice core isotopic records show a gradual warming with small variability until 9000 cal yr B.P. However, two relative cool periods are indicated by the temperature records from Blood Pond and Crooked Pond (Huang et al., 2002). The first cool period is coincident with high percentages of Tsuga pollen, while the second coincides with an Alnus pollen peak around 9800 cal yr B.P. The 8.2 ka event was a major, abrupt climate change event in the early Holocene (Alley and Agustsdottir, 2005; Alley et al., 1997). It is believed that the strongest 16 evidence of this cooler, drier and perhaps windier event should be recovered in eastern North America. However, so far evidence for a strong 8.2 ka event in eastern North America has been sparse. For example, Yu and Eicher (1998) found an oxygen isotope shift in carbonate toward colder values in a lake in southern Ontario, Canada. A lake in central Nova Scotia, Canada showed a short-lived shift to more minerogenic conditions at the time of the event, indicating ecological change in the lake (Spooner et al., 2002). In New England, a sharp drop in loss-on-ignition was detected in two lakes from New Hampshire (Kurek et al., 2004). Shuman et al. (2004) showed that pollen data from North Pond (western Massachusetts) contain a reversal that may be attributed to the 8.2 ka event. There is no clear response of pollen record from Blood Pond to the 8.2 ka cold event. In contrast, !D records of palmitic acid and behenic acid clearly record this cold event (~9000 – 8000 cal yr B.P.). The estimated temperature variation based on our data is 1.7 to 2.8 °C. This cooling is less than suggested by the Greenland ice core (5-7°C) (Alley and Agustsdottir, 2005; Severinghaus et al., 1998), but is consistent with the variation in a lake in southern Ontario, Canada (Yu and Eicher, 1998). The 8.2 ka event at Blood Pond appears to have started earlier and lasted longer than at the Greenland summit. Similar longer-duration cold events have been reported in North America (Spooner et al., 2002) and Europe (Ariztegui et al., 2001; Heiri et al., 2003). However, our dating control is based on linear interpolation between only a few 14C dates, and 17 hence we cannot accurately determine the duration of the event without further detailed radiocarbon dating. A relatively cool period between 7000 and 5000 cal yr B.P. was revealed by the !D records from Blood Pond. !18O records from the Greenland ice core also show slight cooling during the same period (Stuiver et al., 1995). Temperatures remained high from 5000 to 3000 cal yr B.P., and then declined. After ~3000 cal yr B.P., declines in the !D values of behenic acid and Carya pollen percentages suggest a cooling trend. This trend is consistent with pollen and hydrogen isotope records from Crooked Pond (Huang et al., 2002; Shuman et al., 2004). Overall, our compound-specific D/H data provide a new paleoclimate proxy that is particularly sensitive to relatively short-term events, such as the 8.2 ka event !D records of long chain fatty acids and implication for effective moisture balance changes Long chain normal fatty acids (C24 to C32 n-acids) originate from the waxy coating of terrestrial plants which can be transported to lake sediments by wind and streams (Meyers, 2003). Higher plants use soil water as their source water for biosynthesis. As a result, the isotopic compositions of organic compounds derived from higher plants are affected by precipitation-evaporation ratios and plant evapotranspiration. 18 We estimated the lake !D values for the past 15 ka in Blood Pond using !DBA, based on the regression established from surface sediments calibration. We then calculated the fractionation factor of long chain fatty acids relative to lake water. The assumption here is that lake water !D values are the same as those used by terrestrial plants, which is the best we can do since we have no other means to reconstruct soil water !D values. Biosynthesis of leaf waxes from leaf water involves many biochemical steps with the possibility of hydrogen isotope fractionation (Buddenbaum and Shiner, 1997). However, it is the overall fractionation rather than the fractionation associated with any one step that is of interest of us. We therefore calculated the overall hydrogen isotope fractionation using the following equations: " = (1000+!Dlong chain acid)/(1000+!Dlake water) (3) An empirical relationship (" = -0.124h + 1.089) between net hydrogen isotope fractionation factor and growing season relative humidity has been obtained from tree ring cellulose (Yapp and Epstein, 1982), indicating higher value of fractionation factor (") under lower relative humidity. A quantitative relationship between hydrogen isotope fractionation of leaf waxes and relative humidity has not been established. However, study of a loess profile spanning the past 130 ka by Liu and Huang (2005) has demonstrated that !D values of long chain leaf waxes during dry periods are 40 to 50 ‰ 19 higher than during wet periods in the Chinese Loess Plateau. Therefore, the impact of relative humidity on hydrogen isotopic fractionation must be in the same direction for both tree ring cellulose and leaf waxes. The variations of growing season relative humidity estimated from isotope fractionation factor can be compared with the Blood Pond pollen record and well-established climate scenarios (Figure 4). Growing season relative humidity increased in parallel with Picea and Pinus pollen percentages from 15 000 to 13 000 cal yr B.P. Subsequently, the D/H fractionation factor increases during the YDC, indicating lower growing season relative humidity (Figure 4). This finding is inconsistent with lake-level reconstructions for Crooked Pond and Makepeace Cedar Swamp in eastern Massachusetts, which show a rise during the YDC (Newby et al., 2000; Shuman et al., 2001). Our estimates of the growing season relative humidity are related to precipitation during the growing season of plants (spring and summer in New England). In contrast, lakes in New England were recharged mainly by winter precipitation (Shuman et al., 2001). It is possible that a P/E increase in winter and a P/E reduction in spring and summer occurred during the YDC, resulting in the observed discrepancy in isotope and lake level data. From the end of YDC to 9000 cal yr B.P., the isotope data suggest that growing season relative humidity changed from relatively low to high values (Figure 4). This 20 change coincides with an interval of high Pinus and Tsuga pollen percentages, and a subsequent increase in Quercus abundance. The cool event around 8.2 ka also appears to have lower relative humidity based on the fractionation factor of the long chain acids, a finding which is consistent with climate conditions in other regions (Alley and Agustsdottir, 2005). However, the pollen data did not show clear evidence for the 8.2 ka event (Figure 4), perhaps indicating that isotope ratios are a more sensitive proxy for detecting short climate events. After 8000 cal yr B.P., growing season relative humidity remained relatively high, with minor fluctuations around 5000 and 3000 cal yr B.P. This pattern is inconsistent with the lake-level reconstruction from Crooked Pond (Shuman et al., 2004), which suggests decreased effective moisture during that interval. Comparison with !D records of palmitic acid from Crooked Pond Our previously published hydrogen isotope record from Crooked Pond, Massachusetts (Huang et al., 2002) shows a similar general trend of climate variations to that from Blood Pond, particularly the clear climate cooling during the YDC (Figure 5). The mid-Holocene warm period and relatively cool late Holocene were also observed in our isotope data in both lakes, although the Blood Pond data show more variability. Downcore variations of !DBA and !DPA from Blood Pond are very similar to those of !DPA from Crooked Pond. However, there are a few differences between the records from the two lakes. First, !DPA from Crooked Pond did not reveal a clear response to the cool 21 event around 8200 cal yr B.P. Secondly, the full range of !DPA variation (-245.1 to -157.1‰, ~88 ‰) from Crooked Pond is larger than that of !DBA (-167.3 to -140.7 ‰, ~ 27 ‰) and !DPA (-200.7 to -162.0‰, ~38 ‰) from Blood Pond. The differences are likely due to 1) different sampling resolution for the two records; 2) different hydrological conditions at Crooked and Blood Ponds. Blood Pond is surrounded by a large area of shallow-water swamp, which is not present at Crooked Pond. There is intermittent inflow and a continuous outlet at Blood Pond, whereas Crooked Pond is hydrologically closed. 22 CONCLUSIONS Surface calibration using 33 lakes in eastern North America demonstrated that behenic acid (BA, C22 n-acid) captures lake water isotope ratios effectively. Correlation between lake water !D and !DBA is significant (!DBA = 0.8185 " !Dlake water - 140.01, R= 0.898). Consistent fractionation factors derived from the slope and intercept of the equation suggests that behenic acid is mainly produced by submerged and emergent macrophytes. Downcore comparisons show that !DBA is consistent with pollen record from the same sediment core. In Blood Pond, the downcore variations in !DBA and !DPA are similar and appear to reflect primarily temperature changes for the last 15 ka. In contrast, the isotopic fractionation factor variations of terrestrial plant leaf waxes (long chain acids) reflect growing season relative humidity. Our isotope data are highly sensitive to abrupt climate changes, such as the YDC and the 8.2 ka event. Based on our data, the duration of 8.2 ka event in New England may be longer than on the Greenland ice sheet, although higher resolution radiocarbon dating of Blood Pond sediments is needed to further constrain its exact duration. Hydrogen isotope ratios of long chain fatty acids also allowed us to estimate moisture balance changes during the YDC and the 8.2 ka event. Current pollen records from Blood Pond do not show the relatively short-term climate events such as 8.2 ka event. This suggests that our isotope approach is promising for high resolution reconstruction of Holocene 23 climate change in New England. Fatty acids are a particularly promising class of compounds for paleoclimatic and paleoenvironmental reconstructions because of the possibility of simultaneous reconstruction of both temperature and moisture balance changes. Fatty acids are relatively easy to isolate and purify during sampling preparation, minimizing coelution during gas chromatography – isotope ratio mass spectrometry analyses. Although palmitic and behenic acid are not exclusive biomarkers for aquatic plants, our surface sediment calibrations demonstrate that they do capture modern lake water conditions, and our downcore studies suggest their D/H values record paleoclimatic changes. ACKNOWLEDGEMENTS This work was supported by grants from the National Science Foundation (NSF 0318050, 0318123, 0402383) to Y, Huang. We thank Alex Sessions and an anonymous reviewer for critical reviews on the manuscript and helpful discussions with Dana McDonald. 24 REFERENCES Alley R.B. and Agustsdottir A.M. (2005) The 8k event: cause and consequences of a major Holocene abrupt climate change. Quaternary Science Reviews 24, 1123-1149. Alley R.B., Mayewski P.A., Sowers T., Stuiver M., Taylor K.C. and Clark P. U. (1997) Holocene climatic instablity: A prominent, widespread event 8200 yr ago. Geology 25, 483-486. Anderson W.T., Mullins H.T. and Ito E. (1997) Stable isotope record from Seneca Lake, New York: Evidence for a cold paleoclimate following the Younger Dryas. Geology 25, 135-138. Ariztegui D., Chondrogianni C., Lami A., Guilizzoni P. and Lafargue E. (2001) Lacustrine organic matter and the Holocene paleoenvironmental record of Lake Albano (central Italy). Journal of Paleolimnology 26, 283-292. Bartlein P.J., Webb T. III. and Fleri E. C. (1984) Holocene climate change in the northern Midewest: Pollen-derived estimates. Quaternary Research 22, 361-374. Bronk Ramsey C. (2001) Development of the radiocarbon program OxCal. Radiocarbon 43, 355-363. Buddenbaum W.E. and Shiner V.J.J. (1977) Computation of isotope effects on equilibria and rates, in Isotope effects on Enzyme-catalyzed reactions, edited by W. Cleland et al., University Park Press, Baltimore. Burgoyne T.W. and Hayes J.M. (1998) Quantitative production of H2 by pyrolysis of gas 25 chromatographic effluents. Analytical Chemistry, 70, 5136-514. Coolen M.J.L., Muyzer G., Rijpstra W.I.C., Schouten S., Volkman J.K. and Damste J.S.S. (2004) Combined DNA and lipid analyses of sediments reveal changes in Holocene haptophyte and diatom populations in an Antarctic lake. Earth and Planatary Science Letters 223, 225-239. Cranwell P.A. (1982) Lipids of aquatic sediments and sedimenting particulates. Progress in Lipid Research 21, 271-308. Cwynar L.C. and Levesque A.J. (1995) Chironomid evidence for late-Glacial climatic reversals in Maine. Quaternary Research 43, 403-415. D'Andrea W.J. and Huang Y. (2005) Long chain alkenones in Greenland lake sediments: Low !13C values and exceptional abundance. Organic Geochemistry 36, 1234-1241. Dansgaard W. (1964) Stable isotopes in precipitation. Tellus 16, 436-468. Edwards T.W.D. (1993) Interpreting past climate from stable isotofpes in continental organic matter, in Climatic change in continental isotopic records: American Geophysical Union Geophysical Monograph, edited by P. K. Swart, pp. 333-341. Ficken K.J., Li B., Swain D.L. and Eglinton G. (2000) An n-alkane proxy for the sedimentary input of submerged/floating freshwater aquatic macrophytes. Organic Geochemistry 31, 745-749. Hayes J.M., Freeman K.H., Popp B.N. and Hoham C.N. (1990) Compound-specific isotope analyses: A novel tool for reconstruction of ancient biogeochemical 26 processes. Organic Geochemistry 16, 1115-1128. Heiri O., Lotter A.F., Hausmann S. and Kienast F. (2003) A chironomid-based Holocene summer air temperature reconstruction from the Swiss Alps. The Holocene 13, 477-484. Huang Y., Freeman K.H., Eglinton T.I. and Street-Perrott F.A. (1999) !13C analyses of individual lignin phenols in Quaternary lake sediments: A novel proxy for deciphering past terrestrial vegetation changes. Geology 27, 471-474. Huang Y., Shuman B., Wang Y. and Webb T.III. (2002) Hydrogen isotope ratios of palmitic acid in lacustrine sediments record late Quaternary climate variations. Geology 30, 1103-1106. Huang Y., Shuman B., Wang Y. and Webb T.III. (2004) Hydrogen isotope ratios of individual lipids in lake sediments as novel tracers of climatic and environmental change: a surface sediment test. Journal of Paleolimnology 31, 363-375. Huang Y., Street-Perrott F.A., Metcalfe S.E., Brenner M., Moreland M. and Freeman K.H. (2001) Climate change as the dominant control on glacial-interglacial variations in C3 and C4 plant abundance. Science 293, 1467-1471. Krishnamurthy R.V., Syrup K.A., Baskaran M. and Long A. (1995) Late glacial climate record of midwestern United States from the hydrogen isotope ratios of lake orgnaic matter. Science 269, 1565-1567. Kurek J., Cwynar L.C. and Spear R.W. (2004) The 8200 cal yr BP cooling event in eastern North America and the utility of midge analysis for Holocene temperature 27 reconstructions. Quaternary Science Reviews 23, 627-639. Lamb A.L., Leng M.J., Mohammed M.U. and Lamb H. F. (2004) Holocene climate and vegetation change in the Main Ethiopian Rift Valley, inferred from composition (C/N and !13C) of lacustrine organic matter. Quaternary Science Reviews 23, 881-891. Liu W. and Huang Y. (2005) Compound specific D/H ratios and molecular distributions of higher plant leaf waxes as novel paleoenvironmental indicators in the Chinese Loess Plateau. Organic Geochemistry 36, 851-860. Meyers P. A. (2003) Applications of organic geochemistry to paleolimnological reconstructions: a summary of examples from the Laurentian Great Lakes. Organic Geochemistry 34, 261-289. Meyers P. A. and Ishiwatari R. (1993) Lacustrine organic geochemistry - an overview of indicators of organic matter sources and diagenesis in lake sediments. Organic Geochemistry 20, 867-900. Newby P.E., Killoran P., Waldorf M.R., Shuman B., Webb R.S. and Webb T.III. (2000) 14,000 years of sediment, vegetation, and water-level changes at the Makepeace Cedar Swamp, Southeastern Massachusetts. Quaternary Research 53, 352-368. Sauer P.E., Eglinton T.I., Hayes J.M., Schimmelmann A. and Sessions A.L. (2001) Compound-specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Cosmochimica Acta 65, 213-222. 28 Sessions A.L. and Hayes J.M. (2005) Calculation of hydrogen isotopic fractionations in biogeochemical systems. Geochimica et Cosmochimica Acta 69, 593-597. Sessions A.L., Jahnke L.L., Schimmelmann A. and Hayes J.M. (2002) Hydrogen isotope fractionation in lipids of the methane-oxidizing bacterium Methylococcus capsulatus. Geochimica et Cosmochimica Acta 66, 3955-3969. Severinghaus J.P., Sowers T., Brook E.J., Alley R.B. and Bender M.L. (1998) Timing of abrupt climate change at the end of the Younger Dryas interval from thermally fractionated gases in polar ice. Nature 391, 141-146. Shuman B., Bartlein P.J., Logar N., Newby P. and Webb T.III. (2002a) Parallel climate and vegetation responses to the early Holocene collapse of the Laurentide Ice Sheet. Quaternary Science Reviews 21, 1793-1805. Shuman B., Bravo J., Kaye J., Lynch J.A., Newby P. and Webb T.III. (2001) Late Quaternary water-level variations and vegetation history at Crooked Pond, Southeastern Massachusetts. Quaternary Research 56, 401-410. Shuman B., Newby P., Huang Y. and Webb T.III. (2004) Evidence for the close climatic control of New England vegetation history. Ecology 85, 1297-1310. Shuman B., Webb T.III., Bartlein P.J. and Williams J.W. (2002b) The Anatomy of a climatic oscillation: vegetation change in eastern North America during the Younger Dryas chronozone. Quaternary Science Reviews 21, 1777-1791. Spooner I., Douglas M.S.V. and Terrusi L. (2002) Multiproxy evidence of an early Holocene (8.2 kyr) climate oscillation in central Nova Scotia, Canada. Journal of 29 Quaternary Science 17, 639-645. Sternberg L.D.S.L. (1988) D/H ratios of environmental water recorded by D/H ratios of plant lipids. Nature 333, 59-61. Stuiver M., Grootes P.M. and Braziunas T.F. (1995) The GISP2 !18O climate record of the past 16,500 years and the role of sun, ocean, and volcanoes. Quaternary Research 44, 341-354. Yapp C.J. and Epstein S. (1982) A re-examination of cellulose carbon-bound hydrogen dD measurements and some factors affecting plant-water D/H relationships. Geochimica et Cosmochimica Acta 46, 955-965. Yu Z. and Eicher U. (1998) Abrupt climate oscillations during the last deglaciation in central North America. Science 282, 2235-2238. 30 Figure 1. Sampling sites (n=33) for lake surface sediments along north-south and east-west transects, and the locations of Blood Pond and Crooked Pond, Massachusetts. 31 Figure 2. Age model for the sediment core from Blood Pond, Massachusetts. A total of twenty one 14C dates were obtained, but six were excluded due to age inversions (14C date of a sample is younger than a sample above it). 32 Figure 3. (A) Correlation between !D values of behenic acid (BA, C22 n-acid) from surface sediments and !D values of lake water (n=33; VSMOW is Vienna standard mean ocean water) of both transects; (B) Correlation between !DBA and !Dlake water of lakes in the north-south transect. 33 Figure 4. Hydrogen isotope records of behenic acid (!DBA), palmitic acid (!DPA), fractionation factor of C28 n-acid relative to lake water ("C28-lakewater) from Blood Pond, Massachusetts. The pollen records from the same sediment core provide vegetation information. The GISP2 !18O record is shown for comparison. Shaded columns indicate the Younger Dryas chronozone (YDC) and the 8.2 ka cool event. 34 Figure 5. Comparison of isotope records from Blood Pond (Blood !DBA, Blood "C28-lake water), Crooked Pond (Crooked !DPA, Crooked "C28-lake water) and Greenland Ice Sheet isotope record (GISP2 !18O). Blood Pond and Crooked Pond data show similar trends which slightly differ from the GISP2 record, probably suggesting climate differences between New England and the Greenland summit. CHAPTER 2 Hydrogen isotopic variability in leaf waxes of plants Part 1 Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts Co-authors: William J. D’Andrea, Dana MacDonald, Yongsong Huang Published in Organic Geochemistry, 2007, v. 38, p. 977-984, doi:10.1016/j.orggeochem.2006.12.009 35 36 37 ABSTRACT Paleoclimate interpretation of the hydrogen isotope values of plant leaf waxes extracted from sediments requires a thorough understanding of the factors controlling the isotopic ratios. Existing studies have found relatively small variability in hydrogen isotope fractionation among plants of different photosynthetic pathways (C3, C4 and CAM) and between gymnosperms and angiosperms. However, there has been no systematic study so far at a single site to determine how leaf wax hydrogen isotope (D/H) ratios differ in different plant types under the same precipitation and environmental regime. Such data are nevertheless crucial for understanding the impact of past vegetation changes on the sedimentary hydrogen isotope records of leaf waxes. Here we present a study on D/H ratios of leaf waxes from 48 species in 7 types of terrestrial and aquatic C3 plants around Blood Pond, Dudley, Massachusetts. The !D values of leaf waxes differ by as much as 70 ‰ for different plant types, with those from trees and ferns having the highest values and those from grasses having the lowest values. The large isotopic variation indicates that the apparent hydrogen isotopic fractionation between leaf waxes and precipitation is not constant for different plants. Our results indicate that inferring precipitation D/H ratios based on sedimentary leaf waxes are only viable when significant vegetation change is absent or can be accounted for isotopically. 38 INTRODUCTION Compound-specific hydrogen isotope (D/H) ratios of leaf waxes from lake surface sediments across large climatic gradients have been shown to track the D/H ratios of environmental water (Huang et al., 2004; Liu and Huang, 2005) or precipitation (e.g., Sachse et al., 2004, 2006; Shuman et al., in press). Long chain n-alkanes from different sites have been suggested to track D/H ratios of precipitation at nearly constant apparent isotopic fractionation (Saches et al., 2004; 2006). Based on hydrogen isotopic analyses of both aquatic and terrestrial compounds, Hou et al. (2006) and Shuman et al. (in press) have constructed late-glacial to Holocene climate reconstructions that are consistent with existing climate data (Huang et al., 2002; Shuman et al., 2004). However, despite these corroborated paleoclimate reconstructions, the reliability with which D/H ratios of sedimentary leaf waxes record ancient precipitation depends on an important premise: the natural hydrogen isotopic variability among different plants at a given site is relatively small. If, on the other hand, different plants at a given site produce leaf waxes with large hydrogen isotopic differences, vegetation changes through time could confound or even overpower the precipitation !D signal believed to be recorded by sedimentary leaf waxes. D/H ratios of leaf waxes in plants that use different photosynthetic pathways, and in several plant types (gymnosperms, angiosperms, ferns, aquatic plants), have been studied 39 (e.g., Chikaraishi et al., 2003, 2004; Bi et al., 2005; Smith and Freeman, 2006). In contrast to the large carbon isotopic differences observed between C3 and C4 plants (Farquhar et al., 1989), hydrogen isotopic differences between C3 and C4 plants are relatively small (Chikaraishi et al., 2003; Bi et al., 2005). If only grasses are considered, C4 grasses are enriched in deuterium by ~ 20‰ relative to C3 grasses collected from the same site (Smith and Freeman, 2006). The plant samples studied by Chikaraishi et al. (2003) were collected from a wide range of ecosystems, locations, and climate zones making it difficult to ascertain the D/H ratios of plant source waters and to distinguish the effect of plant types from the effect of source water and environmental conditions (e.g., nutrients, soil types, evaporation, relative humidity, sun light etc.) on the hydrogen isotopic fractionations. At different locations, plants may utilize soil water derived from precipitation falling during different seasons, and may incorporate different amounts of ground water for photosynthesis (White et al., 1985). Therefore, determining plant apparent hydrogen isotopic fractionation relative to either mean annual or summer precipitation may introduce significant uncertainty, especially when comparing sites with large climatic, hydrological and micro-environmental differences. Here we present a systematic study of 48 species of terrestrial and aquatic plants (all are C3 plants) around Blood Pond, MA (Hou et al., 2006) to examine the difference in their leaf wax !D values. By collecting plant samples within the close vicinity (within 50 40 m) of Blood Pond, we minimize the impact of various hydrological and environmental factors on leaf wax D/H ratios. We aim to 1) systematically study the !D values of leaf waxes produced by different terrestrial and aquatic plants under the same hydrological and environmental system; 2) evaluate the influence of plant composition on the application of leaf waxes to climate reconstruction; 3) determine the biosynthetic hydrogen isotopic fractionation between !D values of long chain normal acids and alkanes. 41 SAMPLES AND METHODS Samples Blood Pond (42.08oN, 71.96oW, 212.1 m a.s.l.) is a kettle pond located in the south-central Massachusetts town of Dudley (Hou et al., 2006). We collected 74 samples of plant leaves, representing 48 species of terrestrial and aquatic plants around Blood Pond on Aug 11, 2005 (Table 1). There are 11 tree species (35 leaf samples), 9 shrub species (11 leaf samples), 3 vine species (3 leaf samples), 9 herb species (9 leaf samples), 7 grass species (7 samples), 4 fern species (4 leaf samples) and 5 species of aquatic macrophytes (5 leaf samples). All plants in this study are C3 plants. Multiple samples from different heights (6m, 4.5m, 3 m) of individual trees or shrubs (Table 1) were also collected to determine the hydrogen isotopic variability within the same plant. Methods All leaf samples were freeze-dried and then ultrasonically extracted with dichloromethane three times for 15 minutes. The extract was separated into neutral and acid fractions using solid phase extraction (Aminopropyl Bond Elute®). The acid fraction was then methylated using anhydrous 2% HCL in methanol. Hydroxyl acids were 42 removed using silica gel column chromatography (DCM as solvent), in order to further purify the fatty acid methyl esters and avoid chromatographic coelution. The neutral fraction was further separated into hydrocarbon, ketone/aldehyde, and alcohol fractions using hexane, DCM, and a mixed solvent (ethyl acetate:hexane=1:3), respectively. Quantification and identification of compounds were carried out using GC and GC-MS. An HP 6890 GC interfaced to a Finnigan Delta+ XL stable isotope mass spectrometer through a high-temperature pyrolysis reactor was used for hydrogen isotopic analysis (Huang et al., 2002, 2004). The H3+ factor was determined daily prior to sample analysis (average values 2.0 during this study). The precision (1!) of triplicate analyses was <±2‰. The accuracy was routinely checked by an injection of laboratory isotopic standards between every six measurements. "D values obtained from individual acids (as methyl esters) were corrected by mathematically removing the isotopic contributions from added groups before reporting. The "D value of the added methyl group was determined by acidifying and then methylating (along with the samples) the disodium salt of succinic acid with a predetermined "D value (using TC/EA-IRMS) (Huang et al., 2002). There is no kinetic isotopic fractionation and hydrogen exchange during the methylation (Yang and Huang, 2003). 43 RESULTS AND DISCUSSION Hydrogen isotopic variation among plant types Hydrogen isotopic values of C27, C29, C31 n-alkanes and C26, C28 and C30 n-acids of all plant samples are listed in Table 1. Because !D values are strongly correlated for C27, C29 and C31 n-alkanes (Figure 1A, Table 2) and for C26, C28, C30 n-acids (Figure 1B, Table 2), our subsequent discussions will focus on C29 n-alkane and C30 n-acid. To facilitate comparison, we also group plants into 7 types: trees, shrubs, vines, herbs, grasses, ferns and aquatic macrophytes based on their growth habits (Raven et al., 2003; Figure 2). The definitions of growth habits and their acronyms (e.g., BEPO for Betula populifolia Marsh.) used in Table 1 are obtained from the PLANTS database on the United States Department of Agriculture website (USDA, 2006). !D values of leaf waxes from the different plant types sampled at Blood Pond vary significantly (Table 1; Figure 2). Long chain n-acids and n-alkanes from trees and ferns have the highest average !D values, whereas grasses have the lowest. The hydrogen isotopic difference between ferns/trees and grasses is as large as 60 to 70‰. Shrubs, vines, herbs, and macrophytes have intermediate !D values. Within each plant type, the !D values vary among different species of plants. Trees, grasses and herbs show larger 44 isotopic variability than the other types. For example, among the trees, !D values of C30 n-acid range from -190 ‰ (CARYA-2, Carya sp. Nutt) to -121 ‰ (FRAM2(a), Fraxinus americana L., Table.1). The !D values of C30 n-acid in grasses range from -227 ‰ (PHPR3, Phleum pratense L.) to -182‰ (DAGL, Dactylis glomerata L. and JUTE, Juncus tenuis Willd.). The large variability observed for trees and grasses may partially reflect the relatively large number of samples taken from trees and grasses in this study. Factors leading to D/H variability in different plants A number of factors may affect the D/H ratios of leaf waxes. The source water used by plants may differ, even though the regional precipitation is the same. Different trees root to different depths, and may have access to water with different D/H ratios. In general, shallow soil water should have higher D/H ratios due to evaporation (Barnes and Turner, 1998). Trees, having more extensive root systems, may tap into deeper ground water than grasses and herbs. Our data do not, however, suggest that this is an important mechanism controlling the D/H ratios of leaf waxes from trees, grasses, and herbs, since grasses and herbs have on average 60 to 70 ‰ lower !D values than trees, opposite to what would be expected if source water depth exerted a primary control. The higher !D values in trees relative to grasses and herbs may result from 45 differences in the microclimate around tree crown and grasses that lead to different degrees of evapotranspiration. Grasses and herbs are more likely to be shaded by trees and shrubs in the forest. In contrast, most trees receive direct sun light and experience more windy conditions (which also facilitate more evapotranspiration). The temperature at the surface of tree leaves should be higher than that experienced by grasses due to the direct sun light on tree leaves and shading effect on grasses. The deep roots of trees also enhance transpiration (Raven et al., 2003). The relative humidity around tree leaves may also be lower than that around grasses because of the moisture from soil water evaporation. Lower humidity would act to increase the kinetic isotopic effect for trees relative to grasses. The evapotranspiration effect, which enriches deuterium in the leaf water, could thus be smaller for the grasses than for trees. The differences in evapotranspiration rate may therefore be an important factor leading to higher hydrogen isotopic ratios of the leaf waxes of trees relative to grasses. In addition, the difference in leaf morphology affects the isotopic composition of leaf water (Grice et al., 2005), which then may also contribute the !D values of leaf waxes for different plants. It is also possible that differences in plant physiology and biochemistry could contribute to the hydrogen isotopic differences. However, we currently do not have data to support this possibility. Modified Craig-Gordon isotopic models on leaf water (Roden and Ehleringer, 1999) for different plants may help to explain the isotopic difference amongst the plants. 46 Leaf samples collected from different heights of individual plants do not show significant variation (Table 1). The leaves of BELE-2, PIST, NYSY show little change in n-acid !D values with height of the sample. There is a slightly increasing trend (~9 ‰) of !D values with height for BEPO-1 and FRAM2, but the trend is reversed for BEPO-2 (same plant species). Within individual plant types, the variation of !D values of n-acids and n-alkanes may also result from evapotranspiration differences among different species, but further work is needed to test this hypothesis. Implication for reconstructing past precipitation D/H ratios The mean annual !D value of precipitation at Blood Pond (Dudley, MA) is calculated to be -63 ‰, while summer !D is -40 ‰, using the online precipitation isotope calculator (Bowen and Revenaugh, 2003). Since all our samples were collected within 50 m of Blood Pond, all plants from this study have received the same precipitation. Thus, the difference in !D values of the leaf samples (Table 1 and Figure 2) are equivalent to the differences in the apparent hydrogen isotopic fractionation relative to precipitation. The apparent hydrogen isotopic fractionation between leaf waxes and precipitation must be known in order to reconstruct past precipitation D/H ratios from leaf waxes preserved in lake sediment. Based on our data, the apparent H isotopic enrichment relative to summer precipitation for trees ranges from -156 to -85‰ for n-acids (mean = -117‰), 47 and -180 to -115‰ for n-alkanes (mean = -150‰). Grasses show much larger apparent isotopic fractionation, with n-acids ranging from -195 to -148‰ (mean = -171‰), and n-alkanes ranging from -206 to -154 ‰ (mean = -183‰). Lower average !D values for grasses and herbs relative to trees may help explain the low !D values of n-alkanes from lake sediments relative to n-alkanes from surrounding tree leaves reported by Sachse et al. (2006). Sachse et al. (2004) suggest that D/H ratios of leaf wax n-alkanes in surface sediments fractionate relative to precipitation by approximately 130 ‰. However, the results from this study suggest large differences in hydrogen isotopic fractionation of different types of plants at a single site. Leaf waxes are transported into lake sediment primarily by wind. Sedimentary leaf waxes thus integrate those produced by all terrestrial and aquatic plants. Our data indicate the apparent H isotopic enrichment can differ by as much as 60 to 70 ‰ between trees and grasses. Therefore, it is inappropriate to assume constant fractionation between sedimentary leaf waxes and precipitation over intervals where there have been significant changes in vegetation. Naturally, vegetation is often sensitive to climate change. Hence, paleoclimate interpretations based on leaf wax !D values must first consider the isotopic variation induced by changing vegetation types. In regions where vegetation is less variable, such as arid grasslands, it is possible that a relatively invariable H isotopic fractionation could be applied to sedimentary leaf wax 48 records to obtain precipitation D/H ratios. In addition, n-alkanes from sphagna may resemble the environmental water hydrogen isotopes (Xie et al., 2004), which can be used to reconstruct past precipitation variation. H isotopic difference between n-acids and n-alkanes C30 n-acid is the biosynthetic precursor of C29 n-alkane in plants. There is biosynthetic isotopic fractionation during the decarboxylation process. This fractionation for various plants has not been reported in literature. Our data (Table 1, Fig 1C, Fig 2) show that C29 n-alkanes are depleted in D by up to 30 ‰ relative to C30 n-acids. This indicates that decarboxylation discriminates against D, as expected. Grasses and herbs appear to show smaller D depletion of C29 n-alkane relative to C30 n-acid (<13 ‰) than trees, although we cannot exclude the possibility that such difference is a result of sampling bias at this time. C28 n-acid and C27 n-alkane from the same plants show similar difference in their !D values (Table 1, Fig 1D). Chikaraishi et al (2004) reported much larger hydrogen isotopic effect (up to 100 ‰) during enzymatic desaturation of fatty acids, which is apparently larger than !D change during decarboxylation. 49 CONCLUSIONS We systematically studied the hydrogen isotopes of leaf waxes from 48 species of 7 types of terrestrial and aquatic C3 plants around Blood Pond, Massachusetts. The !D values of leaf waxes show large differences among plant types. The difference between average !D values of ferns, trees and grasses is as large as 70‰. We attribute the hydrogen isotopic difference to different degrees of evapotranspiration as a result of microenvironmental differences for different plants, such as sun light exposure, temperature, humidity and turbulence (wind). Differences in plant physiology and biochemical fractionation may also be important, but we currently do not have data to demonstrate their impact on hydrogen isotopic fraction of leaf waxes. Our data indicate that caution must be taken when attempting to use sediment leaf wax D/H ratios to reconstruct past precipitation isotopic ratios. Variable fractionation among different plants must be considered to minimize the effect of vegetation changes under different climate conditions. It is possible that combining isotopic and pollen stratigraphy will allow more accurate paleoclimate interpretations based on leaf wax !D values in sediments. In regions with more monotonic plant types (e.g., arid regions, Liu and Huang, 2005), hydrogen isotopic ratios of sedimentary leaf waxes may more faithfully track precipitation D/H ratios. 50 ACKNOWLEDGEMENTS We would like to thank Koebke Farm (Dudley, MA) for access to Blood Pond. This work was supported by grants from the National Science Foundation (NSF 0318050, 0318123, 0402383) to Y. Huang. REFERENCES Barnes C.J. and Turner J.V. (1998) Isotopic exchange in soil water. In: C. Kendall, J.J. McDonnell (Eds.), Isotope tracers in catchment hydrology (Ed. by C. Kendall, J.J. McDonnell), pp. 137-163. Elsevier. Bi X., Sheng G., Liu X., Li C. and Fu J. (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36, 1405-1417. Bowen G. J. and Revenaugh J. (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39, 1299, doi:10.129/2003WR002086. Chikaraishi Y. and Naraoka H. (2003) Compound-specific dD-d13C analyses of n-alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63, 361-371. 51 Chikaraishi Y., Naraoka H. and Poulson S.R. (2004) Hydrogen and carbon isotopic fractionations of lipid biosynthesis among terrestrial (C3, C4 and CAM) and aquatic plants. Phytochemistry 65, 1369-1381. Farquhar G.D., Ehleringer J.R. and Hubick K.T. (1989) Carbon Isotope Discrimination and Photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40, 503-537. Grice K., Zhou Y., Stuart-Williams H., Wong S.C., Schouten S., Wang S.X. and Farquhar G.D. (2005) D/H of water, lipids and carbohydrates from modern plants grown under controlled conditions (Temperature, light intensity, CO2 concentration and water availability). Organic Geochemistry: Challenges for the 21st Century, 22nd IMOG, Serille, Spain., 734-735. Hou J., Huang Y., Wang Y., Shuman B., Oswald W.W., Faison E. and Foster D.R. (2006) Postglacial climate reconstruction based on compound-specific D/H ratios of fatty acids from Blood Pond, New England, Geochemistry Geophysics Geosystem 7, Q03008, doi:10.1029/2005GC001076. Huang Y., Shuman B., Wang Y. and Webb T.III. (2002) Hydrogen isotope ratios of palmitic acid in lacustrine sediments record late Quaternary climate variations. Geology 30, 1103-1106. Huang Y., Shuman B., Wang Y. and Webb T.III. (2004) Hydrogen isotope ratios of individual lipids in lake sediments as novel tracers of climatic and environmental 52 change: a surface sediment test. Journal of Paleolimnology 31, 363-375. Killops S. and Killops V. (2005) Introduction to Organic Geochemistry. Blackwell Publishing Company. Liu W. and Huang Y. (2005) Compound specific D/H ratios and molecular distributions of higher plant leaf waxes as novel paleoenvironmental indicators in the Chinese Loess Plateau. Organic Geochemistry 36, 851-860. Raven P.H., Evert R.F. and Eichhorn S.E. (2003) Biology of Plants. W.H. Freeman and Company, New York, pp. 466-494. Roden J.S. and Ehleringer J.R. (1999) Observations of hydrogen and oxygen isotopes in leaf water confirm the Craig-Gordon model under wide-ranging environmental conditions. Plant Physiology 120(4), 1165-1173. Sachse D., Radke J. and Gleixner G. (2004) Hydrogen isotope ratios of recent lacustrine sedimentary n-alkanes record modern climate variability. Geochimica et Cosmochimica Acta, 68, 4877-4889. Sachse D., Radke J. and Gleixner G. (2006) !D values of individual n-alkanes from terrestrial plants along a climatic gradient - Implications for the sedimentary biomarker record. Organic Geochemistry 37, 469-483. Shuman B., Huang Y., Newby P. and Wang Y. (2006) Compound-specific isotopic analyses track changes in the seasonality of precipitation in the Northeastern United States at ca. 8200 cal yr BP. Quaternary Science Reviews 25, 2992-3002. 53 Shuman B., Newby P., Huang Y. and Webb T.III. (2004) Evidence for the close climatic control of New England vegetation history. Ecology 85, 1297-1310. Smith F.A. and Freeman K.H. (2006) Influence of physiology and climate on !D of leaf wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70, 1172-1187. USDA NRCS. (2006) The PLANTS Database, 6 March 2006 (http://plants.usda.gov). National Plant Data Center, Baton Rouge, LA 70874-4490 USA. White J.W.C., Cook E.R., Lawrence J.R. and Broecker W.S. (1985) The D/H Ratios of Sap in Trees - Implications for Water Sources and Tree-Ring D/H Ratios. Geochimica Et Cosmochimica Acta 49, 237-246. Xie S.C., Nott C.J., Avsejs L.A., Maddy D., Chambers F.M. and Evershed R.P. (2004) Molecular and isotopic stratigraphy in an ombrotrophic mire for paleoclimate reconstruction. Geochimica Et Cosmochimica Acta 68, 2849-2862. Yang H. and Huang Y. (2003) Preservation of lipid hydrogen isotope ratios in Miocene lacustrine sediments and plant fossils at Clarkia, northern Idaho, USA. Organic Geochemistry 34, 413-423. 54 Figure 1. Correlations between !D values of leaf wax compounds. 55 Figure 2. Hydrogen isotope ratios of leaf waxes collected around Blood Pond. !D values of C29 n-alkanes (diamonds) and C30 n-acids (circles) for each plant are shown, and the respective average values are depicted as horizontal bars. FE - Ferns, TR - Trees, SH - Shrubs, VI - Vines, HB - Herbs, GR - Grasses, AQ - Aquatic Macrophytes. 56 Table 1. !D values of leaf wax compounds extracted from leaf samples Type Symbol Scientific Name n-Alkanes n-Acids C27 C29 C31 C26 C28 C30 TR BEPO-1(a) Betula populifolia Marsh. -191 -182 -177 -153 -159 -131 TR BEPO-1(b) Betula populifolia Marsh. -186 -176 -174 -155 -163 -142 TR BEPO-1(c) Betula populifolia Marsh. -198 -190 -182 -163 -167 -151 TR BEPO-2(a) Betula populifolia Marsh. -192 -178 -176 -147 -161 -135 TR BEPO-2(b) Betula populifolia Marsh. -192 -177 -180 -125 -144 -122 TR QUVE Quercus velutina Lam. -152 -167 -130 -152 -147 TR ACRU-1 Acer rubrum L. -172 -194 -180 -136 -159 -159 TR ACRU-2 Acer rubrum L. -157 -195 -198 -158 -168 -162 TR ACRU-3(a) Acer rubrum L. -206 -213 -158 -173 -172 TR ACRU-3(b) Acer rubrum L. -219 -210 -142 -171 -165 TR BELE-1(a) Betula lenta L. -182 -182 -149 -167 -158 TR BELE-1(b) Betula lenta L. -197 -178 TR BELE-1(c) Betula lenta L. -181 -181 -165 -167 -153 TR BELE-2(a) Betula lenta L. -188 -168 -144 -163 -155 TR BELE-2(b) Betula lenta L. -193 -174 -137 -160 -156 TR BELE-2(c) Betula lenta L. -204 -174 -138 -164 -153 TR CARYA-1(a) Carya sp. Nutt. -194 -172 -158 -162 -160 TR CARYA-1(b) Carya sp. Nutt. -188 -175 -144 -152 -147 TR CARYA-1(c) Carya sp. Nutt. -176 -185 -138 -144 -145 TR CARYA-2 Carya sp. Nutt. -190 -208 -203 -175 -179 -182 TR PIST (a) Pinus strobus L. -183 -192 -195 -184 -188 -181 TR PIST(b) Pinus strobus L. -173 -201 -200 -175 -154 -184 TR PIST(c) Pinus strobus L. -178 -200 -199 -182 -179 -190 TR PRSE(a) Prunus serotina Ehrh. -184 -166 -155 -153 -156 TR PRSE(b) Prunus serotina Ehrh. -201 -173 TR QURU(a) Quercus rubra L. -167 -165 -150 -157 -164 -161 TR QURU(c) Quercus rubra L. -168 -173 -157 -153 -165 -166 TR FRAM2(a) Fraxinus americana L. -185 -162 -109 -124 -121 TR FRAM2(b) Fraxinus americana L. -182 -177 -101 -122 -129 TR FRAM2(c) Fraxinus americana L. -187 -172 -115 -125 -139 TR TSCA(b) Tsuga canadensis L. -159 -160 -155 -159 -156 -158 TR TSCA(c) Tsuga canadensis L. -163 -165 -156 -126 -131 -142 TR NYSY(a) Nyssa sylvatica Marsh. -180 -124 -126 -133 TR NYSY(b) Nyssa sylvatica Marsh. -187 -126 -130 -134 TR NYSY(c) Nyssa sylvatica Marsh. -188 -114 -125 -131 SH HAVI4-1 Hamamelis virginiana L. -166 -188 -140 -155 -158 SH HAVI4-2 Hamamelis virginiana L. -179 -170 -145 -168 -152 SH HAVI4-3 Hamamelis virginiana L. -171 -168 -161 -147 -157 -160 57 SH LOTA Lonicera tatarica L. -185 -179 -141 -151 -164 SH RUAL Rubus allegheniensis -178 -187 -180 -177 -177 -158 SH VIAC Viburnum acerifolium L. -187 -134 -145 -163 SH CLAL3 Clethra alnifolia L. -197 -172 -176 -175 SH LIBE3 Lindera benzoin (L.) -189 -200 -182 -178 -198 -187 SH DEVE Decodon verticillatus (L.) Elliot -149 -179 -197 SH ILVE Ilex verticillata (L.) Gray -160 -163 -157 SH RHODO Rhododendron sp. L. -190 -190 -168 -136 -156 -172 VI SMHE Smilax herbacea L. -141 -175 -170 VI SMRO Smilax rotundifolia L. -202 -175 -185 -172 -154 VI VIAE Vitis aestivalis Michx. -179 -193 -167 -148 -155 -153 HB ASDI Aster Divaricatus L. -176 -181 -176 -177 -175 HB DACA6 Daucus carota L. -188 -170 -156 -167 -186 -184 HB PLMA2 Plantago major L. -204 -208 -210 -213 -211 HB TRPR2 Trifolium pratense L. -196 -207 -190 -176 -185 -198 HB UNKN1 Unknown -165 -181 -174 HB UNKN2 Unknown -188 -190 -186 -162 -193 HB IMCA Impatiens capensis -193 -187 -194 -194 -164 HB MIRI Mimulus ringens L. -173 -168 -144 -155 -156 HB TYLA Typha latifolia L. -195 -195 -189 -170 -164 -191 GR CAPE6 Carex pensylvanica Lam. -193 -210 -217 GR AGRO Agropyron sp. -175 -231 -228 -217 -220 -220 GR DAGL Dactylis glomerata L. -202 -218 -221 -194 -207 -182 GR JUTE Juncus tenuis Willd. -216 -217 -198 -192 -185 -182 GR PHPR3 Phleum pratense L. -232 -226 -217 -218 -226 -227 GR POAC Poaceae -206 -232 -238 -226 -220 -221 GR ALOPE Alopecurus sp. L. -171 -186 -188 -168 -179 -184 FE DEPU2 Dennstaedtia punctilobula -129 -123 -126 FE POAC4 Polystichum acrostichoides -162 -170 -165 -136 -143 -140 FE OSCI Osmunda cinnamomea L. -152 -159 -129 -147 -147 FE ONSE Onoclea sensibilis L. -156 -156 -114 -143 -135 AQ BRSC Brasenia schreberi J. F. Gmel -154 -162 -171 AQ LEMNA Lemna sp. L. -172 -200 -169 -169 -178 AQ NYMPH Nymphaea L. -161 -165 -173 AQ NYMPH Nymphaea L. -169 -173 -183 AQ POCO14 Pontederia cordata L. -188 -179 -151 -159 -165 *Type of plants: TR-Tree, SH-Shrub, VI-Vine, HB-Herb, GR-Grass, FE-Fern, AQ-Aquatic macrophytes; ** Symbols are obtained from USDA PLANTS database (http://plants.usda.gov/). The number following the symbol indicates individual plants. The letters, (a), (b), (c) indicate the height of the leaf samples above the ground (a = 6 m, b = 4.5 m, c = 3 m). For example, BEPO-1(a) indicates the leaf sample was taken from the first Betula populifolia at 6 m above the ground. 58 Part 2 Evidence for water use efficiency as an important factor in determining the !D values of tree leaf waxes Co-authors: William J. D’Andrea, Dana MacDonald, Yongsong Huang Published in Organic Geochemistry, 2007, v. 38, p. 1251-1255, doi:10.1016/j.orggeochem.2007.03.011 59 ABSTRACT D/H ratios of sedimentary leaf waxes can provide useful information about past climate changes. However, factors controlling !D values of higher plant leaf waxes (!Dwax) are poorly understood. Here we show that !Dwax are negatively correlated with !13C values of leaf waxes (!13Cwax) based on study of 35 leaf samples of 11 tree species around Blood Pond, Massachusetts. Our data suggest that plant water-use efficiency exerts an important control on the !Dwax variation among different tree species. 60 INTRODUCTION Hydrogen isotope ratios of plant leaf waxes (!Dwax) from lake and ocean sediments track !D of environmental waters and have potential for paleoclimate reconstructions (e.g., Schefu" et al., 2005; Liu and Huang, 2005; Sachse et al., 2004; Shuman et al., 2006; Pagani et al., 2006). However, values of !Dwax show considerable variation among different plants from a single site receiving the same precipitation (Hou et al., 2007). While existing data suggest plant water-use strategies may affect !D values of plant leaf waxes (Smith and Freeman, 2006; Sachse et al., 2006; Hou et al., 2007), there have been no data directly linking plant water-use efficiency to hydrogen isotope ratios of leaf waxes. Carbon isotope ratios are well-established indicators of plant water-use efficiency, which is defined as the ratio of carbon assimilation, A, to the transpiration water loss of the plants, E (Farquhar et al., 1989; Bacon, 2004), (1) where is the respired proportion of assimilated carbon; pa is the atmospheric CO2 pressure; v is the difference between intercellular and atmospheric water vapor pressure; the factor 1.6 is the ratio of diffusivities of water vapor and CO2 in air; a is the 61 fractionation occurring due to diffusion in air and b is the net fractionation caused by carboxylation; !13C is the carbon isotopic discrimination of plant material relative to ambient CO2; !13C is defined as ("13Ca – "13Cp)/(1 + "13Ca), where "13C a and "13C p represent the "13C values of atmospheric CO2 and plant materials, respectively (Farquhar et al., 1989). Carbon isotope discrimination (!13C) is relevant to WUE via the extent of stomatal conductance and capacity of CO2 fixation during photosynthesis. Smaller stomatal conductance causes smaller !13C and larger WUE, while greater stomatal conductance causes larger !13C and smaller WUE. WUE is a function of plant type and also depends on environmental factors such as humidity, sunlight exposure, and temperature. Here we compare carbon and hydrogen isotope ratios of tree leaf waxes to assess the relationship between tree water-use efficiency and "Dwax. 62 SAMPLES AND METHODS 35 leaf samples from 11 tree species were collected near the shoreline of Blood Pond, Massachusetts in August 2005 (Table 1). All leaf samples were growing below the forest canopy, receiving dappled sunlight throughout the day. Hydrogen isotope ratios of the leaf waxes from these trees have been reported in Hou et al. (2007). Carbon isotope analyses of individual leaf wax compounds were performed using a gas chromatography – combustion – isotope ratio mass spectrometer (Thermo Finnigan). An HP 6890 GC was connected to a Finnigan MAT Delta+-XL mass spectrometer via a combustion interface. Helium was used as the carrier gas, operating at constant flow mode with a rate of 1.1 mL min-1. The oven programming was 40 oC – 200 oC @ 20oC/min – 315 @ 5 oC/min and hold for 20 min. Compounds separated by GC column were converted to CO2 and H2O through a combustion furnace operated at 940 oC and loaded with CuO and Pt wires as oxidant and catalyst, respectively. A small stream of 1% O2 in helium was added right in front of reactor to maintain the oxidation capacity of CuO. Six pulses of CO2 reference gas with known !13C values were injected via interface to the isotope ratio mass spectrometer for the computation of !13C values of sample compounds. Samples were analyzed in duplicate, with standard deviation smaller than ± 0.2 ‰. !13C values obtained from individual acids (as methyl esters) were corrected by mathematically removing the isotopic contributions from added groups. The !13C values are reported relative to the VPDB standard. 63 RESULTS AND DISCUSSION !D and !13C variation in tree leaf waxes !D values of three long chain n-acids (C26, C28, C30) from different trees show an overall range of 60 to 70 ‰ variation (Table 1). C30 n-acids range from -190 to -120 ‰, C28 n-acids from -179 to -116 ‰, and C26 n-acids from -184 to -101 ‰. The leaf waxes from evergreen trees appear to have lower !D values than those from deciduous trees (e.g., the !Dwax values of the evergreen Pinus strobus L. (~ -180 to -190 ‰) are 20 to 50 ‰ lower than !Dwax values of deciduous trees [Table 1]). !13C values of the long chain n-acids (!13Cwax) range from -39.7 to -32.9‰ and show strong inter-correlation among the three n-acids (r = 0.85 for C30 and C28 n-acids, 0.79 for C28 and C26 n-acids, and 0.55 for C30 and C26 n-acids, respectively). The variation in !13C values is not a result of different biosynthetic pathways as all trees are C3 plants. !13Cwax values of different leaf samples from individual trees also show some variability. However, there is no clear correlation between !13Cwax and the height of leaf samples in this study. Hydrogen and carbon isotope ratios show significant negative correlation for the 64 three individual n-acids (Figure 3). The correlation coefficients between !Dwax and !13Cwax are R2 = 0.55 for C30 n-acids (n = 27, p<0.001), 0.40 for C28 n-acids (n = 31, p<0.001), and 0.34 for C26 n-acids (n = 26, p<0.005), respectively. WUE and !13C values of leaf waxes Published data have shown that there is a strong negative correlation between !13C and WUE in field samples. For example, Hubick et al. (1986) demonstrate such a correlation for peanuts grown under wet and dry conditions. Ismail and Hall (1992) show a strong linear relationship between WUE and !13C for different cowpea genotypes. This relationship is also observed among different tree species. For example, higher WUE for evergreen tree species is accompanied by higher "13C values (or smaller !13C) relative to deciduous trees (Garten and Taylor, 1992). The difference in WUE and "13C values between evergreen and deciduous trees was modeled by Marshall and Zhang, 1994. Their results suggest that !13C can be used as an estimate of WUE for plants from a given region where environmental variables (e.g., soil type, temperature, humidity) are relatively constant. As discussed in the introduction, WUE and !13C are negatively correlated, with !13C is negatively correlated with "13Cwax values. Therefore, "13Cwax values show positive 65 correlation with WUE, i.e. higher !13Cwax values should correspond to higher water use efficiency in trees. For example, the !13C values of leaf waxes from Pinus strobus L. are higher than those from Betula lenta L. and Nyssa sylvatica Marsh, suggesting that Pinus strobus L. has a higher water-use efficiency than Betula lenta L. and Nyssa sylvatica Marsh. Leaf waxes of evergreen trees have been shown to have higher !13C values than deciduous trees (Chikaraishi and Naraoka, 2003; Huang et al., 2006). These data are consistent with the previous observation that evergreen species tend to have higher water-use efficiency than co-occurring deciduous species (e.g., DeLucia et al., 1988; Gower and Richards, 1990; DeLucia and Schlesinger, 1991; Garten and Taylor, 1992). WUE and !Dwax The significant negative correlation between !13Cwax and !Dwax (Figure 1) in the leaf samples of this study indicate that tree WUE is likely an important control on hydrogen isotope ratios of leaf waxes. D/H ratios of plant leaf waxes are influenced by WUE in addition to !D values of precipitation. Other factors, such as temperature, relative humidity, and photosynthetic pathways may also affect !Dwax (Smith and Freeman, 2006; Sachse et al., 2006). However, since all our samples were collected within 50 m of Blood Pond’s shoreline, precipitation, temperature, humidity, soil, and source water should be very similar for all trees. The observed hydrogen isotopic 66 variation among different trees is therefore likely related to the WUE of the different trees. In this study, trees with higher WUE (as inferred from higher !13Cwax values) were found to have lower !Dwax values. This is readily understandable. Plants with higher WUE require transpiration of less water to produce the same amount of leaf waxes as plants with lower WUE. As water evaporates from the sub-stomatal cavity, the remaining leaf water becomes enriched in deuterium (Farquhar et al., 1989). Smaller transpiration rates result in less hydrogen-isotopic enrichment of leaf water, resulting in the subsequent synthesis of leaf waxes with lower !D values. For example, !D values of C30 n-acid from Betula lenta L. and Nyssa sylvatica Marsh. are -155‰ and -133‰, respectively. The !D of C30 n-acid from Pinus strobus L. is around -185‰. The higher !D values indicate that Betula and Nyssa have lower WUE than Pinus, consistent with higher !13C values observed in Pinus. Chikaraishi and Naraoka (2006) measured both !D and !13C of plant leaf waxes, but did not observe the relationship we report here. The samples used in Chikaraishi and Naraoka (2006), however, were collected from multiple sites in Japan and in Thailand. Environmental factors such as temperature, humidity, and soil type, at the different sites may affect the carbon and hydrogen isotope ratios of leaf waxes in complex ways, 67 thereby complicating comparisons between !Dwax and !13Cwax. Contrary to our findings, Bi et al. (2005) show a positive correlation between weighted mean !Dwax and !13Cwax values for 9 tree samples from a well maintained botanic garden in China. However, the relationship is defined by one data point relative to a cluster of 8 points. Plants in the botanic garden are grown under varying degrees of artificial irrigation, fertilization and other soil treatment (e.g., adjustment of pH), which could also complicate the relationship between !Dwax and !13Cwax. CONCLUSIONS !13C and !D values of leaf waxes from 35 leaf samples of 11 tree species from Blood Pond, MA show a statistically significant negative correlation, suggesting that water-use efficiency exerts an important control on !D values of leaf waxes in trees. Plants with higher WUE display lower !Dwax values. The correlation is not perfect, reflecting the fact that other independent factors also affect !Dwax. Nevertheless, our study represents the first step to decipher the complex mechanisms controlling the variations of leaf wax !D values in terrestrial plants. 68 ACKNOWLEDGEMENTS We would like to thank Koebke Family (Koebke Farm, Dudley, MA) for access to Blood Pond. We thank Jonathan Nichols for field assistance. We are grateful to Dr. Stefan Schouten and Dr. Enno Schefuss for their constructive comments. This work was supported by grants from the National Science Foundation (NSF 0318050, 0318123, 0402383) to Y. Huang. 69 REFERENCES Bacon M.A. (2004) Water-use efficiency in plant biology. In: M.A. Bacon (Ed.), Water-use efficiency in plant biology. Blackwell, pp. 1-26. Bi X., Sheng G., Liu X., Li C. and Fu J. (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36, 1405-1417. Chikaraishi Y. and Naraoka H. (2003) Compound-specific !D-!13C analyses of n-alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63, 361-371. Chikaraishi K. and Naraoka H. (2006) !13C and !D relationships among the three n-alkyl compound classes (n-alkanoic acid, n-alkane and n-alkanol) of terrestrial higher plants, Organic Geochemistry, doi:10.1016ij.orggeochem.2006.10.003. Collister J.W., Rieley G., Stern B., Eglinton G. and Fry B. (1994) Compound-Specific !13C Analyses of Leaf Lipids from Plants with Differing Carbon-Dioxide Metabolisms. Organic Geochemistry 21(6-7), 619-627. DeLucia E.H., Schlesinger W.H. (1991) Resource-use efficiency and drought tolerance in adjacent great basin and Sierran plants. Ecology 72, 51-58. DeLucia E.H., Schlesinger W.H. and Billings W.D. (1988) Water relations and the maintenance of sierran conifers on hydrothermally altered rock. Ecology 69, 303-311. 70 Farquhar G.D., Hubick K.T., Condon A.G. and Richards R.A. (1989) Carbon isotope fractionation and plant water-use efficiency. In: Rundel P.W., Ehleringer J.R., Nagy K.A. (Eds.), Stable isotopes in ecological research. Springer-Verlag, New York, pp. 21-40. Garten C.T. and Taylor G.E. (1992) Foliar delta C-13 within a temperature decisuous forest – spatial, temporal, and species sources of variation. Oecologia 90, 1-7. Gower S.T. and Richards J.H. (1990) Larches: deciduous conifers in an evergreen world. BioScience 40, 818-826. Hou J., D'Andrea W., MacDonald D. and Huang Y. (2007) Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts (USA). Organic Geochemistry 38, 977-984 doi:10.1016/j.orggeochem.2006.12.009. Huang Y., Shuman B., Wang Y., Webb T., Grimm E.C. and Jacobson G.L. (2006) Climatic and environmental controls on the variation of C3 and C4 plant abundances in central Florida for the past 62,000 years. Palaeogeography, Palaeoclimatology, Palaecology 237, 428-435. Liu W. and Huang Y. (2005) Compound specific D/H ratios and molecular distributions of higher plant leaf waxes as novel paleoenvironmental indicators in the Chinese Loess Plateau. Organic Geochemistry 36, 851-860. Pagani M., Pedentchouk N., Huber M., Sluijs A., Schouten S., Brinkhuis H., Damste 71 J.S.S., Dickens G.R. and Scientists E. (2006) Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum. Nature 442, 671-675. Sachse D., Radke J. and Gleixner G. (2006) !D values of individual n-alkanes from terrestrial plants along a climatic gradient - Implications for the sedimentary biomarker record. Organic Geochemistry 37, 469-483. Saurer M., Siegwolf R.T.W. and Schweingruber F.H. (2004) Carbon isotope discrimination indicates improving water-use efficiency of trees in northern Eurasia over the last 100 years. Global Change Biology 10, 2109-2120. Shuman B., Huang Y., Newby P. and Wang, Y. (2006) Compound-Specific Isotopic Analyses Track Changes in the Seasonality of Precipitation in the Northeastern United States at ca. 8200 cal yr BP. Quaternary Science Reviews 25, 2992-3002. Smith F.A. and Freeman K.H. (2006) Influence of physiology and climate on !D of leaf wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70, 1172-1187. 72 Figure 3. Correlations between !13C and !D values of leaf wax compounds (C30, C28, and C26 n-acids) from natural trees around Blood Pond, Massachusetts. 73 Table 1. !13C and !D of leaf wax compounds (n-acids) from trees around Blood Pond, Massachusetts Height !D* (‰, VSMOW) !13C (‰, VPDB) Scientific Name (m) C26 C28 C30 C26 C28 C30 Betula populifolia Marsh. (1) 6 -153 -159 -131 -33.6 -34.9 -36.2 Betula populifolia Marsh. (1) 4.5 -155 -163 -142 -34.3 -35.1 -37.4 Betula populifolia Marsh. (1) 3 -163 -167 -151 -35.7 -36.6 Betula populifolia Marsh. (2) 6 -147 -161 -135 -34.0 -35.1 -38.7 Betula populifolia Marsh. (2) 4.5 -125 -144 -122 -34.8 -35.7 -39.3 Betula lenta L. (1) 6 -144 -163 -155 -36.1 -34.6 -36.4 Betula lenta L. (1) 4.5 -137 -160 -156 -35.2 -33.7 -35.6 Betula lenta L. (1) 3 -138 -164 -153 -36.2 -35.4 -37.3 Betula lenta L. (2) 6 -149 -167 -158 -33.8 -33.1 -35.1 Betula lenta L. (2) 4.5 -34.2 -33.7 -34.4 Betula lenta L. (2) 3 -165 -167 -153 -33.4 -33.3 -35.3 Quercus velutina Lam. 3 -130 -152 -147 Quercus rubra L. (1) 6 -157 -164 -161 -33.9 -34.1 -34.0 Quercus rubra L. (1) 4.5 -153 -165 -166 -34.1 -34.0 -34.5 Acer rubrum L. (1) 3 -136 -159 -159 -34.0 -34.8 -36.9 Acer rubrum L. (2) 3 -148 -168 -162 -32.9 -35.5 Acer rubrum L. (3) 6 -158 -173 -172 -33.4 -34.3 -35.8 Acer rubrum L. (3) 4.5 -142 -171 -165 -34.8 -34.3 -36.2 Carya sp. Nutt. (1) 6 -158 -162 -160 -33.2 -34.9 Carya sp. Nutt. (1) 4.5 -144 -152 -147 -33.4 -34.4 Carya sp. Nutt. (1) 3 -138 -144 -145 -33.1 -33.3 -34.6 Carya sp. Nutt. (2) 3 -175 -179 -182 Pinus strobus L. (1) 6 -184 -188 -181 -33.0 -33.4 Pinus strobus L. (1) 4.5 -175 -154 -184 -35.9 -32.9 -33.4 Pinus strobus L. (1) 3 -182 -179 -190 -32.8 -33.7 -34.5 Prunus serotina Ehrh. (1) 6 -34.4 -35.2 -36.8 Prunus serotina Ehrh. (1) 4.5 -155 -153 -156 -34.8 -35.6 -37.0 Fraxinus americana L. (1) 6 -109 -124 -121 -35.3 -35.3 Fraxinus americana L. (1) 4.5 -101 -122 -129 -37.3 -36.8 Fraxinus americana L. (1) 3 -115 -125 -139 -38.0 -35.9 -35.8 Tsuga canadensis L. (1) 4.5 -159 -156 -158 -32.8 -33.1 74 Tsuga canadensis L. (1) 3 -126 -131 -142 -33.7 -33.5 Nyssa sylvatica Marsh. (1) 6 -124 -126 -133 -37.3 -38.0 -39.3 Nyssa sylvatica Marsh. (1) 4.5 -126 -130 -134 -37.8 -38.7 -39.7 Nyssa sylvatica Marsh. (1) 3 -114 -125 -131 -38.2 -38.1 -39.1 *!D data first published in Hou et al. (2007) 75 Table 2. The correlation coefficients between !D values of individual leaf waxes Compounds C26 n-acid C28 n-acid C30 n-acid C27 n-alkane C29 n-alkane C31 n-alkane C26 n-acid 1 C28 n-acid 0.91 1 C30 n-acid 0.85 0.86 1 C27 n-alkane 0.54 0.66 0.43 1 C29 n-alkane 0.63 0.60 0.64 0.61 1 C31 n-alkane 0.64 0.67 0.64 0.60 0.87 1 76 CHAPTER 3 Can Sedimentary Leaf Waxes record D/H ratios of Continental Precipitation? Field, Model, and Experimental Assessments Co-authors: William D’Andrea, Yongsong Huang Published in Geochimica et Cosmochimica Acta, 2008, v. 72, p. 3503-3517 doi:10.1016/j.gca.2008.04.030 77 78 79 ABSTRACT D/H ratios of leaf waxes (!Dwax) derived from terrestrial plants and preserved in lake sediments can provide important information on past continental hydrology. Ideally, !Dwax can be used to reconstruct precipitation D/H ratios (!DP) which is a well-established paleoclimate proxy. However, many other factors, such as vegetation and relative humidity (RH), also affect !Dwax variation. How the combination of these factors affects sedimentary !Dwax is unclear. Here, we use a transect of 32 lake surface sediments across large gradients of precipitation, relative humidity, and vegetation in the southwestern United States to study the natural factors affecting sedimentary !Dwax. !D values of C28 n-alkanoic acids show significant correlation with !DP values (R2 = 0.76) with an apparent isotopic enrichment of ~ 99 ± 8 ‰, indicating that sedimentary !Dwax values track overall !DP variation along the entire transect. Leaf waxes produced by plants grown under controlled conditions (RH = 80, 60, 40%) show a small increase in D/H ratios as RH decreases, consistent with prediction from the Craig-Gordon model. However, the isotopic effect of RH on !Dwax along the entire transect is partially countered by the opposing influence of vegetation changes. The correlation between !Dwax and !DP values is significantly higher (R2 = 0.84) in the drier portions of the transect than in the wetter regions (R2 = 0.64). This study suggests that D/H ratios of sedimentary 80 leaf waxes can be used as a proxy for precipitation !D variations, with particularly high fidelity in dry regions, although more studies in other regions will be important to further test this proxy. 81 INTRODUCTION Hydrogen and oxygen isotope ratios (D/H and 18O/16O) of precipitation are among the most effective and quantitative proxies for past continental climate variability. Ice cores (e.g., Grootes et al., 1993; Petit et al., 1999) and speleothems (e.g., Wang et al., 2001; Yuan et al., 2004) are the most common archives for reconstructing precipitation isotopic ratios. While the fidelity of such reconstructions is exceptional, the global distribution of ice cores and speleothems is very limited; speleothems to regions with carbonate bedrock and adequate precipitation; and ice cores to polar and/or high altitude locations. Continental climate is characterized by large regional variability, requiring records from widely distributed sites to assess the underlying mechanisms of variation. The wide geographic distribution of lakes provides excellent opportunities to reconstruct the isotope ratios of past continental precipitation from sediment cores. D/H ratios of aquatic lipids in lake sediments have been shown to track D/H ratios of lake water and precipitation in regions with high precipitation/evaporation (P/E) ratios (Sauer et al., 2001; Huang et al., 2002, 2004; Hou et al., 2006). However, in dry and warm regions, D/H ratios of lake water are strongly affected by evaporation, complicating climatic interpretation of !D records. Temporal variability in D/H ratios of leaf waxes (!Dwax) derived from terrestrial plants and preserved in sediments has been 82 attributed to past precipitation and hydrological changes (Liu and Huang, 2005; Schefu! et al., 2005; Pagani et al., 2006; Shuman et al., 2006; Huang et al., 2007). However, in addition to "D of precipitation ("DP), other factors, such as relative humidity (RH) and vegetation composition, may also affect sedimentary "Dwax. Without a full evaluation of these factors, the reliability and limitations of leaf wax "D as a paleohydrological proxy remains uncertain. Sachse et al. (2004) showed that leaf wax D/H ratios ("Dwax) from lake surface sediments track "DP variation along a 13-lake transect in Europe. However, the European sites used in the study have nearly constant relative humidity (RH = 75-85 %), and relatively invariable, forest-dominated vegetation, and therefore cannot be used to assess whether "Dwax tracks "DP across relative humidity gradients and/or different biomes! While we know that RH has a significant effect on the hydrogen isotopic fractionation of water (Gonfiantini, 1986; Gat, 1996), the effect of RH on the "D values of leaf waxes has not been studied directly. Previous studies reported that the D/H ratios of cellulose increase dramatically as RH decreases (Yapp and Epstein, 1982; Edwards and Fritz, 1986), but other studies (Gray and Song, 1984; Ramesh et al., 1986; White et al., 1994; Terwilliger and DeNiro, 1995) contradict the claim. Most published studies (e.g., Sauer et al., 2001; Liu and Huang, 2005; Sachse et al. 2006; Pagani et al., 2006) have assumed that "Dwax values are also affected by RH in addition to hydrological factors. Furthermore, 83 recent studies have shown that hydrogen isotope enrichment between precipitation and leaf waxes (!wax-p) differs greatly among different types of higher plants (Liu and Huang, 2005; Liu et; 2006; Hou et al., 2007a; Liu and Huang, 2008), with grasses displaying ~40-50‰ lower "D values than trees. Therefore, in order to confidently reconstruct the variability of precipitation "D using leaf waxes, we must determine how RH and vegetation changes affect the relationship between "Dwax and "DP values. In this study, we aim to evaluate the degree to which "Dwax values in lake sediments reflect "DP variation across large, natural environmental gradients of RH, precipitation and vegetation types. If it is to be used as a proxy for reconstructing past changes in "DP, "Dwax must first be shown to track modern precipitation "D values across sites with highly variable RH and vegetation composition. To achieve this goal, we examined lake surface sediments from a 32-lake transect across large gradients in precipitation, RH, and vegetation cover (from Phoenix, AZ to Houston, TX). We also grew plants under controlled conditions to accurately assess the effects of RH and modeled the isotopic effects under both experimental and field conditions. Our integrated field, experimental, and modeling approaches allow us to objectively assess the fidelity of leaf wax "D values as a proxy for "D variation of past continental precipitation. 84 SAMPLES AND METHODS The southwestern transect Surface sediment and lake water samples were collected from 32 lakes spanning large precipitation and relative humidity gradients (extending from Phoenix, Arizona to Houston, Texas) in May 2004 (Figure 1; Table 1). For the purpose of this study (and based on our isotope data described below), the sampling sites along the transect are divided into two general groups according to their geographic setting and mean annual precipitation: 1) the Interior Plains (Texas and northeastern New Mexico, referred to hereafter as “the Plains”), eastward of 104.5 oW, with precipitation >400 mm, and 2) Southern Rocky Mountains and Basin and Range (Arizona and part of New Mexico, hereafter referred to as “Basin and Range”) westward of 104.5 oW, with precipitation <400 mm. The lakes from the Plains are mostly naturally occurring, whereas the sampling sites from the Basin and Range represent reservoirs constructed by damming rivers and streams sourced from adjacent higher elevations. There is a large mean annual precipitation gradient (120 to 1150 mm) and mean annual relative humidity gradient (36 to 82 %) along the transect from Phoenix to Houston (Figure 1A, 1B). Vegetation composition changes dramatically along the 85 transect, from shrubland to grassland to savanna to subtropical forest (Figure 1C). Mean annual temperature along the transect, excluding the highest elevation sites, ranges from 12 to 21oC (Table 1). Three of the lakes (Upper Glacial Lake, Lake #1 and Mary’s Lake) are at much higher elevation and have lower mean annual temperature (Table 1) than the rest. Surface sediments (0-2 cm below the sediment-water interface) were collected from the deepest part of each lake (in reservoirs, deepest sites are typically near the dam site, far from river inlets) using an Ekman Dredge, and were kept cold and in the dark until being frozen at Brown University. Duplicate lake water samples for isotopic analysis were collected using 4 ml glass vials. The vials were filled to capacity and sealed to prevent evaporation. Growth chamber experiments We grew 15 trees (representing 6 species) and 11 grasses (representing 4 species) in an EGC® GC series temperature and humidity-controlled growth chamber. Fraxinus americana L. (white ash), Carya glabra (pignut hickory), Acer rubrum L. (red maple) and Quercus rubra L. (red oak) were collected from the watershed of Blood Pond, Massachusetts. Two coniferous tree species, Thuja Occidentalis (eastern white cedar) and 86 Picea Glauca (alberta spruce) were obtained from a local nursery. Tree seedlings were 30-50 cm in height before transplantation. The trees were transplanted into 8-inch pots using Canadian Growing Mix 2 (Fafard®). Grass seeds, including Zea Mays L. (corn), Dactylis glomerata L. (orchard grass), Alopecurus spp. (foxtail) and Phleum pratense L. (timothy weed), were obtained from a local grass seed distributor. The trees were kept in the greenhouse for 20 days and were irrigated with water of known isotopic composition twice daily before being moved into the growth chamber. The grasses were germinated and cultured in the greenhouse for 20 days before being moved to the growth chamber. The temperature in the growth chamber was kept at 20 oC. RH was set to 80% during the first growing period (Jun 26 - Jul 27, 2006), 60% for the second growing period (Jul 28 - Sep 4, 2006), and 40% for the third growing period (Sep 5 - Oct 11, 2006). A dehumidifier was placed in the growth chamber for experiments with 60 % and 40 % RH settings, in order to assist the system to maintain stable RH levels. On the last day of each growing period, newly grown leaves from trees were collected for !Dwax analysis. There was no new leaf growth on the Carya glabra specimen during the experiment, so we collected preexisting leaves from Carya glabra on the last day of each growing period. For grass specimens, we collected whole leaves on the last day of each growing period. All plants were irrigated with 500 ml of water of known hydrogen isotopic ratio (!D = -49 ‰) every two days. The irrigation water was stored in Reliance 87 Fold-A-Carrier® collapsible containers at 4oC prior to the experiment to ensure the isotope ratios remained constant. Analyses of water !D values D/H ratios of lake water samples and irrigation water were measured using Thermal Conversion/Elemental Analyzer-Isotopic Ratio monitoring Mass Spectrometry (TC/EA-IRMS). Nine 0.2 µl aliquots of each water sample were manually injected into the TC/EA. Standard Mean Ocean Water (SMOW, !D = 0 ‰), Greenland Ice Sheet Precipitation (GISP, !D = -190‰) and Standard Light Antarctic Precipitation (SLAP, !D = -428 ‰) were measured between every 10 sample analyses for isotope calibration (Huang et al., 2002). A regression line (R2 = 1) was established between measured and actual !D values of the three isotope standards, and was used for instrument calibration. The standard deviation for water !D values was < ±1.5‰. Analyses of sediment and plant samples All sediment and leaf samples were freeze-dried. Sediment samples were extracted using an Accelerated Solvent Extractor (ASE200, Dionex) with dichloromethane:methanol (2:1 v/v) at 150oC and 1200 psi for three 15-min cycles. 88 Approximately 5 g of each dry sample were extracted. Leaves were soaked in dichloromethane:methanol (2:1 v/v) and ultrasonicated for 15 minutes to extract lipids. This extraction procedure was repeated three times. The total lipid extract obtained from sediment and plant leaves was separated into neutral and acid fractions using Supelco® Supelclean™ LC-NH2 SPE column and dichloromethane:isopropyl alcohol (2:1 v/v) and ether with 4% acetic acid (v/v), respectively, as eluents. Acid fractions were methylated with 5% anhydrous HCl in methanol at 60 oC for 12 hours. Hydroxyl acids were removed by eluting the methylated samples through silica gel columns with hexane, and the fatty acid methyl esters were collected by elution with dichloromethane. Quantification and identification of compounds was performed by Gas Chromatography - Flame Ionization Detection (GC-FID) and Gas Chromatography-Mass Spectrometry (GC-MS) (HP 6890, Agilent). An HP 6890 GC interfaced to a Finnigan DeltaPlus XL stable isotope spectrometer through a high-temperature pyrolysis reactor was used for hydrogen isotopic analysis (Huang et al., 2004). The compounds separated by the GC were pyrolyzed to H2 and CO at 1445oC. A tank of ultra high purity hydrogen gas with known !D values was used as the isotope standard during the measurements. The H3+ factor was determined daily prior to sample analysis (average value was 2.5 ± 0.04 (1 sigma) during the course of this study). The accuracy for the instrument was routinely checked by an injection of laboratory isotopic standards (C16, C18, C22, C24 89 n-alkanoic acid methyl esters) between every six measurements. The precision (1!) for the four laboratory standards were < ±2‰ throughout the entire process. The precision (1!) for triplicate analyses of all samples was < ±2‰. "D values obtained from individual alkanoic acids (as methyl esters) were corrected by mathematically removing the isotopic contributions from added groups before reporting. The "D value of the added methyl group was determined by acidifying and then methylating (along with the samples) the disodium salt of succinic acid with a predetermined "D value (using TC/EA-IRMS) (Huang et al., 2002). 90 RESULTS Precipitation and lake water !D The D/H ratio of precipitation (!Dp, weighted average values based on monthly precipitation amount) at the location of each lake along the SW transect was calculated from the Online Isotopes in Precipitation Calculator (OIPC, Bowen and Revenaugh, 2003). The OIPC estimates the !DP at a given site by combining an empirical model for isotopic trends related to latitude and altitude with detrended interpolation based on the isotope data from stations in Global Network of Isotopes in Precipitation (Bowen and Revenaugh, 2003). !DP and lake water !D values (!Dlake) show similar variation along the transect (Figure 2; Table 1). Higher !DP and !Dlake values occur in the eastern Plains and western Basin and Range. Lower !DP and !Dlake values are found in the central part of the transect (Southern Rocky Mountains). Lake water samples from the Plains show greater deuterium enrichment relative to precipitation than the lakes from Basin and Range, which have !Dlake values that are very similar to that of precipitation (Figure 2). !D of sedimentary leaf waxes We focus on C26, C28, and C30 n-alkanoic acids, in the lake surface sediment 91 samples as representative compounds for modern leaf waxes. We avoided use of long chain n-alkanes as terrestrial plant biomarkers to minimize the possibility of contamination from fossil fuel products (most lakes in the region are heavily used by recreational boaters and likely receive atmospheric input of combustion products). The hydrogen isotope ratios of the C26, C28, and C30 n-alkanoic acids are strongly inter-correlated (R2 = 1.0 for C28 and C30, R2 = 0.82 for C28 and C26; n = 32, p<0.001). Therefore, in the following discussion !D values of C28 n-alkanoic acid (hereafter referred to as !Dwax) are taken as representative of integrated !D values of terrestrial leaf waxes from plants around the lakes. Sedimentary !Dwax values show significant correlation with !DP values along the entire transect (R2 = 0.76, y = 0.93x – 99.1, n = 32, p<0.001, residual mean squares (RMS) = 122.3; Figure 2, Figure 3A). However, !Dwax and !Dp correlate much more strongly within the Basin and Range (R2 = 0.84, y = 1.69x – 42, n = 18, p<0.001, RMS = 70.2) than they do for the entire transect (R2 = 0.76) or the Great Plains lakes (R2 = 0.64, y = 0.67x – 109.7, n = 14, p<0.001, RMS = 60.6; Figure 3B). Furthermore, a second-order polynomial regression through the entire transect provides an overall better fit for the !Dwax and !DP data (R2 = 0.86, y = - 0.02x2 - 1.06x -147.4, n = 32, p<0.001, RMS = 73.0, Figure 3A) than does the linear regression. The correlation between !Dwax and !Dlake (R2 = 0.46, n = 32, p <0.001, RMS = 280.3) is not as strong as that between !Dwax and !DP. 92 Leaf wax !D for plants from growth chamber !D values of C28 n-alkanoic acids are strongly correlated with C30 n-alkanoic acids for the growth chamber samples (R2 = 0.86). As with the transect samples, !D values of C26 n-alkanoic acids show weaker correlation with C28 and C30 n-acids (R2= 0.61 and 0.44, respectively). This observation suggests that C28 or C30 n-alkanoic acids may be better choices for representative leaf wax compounds in sediment samples. It should also be noted that for certain growth chamber species, !D values of the three n-alkanoic acids show relatively large variation within and among specimens (Table 2), highlighting the natural variability of individual plants. The calculated leaf wax hydrogen isotopic enrichment from source water ("wax-water) for the growth chamber samples show an approximately normal distribution (Figure 4). The average "wax-water values of all plants grown at 80%, 60% and 40% RH are -114 ‰, -111.5 ‰, and -107 ‰, respectively, depicting minor deuterium enrichment as RH decreases. 93 DISCUSSION Controls on !D of precipitation and lake water The patterns of !DP variation along the transect in the southwestern United States result from differing water vapor sources and the continental isotope effect (Gat, 1996). The water vapor for precipitation in the Texas Plains mainly originates from the Gulf of Mexico, while the precipitation in the Basin and Range mainly derives from the eastern Pacific Ocean, off coastal Mexico (Figure 1A; Adams and Comrie, 1997). Due to Rayleigh distillation, as water vapor moves inland the heavier isotopes are preferentially precipitated, resulting in isotopically lighter precipitation with increasing distance from the source. This helps explain the observation that higher !DP values are found at either end of the transect, and lower !DP values are found in the central portion of the transect. The lowest !DP values are found at Mary’s Lake, ~ 105 oW (Figure 2). This site is the farthest from both vapor sources among all sites and at high elevation (2155 m) (Table 1). The !Dlake values along the transect are controlled not only by !DP, but also by the degree of evaporation and local hydrological conditions. Evaporation causes lake water to become isotopically enriched in D and 18O, relative to precipitation (Gat, 1996). The evaporative effect explains the observation that !Dlake values are higher than !DP values 94 within the Great Plains. However, the lake waters from the Basin and Range do not show isotopic enrichment with respect to precipitation (!D values of lakes fall within 95% confidence intervals of estimated !D values for local precipitation; Figure 2). This could be due to the fact that all Basin and Range lakes in this study are reservoirs, representing dammed rivers that originate in adjacent mountains (Table 1). The rivers are fed by precipitation within their catchments, including melt water from wintertime precipitation (e.g., snow) in the mountains. The water delivered to the reservoirs by river input is thus depleted in deuterium relative to precipitation falling directly onto the reservoirs (i.e., our estimated !DP). Subsequent evaporative enrichment in the reservoirs might counter the influence of lower !D values of the winter snow precipitation, and help explain the similarity between !Dlake and the estimated !DP in the Basin and Range (Figure 2). As a further test of our explanation for lake water !D variations described above, we used the Craig-Gordon model to calculate the isotope ratios of lake water at steady state (Craig and Gordon, 1965; Gat, 1996), by assuming that all lake waters had the initial !D values of the precipitation prior to evaporative enrichment. (1) 95 (2) where P, Q and E represent the amount of precipitation, outflow, and evaporation, respectively, and !P, !Q, and !E are their associated hydrogen isotope ratios. !L is the D/H ratio of lake water (note !L = !Q). We adopted the modified Craig-Gordon model described by Gibson and Edwards (2002), (3) Where "* is the equilibrium isotopic enrichment factor ["* = 1158.8(T3/109) - 1620.1(T2/106) + 794.84(T/103) - 161.04 + 2.9992(109/T3), T is temperature (K) (Gibson and Edwards, 2002)]. #* is the equilibrium isotopic fractionation factor, #* = (1 + 10-3"*). "K is the kinetic isotopic enrichment factor ["K = 12.5 $ (1-RH)]. !A is the D/H ratio of atmospheric water vapor, estimated assuming equilibrium with flux-weighted precipitation (!A = #*!P – "*; Gibson and Edwards, 2002). RH is relative humidity, obtained from the NOAA Climate Diagnostic Center (http://www.cdc.noaa.gov/) for individual study sites. Mean annual precipitation and pan evaporation data are from the State Climatological Summaries (http://nndc.noaa.govgetsiteref) for each site (Table 1). Pan evaporation data for lakes in Texas were also obtained from Texas 96 Evaporation/Precipitation website (http://hyper20.twdb.state.tx.us/Evaporation/evap.html), and evaporation from lake surfaces has been estimated as 70% of that measured from evaporation pan (e.g., Gonfiantini, 1986; Gibson and Edwards, 1996; Gibson et al., 1998; Smith and Freeman, 2006). Rearranging Eqs. (1), (2) and (3) yields (4) where, . We can obtain modeled D/H ratios of lake water (!Dmodel = !L) from Eq.4. !Dmodel for all the lakes along the SW transect are plotted in Figure 2. For lakes in the Plains, the !Dmodel values are similar to the measured !Dlake values, consistent with our explanation that these lakes are largely recharged by local precipitation. However, for most lakes/reservoirs in Basin and Range, the !Dmodel values are markedly higher than the measured !Dlake values. The data are again consistent with our explanation that the Basin and Range reservoirs are mainly recharged by rivers rather than by local precipitation. The !D values of river water (sourced at colder temperatures and higher elevations than the reservoirs they recharge) are expected to be lower than the estimated 97 local !DP, resulting in the correspondingly lower !D values for the observed lake waters than the modeled lake waters (the local !DP is used to represent the initial or un-evaporated lake water in the model calculation) (Figure 2). Sedimentary !Dwax as a proxy for !DP: insights from the natural transect Sedimentary !Dwax values are strongly correlated (R2 = 0.76) with mean annual !DP values along the entire transect (Figure 3A). Similar correlation (R2 = 0.77) is observed between !Dwax values and mean summer (April-October) !DP values calculated from OIPC. Water is the ultimate source for hydrogen in plant tissues, and organic matter derived from terrestrial plants should record the isotopic ratio of soil water, which is ultimately recharged by precipitation. The !D values of sedimentary leaf waxes could thus be expected to correlate with !DP at a large scale. However, there are a number of other factors that could significantly impact the relative strength of this correlation. Two prominent factors are: 1) changes in evaporative isotopic enrichment at different RH due to evaporation from soil and leaf surfaces, and 2) variations in the apparent hydrogen fractionation among different vegetation types. Both factors could potentially impact !Dwax so strongly that reconstruction of !DP from !Dwax would become too complicated and inaccurate for paleoclimate applications. 98 Surprisingly, however, even with the dramatic changes in RH and vegetation type along the lake surface sediment transect, the average apparent isotopic enrichments between local precipitation and sedimentary leaf waxes (!wax-p) do not differ significantly over the entire transect (-98.8 ± 7.8 ‰, Figure 5). This finding would suggest that the combined impact of environmental factors (e.g. vegetation cover, RH) is relatively small. Moreover, the stronger correlation between "Dwax and "DP in dry regions (R2 = 0.84) relative to the entire transect or the more humid regions (Figure 3B) may imply that "Dwax tracks "DP variation with greater fidelity in dry regions. We will provide a detailed interpretation for this observation in Sections 4.3 and 4.4 after presenting data from growth chamber and model calculations. Three lakes of the transect (Animas Ditch (AD), Upper Glacial Lake (UGL), and Santa Rosa Lake (SRL)) stand out as having significantly different !wax-p (Figure 5). These exceptions can be understood in the context of their local geographical and hydrological settings. Animas Ditch is located adjacent to a major highway, within an extensive playa (dry lake bed). "D values measured for Animas Ditch ("Dlake = 4 ‰), a small remnant pond, are much higher than other lakes in Arizona ("Dlake ~ -40 to 110 ‰), indicating the influence of very strong evaporation. Therefore, exceptionally strong evaporation of soil water at Animas Ditch could explain the reduced !wax-p. Upper Glacial Lake receives water from nearby glaciers, and both precipitation and lake water show low 99 !D values (-96 and -94 ‰, respectively). Glacial melt water and snowmelt in the catchment of this high-elevation lake may represent a large proportion of the soil water recharge. Plants utilizing this soil water would produce leaf waxes with correspondingly low !D values. Santa Rosa Lake is a recently constructed man-made reservoir, completed in 1981. Disturbance associated with construction activities may complicate the !Dwax signal of surface sediment. Reservoirs are dammed rivers. It is therefore possible that a small fraction of the leaf waxes in sediments might have been contributed from fluvial input. However, we collected all of our sediment samples from the deepest part of the reservoirs (dammed side), minimizing the fluvial components. Therefore, the leaf waxes extracted from the lake sediments in this study should reflect primarily the local vegetation inputs. The effect of RH on !wax-p • RH and !wax-p variation along the transect The apparent isotopic enrichment between local precipitation and sedimentary leaf waxes ("wax-p) is plotted against RH in Figure 5A. There appears to be a small decrease in absolute "wax-p values as RH decreases, although the correlation between RH and "wax-p is relatively weak (R2 = 0.28). To better understand whether soil evaporation or plant 100 transpiration causes the RH effect on !wax-p, we estimated the D/H ratios of soil water ("Dsoil) at all sites using the modified Craig-Gordon model. We then took the fraction of evaporative water into consideration and adopted a binary isotope model (Smith and Freeman, 2006): "Dsoil = (1-fe) # "DP + fe # "Dmodel (5) where "Dmodel (the modeled "D ratios of soil water subjected to evaporation) is calculated from Eq.4; fe is the fraction of soil water subjected to evaporation in bulk soil water. Because accurate measurements for fe are not available for all sites along the SW transect, fe was set to 20 % according to Smith and Freeman (2006) who modeled soil evaporation over a similar relative humidity range. The estimated !wax-soil shows no correlation to RH (R2 = 0.00, Figure 5B). However, because the assumption of constant fe along the SW transect might be too simplistic (fe could be relatively higher in dry regions, such as in Basin and Range, than in wet regions, such as the Plains), we also modeled the soil water under the condition that fe increases linearly from 15-25% for our sites from east to west along the transect. The new model results (shown in Figure 5C) still show no significant correlation between RH and !wax-soil (R2 = 0.07), suggesting that the small effect of RH on !wax-p in our transect samples may be primarily due to soil evaporation rather than plant transpiration. 101 • RH and !wax-p: growth chamber experiments Natural samples are inevitably influenced by many complicated environmental factors that may affect !Dwax, and varying levels of inaccuracy exist in estimates of precipitation !D. To examine the effect of RH on !Dwax while minimizing the complications inherent to natural systems, we conducted controlled plant growth experiments. The plants selected for the growth chamber experiments include broad leaf trees, conifers, C3 grasses and C4 grasses, and represent the typical plant types found in the catchments of lakes in the southwestern United States. The results of the growth experiments indicate that the effect of RH on the apparent hydrogen isotopic enrichment between irrigation water and leaf waxes ("wax-water) is very small (Figure 4). The average "wax-water values for plants grown at 80, 60 and 40% RH are -114 ± 14, -111.5 ± 17, and -107 ± 15 ‰, respectively. Therefore, the "wax-water values only increase by ~7 ‰ over a 40 % change in RH. In order to quantify and to better understand the minor effect of RH on isotopic fractionation between plant leaf waxes and water, we modeled plant leaf water D/H ratios at different humidity levels using the modified Craig-Gordon model of Flanagan et al. (1991) and Roden et al. (2000), which considers the leaf boundary layer and diffusion 102 through stomata (equation 6). We also applied the same isotopic model to simulated plants around the lakes along the SW transect, to examine the RH effect on the isotopic composition of leaf waxes in the lake surface sediment. (6) where Rleaf, Rsoil, Ra refer to the D/H ratios of leaf water subject to evaporation, soil water, atmospheric water vapor, respectively. !* is the equilibrium fractionation factor; !K is kinetic fractionation associated with diffusion in air, and !KB is the kinetic fractionation associated with diffusion through the boundary layer (!KB = (!K)2/3; Roden et al., 2000). ei, es, and ea represent the vapor pressures of intercellular air spaces in the leaf, leaf surface, and the atmosphere, respectively. These vapor pressures were estimated using stomatal conductance and transpiration rate (Ball, 1987; Roden et al., 2000). Eq. 6 estimates the leaf water "D values at the site of evaporation. Because not all of the water within the leaf is subjected to evaporation, a correction is needed for bulk leaf water isotopic ratios ("bulk). (7) 103 where fl is the proportion of the leaf water subjected to evaporative enrichment and ranges from 13 - 33 % (Flanagan et al., 1991); because the accurate measurements of fl are not available, fl is set at 15 % for all sites in this study in order to provide the most conservative estimate. The model results for growth chamber plants and field plants are plotted in Figure 6A and 6B, respectively. To facilitate comparison between the RH effect on the D/H ratios of soil water and leaf water, the precipitation for plants from the field is set to -49 ‰, the same as the irrigation water used in the growth chamber experiments. To be consistent with the plants in the growth chamber, the temperature in the field model was set to 20 oC. However, there are differences between the growth chamber model and the field model. !Dsoil in the growth chamber experiments approximately equals !D of irrigation water because the plants were irrigated every two days, minimizing evaporative effects (note plant transpiration does not fractionate D/H ratios of the remaining water inside the growth chamber). Therefore, the RH effect on soil water in the growth chamber experiment is negligible. However, when modeling field conditions, soil evaporation must be considered. !D of soil water is estimated using Eq.4, where fe is set to 20% according to Smith and Freeman (2006) and as discussed in Section 4.3.1. In other words, the field model takes into consideration both soil evaporation and plant transpiration, whereas the growth chamber model considers only plant transpiration. The modeled 104 !Dleaf water value is -46 ‰ at RH = 80 %, !Dleaf water = -43 ‰ at RH = 60 % and !Dleaf water = -40 ‰ at RH = 40 % for the growth chamber plants. This 6‰ modeled increase in !Dleaf water as RH decreases from 80-40% agrees well with the observed increases in !Dwax of 5.7‰ as RH in the growth chambers was lowered from 80-40% (Figure 4). The effect of RH on the D/H ratios of leaf water is small, as we observed in the transect samples (Figure 5). For the simulated field plants, modeled !Dleaf is slightly higher at corresponding humidity levels, -36, -27, -19 ‰, due to the additional isotope fractionation associated with soil water evaporation. !Dleaf water increases by 17 ‰ when RH decreases from 80 to 40%. This modeled RH effect must represent an average value, as natural desert plants show large variability in their leaf water isotopic ratios (e.g., Gat et al., 2007). Inevitably, growth chamber results can never fully represent all the complex variables of natural settings. The effect of vegetation types Recent studies have shown that there is significant difference in !Dwax among different plant types. Our recent study at Blood Pond, MA shows that !Dwax values from trees are ~40 to 50 ‰ higher than those from grasses (Hou et al. 2007a). Because all plant samples at the Blood Pond site receive the same precipitation, our results suggest significantly larger apparent hydrogen isotopic fractionation during leaf wax synthesis in 105 grasses relative to trees. Similarly, Krull et al. (2006) found that leaf waxes from grasses and grassland soils are ~50 ‰ lower than those from trees and woodland soils. Discrepancy between the !Dwax of grasses and trees was also reported by Liu et al. (2006), although the precipitation !D values in their study were not well constrained. The vegetation change along the SW transect is dramatic, shifting from shrubland to savanna to temperate grassland to subtropical forest (Figure 1C). Relatively small "wax-p would be expected in the eastern Texas Plains due to more extensive forest cover, while relatively large "wax-p would be expected in the western part of the transect due to more extensive grassland, shrubland and CAM plants. However, these expectations are not observed and the large changes in vegetation cover are accompanied by a relatively small change in overall "wax-p. We hypothesize that the opposing hydrogen isotope fractionation effects of RH (as expressed in the above model and our growth experiments) and vegetation type offset one another and result in relatively constant "wax-p across the transect. In order to test our hypothesis, we modeled separately the effects of RH and vegetation cover on !Dwax ratios along the Southwestern US transect. First, we used the soil water model (Eq.5) and leaf water model (Eq.7) to estimate the effect of RH on !Dwax variation, while holding the vegetation constant (assuming 100% tree cover) (Figure 7A). We then calculated the effect of vegetation types on !Dwax values by holding RH constant (assuming RH = 100% for all sites). We used a binary isotopic model to estimate the 106 vegetation effect. Because we were unable to find accurate productivity data for different plant types (e.g., trees, grasses, shrubs, CAM plants) along the transect, we assumed that the relative proportion of trees vs. grasses-shrubs is linearly correlated to RH. CAM plant leaf waxes have mean enrichment factor !wax-p values ~ -145‰ (Chikaraishi and Naraoka, 2003; Chikaraishi et al., 2004). Therefore, similar to an increase in grass-shrub cover, an increase in CAM plants relative to trees should also increase the !wax-p,justifying our use of a simple two end member model to simulate the vegetation effect along the SW transect. The combined vegetation and RH effects are shown in Figure 7A. To simplify the plots, we show the polynomial regression for the measured and modeled D/H ratios of leaf wax (!Dwax(model)) values and !Dp values along the transect. The !Dwax(model) - !Dp curve matches the measured !Dwax – !Dp well (Figure 7A), supporting our hypothesis that the opposing RH and vegetation effects could regulate !Dwax values. Altering the modeled relationship between percent tree cover and RH (by varying the coefficient of the linear relationship from 0.7 to 1.2) only results in minor changes in the modeled !Dwax (Figure 7B, C). Of course, our model is a simplification of natural systems, and many other environmental factors may also play a part in determining the ultimate "wax-p values at any given site. Moreover, there is very little published data showing that the vegetation effect extends to very arid regions. However, we are encouraged to see that 107 this simple model experiment closely matches our observations. It is likely that the RH and vegetation effects are not always perfectly canceled. For example, we observe two distinct relationships between !DP and !Dwax in the transect data (one for the dry portion and one for the wet portion) (Figure 3B). In the Basin and Range, the dominance of grassland and shrubland may produce more leaf waxes with lower D/H ratios, resulting in a vegetation effect that overwhelms the RH effect. This would explain why sites from the Basin and Range show a greater slope in the relationship between !DP and !Dwax than those from Texas Plains (Figure 3B). Vegetation cover may also explain the much stronger correlation between !Dwax and !DP observed in the dry portion of the transect (Figure 3B, R2 = 0.84 for Basin and Range, R2 = 0.64 for the Plains). Diverse types of vegetation, more variable seasonal !DP values, groundwater versus summer precipitation input, and difference in plant growth seasons in the Plains may introduce greater variability to sedimentary !Dwax values, thereby introducing noise to the !Dwax - !DP correlation over the more humid portion of the transect. In contrast, plants in dry regions might only grow in the rainy season, when soil water is available, reducing the variability in source water !D values used by plants during leaf wax biosynthesis. 108 IMPLICATIONS TO PALEOCLIMATE RECONSTRUCTIONS Wet and/or cool regions with high P/E ratio The !D values of aquatic lipids, biosynthesized using lake water, show strong correlation with !D values of lake water at a nearly constant isotopic fractionation (Sauer et al., 2001; Huang et al., 2002, 2004; Hou et al., 2006; Hou et al., 2007b). In cool and wet regions, such as northeastern North America (Huang et al., 2002) and western Europe (Sachse et al., 2004), lake water is mainly recharged by precipitation and/or ground water. Therefore, !D values of lake water more closely represent !D of mean annual precipitation (Figure 8, right panel). However, !Dwax values, especially those from low lying areas, are affected by multiple factors, including multiple water sources (direct precipitation, ground water), seasonality of precipitation, and diversity of vegetation cover. In addition to precipitation, plants can also use the shallow ground water for biosynthesis. The ground water in northeastern North America is mainly recharged in early spring, which adds isotopically-distinct (relative to direct summer precipitation) source water for plants. Diversity of plant types could also introduce significant uncertainties to the sedimentary !Dwax ratios, as different plant types at the same site have been shown to produce leaf waxes with quite different !Dwax values (Krull et al., 2006; Liu et al., 2006; Hou et al., 2007a; Liu and Huang, 2008). Therefore, !Dwax values from 109 higher RH and/or cool regions could, generally, show larger variability relative to !D of precipitation (Figure 8). Dry and/or warm regions with low P/E ratios In relatively dry and warm regions, such as the western part of the transect in this study, the evaporation effect on soil water and leaf water is enhanced due to the low RH (!1, !2 in Figure 8). The net biosynthetic H isotopic fractionation (!bio in Figure 8) also increases relative to wet regions because of the reduced input from trees (Liu et al., 2006; Hou et al., 2007a). In such regions, "D values of aquatic lipids (which record the isotopic composition of lake water) are isotopically enriched relative to precipitation due to strong evaporation. Therefore, in dry regions the "D values of terrestrial leaf waxes may prove to be better proxies for "D of precipitation. Since speleothems tend not to form in areas of insufficient precipitation, leaf waxes extracted from lake sediments may provide the best opportunity for reconstructing the !D of precipitation in arid regions. Saline lakes in arid regions, recharged by mountain glaciers (e.g., Mono Lake in the Western U.S.; Lake Qinghai in China), could represent excellent candidates for paleohydrological reconstruction using the !D of leaf waxes. 110 CONCLUSIONS The hydrogen isotopic composition of higher plant leaf waxes represents an effective proxy for !D of past continental precipitation. The proxy may prove particularly useful in dry regions, such as those with annual precipitation lower than 400 mm shown in this study. It appears that plants grown in dry regions use direct precipitation as source water for biosynthesis, as a means to maximize their water use efficiency. If this is indeed the case, leaf waxes from these plants are likely to track precipitation !D values with high fidelity. While the relationship between and !DP is also strong in wetter regions (such as those with annual precipitation exceeding 400 mm in this study), complicating factors such as multiple water sources (e.g., groundwater vs. summer precipitation), seasonality of precipitation and diversity of vegetation types could lead to reduced fidelity of !Dwax as a paleohydrologic proxy. However, because the sites in this study are restricted to the Southwestern United States, more research is needed to test whether our findings are also applicable to other regions around the world. Our growth chamber experiments indicate that the effect of RH on !Dwax is small. Changing RH from 80 to 40 % led to a mere ~ 7 ‰ increase in hydrogen isotopic enrichment in "wax-water. The experimental results are consistent with output from the latest leaf water model. Based on model results using the natural transect data, soil evaporation 111 appears to make the largest contribution to the hydrogen isotopic variation induced by changes in relative humidity. However, the isotopic effect of relative humidity changes on plant leaf water and leaf waxes appears to be partially cancelled by the opposing isotopic effects of changing vegetation cover. We were able to reproduce the observed relationship between !Dwax and !DP along our southwestern US transect using a simple model by combining the effects of RH with those of vegetation cover. In arid settings, the combination of D/H ratios of terrestrial leaf waxes and aquatic lipids may provide important information concerning the past variability of the hydrologic cycle, including changes to both the isotopic composition of precipitation and lake water evaporation. In wet and/or cool regions, aquatic lipids from lakes in low lying areas are particularly useful for reconstructing past precipitation !D values. Leaf waxes in wet and/or cool regions may show greater variability relative to precipitation. However, if groundwater input could be minimized (e.g. lakes at topographic highs), leaf waxes could also provide valuable information on the !D values of growth season precipitation in wet and/or cool regions. 112 ACKNOWLEDGEMENTS This work was supported by grants from the National Science Foundation (NSF 0318050, 0318123, 0402383) to Y. Huang. We would like to thank the Koebke Family (Dudley, MA) for access to Blood Pond. We thank Dana MacDonald for identifying and collecting the tree seedlings, Fred Jackson and Brian Leib for tending to plants in the greenhouse, David Murray and Joseph Orchardo for assistance during the growth chamber experiments, and Marcelo Alexandre for help during the sample analysis. We thank John Roden for providing the spreadsheet to model the leaf water D/H ratios. REFERENCES Adams D. K. and Comrie A. C. (1997) The North American monsoon. Bulletin of the American Meteorological Society 78, 2197-2213. Ball J. T. (1987) Calculations related to leaf gas exchange. In: Zieger E., Farquhar G. D., and Cowan I. R. Eds.), Stomatal Function. Stanford University Press, Stanford, California. Bowen G. J. and Revenaugh J. (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39, 1299, 113 doi:10.129/2003WR002086. Chikaraishi Y. and Naraoka H. (2003) Compound-specific dD-d13C analyses of n-alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63, 361-371. Chikaraishi Y., Naraoka H. and Poulson S. R. (2004) Hydrogen and carbon isotopic fractionations of lipid biosynthesis among terrestrial (C3, C4 and CAM) and aquatic plants. Phytochemistry 65, 1369-1381. Craig H. and Gordon L. I. (1965) Deuterium and oxygen 18 variations in the ocean and marine atmosphere. In Stable Isotopes in Oceanographic studies and Paleotemperatures (ed. E. Tongiorgi), pp. 9-130. Lab. Geologia Nucleare. Edwards T. D. W. and Fritz P. (1986) Assessing meteoric water composition and relative from 18O and 2H in wood cellulose: Paleoclimatic implications for southern Ontario, Canada. Applied Geochemistry 1, 715-723. Flanagan L. B., Comstock J. P. and Ehleringer J. R. (1991) Comparison of Modeled and Observed Environmental-Influences on the Stable Oxygen and Hydrogen Isotope Composition of Leaf Water in Phaseolus-Vulgaris L. Plant Physiology 96, 588-596. Gat J. R. (1996) Oxygen and hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24, 225-262. Gat J. R., Yakir D., Goodfriend G., Fritz P., Trimborn P., Lipp J., Gev I., Adar E. and 114 Waisel Y. (2007) Stable isotope composition of water in desert plants. Plant and Soil 298, 31-45. Gibson J. J. and Edwards W. T. (1996) Development and validation of an isotopic method for estimating lake evaporation. Hydrological Processes 10, 1369–1382. Gibson J. J., Reid R. and Spence C. (1998) A six-year isotopic record of lake evaporation at a mine site in the Canadian subarctic: results and validation. Hydrological Processes 12, 1779–1792. Gibson J. J. and Edwards T. W. D. (2002) Regional water balance trends and evaporation-transpiration partitioning from a stable isotope survey of lakes in northern Canada. Global Biogeochemical Cycles 16, doi: 10.1029/2001GB001839. Gonfiantini R. (1986) Environmental isotopes in lake studies. In Handbook of Environmental Isotope Geochemistry (ed. P. Fritz and J. C. Fontes), pp. 113-168. Elsevier. Gray J. and Song S. J. (1984) Climatic Implications of the Natural Variations of D/H Ratios in Tree-Ring Cellulose. Earth and Planetary Science Letters 70, 129-138. Grootes P. M., Stuiver M., White J. W. C., Johnsen S. and Jouzel J. (1993) Comparison of Oxygen-Isotope Records from the GISP2 and GRIP Greenland Ice Cores. Nature 366, 552-554. Hou J., Huang Y., Wang Y., Shuman B., Oswald W. W., Faison E. and Foster D. R. (2006) 115 Postglacial climate reconstruction based on compound-specific D/H ratios of fatty acids from Blood Pond, New England. Geochemistry Geophysics Geosystems 7, doi:10.1029/2005GC001076. Hou J., D'Andrea W., MacDonald D. and Huang Y. (2007a) Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts (USA). Organic Geochemistry 38, 977-984 doi:10.1016/j.orggeochem.2006.12.009. Hou J., Huang Y., Oswald W.W., Foster D.R. and Shuman B. (2007b). Centennial-scale compound-specific hydrogen isotope record of Pleistocene - Holocene climate transition from southern New England. Geophysical Research Letters, 34, L19706, doi:10.1029/2007GL030303. Huang Y., Shuman B., Wang Y. and Webb T. III. (2002) Hydrogen isotope ratios of palmitic acid in lacustrine sediments record late Quaternary climate variations. Geology 30, 1103-1106. Huang Y., Shuman B., Wang Y. and Webb T., III. (2004) Hydrogen isotope ratios of individual lipids in lake sediments as novel tracers of climatic and environmental change: a surface sediment test. Journal of Paleolimnology 31, 363-375. Huang Y., Clemens S.C., Liu W., Wang Y. and Prell W.L. (2007) Large scale hydrological change drove the late Miocene C4 plant expansion in the Himalayan foreland and Arabian Peninsula. Geology 35, 531-534. 116 Krull E., Sachse D., Mugler I., Thiele A. and Gleixner G. (2006) Compound-specific !13C and !2H analyses of plant and soil organic matter: A preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology & Biochemistry 38, 3211-3221. Liu W. and Huang Y. (2005) Compound specific D/H ratios and molecular distributions of higher plant leaf waxes as novel paleoenvironmental indicators in the Chinese Loess Plateau. Organic Geochemistry 36, 851-860. Liu Z. and Huang Y. (2008) Hydrogen isotopic compositions of plant leaf lipids are unaffected by a twofold pCO2 change in growth chambers. Organic Geochemistry, doi: 10.1016/j.orggeochem.2008.01.020 Liu W., Yang H. and Li L.W. (2006) Hydrogen isotopic compositions of n-alkanes from terrestrial plants correlate with their ecological life forms. Oecologia 150, 330-338. Pagani M., Pedentchouk N., Huber M., Sluijs A., Schouten S., Brinkhuis H., Damste J. S. S., Dickens G. R. and Scientists E. (2006) Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum. Nature 442, 671-675. Petit J. R., Jouzel J., Raynaud D., Barkov N. I., Barnola J. M., Basile I., Bender M., Chappellaz J., Davis M., Delaygue G., Delmotte M., Kotlyakov V. M., Legrand M., Lipenkov V. Y., Lorius C., Pepin L., Ritz C., Saltzman E. and Stievenard M. (1999) Climate and atmospheric history of the past 420,000 years from the Vostok 117 ice core, Antarctica. Nature 399, 429-436. Roden J. S., Lin G. and Ehleringer J. R. (2000) A mechanistic model for interpretation of hydrogen and oxygen isotope ratios in tree-ring cellulose. Geochimica Et Cosmochimica Acta 64, 21-35. Sachse D., Radke J. and Gleixner G. (2004) Hydrogen isotope ratios of recent lacustrine sedimentary n-alkanes record modern climate variability. Geochimica et Cosmochimica Acta 68, 4877-4889. Sachse D., Radke J. and Gleixner G. (2006) !D values of individual n-alkanes from terrestrial plants along a climatic gradient - Implications for the sedimentary biomarker record. Organic Geochemistry 37, 469-483. Sauer P.E., Eglinton T.I., Hayes J.M., Schimmelmann A. and Sessions A.L. (2001) Compound-specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Cosmochimica Acta 65, 213-222. Schefu! E., Schouten S. and Schneider R. R. (2005) Climatic controls on central African hydrology during the past 20,000 years. Nature 437, 1003-1006. Shuman B., Huang Y., Newby P. and Wang Y. (2006) Compound-specific isotopic analyses track changes in the seasonality of precipitation in the Northeastern United States at ca. 8200 cal yr BP. Quaternary Science Reviews 25, 2992-3002. Smith F.A. and Freeman K.H. (2006) Influence of physiology and climate on dD of leaf 118 wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70, 1172-1187. Terwilliger V.J. and DeNiro M.J. (1995) Hydrogen isotope fractionation in wood-producing avocado seedlings: Biological constraints to paleoclimatic interpretations of delta D values in tree ring cellulose. Geochimica et Cosmochimica Acta 59, 5199-5207. Wang Y.J., Cheng H., Edwards R.L., An Z.S., Wu J.Y., Shen C.C. and Dorale J. A. (2001) A high-resolution absolute-dated late Pleistocene monsoon record from Hulu Cave, China. Science 294, 2345-2348. White J.W.C. (1989) Stable isotope ratios in plants: A review of current theory and some potential applications. In Stable Isotopes in Ecological Research (ed. P. W. Rundel et al.), pp. 142-162. Springer-verlag. White J.W.C., Lawrence J.R. and Broecker W. S. (1994) Modeling and Interpreting D/H Ratios in Tree-Rings - a Test-Case of White-Pine in the Northeastern United-States. Geochimica et Cosmochimica Acta 58, 851-862. Yapp C.J. and Epstein S. (1982) A re-examination of cellulose carbon-bound hydrogen dD measurements and some factors affecting plant-water D/H relationships. Geochimica et Cosmochimica Acta 46, 955-965. Yuan D.X., Cheng H., Edwards R.L., Dykoski C.A., Kelly M.J., Zhang M., Qing J., Lin Y., Wang Y., Wu J., Dorale J.A., An Z.S. and Cai Y. (2004) Timing, during, and 119 transitions of last interglacial Asian monsoon. Science 304, 575-578. 120 Table 1. Geographical lake data and measured hydrogen isotopic ratios of lake water and sedimentary leaf waxes. name State Lat (o) Long(o) Elv (m) RH P (mm) E (mm) T (oC) !DP !DLake !DC26 !DC28 !DC30 Horseshoe Lake* AZ 33.99 -111.72 613 45.0 368 1615 16.2 -67 -73 -173 -161 -162 Bartelett Lake* AZ 33.84 -111.64 527 44.0 500 1637 13.5 -65 -70 -164 -154 -155 Saguaro Lake* AZ 33.57 -111.54 472 41.0 487 1717 13.2 -64 -49 -151 -150 -151 Apache Lake* AZ 33.57 -111.26 567 41.0 304 1731 21.7 -65 -52 -154 -147 -147 Roosevolt Lake* AZ 33.68 -111.10 655 43.0 234 1711 12.3 -67 -67 -164 -156 -157 San Carlos Reservoir* AZ 33.17 -110.52 728 40.0 348 1665 20.4 -66 -59 -167 -161 -163 Cluff Pond* AZ 32.81 -109.86 1006 37.0 463 1743 18.1 -69 -39 -150 -152 -153 Roper Lake* AZ 32.76 -109.72 948 36.0 512 1734 11.3 -68 -57 -159 -157 -158 Dankworth Pond* AZ 32.71 -109.71 978 36.0 503 1770 11.3 -68 -80 -166 -154 -155 Crescent Lake* AZ 33.00 -109.00 1975 42.5 305 1650 15.1 -82 -88 -184 -176 -177 Animas Ditch* AZ 32.28 -108.89 1268 43.0 305 1715 15.6 -70 4 -142 -141 -142 Caballo Lake* NM 32.91 -108.31 1271 47.0 258 1572 15.7 -71 -66 -159 -160 -162 Elephant Butt Lake* NM 33.17 -107.21 1335 54.0 215 1499 13.1 -71 -78 -183 -176 -178 Lake Escondida* NM 34.11 -106.89 1423 60.0 127 1383 13.9 -74 -80 -174 -174 -175 Upper Glacier NM 36.82 -105.05 2524 68.0 358 1246 9.7 -96 -94 -212 -215 -219 Lake #1 NM 36.08 -105.04 2420 65.0 396 1246 4.4 -92 -49 -177 -191 -196 Mary's Lake NM 36.00 -105.00 2155 63.0 - - 11.0 -88 -27 -191 -186 -188 Santa Rosa Lake* NM 35.01 -104.69 1466 51.0 394 1537 14.4 -75 -83 -192 -190 -193 Conchas Lake* NM 35.39 -104.19 1283 53.0 419 1547 11.9 -73 -21 -168 -158 -159 UTE Lake* NM 35.33 -103.43 1155 53.0 194 1547 12.1 -69 -8 -165 -163 -165 Buffalo Spring Creek TX 33.52 -101.71 919 51.0 583 1699 15.6 -56 -39 -131 -136 -136 121 Lake Meredith* TX 35.65 -101.63 616 57.0 455 1654 12.9 -59 -9 -152 -154 -155 Lake Colorado City TX 32.34 -100.92 631 53.0 338 1589 17.6 -48 -3 -106 -134 -134 Champion Creek TX 32.29 -100.86 616 53.0 615 1589 16.7 -48 -18 -126 -143 -143 EV Spence Reservoir TX 32.00 -100.67 607 54.0 483 1589 17.4 -47 -7 -105 -129 -128 Lake Nasworth TX 31.37 -100.60 565 56.0 483 1663 17.4 -45 9 -108 -139 -140 Lake Junction TX 30.49 -99.46 579 61.0 702 1461 18.2 -41 -34 -141 -142 -142 Lake Medina* TX 29.57 -98.70 324 67.0 611 1375 18.5 -35 -18 -150 -162 -163 Lake Bastrop TX 30.18 -97.29 137 73.0 850 1341 19.4 -31 16 -126 -127 -126 Cedar Creek Lake TX 29.93 -96.73 98 76.0 995 1138 20.4 -30 13 -148 -141 -142 Somerville Lake TX 30.33 -96.60 73 76.0 1108 1138 19.6 -29 -7 -116 -124 -124 Lake Houston TX 29.70 -95.17 13 82.0 1150 1175 20.7 -27 -26 -135 -134 -134 *Dammed reservoir; Elv - elevation; RH – Mean annual relative humidity; P – mean annual precipitation; E - evaporation; T - temperature; !DP - mean annual !D of precipitation; !Dlake – !D of lake water; !DC26, !DC28, !DC28 – !D values of C26, C28, C30 n-acids. 122 Table 2. !D values of long chain n-alkanoic acids and mean values ± standard deviation from individual species cultured in growth chamber Life RH=80% RH=60% RH=40% Species form C26 C28 C30 C26 C28 C30 C26 C28 C30 Fraxinus americana tree -156 -142 -136 - -141 -129 -136 -139 -133 Fraxinus americana tree -160 -162 -160 - -145 -152 -134 -156 -141 L Fraxinus americana tree -163 -152 -155 - -135 -136 -132 -139 -139 L Fraxinus americana tree -143 -134 -132 - -134 -121 - -121 -125 L Mean for F.A. -156±9 -147±13 -146±14 - -139±5 -135±13 -134±2 -139±14 -134±8 L Carya glabra tree -157 -153 -142 - -152 -144 - -152 -140 Acer rubrum L. tree -171 -167 -160 -149 -157 -148 -138 -144 -133 Acer rubrum L. tree -169 -163 -151 -138 -143 -135 -141 -146 -135 Acer rubrum L. tree -162 -154 -147 -128 -136 -136 -139 -150 -144 Mean for A.R. -168±5 -161±7 -153±7 -138±11 -145±11 -140±7 -139±1 -146±3 -137±6 Quercus rubra L. tree -183 -171 -166 -169 -177 -169 -167 -173 -166 Quercus rubra L. tree -170 -163 -159 -146 -159 -159 -145 -153 -153 Quercus rubra L. tree -155 -165 -170 -150 -172 -178 -125 -167 -173 Mean for Q.R. -169±14 -167±4 -165±5 -155±12 -170±9 -168±9 -145±21 -164±11 -164±10 Thuja occidentalis tree -154 -134 -128 -141 -136 -125 -140 -139 -133 Thuja occidentalis tree -138 -133 -134 -143 -140 -131 -140 -141 -136 Mean for T.O. -146±11 -134±1 -131±4 -142±1 -138±3 -128±4 -140±0 -140±1 -135±2 Picea glauca tree -158 -141 -139 -152 -138 -138 -156 -147 -153 Picea glauca tree -161 -146 -145 -156 -141 -139 -151 -144 -144 Mean for P.G. -159±2 -143±4 -142±4 -154±3 -139±2 -139±1 -154±4 -145±3 -149±6 Zea mays L. grass -165 -156 -160 -162 -154 -151 -158 -157 -166 Zea mays L. grass -164 -154 -146 -159 -159 -164 -155 -164 -163 Mean for Z.M. -164±0 -155±2 -153±10 -161±2 -156±3 -157±9 -157±2 -160±5 -165±3 Dactylis glomerata L. grass -208 -190 -178 - -192 -189 - -188 -183 Dactylis glomerata L. grass - -191 -189 - -194 -183 - -174 -170 Mean for D.G. -208±0 -190±1 -184±7 - -193±1 -186±4 - -181±10 -177±9 Alopecurus spp. grass -162 -158 -147 -170 -167 -160 - - - Alopecurus spp. grass -169 -162 -149 -161 -164 -159 - - - Alopecurus spp. grass -165 -160 -157 -157 -163 -159 - - - Mean for A. -166±4 -160±2 -151±5 -163±6 -165±2 -159±1 - - - Phleum pratense L. grass -166 -160 -150 -166 -159 -143 -147 -144 -135 Phleum pratense L. grass -166 -152 -150 -159 -158 -158 -141 -138 -131 Phleum pratense L. grass -169 -160 -132 -158 -152 -130 -146 -144 -136 Mean forP.P. -167±2 -157±4 -144±10 -161±4 -156±4 -144±14 -145±3 -142±4 -134±3 123 124 Figure 1. The modern lake transect in the southwestern United States. (A) Precipitation gradient (mean annual precipitation (mm/yr) from 1961 to 1990; data from NOAA Cooperative Station Normals climate observations); (B) relative humidity (RH) gradient and mean annual relative humidity (%) (data from http://www.cdc.noaa.gov); (C) vegetation cover (data from http://www.nationalatlas.gov); ST – Subtropical forests; PR – Prairie; T/ST D – Tropical/subtropical Desert; T/ST S – Tropical/subtropical Steppe; M/A Z Mountain with Altitudinal Zonation. The arrows in (A) indicate the moisture source (Tropical Pacific Ocean from west, Gulf of Mexico from east) for precipitation in this region. The lakes are divided into two groups according to the geographic division and mean annual precipitation: Plains (including lakes from Texas); Basin and range (including lakes from New Mexico and Arizona). 125 Figure 2. Hydrogen isotope data for the SW transect: !Dwax= measured D/H ratios of sedimentary leaf waxes; !DP= D/H ratios of precipitation calculated from OIPC; !Dlake= measured D/H ratios of lake water; !Dmodel= modeled D/H ratios of lake water; 95% CF (!DP)= 95% confidence interval for !DP; 95% CF (!Dwax)= 95% confidence interval for !Dwax. 126 Figure 3. Correlation between !Dwax and !DP for the lake surface sediment samples along the SW transect. (A) Correlation for the entire data set showing both linear and polynomial regressions. (B) Correlation for data from distinct geographical provinces (Basin and Range, and Plains). 95% confidence intervals are shown. 127 Figure 4. Differences in the apparent isotopic enrichment of leaf waxes for plants grown in growth chamber at three humidity levels. The left-hand plots depict changes in enrichment factor (!wax-water) as RH changes from 80% to 40% (A: RH = 80%; B: RH = 60%; C: RH = 40%). The vertical dashed line represents the average !wax-water value at RH = 80%. The righthand plots show hydrogen isotopic differences (! values) between different RH levels. (D = "D80-60%; E = "D60-40%; F = "D80-40%). 128 Figure 5. The correlation between RH and apparent isotopic enrichment of sedimentary leaf waxes relative to (A) precipitation (!wax-p), (B) modeled soil water (!wax-soil water) assuming fe = 15%, (C) modeled soil water (!wax-soil water) assuming a linear east to west gradient of fe from 15% to 25% for our study sites. !wax-soil water = [(D/H)wax/(D/H)soil water – 1] " 1000 ‰. Three of the lakes, AD (Animas Ditch), UGL (Upper Glacial Lake) and SRL (Santa Rosa Lake) show different !wax-p values (see text for discussion). 129 Figure 6. Modeled D/H ratios of soil water and leaf water for plants (A) in the growth chamber and (B) in the field. To facilitate the comparison between the plants from the growth chamber and the field, the D/H ratio of precipitation is set to the !D value of irrigation water in the growth chamber (-49 ‰). 130 131 Figure 7. (A) Models showing the opposing effects of RH and vegetation on the sedimentary !Dwax. The data for RH effect (circles) are obtained by assuming the production percentage of tree leaf waxes is 100%, whereas the data for vegetation effect (triangles) are obtained by assuming RH is 100% for all lakes. The average of the RH effect and vegetation effect for all sites (!Dwax(model), dashed line) is then compared with measured D/H ratios of leaf waxes from the lake sediments (!Dwax, black line). To facilitate comparison, second order polynomial regressions are shown for all four datasets. (B) and (C) Sensitivity tests for the model combining RH and vegetation effects. As the percentage of tree cover increases from 0.7RH (B) to 1.2RH (C), the modeled !Dwax vs. !Dp relationship does not change significantly. 132 Figure 8. The schematic variation of D/H ratios for low RH and/or warm regions (left panel), and high RH and/or cool regions (right panel; This panel is adapted from Sachse et al. (2006)). !1, !2, !’1, !’2 represent the evaporation effect from precipitation to soil water, then to leaf water; !bio and !’bio are the hydrogen isotopic enrichment during biosynthesis. !aq is the biosynthesis effect for aquatic plants. !wax-p and !’wax-p are the apparent hydrogen isotopic enrichment from precipitation to leaf waxes. E is the isotopic enrichment between precipitation and lake water due to evaporation. CHAPTER 4 Centennial-scale compound-specific hydrogen isotope record of Pleistocene - Holocene climate transition from southern New England Co-authors: Yongsong Huang, W. Wyatt Oswald, David R. Foster, Bryan Shuman Published in Geophysical Research Letters, 2007, v. 34, L19706, doi:10.1029/2007GL030303 133 134 135 ABSTRACT Northeastern North America experienced major climate shifts during the Pleistocene – Holocene transition. However, there have been no high-resolution isotopic records of climate change from this region. Here, we present a centennial- scale record of climate change during the transition based on D/H ratios of behenic acid (C22 n-acid) or !DBA from a sediment core in Blood Pond, Massachusetts. Surface calibrations from a transect of 19 lakes in eastern North America show that !DBA values track mean annual atmospheric temperature variations. The abrupt climate events observed in Blood Pond records show remarkable similarity with Greenland ice core !18O records during the Pleistocene. During the early Holocene, the northeastern North America !DBA record was more variable than Greenland, possibly due to the close proximity of the Laurentide ice sheet, and impact of freshwater outbursts as the ice sheet rapidly retreated. 136 INTRODUCTION The Pleistocene - Holocene transition is characterized by abrupt climatic fluctuations around North Atlantic Ocean (e.g., Stuiver et al., 1995; Hughen et al., 1996). However, the spatial variations in the timing, amplitude, and phasing of the abrupt events are less understood on the adjacent continents. This is particularly true for the northeastern North America where the driving forces for climate were particularly complex, comprising a combination of changes in North Atlantic sea surface temperature, Laurentide ice sheet (LIS) extent, atmospheric composition, and insolation (e.g., Webb et al., 1993). Centennial-scale quantitative records from the northeastern North America are thus extremely important for better understanding marine-terrestrial-atmosphere- cryosphere connections and regional climatic responses. Existing paleoclimate records from the northeastern North America are mainly based on paleoecological approaches, such as assemblages of pollen (e.g., Peteet et al., 1990; Webb et al., 1993; Shuman et al., 2002a) and Chironomidae (midge, e.g., Cwynar and Spear, 2001; Walker et al., 1997) from lake sediments. However, the possibility of transient vegetation responses to short-term, small-magnitude climate variations (e.g., Davis and Botkin, 1985), and the non-analogue conditions for Chironomidae (e.g., Kurek et al., 2002) could increase the difficulty of using pollen and Chironomidae data to assess abrupt (6 °C. This is unlikely, given that our estimated temperature change during the YD event is approximately the same amplitude (Hou et al., 2007). We attribute our observed isotopic change as a combination of seasonal shift in precipitation and temperature around 9.2 ka, with the former factor predominating. We have no means to separate these two effects at this time, and therefore designate changes in !DBA to precipitation-weighted temperature, which integrates both factors. The duration of the 9.2 ka event in Blood Pond records defined by CLIM-X-DETECT (Mudelsee, 2006) spans 9.3 to 9.1 ka, which is consistent with the visual observation. The amplitude defined by CLIM-X-DETECT is ~ 3°C, which is slightly smaller than the visual observation. Therefore, the 9.2 ka event revealed by Blood Pond !DBA records spans about 200 years (9.3-9.1 ka) and the precipitation weighted temperature variability is about 3°C cooler than the millennial-scale background temperature during early Holocene. Suggestions of a ~9.2 ka climate reversal are recently found in Wildwood Pond, New York, where an evident sandy layer suggests a brief period of cold and dry climate around 9.2 kyr BP (Oswald et al., in review). The Blood Pond temperature record also reveals a clear climate reversal from 8.45 to 8.3 kyr BP, spanning ~ 150 yrs (Figs. 1 & 2). The maximal precipitation weighted temperature change is ~ 3 oC, or 1.5 oC relative to millennial background temperature based on the CLIM-X-DETECT program. Many published lake sediment records in northeastern North America have identified climate reversals at ~ 8.4 to 8.2 ka. For ! ! 163 example, !D records of both aquatic (C16 n-acid) and terrestrial (C28 n-acid) biomarkers from Berry and Crooked Ponds, Massachusetts suggest a multi-century period of cold conditions from 8.4 to 8.0 ka (Shuman et al., 2006). Pollen records from North Pond (Shuman et al., 2004) and Makepeace Cedar Swamp (Newby et al., 2000) also reveal abrupt climate changes around this time. Midge assemblages and LOI records from two lakes in White Mountain also reflect cooling between 8.4 and 8.2 kyr BP (Kurek et al., 2004). LOI and sediment grain size records from Brown’s Lake in Ohio also showed abrupt variations from 8.9 to 8.0 ka (Lutz et al., 2008). Due to the uncertainties in radiocarbon chronology, it is likely that all these observed climate reversals in northeastern North America occurred around the same time and correspond directly to the so called 8.2 event in Greenland (Alley et al., 1997) and other regions of the world (e.g., Alley and Agustsdottir, 2005). However, it is also likely that the climate reversal at New England indeed proceeded the 8.2 event on top of Greenland ice sheet by ~ 200 years, because a number of independently dated lake records show a similar timing of climate reversal at ~ 8.4 ka (Shuman et al., 2002; Shuman et al., 2006). Most importantly, the freshwater release that has been suggested to trigger the “8.2 event” occurred at 8.4 ka rather than 8.2 ka (Teller and Leverington, 2004). Therefore, it is possible that the climate reversal in New England show different sensitivity and timing of response to the 8.4 ka freshwater release from Greenland ice core records. We will fully discuss possible scenarios in the next section of this paper. !D records of leaf waxes (!Dwax) at Blood Pond show significantly different variability from !DBA records during the early Holocene (Fig.2). !Dwax records revealed ! ! 164 gradually decreasing trend from 10 to 7 ka by ~ 10‰, with abrupt increase of 10‰ around 9.3 – 9.1 ka, and 10‰ at 8.5 – 8.35 ka. Higher plants use a combination of summer precipitation and ground water as their source water in the northeastern United States (White et al., 1985; Hou et al., 2008), with summer precipitation has much higher D/H ratios than groundwater. The gradual decrease trend of !Dwax records from 10 to 7 ka probably suggest a slowly decreased contribution of summer precipitation and/or increased contribution of groundwater to the plant source water. This is supported by the increase in regional lake level during the same period (Fig.2; Shuman et al., 2001; Newby et al., 2000). The abrupt increase in !Dwax around 9.3 – 9.1 and 8.5 – 8.5 ka may imply abrupt increases in contribution from summer precipitation as the higher plant source water. ! ! 165 MECHANISMS FOR THE ABRUPT CLIMATE SHIFTS DURING THE EARLY HOLOCENE The slowdown in thermohaline circulation in North Atlantic Ocean caused by the freshwater releases from the proglacial lakes (Lake Agassiz and Ojibway) have been suggested to be responsible for the 8.2 event (e.g., Barber et al., 1999; Alley and Agustsdottir, 2005) and the recently suggested 9.2 ka event (Fleitmann et al., 2008). The slowdown in THC resulted in a decrease of the northward heat transport by the North Atlantic Ocean, probably leading to increased sea-ice cover in northern North Atlantic Ocean (Wiersma and Renssen, 2006). The increased sea-ice cover would result in steeper thermal gradient between high and low latitudes, which decreases the precipitation at high latitudes (including Europe, North Atlantic Ocean, Asia and North America) associated with modifications to atmospheric circulation (Renssen et al., 2002;Wiersma and Renssen, 2006). The climate effect of the slowdown in THC as a result of meltwater outburst have also been simulated in various models (e.g., Renssen et al., 2002; Alley and Agustsdottir, 2005; Wiersma and Renssen, 2006), suggesting generally cool and dry climate in Northern Hemisphere around 8.2 ka. Fleitmann et al. (2008) suggest that there might be a worldwide “9.2 ka event”, as shown by the widespread and nearly synchronous climatic perturbation during the early Holocene. The evidence for the possible existence of such an event is based on the alignment of a set of high-quality climate records across climate zones, including speleothem records from Dongge Cave in China (Dykoski et al., 2005) and Qunf Cave in ! ! 166 Oman (Cai et al., 2008), Greenland ice cores (NGRIP, GRIP and DYE-3) (Vinther et al., 2006), and oxygen isotope records from Lake Ammersee, Germany (von Grafenstein et al., 1999). Fleitmann et al. (2008) suggest that freshwater release to the North Atlantic Ocean centered at 9.17±0.15 ka caused the slowdown in THC, which then induced widespread climate changes. Our hydrogen isotopic records of the C22 and C28 n-acids from Blood Pond represent the first quantitative indicators for abrupt climatic reversal at ~ 9.2 ka in the northeastern North America, where the climate is likely to be highly responsive to meltwater releases, such as demonstrated by the large cooling during the Younger Dryas event (Webb et al., 1993; Hou et al., 2007). The abrupt isotopic shift at 9.2 ka from Blood Pond is concurrent with Greenland ice core records (Vinther et al., 2006), lake sediment records (e.g., von Grafenstein et al., 1999), speleothem records(e.g., Dongge Cave, China (Dykoski et al., 2005), Qunf Cave, Oman (Fleitmann et al., 2007), and Dongshiya Cave, China (Cai et al., 2008) (Fig.1). There is a significant conflict if one assumes that amount of meltwater released into north Atlantic is proportional to the degree of climate reversal. The volume of the meltwater outburst triggered the 8.2 ka event was the largest freshwater release during the early Holocene (5.2 Sv if released within one year, 1 Sv = 106 m3/s) (Teller and Leverington, 2004). The volume of the freshwater release centered at 9.17 ka was only ~5% of the meltwater outburst triggered 8.2 ka event. However, the climate responses to the two freshwater release events are very different for the currently available, high resolution records during the early Holocene. The precipitation weighted temperature variability at 9.2 ka is at least 2 times larger than that at 8.4 ka in the Blood Pond record ! ! 167 (Figs.1 & 2). There are several records around the world that show greater 9.2 ka climate response than the response at 8.2 ka, including the speleothem !18O records from Dongge Cave, China which show 1‰ and 0.5‰ for 9.2 and 8.2 ka events respectively (Dykoski et al., 2005) and the ostracod !18O records from Lake Ammersee which show 1.0‰ and 0.7‰ variation for the two abrupt events (von Grafenstein et al., 1999). However, the ice core oxygen isotope records from Greenland show very similar response for the climate change at 9.2 and 8.2 ka (Fig.1; Vinther et al., 2006). ! ! 168 AMPLITUDES OF THE ABRUPT CLIMATE SHIFTS DURING THE EARLY HOLOCENE We attribute the amplitude difference between the two abrupt climate reversals in northeastern North American continent during the early Holocene primarily to two possible factors 1) the difference in geographic distribution of Laurentide Ice sheet (LIS) and associated atmospheric circulation at 9.2 ka and 8.2 ka (Fig.3); and 2) to lesser extent, different meltwater paths at 9.2 and 8.4 ka. The LIS extent varied dramatically in North American continent during the 9.2 ka event and the 8.2 ka event, in contrast to the relatively invariable Greenland Ice Sheet (Fig.3; Peltier, 1994). The relatively large and southerly LIS at 9.2 kyr BP may have created a strong anticyclone system (Fig.3A) that can effectively reduce the moisture input from Gulf of Mexico into northeastern North America. This could also interfere with the heat transport by Gulf Stream to the north, leading THC to a less stable status. The freshwater release into North Atlantic Ocean around 9.2 ka would further slowdown the THC, which then amplified the effect of anticyclone on the climate in this region. However, as the LIS collapsed around 8.4 ka (Fig.3B), the anticyclone system over the North American continent would disappear or be very weak, which cannot effectively block the moisture from Gulf of Mexico. The THC would be more stable as the interference from the anticyclone was minimal. The slowdown in THC as a result of freshwater release into North Atlantic Ocean around 8.4 ka would influence the climate in northeastern North American continent to a less extent because of lack of positive feedback to the anticyclone system around 9.2 ka. Therefore, the climate response in New England to freshwater release is likely amplified at 9.2 ka ! ! 169 than at 8.4 ka, about doubled as exemplified by Blood Pond temperature records. The positive feedback between the anticyclone over North America and slowdown in THC would be propagated to Europe and Asia via atmospheric circulation. In comparison, the climate responses of Greenland at 9.2 ka and during the 8.2 ka event were not significantly different because the Greenland ice sheet did not change in size significantly from 9.2 to 8.2 ka. The amplitude difference between the 8.2 and 9.2 ka events could also result from the routes of the freshwater release events from the proglacial lakes. The freshwater release occurred at 9.17 ka routed east through Nipigon Basin, the Great Lakes to the North Atlantic Ocean, a routing that injects freshwater close to key areas of North Atlantic Deep Water formation (Teller and Leverington, 2004). However, the final freshwater release, triggering the 8.2 ka event, routed north to Hudson Bay. It is likely that the former route may have a greater influence on the stability of the thermohaline circulation. However, there is currently no direct evidence to support the above assumptions. Climate models have so far mostly focused on the 8.2 ka events when the Laurentide Ice Sheet (LIS) was very small or nearly absent. There are currently no models to simulate the climate effect of the meltwater outburst which triggered to the 9.2 ka event, when the LIS was still relatively large in North American continent. High spatial resolution GCM models could probably provide more insightful ideas. Moreover, high resolution (both spatial and temporal) climate records from North Atlantic Ocean and surrounding ! ! 170 continents will also be helpful to resolve the different regional response to the meltwater pulse from the proglacial lakes. ! ! 171 ABRUPT CHANGES IN PRECIPITATION SEASONALITY AT 9.2 AND 8.4 KA IN NEW ENGLAND The hydrogen isotope records from Blood Pond suggest dramatic changes in precipitation seasonality (relative proportion of summer and winter precipitation) in northeastern North America during the early Holocene, which was caused by the changes in THC and associated atmospheric circulation. The precipitation seasonality strongly influences the D/H ratios of C22 n-alkanoic acid, which reflects the D/H ratios of lake water (Hou et al., 2006) since Blood Pond is recharged by ground water that represents annual mean precipitation. On the other side, the precipitation seasonality also affects the !D values of leaf waxes from lake sediment, as the higher plants use soil water for biosynthesis. The soil water in New England represents a combination of ground water and summer precipitation (White et al., 1985). During 10 - 9 ka, the LIS and associated anticyclone (high pressure system) prevented the moisture originated from Gulf of Mexico transporting to northeastern North America (Fig.3A). This leads to less mean annual precipitation, but mostly decreased summer precipitation, which in turn increased the proportion of winter precipitation. The LIS collapsed around 8.4 ka, therefore the associated anticyclone became very small or disappeared, and then the water vapor from Gulf of Mexico could freely penetrate into the northeastern North American continent (Fig.3B). This would lead increase in annual precipitation, but mostly increased summer precipitation, which in turn decreased the proportion of winter precipitation. The precipitation pattern would affect the isotopic composition of groundwater and lake water. ! ! 172 In northeastern North American continent, the groundwater isotopic ratios are generally equivalent to annual mean precipitation (Darling and Bath, 1988). The general trends in !D records of C22 and C28 n-acids can be explained by the changes in precipitation seasonality. The lower C22 n-acid !D values and higher C28 n- acid !D values between 10 and 8.5 ka can be attributed to less annual precipitation (decreased proportion of summer precipitation and increased proportion of winter precipitation). Because of less mean annual precipitation, the groundwater and lake level were relatively low, the terrestrial plants mainly used summer precipitation for biosynthesis. Summer precipitation was also enriched due to the amount effect (since smaller amount of low latitude moisture will reach our study site, representing more evolved moisture along Rayleigh fractionation path). This results in higher D/H ratios in plant source water and hence the leaf waxes. The lake water in New England, recharged mainly by groundwater and cool season precipitation, would be isotopically lighter, which is recorded by C22 n-acid. After 8.4 ka when LIS collapsed, the increased annual precipitation (mainly increase in summer precipitation) led elevated groundwater and lake level. The terrestrial plants could use both groundwater and summer precipitation for biosynthesis, which results in lower D/H ratios in leaf waxes. The generally high !D values of C22 n-acid also suggest that the lake water was isotopically higher because of more contribution from summer precipitation. The changes in precipitation seasonality are supported by the lake level records from Crooked Pond (Shuman et al., 2001) and Makepeace Cedar Swamp (Newby et al., 2000) (Fig.2). ! ! 173 The abrupt changes in !D records at Blood Pond during the 8.2 ka and 9.2 ka events indicate the dramatic changes in precipitation seasonality in northeastern North America. The lower C22 n-acid !D values (representing depleted lake water) and higher C28 n-acid !D values (more contribution of summer precipitation) suggest extreme seasonality at the two abrupt climate reversals during the early Holocene. The less annual precipitation was dominated by winter precipitation. Simulation of precipitation seasonality could also produce similar changes in lake water isotopic composition and precipitation-weighted temperature. The average !D value of modern warm season precipitation is ~82‰, and the cool season precipitation is ~ 40‰, which were calculated from the online isotopes in precipitation calculator (OIPC, Bowen and Revenaugh, 2003). A binary isotope model shows that 20% decrease in summer precipitation would produce a 13‰ decrease in mean annual precipitation if the annual precipitation amount does change, which corresponds to variability of 2.5°C in precipitation weighted temperature. ! ! 174 CONCLUSION We report the first quantitative record of an abrupt and prominent climatic reversal at 9.2 kyr BP based on hydrogen isotopic records from Blood Pond, northeastern North America. The 9.2 ka event spans ~200 yr, from 9.3 to 9.1 kyr BP, with an average ~3 °C precipitation weighted cooling against the millennium background. The “8.2 ka event” was also shown in the Blood Pond records, with a smaller response, ~1.5 °C cooling over 150 yrs and earlier timing (8.45-8.3 kyr BP) than Greenland. Both abrupt climate shifts could be triggered by the slowdown of thermohaline circulation as a result of meltwater outbursts into North Atlantic Ocean from proglacial lakes (Lake Agassiz and Ojibway) in North American continent. The difference in extent of Laurentide Ice Sheet and associated atmospheric circulation during the two abrupt climate shifts at 9.2 ka and 8.4 ka was critical in determining the climate response as a result of meltwater releases. The climate was much more sensitive to the meltwater at 9.2 ka when LIS was and large and in a more southerly position. Even though the meltwater volume at 9.2 ka was only ~5 % of that at 8.4 ka, the changes in precipitation weighted temperature was more than twice as large. The main characteristic of the 9.2 ka climate reversal in New England is the dramatic shift in precipitation seasonality. High spatial resolution GCM models, and high resolution (both spatial and temporal) climate records from North Atlantic Ocean and surrounding continents will be helpful to resolve the different regional response to the meltwater pulse from the proglacial lakes when boundary conditions were different during the early Holocene. ! ! 175 ACKNOWLEDGEMENTS This work was supported by grants from the National Science Foundation (NSF 0318050, 0318123, 0402383) to Y. Huang. REFERENCE Alley R.B. (2000) The Younger Dryas cold interval as viewed from central Greenland. Quaternary Science Reviews 19, 213-226. Alley R.B. and Agustsdottir A.M. (2005) The 8k event, cause and consequences of a major Holocene abrupt climate change. Quaternary Science Reviews 24,1123- 1149. Alley R.B., Marotzke J., Nordhaus W.D., Overpeck J.T., Peteet D.M., Pielke R.A.J., Pierrehumbert R.T., Rhines P.B., Stocker T.F., Talley L.D. and Wallace J.M. (2003) Abrupt climate change. Science 299, 2005-2010, doi:10.1126/science.1081056. Alley R.B., Mayewski P.A., Sowers T., Stuiver M., Taylor K.C. and Clark P.U. (1997) Holocene climatic instablity: A prominent, widespread event 8200 yr ago. Geology 25, 483-486. Banner J.L., Guilfoyle A., James E.W., Stern L.A. and Musgrove M. (2007) Seasonal variations in modern speleothem calcite growth in Central Texas, U.S.A. Journal of Sedimentary Research 77, 615-622. ! ! 176 Barber D.C., Dyke A., Hillaire-Marcel C., Jennings A.E., Andrews J.T., Kergin M.W., Bilodeau G., McNeely R., Southon J., Morehead M.D. and Gagnon J.M. (1999) Forcing of the cold event of 8200 years ago by catasrophic drainage of Laurentide lakes. Nature 400, 344-348, doi:10.1038/22504. Bowen G.J. and Revenaugh J. (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39, 1299, doi:10.129/2003WR002086. Broecker W.S. (2007) Abrupt climate change revistied. Global and Planetary Change 54, 211-215, doi:10.1016/j.gloplacha.2006.06.019. Cai B., Edwards R.L., Cheng H., Tan M., Wang X. and Liu T. (2008) A dry episode during the Younger Dryas and centennial-scale weak monsoon events during the early Holocene, A high-resolution stalagmite record from southeast of the Loess Plateau, China. Geophysical Research Letters 35, doi:10.1029/2007GL030986. Clark P.U., Pisias N.G., Stocker T.F. and Weaver A.J. (2002) The role of the thermohaline circulation in abrupt climate change. Nature 415, 863-869. Darling W.G. and Bath A.H. (1988) A stable isotope study o recharge processes in the English chalk. Journal of Hydrology 101, 31-46. Davis M.B. and Botkin D.B. (1985) Sensitivity of cool-temperate forests and their fossil pollen record to rapid temperature-change. Quaternary Research 23, 327-340. Denton G.H., Alley R.B., Comer G.C. and Broecker W.S. (2005) The role of seasonality in abrupt climate change. Quaternary Science Reviews 24, 1159-1182. Denton G.H. and Karlén W. (1973) Holocene climatic variations, their pattern and possible cause. Quaternary Research 3, 155-205. ! ! 177 Dykoski C.A., Edwards R.L., Cheng H., Yuan D., Cai Y., Zhang M., Li Y., Qing J., An Z. and Revenaugh J. (2005) A high-resolution, absolute-dated Holocene and deglacial Asian monsoon record from Dongge Cave, China. Earth and Planetary Science Letters 233, 71-86, doi:10.1016/j.epsl.2005.01.036. Ficken K.J., Li B., Swain D.L. and Eglinton G. (2000) An n-alkane proxy for the sedimentary input of submerged/floating freshwater aquatic macrophytes. Organic Geochemistry 31, 745-749. Fleitmann D., Burns S.J., Mangini A., Mudelsee M., Kramers J., Villa I., Neff U., Al- Subbary A.A., Buettner A., Hippler D. and Matter A. (2007) Holocene ITCZ and Indian monsoon dynamics recorded in stalagmites from Oman and Yemen (Socotra). Quaternary Science Reviews 26, 170-188, doi:10.1016/j.quascirv.2006.04.012. Fleitmann D., Mudelsee M., Burns S.J., Bradley R.S., Kramers J. and Matter A. (2008) Evidence for a widespread climatic anomaly at around 9.2 ka before present. Paleoceanography 23, doi:10.1029/2007PA001519. Hou J., D'Andrea W.J. and Huang Y. (2008) Can sedimentary leaf waxes record D/H ratios of continental precipitation? Field, model, and experimental assessments. Geochimica et Cosmochimica Acta 72, doi:10.1016/j.gca.2008.04.030. Hou J., Huang Y., Wang Y., Shuman B., Oswald W.W., Faison E. and Foster D.R. (2006) Postglacial climate reconstruction based on compound-specific D/H ratios of fatty acids from Blood Pond, New England. Geochemistry Geophysics Geosystems 7, doi:10.1029/2005GC001076. ! ! 178 Hou J., Huang Y., Oswald W.W., Foster D.R. and Shuman B. (2007) Centennial-scale compound-specific hydrogen isotope record of Pleistocene-Holocene climate transition from southern New England. Geophysical Research Letters 34, doi:10.1029/2007GL030303. Kurek J., Cwynar L.C. and Spear R.W. (2004) The 8200 cal yr BP cooling event in eastern North America and the utility of midge analysis for Holocene temperature reconstructions. Quaternary Science Reviews 23, 627-639. Lanzante J.R. (1996) Resistant, robust and non-parametric techniques for the analysis of climate data, theory and examples, including applications to historical radiosonde station data. International Journal of Climatology 16, 1197-1226. Lie Ø. and Paasche Ø. (2006) How extreme was northern hemisphere seasonality during the Younger Dryas. Quaternary Science Reviews 25, 404-407. Lutz B., Wiles G., Lowell T. and Michaels J. (2007) The 8.2 ka abrupt climate change event in Brown's Lake, northeast Ohio. Quaternary Research 67, 292-296, doi:10.1016/j.yqres.2006.08.007. Mickler P.J., Stern L.A. and Banner J.L. (2006) Large kinetic isotope effects in modern speleothems. Geological Society of America Bulletin 118, 65-81. Mudelsee M. (2006) CLIM-X-DETECT, A Fortran 90 program for robust detection of extremes against a time-dependent background in climate records. Computers & Geosciences 32, 141-144, doi:10.1016/j.cageo.200505.010. Newby P.E., Killoran P., Waldorf M.R., Shuman B., Webb R.S., Webb T.III. (2000) 14,000 years of sediment, vegetation, and water-level changes at the Makepeace Cedar Swamp, Southeastern Massachusetts. Quaternary Research 53, 352-368. ! ! 179 Overpeck J.T. and Cole J.E. (2006) Abrupt change in Earth's climate system. Annual review of Environment and Resources 31, 1-31. Peltier W.R. (1994) Ice age paleotopography. Science 265, 195-201. Renssen H., Goosse H. and Fichefet T. (2002) Modeling the effect of freshwater pulses on the early Holocene climate, the influence of high frequency climate variability. Paleoceanography 17, doi:10.1029/2001PA000649. Shuman B., Bartlein P.J. and Webb T.III. (2005) The magnitudes of millennial- and orbital-scale climatic change in eastern North America during the Late Quanternary. Quaternary Science Reviews 24, 2194-2206. Shuman B., Bravo J., Kaye J., Lynch J.A., Newby P. and Webb T.III. (2001) Late Quaternary water-level variations and vegetation history at Crooked Pond, Southeastern Massachusetts. Quaternary Research 56, 401-410. Shuman B. and Donnelly J.P. (2006) The influence of seasonal precipitation and temperature regimes on lake levels in the northeastern United States during the Holocene. Quaternary Research 65, 44-56. Shuman B., Newby P., Huang Y. and Webb T.III. (2004) Evidence for the close climatic control of New England vegetation history. Ecology 85, 1297-1310. Shuman B., Huang Y., Newby P. and Wang Y. (2006) Compound-specific isotopic analyses track changes in the seasonality of precipitation in the Northeastern United States at ca. 8200 cal yr BP. Quaternary Science Reviews 25, 2992-3002. Teller J.T. and Leverington D.W. (2004) Glacial Lake Agassiz, A 5000 yr history of change and its relationship to the delta O-18 record of Greenland. Geological Society of America Bulletin 116, 729-742. ! ! 180 Vinther B.M., Clausen H.B., Johnsen S.J., Rasmussen S.O., Andersen K.K., Buchardt S.L., Dahl-Jensen D., Seierstad I.K., Siggaard-Andersen M.L., Steffensen J.P., Svensson A., Olsen J. and Heinemeier J. (2006) A synchronized dating of three Greenland ice cores throughout the Holocene. Journal of Geophysical Research 111, doi:10.1029/2005JD006921. von Grafenstein U., Erlenkeuser H., Brauer A., Jouzel J. and Johnsen S.J. (1999) A mid- European decadal isotope-climate record from 15,500 to 5000 years BP. Science 284, 1654-1657. Webb T.III., Bartlein P.J., Harrison S.P. and Spaulding W.G. (1993) Vegetation, lake levels, and climate in Eastern North America for the past 18,000 years, in Global Climates since the Last Glacial Maximum, edited by H.E. Wright Jr, J. E. Kutzbach, T. Webb, III, W. F. Ruddiman, F. A. Street-Perrot, and P. J. Bartlein, University of Minnesota Press, Minneapolis, 415-467. Wiersma A.P. and Renssen H. (2006) Model-data comparison for the 8.2 ka BP even, confirmation of a forcing mechanism by catastrophic drainage of Laurentide Lakes. Quaternary Science Reviews 25, 63-88, doi:10.1016/j.quascirev.2005.07.009. ! ! 181 ! Figure 1. The 9.2 and 8.2 ka events were revealed in Blood Pond !D records, speleothem !18O records and Greenland ice core !18O records. Also shown the meltwater outbursts from proglacial lakes during the early Holocene. ! ! 182 ! ! Figure 2. Comparison between !D records of C22 and C28 n-alkanoic acids. D/H ratios of C22 n-alkanoic acid represent the precipitation-weighted temperature; D/H ratios of C28 n-alkanoic acid reflect the available water for plant biosynthesis. Lake level records from Crooked Pond (Shuman et al., 2001) are also shown to indicate the elevated lake level from 10 to 7 ka. ! ! 183 Figure 3. The extent of Laurentide Ice Sheet (LIS) in North American continent between 10 and 9 kyr BP (left), between 8 and 7 kyr BP (right), and associated anticyclone (marked as H) over LIS between 10 and 9 kyr BP. The Greenland ice sheet did not show significant variation during the early Holocene. The triangles show the location of Blood Pond in northeastern North America. The green arrows indicate the trajectories of water vapor originated from Gulf of Mexico (see text for discussion). ! ! 184 SUPPLEMENTARY MATERIALS 1. METHODS We examined a sediment core from the Blood Pond, Massachusetts, the United States. The age model of the core was established by C-14 dating of bulk sediments and plant macrofossils and published previously (Hou et al., 2006). Ages are reported as thousands of calibrated years before present (kyr BP). Average sedimentation rate for the study section in this study is 1.2 mm/yr. Each sample spans 1 cm, and the distance between two neighboring samples is 2 cm, which leads to the average sampling resolution of ~25 yr. Compound-specific hydrogen isotope analysis of n-alkanoic acids was performed using gas chromatography – isotope ratio monitoring mass spectrometry. Each sample was run in triplicate and the standard deviation of all samples is <± 2‰. Method details and the isotope are available in supplemental material. Detailed procedures and instrumental conditions for the isotopic analyses are reported previously (Huang et al., 2002, 2004). In order to detect the duration and amplitude of the abrupt climate reversal, we performed a robust, statistical method to detect the anomaly in order to avoid highly inflated values in presence of extremes (Lanzante, 1996). A Fortran 90 program CLIM- X-DETECT (Mudelsee, 2006), specifically designed for the purpose of anomaly detection, is used to calculate the running median (2k+1 window point) as estimator of the time-dependent trend and the running median of absolute distances to the median (MAD) as estimator of the time-dependent variability. The 95% confidence band, which ! ! 185 is employed to define the extremes detection threshold, is given by median ± 2.96MAD. The duration of an anomaly is given by the time points that the detection threshold is crossed. The size of an anomaly is the maximum of the peak value minus median, divided by the MAD. We selected k = 45, which leads to average window width on the order of 1000 years. It permits us to explore millennial-scale background and variability. 2. CHRONOLOGY The chronology of the Blood Pond sediment core during the early Holocene was based on five radiocarbon dates (Figure S1). ! ! 186 Figure S1. Five radiocarbon dates provide the chronology for the Blood Pond sediment core during the early Holocene. ! ! 187 Figure S2. Comparison between Blood Pond hydrogen isotope records during the early Holocene and records from the late Pleistocene. The late Pleistocene records show prominent variability, while the early Holocene records show less variability. However, two abrupt climate shifts (9.2 and 8.4 ka) were shown during the early Holocene. ! CHAPTER 6 Solar activity induced pronounced temperature oscillations during the late Pleistocene to the early Holocene in the northeastern North America Co-author: Yongsong Huang To be submitted to Science 189 190 191 Sensitivity of the Earth’s climate to the solar activity has profound implications for modeling future climate changes, but is a subject of intense scientific debate. GCM modeling considering solar forcing can successfully simulate climatic variations in the past millennia such as the Little Ice Age (LIA), but volcanic forcing during the same time period can also yield similar climatic effect and confound interpretations. Here we present decadal to centennial resolution compound-specific hydrogen isotopic records from two lakes in the northeastern North America to investigate the relationship between solar activity and temperature changes during the late Pleistocene to early Holocene. Our temperature reconstructions from the two lakes 100 km apart in New England are highly consistent with each other and agree well with established general climatic scenarios. More importantly, our records contain centennial-scale cyclicities related to the solar cycles (88 and 232 yr), indicating strong links between the pronounced 1 to 2oC temperature oscillations and solar activity during the late Pleistocene. Our results strongly support the presence of an internal amplification mechanism for the solar forcing that is capable of causing disproportionally large climatic responses with relative small changes in the incoming solar radiation. 192 The causes and controlling mechanisms for the Earth climatic change on different time scales have been subjected to intensive study in order to better predict future climatic changes. On relatively long time scale (several thousands to tens and hundreds of thousands years) in the past few million years, changes in the Earth orbital configuration played a dominant role in the prominent climatic oscillations such as the glacial – interglacial variations (e.g., Imbrie et al., 1992). However, the mechanisms for the climate changes of relatively short time (e.g., decadal to centennial), which are more relevant to the human society, remain much less understood. Solar forcing has been proposed to be one plausible mechanism for the decadal- to centennial- scale climate changes, such as those observed during the well known little ice age (LIA, mid-1600s to early 1700s) and medieval warm period (MWP, 1100s-1500s) (Bard et al., 2000). However, the impact of solar variability on climate changes has been highly controversial, because of the minute amplitude of solar irradiance variation over decadal to centennial time scales (e.g., decrease by only ~ 0.1 % of the total irradiance during LIA; Bard et al., 2000). Although solar-forced climate model have successfully simulated the climate changes on northern hemisphere continents during LIA by involving changes of ozone chemistry and the North Atlantic Oscillation or NAO (Shindell et al., 2001), the credibility of the solar forcing of the LIA is questioned by the evidence for the coincident enhanced volcanic activity (e.g., Mann et al., 2005; Briffa et al., 1998). Increased volcanic eruptions would produce similar cooling effects during the LIA as changes in solar forcing, making the two factors virtually impossible to distinguish. 193 Here we present new evidence for prominent solar forcing of the temperature variations in New England based on a high-resolution hydrogen isotopic record from Rocky Pond, MA (Fig.1). Rocky Pond sediment has high sedimentation rate (2 m/kyr) during the late Pleistocene to early Holocene transition, allowing decadal scale down core analysis (see Fig.S1 in Supplementary Material). The modeling results of Shindell et al. (2001) indicate that the Northeastern North America is an exceptionally sensitive region to the solar variability on Earth, because of internal amplification mechanisms related to the North Atlantic Oscillation. For example, the relatively small variations in solar irradiance during the LIA (~0.1 %) induced up to 1.0 °C temperature change, especially during the winter (Shindell et al., 2001). However, so far there have been no paleo- temperature records from northeastern North America beyond the last 1000 years that have sufficient time resolution to detect the solar forcing on the climate change, and to support the model results. Because our new isotopic record is from the late Pleistocene to early Holocene, we avoid the overlap with volcanic forcing factor during the LIA that has so far confounded the interpretation (Briffa et al., 1998). In this paper, we will first demonstrate that our hydrogen isotopic data from Rocky Pond record the regional temperature changes during the late Pleistocene to early Holocene, and are consistent with previous published results. We will then show that the spectral components in our records are related to the solar variability. Our results strongly support that solar cycles can indeed drive disproportionably large climatic oscillations through Earth’s internal feedback mechanisms. 194 We measured hydrogen isotope (D/H) ratios of C22 n-alkanoic acid (behenic acid) from solvent extractable fraction and sediment bound C16 n-alkanoic acid (palmitic acid) from the Rocky Pond sediment core (Huang et al., 2004). D/H ratios of C22 n-alkanoic acid (!DBA) have been demonstrated to be a reliable temperature proxy (!DBA = 4.3T - 208.4‰, Hou et al., 2007). D/H ratios of bound C16 n-alkanoic acid (!DPA) from lake surface sediment track lake water D/H ratios with constant isotopic fractionation along lake transects in eastern North America (Huang et al., 2004). !DPA also show strong correlation with annual mean temperature variability (!DPA = 3.3 T – 202, R2 = 0.95, p <0.001. See supplementary materials for method and data). Temperature variations during the Pleistocene – early Holocene transition inferred from C22 and C16 n-alkanoic acids at Rocky Pond show remarkable similarity to those of Blood Pond and record well-known climate fluctuations (Fig.2). For example, the Bølling and Allerød warm periods were indicated by the increase in !D values and temperature. The short cool periods, Older Dryas (OD, ~14.1 ka) and Intra-Allerød cold period (IACP, ~13.5 ka) during the late Pleistocene, which were indicated by single samples in Blood Pond records, were clearly revealed in Rocky Pond records (Fig.2). The amplitude of OD and IACP were similar, ~3 oC cooler than Allerød warm period. The most significant temperature variability occurred at the beginning and end of the Younger Dryas (YD). All records show about ~6-7 °C cooler than the early Holocene. The Preboreal Oscillation (PBO), an abrupt and short cool period around 11.4 ka, was also revealed by multiple samples in Rocky Pond records, which is similar to OD and IACP. Temperature increased ~ 6 oC abruptly at the end of YD. Following PBO, temperature increased ~ 6 oC 195 gradually until 10 ka. The consistent results from two different lakes further support that !DBA and !DPA are reliable temperature proxies and record regional temperature variability. There are some differences between the temperature records from Rocky Pond and Blood Pond. Notably, the Rocky Pond temperatures are generally ~2oC higher than Blood Pond. The temperature difference is likely due to the proximity of Rocky Pond to the Atlantic coast (100 km). The modern temperature around Rocky Pond is 1.5 °C higher than Blood Pond. Blood Pond record also shows slightly smaller temperature fluctuations during OD, Allerød warm period and IACP than Rocky Pond record. This probably results from the lower temporal resolution of the Blood Pond records. In addition, based on independent radiocarbon dates, the end of YD at Rocky Pond is ~ 11.9 ka, about 100 years earlier than the Blood Pond record, which the YD ends ~11.8 ka. We will discuss this chronology issue in the next section. Chronology for the two lake sediment cores was established independently by 5 radiocarbon dates from two lakes, respectively, between 15 to 10 ka (Fig.2). Ages between individual dates are linearly interpolated. Unfortunately, one of the dates (11995.5±245.5 yr) for the Rocky Pond prior to the termination of YD has large errors due to the coincidence with the radiocarbon plateau (Goslar et al., 1999). We have previously compared with our Blood Pond chronology and isotopic records with pollen data (Hou et al., 2007). Our results indicate Blood Pond chronology is accurate, especially for the termination of the YD. In order to avoid inconsistencies for the spectral analysis, we tuned the end of YD as indicated by !DBA at Rocky Pond to the same timing as Blood Pond (Fig.2). The revised chronology for Rocky Pond is consistent with 196 regional paleoecological and paleoclimate records (e.g., Hou et al., 2007; Shuman et al., 2002; Yu et al., 1997; Ellis et al., 2004). The !D records from Rocky Pond were re- plotted using the tuned chronology in Fig.1s in the supplementary materials. We performed singular spectrum analysis (multi-taper method) on the temperature records from both Rocky and Blood Pond. Both !DBA and !DPA records from Rocky Pond show prominent spectral peaks centered at 504-, 232-, 131- and 88 yr (Fig.3A, B). Temperature inferred from C22 n-alkanoic acid show additional peaks compared with C16 n-alkanoic acid record, centering at 805-, 427-, 299-, and 105-yr. Temperature records from Blood Pond also show spectral peak at 427- and 232 yr (Fig.3). The absence of 88 yr cycle at Blood pond is due to the relatively low time resolution (~80 yr). The spectra are confirmed by the Gaussian band-passing of all temperature records from Rocky and Blood Ponds (see Fig.S3 in supplementary materials). The spectra peaks (88 and 232 yr) shown in Rocky and Blood Pond temperature records are well documented by the atmospheric 14C production record, which is a proxy for solar variability (Damon and Sonett, 1990; Solanki et al., 2004, Fig.3). Moreover, the temperature variation from the northeastern North America appears to be coherent with time series of the cosmogenic nuclides, 14C (Fig.4). Notably, the 232 yr cycle in both lakes shows stronger amplitude during the middle of Younger Dryas, whereas the 232 yr cycle in 14C production record shows small amplitude and opposite trend during the same interval. This suggests a higher sensitivity of the regional climate during colder times. 197 The temperature in northeastern North America increased during periods of strong solar activity, as indicated by the lower 14C production (Fig.4). The lower temperature was correlated with the periods of less solar activity. Our results suggest that the centennial- scale climate variability during the late Pleistocene to early Holocene from northeastern North America is closely related to solar activity. The mechanisms how the relatively small amplitude irradiation change of the sunspot cycle may cause disproportionally large climatic response in the northern hemisphere continents have been discussed by Shindell et al. (1999, 2001) and Bard et al. (2000), Bard and Frank (2006). During the Little Ice Age, for example, the global average temperature changes that can be modeled by solar forcing alone are relatively small (~ 0.3 to 0.4oC lower during the Maunder Minimum). However, a key component for the amplified temperature change in the northern hemisphere continents is the North Atlantic Oscillation (NAO) (Shindell et al., 1999; Shindell et al., 2001). The reduced solar irradiance leads to lower tropical and subtropical sea surface temperature (SST), reduced latitudinal temperature gradient at 100 to 200 mbar, and diminished stratospheric westerly winds. These changes then result in decreased angular momentum transport to high latitudes and reduced tropospheric westerlies and associated temperature and pressure changes corresponding to a low NAO/AO (Shindell et al., 1999a; Shindell et al., 2001a). The dynamic feedbacks of ozone chemistry in troposphere are also involved in the mechanism to enhance or dampen the climate response to solar forcing (Shindell et al., 1999b; Shindell et al., 2001b). This is because the solar irradiation variation is not spectrally averaged: the relative amplitude of the UV variability is an order of magnitude 198 larger than the total irradiance changes. When solar input increases, the strong UV variations lead to enhanced ozone formation in the stratosphere (Gille et al., 1984) and excess UV adsorption by ozone, which causes additional heating the stratosphere. Therefore, changes in ozone chemistry are a positive feedback to the solar forcing. In the high-index NAO (positive phase), the eastern United States coast experiences mild and wet winter because the increased atmospheric pressure difference between subtropical high and polar low deviates the storm track on a more northerly track. On the other hand, in the low-index NAO (negative phase), the eastern United States coast experience much cold climate because the reduced press gradient bring the storm track more west-east pathway (Hurrell et al., 2003). Although the NAO is a mode of variability internal to the atmosphere, model results and paleoclimate records suggest the mean state of NAO also show centennial-scale variability. For example, lower SST records from Sargasso Sea and warm temperature in the area off Newfoundland during 16th to 19th centuries coincides with the reduced NAO during the same period (Keigwin and Pickart 1999). The GCM simulations also suggest low NAO index during the Maunder Minimum (Shindell et al., 2001b). Therefore, the decadal to centennial-scale temperature variability in northeastern North America can be attributed to solar irradiance changes combined with the changes in ozone chemistry, but strongly amplified by the changes in the mean state of NAO. 199 Our records provide the first evidence for a robust connection between solar irradiance and centennial-scale temperature variability during the late Pleistocene to early Holocene. In contrast to the records for the past millennia that contain the confounding effects of solar and volcanic forcing, our records extends ~5000 years and eliminates the fortuitous coincidence between the two confounding factors. Therefore, our results strongly supports that the relatively small solar forcing can lead to significant decadal to centennial scale climate change, especially on northern hemisphere continents. The amplification appears to be much stronger during the colder times (such as during the middle of YD in this study). 200 REFERENCE Bard E. and Frank M. (2006) Climate change and solar variability: what's new under the Sun?. Earth and Planetary Science Letters 248, 1-14, doi:10.1016/j.epsl.2006.06.016. Bard E., Raisbeck G., Yiou F. and Jouzel J. (2000) Solar irradiance during the last 1200 years based on cosmogenic nuclides. Tellus 52B, 985-992. Briffa K.R., Jones P.D., Schweingruber F.H. and Osborn T.J. (1998) Influence of volcanic eruptions on Northern Hemisphere summer temperature over the past 600 years. Nature 393, 450-455. Damon P.E. and Sonett C.P. (1991) Solar and terrestrial components of the atmospheric 14C variation spectrum, in Sonett, C.P., Giampapa, M.S., and Matthews, M.S., eds., The Sun in time: Tucson, The University of Arizona Press, 360-388. Ellis K.G., Mullins H.T. and Patterson W.P. (2004) Deglacial to middle Holocene (16,000 to 6000 calendar years BP) climate change in the northeastern United States inferred from multi-proxy stable isotope data, Seneca Lake, New York. Journal of Paleolimnology 31, 343-361. Gille J.C., Smythe, C.M. and Heath D.F. (1984) Observed ozone response to variations in solar ultraviolet radiation. Science 225, 315-317. 201 Goslar T., Wohlfarth B., Bjorck S., Possnert G. and Bjorck J. (1999) Variations of atmospheric 14C concentrations over the Allerod - Younger Dryas transition. Climate Dynamics 15, 29-42, doi:10.1007/s003820050266. Hou J., Huang Y., Oswald W.W., Foster D.R. and Shuman B. (2007) Centennial-scale compound-specific hydrogen isotope record of Pleistocene-Holocene climate transition from southern New England. Geophysical Research Letters 34, doi:10.1029/2007GL030303. Huang Y., Shuman B., Wang Y. and Webb T.III. (2004) Hydrogen isotope ratios of individual lipids in lake sediments as novel tracers of climatic and environmental change: a surface sediment test. Journal of Paleolimnology 31, 363-375. Hurrell J.W., Kushnir Y., Ottersen G. and Visbeck M. (2003) An overview of the North Atlantic Oscillation, in Hurrell, J.W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation - Climatic significance and environmental impact: (Geophysical Monograph; 134): Washington DC, American Geophysical Union, 1-35. Imbrie J., Boyle E.A., Clemens S.C., Duffy A., Howard W.R., Kukla G., Kutzbach J., Martinson D.G., McIntyre A., Mix A.C., Molfino B., Morley J.J., Peterson L.C., Pisias N.G., Press W.L., Raymo M.E., Shackleton N.J. and Toggweiler J.R. (1992) On the structure and origin of major glaciation cycles. I. Linear response to Milankovitch forcing. Paleoceanography 7, 701-738. 202 Keigwin L.D. and Pickart R.S. (1999) Slope water current over the Laurentian Fan on the interannual to millennial time scales. Science 286, 520-523, doi:10.1126/science.286.5439.520. Mann M.E., Cane M.A., Zebiak S.E. and Clement, A. (2005) Volcanic and Solar Forcing of the Tropical Pacific over the Past 1000 Years. Journal of Climate 18, 417 - 456. Shindell D.T., Schmidt G.A., Miller R.L. and Rind D. (2001a) Northern Hemisphere winter climate response to greenhouse gas, ozone, solar and volanic forcing. Journal of Geophysical Research 106, 7193-7210. Shindell D.T., Schmidt G.A., Mann M.E., Rind D. and Waple A. (2001b) Solar forcing of regional climate change during th Maunder Minimum. Science 294, 2149-2152, doi:10.1126/science.294.5528.2149. Shindell D.T., Miller R.L., Schmidt G.A. and Pandolfo L. (1999a) Simulation of recent northern winter climate trends by greenhouse-gas forcing. Nature 399, 452-455. Shindell D.T., Rind D., Balachandran N., Lean J. and Lonergan P. (1999b) Solar cycle variability, ozone, and climate: Science 284, 305-308, doi:10.1126/science.284.5412.305. Shuman B., Bartlein P.J., Logar N., Newby P. and Webb T.III. (2002) Parallel climate and vegetation responses to the early Holocene collapse of the Laurentide Ice Sheet. Quaternary Science Reviews 21, 1793-1805. 203 Solanki S.K., Usoskin I.G., Kromer B., Schussler M. and Beer J. (2004) Unusual activity of the Sun during recent decaes compared to the previous 11,000 years. Nature 431, 1084-1087, doi:10.1038/nature02995. Stuiver M., Reimer P.J., Bard E., Beck J.W., Burr G.S., Hughen K.A., Kromer B., McCormac F.G., Plicht J.v.d. and Spurk M. (1998) INTCAL98 radiocarbon age calibration, 24,000-0 cal BP. Radiocarbon 40, 1041-1083. Yu Z., McAndrews J.H. and Eicher U. (1997) Middle Holocene dry climate caused by change in atmospheric circulation patterns: Evidence from lake levels and stable isotopes. Geology 25, 251-254. Zanchettin D., Aubino A., Traverso T. and Tomasino M. (2008) Impact of variations in solar activity on hydrological decadal patterns in northern Italy. Journal of Geophysical Research 113, D12102, doi:10.1029/2007JD009157. 204 Figure 1. Location of the two lakes in northeastern North America. The two lakes are about 100 km apart, with Rocky Pond is close to the coast. 205 206 Figure 2. Hydrogen isotope records and inferred temperature from lakes in northeastern North America, (A) Rocky Pond C16 n-alkanoic acid; (B) Rocky Pond C22 n- alkanoic acid; (C) Blood Pond C22 n-alkanoic acid. The Y-axes at the left side are of hydrogen isotopic scale, the Y-axes at the right side are of temperature scale which were calculated based on the correlation between !DBA, !DPA and mean annual surface air temperature (T) (!DBA = 4.3T -208.4‰, !DPA = 3.3 T – 202). The GISP2 !18O records are also shown for comparison (D). 207 Figure 3. Spectral components of the temperature records from lakes in northeastern North America, (bottom to top), Rocky Pond C16, Rocky Pond C22, and Blood Pond C22. The sunspot number records from Solanki et al. (2004) is also shown for comparison. All spectral analyses were performed by Singular Spectrum Analysis - MultiTaper Method (SSA-MTM) Toolkit. 208 Figure 4. Band pass of two prominent solar cycles (232 and 88 years) of the Blood Pond C22, Rocky Pond C16, C22 and radiocarbon data (Stuiver et al., 1998). The hydrogen isotope data (orange curves) are also plotted to show the corresponding cyclicities. Notably, the 232 yr cycle in both lakes shows stronger amplitude during the middle of Younger Dryas, whereas the 232 yr cycle in 14C production record shows small amplitude and opposite trend during the same interval. 209 SUPPLEMENTARY MATERIALS 1. METHODS We examined lower part of the sediment core (15-10 ka) from Rocky Pond in northeastern North America (Fig.1). The age model of the core was established by five radiocarbon dates. The sedimentation rate was very high during the study period (~ 2mm/yr) (Fig.1). All sediment and leaf samples were freeze-dried. Sediment samples were extracted using an Accelerated Solvent Extractor (ASE200, Dionex) with dichloromethane: methanol (9:1 v/v) at 150oC and 1200 psi for three 15-min cycles to get solvent-extractable lipids. Approximately 0.5-1 g of each dry sample was extracted. The bound lipids from sediment were released by saponifying the ASE extracted sediments under reflux using 2 N KOH/(Methanol + 5%water) (e.g., Meyers and Ishiwatari 1993). The solution was then acidified and extracted with hexane. The total lipid extract obtained from sediment was separated into neutral and acid fractions using Supelco® Supelclean™ LC-NH2 silica gel using dichloromethane : isopropyl alcohol (2:1 v/v) and ether with 4% acetic acid (v/v), respectively. Acid fractions were methylated with 5% anhydrous HCl in methanol at 60 oC for 12 hours. Hydroxyl acids were removed by eluting the samples through silica gel columns using hexane, in order to further purify the fatty acid methyl esters to avoid chromatographic coelution. Quantification and identification of compounds was performed by Gas Chromatography - Flame Ionization Detection (GC-FID) and Gas Chromatography-Mass Spectrometry (GC-MS) (HP 6890, Agilent). An HP 6890 GC interfaced to a Finnigan 210 DeltaPlus XL stable isotope spectrometer through a high-temperature pyrolysis reactor was used for hydrogen isotopic analysis (Huang et al., 2004). The compounds separated by the GC were pyrolyzed to H2 and CO at 1445oC. A tank of ultra high purity hydrogen gas with known !D values was used as the isotope standard during the measurements. The H3+ factor was determined daily prior to sample analysis (average value was 2.5 during the course of this study). The accuracy for the instrument was routinely checked by an injection of laboratory isotopic standards (C16, C18, C22, C24 n-acid methyl esters) between every six measurements. The precisions (1") for the four laboratory standards were < ±2‰ throughout the entire process. The precision (1") for triplicate analyses of all samples was < ±2‰. !D values obtained from individual acids (as methyl esters) were corrected by mathematically removing the isotopic contributions from added groups before reporting. The !D value of the added methyl group was determined by acidifying and then methylating (along with the samples) the disodium salt of succinic acid with a predetermined !D value (using TC/EA-IRMS) (Huang et al., 2002). 2. CALIBRATION OF SEDIMENT BOUND C16 n-ALKANOIC ACID !D AS TEMPERATURE PROXY We determined the relationship between !DPA and T along the N-S transect across a temperature gradient (2 - 23oC) in eastern North America (Fig.S1). !DPA = 3.3 T – 202, R2 = 0.95, p <0.001 (1) Based on this relationship, a 1oC change in T corresponds to ~ 3.3 ± 0.2‰ variation in !DBA. 211 The temperature dependence of !DPA (3.3 ‰/oC) is smaller than that of !DBA (4.3‰/oC) along the same transect in North America (Hou et al., 2007). This may be attributed to the sources of bound lipids and their isotopic fractionation during the production and diagenesis. Cranwell (1981) showed that the bound lipids in lake sediments were mainly produced from algal and microbial sources. Furthermore, Albaigés et al. (1984) demonstrated that the isotopic fractionation of bound lipids from microbial sources is much larger than the algal sources, which result in smaller isotopic fractionation factor. Nevertheless, the significant correlation between !DPA and surface air temperature along the modern lake transect suggest that !DPA could be a reliable proxy for surface air temperature. We estimated and compared the temperature variability from both C22 and C16 n-alkanoic acids from Rocky Pond during the late Pleistocene, which was also compared with regional temperature records. 212 Figure S1. The Rocky Pond chronology based on five radiocarbon dates (left), and the sedimentation rate (right). 213 Figure S2. Map showing the location of the north-south transect of lakes (triangles) for calibration of C16 n-alkanoic acid !D values in eastern North America, the GNIP stations (solid circles) for comparison with temperature dependence of precipitation !D variation, and the study sites in the paper, Blood Pond and Rocky Pond (stars). 214 Figure S3. Calibration of !D values of sediment bound C16 n-alkanoic acid as a proxy for surface temperature. The transect of the lakes are shown in Figure S2. The relationship between !DPA and temperature is y = 3.3x-202, which indicates the temperature dependence of the !DPA is 3.3‰/°C. 215 Figure S4-A 216 Figure S4-B 217 Figure S4-C Figure S4. Band pass (Gaussian filter) of temperature records inferred from Rocky Pond C22 (A), C16 (B), and Blood Pond C22 (C). The overlap of band passed signal and original signal confirm the prominent spectral components in the temperature records from the two lakes. 218 Figure S5. Band pass (Gaussian filter) of Rocky C22, C16 !D records, Blood Pond C22 !D records and 14C records (Stuvier et al., 1998). The in phase variation of all records suggests that the temperature records related with solar activity. ! 220! ! 221! Appendix 1. Chapter 4: Hydrogen isotope and inferred temperature records during the late Pleistocene to early Holocene from Blood Pond, northeastern North America. ! Standard Temperature Depth Age C22 n-acid !D Deviation of 3 inferred from Sample I.D. measurements C22 !D (cm) (yr BP) (‰, VSMOW) (‰, VSMOW) (°C) BP1418 1418 16320 -164 1.8 10.4 BP1410 1410 16135 -157 0.2 12.1 BP1402 1402 15950 -149 1.6 13.9 BP1394 1394 15765 -161 1.8 11.1 BP1386 1386 15580 -143 1.5 15.3 BP1378 1378 15395 -154 1.8 12.5 BP1370 1370 15210 -151 1.5 13.4 BP1362 1362 15025 -143 1.5 15.1 BP1354 1354 14840 -144 1.5 15.1 BP1346 1346 14675 -147 1.5 14.3 BP1338 1338 14544 -145 1.7 14.7 BP1330 1330 14413 -155 1.0 12.4 BP1322 1322 14283 -150 1.2 13.7 BP1314 1314 14152 -152 1.2 13.1 BP1306 1306 14021 -151 2.0 13.4 BP1298 1298 13890 -150 0.8 13.5 BP1290 1290 13759 -151 1.3 13.4 BP1282 1282 13628 -150 1.7 13.6 BP1274 1274 13497 -155 1.9 12.5 BP1266 1266 13366 -150 1.3 13.6 BP1258 1258 13297 -146 1.7 14.6 BP1250 1250 13237 -160 1.1 11.2 BP1242 1242 13176 -160 1.9 11.2 BP1234 1234 13116 -168 1.2 9.4 ! 222! BP1226 1226 13056 -168 1.2 9.4 BP1218 1218 12996 -165 1.4 10.1 BP1210 1210 12935 -167 1.5 9.7 BP1202 1202 12875 -159 1.2 11.6 BP1194 1194 12815 -162 1.8 10.8 BP1186 1186 12755 -168 2.0 9.3 BP1178 1178 12694 -168 1.2 9.4 BP1170 1170 12634 -162 0.1 10.7 BP1162 1162 12574 -155 1.5 12.5 BP1154 1154 12514 -168 1.6 9.4 BP1146 1146 12453 -156 1.7 12.1 BP1138 1138 12393 -159 0.9 11.5 BP1130 1130 12333 -173 1.0 8.1 BP1122 1122 12273 -157 1.1 12.0 BP1114 1114 12212 -171 0.2 8.8 BP1106 1106 12152 -167 1.8 9.6 BP1098 1098 12092 -168 1.4 9.4 BP1090 1090 12031 -167 0.4 9.5 BP1082 1082 11971 -149 0.9 13.7 BP1074 1074 11911 -160 1.8 11.3 BP1066 1066 11851 -172 0.6 8.4 BP1058 1058 11790 -163 1.5 10.5 BP1050 1050 11730 -152 1.9 13.1 BP1042 1042 11670 -146 1.6 14.6 BP1034 1034 11610 -159 0.2 11.5 BP1026 1026 11549 -163 0.1 10.6 BP1018 1018 11489 -157 1.8 12.0 BP1010 1010 11429 -157 0.6 11.9 BP1002 1002 11369 -141 1.7 15.6 BP994 994 11308 -153 1.5 13.0 ! 223! BP986 986 11248 -152 2.0 13.2 BP978 978 11188 -158 1.5 11.7 BP970 970 11128 -154 1.6 12.6 BP962 962 11067 -149 1.7 13.7 BP954 954 11007 -152 1.8 13.0 BP946 946 10958 -156 0.4 12.1 BP938 938 10910 -152 1.1 13.0 BP930 930 10862 -157 1.3 11.9 BP924 924 10826 -156 0.5 12.2 BP916 916 10778 -150 1.0 13.6 BP908 908 10730 -151 1.9 13.2 BP900 900 10683 -155 1.5 12.5 BP892 892 10635 -153 1.1 12.8 BP884 884 10587 -158 1.8 11.6 BP876 876 10539 -154 1.4 12.6 BP868 868 10491 -152 1.7 13.0 BP860 860 10443 -146 1.5 14.5 BP852 852 10396 -147 1.9 14.3 BP844 844 10348 -150 0.4 13.5 BP836 836 10300 -152 1.5 13.0 BP828 828 10252 -161 2.0 11.0 BP820 820 10204 -152 0.2 13.0 BP812 812 10156 -161 2.0 11.0 BP804 804 10108 -164 1.8 10.4 BP796 796 10062 -151 0.9 13.4 BP788 788 10017 -141 2.0 15.6 BP780 780 9972 -149 1.7 13.7 BP764 764 9881 -145 1.7 14.9 BP756 756 9836 -147 2.0 14.4 BP740 740 9746 -158 1.7 11.8 ! 224! BP706 706 9554 -154 1.5 12.7 BP698 698 9509 -153 1.4 12.9 BP690 690 9464 -155 1.5 12.3 BP682 682 9418 -150 1.2 13.5 BP674 674 9373 -147 0.4 14.3 BP666 666 9328 -154 1.9 12.7 BP658 658 9283 -157 0.6 11.9 BP650 650 9238 -148 1.9 14.1 BP642 642 9183 -140 1.4 15.9 BP634 634 9111 -146 0.7 14.6 BP626 626 9040 -148 1.1 14.1 BP618 618 8968 -151 1.2 13.4 BP610 610 8897 -156 1.2 12.1 BP602 602 8825 -153 0.9 12.8 BP594 594 8754 -154 0.7 12.7 BP586 586 8682 -149 1.1 13.7 BP578 578 8611 -159 1.5 11.6 BP570 570 8539 -148 0.9 14.0 BP562 562 8468 -157 0.3 12.0 BP554 554 8396 -150 1.1 13.6 BP546 546 8325 -146 1.1 14.5 BP538 538 8253 -150 1.6 13.7 BP530 530 8181 -149 1.2 13.8 BP522 522 8110 -151 1.3 13.3 BP514 514 8038 -150 1.2 13.7 ! ! ! 225! Appendix 2. Chapter 5: Hydrogen isotope and inferred temperature records during the early Holocene from Blood Pond, northeastern North America. ! C22 n-acid !D C28 n-acid dD C22 n- Temperature Standard C28 n- Standard Age acid !D inferred from Sample Depth Deviation of acid !D Deviation of C22 !D I.D. (cm) (‰, three (‰, three (yr BP) VSMOW) measurements (°C) VSMOW) measurements (‰, VSMOW) (‰, VSMOW) BP458 458 7236 -143 1.9 15.2 -162 0.6 BP460 460 7281 -143 1.8 15.2 -158 0.9 BP462 462 7327 -148 2.0 13.9 -158 0.8 BP464 464 7372 -151 1.5 13.3 -159 1.0 BP466 466 7417 -144 1.5 14.9 -160 0.6 BP468 468 7463 -152 1.9 13.1 -161 0.8 BP470 470 7508 -150 0.2 13.6 -157 0.3 BP472 472 7553 -153 0.9 13.0 -157 1.2 BP474 474 7599 -143 1.9 15.1 -153 1.5 BP476 476 7644 -146 1.5 14.6 -152 0.6 BP478 478 7689 -149 1.2 13.8 -153 0.7 BP480 480 7735 -151 1.7 13.4 -155 1.2 BP482 482 7752 -146 0.9 14.5 -155 1.2 BP484 484 7770 -147 0.9 14.4 -154 0.6 BP486 486 7788 -151 1.3 13.4 -156 2.0 BP488 488 7806 -147 1.9 14.3 -158 1.1 BP490 490 7824 -142 1.5 15.3 -157 1.1 BP492 492 7842 -143 2.0 15.3 -155 0.9 BP494 494 7860 -136 1.8 16.9 -153 1.5 BP496 496 7878 -153 1.4 12.9 -155 2.0 BP498 498 7895 -143 2.0 15.2 -156 0.7 BP500 500 7913 -147 2.0 14.3 -156 1.2 BP502 502 7931 -143 2.0 15.1 -153 1.4 ! 226! BP504 504 7949 -148 1.5 14.1 -154 1.0 BP506 506 7967 -142 1.3 15.3 -158 1.5 BP508 508 7985 -139 1.7 16.1 -157 1.7 BP510 510 8003 -140 1.8 15.8 -156 1.6 BP512 512 8021 -144 1.9 15.1 -157 1.4 BP514 514 8038 -141 2.0 15.6 -156 0.5 BP516 516 8056 -143 1.5 15.2 -156 1.0 BP518 518 8074 -146 1.1 14.5 -152 0.6 BP520 520 8092 -139 1.5 16.1 -150 0.4 BP522 522 8110 -143 1.2 15.2 -152 2.0 BP524 524 8128 -134 0.6 17.3 -152 2.0 BP526 526 8146 -143 1.8 15.3 -155 1.8 BP528 528 8164 -136 0.7 16.9 -156 1.2 BP530 530 8181 -141 1.8 15.7 -151 1.5 BP532 532 8199 -137 0.9 16.5 -159 1.6 BP534 534 8217 -145 1.6 14.8 -161 1.0 BP536 536 8235 -140 1.8 15.9 -159 1.6 BP538 538 8253 -142 2.0 15.6 -156 1.1 BP540 540 8271 -137 1.8 16.6 -152 1.3 BP542 542 8289 -137 1.9 16.6 -150 0.9 BP544 544 8307 -142 0.5 15.5 -149 1.5 BP546 546 8325 -138 2.0 16.4 -153 1.4 BP548 548 8342 -131 1.5 18.0 -146 1.1 BP550 550 8360 -148 1.3 14.1 -151 0.4 BP552 552 8378 -142 0.8 15.3 -148 2.0 BP556 556 8414 -152 1.5 13.2 -149 1.9 BP558 558 8432 -155 2.0 12.4 -153 0.4 BP560 560 8450 -138 0.2 16.4 -150 1.5 BP562 562 8468 -149 1.4 13.7 -151 0.8 BP564 564 8485 -142 0.9 15.5 -150 0.8 ! 227! BP566 566 8503 -141 1.7 15.6 -149 0.1 BP568 568 8521 -146 1.6 14.5 -150 0.7 BP570 570 8539 -140 0.9 16.0 -148 0.8 BP572 572 8557 -135 1.5 17.1 -150 0.2 BP574 574 8575 -140 1.2 15.8 -151 1.1 BP576 576 8593 -142 2.0 15.5 -151 2.0 BP578 578 8611 -151 2.0 13.3 -154 1.5 BP580 580 8628 -134 1.9 17.4 -152 0.6 BP582 582 8646 -144 1.2 14.9 -152 1.2 BP584 584 8664 -153 1.5 12.8 -149 0.9 BP586 586 8682 0.9 -143 0.6 BP588 588 8700 -147 1.5 14.2 -152 0.9 BP590 590 8718 -149 0.6 13.8 -155 1.1 BP592 592 8736 -146 0.1 14.4 -148 1.1 BP594 594 8754 -146 0.2 14.6 -155 1.6 BP596 596 8771 -145 0.8 14.8 -153 0.3 BP598 598 8789 -152 1.6 13.1 -154 1.1 BP600 600 8807 -153 1.9 13.0 -152 1.1 BP602 602 8825 -145 0.5 14.7 -155 0.3 BP604 604 8843 -165 2.0 10.1 -150 1.1 BP606 606 8861 -146 2.0 14.6 -150 1.7 BP608 608 8879 -153 0.9 12.8 -149 1.2 BP610 610 8897 -149 1.0 13.9 -153 0.5 BP612 612 8915 -147 1.5 14.3 -148 1.0 BP614 614 8932 -146 0.3 14.5 -148 1.3 BP616 616 8950 -150 0.7 13.5 -148 0.9 BP618 618 8968 -143 1.5 15.2 -151 1.4 BP620 620 8986 -161 1.3 11.1 -144 1.5 BP622 622 9004 -144 0.5 14.9 -144 1.3 BP626 626 9040 -140 1.5 16.0 -147 0.8 ! 228! BP630 630 9075 -150 1.9 13.5 -142 1.5 BP632 632 9093 -166 0.1 9.9 -148 1.9 BP634 634 9111 -137 2.0 16.6 -150 1.8 BP636 636 9129 -165 1.5 10.0 -145 1.6 BP638 638 9147 -166 1.3 10.0 -145 0.7 BP640 640 9165 -161 1.5 11.0 -140 0.3 BP642 642 9183 0.1 -144 0.6 BP644 644 9201 -170 1.0 8.8 -146 0.5 BP646 646 9218 -176 1.0 7.6 -145 2.0 BP648 648 9236 -160 1.1 11.3 -147 1.0 BP650 650 9238 -140 5.0 16.0 -147 0.8 BP652 652 9249 -158 0.6 11.7 -147 0.6 BP654 654 9260 -167 1.4 9.7 -147 1.4 BP658 658 9283 -149 0.4 13.7 -152 0.8 BP660 660 9294 -150 0.2 13.7 -152 1.6 BP662 662 9306 -145 0.9 14.7 -152 1.5 BP664 664 9317 -147 1.5 14.2 -154 0.9 BP666 666 9328 -146 2.0 14.6 -151 1.1 BP668 668 9339 -141 1.5 15.6 -151 0.4 BP670 670 9351 -141 1.6 15.7 -152 2.0 BP672 672 9362 -143 1.7 15.1 -151 1.6 BP674 674 9373 -139 1.8 16.2 -151 1.5 BP678 678 9396 -146 0.4 14.6 -150 1.2 BP680 680 9407 -144 1.1 14.9 -150 1.5 BP682 682 9418 -142 1.3 15.4 -152 1.2 BP684 684 9430 -150 0.5 13.5 -150 1.0 BP686 686 9441 -146 1.0 14.5 -152 1.6 BP688 688 9452 -143 1.9 15.2 -148 1.1 BP690 690 9464 -148 1.5 14.1 -151 1.3 BP692 692 9475 -153 1.1 12.9 -147 0.9 ! 229! BP694 694 9486 -145 1.8 14.7 -150 1.5 BP696 696 9497 -149 1.4 13.9 -151 1.4 BP698 698 9509 -145 1.7 14.7 -149 1.1 BP700 700 9520 -151 1.5 13.3 -147 0.4 BP702 702 9531 -167 1.9 9.6 -149 2.0 BP704 704 9543 -151 0.4 13.3 -138 1.9 BP706 706 9554 -165 1.5 10.0 -148 0.4 BP710 710 9576 -147 2.0 14.3 -149 1.5 BP712 712 9588 -150 0.2 13.5 -147 0.8 BP716 716 9610 -150 2.0 13.6 -149 0.8 BP718 718 9622 -153 1.8 12.8 -149 0.1 BP720 720 9633 -146 0.9 14.6 -150 0.7 BP722 722 9644 -146 2.0 14.4 -148 0.8 BP724 724 9656 -144 1.7 14.9 -149 0.2 BP726 726 9667 -145 1.2 14.8 -149 1.1 BP728 728 9678 -152 1.7 13.0 -148 2.0 BP730 730 9689 -145 1.7 14.6 -152 1.5 ! ! ! ! 230! Appendix 3. Chapter 6: Hydrogen isotope and inferred temperature records from Blood Pond, northeastern North America. ! !"#$%&'(!Hydrogen isotope of C22 n-acid and inferred temperature records from Blood Pond, northeastern North America.! ! C22 n-acid !D Temperature Standard Deviation inferred from C22 Depth Age C22 n-acid Sample of three !D (‰, !D I.D. measurements (cm) (yr BP) VSMOW) ‰, VSMOW) (°C) RP1075 1075 14572 -158 2.0 11.8 RP1070 1070 14549 -149 1.5 13.8 RP1065 1065 14525 -157 2.0 11.9 RP1040 1040 14409 -146 1.3 14.5 RP1035 1035 14386 -150 1.2 13.6 RP1030 1030 14362 -137 0.5 16.6 RP1025 1025 14339 -142 1.1 15.5 RP1020 1020 14316 -135 1.7 17.0 RP1016 1016 14297 -135 0.1 17.2 RP1005 1005 14246 -144 1.2 15.1 RP1000 1000 14223 -142 1.4 15.4 RP995 995 14199 -144 2.0 15.0 RP990 990 14176 -140 1.5 15.9 RP986 986 14157 -148 1.2 14.1 RP980 980 14129 -154 1.1 12.7 RP970 970 14083 -144 1.6 14.9 RP967 967 14069 -149 2.0 13.8 RP960 960 14036 -144 1.7 15.0 RP955 955 14013 -146 1.6 14.5 RP951 951 13994 -136 1.2 16.9 ! 231! RP940 940 13943 -145 2.0 14.7 RP935 935 13920 -141 2.0 15.7 RP930 930 13896 -140 2.0 16.0 RP925 925 13873 -152 1.4 13.2 RP921 921 13855 -147 1.6 14.3 RP915 915 13827 -151 1.2 13.3 RP905 905 13780 -144 2.0 15.0 RP900 900 13757 -144 1.2 14.9 RP895 895 13733 -149 1.2 13.8 RP891 891 13715 -142 1.5 15.6 RP885 885 13687 -147 0.4 14.2 RP880 880 13664 -137 2.0 16.7 RP872 872 13626 -151 2.0 13.3 RP865 865 13598 -148 0.3 14.1 RP860 860 13579 -147 2.0 14.2 RP856 856 13564 -156 2.0 12.3 RP853 853 13553 -156 0.4 12.3 RP842 842 13512 -152 2.0 13.1 RP837 837 13493 -154 2.0 12.7 RP832 832 13474 -149 1.6 13.7 RP827 827 13456 -149 1.4 13.8 RP823 823 13441 -152 1.3 13.1 RP818 818 13422 -150 1.8 13.7 RP808 808 13384 -148 1.1 14.1 RP803 803 13365 -147 1.0 14.3 RP798 798 13347 -147 2.0 14.3 RP794 794 13332 -146 1.0 14.6 RP788 788 13309 -143 2.0 15.3 RP783 783 13290 -142 1.4 15.4 RP775 775 13260 -150 1.0 13.5 ! 232! RP768 768 13234 -146 1.5 14.6 RP761 761 13208 -136 2.0 16.8 RP757 757 13193 -145 1.1 14.7 RP751 751 13170 -139 1.5 16.2 RP746 746 13152 -159 2.0 11.5 RP736 736 13114 -151 2.0 13.4 RP731 731 13095 -157 2.0 11.9 RP727 727 13080 -138 0.3 16.3 RP721 721 13058 -153 2.0 12.9 RP711 711 13020 -155 0.9 12.3 RP706 706 13001 -148 1.0 14.2 RP701 701 12983 -157 2.0 11.9 RP697 697 12968 -157 1.4 12.0 RP691 691 12945 -159 2.0 11.4 RP686 686 12926 -158 1.4 11.7 RP678 678 12896 -157 1.5 11.9 RP671 671 12848 -155 1.7 12.5 RP667 667 12818 -162 1.6 10.8 RP665 665 12803 -164 0.2 10.3 RP659 659 12758 -158 1.6 11.8 RP655 655 12729 -167 1.1 9.5 RP644 644 12647 -165 0.7 10.0 RP639 639 12610 -165 0.2 10.0 RP634 634 12572 -152 0.2 13.2 RP629 629 12535 -158 2.0 11.8 RP625 625 12505 -159 2.0 11.5 RP614 614 12423 -168 2.0 9.3 RP609 609 12386 -163 2.0 10.6 RP604 604 12349 -171 2.0 8.7 RP599 599 12312 -162 2.0 10.7 ! 233! RP595 595 12282 -164 2.0 10.4 RP589 589 12237 -166 0.7 9.9 RP579 579 12163 -154 1.2 12.6 RP576 576 12141 -158 0.1 11.8 RP569 569 12088 -169 1.1 9.1 RP565 565 12059 -160 0.1 11.2 RP560 560 12021 -164 1.2 10.2 RP556 556 11992 -166 1.4 9.8 RP545 545 11910 -163 1.5 10.5 RP540 540 11873 -150 1.3 13.6 RP535 535 11835 -155 2.0 12.5 RP530 530 11798 -145 2.0 14.7 RP526 526 11768 -141 1.1 15.6 RP520 520 11724 -163 1.0 10.5 RP510 510 11649 -155 1.0 12.4 RP505 505 11612 -159 1.0 11.6 RP500 500 11575 -156 1.0 12.2 RP496 496 11545 -162 1.1 10.8 RP490 490 11500 -145 1.8 14.7 RP480 480 11426 -159 1.0 11.6 RP477 477 11403 -148 2.0 14.0 RP470 470 11351 -154 2.0 12.6 RP466 466 11322 -143 1.5 15.3 RP461 461 11284 -132 2.0 17.8 RP457 457 11255 -127 2.0 18.8 RP446 446 11173 -126 1.2 19.2 RP441 441 11135 -140 1.1 16.0 RP436 436 11098 -130 2.0 18.1 RP431 431 11061 -137 2.0 16.5 RP427 427 11031 -138 1.5 16.3 ! 234! RP421 421 10987 -136 2.0 16.9 RP411 411 10912 -133 1.5 17.4 RP406 406 10875 -138 1.7 16.4 RP401 401 10838 -143 1.4 15.2 RP397 397 10808 -140 1.1 15.9 RP391 391 10763 -142 0.9 15.5 RP381 381 10689 -136 1.5 16.9 RP371 371 10578 -143 1.3 15.2 RP367 367 10520 -136 2.0 16.9 RP362 362 10446 -127 1.2 19.0 RP359 359 10402 -130 2.0 18.1 RP356 356 10358 -134 2.0 17.4 RP353 353 10314 -136 2.0 16.8 ! ! ! 235! !"#$%&'(!Hydrogen isotope of sediment bound C16 n-acid and inferred temperature records from Blood Pond, northeastern North America.! ! ! C16 n-acid !D Temperature Age C16 n-acid Standard Deviation Sample Depth inferred from C16 !D (‰, of three I.D. (cm) !D (yr BP) VSMOW) measurements ‰, VSMOW) (°C) RP1049 1049 14447 -157 0.2 13.5 RP1043 1043 14420 -151 1.8 15.3 RP1038 1038 14396 -157 0.4 13.6 RP1033 1033 14373 -158 1.1 13.2 RP1028 1028 14350 -153 0.6 14.9 RP1023 1023 14327 -160 1.1 12.7 RP1019 1019 14308 -165 2.0 9.6 RP1013 1013 14280 -156 1.2 14.1 RP1008 1008 14257 -155 1.3 14.3 RP1003 1003 14233 -151 1.5 15.5 RP998 998 14210 -157 1.2 13.5 RP993 993 14187 -165 1.5 11.2 RP989 989 14168 -151 0.2 15.5 RP983 983 14140 -168 2.0 10.4 RP978 978 14117 -167 1.0 10.5 RP973 973 14094 -159 1.1 13.2 RP970 970 14080 -152 0.4 15.2 RP963 963 14047 -151 1.0 15.4 RP958 958 14024 -151 1.0 15.6 RP954 954 14005 -146 1.6 17.0 RP948 948 13978 -150 0.9 15.8 ! 236! RP943 943 13954 -157 1.4 13.6 RP938 938 13931 -154 1.3 14.5 RP933 933 13908 -150 0.9 15.7 RP928 928 13885 -145 0.9 17.4 RP924 924 13866 -149 0.8 16.0 RP918 918 13838 -149 0.2 16.2 RP913 913 13815 -154 0.5 14.6 RP908 908 13791 -155 1.7 14.4 RP903 903 13768 -156 0.9 13.8 RP898 898 13745 -152 1.3 15.1 RP894 894 13726 -156 1.9 13.9 RP888 888 13698 -156 0.5 13.8 RP883 883 13675 -161 0.8 12.3 RP878 878 13652 -155 1.0 14.3 RP875 875 13638 -155 1.1 14.2 RP868 868 13608 -155 1.2 14.2 RP863 863 13589 -160 1.3 12.8 RP859 859 13574 -167 1.7 10.6 RP856 856 13563 -168 0.5 10.3 RP850 850 13540 -167 0.8 10.7 RP845 845 13521 -164 1.0 11.6 RP840 840 13503 -168 0.9 10.2 RP835 835 13484 -155 0.2 14.1 RP830 830 13465 -161 1.5 12.4 RP826 826 13450 -160 0.2 12.6 RP820 820 13427 -168 0.4 10.3 RP815 815 13409 -157 1.2 13.5 RP810 810 13390 -163 0.2 11.7 ! 237! RP805 805 13371 -159 0.9 13.1 RP800 800 13352 -159 0.6 13.0 RP796 796 13337 -168 1.0 10.4 RP790 790 13315 -156 1.6 14.0 RP785 785 13296 -154 0.7 14.5 RP780 780 13277 -150 0.1 15.7 RP777 777 13266 -152 1.3 15.0 RP770 770 13240 -145 1.3 17.2 RP765 765 13221 -155 1.9 14.2 RP764 764 13217 -164 1.2 11.4 RP760 760 13202 -155 0.5 14.2 RP754 754 13180 -159 1.9 13.0 RP749 749 13161 -165 1.5 11.3 RP744 744 13142 -164 1.2 11.5 RP739 739 13123 -158 1.0 13.5 RP734 734 13105 -149 1.6 16.0 RP730 730 13090 -164 0.9 11.5 RP724 724 13067 -166 0.7 11.0 RP719 719 13048 -159 1.4 13.1 RP714 714 13030 -150 0.1 15.7 RP709 709 13011 -166 1.8 11.0 RP704 704 12992 -154 0.6 14.6 RP700 700 12977 -153 0.3 14.9 RP694 694 12954 -170 0.1 9.6 RP689 689 12936 -164 2.0 11.5 RP684 684 12917 -180 2.4 6.6 RP681 681 12906 -165 1.8 11.2 RP674 674 12877 -165 0.9 11.1 ! 238! RP670 670 12857 -173 2.0 8.9 RP668 668 12847 -168 1.8 10.2 RP662 662 12816 -171 1.4 9.4 RP658 658 12796 -176 1.5 7.8 RP652 652 12765 -171 1.2 9.3 RP647 647 12740 -167 2.0 10.6 RP642 642 12715 -167 0.8 10.6 RP637 637 12689 -169 1.8 10.0 RP632 632 12664 -168 1.5 10.4 RP628 628 12643 -167 0.4 10.7 RP622 622 12613 -171 1.2 9.4 RP617 617 12588 -177 1.9 7.5 RP612 612 12562 -171 1.8 9.3 RP607 607 12537 -166 2.0 10.9 RP602 602 12511 -174 1.4 8.4 RP598 598 12491 -175 0.8 8.3 RP592 592 12461 -181 1.7 6.3 RP587 587 12435 -170 2.0 9.8 RP582 582 12410 -167 1.1 10.5 RP579 579 12394 -170 0.9 9.8 RP572 572 12359 -165 1.4 11.3 RP568 568 12339 -166 1.4 10.9 RP563 563 12313 -167 1.3 10.5 RP559 559 12293 -174 1.5 8.5 RP553 553 12262 -171 1.4 9.3 RP548 548 12237 -174 1.1 8.4 RP543 543 12211 -169 0.5 10.1 RP538 538 12186 -170 0.3 9.7 ! 239! RP533 533 12161 -170 0.2 9.6 RP529 529 12140 -175 1.4 8.3 RP523 523 12110 -189 1.6 3.9 RP518 518 12084 -173 0.2 8.7 RP513 513 12059 -168 1.0 10.4 RP508 508 12034 -181 0.2 6.2 RP503 503 12008 -166 0.5 11.0 RP499 499 11979 -170 1.7 9.5 RP493 493 11914 -165 0.4 14.3 RP488 488 11860 -159 1.3 13.0 RP483 483 11806 -156 0.6 13.9 RP480 480 11773 -161 1.1 12.6 RP473 473 11697 -153 0.1 14.8 RP468 468 11643 -164 0.3 13.0 RP464 464 11600 -168 1.4 11.8 RP460 460 11557 -167 0.5 12.2 RP454 454 11492 -158 1.8 14.4 RP449 449 11437 -161 0.2 11.6 RP444 444 11383 -165 1.8 10.4 RP439 439 11329 -160 1.4 11.9 RP434 434 11275 -156 0.7 14.1 RP430 430 11231 -158 2.0 13.5 RP424 424 11166 -154 1.9 14.5 RP419 419 11112 -154 1.2 14.5 RP414 414 11058 -153 1.2 14.9 RP409 409 11004 -150 1.9 15.6 RP404 404 10950 -149 0.2 16.2 RP400 400 10906 -150 1.5 15.8 ! 240! !