USF Libraries
USF Digital Collections

Tropical Atlantic and Caribbean climate variations during the past eight centuries

MISSING IMAGE

Material Information

Title:
Tropical Atlantic and Caribbean climate variations during the past eight centuries
Physical Description:
Book
Language:
English
Creator:
Kilbourne, Kelly Halimeda
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Coral
Oxygen isotope
Stontium
Radiocarbon
Little ice age
Dissertations, Academic -- Marine Science -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The Caribbean Sea is a key region from which to generate paleoclimate records because instrumental temperature data indicate that surface temperatures in the Caribbean region are correlated with global surface temperature. Heat and salt fluxes in the Caribbean have been implicated in major reorganizations in Atlantic Meridional Overturning Circulation on glacial time scales Schmidt et al., 2004 and it has been proposed that the Tropical Atlantic (including the Caribbean Sea) may play a role in smaller-scale changes in the Atlantic basin Vellinga and Wu, 2004. A theme of past Caribbean paleoclimate research has been to hypothesize about the interaction between mean climate state and seasonality, but little work has actually been done with sub-annually resolved climate proxies, and many questions still remain unanswered.This work focuses on reconstructing ocean conditions in the northern Caribbean from the geochemistry of corals growing offshore of southwestern Puerto Rico (17.9°N, 67.0°W). Annually resolved records of Sr/Ca and δ18O were generated for the years 2004 to 1751 from one continuous core. The same annual samples were analyzed for Δ14C between 2004 and 1950, and every 5 years between 1955-1751. Short (14-4 years) monthly-resolved records of δ18O and Sr/Ca were generated from this core and two other cores to investigate the role of seasonal variability during mean climate state variations. Substantial multidecadal variability in delta-18-O and Delta-14-C was found to correlate temporally with the intensity of the trade winds during recent times and over the last 250 years.Strong trade winds are associated with isotopic depletion in the coral geochemistry with respect to both 18-O and 14-C, and this is interpreted as an increase in the amount of equatorial or southern Caribbean Sea water in the northern Caribbean. Other findings include a 2 degree C cooling in the Caribbean during the Maunder solar minimum and no change in coral delta-18-O seasonality during significant mean state variations. Inter-colony geochemical variability in the coral species Montastraea faveolata was quantified, and the median difference between Sr/Ca and delta-18-O in corals growing on the same reef at the same time is 0.047 mmol/mol and 0.11 permil, respectively.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Kelly Halimeda Kilbourne.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 171 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001930852
oclc - 213466439
usfldc doi - E14-SFE0001724
usfldc handle - e14.1724
System ID:
SFS0026042:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200397Ka 4500
controlfield tag 001 001930852
005 20090423083311.0
007 cr mnu uuuuu
008 080319s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001724
035
(OCoLC)213466439
040
FHM
c FHM
049
FHMM
090
GC11.2 (ONLINE)
1 100
Kilbourne, Kelly Halimeda
0 245
Tropical Atlantic and Caribbean climate variations during the past eight centuries
h [electronic resource] /
by Kelly Halimeda Kilbourne.
260
[Tampa, Fla] :
b University of South Florida,
2006.
520
ABSTRACT: The Caribbean Sea is a key region from which to generate paleoclimate records because instrumental temperature data indicate that surface temperatures in the Caribbean region are correlated with global surface temperature. Heat and salt fluxes in the Caribbean have been implicated in major reorganizations in Atlantic Meridional Overturning Circulation on glacial time scales [Schmidt et al., 2004] and it has been proposed that the Tropical Atlantic (including the Caribbean Sea) may play a role in smaller-scale changes in the Atlantic basin [Vellinga and Wu, 2004]. A theme of past Caribbean paleoclimate research has been to hypothesize about the interaction between mean climate state and seasonality, but little work has actually been done with sub-annually resolved climate proxies, and many questions still remain unanswered.This work focuses on reconstructing ocean conditions in the northern Caribbean from the geochemistry of corals growing offshore of southwestern Puerto Rico (17.9N, 67.0W). Annually resolved records of Sr/Ca and 18O were generated for the years 2004 to 1751 from one continuous core. The same annual samples were analyzed for 14C between 2004 and 1950, and every 5 years between 1955-1751. Short (14-4 years) monthly-resolved records of 18O and Sr/Ca were generated from this core and two other cores to investigate the role of seasonal variability during mean climate state variations. Substantial multidecadal variability in delta-18-O and Delta-14-C was found to correlate temporally with the intensity of the trade winds during recent times and over the last 250 years.Strong trade winds are associated with isotopic depletion in the coral geochemistry with respect to both 18-O and 14-C, and this is interpreted as an increase in the amount of equatorial or southern Caribbean Sea water in the northern Caribbean. Other findings include a 2 degree C cooling in the Caribbean during the Maunder solar minimum and no change in coral delta-18-O seasonality during significant mean state variations. Inter-colony geochemical variability in the coral species Montastraea faveolata was quantified, and the median difference between Sr/Ca and delta-18-O in corals growing on the same reef at the same time is 0.047 mmol/mol and 0.11 permil, respectively.
502
Dissertation (Ph.D.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 171 pages.
Includes vita.
590
Adviser: Terrence M. Quinn, Ph.D.
653
Coral.
Oxygen isotope.
Stontium.
Radiocarbon.
Little ice age.
690
Dissertations, Academic
z USF
x Marine Science
Doctoral.
773
t USF Electronic Theses and Dissertations.
949
FTS
SFERS
ETD
GC11.2 (ONLINE)
4 856
u http://digital.lib.usf.edu/?e14.1724



PAGE 1

Tropical Atlantic and Caribbean Climate Variations During the Past Eight Centuries by Kelly Halimeda Kilbourne A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Marine Science University of South Florida Major Professor: Terrence M. Quinn, P h D Benjamin P. Flower, Ph.D. Thomas P. Guilderson, Ph.D. Frank Muller-Karger, Ph.D. Robert S. Webb, Ph.D. Robert H. Weisberg, Ph.D Date of Approval: September 18, 2006 Keywords: coral, oxygen isotope, strontium, radiocarbon, little ice age Copyright 2006 Kelly Halimeda Kilbourne

PAGE 2

DEDICATION This dissertation is dedicated to the memory of my father, Dr. Richard T. Kilbourne.

PAGE 3

ACKNOWLEDGEMENT I would like to acknowledge the committee members of both my Masters and Ph.D., Terrence Quinn, Benjamin Flower, Gary Mitchum, David Hollander, Robert Webb, Robert Weisberg, Frank Muller-Karger, and Thomas Guilderson, who have each guided my growth as a scientist through our interactions. The past and present students of the Paleo lab have made working at USF a pleasure, especially Jennifer Smith and Heather Hill who were always there for discussing an intellectual conundrum or for a mental break. Many thanks are deserved by Jyotika Virmani, Denis Mayer, Christina Holland, and Kara Sedwick for long discussions on the physics of climate and occasionally Matlab code. Fieldwork for this study was made possible by Fred Taylor of the University of Texas, dive master Milton Carlo of the University of Puerto Rico and Captain Angel Nazario. Peter Swart deserves thanks for help with fieldwork in Puerto Rico and for leading the research cruise to the Lesser Antilles where I first learned to drill cores. Amos Winter and Johan Nyberg are thanked for supplying samples and for their intellectual work towards understanding past climate in Puerto Rico. Kevin Helmle of Nova Southeastern University cut and x-rayed all of the cores used in this study. Funding for this project is from National Science Foundation grant OCE-0327420 and the Elsie and William Knight Oceanographic Fellowship. Radiocarbon analyses were performed under the auspices of the U.S. Department of Energy by the University of California Lawrence Livermore National Laboratory (contract W-7405-Eng-48).

PAGE 4

NOTE TO READER The original of this document contains color that is necessary for understanding the data. The original dissertation is on file with the USF library in Tampa, Florida

PAGE 5

i TABLE OF CONTENTS LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Caribbean and Tropical Atlantic Ocean Surface Circulation . . . . . . . . . . . . . . 6 1.3 Atlantic and Caribbean Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.4 Dissertation Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2 DECADALTO INTERANNUAL-SCALE SOURCE WATER VARIATIONS IN THE CARIBBEAN SEA RECORDED BY PUERTO RICAN CORAL RADIOCARBON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2.1 Coral 14 C as a water mass tracer in the Caribbean . . . . . . . . . . . . . . . . 18 2.2.2 Analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.2.3 A mixing model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.3.1 Pre-Bomb Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.3.2 Post-Bomb Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

PAGE 6

ii 2.4.1 Pre-Bomb Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.4.2 Water Mass Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 2.4.3 Mechanisms for Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3 PALEOCLIMATE PROXY PERSPECTIVE ON CARIBBEAN CLIMATE SINCE THE YEAR 1751 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 3.2 Analytical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 3.3 The Modern Climate Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.4 Geochemical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.5.1 Documentation of the Climate Signal . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.5.2 Long-Term Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.5.3 Multidecadal Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 3.5.4 Interannual Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 3.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 4 CORAL GEOCHEMICAL RECORDS OF CLIMATE IN THE CARIBBEAN OVER THE LAST 800 YEARS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 4.4.1 Possible Non-Climatic Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

PAGE 7

iii 4.4.2 Centennial Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.4.3 Seasonality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 5 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 6 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 APPENDICIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Appendix 1: Annual Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Appendix 2: High Resolution Isotope Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Appendix 3: Modern Monthly Data 2004-1993 C.E. . . . . . . . . . . . . . . . . . . . . . . . 145 Appendix 4: Monthly Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Appendix 5: Sr/Ca Replication Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 ABOUT THE AUTHOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . END PAGE

PAGE 8

iv LIST OF FIGURES Figure 1: Sea surface temperature (SST) anomalies (SSTA) from a local record taken daily near the study site (dashed line) tracks larger patterns in regional and hemispheric temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Figure 2: Seasonal SST averaged over 1948-2006 from NCEP/NCAR reanalysis data showing the seasonal progression the highest SST values from near the equator during the DJF and MAM to the Caribbean sea during JJA and SON . . . . . 9 Figure 3 Seasonal salinity fields in the Atlantic Ocean from the World Ocean Atlas [Conkright et al., 2001] show that the Caribbean Sea has steep salinity gradients with fresher water in the southern Caribbean Sea and saltier water in the northern subtropical Atlantic Ocean. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 4: Seasonal mean Atlantic-region surface precipitation rate (mm/day) from NCEP/NCAR Reanalysis data spanning the period Jan 1948 to August 2006. . . . . 11 Figure 5: Seasonal average surface vector wind (m/s) over the Atlantic for the period 1948-2006 from NCEP/NCAR reanalysis data. . . . . . . . . . . . . . . . . . . . . . 12 Figure 6: Radiocarbon values of seawater DIC between 1981 and 1983 [ Ostlund and Grall 1987; Nydal and Lvseth 1996] plotted with coral radiocarbon values at Bermuda, Puerto Rico and Cabo for 1983 outlined in black. . . . . . . . . . . 19 Figure 7: Caribbean and western Atlantic coral 14 C records. . . . . . . . . . . . . . . . . . . . . 24 Figure 8: Puerto Rico coral 14 C (black circles) and atmospheric 14 C (grey circles) from tree ring records [ Reimer et al. 2004] spanning 1751 to 1960. . . . . . . . . . . . 25

PAGE 9

v Figure 9: Coral 14 C records from various sites around the western Atlantic capture a strong depletion between 1950 and 1953 in 5 different corals (Glovers Reef, Belize; Puerto Rico; Punto de Galinha, Brazil; Pickles Reef, Florida Keys; The Rocks, Florida Keys). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 10: Post-bomb radiocarbon records from corals and seawater measurements of various water masses as determined by the sampling location [ Ostlund and Grall 1987]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 11: Results of a mixing model describing the sources of water bathing southwestern Puerto Rico using the Cabo coral 14 C as the equatorial end member and the Bermuda coral 14 C as the subtropical end member (lower panel). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 12: Difference between mean NCEP/NCAR reanalysis annual surface vector winds for 1963-1970 and 1977-1983. Equatorial winds were stronger during 1977-1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Figure 13: Spatial correlation between NCEP/NCAR reanalysis zonal wind velocity (m/s) and the coral-derived mixing model between 1963 and 1983, indicating that the observed changes in 14 C at Puerto Rico are strongly correlated to the trade winds over a wide region of the tropical Atlantic. . . . . . . . . . . . . . . . . . . . . 38 Figure 14: Time series of annual zonal wind velocity averaged over the boxed areas in Figure 13 and the results of the mixing model demonstrate that the trade winds were stronger during the latter portion of the study period. . . . . . . . . . 39 Figure 15: Surface temperature (black) and salinity (grey, upper graph) data from the CaTS station located at 67W and 17 37 N and the 18 Ow (grey, lower

PAGE 10

vi graph) implied by the salinity data and the local 18 Ow-salinity relationship from section1.1 [ Watanabe et al. 2001]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Figure 16: Annual variations in 18 O (, gray) and Sr/Ca (mmol/mol, black) from a Montastraea faveolata coral from southwestern Puerto Rico. . . . . . . . . . . . . . . . . 53 Figure 17: Coral 18 O with the first three principle components calculated by singular spectrum analysis with a window length of 25 [ Dettinger et al. 1995; Ghil et al. 2002] illustrating that the trend in the data is robust. . . . . . . . . . . . . . . 53 Figure 18: Comparison of annual coral 18 O data with the first principle components of singular spectrum analyses [ Dettinger et al. 1995; Ghil et al. 2002] carried out with five different window (M) lengths: 15, 20, 25, 50, and 100. . . . . . . . . . . 54 Figure 19: Puerto Rico coral 18 O (thin black line) and Cariaco basin G. bulloides abundance (thin grey line) from Black [1999] detrended and normalized to unit variance and zero mean. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Figure 20: Power spectra of time series of Puerto Rico coral 18 O (left) and Cariaco Basin G. bulloides abundance (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 21: Geochemical results from the top of coral core 04LPTA, including 18 O, 13 C, Sr/Ca, and Mg/Ca. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure 22: A) Comparison of monthly resolved coral 18 O (black) with a local record of monthly SST(gray), which was calculated from daily measurements from the Magueyes Island dock. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Figure 23: High resolution coral 18 O data compared with approximately monthly samples over a year spanning 1998 to 1999 show no significant depletion in 18 O during the fall of 1998 when Hurricane Georges passed over the area. . . . . . 60

PAGE 11

vii Figure 25: Time series and regression of annual extension rates (grey) with Sr/Ca (black) for a Puerto Rican coral. Sr/Ca is essentially independent of extension rates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Figure 26: Coral Sr/Ca (mmol/mol) from four different coral heads growing at the same time on the same reef. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Figure 27: Paleoclimate records from the Caribbean and Eastern Pacific showing 2C cooler temperatures in the 18 th century relative to todays temperature. . . . . . 64 Figure 28: Multidecadal variability in Atlantic records illustrating the strong coherence between the AMO index and proxy data from tree rings (Gray AMO index) and coral 18 O (VPDB) during the instrumental record, and the disconnect between the proxy data during the pre-instrumental time. . . . . . . . . . . 69 Figure 29: Wavelet analysis of Puerto Rican coral 18 O showing the dominance of the 60 year cycle in the data and showing the evolution of the variance at lower periods (36 years, 4.7 years, and 2.3 years). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure 30: X-radiograph of core E1P illustrating the unconformity in the late 19 th century and the uranium-series date at the bottom. . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure 31: X-radiographs of corals used in this study that have not previously been published. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Figure 32: Geochemical cross plots of coral Sr/Ca, 18 O and 13 C indicate that most of the older coral material is not diagenetically altered. . . . . . . . . . . . . . . . . . . . . . 88 Figure 33: Coral Sr/Ca and 18 O records (~monthly resolution) from Puerto Rico spanning more than 8 centuries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

PAGE 12

viii Figure 34: Mean coral 18 O (black diamonds) and Sr/Ca (grey squares) values from Puerto Rican corals plotted with annual values from core 04LPTA (dashed lines). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Figure 35: Surface temperature and salinity changes indicated by the coral geochemistry as calculated using the method of Kilbourne et al. [2004]. . . . . . . . . 92 Figure 36: The seasonal cycles of 18 O during 8 time periods from two continuous coral cores with the means removed to facilitate comparison. . . . . . . . . . . . . . . . . 93 Figure 37: High resolution paleoclimate records from the Caribbean and South America showing transition periods at the beginning of the Sporer solar minimum and during the Maunder minimum. . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

PAGE 13

ix LIST OF TABLES Table 1: Mean surface 14 C values collected between January 1955 and December 1957. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Table 2: Correlation coefficients (r) of annual SSTA scaling up from our local study site, to the Northern Hemisphere average. . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Table 3: Absolute value of the differences between the geochemistry of coral heads growing at the same time and in the same area. Sr/Ca units are mmol/mol and 18 O units are VPDB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

PAGE 14

x ABSTRACT The Caribbean Sea is a key region from which to generate paleoclimate records because instrumental temperature data indicate that surface temperatures in the Caribbean region are correlated with global surface temperature. Heat and salt fluxes in the Caribbean have been implicated in major reorganizations in Atlantic Meridional Overturning Circulation on glacial time scales [ Schmidt et al. 2004] and it has been proposed that the Tropical Atlantic (including the Caribbean Sea) may play a role in smaller-scale changes in the Atlantic basin [ Vellinga and Wu 2004]. A theme of past Caribbean paleoclimate research has been to hypothesize about the interaction between mean climate state and seasonality, but little work has actually been done with sub-annually resolved climate proxies, and many questions still remain unanswered. This work focuses on reconstructing ocean conditions in the northern Caribbean from the geochemistry of corals growing offshore of southwestern Puerto Rico (17.9N, 67.0W). Annually resolved records of Sr/Ca and 18 O were generated for the years 2004 to 1751 from one continuous core. The same annual samples were analyzed for 14 C between 2004 and 1950, and every 5 years between 1955-1751. Short (14-4 years) monthly-resolved records of 18 O and Sr/Ca were generated from this core and two other cores to investigate the role of seasonal variability during mean climate state variations.

PAGE 15

xi Substantial multidecadal variability in 18 O and 14 C was found to correlate temporally with the intensity of the trade winds during recent times and over the last 250 years. Strong trade winds are associated with isotopic depletion in the coral geochemistry with respect to both 18 O and 14 C, and this is interpreted as an increase in the amount of equatorial or southern Caribbean Sea water in the northern Caribbean. Other findings include a 2C cooling in the Caribbean during the Maunder solar minimum and no change in coral 18 O seasonality during significant mean state variations. Inter-colony geochemical variability in the coral species Montastraea faveolata was quantified, and the median difference between Sr/Ca and 18 O in corals growing on the same reef at the same time is 0.047 mmol/mol and 0.11 respectively.

PAGE 16

1 1 INTRODUCTION 1.1 Overview Predicting climate change requires understanding of the climate system. Climate change on time scales of decades to centuries are particularly important because those are the scales over which we try to predict anthropogenic effects on climate [ Ramaswamy et al. 2001]. Widespread instrumental records of climate that can be used to understand the climate system do not span much more than the 20 th century, whereas net radiative forcing due to greenhouse gases has increased by about 1.5 W/m 2 since 1800 C.E. [ Ramaswamy et al. 2001] (C.E. stands for Common Era and is the accepted secular replacement for A.D. or Anno Domini and C.E. will be used throughout this document). Thus, the instrumental data are recording an already perturbed climate signal. Understanding climate changes before substantial anthropogenic impact provides tests of our understanding and of our ability to predict future climate change [ Otto-Bliesner et al. 2006]. Documenting past changes in the climate system is the first step toward understanding it. Paleoclimatology is a discipline that is focused on documenting and understanding Earths climate history using non-instrumental sources of climatic information often called climate proxies. Proxy records take much time and effort to obtain and interpret, so researchers must carefully target the sample site and proxy variables to answer their

PAGE 17

2 science questions. Studies focused on large-scale climate processes such as this one must ennsure that the local climate is representative of a large region or large-scale processes so that broad spatial and temporal patterns can be identified from a small network of records [ Evans et al. 2002; Holland et al. 2006]. One location that has been previously identified as climatologically important is the Caribbean Sea [ Bradley 1996]. Although the importance of this region was originally identified based on a high correlation to global temperature, the Caribbean Sea is involved with many important climate processes. The tropical Atlantic, including the Caribbean, is influenced by (and possibly influences) at least four major climate phenomena: North Atlantic Oscillation (NAO), the Pacific El Nio-Southern Oscillation (ENSO), Atlantic Meridional Overturning Circulation (MOC) which may be related to the Atlantic Multidecadal Oscillation (AMO), and the Atlantic cross-equatorial sea surface temperature gradient [e.g., Marshall et al. 2001and references therein]. The link between the Caribbean and MOC may be strong, as proposed by Vellinga and Wu [2004], and has been demonstrated for glacial time scales by Schmidt [2004] Decadalto centennialscale processes in the Caribbean are poorly known, in part because of the temporal limitations of instrumental data, and in part because of a lack of paleoclimate data. The primary way to increase our understanding of the role that the Caribbean plays in long-term climate processes is to generate more paleoclimate data designed to fill in our knowledge gaps. Specific climate-related questions investigated in this dissertation with new coralbased paleoclimate records from Puerto Rico are:

PAGE 18

3 1. Have there been detectible changes in the proportion of North Atlantic subtropical gyre water and South Atlantic equatorial water entering into the Caribbean Sea during any time in the last 250 years? If so, are the observed changes related to other climatic variables? 2. Was the northern Caribbean Sea indeed 2C cooler than today during the Little Ice Age as has been proposed by earlier studies? Was there a substantial salinity anomaly during the same period? 3. Is there substantial decadal variability in northern Caribbean Sea surface temperature or salinity? If so, can it be related to other Atlantic climate processes? 4. How stable were the conditions in the Caribbean through the global climatic changes of the last 1000 years? 5. Did changes in seasonality play a role in any observed mean state changes? Corals have several characteristics making them excellent archives of climate variability. First, their annual density bands make an effective time stamp, putting their geochemical fluctuations into a meaningful temporal context [ Knutson et al. 1972]. Second, they can be sampled at monthly intervals, giving seasonal resolution to time series produced from them. Scleractinian corals are composed of the aragonitic form of calcium carbonate, which is amenable to many geochemical analyses, and can record a variety of independent, climate-related variables such as: Sr/Ca (temperature), 18 O (temperature and hydrologic variations), 13 C (productivity, upwelling, cloud cover, etc), and radiocarbon (ocean circulation). The details of reconstructing climate variability

PAGE 19

4 from coral geochemical records are not as simple as the above statement makes it seem, but many studies have shown that it can be done well, if appropriate care is taken [e.g., Druffel 1997a; Gagan et al. 2000; Druffel 2002; Corrge 2006] Each major section of this dissertation addresses the fidelity of the climate signal before interpreting the coral geochemical signal. Puerto Rico is a good place to reconstruct a coral-based history of tropical Atlantic climate variability because its climate is influenced by several climatic phenomena that strongly influence Atlantic-sector conditions. Temperature variations in southwestern Puerto Rico track those in the Western Hemisphere Warm Pool (WHWP), and the northern hemisphere, on interannual and interdecadal time scales (Figure 1), this is quantified in Table 2 of Section 3.3). The strong relation indicates that variability in a coral proxy-temperature record from this area will be a good estimate for past WHWP behavior and could be a valuable addition to studies of past global mean temperature. Flushed by northern Caribbean surface waters, corals growing on the reefs offshore of southwestern Puerto Rico will record the isotopic composition of the dissolved inorganic carbon in their surroundings [ Druffel and Linick 1978; Druffel 1997a]. A coral-based 14 C history of surface waters in this region could provide the data necessary to assess the proportion of isotopically distinct water masses entering the Caribbean Sea from the Northern Hemisphere and the Southern Hemisphere. The interpretation of changes in ocean mixing or current patterns from the 14 C data could be strengthened when combined with temperature and salinity information from stable isotope and Sr/Ca data.

PAGE 20

5 -1.0 0.0 1.0 1900 1920 1940 1960 1980 2000 SST anomaly (C) La Parguera SSTA sw Puerto Rico SSTA WHWP SSTA N.Hem Avg. (Mann et al., 1999) Figure 1: Sea surface temperature (SST) anomalies (SSTA) from a local record taken daily near the study site (dashed line) tracks larger patterns in regional and hemispheric temperature. Gridded sea surface temperature data from a 1x1 grid box including the study site (HadISST, Rayner et al., 2003; centered on 17.5N, 67.5W) are depicted in grey. Regional average sea surface temperatures from the Western Hemisphere Warm Pool (also HadISST, [ Rayner et al. 2003] ; averaged over 7-27N, and 110-50W) are depicted with dark blue, and the Northern Hemisphere surface temperature reconstruction by Mann et al. [1999] using paleoclimate proxy data is red. The base period in each instance is the full record: 1966-2002 for the local SST record, 1871-2003 for the Hadley Center data, and 1902-1980 for the Mann et al. [1999] reconstruction. The local record has been shifted to have a similar mean anomaly over the period of overlap. Puerto Rico also lies along a salinity gradient between more saline northern subtropical water to the north and the less saline Caribbean to the south [ Corredor et al. 2003]. Seasonally, meridional variability of this front induces large (up to ~2psu) salinity fluctuations in waters south of Puerto Rico [ Corredor and Morell 2001] Decadalto centennialscale changes in trade wind strength and regional evaporation to precipitation ratios have the potential to cause shifts in the salinity front in the northern Caribbean and therefore strongly influence the salinity of the waters surrounding Puerto Rico. The oxygen isotopic composition of the seawater and the salinity covary in this region with the following relationship from Watanabe et al. [2001] who measured water samples during every season over several years in the 1990s:

PAGE 21

6 18 O w = 0.20 ( 0.03 ) SSS # 6.5( 0.7) (r=0.93, N=20) Seawater oxygen isotopic composition ( 18 Ow) and salinity are positively correlated because evaporation preferentially removes the light isotope (salinity increases and seawater 18 O increases) and precipitation comes from the isotopically depleted vapor (salinity decreases and seawater 18 O decreases). River runoff ultimately comes from precipitation and also is depleted with respect to 18 O. 1.2 Caribbean and Tropical Atlantic Ocean Surface Circulation Wind stress acting on the ocean in the Atlantic sets up four central gyres that can be determined theoretically from the Sverdrup transport [ Mayer and Weisberg 1993]. They are the Northern Hemisphere anticyclonic subtropical gyre, the cyclonic tropical gyre just north of the equator, the clockwise equatorial gyre straddling the equator, and the counter clockwise Southern Hemisphere subtropical gyre. The surface currents in the tropics making up these gyre circulations can be seen in satellite-tracked drifter trajectory data [ Fratantoni 2001] and are from North to South: the westward flowing North Equatorial Current, the eastward flowing North Equatorial Counter Current, and the westward flowing South Equatorial Current. The South Equatorial Current impinges upon the South American continent, and a northward branch joins the North Brazil Current that runs northwestward along the north coast of Brazil until it retroflects eastward into the North Equatorial Counter Current [ Lumpkin and Garzoli 2005] Components of deeper circulation systems also pass through the above described surface circulation system. The Subtropical Cells (STCs) are like the oceanic version of Hadley Circulation with wintertime subduction in the subtropics, equatorward transport

PAGE 22

7 in the pycnocline, upwelling at the equator, and poleward transport back to the subtropics in the surface [ Zhang et al. 2003]. The water subducting in the subtropics is know as salinity maximum water (SMW), subtropical-mode water, or subtropical under water (SUW). Another major circulation system to influence the tropical Atlantic surface currents is the Atlantic Meridional Overturning Circulation (MOC) [ Wunch 2002] Water in this system sinks to depth in the North Atlantic, is exported to the South Atlantic, and is replaced by surface water coming from the South Atlantic into the Caribbean and northward with the Florida Current [ Schmitz and Richardson 1991; Schmitz and MCCartney 1993]. Of the approximately 13 Sverdrups (Sv) exported to the South Atlantic in MOC, 7 Sv returns in the upper 50-100m of the water column (potential temperature >24C), 1 Sv returns with potential temperatures between 12C and 24C, and 5 Sv returns below that in the 7Cto 12C range. The three main mechanisms identified for return flow to enter the Caribbean are: 1) entrainment from the South Equatorial Current to the North Brazil Current and North Equatorial Counter Current, northward flow in the interior of the basin by Ekman transport, subduction in the subtropics and finally into the Caribbean [ Halliwell et al. 2003] 2) transport within eddies shed off of the North Brazil Current retroflection [ Richardson et al. 1994], and 3) direct flow into Caribbean from the currents off the northern coast of south America [ Johns et al. 2002] 1.3 Atlantic and Caribbean Climate Seasonal fields of sea surface temperature (SST), sea surface salinity (SSS), precipitation rate, and surface wind velocity in Figures 2,3, 4, and 5 give the regional

PAGE 23

8 context for climate in Puerto Rico discussed below. Sea surface temperature is highest in the Caribbean Sea and Gulf of Mexico during the northern hemisphere summer and fall, whereas SST is highest around the equator during the southern hemisphere summer and fall. The high temperatures are biased toward the western part of the basin such that the Caribbean Sea is always warmer than at similar latitudes on the eastern side of the Atlantic. The 28C isotherm seasonally encompasses the Caribbean and Gulf of Mexico, creating what has been called the Western Hemisphere Warm Pool [ Wang and Enfield 2001, 2003], analogous to the Western Pacific Warm Pool. The WHWP is a source of sensible and latent heat to the atmosphere in the region [ Wang and Enfield 2001, 2003]. The Inter-tropical Convergence Zone (ITCZ) can be defined in many ways, including surface wind convergence, precipitation rate, and convective activity. The seasonal location of the ITCZ can be identified in Figures 3,4, and 5 representing SSS, precipitation rate and surface wind velocity. The weakest winds correspond with a definite area of convergence over the Atlantic Ocean in Figure 5. However, the winds are relatively weak over South America and there is no zonally linear feature that one could call the ITCZ in any season. The precipitation rate data in Figure 3 similarly show a zonal band of high values over the tropical ocean and relatively disorganized higher values over much of northern South America, Central America, and the Caribbean. The breakdown of the linear ITCZ over land illustrates the strong effects that land and vegetation have on classical ITCZ activity in this region. Much of the rainfall in Central America and northern South America is monsoonal and cannot be considered to be from the ITCZ directly, although it may be related to ITCZ activity.

PAGE 24

DJF MAM JJA SON Figure 2: Seasonal SST averaged over 1948-2006 from NCEP/NCAR reanalysis data showing the seasonal progression the highest SST values from near the equator during the DJF and MAM to the Caribbean sea during JJA and SON. 9

PAGE 25

32 33 34 34 34 35 35 35 35 35 36 36 36 37 37 JAS 32 33 34 35 35 35 35 35 36 36 36 37 37 OND0 33 34 34 35 35 35 35 35 36 36 36 36 36 36 36 37 37 JFM00 32 33 33 34 34 34 35 35 35 35 35 36 36 36 36 36 36 37 AMJ00000 Figure 3: Seasonal salinity fields in the Atlantic Ocean from the World Ocean Atlas [Conkright et al., 2001] show that the Caribbean Sea has steep salinity gradients with fresher water in the southern Caribbean Sea and saltier water in the northern subtropical Atlantic Ocean. 10

PAGE 26

DJF MAM Figure 4: Seasonal mean Atlantic-region surface precipitation rate (mm/day) from NCEP/NCAR Reanalysis data spanning the period Jan 1948 to August 2006. The data illustrate that northern South America and the Caribbean have precipitation rate patterns that are irregular both spatially and temporally in comparison with the ITCZ regions over the ocean. 11 JJA SON

PAGE 27

DJF Figure 5: Seasonal average surface vector wind (m/s) over the Atlantic for the period 1948-2006 from NCEP/NCAR reanalysis data. The tropical easterlies, higher latitudewesterlies, and equatorial convergence are visible. Puerto Rico is in the trade wind belt throughout the year. MAM JJA SON12

PAGE 28

13 Surface salinity in the Atlantic Ocean generally follows the patterns expected from precipitation with one important deviation. The salinity in the Caribbean is much lower than one would expect from just the precipitation over it. River runoff from South America contributes substantially to the observed salinity field [ Mller-Karger et al. 1989; Corredor and Morell 2001; Corredor et al. 2003]. The Amazon and Orinoco rivers are the major rivers in this part of the world. Runoff from the Amazon river basin can come into the Caribbean through the North Brazil and Guyana Currents. The Amazon river plume does not make it directly into the Caribbean Sea between June and December when it is incorporated into the North Brazil Current retroflection while the retroflection is strong [ Muller-Karger et al. 1995]. The Orinoco river has maximum discharge rates in June and a low-salinity plume originating from the mouth of the Orinoco often stretches across the Caribbean to Puerto Rico in September through November [ Mller-Karger et al. 1989; Hu et al. 2004] 1.4 Dissertation Organization This dissertation is split into three main sections based on the kind of data used and on the time interval of focus. All of the sections address the climate history of the Caribbean as recorded in the geochemistry of corals from a site offshore from the town of La Parguera in southwestern Puerto Rico. The first section explores ocean circulation changes implied by a radiocarbon record from a 254-year old coral in an attempt to answer the first part of question one from section 1.1. The next section investigates surface ocean temperature and salinity changes over interannual to centennial timescales at the site during the last ~250 years. This section will address the second part of question one as well as questions two and three. The last main section attempts to

PAGE 29

14 identify possible mean climate state changes over the last 800 years and address longstanding questions about the possible influence of seasonality during mean state shifts, the subjects of questions four and five.

PAGE 30

15 2 DECADALTO INTERANNUAL-SCALE SOURCE WATER VARIATIONS IN THE CARIBBEAN SEA RECORDED BY PUERTO RICAN CORAL RADIOCARBON 2.1 Introduction Sverdrup theory [ Sverdrup 1947] predicts that a southward shallow flow in the central portion of the North Atlantic be compensated by an equal rate of northward flow at the western boundary. Measurements of the northward transport of the Florida Current [ Schmitz and Richardson 1968; Niiler and Richardson 1973; Leaman et al. 1987] are equal to the amount predicted by Sverdrup theory alone [ Leetmaa et al. 1977]. This is problematic because the return flow of meridional overturning circulation (MOC) theoretically moves northward in the western boundary current too [ Schmitz and Richardson 1991] so the total northward transport should be greater than the wind driven transport alone, and should equal the total of MOC plus the wind-driven transport. Waters forming the well-defined western boundary current along North America (Florida Current and Gulf Stream system) coalesce in the Caribbean Sea and exit the Caribbean Sea through the Yucatan Straits as a well-defined, rapidly moving western boundary current. Hydrographic surveys of the Caribbean inflow indicating that the inflow to the Caribbean comes from primarily South Atlantic origins [ Wust 1964], and from primarily North Atlantic origins [ Parr 1937; Metcalf et al. 1971; Mazeika et al. 1983], led Wilson

PAGE 31

16 and Johns [1997] to conclude that the amount of water coming into the Caribbean from the South and North Atlantic may be variable with time. Recent attention to the possibility of abrupt climate change in the Northern Hemisphere driven by MOC shut-down [ Broecker 1998; Ganopolski and Rahmstorf 2001], and identification of the Atlantic Multidecadal Oscillation [ Schlesinger and Ramankutty 1994] which may be related to smaller changes in the MOC [ Delworth and Mann 2000] gives new importance to the time-varying sources of Caribbean inflow. Continuity requires that increased (decreased) deep oceanic convection at high latitude be compensated by increased (decreased) return flow at the surface, which must come across the equator from the South Atlantic, in order to be transported northward. Thus, the firstorder hypothesis is that time varying MOC strength should be related to the proportion of North Atlantic versus South Atlantic waters entering the Caribbean Sea, where much of the waters making up the Gulf Stream originate. A testable prediction can be made regarding the timing of changes in the proportion of water from northern and equatorial sources based on a link between MOC and multidecadal-scale temperature changes in the North Atlantic. The surface return flow brings heat to the North Atlantic, and an increase in the volume of flow could theoretically build up heat in the North Atlantic, whereas a decrease in the volume of flow would decrease the heat content of the North Atlantic. If MOC and AMO are indeed related in this way, the volume of return flow into the North Atlantic should be related to the derivative of the AMO curve such that times of positive (negative) slope would have higher (lower) volumes of MOC return flow bringing more (less) heat into the North Atlantic. Connecting this idea with circulation in the Caribbean Sea: times

PAGE 32

17 with a positive (negative) AMO slope should have higher (lower) volumes of MOC return flow, and thus more (less) South Atlantic water entering the Caribbean Sea. Documenting the time-varying inflow to the Caribbean is difficult. Despite multiyear sampling efforts [ Johns et al. 2002], and many previous hydrographic surveys in the region, the Caribbean inflow is highly under sampled. What is needed is a multi-decadal record of a water-mass tracer that integrates the high frequency variability such as eddies and tides, but resolves the interannualto decadal-scale variability. Massive corals provide an ideal archive of such information because their skeletons record the 14 C of the dissolved inorganic carbon (DIC) in the water as they grow, which can be used as a water-mass tracer [e.g., Druffel and Linick 1978; Druffel and Griffin 1993; Guilderson et al. 1998; Guilderson et al. 2004]. They can be long-lived (centuries), and their skeletons have annual density bands, much like tree rings, which can be counted to provide excellent age control for the measured geochemical tracer time series. Additionally, coral skeletal material contains multiple independent records of environmental conditions, two of which are commonly used to reconstruct temperature (Sr/Ca and 18 O) and salinity ( 18 O). We use a 250-year long 14 C record sampled at annual resolution from a coral growing off southwestern Puerto Rico to reconstruct surface water mass changes in the northern Caribbean. Combining our results with those of previous studies, we are able to identify a widespread 14 C event in the circum-Caribbean region during the early 1950s, and create a 20-year time series of the proportion of equatorial versus North Atlantic subtropical waters influencing southwestern Puerto Rico.

PAGE 33

18 2.2 Methods 2.2.1 Coral 14 C as a water mass tracer in the Caribbean Radiocarbon is continuously produced in the atmosphere by the bombardment of 14 N atoms with neutrons. The atmospheric 14 C equilibrates with the surface ocean through gas exchange with an isotopic equilibration time of about ten years. Use of radiocarbon in seawater dissolved inorganic carbon as a water-mass tracer began with some of the first oceanic radiocarbon measurements in the 1950s [ Broecker and Olson 1961], and coral 14 C has been used as a water-mass tracer since the late 1970s [e.g., Druffel and Linick 1978; Druffel 1980, 1989; Guilderson et al. 1998; Druffel 2002; Guilderson et al. 2004]. Corals ultimately utilize seawater DIC for the carbon in their skeletons, so coral skeletal 14 C can be a proxy for seawater 14 C after using 13 C to correct for the mass-dependent fractionation that occurs during calcification. Good agreement between radiocarbon measurements taken just off the southwestern coast of Puerto Rico as part of a World Ocean Circulation Experiment (WOCE) cruise in the summer of 1997 (75.9 4.2 ) and coral 14 C (81.0 3.3) provides further evidence for the fidelity of coral 14 C to record isotopic ratios of seawater DIC. Only a few coral 14 C time series from the Atlantic exist, and they are from Brazil, Cape Verde, Venezuela, Belize, the Florida Keys, and Bermuda [ Druffel 1980; Druffel 1982; Druffel 1989, 1996; Druffel 1997b; Guilderson et al. 2005]. The primary controls on surface ocean 14 C are equilibration with the atmosphere, advection, diffusion, and convective mixing [ Mahadevan 2001]. Major water sampling programs such as the Geochemical Ocean Sections Study (GEOSECS) and WOCE, as well as the efforts of smaller groups document persistent features of the surface 14 C in

PAGE 34

255075100125150175 80W 60W 40W 20W 0 20S EQ 20N 40NFloridaKeys NECCSECNBCCCFlorida CurrentGulf StreamNEC SEC 19Figure 6: Radiocarbon values of seawater DIC between 1981 and 1983 [Ostlund and Grall, 1987; Nydal and Lvseth, 1996] plotted with coral radiocarbon values at Bermuda, Puerto Rico and Cabo for 1983 outlined in black. The high level of agreement between these two datasets demonstrates that coral radiocarbon records from these areas accu-rately reflect the 14C of the surrounding seawater. Some major currents (CC Caribbean Current, NBC North Brazil Current, NEC North Equatorial Current, NECC North Equatorial Counter Current, SEC South Equatorial Current) are labeled and schemati-cally represented by the purple lines and arrows, after the drifter study of [Fratantoni, 2001]. Other sample sites mentioned in the text are marked with Xs and labeled.the Atlantic, illustrated in Figure 6. Water in the subtropical gyres has the highest 14C because it has equilibrated most with the atmosphere. In contrast, water in the equatorial regions is influenced by vertical mixing with 14C-depleted water through upwelling off the coast of Africa, and therefore has consistently lower 14C. The contrast between the North Atlantic gyre and the south equatorial region was 10 before about 1957, when nuclear bomb-produced 14C began to substantially affect the surface ocean (Table 1; Cabo/AbrolhosGloversReefXXXPorto de GalinhasPuertoRicoBermuda

PAGE 35

20 Latitude Range Average 14 C (permil) Standard Error Number of measurements 15N to 40N -49 3 11 1N to 20S -59 3 8 Table 1: Mean surface 14 C values collected between January 1955 and December 1957. Data are from Broecker and Olson [1961] [ Broecker and Olson 1961]). The gradient has steepened since that time because the surface waters equilibrated with the bomb carbon spike in 14 C, while the deeper waters were essentially unaffected (this depends on the source of the deeper water). Southwestern Puerto Rico is an ideal site to monitor seawater 14 C. The prevailing northeasterly trade winds make the local circulation regime more prone to downwelling than upwelling, so that coral 14 C variations are not likely to reflect a local upwelling signal. The coastal geomorphology is such that no significant rivers flow into the coastal region at this location, minimizing the potential impacts of local runoff. Furthermore, the wide continental shelf in this area permitted sampling of corals from a few kilometers offshore, further minimizing the probability of impacts from coastal processes and increasing the likelihood that the measurements are representative of the northern Caribbean Sea. Fed from the northern subtropical, tropical and equatorial gyres, as defined by Mayer and Weisberg [1993], northern Caribbean water 14 C should be controlled by the proportion of water from each of the sources and the 14 C value of the sources. If one can constrain the time-varying 14 C content of each of the source waters, then one can theoretically calculate the proportion of each water mass influencing a 14 C record of the northern Caribbean. Existing coral records from Bermuda and Brazil enable the estimation of the North Atlantic subtropical and equatorial end members so that such a

PAGE 36

21 calculation can be made with the Puerto Rico coral 14 C record. For the purposes of this paper, equatorial water refers to undifferentiated water from both the tropical and equatorial gyres, the two of which have lower 14 C values because of the influence of upwelling. 2.2.2 Analytical methods A 245-cm core was collected in ~5 m of water from a massive coral head of Montastraea faviolata at Turrumote Reef (17.933N, 67.001W) offshore from La Parguera, Puerto Rico in August of 2004. The coral core was cut into 5-mm thick slabs along the primary growth axis, cleaned with deionized water in a sonicator, and Xradiographed to display the annual density banding. Nominally annual samples of coral powder were milled with a Dremel tool mounted on a computer-controlled drilling stage using the X-radiographs as a guide to ensure that one density band couplet was sampled for each sample. Each sample is composed of a dense band and the less dense band laid down just previously, and labeled with the year the dense band was created. This results in annual samples approximately centered on January of the year with which they are labeled, assuming an early summer date for the formation of the dense band [ Watanabe et al. 2002]. Each annual sample was analyzed for 14 C between 2004 and 1950, and before 1950, every 5 th year was analyzed back to 1755, with the earliest sample representing 1751. The samples were prepared and analyzed for radiocarbon at Lawrence Livermore National Laboratory Center for Accelerator Mass Spectrometry [ Davis et al. 1990]. Coral aragonite was converted to graphite by reacting 2.5-7.5 mg aliquots of sample with 85% phosphoric acid in evacuated Vacutainers and graphitizing the resulting CO 2

PAGE 37

22 using an Fe catalyst and H 2 as the reducing agent [ Vogel et al. 1987]. A background subtraction was determined using 14 C-free calcite, and analytical precision was monitored with standard reference materials analyzed with the samples. Average precision was 0.32% (fractional error) with reported errors on each sample averaging 3.2. Results are reported in age-corrected 14 C as defined by Stuiver and Polach [1977], and are corrected for a uniform 13 C value of -2 2.2.3 A mixing model A two end-member mixing model is calculated describing the percentage of subtropical water, represented by the Bermuda coral of Druffel [1989], versus equatorial water, represented by the Abrolhos Bank, Brazil coral record of Druffel [1996] referred to as Cabo. The Puerto Rico coral is described as a percentage along the mixing line between the Cabo coral and the Bermuda coral for each year in the model. The model begins in 1963, when the gradient between the equatorial and subtropical waters becomes great enough that we can differentiate the two water masses from the corals with confidence from single-point measurements. The Cabo 14 C data are centered on 19xx.5, and the Bermuda data are centered on 19xx.8 (except 1983.3), so the dates are rounded up (except 1983.3 which is rounded down) in order to compare with the Puerto Rico coral data which are centered on the beginning of the year (i.e., 19xx.0). An alternate method was also calculated whereby the Bermuda and Brazil data were interpolated so that they were centered on the beginning of the year, like the Puerto Rico data, however the results were essentially the same, so they are not shown. The Cabo coral was collected in December of 1982 and the Bermuda coral in 1983, so the last nominal year of the model is 1983.

PAGE 38

23 The Cabo coral site is situated at 17.5S and 39.3W and the South Brazil Current, which runs past this site, is fed by the South Equatorial Current (SEC; Figure 6). The 14 C of the SEC would be ideal to monitor directly as an end member for this mixing exercise, however a continuous coral time series of 14 C has not been made from an appropriately situated coral, so the Cabo site is the best we can do at this time. Hydrographic data indicate that although the 14 C at this site is similar to that of the SEC, the 14 C in this region can be slightly higher than in the SEC [ Stuiver and Ostlund 1980; Takahashi et al. 1995] presumably because of the influence of waters from the subtropical South Atlantic. The mixing model using these data should be considered a maximum estimate of the amount of equatorial influence at Puerto Rico in light of these facts. 2.3 Results 2.3.1 Pre-Bomb Variability The entire coral 14 C record from Puerto Rico is shown in Figure 7 along with other Atlantic coral records for context. The first thing that one notices about these records is the signal of bomb carbon represented by an enormous increase in 14 C beginning in the late 1950s, which peaks in the early to middle 1970s. The pre-bomb average at Puerto Rico is -51 1.2 (1751-1950, 2 # standard error, N=41), with a standard deviation of 3.7 (1 # ). This is less than the pre-bomb average of the North Atlantic gyre measured from corals at Bermuda -45 1.6 (1885-1953, 2 # standard error, N=35) by Druffel [1997b], as expected from the gross distribution of 14 C in the ocean. The small negative trend in the data between 1751 and 1950 (-0.0095 /year,

PAGE 39

24 -100 -50 0 50 100 150 200 1750 1800 1850 1900 1950 2000 PR 14 C Glovers 14 C Bmda97 14 C Bmda 14 C Cabo 14 C Abrolhos 14 C P.Galinha 14 C PicklesReef 14 C Rocks 14 C 1751-1950 14 C Trend Coral 14 C -0.0095 0.0098 /year Figure 7: Caribbean and western Atlantic coral 14 C records. New data from Puerto Rico are shown as solid red circles. Data from western boundary current sites in Belize (Glovers) and the Florida Keys (Pickles Reef and The Rocks) are denoted by open symbols. Records from the South Atlantic are plotted in grey (Cabo, Abrolhos, and Punto Galinha), and gyre sites are represented by solid black symbols (Bermuda). The trend line represents the ordinary least squares regression on the Puerto Rico coral data from 1751 to 1950, with the slope and 95% confidence intervals on the slope, illustrating that the Suess effect is not a significant portion of the variance in 14 C at this location before 1950. p=0.34) is not significant because of the large amount of higher frequency variance in the records (Figure 7). Limiting the regression period to between 1900 and 1950 produced a significant slope of -0.15.05 /year (p=0.02) over the period when atmospheric 14 C was dropping precipitously (Figure 8) due to the input of 14 C-free CO 2 from fossil-fuel burning [Suess effect, Suess 1953]. The total range of pre-bomb values at Puerto Rico is -59 to -43 (Figure 8), less than the range found in Bermuda and the Florida Keys [-34 to -64 Druffel 1997b]. Typical variations in the pre-bomb 14 C record at Puerto Rico are less than

PAGE 40

25 -70 -60 -50 -40 -30 -30 -20 -10 0 10 1750 1800 1850 1900 1950 PR 14 C Atm 14 C Puerto Rico Coral 14 C Atmospheric 14 C Figure 8: Puerto Rico coral 14 C (black circles) and atmospheric 14 C (grey circles) from tree ring records [ Reimer et al. 2004] spanning 1751 to 1960. Annual samples are reported every 5 th year until 1950, when values are reported for each year. Large and relatively rapid variations occur in the ocean that cannot be explained from atmospheric forcing, indicating that horizontal and vertical mixing, or a combination of the two are driving most of the coral 14 C variability. 10 between samples, but these are still substantially greater than the atmospheric variability from tree ring records (Figure 8), and therefore must be evidence of vertical and/or horizontal advection. The rapid depletion between 1950 and 1951 that is sustained until the mid-1950s when the bomb-carbon spike enters the system, is of particular interest because the Puerto Rico record is annually resolved during this period, and we have multiple coral records from around the Atlantic at this time, allowing for a more complete spatial picture of the event (Figure 9). This event will be discussed further below.

PAGE 41

26 PR 14 C Glovers 14 C Bmda97 14 C Bmda 14 C Cabo 14 C Abrolhos 14 C P.Galinha 14 C PicklesReef 14 C Rocks 14 C -70 -65 -60 -55 -50 -45 -40 -35 -30 1920 1925 1930 1935 1940 1945 1950 1955 1960 Caribbean 14 C Gyre 14 C SEC 14 C 14 C () Figure 9: Coral 14 C records from various sites around the western Atlantic capture a strong depletion between 1950 and 1953 in 5 different corals (Glovers Reef, Belize; Puerto Rico; Punto de Galinha, Brazil; Pickles Reef, Florida Keys; The Rocks, Florida Keys). Symbols are the same as Figure 7, with the addition of blue symbols representing average seawater 14 C measurements from three different water masses between 1955 and 1956 [ Broecker and Olson 1961] Error bars represent analytical error. 2.3.2 Post-Bomb Variability Radiocarbon values in the Puerto Rico coral increase rapidly beyond pre-bomb values beginning in 1959 (Figure 10). The rate of increase levels off between 1960 and 1963, after which radiocarbon values continue to increase steadily until they peak around 130 in the early 1970s. The absolute maximum value is 147 reached in 1976 as part of a two-year peak that rises above the sustained maximum around 130 The rate of decay between 1972 and 2005 can be estimated at -2.4 /year (r 2 =0.94) by ordinary least squares linear regression. The rate of decay is more correctly estimated with an exponential fit, resulting in the following equation: 14 C = 7.34 # 10 22 e $ 0.02417 yearAD r 2 =0.94

PAGE 42

27 -100 -50 0 50 100 150 200 1950 1960 1970 1980 1990 2000 PR 14 C Glovers 14 C Bmda97 14 C Bmda 14 C Cabo 14 C Abrolhos 14 C P.Galinha 14 C PicklesReef 14 C Rocks 14 C 14 C () BrazilCurrent 14 C Caribbean 14 C Gyre 14 C Gyre/NBC 14 C NBC 14 C NEC 14 C NECC 14 C PR 14 C SEC 14 C Figure 10: Post-bomb radiocarbon records from corals and seawater measurements of various water masses as determined by the sampling location [ Ostlund and Grall 1987] Coral symbols are the same as in Figure 7. The data were binned into water masses according to the sampling location and the location of currents as depicted in Figure 6. A two end-member mixing model describing the percentage of subtropical water versus equatorial water in the northern Caribbean between 1963 and 1983 can be calculated with the coral 14 C data as described above (Figure 11). The earlier period from 1963-1970 is consistent with a subtropical origin for most of the water bathing southwestern Puerto Rico, and the latter period from 1977-1983 is interpreted as representing primarily equatorial water at the site. The two periods are separated by a transitional period between 1970 and 1976 with large interannual changes in the relative volume of either water mass. Specifically, the proportion of equatorial water increases sharply in 1970 and remains high except for 1975 and 1976 when the two-year peak in Puerto Rico coral 14 C (1975-1976) is interpreted as two years of strong northern subtropical influence. The increasing influence of equatorial water at Puerto Rico over

PAGE 43

28 0 20 40 60 80 100 1940 1950 1960 1970 1980 1990 -100 -80 -60 -40 -20 0 Percent Equatorial Percent Subtropical Subtropical Dominates Equatorial Dominates Transition Period -0.3 -0.2 -0.1 0 0.1 0.2 0.3 1940 1950 1960 1970 1980 1990 AMO index (C) More Subtropical Predicted in Carib. More Equatorial Predicted in Carib. Figure 11: Results of a mixing model describing the sources of water bathing southwestern Puerto Rico using the Cabo coral 14 C as the equatorial end member and the Bermuda coral 14 C as the subtropical end member (lower panel). Coral radiocarbon data indicate that subtropical waters dominate the signal before 1971, and equatorial waters dominate the signal after 1976. The shift in coral 14 C is coincident with a sign change in the derivative of the AMO (upper panel) and with an increase in northern hemisphere trade wind strength, indicating that Ekman transport may play a role in the delivery of water into the Northern Caribbean. this time period is a robust signal that does not disappear when the phasing of the coral records is changed in the model to represent the time taken by water parcels to travel between the study areas. It should be noted again that the percentage of equatorial water calculated from this model is a maximal estimate because of the possibility that the Cabo coral is recording a mixed equatorial and South Atlantic gyre signal. The Cabo coral is therefore probably biased towards higher 14 C values than the actual equatorial water end member. An estimate of the effects of this bias on the mixing model calculations was made using GEOSECS seawater 14 C measurements from the fall of 1972, the only time that we have both gyre and south equatorial current seawater 14 C measurements in the same year. The results of this test indicate that our end member corals capture a range between 16%

PAGE 44

29 to 53% equatorial water and 47% to 84% North Atlantic gyre water, instead of zero to 100% of each water mass as depicted in the model for simplicity. 2.4 Discussion 2.4.1 Pre-Bomb Variability The Suess effect is not resolvable with these data because of the large interannualto interdecadalscale fluctuations (Figure 8). It is possible that the significant trend between 1900 and 1950 (-0.15.05 /year, p=0.02) is indeed due to anthropogenic input of 14 C-depleted carbon to the system, and that the circulation changes do not influence the 14 C during that time period. However, the significance of the trend is highly dependent on the period of regression. Extending the regression period to 18801950, results in a completely insignificant trend (-0.02.10 /year, p = 0.70), and extending the regression period by just two years to 1900-1952 increases the slope considerably (-0.25.12 /year, p=0.002). Therefore, we conclude that 14 C changes due to horizontal and vertical advection in the Northern Caribbean swamp the atmospheric signal of hydrocarbon use during the pre-bomb era. Early work with 14 C from corals that grew in the Florida Keys and Belize identified a decreasing trend between 1900 and the 1950s that was identified as the Suess effect [ Druffel and Linick 1978; Druffel 1980]. However, later work has identified a strong depletion in the Florida Keys corals between the 1940s and the 1950s that has been attributed to ocean dynamics rather than equilibration with atmospheric radiocarbon [ Druffel 1997a; Druffel 1997b; Guilderson et al. 2005]. A closer look at radiocarbon records from this time indicates that out of seven records to span the 1950s, five of them exhibit a rapid depletion to about -65 between 1950 and 1953 (Figure 9). The records

PAGE 45

30 exhibiting this behavior come from Porto de Galinhas, Brazil (8S), Puerto Rico (17N), Glovers Reef, Belize (17N), Pickles Reef and the Rocks Reef in the Florida Keys (25N). The two records that do not exhibit this strong depletion are from the corals that are located the farthest north and the farthest south, Bermuda (34N) and Cabo (17S). The event seems to have started rapidly because the records with values for 1950 (Puerto Rico, Pickles Reef and Glovers Reef) show the seawater 14 C to be close to in that year, but the next sample in all three records (1951, 1952, and 1952.5 respectively) was close to Ocean dynamics must be at play given that the isotopic equilibration time between the atmosphere and the surface ocean is on the order of 10 years. Four scenarios are possible to explain the rapid drop in 14 C at all of these sites. The first explanation is that the depletions are simply measurement error. Previous studies likely did not make much of the 10 depletion because one sigma error bars on 14 C measurements hover around 3-4 and the time resolution is such for most of the records that the early 1950s are represented by only one or two points. The likelihood of five independent records having an analytical hiccup at the same time is very low, so we choose to reject this first explanation. The second possibility is local ventilation of low 14 C waters at each of these sites. Simultaneous local upwelling events at all of these sites, which are in totally different surface wind regimes and which have different propensities toward upwelling, seems highly unlikely, so we also choose to reject this explanation. A third possible scenario is that a large increase in equatorial upwelling lowered the 14 C value of the equatorial end-member that mixes into the Caribbean and Gulf Stream waters. The Porto de Galinhas coral, which is representative of the SEC, has a minimum value of -64 in 1952, indicating that the equatorial end-member was not

PAGE 46

31 more depleted with respect to 14 C than the waters bathing the other sites. Thus, the third scenario is also rejected, leaving one remaining scenario to be explored. The fourth and most plausible explanation is that all of these sites were strongly influenced by low 14 C equatorial surface waters. This implies that the proportion of equatorial versus northern subtropical water in the surface inflow of the Caribbean is highly variable on interannual time scales. Sill depths in the Antilles passages are generally equal to or greater than 400 m, permitting transport at intermediate depths as well as at the surface. Despite this fact, increasing the surface transport of equatorial waters into the Caribbean, may have an impact on the total amount of South Atlantic water entering the Caribbean because ~50% of the 13 Sv MOC return flow coming from the South Atlantic does so in the topmost 100 meters [ Schmitz and Richardson 1991]. The duration of this increased equatorial-water transport is unknown because the 14 C at all sites begins to rise rapidly in the mid 1950s due to enormous atmospheric input from thermonuclear bomb testing. 2.4.2 Water Mass Changes The results of the mixing model (Figure 11) are intriguing because they indicate that the proportion of water from equatorial sources versus northern subtropical sources bathing southwestern Puerto Rico changes on interannual to decadal time scales as hypothesized by Wilson and Johns [1997]. It must be noted that the low 14 C, equatorially sourced, water is a mixture of North Atlantic and South Atlantic water, and is not a pure South Atlantic signal, but it is likely that increases in the amount of equatorial water also means an increase in the amount of South Atlantic water that is carried with it.

PAGE 47

32 There are two ways of looking at the 14 C variability in the Puerto Rico coral: 1) the signal is that of a well-mixed Caribbean surface flow and represents the average proportion of water parcels from each source in the Caribbean, or, more conservatively, 2) the signal is that of the northern Caribbean, just south of Puerto Rico and signifies the sources of water in that region only. The only way to distinguish which hypothesis is correct is to look at the spatial distribution of 14 C in the Caribbean, either in the modern ocean or from the geologic record. Seawater 14 C records are too rare in the Caribbean to be able to determine any spatial patterns, especially if they are also time variant, so the best way to achieve this goal is to examine 14 C time series from other corals in the Caribbean. The only other long time series in the Caribbean is the Glovers Reef record from Belize [ Druffel 1980], which is situated along the eastern edge of the Yucatan Peninsula, just south of where the narrowly focused Caribbean Current impinges on the continent [ Richardson 2005] The 14 C record from the Belize coral has been interpreted as a signal from the western boundary current because of this fact [ Druffel 1989], and as such, it may be more of a mixed Caribbean signal rather than representative of the spatial variability due to different source waters. More records from the eastern Caribbean are needed to address the issue of spatial variability of 14 C in the Caribbean as waters enter the basin. We conservatively assume that the Puerto Rico coral 14 C record is representative of the northern Caribbean at approximately 67 longitude for the rest of our analysis. One aspect of the Belize record is worth a digression. This coral shows much higher values in the 1960s than either the Bermuda or the Puerto Rico corals (Figure 10). Errors in the coral age models of only 2 years could explain the difference, but this is

PAGE 48

33 unlikely since the 1960s are in the top 20 years of the Belize and Bermunda records where the likelihood of density band counting errors is low. The primary errors in density band counting occur over core section breaks and in sections of coral with contorted growth directions, both of which do not often occur in the top few decimeters of a core. Druffel [1989] attributed the difference between the Belize and Bermuda coral 14 C records to the formation of 18 mode water in the gyre influencing the Bermuda coral. However, the gyre water feeds into the Caribbean along with lower 14 C equatorial water, so the Belize coral should have lower 14 C values than the Bermuda coral if the Belize coral actually represents a mixed Caribbean water signal. Since the Belize coral 14 C is higher than the Bermuda and Puerto Rico coral 14 C records, 14 C must be added to the system in the Caribbean. Therefore, the surface water bathing Glovers Reef must be equilibrating with the atmosphere relatively rapidly. If one looks into more detail at the circulation surrounding Glovers Reef in the Gulf of Honduras, one finds that the reef platform is bathed by a small cyclonic gyre with relatively weak and spatially variable currents [ Centurioni and Niiler 2003; Tang et al. 2006]. A logical explanation for the relative enrichment of the Glovers Reef coral is that the relative isolation of the Gulf of Honduras, the strong thermal stratification, and the resultant very shallow mixed layer permitted the very surface waters around Glovers Reef to equilibrate with the atmosphere more rapidly than the deeply mixed northern subtropical gyre. This explains why Puerto Rico, relatively close to the Caribbean source waters, was not similarly enriched with respect to 14 C during the 1960s. The two-year enrichment of Puerto Rico coral 14 C in 1975 and 1976 indicates a short period when northern subtropical water strongly influenced the northern Caribbean

PAGE 49

34 during a longer-term period of more equatorial influence. The interannual event coincides with an El-Nino Southern Oscillation (ENSO) cold-phase event (La Nina) in the Pacific and an extensive cold anomaly in the northern Tropical Atlantic with seasonal SST anomalies less than C in the Caribbean. However, the proximal cause of the event remains unknown despite examinations of Atlantic region seasonal sea level pressure, wind velocity, and sea surface temperature anomalies over the calendar years of 1974-1977 (compared to the seasonal means between 1948-2005). Further investigation is needed to determine the causes of this event and other interannual-scale events (e.g., the early 1950s depletion) in the record. Decadal-scale water mass changes predicted by our mixing model are consistent with AMO-related MOC changes. The short length of the model output compared to variability in the AMO prohibits a quantitative analysis, however it is worth noting the qualitative agreement. Vellinga and Wood [2002] demonstrate that when MOC shuts down, the northern North Atlantic becomes much cooler and the tropical North Atlantic becomes saltier due to a southward shift in tropical precipitation patterns. It is possible that similar, if less pronounced, changes occur on multidecadal timescales. The idea is that vigorous MOC activity transports heat from the South Atlantic to the North Atlantic during the positive AMO phase and the resulting weaker MOC leads to a cooler North Atlantic during the negative AMO phase. If this model is correct, the decay of the positive phase should be associated with less vigorous MOC and therefore less low! 14 C equatorial water returning into the Caribbean. Conversely, the decay of the negative phase of the AMO should be associated with stronger overturning circulation and thus more low! 14 C equatorial water

PAGE 50

35 flowing into the Caribbean. Figure 11 shows that during the early period of our model, 1963-1975, one would predict less equatorial water in the Caribbean and during the latter period, 1975-1983, one would predict more equatorial water, exactly the pattern observed in the mixing model. The data are suggestive that the observed circulation changes might be coincident with the AMO, but much longer time series of similar data are needed to prove a consistent relationship. 2.4.3 Mechanisms for Transport The Puerto Rico coral was bathed by seawater with lower 14 C values during the early 1980s compared to the 1960s and the source of that 14 C depletion is likely to be waters from the equatorial region where upwelling is a consistent source of 14 C-depleted DIC. Current theories of how water is transported northward across the equator include: 1) eddy shedding from the North Brazil Current Retroflection [ Johns et al. 1990; Didden and Schott 1993; Richardson et al. 1994], 2) direct flow into the southern Caribbean by the seasonal northwestward flowing Guyana Current via the North Brazil Current [ Lumpkin and Garzoli 2005], and 3) northward transport of surface waters by Ekman drift in the ocean interior [ Mayer and Weisberg 1993; Halliwell et al. 2003; Lumpkin and Garzoli 2005]. In order to get equatorial 14 C-depleted waters around Puerto Rico, it has to be transported northward to the latitude of Puerto Rico, either before or after entering the Caribbean. Using steady-state assumptions, northward surface transport at this latitude (17N) is primarily by wind-driven Ekman drift [ Mayer and Weisberg 1993], thus it follows logically that the coral 14 C signal in Puerto Rico might be related to the trade-wind strength on decadal timescales.

PAGE 51

36 To test this theory, we compared the tropical Atlantic wind velocities between the early period (1963-1970) when subtropical waters dominate the Puerto Rico coral signal, and the late period (1977-1983) when equatorial waters dominate the Puerto Rico coral signal by creating a map of the differences in NCEP wind velocities between the two time periods (Figure 12). As expected from our hypothesis, the trade winds were stronger during the latter period, consistent with the proposed increase in Ekman transport during that time. To further investigate the relationship between the wind data and the results of our mixing model, we created a spatial correlation map of NCEP zonal wind velocity with our mixing model results (Figure 13). The largest region of strong negative correlation is very similar to the region of the largest difference in wind velocity between 1963-1970 and 1977-1983, and roughly encompasses the northern trade wind belt. Annually resolved zonal wind speed data for two regions in the Atlantic are shown along with the mixing model results in Figure 14 to demonstrate that the difference between the two periods reflects a robust change in the winds with time. The time series of the winds shows a trend over the same period as the coral 14 C data and the winds from the equatorial box are strongly correlated to the model (r 2 = 0.69). Serial autocorrelation inherent in data with similar trends ensures that the correlation coefficient is not statistically significant (p=0.66), but the temporal correlation of the theoretical forcing and the predicted outcome is suggestive that increased Ekman transport is a viable explanation of the 14 C data.

PAGE 52

37 Figure 12: Difference between mean NCEP/NCAR reanalysis annual surface vector winds for 1963-1970 and 1977-1983. Equatorial winds were stronger during 1977-1983.

PAGE 53

38 Figure 13: Spatial correlation between NCEP/NCAR reanalysis zonal wind velocity (m/s) and the coral-derived mixing model between 1963 and 1983, indicating that the observed changes in 14 C at Puerto Rico are strongly correlated to the trade winds over a wide region of the tropical Atlantic. Negative contours are dashed lines, positive contours are solid lines, and the contour interval is 0.2. The zonal component of the winds from each of the two regions indicated by boxes is shown in Figure 14.

PAGE 54

39 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -20 0 20 40 60 80 100 120 1950 1960 1970 1980 1990 2000 Annual Zonal-Wind Component (m/s) Percent Subtropical 5N-5S 15-25W -6.6 -6.4 -6.2 -6.0 -5.8 -5.6 -5.4 -5.2 -20 0 20 40 60 80 100 120 1950 1960 1970 1980 1990 2000 10-20N 60-70W Figure 14: Time series of annual zonal wind velocity averaged over the boxed areas in Figure 13 and the results of the mixing model demonstrate that the trade winds were stronger during the latter portion of the study period. The data are not inconsistent with Ekman transport control on the amount of low 14 C waters affecting Puerto Rico. Longer coral time series from key locations are needed to further substantiate the connection between wind speed and coral 14 C proposed in this study. 2.5 Conclusions 1. The Suess effect is not apparent in coral 14 C from southwestern Puerto Rico. Interannualto interdecadal-scale variations in 14 C are large and make the longterm trends in the data insignificant. 2. The radiocarbon content of the waters bathing southwestern Puerto Rico changes on interannual to multidecadal time scales in response to ocean dynamics. 3. A major contributor to this variance is the proportion of equatorial versus subtropical water in this area. The timing of water mass changes between 1964 and 1983 is consistent with predictions based on the coupling of the Atlantic Multidecadal Oscillation with Meridional Overturning Circulation.

PAGE 55

40 4. We propose that the observed changes in the northern extent of equatorial water masses are caused by an increase in trade winds and the accompanying increase in northward Ekman transport of the surface waters.

PAGE 56

41 3 PALEOCLIMATE PROXY PERSPECTIVE ON CARIBBEAN CLIMATE SINCE THE YEAR 1751 3.1 Introduction A major challenge facing the paleoclimate community is the robust reconstruction of global climate trends from select paleoclimate records that are not universally distributed in time and space. Quantitative analysis of modern climate variability can used to help guide site selection by identifying sites that are highly correlated with broad regions of the globe. Southwestern Puerto Rico, is an optimal site for global paleotemperature reconstruction because its annual surface temperature is highly correlated with northern hemisphere annual surface temperature during the instrumental period (Figure 1). The region also has a distinct salinity gradient between the relatively fresh seawater of the Caribbean Sea and the relatively saline seawater of the North Atlantic subtropical gyre [ Levitus 1982; Conkright et al. 2002]. Caribbean surface water is influenced by both continental runoff and direct precipitation, both of which are related to the movements of the ITCZ [ Corredor and Morell 2001; Corredor et al. 2003]. The compound sources of freshwater in the Caribbean Sea and the strongly evaporative subtropical Atlantic to the north make northern Caribbean salinity especially sensitive to changes in the northeast trade winds and tropical convection. Previous work in the Caribbean region has established that the climate of last 3000 years is characterized by a drying trend that has been interpreted to reflect a

PAGE 57

42 southward migration of the mean ITCZ position since the middle Holocene [e.g., Hodell et al. 1991; Higuera-Gundy et al. 1999; Haug et al. 2001a; Black et al. 2004]. Higher resolution studies show substantial decadal to centennial variability in temperature and hydrologic balance on top of this trend [ Black et al. 1999; Haase-Schramm et al. 2003; Hodell et al. 2005a; Hodell et al. 2005b] Solar variability has been implicated as a driver for at least some of the high frequency variability ranging from the 206-year to 11year period cycles. [ Black et al. 1999; Hodell et al. 2001; Nyberg et al. 2001; Nyberg et al. 2002; Haase-Schramm et al. 2003; Black et al. 2004], but processes internal to the climate system have not been ruled out and some of the variance is surely not driven by solar forcing (see the discussion by Black et al. [2004] ). One internal process to consider involves multidecadal changes in North Atlantic sea surface temperature[ Schlesinger and Ramankutty 1994] called the Atlantic Multidecadal Oscillation (AMO, [ Kerr 2000]), which may be related to changes in thermohaline circulation [ Delworth and Mann 2000] It has been suggested that the phase of the AMO is linked with atmospheric circulation and precipitation patterns over the North American continent [ Enfield et al. 2001], and that it is linked with tropical storm frequency and intensity [ Landsea et al. 1999; Goldenberg et al. 2001]. Others contend that the AMO is not a dominant feature in the tropical Atlantic and instead attribute recent increases in hurricane activity solely to global temperature increases [ Emanuel 2005; Webster et al. 2005; Mann and Emanuel 2006]. This latter argument is strengthened by the fact that midto high-latitude tree-ring records have been the primary source of informationto document the AMO as a persistent feature over multiple centuries [ Delworth and Mann 2000; Gray et al. 2004]. Thus, it is possible that the

PAGE 58

43 records are biased towards an extra-tropical signal and that the AMO does not affect the tropics as much, despite Herculean efforts to chose a representative sampling set and quantify the errors in the analysis [see Mann et al. 1998]. One way to address at least part of this dispute is to investigate tropical Atlantic paleoclimate records for evidence of AMO-related variability. Existence of a strong tropical AMO signal before significant anthropogenic warming would support the idea that the AMO is indeed a natural part of the climate system in the tropical Atlantic, although its impact on tropical storm frequency or intensity and causes of recent apparent changes in Atlantic tropical storms would remain open lines of investigation. An appropriate record to test for the influence of the AMO on tropical Atlantic climate must overlap with the instrumental record, be multicentury in length, and have a temporal resolution finer than about 15 years per sample (to be able to resolve periods of at least 30 years). Records from two locations in the tropical Atlantic are appropriate to answer this question but the data do not put the issue to rest. The sclerosponge 18 O and Sr/Ca records from Jamaica and San Pedro Bank [ Haase-Schramm et al. 2003] are from two specimens, one from 20 m depth and the other from 125 m, and they show fairly strong multidecadal variability [ Haase-Schramm et al. 2003] The foraminiferal abundance and geochemical records of Black and co-authors [ Black et al. 1999; Black et al. 2004] are also appropriate for investigating an AMO signal in the tropical Atlantic. Spectral analysis of the foraminiferal-based records (800 years long and 2000 years long respectively) show significant variance at very low frequencies (periods >100 years) and at 10-13 years periods, but not at the multidecadal periods expected for the AMO. These

PAGE 59

44 studies have made large contributions toward understanding tropical Atlantic and Caribbean climate history. However, the existing data are contradictory and more records from the region are needed to document what role, if any, the AMO plays in tropical Atlantic climate variability. Another issue that can be addressed by multicentury Caribbean paleoclimate records is the extent of tropical cooling during the Little Ice Age, approximately between 1400-1850 C.E. It is clear that there were some substantial globally distributed climate anomalies during this time [e.g., Mayewski et al. 2004], but it is also clear that different regions responded differently at different times [ Jones and Mann 2004]. One thing that we still do not know well for this time interval is the mean climate state of the tropical Atlantic, especially in terms of SST. Again, the results of existing research do not provide a straightforward answer to this question. Oxygen isotope data from marine sediment cores from the Sargasso Sea (33.693N, 57.612W) indicate approximately 1C of cooling [ Keigwin 1996] but that is a subtropical region, not a tropical one. Planktonic foraminiferal oxygen isotope records from the upwelling-prone Cariaco basin show a signal during 1400-1850 C.E. which may be more indicative of trade wind activity and hydrologic balance than regional SST [ Black et al. 2004]. Sclerosponge 18 O and Sr/Ca records from Jamaica indicate that mixed-layer temperatures were about 2C cooler than today during the Maunder (~16501750 C.E.) and Dalton (~1775-1850 C.E.) sunspot minima [ Haase-Schramm et al. 2003; Haase-Schramm et al. 2005] However, those geochemical records do not correlate particularly well with instrumental SST variability over the past 140 years and are suspected of being substantially influenced by sub-surface advective processes [ Haase

PAGE 60

45 Schramm et al. 2003]. The only other SST records from the Caribbean Sea between 1400 C.E. and 1850 C.E. are short coral records from Puerto Rico [ Winter et al. 2000; Watanabe et al. 2001], which indicate 2-3C cooling during the Dalton and Maunder sunspot minima (specifically the coral records span1699-1703 C.E., 1700-1705 C.E., 1780-1785 C.E., and 1810-1815 C.E.) Each of these records is only 5 years long and therefore lack context within the decadal and centennial timescales. In short, few records exist and replication is needed before we can confidently declare what climate conditions were like in the Caribbean over the last few centuries. Confirming the extent of tropical cooling in recent centuries is important because of the implications for the sensitivity of the climate system to small forcing. It is generally thought that the climate at Earths poles changes more drastically than the climate of the tropics in response to the same forcing because of albedo feedbacks and possibly because of atmospheric heat transport of polar amplification [ Cai 2005, 2006]. This concept, known as polar amplification is consistent with what is known about Earth during glacial times, with the current observed response of the climate system to CO 2 forcing, and with both simple and dynamically state-of-of-the-art climate models. Higher latitude cooling during the Little Ice Age is thought to have been on the order of 1-3C [ Overpeck et al. 1997 and references therein] so by the principle of polar amplification, one would expect much less change in the tropics. A two-degree cooling of the tropical Atlantic during the Little Ice Age, if confirmed, contradicts a fundamental idea about how the climate system works. Explaining such a phenomenon could considerably further our understanding of the climate system.

PAGE 61

46 This study addresses the outstanding issues of multidecadal variability and mean temperatures by producing a new 254-year coral-based paleoclimate record from southwestern Puerto Rico, near the same site as previous coral studies in Puerto Rico. Annual coral samples are analyzed for Sr/Ca and 18 O in order to produce records of thermal and hydrologic variability in the northern Caribbean since 1751 C.E. The primary goal of this study is to document climatic variability of the region and to put that variability into the context of existing marine and terrestrial paleoclimate records from the greater Caribbean region. 3.2 Analytical Methods A 245-cm long core of Montastraea faveolata was collected at Turrumote Reef (17.933N, 67.001W) offshore from La Parguera, Puerto Rico in August of 2004. The coral core, named 04 LPT A, was cut into 5-mm wide slabs along the primary growth axis, washed with tap water, and X-radiographed at Nova Southeastern University. The slabs were then cleaned with deionized water in a sonicator and nominally annual samples of coral powder were milled with a Dremel tool mounted on a computeroperated drilling stage using the X-radiographs as a guide to ensure that one density band couplet was sampled for each sample. Each sample is composed of powder collected from a dense band and the contiguous less dense band down core. At this location Montastraea high density bands tend to be laid down in the early summer [ Watanabe et al. 2002], so the samples are approximately centered on January of the year with which they are labeled. The annual coral powders were homogenized by grinding with an agate mortar and pestle until the sample was the consistency of cornstarch (~10-20 seconds). The top portion of the coral slab was also milled at approximately monthly resolution. A

PAGE 62

47 0.7mm diameter bit was used to drill a continuous 0.7mm wide path, removing samples every 0.6 mm down the path. Care was taken to extract material from the thecal walls and avoid other skeletal parts when sampling at both monthly and annual resolutions. The age model for these data was created by counting annual density bands. Where the core was broken into pieces, great care was taken to reconstruct the exact position of the pieces. The cut core pieces were stacked back together into round core sections with spacers representing the material lost in cutting the cores and with a lifesize copy of the X-radiograph taped to the micro-drilled slab. The two pieces of core were then put back together by aligning the corallites so that there were no gaps, and the X-radiographs on both sides of the core break were marked exactly where they met. We are confident that we accurately matched density bands over the core breaks using this method. Once the breaks were aligned, we used the X-radiographs to count density bands multiple times. We estimate an error of year at the bottom of the core. Sr/Ca, Mg/Ca, 18 O, and 13 C were analyzed at the University of South Florida, College of Marine Science from aliquots of the coral powder, using a Perkin Elmer Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) for the elemental ratios, and a ThermoFinnigan Delta Plus XL Mass Spectrometer with a Kiel Carbonate sample preparation device for the isotopic measurements. Stable isotope measurements are obtained by reaction of 35-100 g of aragonite with 100% phosphoric acid at 70C and the results are corrected to permil units relative to VPDB using a calcitebased correction. Long-term precision on 18 O and 13 C analyses is 0.06 and 0.03 (1 # ) respectively, determined by 6 measurements of NBS19 standard run daily with every set of 40 samples. Elemental data were obtained by dissolving approximately 75

PAGE 63

48 200 g of coral aragonite in enough 2% trace-metal grade HNO 3 to bring the concentration of calcium to 20 ppm (~1.5 4 ml) and analyzing the resulting solutions by ICP-OES using the method of [ Schrag 1999]. The precision of the Sr/Ca and Mg/Ca measurements is 0.15% or 0.013 mmol/mol and 0.25% or 0.010 mmol/mol (1 # N=145) respectively determined by a laboratory standard solution made with Porites lutea coral powder and run with the samples. The same precision was also obtained with a gravimetric standard solution run with the samples (N=97). The elemental analysis of the annual coral samples was replicated with separate aliquots of coral powder. The average absolute difference between any individual Sr/Ca measurement was 0.03 mmol/mol and 95.5% (approximately 2 # ) of the replicates were within 0.07 mmol/mol of each other. The difference between this replication precision and the analytical precision determined by standards run with the samples is that the replication precision includes error related to measuring a small aliquot of each 5-10 mg sample, which may not have been completely homogenized and which represents an entire years worth of temperature variability. The reported data are the average of both replicates and therefore have a 2 # precision of 0.05 mmol/mol (0.07 divided by the square root of 2). 3.3 The Modern Climate Signal Puerto Rican climate is typically tropical. The annual cycle of SST averages 27.9C with a 3.2C seasonal cycle between 1966 and 2002, measured from a bucket of water taken off the Magueyes Island dock every morning, about 3 km from our study site[ Winter et al. 1998] Peak SST occurs August through October and the lowest SST is January to March. A pronounced dry season occurs December-March and a bimodal wet

PAGE 64

49 season starts in April and ends in November, with a break in the rains during June and July, sometimes called Los Veranillos (Magueyes Island Precipitation record from the Southeast Regional Climate Center http://www.sercc.com/index.html ) Peak rainfall occurs in September through November. Analysis of salinity and temperature data (Figure 15) from the 1990s at the Caribbean Time Series (CaTS) station [ Corredor and Morell 2001] offshore from the coral study site indicates that the regional salinity minimum lags the temperature maximum by 2 months. The phasing of the SST and SSS annual cycles amplify the amplitude of coral 18 O measurements since the seasonal increases in temperature, which decrease coral 18 O, co-vary with decreases in salinity that also decrease coral 18 O. The average annual cycle in SSS spans 1.4 psu in the short 31 32 33 34 35 36 6/1/94 6/1/95 5/31/96 5/31/97 6/1/98 6/1/99 5/31/00 5/31/01 6/1/02 Date Surface Salinity (psu) 25 26 27 28 29 30 31 32 33 34 Surface Temperature (C) 0.1 0.3 0.5 0.7 6/1/94 6/1/95 5/31/96 5/31/97 6/1/98 6/1/99 5/31/00 5/31/01 6/1/02 Date Seawater 18 O ( SMOW) Figure 15: Surface temperature (black) and salinity (grey, upper graph) data from the CaTS station located at 67W and 17 37 N and the 18 Ow (grey, lower graph) implied by the salinity data and the local 18 Ow-salinity relationship from section1.1 [ Watanabe et al. 2001].

PAGE 65

50 CaTS time series, but the total range in that time series is 2.6 psu (which occurred in a single year), illustrating the potential for substantial salinity variations in this region. Puerto Rican climate is also connected to global-scale climate patterns (Table 2). SST variations from Magueyes Island [ Winter et al. 1998] correlate significantly with annual temperature from a single grid box including the study site [HadISST 1.1, Rayner et al. 2003], with the average temperature of the Western Hemisphere Warm Pool from the same dataset [ Wang and Enfield 2001, 2003] and with a Northern Hemisphere average [ Mann et al. 1999] The Caribbean Sea, including the waters just south of Puerto Rico, is influenced by continental runoff from the Orinoco River and possibly the Amazon River year round, but is seasonally inundated by runoff from the Orinoco River [ Froelich et al. 1978; Mller-Karger et al. 1989; Corredor and Morell 2001]. The peak influence of the Orinoco River in the northern Caribbean tends to occur in October, but is common during September and November [ Mller-Karger et al. 1989] and is coincident with the local peak in precipitation during September through November. This causes a potential complication in interpreting paleoclimate records from this area. Correlation Coefficient (r) WHWP Puerto Rico La Parguera Northern Hemisphere 0.78 0.66 0.70 WHWP 0.89 0.79 Puerto Rico 0.75 Table 2: Correlation coefficients (r) of annual SSTA scaling up from our local study site, to the Northern Hemisphere average. All correlation coefficients were calculated over the entire period of overlap of the data and all are significant above the 99% confidence level. Northern Hemisphere is from the Mann et al. [1999] compilation of instrumental data from 1902 to 1998. WHWP stands for Western Hemisphere Warm Pool [ Wang and Enfield 2001, 2003] and is the average SST from the Hadley Centers HadISST 1.1 data product spanning 110 W to 50 W, 7 N to 27 N, and 1870-2002. La Parguera represents the daily temperature measurements from the Magueyes Island dock that span 1965 to 2002.

PAGE 66

51 How can one differentiate local precipitation changes from changes in runoff from South America? The answer is that one does not have to, because at the most basic level both are related to the same phenomenon, ITCZ activity. The climatological average outgoing long-wave radiation shows a broad swath of low values over the Caribbean and northern South America during June-August, corresponding to higher clouds, deeper convection and higher rainfall rates in the region [ Poveda et al. 2006]. A study of the ITCZ from satellite imagery of highly reflective clouds shows that the ITCZ is rather diffuse over northern South America and the Caribbean from June through October, when it is its most northerly position [ Waliser and Gautier 1993]. Substantially higher salinities north of Puerto Rico indicate that it is on the northern edge of the region strongly influenced by tropical precipitation and runoff [ Corredor and Morell 2001], and it is thus an ideal location to detect meridional changes in the Hadley cell circulation, including the ITCZ and trade wind belt. As mentioned previously, the local (regional) 18 Ow and SSS covary with the following relationship: 0.20 0.03 per psu [ Watanabe et al. 2001; Watanabe et al. 2002]. The oxygen isotopic composition of seawater at the Pinnacles Reef, near our study site, is highest in the late spring before the heavy late season rains start, drops precipitously from July-November due to a combination of local rainfall and continental runoff, and begins increasing again in the boreal winter [ Watanabe et al. 2001; Watanabe et al. 2002]. The total seasonal amplitude is about 0.5. The 18 Ow SSS relationship is equivalent to conservative mixing between a -6.54 freshwater endmember and a 0.6 salt water (35 psu). This is very similar to mixing lines measured offshore of the Amazon River [ Karr and Showers 2002] and from the southern

PAGE 67

52 Caribbean[ Fairbanks et al. 1992], but quite unlike the mixing line between standard mean ocean water and precipitation from San Juan, Puerto Rico (-2) or Maracay, Venezuela (-3.4) [ Rozanski et al. 1993]. Unfortunately data addressing the oxygen isotopic composition of the Orinoco river could not be found. 3.4 Geochemical Results The time series of annual coral 18 O and Sr/Ca variations from 2004 to 1751 C.E. have mean values of .21 and 9.12 mmol/mol, respectively (Figure 16). Both records are dominated by a centennial-scale trend of decreasing 18 O and Sr/Ca values from the beginning of the record to the present (Figure 16, Figure 17, and Figure 18). The magnitudes of the trends, calculated by ordinary least squares regression are 0.086 mmol/mol and 0.44 equivalent to 1.8 C and 2.0 C respectively, if the 18 O is interpreted completely in terms of temperature (using 0.047 mmol/mol C -1 and 0.22 C -1 [ Leder et al. 1996; Swart et al. 2002], relationships that will be used throughout this paper). The errors on the temperature estimates are large ( 2.2C and 1.4C respectively) if properly propagated to include the error in the regression line and the error in the original Sr/Ca-SST and 18 O-SST calibrations (analytical error is implicitly included as well). Most of the error comes from the regression error, which is due to the higher frequency variability in the coral 18 O and Sr/Ca records. A similar, but more precise estimate of the trend comes from subtracting the mean values from the first 10 years of the records (2004-1995) and the last 10 years of the records (1760-1751). The mean values of 18 O and Sr/Ca from the first 10 years (.57 0.09 and 9.05 0.03 mmol/mol) and last 10 years (.97 0.07 and 9.15 0.02 mmol/mol) are statistically

PAGE 68

53 -4.8 -4.4 -4.0 -3.6 -3.2 8.8 8.9 9.0 9.1 9.2 1750 1800 1850 1900 1950 2000 18 O ( VPDB) Sr/Ca (mmol/mol) Warm Cool Warm/ Wet Cool/ Dry Figure 16: Annual variations in 18 O (, gray) and Sr/Ca (mmol/mol, black) from a Montastraea faveolata coral from southwestern Puerto Rico. Both axes are scaled to equal 10C using 0.047 mmol/mol C -1 and 0.22 C -1 [ Leder et al. 1996; Swart et al. 2002]. The long-term trends in coral Sr/Ca and 18 O are highlighted and have magnitudes of 0.086 mmol/mol (~1.8C) and 0.44 (2.0C), respectively. Hence our coral records suggest that there has been ~2C of warming of SST in this region since 1751 C.E. The larger interannual variance in Sr/Ca relative to 18 O is interpreted as noise in Sr/Ca record rather than true temperature variations. -0.6 -0.4 -0.2 0 0.2 0.4 0.6 1750 1800 1850 1900 1950 2000 Coral 18 O ( VPDB), mean removed PuertoRico Coral d18O PC1 (42.6%) PC2 (9.9%) PC3 (7.4%) Figure 17: Coral 18 O with the first three principle components calculated by singular spectrum analysis with a window length of 25 [ Dettinger et al. 1995; Ghil et al. 2002] illustrating that the trend in the data is robust. Singular spectrum analysis assumes no a priori model of the variability, as opposed to linear regression.

PAGE 69

54 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 1750 1800 1850 1900 1950 2000 Coral 18 O ( VPDB) Puerto Rico Coral M_15 M_100 M_50 M_25 M_20 Figure 18: Comparison of annual coral 18 O data with the first principle components of singular spectrum analyses [ Dettinger et al. 1995; Ghil et al. 2002] carried out with five different window (M) lengths: 15, 20, 25, 50, and 100. The results from a singular spectrum analysis are highly dependent on the window length and the results for many different window lengths confirm that a trend and multidecadal variability are the primary sources of variance in the coral 18 O record. different. The error reported for the mean values are two standard errors, calculated by dividing two standard deviations by the number of data points. The difference between the earliest and latest portions of the record is 0.10 0.04 mmol/mol and 0.6 0.1 These differences translate into 2.1 0.8 C and 2.7 0.5 C for Sr/Ca and 18 O respectively and are larger than that reflected in the long-term regression line because values for the most recent years depart from the regression line ( 18 O and Sr/Ca have both been slightly lower than expected by the long-term trend). A multidecadal oscillation is visible in a plot of the 18 O data and is confirmed by both SSA and Blackman-Tukey spectral analysis on the normalized and de-trended data (Figure 19, [ Dettinger et al. 1995; Ghil et al. 2002]). The dominant period is 60 years

PAGE 70

55 per cycle, but other periods with highly significant variance (>99% confidence level when tested against red noise) are 36.6 years, 4.7 years, and 2.3 years (Figure 20). 3.5 Discussion 3.5.1 Documentation of the Climate Signal Before the coral data are interpreted in terms of a climate signal, the coral records must be established as faithfully recording the expected climate signal. Approximately monthly data from the top 10 years of the core are shown in Figure 21. All variables measured are plotted, although we focus on the Sr/Ca and 18 O in the longer record. The -3 -2 -1 0 1 2 3 4 -2 -1 0 1 2 3 4 5 6 1750 1800 1850 1900 1950 2000 Normalized, Detrended Puerto Rico Coral 18 O Normalized Cariaco Basin % G. bulloides Figure 19: Puerto Rico coral 18 O (thin black line) and Cariaco basin G. bulloides abundance (thin grey line) from Black [1999] detrended and normalized to unit variance and zero mean. Interdecadal variability is emphasized by the reconstruction (thick black line) of the large interdecadal spectral peak (60 year period) from a Blackman-Tukey spectral analysis performed using the SSA-MTM toolkit [ Dettinger et al. 1995; Ghil et al. 2002]. To emphasize the robust nature of this signal, the first principal component of a singular spectrum analysis (25-year window) on the detrended data was also used to reconstruct the time series (dotted thick line). The SSA results were similar using a wide range of window lengths. Note that the G. bulloides abundance always peaks just before the Puerto Rico coral 18 O reaches a minimum, shown by the grey bars.

PAGE 71

56 0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 Frequency (cycles/year) Power Power Spectrum 90% 95% 99% 60.2 years 36.6 years 4.7 years 2.3 years 0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 Frequency (cycles/year) Power Power Spectrum 90% 95% 99% 56.9 years 29.3 years 5.2 years 2.3 years Figure 20: Power spectra of time series of Puerto Rico coral 18 O (left) and Cariaco Basin G. bulloides abundance (right). The Puerto Rico coral 18 O shows a strong peak of spectral power above the 99% confidence level at 0.0166 cycles/year (60.2 year period), with other highly significant peaks at 0.0273 cycles/year (36.6 years), 0.2128 cycles/year (4.7 years), and 0.4348 cycles/year (2.3 years). Only the first 249 years of the foraminiferal record were used in the analysis (1988 to 1740) and the data were resampled to annual resolution, so that the spectra were more comparable. The resulting spectrum has significant peaks at approximately the same frequencies as the coral-based spectrum. strongest annual cycles are in 18 O, not surprising given the amplifying effects of covarying salinity and temperature on 18 O. Temperature-driven seasonal changes in Sr/Ca are obscured and this will be discussed below. Unusually, the carbon isotope record does not show a strong annual cycle either. Carbon isotopes in corals are complex and have been interpreted as indicators of ocean dissolved inorganic carbon 13 C, cloud cover, respiration/photosynthesis within the coral, and biomineralization processes. For a review of the literature, see Druffel [1997a]. Mg/Ca has strong peaks that correlate with bands of endolithic algae, but does not display a strong seasonal signal even after the larger peaks are removed. We will not be relying on Mg/Ca or 13 C in the rest of this

PAGE 72

57 0 5 10 15 20 25 30 0 2 4 6 8 10 12 Mg/Ca (mmol/mol) Depth Down Core (cm) 8.9 9.0 9.1 9.2 9.3 9.4 9.5 Sr/Ca (mmol/mol) -5.5 -5.0 -4.5 -4.0 -3.5 0 2 4 6 8 10 12 18 O (VPDB) -5.0 -4.0 -3.0 -2.0 -1.0 0.0 13 C (VPDB) Figure 21: Geochemical results from the top of coral core 04LPTA, including 18 O, 13 C, Sr/Ca, and Mg/Ca. Oxygen isotopes have the clearest annual cycle and were used to create an age model by tuning the record to monthly sea surface temperature. This age model was confirmed via comparison with annual density bands. analysis, so they will not be discussed further, however the Sr/Ca and 18 O data deserve more scrutiny and discussion. The coral 18 O data compare well with local SST and have a correlation coefficient of r=0.75 (Figure 22A), which is significant (p<0.01) even assuming only one independent data point per year. A seawater 18 O record generated using the local temperature data, the calibration equation of Leder et al. [1996], and the calculation

PAGE 73

58 method of Gagan et al., [1998] is plotted in Figure 22B. This method uses an equation in the form of T=mO+b, where T is the temperature and O is the oxygen isotopic composition of a coral, to calculate the paleotemperature estimate ignoring 18 Ow changes. The difference between this estimate of temperature and an independent measure of temperature, in this case the instrumental temperature record, is a measure of the influence of 18 Ow on the paleotemperature estimate. The units of the residual are subsequently converted to per mil using the slope (m) of the paleotemperature equation used previously, in this case, the equation of Leder et al. [1996]. The measured 18 Ow data correspond well with the calculated 18 Ow within the estimated error of the age model ( 2 months) and the propagated error of the calculations ( 0.3 ), depicted in the upper left hand corner of Figure 22B. This is especially impressive given that the 18 O record has been tuned to temperature and therefore is not independent of the temperature, as one would prefer for these calculations. This also shows the robust nature of the Leder et al. [1996] calibration equation for use at this site. Times when the coral 18 O appears to not follow temperature tend to have substantial precipitation anomalies. For example, the pronounced drought in 1997 and 1998 is well captured in the 18 O data. Hurricane-produced rainfall effects are not apparent in the geochemical time series, and this is probably because hurricanes and their associated rainfall are relatively short-lived events. We tested this by drilling19981999 at 50 samples per year and analyzing the resultant powders for 18 O, looking for a strong depletion during Hurricane Georges, which dumped 23.4 cm of rain in three days in the town of Lajas, just 12 km north of La Parguera. No significant depletion in coral 18 O was observed in these high-resolution data (

PAGE 74

59 Figure 23), nor in the monthly data 24 25 26 27 28 29 30 31 1992 1994 1996 1998 2000 2002 2004 2006 La Parguera Dock SST (C) -5.20 -4.98 -4.76 -4.54 -4.32 -4.10 -3.88 -3.66 18 O ( PDB) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1992 1994 1996 1998 2000 2002 2004 2006 Seawater 18 O ( VPDB) 0 5 10 15 20 25 30 35 40 45 50 Precipitation (cm) Hortense Georges Figure 22: A) Comparison of monthly resolved coral 18 O (black) with a local record of monthly SST(gray), which was calculated from daily measurements from the Magueyes Island dock. B) Variations in seawater 18 O, calculated from the data in A using the method of Gagan [1998] and the calibration equation of Leder et al. [1996] (thick, gray line). The error barin the upper left represents a month estimated error in the age model (x-axis) and the error propagated through the seawater 18 O calculations that includes analytical error as well as calibration uncertainty (y-axis). Black diamonds represent seawater 18 O measurements from Watanabe and coauthors [ Watanabe et al. 2001]. Precipitation data are monthly rainfall totals from the Lajas and Magueyes Island rain-gauge stations averaged together so that the occasional gap in each one is made up for by the other (thin grey line). The monthly precipitation record has been smoothed A B

PAGE 75

60 with a 3 month running average (thick black line). The timing of the passages of hurricanes Hortense and Georges are represented by vertical lines. -5 -4.8 -4.6 -4.4 -4.2 -4 -3.8 -3.6 1/1/97 1/1/98 1/1/99 1/1/00 Date Coral Oxygen Isotopes ( VPDB) Approx. Fortnightly Samples Approx. Monthly Samples 3 per. Mov. Avg. (Approx. Fortnightly Samples) Figure 23: High resolution coral 18 O data compared with approximately monthly samples over a year spanning 1998 to 1999 show no significant depletion in 18 O during the fall of 1998 when Hurricane Georges passed over the area. (Figure 22). The only other named tropical storm to pass within 75 km of La Parguera during 1993-2003 (Hurricane Hortense) did not produce much of a rainfall anomaly and did not register as an isotopic depletion in the coral 18 O (Figure 22). In summary, the 18 O appears to be recording a combined signal of variability in SST and seawater 18 O, however individual short-lived tropical storm events do not substantially influence the record. The poor correlation between Sr/Ca and temperature in the youngest part of the record is investigated here by going though possible explanations given that Sr/Ca measured in corals can be influenced by skeletal sampling method, analytical procedure, seawater Sr/Ca values, poorly understood biomineralization processes (including growth rates), and of course temperature (the primary control in most cases). The variance in the Sr/Ca record is much larger than analytical error as measured repeatedly by a coral-based standard solution, so analytical error is not likely the major source of the Sr/Ca variability. Seawater Sr/Ca variability can be ruled out as well because seawater chemistry should affect all corals in an area and other authors looking at a different

PAGE 76

61 species of coral ( Acropora palmata ) from the same reef area have found Sr/Ca cycles which clearly relate to SST [ Gallup et al. 2006]. Annual extension rates were tested by regressing them onto the annual Sr/Ca data for all of the record, resulting in a statistically significant equation of Sr/Ca = -0.007 .002 *(extension in mm) + 9.1769 r 2 = 0.04, n = 253, p < 0.001 (Figure 24). If the data are corrected using this equation, the average absolute value of the correction factor is 0.007mmol/mol, a value that is even less than the analytical error, so although growth rate effects cannot be completely ruled out without measuring calcification, it seems an unlikely source of the problem identified in this coral Sr/Ca record. The reason for the noisy Sr/Ca could be that it came from an aberrant individual coral head, but that possibility was discounted by sampling the top 5.25 centimeters from cores of four other coral heads gathered at the same time and from the same area (Turrumote Reef and Pinnacles Reef) as the one in this study. Each coral produced similarly noisy data sets that did not have clear annual cycles (Figure 25). An 8.55 8.65 8.75 8.85 8.95 9.05 9.15 9.25 1750 1800 1850 1900 1950 2000 Sr/Ca (mmol/mol) 0 2 4 6 8 10 12 14 16 Annual Extension Rate (mm/year) Figure 24: Time series and regression of annual extension rates (grey) with Sr/Ca (black) for a Puerto Rican coral. Sr/Ca is essentially independent of extension rates. y = -0.0071x + 9.1769 R 2 = 0.0432 8.9 9.0 9.1 9.2 9.3 4 6 8 10 12 14 16

PAGE 77

62 8.8 9.0 9.2 9.4 0 1 2 3 4 5 6 04LPPB1 8.8 9.0 9.2 9.4 04LPPC1 8.8 9.0 9.2 9.4 04LPPD1 8.8 9.0 9.2 9.4 0 1 2 3 4 5 6 04LPPH1 Core Depth (cm) Figure 25: Coral Sr/Ca (mmol/mol) from four different coral heads growing at the same time on the same reef. Annual cycles are not apparent in any of the cores, but the mean values are similar. explanation for the noisy Sr/Ca data is still lacking despite testing for a variety of possibilities. Further work is needed to explain these data. The multiple coral reproducibility test described above had an interesting positive outcome. The mean Sr/Ca value of the five coral time series was quite reproducible (9.12.03 mmol/mol, 1 # ), despite the variability in the data. We interpret this to indicate that while we cannot rely on Sr/Ca produced by our methods to reflect seasonal variability using the monthly resolved time series, the mean over many years could still reflect a robust temperature signal. 3.5.2 Long-Term Trends The long-term Sr/Ca trend in the Puerto Rican coral time series indicates ~2C of warming in the Caribbean since 1751 C.E. This result is consistent with the findings of previous work in the area [ Winter et al. 2000; Watanabe et al. 2001; Haase-Schramm et

PAGE 78

63 al. 2003], and provides replication and context for the previous work. The 2C difference between the 18 th and the earliest part of the 21 st centuries is primarily due to a long-term trend as evidenced by both the Jamaican sclerosponge and Puerto Rico coral Sr/Ca records (Figure 26). According to the sclerosponge record, the 18 th century was the beginning of the trend, but our coral record does not extend far enough back in time to confirm that. Future studies on existing older coral cores could confirm the temporal extent of the trend. There is some reason to believe that the regional evaporation-to-precipitation ratios may have changed over this period as well. The trend in the oxygen isotope data from our Puerto Rican coral is consistently slightly steeper than the Sr/Ca trend when both are converted to temperature, no matter how the trend is calculated. The greater change in 18 O indicates that seawater 18 O might have decreased with time, amplifying the 18 O signal due to warming temperatures over the same period. Using the ordinary least squares regression of the data as a measure of the trend as described in the results section, and ignoring the propagated errors, indicates a 0.2C difference in the Sr/Cabased temperature estimate and the 18 O-based estimate. That translates to a 0.5 psu salinity decrease over the length of the record using the local 18 O-SSS relationship of 0.20/psu. It is possible that the local 18 O-SSS relationship changed through time, but available 18 Ow and salinity data from the North Atlantic subtropical gyre [ Schmidt et al. 1999] and from work in the Lesser Antilles of the Caribbean [ Swart et al. 2003] indicate that the 18 Ow-SSS slopes are similar (0.2 /psu) in the surrounding regions. If the fresh water end member were to change by a substantial change, then the 18 Ow-SSS slope would change by only .05 (0.14 to 0.25 /psu). Thus it is assumed for the

PAGE 79

64 0 500 1000 1500 2000 2500 Cariaco G. bulloides abundance 9.2 9.4 9.6 9.8 10.0 Jamaica Sponge Sr/Ca (mmol/mol) -6.4 -6.0 -5.6 -5.2 Panama Coral 18 O ( VPDB) -4.8 -4.4 -4.0 -3.6 1700 1750 1800 1850 1900 1950 2000 Puerto Rico 18 O ( VPDB) -1.0 -0.9 -0.8 -0.7 -0.6 1700 1750 1800 1850 1900 1950 2000 Jamaica Sponge 18 O ( VPDB) Little Ice Age Figure 26: Paleoclimate records from the Caribbean and Eastern Pacific showing 2C cooler temperatures in the 18 th century relative to todays temperature. The Puerto Rico record is from our coral, the Panama coral record is from Linsley et al. [1994] the Cariaco foraminiferal abundance record is from Black et al. [1999] and the two Jamaican Sponge records are from Haase-Schramm et al. [2003]. The long-term trends are emphasized by the inclusion of ordinary least squares regression lines. present application that no significant change in 18 Ow-SSS slope has occurred. The observed freshening trend is consistent with trends since 1750 C.E. in Panamanian coral 18 O [ Linsley et al. 1994], Cariaco Basin Globigerina bulloides

PAGE 80

65 abundance [ Black et al. 1999], and foraminiferal-based 18 Ow changes in the Florida Current [ Lund and Curry 2006]. The coral 18 O record from Panama spans the years 1707 to 1984 and is driven primarily by ITCZ-related rainfall on seasonal time-scales [ Linsley et al. 1994] It has a trend of the same magnitude and sense as the Puerto Rican coral 18 O (Figure 26), and although an independent measure of temperature change is needed to determine exactly how much of the trend is due to seawater 18 O and how much is due to temperature, it is likely a combined signal that includes some salinity decrease. Further evidence for freshening comes from the Cariaco Basin. A slight decrease in the abundance of G. bulloides in the Cariaco Basin since the 1650s indicates a trend toward less intense upwelling, decreasing trade winds, and increasing precipitation [ Black et al. 1999], although the trend is slight and not statistically significant. Paired analyses of Mg/Ca and 18 O on foraminifera from near the Dry Tortugas, Florida, indicates a drop in 18 Ow since about 250 years ago [ Lund and Curry 2006]. This area is generally bathed by the Florida Current, which comes directly from the Caribbean and flows rapidly through the Gulf of Mexico as the Loop Current, so the data likely primarily represent conditions in the Caribbean with only a small influence from the Gulf of Mexico. In contrast to the freshening trends discussed above, other paleoclimate records suggest the opposite trend, but easily can be explained in the context of our coral results. For example, Jamaican sclerosponge 18 O and Sr/Ca records [ Haase-Schramm et al. 2003] indicate that 18 Ow may have actually increased substantially in the Northern Caribbean (Figure 26). However, the sclerosponge grew at 20m water depth, close to the modern day pycnocline [ Haase-Schramm et al. 2003] separating the isotopically heavier

PAGE 81

66 Subtropical Under Water from the isotopically lighter Caribbean Surface Water. It is possible that changes in the density structure of the overlaying water, changes in wind stress, or current intensities caused the sclerosponge to be bathed with more Subtropical Under Water, thus increasing its 18 O value. Other records that do not show a substantial freshening trend over the period from 1750 to present are the Cariaco G. ruber 18 O record [ Black et al. 2004] and Yucatan lake 18 O records [ Hodell et al. 2005b]. The Cariaco Basin G. ruber 18 O record, continues a millennial-scale drying trend during the last 250 years, and shows a very strong increasing trend after about 1950. However, the Cariaco Basin is an upwelling-prone system that may have been influenced by the 18 O of subsurface water to varying degrees through time like the Jamaican sclerosponge record. Lake records from the Yucatan show that dry conditions dominated starting in the 15 th century, but display only a slight, if any, return to more moist conditions since 1750 C.E. In summary, although the data are not completely consistent, most records indicate that regional surface 18 Ow may have decreased as temperatures increased since the 18 th century. Why might the Caribbean have been ~2C cooler and 0.5 psu saltier in the 18 th century? High latitude cooling in the Northern Hemisphere during this time period [e.g., Moberg et al. 2005] may have increased the thermal gradient, thus strengthening the northeast trade winds, cooling the northern tropics, and causing a decrease in the amount of ITCZ-related rainfall either by a decrease in convective activity or a southward movement of the ITCZ [ Haug et al. 2001b]. Inspection of NCEP reanalysis [ Kalnay et al. 1996] wind and SST monthly anomaly data confirmed that persistent (>6 months) SST anomalies of > 1C have occurred in Caribbean over the last 50 years, apparently in

PAGE 82

67 response to anomalous winds (not shown). If the wind over the Caribbean was consistently stronger because of a centennial-scale anomaly, it seems reasonable that a similar cooling response would occur. This is by no means a definitive explanation. However, our data provide a reference temperature to be reproduced by climate modeling in order to explore mechanistic explanations while taking into account the complex interactions of the climate system. In general, the trend since 1750 is consistent with previous work indicating a substantial radiative forcing (solar and volcanic) mechanism for centennial-scale variability [e.g., Crowley 2000a]. 3.5.3 Multidecadal Variability 3.5.3.1 Tropical Patterns Interdecadal variability in the Puerto Rico 18 O record is strong (Figure 19). A cycle with a 60-year period is visibly apparent and highly significant (p < 0.01) using spectral analysis with a Blackman-Tukey window, testing against a red-noise background [ Ghil et al. 2002]. The first principal component of a Singular Spectrum Analysis on the detrended Puerto Rico coral 18 O results in essentially the same multidecadal signal [ Ghil et al. 2002], providing additional support for the robustness of the signal (Figure 19). The next question is, how much of the coral 18 O oscillation is temperature, and how much is 18 Ow? The most obvious way to answer that question is to use the Sr/Ca data from the same core. However, Sr/Ca in this coral does not correlate well with decadal-scale changes in temperature when compared with either the local SST record [ Winter et al. 1998], or the HadISST 1.1 SST data for this region [ Rayner et al. 2003] This was tested

PAGE 83

68 by subtracting the climatology from the monthly SST data, then smoothing the resultant anomalies with a 10-year running average, and similarly smoothing the annual Sr/Ca data before performing the comparison. The range of the smoothed Sr/Ca over the instrumental period is equivalent to 2.7C, much greater than the 0.5C range in the temperature anomaly over the same period. There is little covariance between the records, leading us to conclude that the noise in the Sr/Ca signal is too great and obscures the temperature signal, even on decadal scales. Instrumental SST can be used instead of Sr/Ca to place constraints on the roles of temperature and salinity in the 18 O record. Annual average SST was calculated for 2002 through 1871 using July through June monthly data in order to mimic the time represented by the annual coral samples. The SST data were detrended, expressed as anomalies from the 2002-1871 mean value, and smoothed with a 10-year running average. The coral 18 O data were similarly detrended and smoothed. The resulting curves isolate the decadal variability and are very similar to those in Figure 27 except that the curves in Figure 27 have been normalized to have a standard deviation of one. The total range of variation in the 18 O record is 0.39 whereas the total variation in the SST is 0.56 C. Using 0.22C -1 [ Leder et al. 1996], the SST range can account for 0.12 or 30% of the 18 O range, leaving 0.27 or 70% to be accounted for by 18 Ow and thus salinity changes. The salinity change implied in 0.27 is 1.3 psu assuming no change in the 18 O-SSS relationship with time. Seasonal salinity changes in southwestern Puerto Rico are due to the competing influences of evaporative trade winds and ITCZ-related runoff and precipitation [ Corredor and Morell 2001]. Since Puerto Rico lies on a relatively steep north-south salinity gradient [ Corredor et al. 2003], it is reasonable to

PAGE 84

69 -2.5 -1.5 -0.5 0.5 1.5 2.5 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 Standard Deviation Units -2.5 -1.5 -0.5 0.5 1.5 2.5 Standard Deviation Units AMO Index SST anomaly Gray AMO Index d18O ( VPDB) Figure 27: Multidecadal variability in Atlantic records illustrating the strong coherence between the AMO index and proxy data from tree rings (Gray AMO index) and coral 18 O (VPDB) during the instrumental record, and the disconnect between the proxy data during the pre-instrumental time. The AMO index is calculated using Kaplan SST V2 [ Kaplan et al. 1998] after the method of [ Enfield et al. 2001] and was downloaded from http://www.cdc.noaa.gov/Timeseries/AMO/ SST anomaly is the monthly SST anomaly for the 1 square encompassing our study area from the HadISST 1.1 dataset similarly detrended and smoothed. Gray AMO Index is the tree ring-based reconstruction of the AMO from [ Gray et al. 2004], and 18 O (VPDB) is the annual Puerto Rican coral 18 O detrended and smoothed with a 10-year box-car filter. All of the data have been centered, reduced to annual resolution if originally monthly, and normalized to one standard deviation. expect substantial changes on multidecadal time scales. Measurements of local 18 Ow during the 1990s [ Watanabe et al. 2001] indicate the average local 18 Ow value to be 0.60 VPDB with a range of 0.2 to 1.2 whereas in the subtropical gyre to the north, where evaporation is the dominant process, 18 Ow ranges between 0.9 and 1.3 (between 60.82 to 18 W and 39 to 20 N, < 40 m water depth, [ Schmidt et al. 1999]). A southward shift in the Hadley circulation or an intensification of the trade winds could easily drive mean seawater conditions at Puerto Rico toward saltier and more isotopically enriched conditions. However, one can also envision a scenario whereby Puerto Rican seawater becomes fresher or saltier depending on how much continentally influenced Instrumental record

PAGE 85

70 southern Caribbean water is pushed northward by Ekman transport. In that case, an increase in the trade wind strength would increase the northward Ekman transport and cause a decrease in the local 18 Ow. One clue for the cause of the observed multidecadal changes in Puerto Rico salinity comes from the Cariaco G. bulloides abundance record of Black et al. [1999], which represents a record of upwelling and thus trade wind strength in the southern Caribbean. Their record displays multidecadal variability that is very similar to our coral 18 O record (Figure 19), and that is highly significant in a spectral analysis of the data since 1750 (Figure 20). Periods of upwelling in the Cariaco basin identified by increased G. bulloides abundance tend to be associated with periods when Puerto Rico 18 O decreases. This is consistent with strengthening northeasterly trade winds causing increased upwelling along the northern coast of South America, and causing increased northward Ekman transport of low salinity surface water. Supporting this idea is the fact that the samples with the lowest 18 O values in each cycle tend to also have the lowest 14 C values, indicating more equatorial water surrounding Puerto Rico when 18 O values are low (see Kilbourne et al., [2006b] for further explanation). Earlier work on the G. bulloides record used the entire 800-year record in a spectral analysis and did not find significant variance in the multidecadal band. This indicates that while the 60-year period variance is important over at least the last 250 years, other modes of variability dominate the record at different times. The timedependent nature of the spectral analysis leads to the hypothesis that the mechanism causing the 60-year oscillation is sensitive to background climate and has not always been as active in the last 800 years as it has in the last 250 years.

PAGE 86

71 3.5.3.2 Tropical and Extra-Tropical Comparison The multidecadal oscillation in the Puerto Rican coral 18 O record is strikingly similar to instrumental records of hemispheric SST anomalies in the North Atlantic (Figure 27) known as the Atlantic Multidecadal Oscillation or AMO [ Kerr 2000]. Climate modeling studies have shown that changes in the strength of the Atlantic meridional overturning circulation can cause hemispheric temperature anomalies similar to the observed pattern of the AMO [e.g., Delworth and Mann 2000; Vellinga and Wood 2002]. A precise mechanism for the AMO is still being investigated in a variety of modeling efforts [ Delworth et al. 1993; Timmermann et al. 1998; Delworth and Greatbatch 2000; Delworth and Mann 2000; Vellinga and Wu 2004; Dijkstra et al. 2006]. Despite uncertainties about the mechanism driving the AMO, all agree that the changes in Atlantic meridional overturning circulation (a.k.a. thermohaline circulation) are linked to changes in atmospheric circulation. The AMO has been linked to specific atmospheric phenomena such as precipitation patterns in North America [ Enfield et al. 2001] and tropical storm frequency and intensity [ Landsea et al. 1999; Goldenberg et al. 2001]. It is therefore not surprising that we see an AMO-like signal in our coral 18 O manifest as a signal of another atmospheric phenomenon trade wind strength. The instrumental record of the AMO is not long enough to determine if it is a true oscillation, so researchers have turned to climate proxy records such as tree-rings, ice cores, and corals to look for evidence of a persistent pattern. The two studies that have found an AMO-like signal rely heavily on tree-ring records from north of 30N [ Delworth and Mann 2000; Gray et al. 2004]. This leaves open for debate whether or not the AMO has an impact on the tropical Atlantic. Our data indicate that indeed the

PAGE 87

72 AMO does have a significant impact on the tropical Atlantic, at least over the last 250 years. It is interesting to compare our coral-based record of AMO-related variability with the tree-ring based record of Gray et al. [2004] (Figure 27). Both proxies track the instrumental record of North Atlantic SST anomaly well, but there seems to be a disconnect prior to about 1860 C.E., when the instrumental record begins. What could cause two previously unrelated records to suddenly become coherent beginning in ~1860 C.E.? Some aspect of the climate system must have changed and must have affected one proxy more than the other. Since the coral record is periodic and seems to have a relatively constant character, we assume that the change occurred in the higher latitude records. The Gray et al. [2004] record is comprised of tree-ring records from the southeast United States, Fennoscandia and around the Mediterranean. Tree rings often record moisture availability as well as temperature [e.g., Cook et al. 1998]. All of these areas are highly sensitive to changes in the Westerlies and the position of the average storm track on interannual scales [ Marshall et al. 2001] It stands to reason that they would be sensitive to such changes on even longer time scales. One possibility to explain our divergent AMO records prior to 1860 C.E. is that centennial-scale variability could interact with the multidecadal mode enough to mask it in the extra-tropical records. For example, a particularly active North Atlantic polar front is hypothesized to have induced more storms in the North Atlantic region during the Little Ice Age [ Maasch et al. 2005] and that could have affected the tree-ring records.

PAGE 88

73 3.5.4 Interannual Variability The strong spectral peaks at 4.7 and 2.3 years (Figure 20) coincide with the variance in the quasi-quadrennial and quasi-biennial modes of El Nio-Southern Oscillation (ENSO) variability [e. g., Rasmusson et al. 1990; Jiang et al. 1995]. The higher frequency peak is too close to the Nyquist frequency (1 cycle/ 2 years) to interpret with confidence, but the 4.7-year peak is probably meaningful. ENSO variability in the Caribbean occurs by teleconnection that may be dependent on conditions within the Atlantic at the time of an ENSO event [ Giannini et al. 2001b; Wang 2001; Alexander and Scott 2002; Wang 2006] It has been proposed that the strength of ENSO influence in the tropical Atlantic varies with time and was much weaker in the early part of the 20 th century than in the latter part [ Diaz et al. 2001]. A wavelet analysis [ Torrence and Compo 1998] of our coral 18 O data shows that the strongest interannual variability occurs during the earliest 50 years of the record, and after that, some time periods show more interannual variance than others (Figure 28). The periods of significant interannual variability do not change if the first 60 years are excluded from the analysis. Dong et al. [2006] suggested that changes in Atlantic SST (i.e., the AMO) could play a role in the modulation of ENSO on multidecadal time scales. We do not see a strong relationship between multidecadal variability and interannual variability in our data, but our study site is connected to the ENSO by an atmospheric bridge [ Giannini et al. 2001b], so there may be a relationship that we just cannot detect. A similar evolutionary spectral analysis of a coral record from the Galapagos islands extending back to nearly 1600 C.E. demonstrates centennial-scale shifts in the frequency of ENSO [ Dunbar et al. 1994], but does not show strong multidecadal modulation of the

PAGE 89

74 Figure 28: Wavelet analysis of Puerto Rican coral 18 O showing the dominance of the 60 year cycle in the data and showing the evolution of the variance at lower periods (36 years, 4.7 years, and 2.3 years). The linear trend was removed before performing the analysis and significance was tested using a red-noise background spectrum. The bold black outlines represent areas that are significant at the 95% confidence level, and the cone of influence represents the area possibly influenced by edge effects. Much of the interannual variance occurs between 1750 and 1800. However, the regions of significant interannual variance in the rest of the record do not change when the 1750-1809 interval is excluded from the analysis.

PAGE 90

75 ENSO signal. Interestingly, the Galapagos record indicates that the 1700s were unusually warm and had fewer ENSO events compared to the preceding and subsequent centuries [ Dunbar et al. 1994] This contrasts sharply with the strong interannual variability seen at this time in our Caribbean coral, and may indicate that the interannual variability seen in our coral at that time has an Atlantic origin. 3.6 Conclusions We have produced a 253-year record of coral 18 O and Sr/Ca variations in the northern Caribbean Sea (southwestern Puerto Rico), which we interpret in terms of SST and SSS variations. The long-term trends in both data sets indicate that conditions were ~2 C cooler in the 18 th century compared to today. This study confirms previous studies suggesting substantial cooling in the Caribbean during the Little Ice Age [ Watanabe et al. 2001; Haase-Schramm et al. 2003] The data indicate that salinities may also have been slightly higher 250 years ago, but that is not as clear because the magnitude of change (~0.5 psu) is small compared to the errors in calculating it, and some inconsistencies exist among the different data sets for the region. Multidecadal variability in our data has a strong 60-year cycle that we interpret to be linked with changes in trade wind strength and the associated advection of low-salinity water from the south by Ekman transport. Temperature variations play only a minor role in the observed variability, but salinity variations are ~1.3 psu on these time scales. The timing of the multidecadal oscillations is consistent with a linkage to the AMO. If true, our data indicate the persistence of the AMO back to at least 1750, and indicate that the tropics are indeed affected by AMO variability, contrary to a recent report [ Mann and Emanuel 2006] The difference between a tree ring-based AMO reconstruction [ Gray et

PAGE 91

76 al. 2004] and our multidecadal signal is attributed to substantial atmospheric reorganization at mid to high northern latitudes during the Little Ice Age. We hypothesize that the centennial-scale variability masks any AMO signal in the tree-ring records because the regions where they came from are highly sensitive to atmospheric circulation patterns in the North Atlantic. The spectral characteristics of interannual variability in our core indicate significant variance at the frequencies of the quasi-quadrennial and quasi-biennial modes of ENSO. Interannual variance in our record has the most power between 1750 and 1800, a time when ENSO variability in the eastern Pacific was at a minimum, according to a coral from the Galapagos [ Dunbar et al. 1994]. This may indicate that the strong interannual variations in our record actually have an Atlantic origin. Further paleoclimate work is needed in this region to address several issues. Substantial cooling during recent centuries in the Caribbean raises again the question of tropical climate sensitivity. Older high-resolution records could tell us about the persistence of the cool temperatures found by our study and permit an assessment of the role of solar variability in Caribbean climate. Older records could also address the question of how long the multidecadal oscillations have been operative. As always, replication of paleoclimate results is important, and a study that replicated the multidecadal modulation of interannual variability found in our work could make much more substantial statements about teleconnections with ENSO and about Atlantic interannual variability.

PAGE 92

77 4 CORAL GEOCHEMICAL RECORDS OF CLIMATE IN THE CARIBBEAN OVER THE LAST 800 YEARS 4.1 Introduction Global climate of the last 1000 years can be characterized by three distinct periods, an early warmer period, known as the Medieval Warm Period (MWP, [ Lamb 1965]), a later cooler period, known as the Little Ice Age (although controversy still surrounds the use of this term [ Matthews and Briffa 2005]), and the most recent ~200 years dubbed the Anthropocene because of the impacts of human activities on Earths processes [ Crutzen 2002]. The spatial and temporal heterogeneity of paleoclimate signals [e.g., Jones and Mann 2004] has lead to some discussion about the global nature and timing of these events [ Bradley et al. 2001; Broecker 2001]. However, these relatively warm and cool phases consistently appear in hemispheric and global reconstructions of Earths temperature [e.g., Mann et al. 1999; von Storch et al. 2004; Moberg et al. 2005; D'Arrigo et al. 2006] and a growing consensus is establishing that heterogeneous local and regional paleoclimate signals [ Jones and Mann 2004] are not inconsistent with global scale changes but are in fact to be expected because of the role that internal variability plays in climate on these time scales [ Goosse et al. 2005]. The causes of past global mean surface temperature fluctuations have been investigated using simple climate models and it has been found that solar and volcanic forcing can explain the general pattern of global temperatures prior to the 20 th century

PAGE 93

78 [ Crowley 2000b; Bertrand et al. 2002; Bauer et al. 2003; Gerber et al. 2003]. Yet, the climate system contains many complex feedbacks and processes that are not resolved by these simple models, so high resolution coupled climate models of the last 1000 years are needed to explore what portions of the climate system actually changed in response to known perturbations. In other words, we need models to explore the causes of the observed regional climate variations and to differentiate internal variability from forced variability [ Widmann and Tett 2003]. A primary way to validate climate change mechanisms proposed from modeling-based research is to compare model results with paleoclimate data. To do that, we need paleoclimate data from key regions that will help identify the large-scale processes. One such key region is the Caribbean Sea because it is located at the intersection of many important climate processes. The tropical Atlantic, including the Caribbean, is influenced by (and possibly influences) at least four major climate phenomena: North Atlantic Oscillation/ Arctic Oscillation (NAO/AO), teleconnections with the Pacific El Nio Southern Oscillation (ENSO), meridional overturning circulation (MOC), and tropical Atlantic sea surface temperature (SST) variability interacting with the Atlantic sector trade winds (Tropical Atlantic Variability, TAV) [ Enfield and Mayer 1997; Enfield and Alfaro 1999; Chang et al. 2001; George and Saunders 2001; Marshall et al. 2001; Wang 2001; Alexander and Scott 2002]. Additionally, the Caribbean Sea comprises most of the Western Hemisphere Warm Pool (WHWP), which is the Atlantic analog to the Western Pacific Warm Pool, and includes the Gulf of Mexico, Caribbean Sea, and a portion of the Eastern Pacific [ Wang and Enfield 2001, 2003] This large region of the ocean where SSTs seasonally exceed 28.5C provides a heat source to the

PAGE 94

79 Intertropical Convergence Zone (ITCZ) and a moisture source to central North America during the boreal summer [ Wang and Enfield 2001, 2003]. Variations in the WHWP occur at high temperatures where small changes can strongly influence organized convection, and long-term changes in surface ocean properties may be propagated to higher latitudes [ Sutton and Allen 1997]. A few marine records from the Caribbean span the last 1000 years. The primary information comes from several sources: sediment cores from the Cariaco Basin offshore of Venezuela [ Black et al. 1999; Haug et al. 2001b; Black et al. 2004; Peterson and Haug 2006]; sclerosponges from offshore of Jamaica and Pedro Bank [ Haase-Schramm et al. 2003; Haase-Schramm et al. 2005]; sediment cores from offshore of Puerto Rico [ Nyberg et al. 2001; Nyberg et al. 2002]; and short coral records also from offshore of Puerto Rico [ Winter et al. 2000; Watanabe et al. 2001]. Research results from the study of marine sediment from the Caribbean identify a drying trend over the last 2000 years [ Black et al. 2004] upon which is superimposed decadal and centennial variability. In general this millennial-scale trend is supported by non-marine records from the region [e.g., Hodell et al. 1991; Black et al. 1999; Higuera-Gundy et al. 1999; Haug et al. 2001b; Haase-Schramm et al. 2003; Black et al. 2004; Lachniet et al. 2004; HaaseSchramm et al. 2005; Hodell et al. 2005b; Peterson and Haug 2006] Shorter records highlight the substantial decadal and centennial variability in the region [ Winter et al. 2000; Watanabe et al. 2001; Haase-Schramm et al. 2005; Kilbourne et al. 2006a]. Surface temperatures during the Maunder solar minimum (approximately 1650-1750 C.E.) appear to have been 2 C cooler than today in the northern Caribbean [ Winter et al. 2000; Watanabe et al. 2001; Haase-Schramm et al. 2003; Haase-Schramm et al. 2005;

PAGE 95

80 Kilbourne et al. 2006a]. According to the Jamaican sclerosponge record [ HaaseSchramm et al. 2003], which is the longest continuous high-resolution marine record in the northern Caribbean, the cool Maunder Minimum is the culmination of a cooling trend that started by ~1350 C.E. (when the sclerosponge record begins) and the cool period is followed by a rapid warming toward todays temperatures. Some doubt still remains about the conditions in the Northern Caribbean before the 1700s because the Jamaican sclerosponge was collected from 20m water depth, near the modern day pycnocline and it has been hypothesized that subsurface advective processes may have influenced the record [ Haase-Schramm et al. 2003] Paleoclimate researchers working in the circum-Caribbean region have been very cognizant of the effects that seasonality can play in mean state changes when investigating mechanisms for past climate variability. For example, a proxy indicating a relatively cool mean temperature does not necessarily mean colder winters and cooler summers, but could mean that summer temperature was the same, but winter was cooler. One idea that has taken root is that dry periods are caused by a change in the seasonal migration of the ITCZ, whereby the mean position of the ITCZ is shifted southward [ Hodell et al. 1991; Haug et al. 2001b; Wang et al. 2004; Hodell et al. 2005b; Peterson and Haug 2006] Few paleoclimate archives truly have subannual resolution in order to test this idea, although some have tried to use seasonally biased proxies to explore seasonality in the Caribbean [ Nyberg et al. 2001; Nyberg et al. 2002; Black et al. 2004]. Black et al. [2004] concluded that seasonal changes in upwelling intensity, driven by the northeast trade winds and changes in regional precipitation patterns led to the observed

PAGE 96

81 geochemical differences between summer and winter foraminifera species. Over the last 1000 years, the difference between the seasonally biased species was at a maximum from~1300 C.E. to ~1500 C.E. Similarly, Nyberg et al. [2002] found an increase in seasonal temperature range from 1300 C.E. to 1600 C.E. offshore of Puerto Rico using an artificial neural network trained to identify winter and summer temperatures from an assemblage of 26 foraminiferal species. Do these findings hold up when tested against a truly seasonally resolved climate proxy? Massive scleractinian corals provide an opportunity to find out. They tend to grow at rates that commonly permit sampling > 12 times per year, and the geochemistry of their skeletons can provide information on SST and seawater 18 O [e.g., Gagan et al. 1998], which is correlated to SSS [e.g., Fairbanks et al. 1997; Watanabe et al. 2002]. Corals from Puerto Rico have been used to obtain seasonally resolved records of skeletal 18 O for a combined SST and seawater 18 O ( 18 Ow) signal [ Winter et al. 2000] Paired coral Mg/Ca and 18 O have been used at the same location for separate SST and 18 Ow contributions [ Watanabe et al. 2001]. Winter et al. [2000] showed no significant change in the seasonal cycle of coral 18 O during a few short time periods over the last 300 years (1700-1705 C.E., 1780-1785 C.E., 1810-1815 C.E., and 1984-1989 C.E.). However, coral 18 O is a combined signal of SST and 18 Ow, with coral 18 O inversely related to temperature and positively related to 18 Ow, so that in-phase changes in both could actually mask an environmental change in the coral signal. Paired analysis of an independent SST proxy gets around this dilemma [ Gagan et al. 1998], and the study by Watanabe et al. [2001] shows a small increase in the seasonal 18 Ow range during 16991703 C.E. compared to 1987-1993 C.E., although the significance of the observed change

PAGE 97

82 is neither quantified nor explored. Despite these initial studies, the primary question regarding changes in seasonal variability remains open. Were there significant changes in seasonal patterns of temperature or salinity in the Caribbean during times of difference mean climate states? To address the issues of mean climate state and seasonal variability in the Caribbean, we have generated ~monthly resolved Sr/Ca and 18 O records from massive Puerto Rican corals over 13 time intervals between 1172 C.E. and 2004 C.E. Our goal is to provide data addressing mean SST and 18 Ow of the northern Caribbean over the last ~800 years and to investigate the possibility of changes in the seasonal range of SST and 18 Ow. We chose southwestern Puerto Rico because surface-ocean conditions have been extensively studied in the area, and several environmental monitoring sites exist in this region which provide access to SSS, SST, 18 Ow, and meteorological data [ Winter et al. 1998; Corredor and Morell 2001; Watanabe et al. 2001]. Corals from this area have also been proven to reliably record climate conditions in the surface ocean [ Watanabe et al. 2002]. 4.2 Methods The coral cores used in this study were drilled using a hydraulic drill at different times from the same reef track offshore from La Parguera, Puerto Rico. The first core, called E1P (Figure 29), was drilled in June of 1994 from a live coral at the Pinnacles reef site (17.933N, 67.012W) and has been previously studied [ Watanabe et al. 2001; Watanabe et al. 2002]. The core has an unconformity in the 1880s that was previously assumed to represent only ~10 years. Uranium-series determinations made at the

PAGE 98

83 Figure 29: X-radiograph of core E1P illustrating the unconformity in the late 19 th century and the uranium-series date at the bottom. Density band counts from the dated section indicate that the unconformity represents about 200years. University of Minnesota indicate that the unconformity actually represents about 200 years. A long section of the core is below the unconformity, the bottom has been dated to 1462 C.E., and density-band counting from the bottom places the beginning of the unconformity in the 1670s. This has implications for the previous work done on this core

PAGE 99

84 that will be discussed below. The very bottom of core E1P appears to have broken off during drilling and been abraded so that it is no longer continuous with the rest of the core. This piece has been independently dated, resulting in an age of 1446 C.E. The second core is discontinuous and represents multiple coral heads. Uranium-series dates of the pieces of this core used in this study indicate ages ranging from 1172 32 C.E. to 1429 7 C.E (Figure 30). The third core used in this study, core 04LPTA (Figure 30), was obtained from a live coral head in 2004 at the Turrumote Reef site (17.934N, 67.001W) and represents 254 years of continuous coral growth. Annual Sr/C, 18 O and 14 C from this coral core were generated and used in two recently completed studies [ Kilbourne et al. 2006a; Kilbourne et al. 2006b]. All of the cores were cut into slabs and X-radiographed before further analysis were performed. At the University of South Florida, all of the slabs were sonicated in deionized water and dried in an oven set to 70C. A continuous 0.7 mm-wide path was drilled with a 0.7 mm diameter dental bit in a Dremel tool mounted on a computer controlled drilling stage, taking aliquots of powdered coral material every 0.6 mm for most of the data segments. A 0.5 mm diameter bit was used for coral pieces LP7, LP8.2, LP10.2, LP11 and LP12. Samples LP12 and LP10.2 were cut nearly perpendicular to the growth axis, therefore these slabs were sub-sampled by discrete samples along several adjacent corallite thecal walls and they are not continuously sampled time series like the other samples. The very bottom piece of core E1P (labeled LPD4) was also suboptimally cut, but this piece was

PAGE 100

85 Figure 30: X-radiographs of corals used in this study that have not previously been published. Core E1P used in this study is shown in Watanabe et al. [2001] and Figure 29. Dark region has secondary aragonite fill.

PAGE 101

86 broken in such a way as to permit time transgressive sampling of a thecal wall by hand (not computer controlled) with the Dremel tool along the edge of the slab. An attempt was made to take aliquots every 0.6 mm to maintain consistency. The resulting powders were analyzed for Sr/Ca, Mg/Ca, 18 O, and 13 C at the University of South Florida Paleoclimatology Paleoceanography and Biogeochemistry laboratory facility. Isotopic data were obtained by dissolving 35-100 g of coral powder in 100% phosphoric acid at 70C in a ThermoFinnigan Kiel Carbonate Device. The resultant CO 2 gas was analyzed by an attached ThermoFinnigan Delta Plus XL Mass Spectrometer. The calcite standard NBS 19 was used for correcting the data to permil units relative to VPDB. Long-term precision (1 # ) on the NBS 19 standards run with each set of samples is 0.03 and 0.06 for carbon and oxygen, respectively. Elemental ratios were obtained by dissolving 75-250 g of coral powder in a sufficient volume of trace-metal grade 2% nitric acid to make a ~20 ppm calcium solution. The solutions were analyzed using a Perkin Elmer Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES). The data are corrected with a matrix-matched gravimetric standard run between each sample [ Schrag 1999] The gravimetric standard used at the University of South Florida was cross-calibrated by thermal ionization mass spectrometry at the University of Minnesota. ICP-OES precision for Sr/Ca and Mg/Ca ratios (1 # ) is 0.21% (0.019 mmol/mol) and 0.84% (0.033 mmol/mol) respectively, measured by replicate analyses of a coral solution run with the samples. Seasonal cycles were defined from the data by first converting the depth series to a time series by matching minima and maxima in 18 O with maxima and minima in a climatological SST cycle (average January, average February, etc). The resulting time

PAGE 102

87 series were then re-sampled to obtain 12 evenly spaced samples per year using the software program AnalySeries [ Paillard et al. 1996]. A climatology could then be calculated by averaging each month over the entire series to obtain the average January value, average February value, etc. Coral Sr/Ca did not display obvious seasonal cycles in the raw data and so no climatologies were calculated for Sr/Ca. Many possible reasons were considered for the lack of seasonal cycles in our Sr/Ca data but an adequate explanation remains to be found. Potential secondary alteration of the samples was investigated by petrographic inspection of thin sections, scanning electron microscopy (SEM) and geochemical cross plots as described in Quinn and Taylor [2006] Thin sections were made of selected portions of core 04LPTA, however a lack of available material precluded the destruction of the older coral slabs for thin section work. 4.3 Results Corals used in this study are primarily comprised of pristine coral aragonite; evidence of diagenetic alteration was noted in only a few isolated places. Secondary aragonite and micrite fill was found in the extreme lower section of core 04LPTA (Figure 30) and was easily avoided in analysis. X-radiographs of all of the cores were used to identify denser sections to be avoided during sampling because of the existence of infilling aragonite growth. Geochemical cross plots of all of the data enabled the identification of an altered section in piece LP11 (Figure 31). The time series of LP11 had distinct periods with large swings in 18 O on a monthly basis that were on the order of 0.5 to 1 equivalent to temperature swings equivalent to the entire present-day seasonal cycle. Close inspection of the slab using a standard light microscope revealed

PAGE 103

88 8.9 9 9.1 9.2 9.3 9.4 9.5 9.6 9.7 -5 -4 -3 -2 -1 0 Sr/Ca (mmol/mol) 18 O (VPDB) Pristine coral 04LPTA Older coral Inorganic Aragonite Altered coral -4 -2 0 2 4 -5 -4 -3 -2 -1 0 04LPTA 13 C 18 O (VPDB) Inorganic Aragonite Pristine coral 04LPTA Older coral Altered coral Figure 31: Geochemical cross plots of coral Sr/Ca, 18 O and 13 C indicate that most of the older coral material is not diagenetically altered. Data not used in this study because of secondary aragonite drift towards the values of inorganic aragonite precipitated in equilibrium with the local conditions. Inorganic aragonite values were calculated using the equations of Morse and MacKenzie [1990] for 13 C, Grossman and Ku [1986] for 18 O, and Smith et al. [1979] for Sr/Ca with local temperature data [ Winter et al. 1998] slight discoloration and a patchy sugary coating on those parts of the coral displaying an aberrant geochemical signature. Removal of the data from the altered sections removed the data points in the cross plots that displayed evidence of secondary aragonite (Figure 31). SEM images of pieces of LP11 and LP12 show small amounts of secondary aragonite,mostly on the septa, which we specifically avoid during sampling. We estimate that secondary aragonite contamination could comprise up to 1-2% of our samples in these two core pieces. We are confident that other altered sections would have been caught by our inspection and that the remaining data are not affected by diagenetic alteration.

PAGE 104

89 Sr/Ca and 18 O data generated for this study are shown in Figure 32 and the mean values are compared with annual data from [ Kilbourne et al. 2006a] on a single time axis in Figure 33. The mean values of the monthly data and the annual data compare well for the 2004 C.E. to 1751 C.E. period. Having a continuous record at the beginning of the record enables us to better characterize the variability we may be missing as result of using relatively short geochemical time series over the older parts of the record. The multicentury trends in both 18 O and Sr/Ca are captured by the four data windows despite the significant multidecadal variability in the annual data (discussed in [ Kilbourne et al. 2006a]). We used the annual data to explore how representative a short time series can Geochemistry of Puerto Rican Corals over the past ~800 years 8.9 9.0 9.1 9.2 9.3 9.4 9.5 0 200 400 600 800 1000 1200 1400 1600 1800 Sample Index, positive towards past time Sr/Ca (mmol/mol) -5.1 -4.7 -4.3 -3.9 -3.5 -3.1 -2.7 -2.3 Coral 18 O ( VPDB) 1172 1446+4 1429 1427 1368 1274 1993-2004 1902-1912 1620-1628 1665-1673 1560-1568 1750-1763 1484-1493 1852-1856 04LPTA E1P LP11 LP12 LP8.2 LP10.2 LP7 Figure 32: Coral Sr/Ca and 18 O records (~monthly resolution) from Puerto Rico spanning more than 8 centuries. Different coral heads are represented by black bars at the bottom of the graph that are labeled with the coral sample name. The data are labeled with the period of coral growth represented. The dates for coral 04LPTA are from density band counting from the top of the core, whereas the dates for core E1P are density band counted from a Uranium-series date at the bottom of the core (1462 5 years), and the dates for the shorter segments are the Uranium-series dates on the individual pieces of the core. The oldest segment of core EIP (named LPD4) is independently dated (1446) and not part of the layer counting because it appears to have broken off during drilling and been ground down considerably.

PAGE 105

90 8.900 8.994 9.088 9.182 9.276 9.370 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Year (C.E.) 18 O ( VPDB) -4.80 -4.36 -3.92 -3.48 -3.04 -2.60 Sr/Ca (mmol/mol) 04LPTA EIP Figure 33: Mean coral 18 O (black diamonds) and Sr/Ca (grey squares) values from Puerto Rican corals plotted with annual values from core 04LPTA (dashed lines). The error bars on the mean values depict the standard error of the means. The other error bars in the right and left bottom corners represent the median error expected from inter-colony variability, determined by replicate measurements of Montastraea faveolata from the same reef in several studies (Table 1). Both axes have been scaled to equal 10C using the calibration equations of Leder et al. [1996] and Swart et al. [2002] be of a century mean. To do so, we split our 18 O and Sr/Ca time series into four overlapping centuries and calculated the mean and standard deviation (s) for each century. Then we computed 5-year averages along the entire 254-year record. All of the 5-year averages were within 2 # of the centennial mean in which they lay, with 25% to 8% of the 5-year means equal to or outside of 1 # from the centennial mean. Of those 5year means that were outside of 1 # all had 1 # error bars overlapping with the 1 # error on the centennial mean. The seasonal cycle does not contribute to the variance in these data because they are annual, so a similar exercise with monthly data would have larger error bars that would overlap even further. We are therefore confident that our short, 4

PAGE 106

91 12 year time series with monthly resolution are representative of centennial means within the errors of the calculations. Sr/Ca and 18 O data were converted to temperature and 18 Ow differences from modern values by the calculation method of Kilbourne et al. [2004], using the calibration slopes of Leder et al. [1996] for the coral 18 O-SST relationship (0.22/C), and Swart et al. [2002] for the Sr/Ca-SST relationship (0.047 mmol/mol/C, Figure 34A). These calculations can be converted to salinity values because the local 18 O-SSS relationship has been well documented to be 0.20 /psu [ Watanabe et al. 2001], but we refrain from doing so because of the possibility of non-stationarity in the 18 O-SSS relationship. The results of our analysis indicate that SSTs were ~1-3 cooler than today during most of the last 800 years. Salinity has a more distinct signal with relatively minor changes in salinity over the most recent 250 years, a distinctly fresher period between about 1440 C.E. and 1670 C.E., and a distinctly saltier period before that. The amplitude of the annual cycles in coral 18 O ranged between 0.46 and 0.80, and averaged 0.60 (Figure 35). Excluding the two lowest and highest values, the amplitudes only ranged between 0.50 and 0.64 and were not significantly different from each other. The two time intervals showing significantly different seasonal amplitudes were centered on 1854 and 1907. Mean 18 O during the 1852-1856 and 1902-1912 intervals were -4.11.19 and 4.12.09, respectively. Mean Sr/Ca during the same time intervals were 9.12.04 and 9.11.03, respectively. Those two periods with different amplitudes have the same mean values, whereas other periods with

PAGE 107

92 -6 -4 -2 0 2 4 1100 1300 1500 1700 1900 SST Difference from 2004-1993 (C) -1.20 -0.80 -0.40 0.00 0.40 0.80 18 Ow Diff. from 2004-1993 ( ) SST d18Ow -6 -4 -2 0 2 4 1100 1300 1500 1700 1900 Date (years C.E.) SST Difference from 2004-1993 (C) -1.20 -0.80 -0.40 0.00 0.40 0.80 18 Ow Diff. from 2004-1993 () Figure 34: Surface temperature and salinity changes indicated by the coral geochemistry as calculated using the method of Kilbourne et al. [2004] Panel A shows the results assuming that inter-coral variability is not an issue. SST hovers between 1 C and 3 C cooler than the modern comparison period of 2004-1993 and the regional salinity variations show a distinct freshening between about 1440 and 1670 C.E. with a distinctly saltier period during the preceding 270 years. Panel B shows the results with the assumption that the coral Sr/Ca values in core E1P are offset by .08 mmol/mol from the other cores. Replication of the data between 1446 and 1670 with a different coral is needed to determine which scenario is correct. The error bars in both A and B are 0.5C and represent approximately one standard deviation of the SST and SSS estimates. This value was chosen by propagating the uncertainty in the mean values (excluding intercoral variability) and the calibration equations through the SSS and SST calculations as described in Schmidt [1999] and used in Kilbourne et al. [2004]. The actual calculated errors are not exactly 0.50, but they hover around .5C and psu (1 # ), depending primarily on the number of years used to calculate the mean values. A B

PAGE 108

93 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Coral 18 O ( VPDB) 1489 1564 1624 1669 1757 1854 1907 1999 Figure 35: The seasonal cycles of 18 O during 8 time periods from two continuous coral cores with the means removed to facilitate comparison. Only two time intervals show seasonal amplitudes that are significantly different from the rest; one centered on 1854 and the other centered on 1907. Mean coral 18 O and Sr/Ca during those two periods with different amplitudes are essentially the same, whereas mean conditions during other periods are significantly different although the seasonal coral 18 O ranges are the same. This implies that on centennial scales at least, seasonal coral 18 O range is independent of mean conditions in the northern Caribbean. significantly different means have the same seasonal 18 O ranges. The earliest dated samples were not included in this analysis for different reasons. Samples LP12 and LP10.2 did not have depth series because the slabs were cut essentially perpendicular to the growth axis, so no seasonal cycles were calculated. Samples LPD4, LP7, LP 8.2, and LP11 did not have clear seasonal cycles and were thus not used in the analysis.

PAGE 109

94 4.4 Discussion 4.4.1 Possible Non-Climatic Signals Before considering the climatic implications for our results, we must entertain the possibility that our results describe a non-climatic signal. Growth-dependent geochemical variations have been studied with contradictory results, and no consensus has been reached. Some studies indicate substantial growth-rate effects in corals [e.g., Goreau 1977; de Villiers et al. 1995; Felis et al. 2003; Goodkin et al. 2005], whereas others indicate none [ Alibert and McCulloch 1997; Gagan et al. 1998; Marshall and McCulloch 2002]. Growth-rate effects are not likely the cause of the observed differences between the two long cores used in this study because the X-radiographs reveal that they have fairly constant and equal extension rates and significant growth-rate effects are primarily observed in corals with substantially different extension rates [e.g., Felis et al. 2003]. Growth-rate effects are possible for the pieces of discontinuous coral core for multiple reasons. The X-radiographs reveal that the extension rates in those pieces are substantially slower than in either of the two continuous cores although quantitative growth rate information for the short cores is lacking because the lack of a seasonal cycle in the geochemical data precluded using the 18 O to create an age model. Slow growth rates have been related to higher Sr/Ca values [ de Villiers et al. 1995], higher 18 O values [ Felis et al. 2003], and a positive relationship between 18 O and 13 C [ McConnaughey 1989a, b]. The geochemistry in every of piece of the oldest, discontinuous core exhibits all three of the above characteristics. This fact combined with the lack of annual cycles in 18 O data, leads to the conclusion that the observed geochemical variations in the oldest core may be non-climatic in origin.

PAGE 110

95 Returning to diagenesis as a potential source of a non-climatic signal, we created a numerical model of mean coral Sr/Ca and 18 O values in the presence of varying amounts of secondary aragonite. Using this model we determined the temperature and salinity values that would be calculated if altered material was incorporated in our analyses. To get 2C cooling, as is recorded by our coral Sr/Ca for much of the last few centuries, when in fact temperatures were the same as today, one would have to have a coral with 20% secondary aragonite. It is unlikely that we did not identify a sample with so much secondary aragonite, and a more plausible incorporation of up to 3% secondary material in a sample would only affect the SST and 18 Ow results by -0.3C and 0.06, respectively. Thus, it is concluded that diagenetic effects are not the primary source of the geochemical signal in the corals used in this study. The most potentially problematic issue is that the entire freshening period during the 15 th to 17 th centuries is manifest in only one coral colony. Additionally, Sr/Ca-SST estimates for this period are the lowest of the entire record, between 3C and 4C less than today, which seems excessively cool for this region and time period. This leads to the question of inter-coral variability. How representative is the mean value of any given coral? Previous work addressing this question has primarily been focused on Indo-Pacific corals, specifically the genus Porites [ Tudhope et al. 1996; Alibert and McCulloch 1997; Gagan et al. 1998; Guilderson and Schrag 1999; Linsley et al. 1999; Hendy et al. 2002; Stephans et al. 2004]. To answer this question for the Atlantic coral species Montastraea faveolata, we have gathered together replicate measurements of 18 O and Sr/Ca from separate colonies growing on the same reefs at the same time, and have calculated the differences between the mean geochemical values.

PAGE 111

96 We did not take into account any other factors that could potentially influence the geochemistry such as reef microenvironment, colony depth, and growth rates, so our results probably over estimate inter-coral geochemical variability, but probably not by much since the replicate cores in each study were chosen because they were similar to each other. The results of our analysis of inter-coral geochemical variability indicate that in 11 out of 12 cases, differences between coral colonies are 0.22 (1.0C, 1.1 psu) for 18 O (Table 3). This is similar to the reproducibility found by Cobb et al. [2003] for the genus Porites The median difference between coral heads is 0.11 (0.5C). Differences in Sr/Ca between coral heads are 0.073 mmol/mol (1.6C) in 10 out of 12 cases and have a median difference of 0.047 mmol/mol (1.0C) (Table 3). The number of replicate cores available (12) is hardly a statistically robust sample population, but it is the best estimate available. Inter-coral variability of 18 O is less than that of Sr/Ca in terms of temperature according to our reproducibility test. The Puerto Rican coral data indicate that the mean 18 O between cores E1P and 04LPTA is not substantially different (Figure 32 and Figure 33). They have a similar ranges and similar variability on the centennial scale. Sr/Ca on the other hand is substantially higher in core E1P than in core 04LPTA (Figure 32 and Figure 33) and the temperatures implied by such high Sr/Ca values are suspiciously low ( Figure 34 A, up to 4C below the 2004-1993 mean). This leads us to think that if there is a mean offset between cores that is not due to climate, it is in the Sr/Ca data. A difference in Sr/Ca values between two cores without a difference in 18 O is not

PAGE 112

97 Absolute 18 O difference Author Number of Years 0.11 Smith et al., 2006 22 0.34 Smith et al., 2006 22 0.22 Smith et al., 2006 22 0.04 Smith et al., 2006 41 0.08 Watanabe et al., 2002 5 0.19 Leder et al., 1996 2 0.11 Leder et al., 1996 2 0.09 Leder et al., 1996 2 0.03 Leder et al., 1996 2 0.22 Leder et al., 1996 2 0.14 Leder et al., 1996 2 0.01 Stair 2006 ~32 Absolute Sr/Ca difference Author Numbe r of Years 0.032 This Study ~5 0.073 This Study ~5 0.020 This Study ~5 0.102 This Study ~5 0.041 This Study ~5 0.012 This Study ~5 0.070 This Study ~5 0.053 This Study ~5 0.029 This Study ~5 0.082 This Study ~5 0.037 Smith et al., 2006 41 0.150 Stair 2006 ~32 Table 3: Absolute value of the differences between the geochemistry of coral heads growing at the same time and in the same area. Sr/Ca units are mmol/mol and 18 O units are VPDB. When the number of years is approximate, the data have not been converted from a depth series to a time series, but the time span of the record was estimated based either from growth rate or from annual geochemical cycles. The average difference between coral heads growing at the same time in the same area is 0.11 and 0.047 mmol/mol for 18 O and Sr/Ca, respectively. unprecedented, as the 41-year records from Smith et al. [2006] and the ~32-year records from Stair [ Stair 2006] attest to in Table 3. If the Sr/Ca values of E1P are shifted until they have the same mean as core 04LPTA (an adjustment of .08 mmol/mol), then the freshening of the 15 th to 17 th centuries disappears when the Sr/Ca and 18 O are used to calculate SST and 18 Ow ( Figure 34 B). Temperatures in this scenario are as warm as today during the 16 th century, but up to 2C cooler for most of the rest of the time. One way to constrain the mean Sr/Ca value of core E1P is to make the temperature found by our study equal that found by Watanabe et al., [2001] from the same core, but based on coral Mg/Ca-SST. Without the aid of Uranium-series dating, Watanabe et al., [2001] counted annual density bands from the top of the coral, across an

PAGE 113

98 unconformity, and sampled density bands assigned to the years 1699-1703 C.E. With the same core and a Uranium-series date at the bottom (1462 5 C.E.), we counted annual density bands up from the bottom and determined that the Watanabe et al. [2001] study actually sampled the years 1499-1503 5 C.E. a difference of ~ 200 years. We analyzed the years 1484-1493 C.E in our study, and while these are not the exact same years as in the Watanabe et al. [2001] study, they are within 6 years of the time interval of these authors Mg/Ca data. Using their Mg/Ca-SST relationship (0.28 mmol/mol C -1 ) and mean Mg/Ca values for 1499-1503 (3.83 mmol/mol) and 1988-1993 (4.47 mmol/mol), we adjusted the E1P Sr/Ca values by .024 mmol/mol to be consistent with the Mg/CaSST estimates. The results of this exercise are between the end-member estimates of Figure 34 and so we are confident that actual conditions were somewhere between our two scenarios (no shift in the Sr/Ca mean between cores, and a 0.08 mmol/mol shift) 4.4.2 Centennial Variability Several robust interpretations are possible based on the mean geochemical values in the different corals used in this study, despite the aforementioned uncertainties. The beginning of our record (1172 to 1429) is marked by cooler and drier conditions in Puerto Rico, as evidenced by geochemical data from 5 different coral pieces. The magnitude of cooling and drying may have been exaggerated by the inclusion of minor amounts of secondary aragonite as described above, but the average difference from modern is -1.9C and +0.27, of which only about .3C and 0.06 could plausibly be the result of the presence of secondary aragonite. Further exaggeration of the mean state

PAGE 114

99 may also have occurred due to kinetic growth-rate effects and so we give a climatic interpretation of these 5 coral pieces with caution. Many climatic phenomena could cause substantial cooling and drying in this area. As is the case with all paleoclimate records, it is possible that local phenomena are the cause of the observed changes. The high correlation of local SST to regional and hemispheric SST over the last ~35 years [ Kilbourne et al. 2006a] indicates that our local signal is representative of larger climate phenomena. Additionally, our study site is in the southwestern corner of Puerto Rico, a region that is least influenced by the orographic rains that comprise most of Puerto Rican precipitation [ Carter et al. 2000] and is therefore more representative of the open Caribbean Sea. Conditions in the Caribbean Sea on these time scales could be controlled by ocean circulation, mean changes in airsea interactions (possibly forced by oceanic or solar variability), and solar variability. Paleoclimate records from the region show contradictory signals with some records indicating drier conditions, whereas others indicate wetter conditions; no records that explicitly represent temperature are presently available. Lachniet et al. [2004] found drier conditions compared to today in Panama starting ~900 C.E. and continuing to the end of their record in the 1300s. Hodell et al. [2005b] summarized many years of research in Central America and found relatively moist conditions on the northern Yucatan peninsula before 1400 C. Titanium concentrations in marine sediments from the Cariaco Basin also indicate wetter conditions before 1400 C.E., certainly more moist conditions than the 1600 to 1800 C.E. interval [ Haug et al. 2001b]. It is interesting that we observe relatively cool SSTs during the time known as the Medieval Warm Period,

PAGE 115

100 especially since regional temperatures in the Caribbean track global temperatures so well during the instrumental period (back to 1870, [ Kilbourne et al. 2006a]). The climate state of the Caribbean Sea is affected by many factors including: the temperature gradient between the Caribbean and the Pacific [ Wang 2006] the state of atmospheric circulation at high northern latitudes [ Giannini et al. 2001a], the state of ENSO and global Walker Circulation [ Enfield and Mayer 1997; Giannini et al. 2001a; Giannini et al. 2001b; Taylor et al. 2002] the cross equatorial temperature gradient [ Nobre and Shukla 1996] and the state of Atlantic Meridional Overturning Circulation [ Vellinga and Wood 2002] Speculation on the causes of cooler and drier conditions in the northern Caribbean at this time without knowing the state of all of the other parameters is highly speculative. Another feature of our record is a substantial freshening between 1426 and 1446 in southwestern Puerto Rico compared with the previous period. This coincides with the onset of the Sporer sunspot minimum (Figure 36) and with a freshening period found in previous sediment core work from offshore of south-central Puerto Rico [ Nyberg et al. 2001]. As in the interpretation of the sediment core record, the initial freshening is followed by an increase in salinity thereafter. Freshening at our site in southwestern Puerto Rico can be caused by multiple mechanisms. A local change in the pattern of onshore breezes and afternoon thunder showers is possible, but not likely because our sampling site is on the outermost reef about 3 km offshore and is thus bathed in open ocean conditions. One mechanism is simply decreasing the regional evaporation to precipitation ratios, thereby freshening the surface water. Another proposed mechanism is to increase trade wind strength, thereby transporting more fresh water runoff from the

PAGE 116

101 southern Caribbean into the northern Caribbean [ Kilbourne et al. 2006a; Kilbourne et al. 2006b]. 1356 1360 1364 1368 1372 1000 1200 1400 1600 1800 2000 Solar Irradiance (W/m 2 ) 10000 20000 30000 40000 Quelccaya Core 1 Particles >1.59 mm -4 -2 0 2 Puerto Rico Coral SSTA (C) and SSSA (psu) 0.05 0.10 0.15 0.20 0.25 Cariaco Basin % Titanium 0 1000 2000 3000 Cariaco Basin G. bulloides 1.0 1.2 1.4 1.6 1.8 2.0 Quelccaya Core 1 Accumulation Rate (m) 9.2 9.4 9.6 9.8 10 Jamaica Sr/Ca (mmol/mol) Wolf Sporer Maunder Dalton -1.0 -0.9 -0.8 -0.7 -0.6 1000 1200 1400 1600 1800 2000 Jamaica 18 O ( VPDB) Figure 36: High resolution paleoclimate records from the Caribbean and South America showing transition periods at the beginning of the Sporer solar minimum and during the Maunder minimum. The solar irradiance record is based on the covariance of 10 Be and 14 C records [ Bard et al. 2000] and is scaled to the estimate of Maunder minimum irradiance reduction of Cliver et al. [1998]. Puerto Rico coral estimates of SST anomaly (SSTA, open circles) and SSS anomaly (SSSA, black squares) from 1993-2004. The G. bulloides abundance record from the Cariaco Basin indicates trade wind strength, but may also be controlled by nutrient dynamics in the basin [ Black et al. 1999]. Titanium concentrations in Cariaco Basin sediments are controlled by runoff of ITCZ-related rainfall [ Haug et al. 2001b]. Particles and accumulation rates in Quelccaya ice core 1 [ Thompson et al. 1984a; Thompson et al. 1984b; Thompson et al. 1985; Thompson et al. 1986; Thompson and Mosley-Thompson 1987] have been smoothed with a 10-year window and generally increase between 1400 and 1650 indicating increased atmospheric circulation [ Thompson and Mosley-Thompson 1989] particularly increased easterly trade winds which bring moisture for precipitation on the Peruvian Altiplano [ Garreaud et al.

PAGE 117

102 2003]. The Jamaican Sr/Ca and 18 O records are from a sclerosponge collected from 20 m water depth, on the edge of the pycnocline between Caribbean Surface Water and Subtropical Under Water [ Haase-Schramm et al. 2003]. The latter mechanism is favored to explain this freshening event because evidence from surrounding regions supports an increase in the trade wind strength at this time (Figure 36). Lake records from the Yucatan Peninsula indicate a substantially drier period after about 1400 C. E. [ Hodell et al. 2005b]. The Cariaco G. bulloides abundance suddenly jumps at 1400 C. E. indicating an increase in trade wind-induced upwelling [ Black et al. 2004]. The Quelccaya ice core in Peru [ Thompson et al. 1986] has an increase in particles, especially the large particles, at the same time and begins to have an increase in the accumulation rates, indicating that there was more vigorous circulation (more moisture and more dust transported to the plateau). The Jamaican sclerosponge record [ Haase-Schramm et al. 2003] shows an increase in the 18 O of seawater 20m deep without a concomitant increase in Sr/Ca, indicating that the subsurface waters were more enriched in 18 O. The sclerosponge grew at a depth just at the pycnocline between Caribbean Surface Water and Subtropical Under Water (SUW) [ Haase-Schramm et al. 2003]. A decrease in the values there may be related to an increase in the salinity of the SUW and hence the subtropical gyre. This inference is supported by sediment core data from the Bermuda rise described below [ Keigwin 1996] Interestingly, the Cariaco Ti record [ Haug et al. 2001b] does not show a substantial change around 1400, but instead exhibits a drying trend starting in the 1300s and culminating in the 1600s. Strengthening of the trade winds in the tropics occurs at essentially the same time as higher latitude changes associated with the Little Ice Age. Sediment cores from the Bermuda Rise indicate an increase in atmospheric storminess by an increase in

PAGE 118

103 terrigenous deposition at that time. The isotopic composition of surface-dwelling foraminifera displays a coeval increase in 18 O on the order of 0.4, suggesting cooler and saltier conditions. The GISP2 ice core shows an increase in Cl and Na + between about 1400 and 1450 C.E. [ O'Brien et al. 1995; Kreutz et al. 1997; Nyberg et al. 2001]. Ice cores from both Greenland and Antarctica show increases in Ca + ion concentrations [ Kreutz et al. 1997]. These changes have been linked to an intensification of the polar vortex and a subsequent increased storminess at high north latitudes [ Kreutz et al. 1997]. These high latitude changes may be directly related to solar forcing of polar atmospheric circulation, thus providing a mechanism for the observed atmospheric reorganization at this time [ Mayewski et al. 2004; Maasch et al. 2005; Mayewski et al. 2005]. Another relative freshening event occurred at our site between 1628 and 1665 C.E. This change was associated with a relative cooling event and the onset of the Maunder solar minimum. The changes taking place in other records at this time include a peak in G. bulloides abundance in the Cariaco Basin (indicating stronger trade winds), a decrease in Ti in the Cariaco sediments (indicating less ITCZ rainfall over northern South America), and a decrease in large particles in the Quelccaya ice core (indicating less dust transport as accumulation rates remain steady). The Caribbean data combine to indicate stronger northern hemisphere trade winds that cooled the Caribbean, transported fresher water to southern Puerto Rico and decreased the amount of rainfall over northern South America. Given the potential errors in the age models of all of these records, these events may or may not have been coincident. What is clear and robust from all of these data is that the Maunder Minimum marked a period of transition. In most cases, the transition is

PAGE 119

104 towards conditions as they were before 1400 C.E., but subsequently most of the records also show a trend toward modern conditions. 4.4.3 Seasonality The seasonal range of coral 18 O does not significantly change with mean state in our cores (Figure 35). Most of the seasonal cycles calculated from our coral 18 O were not significantly different from each other and the two periods with different amplitudes (1852-1856 and 1902-1912) have the same mean values. However, coral 18 O contains a combined signal of temperature and salinity, and it is possible that independent changes in each variable could cancel each other out. Further work is need with a temperature proxy that can reliably resolve annual cycles in order to separate seawater 18 O changes from temperature changes. To that end, we continue to experiment with different coral micro-drilling strategies in an attempt to eradicate the noise from the Sr/Ca signal. Future researchers investigating this subject should keep in mind that potential changes to the seasonal range may be small, therefore many more seasons may need to be averaged in order to minimize the error bars and detect changes in seasonal range with statistical certainty. 4.5 Conclusions We have produced coral 18 O and Sr/Ca records during 13 time intervals from massive corals that grew in southwestern Puerto Rico with the purpose of reconstructing mean conditions and seasonal cycles over the past 800 years. A compilation of 18 O and Sr/Ca records from Montastraea faviolata which grew at the same time on the same reef, was created in order to assess the geochemical variability inherent in a population of

PAGE 120

105 coral heads exposed to the same conditions. The compilation showed that the average difference between two cores was 0.11 in 18 O and 0.047 mmol/mol in Sr/Ca. Climatic conditions during the early part of our record, between 1172 C.E. and 1429 C.E., were about 2C cooler and 18 Ow was 0.4 higher (~2 psu drier) than present day, although minor diagenesis and growth-rate effects may give these corals a non-climatic signal. Geochemical data from the middle portion of our record, between 1446 and 1670 indicates that SST could have been anywhere from 4C cooler than today to the same as today, depending on how representative our coral specimens Sr/Ca is compared to others growing at the same time under the same conditions. The same uncertainty places 18 Ow changes between about +0.2 to .4, corresponding with +1 psu to psu if the local 18 O-SSS relationship is time invariant. The most recent segment of our record, from a single core representing growth between 1751 and 2004, indicates that conditions at the beginning of the record were about 2C cooler than today, as has been indicated in previous work [ Kilbourne et al. 2006a]. The seasonal range in 18 O from our corals shows no significant variability with mean state. Future studies must use a precise independent SST proxy to separate SST from 18 Ow changes and use longer records to detect potentially small changes in the seasonal cycle.

PAGE 121

106 5 CONCLUSIONS Five questions were posed at the beginning of this research and it is useful to summarize how well they were answered. The questions were: 1. Have there been detectible changes in the proportion of North Atlantic subtropical gyre water and South Atlantic equatorial water entering into the Caribbean Sea during any time in the last 250 years? If so, are the observed changes related to other climatic variables? 2. Was the northern Caribbean Sea indeed 2C cooler than today during the Little Ice Age as has been proposed by a earlier studies? Was there a substantial salinity anomaly during the same period? 3. Is there substantial decadal variability in northern Caribbean Sea surface temperature or salinity? If so, can it be related to other Atlantic climate processes? 4. How stable were the conditions in the Caribbean through the global climatic changes of the last 1000 years? 5. Did changes in seasonality play a role in any observed mean state changes? The answer to the first question is yes, a change in the proportion of North Atlantic subtropical gyre water and South Atlantic Equatorial water in the Caribbean Sea was detected south of Puerto Rico. Around 1970 the coral 14 C exhibits a shift from higher values that are similar to those of a coral growing in Bermuda, to lower values that are more similar to those of a coral growing in Brazil. Trade wind strength increases at the same time, so it is proposed that ekman transport is responsible for bringing more South

PAGE 122

107 Atlantic water towards Puerto Rico. Earlier changes in coral 14 C from Puerto Rico may represent similar shifts, because they are too sudden and large to be explained by equilibration with the atmosphere. Further work is needed to show if increasing the amount of water with South Atlantic origins in the surface leads to increased transport of such water into the Caribbean Sea. If the observed water mass changes are related actual volume transport, and thus meridional overturning circulation and heat balance between the hemispheres, one would predict a change in the Puerto Rico coral 14 C during the 1990s as well. New coral 14 C records from sites bathed by northern subtropical gyre water and South Equatorial Current water would extend the time period over which a mixing model can be made, in order to test the theory. Detection of earlier shifts, before the bomb-carbon spike, requires higher precision measurements because of the reduced gradient between the two end members, but is not outside analytical possibility and could be valuable in determining past changes in thermohaline circulation. Question 2 asked if the Caribbean was indeed 2C cooler during the 1700s and if there had been any detectable 18 Ow changes since then. Annual coral 18 O and Sr/Ca data spanning 1751-2004 addressed this issue. The data indicated that conditions were about 2C cooler in the 1750s compared with the mid-1990s to 2000s. The temperature change occurred as a steady trend over the whole record. No significant centennial-scale trend in 18 Ow (i.e., salinity) was found. Significant decadal-scale variability was found in the geochemical data, answering the third question. The annual coral 18 O record demonstrated variability with a 60-year period that consistently tracks a trade-wind climate proxy. The timing of the 14 C shift,

PAGE 123

108 interpreted as a change in the origin of the local water mass, is coincident with one of the observed shifts in 18 O and the two are thought to have similar origins. The surface water south of Puerto Rico, in the Caribbean and in the equatorial Atlantic tends to be lower in salinity and therefore have a lower 18 O value. Increasing the trade wind strength should increase the northward Ekman transport in the tropics and thereby push more low salinity, low 18 O, and low 14 C water northward to be carried into the Caribbean by the North Equatorial Current. The amplitude of the multidecadal shifts in 18 O is about .1, translating to a total amplitude of 0.2 or 1psu if the 18 O-SSS slope did not change. The fourth and fifth questions concerned the possibility of changing climatic conditions over the last 1000 years and the possibility that any observed mean state change might be related to changes in seasonal patterns. Of the 5 questions posed, these two were the least satisfyingly answered. Coral 18 O and Sr/Ca were generated for this section from two long, continuous cores spanning the years 2004-1751 and 1670 1446, and from one discontinuous core with five pieces dated to between 1429 to 1171. Thus, the data span the last 800 years instead of the last 1000 years. Possible diagenetic and kinetic effects on the geochemistry and the demonstrated uncertainty in the mean Sr/Ca value of any particular coral necessitates using caution when interpreting the data from the discontinuous core, i.e., the data before 1446 C.E. Changes within a single coral core are considered robust and therefore one can have greater confidence in the results from the cores spanning 2004 to 1751 and 1670 to 1446. The climatic interpretations are as follows. The oldest five coral pieces had geochemical values indicating consistently ~2C cooler and drier conditions in the late

PAGE 124

109 1100s through the early 1400s compared with recent times (2004-1993 C.E.), although growth-rate effects and minor amounts of secondary aragonite might have exaggerated the observed climatic changes in these cores. Geochemical data from continuous core EIP indicate a 2C temperature range and a 0.2 18 Ow range during the time of coral growth, between the mid 1400s and the late 1600s. These ranges are the same as found for the modern coral spanning 2004-1751 C.E. Uncertainty in the Sr/Ca values between corals leads to large climatic uncertainty when the data are compared to the modern coral core, resulting in SST values anywhere between 4C cooler to the same as today, and 18 Ow values between +0.2 to .4 different from today. Seasonal variability of coral 18 O did not change significantly despite substantial changes in mean climatic state and coral Sr/Ca did not have strong enough annual cycles to interpret seasonality. No change in coral 18 O could be interpreted as no change in seasonality of 18 Ow and SST, or as a change in both 18 Ow and SST such that their signals cancel in the coral 18 O. Corals that overlap in time and that are sampled for longer periods at monthly resolution could eventually be used to reconstruct past mean climate conditions and potential changes in seasonality with a great deal of confidence.

PAGE 125

110 6 REFERENCES Alexander, M., and J. Scott, The influence of ENSO on air-sea interaction in the Atlantic, Geophysical Research Letters 29, 1701, 2002. Alibert, C., and M. T. McCulloch, Strontium/calcium ratios in modern Porites corals from the Great Barrier Reef as proxy for sea surface temperature: Calibration of the thermometer and monitoring of ENSO, Paleoceanography 12, 345-363, 1997. Bard, E., G. Raisbeck, F. Yiou, and J. Jouzel, Solar irradiance during the last 1200 years based on cosmogenic nuclides, Tellus 52B, 985-992, 2000. Bauer, E., M. Claussen, V. Brovkin, and A. Huenerbein, Assessing climate forcings of the Earth system for the past millennium, Geophysical Research Letters 30, 1276, 2003. Bertrand, C., M.-F. Loutre, M. Crucifix, and A. Berger, Climate of the last millennium: a sensitivity study, Tellus 54A, 221-244, 2002. Black, D. E., L. C. Peterson, J. T. Overpeck, A. Kaplan, M. N. Evans, and M. Kashgarian, Eight centuries of North Atlantic ocean atmosphere variability, Science 286, 1709-1713, 1999. Black, D. E., R. C. Thunell, A. Kaplan, L. C. Peterson, and E. J. Tappa, A 2000-year record of Caribbean and tropical North Atlantic hydrographic variability, Paleoceanography 19, PA2022, 2004.

PAGE 126

111 Bradley, R. S., Are there optimum sites for global paleotemperature reconstruction? in Climatic variations and forcing mechanisms of the last 2000 years edited by Jones, P. D., R. S. Bradley and J. Jouzel, pp. 603-624, Springer-Verlag, Berlin, 1996. Bradley, R. S., K. R. Briffa, T. J. Crowley, M. K. Hughes, P. D. Jones, and M. E. Mann, The scope of Medieval Warming, Science 292, 2011-2012, 2001. Broecker, W. S., Paleocean circulation during the last deglaciation: A bipolar seesaw?, Paleoceanography 13, 119-121, 1998. Broecker, W. S., Was the Medieval Warm Period global?, Science 291, 1497-1499, 2001. Broecker, W. S., and E. A. Olson, Lamont radiocarbon measurements VIII, Radiocarbon 3, 176-204, 1961. Cai, M., Dynamical amplification of polar warming, Geophysical Research Letters 32, L22710, 2005. Cai, M., Dynamical greenhouse-plus feedback and polar warming amplification Part 1: A dry radiative-transportive climate model, Climate Dynamics 26, 661-675, 2006. Carter, M. M., J. B. Elsner, and S. P. Bennett, A quantitative precipitation forecast experiment for Puerto Rico, Journal of Hydrology 239, 162-178, 2000. Centurioni, L. R., and P. P. Niiler, On the surface currents of the Caribbean Sea, Geophysical Research Letters 30, 2003. Chang, P., L. Ji, and R. Saravanan, A hybrid coupled model study of tropical Atlantic variability, Journal of Climate 14, 361-390, 2001.

PAGE 127

112 Cliver, E. W., V. Boriakoff, and J. Feynman, Solar variability and climate change: geomagnetic aa index and global surface temperature, Geophysical Research Letters 25, 1035-1038, 1998. Cobb, K., M., C. D. Charles, H. Cheng, and R. L. Edwards, El Nino/Southern Oscillation and tropical Pacific climate during the last millennium, Nature 424, 271-276, 2003. Conkright, M. E., R. A. Locarnini, H. E. Garcia, T. D. O'Brien, T. P. Boyer, C. Stephans, and J. I. Antonov, World Ocean Atlas 2001: Objective Analysis, Data Statistics, and Figures, CD-ROM Documentation, 17 pp., National Oceanographic Data Center, Silver Spring, MD, 2002. Cook, E. R., R. D. D'Arrigo, and K. R. Briffa, A reconstruction of the North Atlantic oscillation using tree-ring chronologies from North America and Europe, The Holocene 8, 9-17, 1998. Corredor, J. E., and J. M. Morell, Seasonal variation of physical and biogeochemical features in eastern Caribbean Surface Water, Journal of Geophysical Research 106, 4517-4525, 2001. Corredor, J. E., J. M. Morell, J. Lopez, R. Armstrong, A. Dieppa, C. Cabanillas, A. Cabrera, and V. Hensley, Remote continental forcing of phytoplankton biogeochemistry: Observations across the "Caribbean-Atlantic front", Geophysical Research Letters 30, 2057, 2003. Corrge, T., Sea surface temperature and salinity reconstruction from coral geochemical tracers, Palaeogeography Palaeoclimatology Palaeoecology 232, 408-428, 2006.

PAGE 128

113 Crowley, T. J., Causes of climate change over the past 1000 years, Science 289, 270-277, 2000a. Crowley, T. J., Causes of climate change over the past 1000 years, Science 270, 270-277, 2000b. Crutzen, P. I., Geology of mankind, Nature 415, 23, 2002. D'Arrigo, R., R. Wilson, and G. Jacoby, On the long-term context for late twentieth century warming, Journal of Geophysical Research-Atmospheres 111, 2006. Davis, J. C., I. D. Proctor, J. R. Southon, M. W. Caffee, D. W. Heikkinen, M. L. Roberts, T. L. Moore, K. W. Turteltaub, D. E. Nelson, D. H. Loyd, and J. S. Vogel, LLNL/UC AMS facility and research program, Nuclear Instruments and Methods in Physics Research B 52B, 269-272, 1990. de Villiers, S., B. K. Nelson, and A. R. Chivas, Biological controls on coral Sr/ca and d18O reconstructions of sea surface temperature, Science 269, 1247-1249, 1995. Delworth, T., S. Manabe, and R. J. Stouffer, Interdecadal Variations of the Thermohaline Circulation in a Coupled Ocean-Atmosphere Model, Journal of Climate 6, 19932011, 1993. Delworth, T. L., and R. J. Greatbatch, Multidecadal thermohaline circulation variability driven by atmospheric surface flux forcing, Journal of Climate 13, 1481-1495, 2000. Delworth, T. L., and M. E. Mann, Observed and simulated multidecadal variability in the Northern Hemisphere, Climate Dynamics 16, 661-676, 2000.

PAGE 129

114 Dettinger, M. D., M. Ghil, C. M. Strong, W. Weibel, and P. Yiou, Software expedites singular-spectrum analysis of noisy time series, EOS, Transactions, American Geophysical Union 76, 12,14,21, 1995. Diaz, H. F., M. P. Hoerling, and J. K. Eischeid, ENSO variability, teleconnections and climate change, International Journal of Climatology 21, 1845-1862, 2001. Didden, N., and F. Schott, Eddies in the North Brazil Current retroflection region observed by Geosat altimetry, Journal of Geophysical Research 98, 2012120131, 1993. Dijkstra, H. A., L. T. Raa, M. Schmeits, and J. Gerrits, On the physics of the atlantic multidecadal oscillation, Ocean Dynamics 56, 36-50, 2006. Dong, B., R. T. Sutton, and A. A. Scaife, Multidecadal modulation of El Nino-Southern Oscillation (ENSO) variance by Atlantic Ocean sea surface temperatures, Geophysical Research Letters 33, L08705, 2006. Druffel, E. R. M., Radiocarbon in annual coral rings of Belize and Florida, Radiocarbon 22, 363-371, 1980. Druffel, E. R. M., Banded corals; changes in oceanic carbon-14 during the Little Ice Age, Science 218, 13-19, 1982. Druffel, E. R. M., Decadal time scale variability of ventilation in the North Atlantic: high-precision measurements of bomb radiocarbon in banded corals, Journal of Geophysical Research 94, 3271-3285, 1989. Druffel, E. R. M., Post-bomb radiocarbon records of surface corals from the tropical Atlantic Ocean, Radiocarbon 38, 563-572, 1996.

PAGE 130

115 Druffel, E. R. M., Geochemistry of corals: proxies of past ocean chemistry, ocean circulation and climate, Proceedings of the National Academy of Sciences of the U.S.A. 94, 8354-8361, 1997a. Druffel, E. R. M., Pulses of rapid ventilation in the North Atlantic surface ocean during the past century, Science 275, 1454-1457, 1997b. Druffel, E. R. M., Radiocarbon in Corals: Records of the Carbon Cycle, Surface Circulation and Climate, Oceanography 15, 122-127, 2002. Druffel, E. R. M., and S. Griffin, Large variations of surface ocean radiocarbon: evidence of circulation changes in the southwestern Pacific, Journal of Geophysical Research 98, 20,249-220,259, 1993. Druffel, E. R. M., and T. W. Linick, Radiocarbon in annual coral rings of Florida, Geophysical Research Letters 5, 913-916, 1978. Dunbar, R. B., G. M. Wellington, M. W. Colgan, and P. W. Glynn, Eastern Pacific sea surface temperature since 1600 A.D.: The 18 O record of climate variability in Galapagos corals, Paleoceanography 9, 291-315, 1994. Emanuel, K. A., Increasing destructiveness of tropical cyclones over the past 30 years, Nature 436, 602-607, 2005. Enfield, D. B., and E. J. Alfaro, The dependence of Caribbean rainfall on the interaction of the tropical Atlantic and Pacific Oceans, Journal of Climate 12, 2093-2103, 1999. Enfield, D. B., and D. A. Mayer, Tropical Atlantic sea surface temperature variability and its relation to El Nio-Southern Oscillation, Journal of Geophysical Research 102, 929-945, 1997.

PAGE 131

116 Enfield, D. B., A. M. Mestas-Nuez, and P. J. Trimble, The Atlantic multidecadal oscillation and its relation to rainfall and river flows in the continental U.S., Geophysical Research Letters 28, 2077-2080, 2001. Evans, M. N., A. Kaplan, and M. A. Cane, Pacific sea surface temperature field reconstruction from coral d18O data using reduced space objective analysis, Paleoceanography 17, 1007, 2002. Fairbanks, R. G., C. D. Charles, and J. D. Wright, Origin of global meltwater pulses. in Radiocarbon after Four Decades edited by Taylor, R. E., pp. 473-500, SpringerVerlag, Berlin, 1992. Fairbanks, R. G., M. N. Evans, J. L. Rubenstone, R. A. Mortlock, K. Broad, M. D. Moore, and C. D. Charles, Evaluating climate indices and their geochemical proxies measured in corals, Coral Reefs 16, S93-S100, 1997. Felis, T., J. Ptzold, and Y. Loya, Mean oxygen-isotope signatures in Porites spp. corals: inter-colony variability and correction for extension-rate effects, Coral Reefs 22, 328-336, 2003. Fratantoni, D. M., North Atlantic surface circulation during the 1990's observed with satellite-tracked drifters, Journal of Geophysical Research 106, 22,067-022,093, 2001. Froelich, P. N., Jr., D. K. Atwood, and G. S. Giese, Influence of Amazon River discharge on surface salinity and dissolved silicate concentration in the Caribbean Sea, Deep Sea Research 25, 1978.

PAGE 132

117 Gagan, M. K., L. K. Ayliffe, J. W. Beck, J. E. Cole, E. R. M. Druffel, R. B. Dunbar, and D. P. Schrag, New views of tropical paleoclimates from corals, Quaternary Science Reviews 19, 45-64, 2000. Gagan, M. K., L. K. Ayliffe, D. Hopley, J. A. Cali, G. E. Mortimer, J. Chappell, M. T. McCulloch, and J. M. Head, Temperature and surface ocean water balance of the Mid-Holocene tropical western Pacific, Science 279, 1014-1018, 1998. Gallup, C. D., D. M. Olson, R. L. Edwards, L. M. Gruhn, A. Winter, and F. W. Taylor, Sr/Ca-Sea surface temperature calibration in the branching Caribbean coral Acropora palmata, Geophysical Research Letters 33, 2006. Ganopolski, A., and S. Rahmstorf, Rapid changes of glacial climate simulated in a coupled climate model, Nature 409, 153-158, 2001. Garreaud, R., M. Vuille, and A. C. Clement, The climate of the Altiplano: observed current conditions and mechanism of past changes, Palaeogeography Palaeoclimatology Palaeoecology 194, 5-22, 2003. George, S. E., and M. A. Saunders, North Atlantic Oscillation impact on tropical north Atlantic winter atmospheric variability, Geophysical Research Letters 28, 10151018, 2001. Gerber, S., F. Joos, P. Brgger, T. F. Stocker, M. E. Mann, S. Stich, and M. Scholze, Constraining temperature variations over the last millennium by comparing simulated and observed atmospheric CO 2 Climate Dynamics 20, 281-299, 2003. Ghil, M., M. R. Allan, M. D. Dettinger, K. Ide, D. Kondrashov, M. E. Mann, A. W. Robertson, A. Saunders, Y. Tian, F. Varadi, and P. Yiou, Advanced spectral methods for climatic time series, Reviews of Geophysics 40, 1-1 to 1-41, 2002.

PAGE 133

118 Giannini, A., M. A. Cane, and Y. Kushnir, Interdecadal changes in the ENSO teleconnection to the Caribbean region and the North Atlantic Oscillation, Journal of Climate 14, 2867-2879, 2001a. Giannini, A., J. C. H. Chiang, M. A. Cane, Y. Kushnir, and R. Seager, The ENSO teleconnection to the tropical Atlantic Ocean: contributions of the remote and local SSTs to rainfall variability in the tropical Americas, Journal of Climate 14, 4530-4544, 2001b. Goldenberg, S. B., C. W. Landsea, A. M. Mestas-Nunez, and W. M. Gray, The recent increase in Atlantic hurricane activity: causes and implications, Science 293, 474479, 2001. Goodkin, N. F., K. A. Hughen, A. L. Cohen, and S. R. Smith, Record of Little Ice Age sea surface temperatures at Bermuda using a growth-dependent calibration of coral Sr/Ca, Paleoceanography 20, 2005. Goosse, H., H. Renssen, A. Timmermann, and R. S. Bradley, Internal and forced climate variability during the last millennium: a model-data comparison using ensemble simulations, Quaternary Science Reviews 24, 1345-1360, 2005. Goreau, T. J., Coral skeletal chemistry: physiological and environmental reculation of stable isotopes and trace metals in Monastrea annularis Proceedings of the Royal Society of London B 196, 291-315, 1977. Gray, S. T., L. J. Graumlich, J. L. Betancourt, and G. T. Pederson, A tree-ring based reconstruction of the Atlantic Multidecadal Oscillation since 1567 AD, Geophysical Research Letters 31, 2004.

PAGE 134

119 Grossman, E. L., and T. L. Ku, Oxygen and carbon isotope fractionation in biogenic aragonite: temperature effects, Chemical Geology 59, 59-74, 1986. Guilderson, T. P., J. E. Cole, and J. R. Southon, Pre-bomb 14 C variability and the Suess effect in Cariaco Basin surface waters as recorded in hermatypic corals, Radiocarbon 47, 57-65, 2005. Guilderson, T. P., and D. P. Schrag, Reliability of coral isotope records from the western Pacific warm pool: A comparison using age-optimized records, Paleoceanography 14, 1999. Guilderson, T. P., D. P. Schrag, and M. A. Cane, Surface water mixing in the Solomon Sea as documented by a high-resolution coral 14 C record, Journal of Climate 17, 1147-1156, 2004. Guilderson, T. P., D. P. Schrag, M. Kashgarian, and J. Southon, Radiocarbon variability in the western equatorial Pacific inferred from a high-resolution coral record from Nauru Island, Journal of Geophysical Research 103, 24,641-624,650, 1998. Haase-Schramm, A., F. Bhm, A. Eisenhauer, W.-C. Dullo, M. M. Joachimski, B. Hansen, and J. Reitner, Sr/Ca ratios and oxygen isotopes from sclerosponges: Temperature history of the Caribbean mixed later and thermocline during the Little Ice Age, Paleoceanography 18, 1073, 2003. Haase-Schramm, A., F. Bhm, A. Eisenhauer, D. Garbe-Schnberg, W.-C. Dullot, and J. Reitner, Annual to interannual temperature variability in the Caribbean during the Maunder sunspot minimum, Paleoceanography 20, PA4015, 2005. Halliwell, G. R., R. H. Weisberg, and D. A. Mayer, A synthetic float analysis of upperlimb meridional overturning circulation interior ocean pathways in the

PAGE 135

120 tropical/subtropical Atlantic. in Interhemispheric Water Exchange in the Atlantic Ocean edited by Goni, G. J. and P. Malanotte-Rizzoli, pp. 93-136, Elsevier, Amsterdam, 2003. Haug, G. H., K. A. Hughen, D. M. Sigman, L. C. Peterson, and U. Rhl, Southward migration of the Intertropical Convergence Zone though the Holocene, Science 293, 1304-1308, 2001a. Haug, G. H., K. A. Hughen, D. M. Sigman, L. C. Peterson, and U. Rhl, Southward migration of the Intertropical Convergence Zone through the Holocene, Science 293, 1304-1308, 2001b. Hendy, E. J., M. K. Gagan, C. A. Alibert, M. T. McCulloch, J. M. Lough, and P. J. Isdale, Abrupt decrease in tropical Pacific sea surface salinity at the end of Little Ice Age, Science 295, 1511-1514, 2002. Higuera-Gundy, A., M. Brenner, D. A. Hodell, J. H. Curtis, B. W. Leyden, and M. W. Binford, A 10,300 C-14 yr record of climate and vegetation change from Haiti, Quaternary Research 52, 159-170, 1999. Hodell, D. A., M. Brenner, and J. H. Curtis, Terminal Classic drought in the northern Maya lowlands inferred from multiple sediment cores in Lake Chichancanab (Mexico), Quaternary Science Reviews 24, 1413-1427, 2005a. Hodell, D. A., M. Brenner, J. H. Curtis, and T. Guilderson, Solar forcing of drought frequency in the Maya lowlands, Science 292, 1367-1370, 2001. Hodell, D. A., M. Brenner, J. H. Curtis, R. Medina-Gonzales, E. Ildefonso-Chan Can, A. Albornaz-Pat, and T. P. Guilderson, Climate change on the Yucatan Peninsula during the Little Ice Age, Quaternary Research 63, 109-121, 2005b.

PAGE 136

121 Hodell, D. A., J. H. Curtis, G. A. Jones, A. Higueragundy, M. Brenner, M. W. Binford, and K. T. Dorsey, Reconstruction of Caribbean Climate Change over the Past 10,500 Years, Nature 352, 790-793, 1991. Holland, C. L., R. B. Scott, S.-I. An, and F. W. Taylor, Propagating decadal sea surface temperature signal identified in modern proxy records of the tropical Pacific, Climate Dynamics in press, 2006. Hu, C., E. T. Montgomery, R. W. Schmitt, and F. E. Muller-Karger, The dispersal of the Amazon and Orinoco river water in the Tropical Atalntic and Caribbean Sea: Observation from space and S-PALACE floats, Deep Sea Research Part II 51, 1151-1171, 2004. Jiang, N., J. D. Neelin, and M. Ghil, Quasi-quadrennial and quasi-biennial variability in the equatorial Pacific, Climate Dynamics 12, 101-112, 1995. Johns, W. E., T. N. Lee, F. A. Schott, R. J. Zantopp, and R. H. Evans, The North Brazil Current retroflection seasonal structure and eddy variability, Journal of Geophysical Research 95, 22103-22120, 1990. Johns, W. E., T. L. Townsend, D. M. Fratantoni, and W. D. Wilson, On the Atlantic inflow to the Caribbean Sea, Deep-Sea Research I 49, 211-243, 2002. Jones, P. D., and M. E. Mann, Climate over past millennia, Reviews of Geophysics 42, 2004. Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K. C. Mo, C. Ropelewski, J. Wang, A. Leetmaa, R. Reynolds, R.

PAGE 137

122 Jenne, and D. Joseph, The NCEP/NCAR 40-year reanalysis project, Bulletin of the American Meteorological Society 77, 437-471, 1996. Kaplan, A., M. A. Cane, Y. Kushnir, A. Clement, M. Blumenthal, and B. Rajagopalan, Analyses of global sea surface temperature 1856-1991, Journal of Geophysical Research 103, 18567-18589, 1998. Karr, J. D., and W. J. Showers, Stable oxygen and hydrogen isotopic tracers in Amazon shelf waters during Amasseds, Oceanologica Acta 25, 71-78, 2002. Keigwin, L. D., The Little Ice Age and Medieval Warm Period in the Sargasso Sea, Science 274, 1504-1508, 1996. Kerr, R. A., A North Atlantic climate pacemaker for the centuries, Science 288, 19841985, 2000. Kilbourne, K. H., T. M. Quinn, T. P. Guilderson, F. W. Taylor, A. Winter, and R. S. Webb, Paleoclimate proxy perspective on Caribbean climate since the year 1750, Paleoceanography? in prep, 2006a. Kilbourne, K. H., T. M. Quinn, T. P. Guilderson, R. S. Webb, and F. W. Taylor, Decadalto interannual-scale source water variations in the Caribbean Sea recorded by Puerto Rican coral radiocarbon, Climate Dynamics Submitted, 2006b. Kilbourne, K. H., T. M. Quinn, and F. W. Taylor, A fossil coral perspective on western tropical Pacific climate ~350 ka, Paleoceanography 19, 2004. Knutson, D. W., R. W. Buddemeir, and S. V. Smith, Coral chronometers; seasonal growth bands in reef coral, Science 177, 270-272, 1972.

PAGE 138

123 Kreutz, K. J., P. A. Mayewski, L. D. Meeker, M. S. Twickler, S. I. Whitlow, and I. I. Pittalwala, Bipolar changes in atmospheric circulation during the Little Ice Age, Science 277, 1294-1296, 1997. Lachniet, M. S., S. J. Burns, D. R. Piperno, Y. Asmerom, V. J. Polyak, C. M. Moy, and K. Christenson, A 1500-year El Nino/Southern Oscillation and rainfall history for the Isthmus of Panama from speleothem calcite, Journal of Geophysical Research-Atmospheres 109, 2004. Lamb, H. H., The early medieval warm epoch and its sequel, Palaeogeography Palaeoclimatology Palaeoecology 1, 13-37, 1965. Landsea, C. W., R. A. Pielke, A. Mestas-Nunez, and J. A. Knaff, Atlantic basin hurricanes: Indices of climatic changes, Climatic Change 42, 89-129, 1999. Leaman, K. D., R. L. Molinari, and P. S. Vertes, Structure and Variability of the Florida Current at 27N: April 1982-July 1984, Journal of Physical Oceanography 17, 565-583, 1987. Leder, J. J., P. K. Swart, A. M. Szmant, and R. E. Dodge, The origin of variation in the isotopic record of scleractinian corals: I. Oxygen, Geochimica et Cosmochimica Acta 60, 2857-2870, 1996. Leetmaa, A., P. Niiler, and H. Stommel, Does the Sverdrup Relation account for the MidAtlantic circulation?, Journal of Marine Research 35, 1-10, 1977. Levitus, S. E., Climatological atlas of the world ocean, NOAA Professional Paper 13, U.S. Government Printing Office, Washington DC, 1982.

PAGE 139

124 Linsley, B. K., R. B. Dunbar, G. M. Wellington, and D. A. Mucciarone, A coral-based reconstruction of Intertropical Convergence Zone variability over Central America since 1707, Journal of Geophysical Research 99, 9977-9994, 1994. Linsley, B. K., R. G. Messier, and R. B. Dunbar, Assessing between-colony oxygen isotope variability in the coral Porites lobata at Cliperton Atoll, Coral Reefs 18, 13-27, 1999. Lumpkin, R., and S. L. Garzoli, Near-surface circulation in the tropical Atlantic Ocean, Deep-Sea Research Part I-Oceanographic Research Papers 52, 495-518, 2005. Lund, D. C., and W. Curry, Florida Current surface temperature and salinity variability during the last millennium, Paleoceanography 21, 2006. Maasch, K. A., P. A. Mayewski, E. J. Rohling, J. C. Stager, W. Karln, L. D. Meeker, and E. A. Meyerson, A 2000-year context for modern climate change, Geografiska Annaler 87A, 7-15, 2005. Mahadevan, A., An analysis of bomb radiocarbon trends in the Pacific, Marine Chemistry 73, 273-290, 2001. Mann, M. E., R. S. Bradley, and M. K. Hughes, Global-scale temperature patterns and climate forcing over the past six centuries, Nature 392, 779-787, 1998. Mann, M. E., R. S. Bradley, and M. K. Hughes, Northern Hemisphere temperatures during the past millennium: inferences, uncertainties, and limitations, Geophysical Research Letters 26, 759-762, 1999. Mann, M. E., and K. A. Emanuel, Atlantic hurricane trends linked to climate change, EOS, Transactions, American Geophysical Union 87, 233, 2006.

PAGE 140

125 Marshall, J., Y. Kushnir, D. S. Battisti, P. Chang, A. Czaja, R. Dickson, J. Hurrell, M. McCartney, R. Saravanan, and M. Visbeck, North Atlantic climate variability: phenomena impacts and mechanisms, International Journal of Climatology 21, 1863-1898, 2001. Marshall, J. F., and M. T. McCulloch, An assessment of the Sr/Ca ratio in shallow water hermatypic corals as proxy for sea surface temperature, Geochimica et Cosmochimica Acta 66, 3263-3280, 2002. Matthews, J. A., and K. R. Briffa, The 'Little Ice Age': Re-evaluation of an evolving concept, Geografiska Annaler Series a-Physical Geography 87A, 17-36, 2005. Mayer, D. A., and R. H. Weisberg, A description of COADS surface meteorological fields and the implied Sverdrup transports for the Atlantic Ocean from 30S to 60N, Journal of Physical Oceanography 23, 2201-2221, 1993. Mayewski, P. A., K. A. Maasch, Y. Yan, S. Kang, E. A. Meyerson, S. B. Sneed, S. D. Kaspari, D. A. Dixon, E. C. Osterberg, V. I. Morgan, T. van Ommen, and M. A. J. Curran, Solar forcing of the polar atmosphere data, mechanism, and implication, submmitted 2005. Mayewski, P. A., E. J. Rohling, J. C. Stager, W. Karlen, K. A. Maasch, L. D. Meeker, E. A. Meyerson, F. Gasse, S. van Kreveld, K. Holmgren, J. Lee-Thorp, G. Rosqvist, F. Rack, M. Staubwasser, R. R. Schneider, and E. J. Steig, Holocene climate variability, Quaternary Research 62, 243-255, 2004. Mazeika, P. A., T. H. Kinder, and D. A. Burns, Measurements of subtidal flow in the Lesser Antilles passages, Journal of Geophysical Research 88, 7783-4488, 1983.

PAGE 141

126 McConnaughey, T., 13 C and 18 O isotopic disequilibrium in biological carbonate: I. In vitro simulation of kinetic isotope effects, Geochimica et Cosmochimica Acta 53, 163171, 1989a. McConnaughey, T., 13 C and 18 O isotopic disequilibrium in biological carbonates: II. Patterns, Geochimica et Cosmochimica Acta 53, 151-162, 1989b. Metcalf, W. G., M. C. Stalcup, and M. E. Zemanovic, Hydrographic station data from Atlantis II Cruise 56 to the southeastern approaches to the Caribbean Sea, February-April 1970., Rep. WHOI 71-13 103 pp., Woods Hole Oceanographic Institution, Woods Hole, MA, 1971. Moberg, A., D. M. Sonechkin, K. Holmgren, N. M. Datsenko, and W. Karlen, Highly variable Northern Hemisphere temperatures reconstructed from lowand highresolution proxy data, Nature 433, 613-617, 2005. Morse, J. W., and F. T. MacKenzie, Geochemistry of Sedimentary Carbonates 705 pp., Elsevier Science Ltd., New York, 1990. Mller-Karger, F. E., C. R. McClain, T. R. Fisher, W. E. Esaias, and R. Varela, Pigment distribution in the Caribbean Sea: Observations from space, Progress in Oceanography 23, 23-64, 1989. Muller-Karger, F. E., P. L. Richardson, and D. McGillicuddy, On the offshore dispersal of the Amazon's plume in the North Atlantic: Comments on the paper by A. Longhurst, "Seasonal cooling and blooming in tropical oceans", Deep Sea Research 42, 2127-2137, 1995. Niiler, P. P., and W. S. Richardson, Seasonal variability in the Florida Current, Journal of Marine Research 21, 144-167, 1973.

PAGE 142

127 Nobre, P., and J. Shukla, Variations of sea surface temperature, wind stress, and rainfall over the tropical Atlantic and South America, Journal of Climate 9, 2464-2479, 1996. Nyberg, J., A. Kuijpers, B. A. Malmgren, and H. Kunzendorf, Late Holocene changes in precipitation and hydrography recorded in marine sediments from the northeastern Caribbean Sea, Quaternary Research 56, 87-102, 2001. Nyberg, J., B. A. Malmgren, A. Kuijpers, and A. Winter, A centennial-scale variability of tropical North Atlantic surface hydrography during the late Holocene, Palaeogeography Palaeoclimatology Palaeoecology 183, 25-41, 2002. Nydal, R., and K. Lvseth, Carbon-14 Measurements in Atmospheric CO2 from Northern and Southern Hemisphere Sites, 1962-1993, Rep. ORNL/CDIAC-93, NDP-057 67 pp., Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1996. O'Brien, S. R., P. A. Mayewski, L. D. Meeker, D. A. Meese, M. S. Twickler, and S. I. Whitlow, Complexity of Holocene climate as reconstructed from a Greenland ice core, Science 270, 1962-1964, 1995. Ostlund, H. G., and C. Grall, North and Tropical Atlantic tritium and radiocarbon, 277 pp., R.S.M.A.S., University of Miami, Miami, Florida, 1987. Otto-Bliesner, B. L., R. Tomas, E. C. Brady, C. Ammann, Z. Kothavala, and G. Clauzet, Climate sensitivity of moderate and low resolution versions of CCSM3 to preindustrial forcings, Journal of Climate CCSM special issue, in Press, 2006. Overpeck, J. T., K. A. Hughen, D. Hardy, R. S. Bradley, R. Case, M. Douglas, B. Finney, K. Gajewski, G. Jacoby, A. Jennings, S. Lamoureux, A. Lasca, G. MacDonald, J.

PAGE 143

128 Moore, M. Retelle, S. Smith, A. Wolfe, and G. Zielinski, Arctic environmental change of the last four centuries, Science 2778, 1251-1256, 1997. Paillard, D., L. Labeyrie, and P. Yiou, Macintosh program performs time-series analysis, EOS, Transactions, American Geophysical Union 77, 379, 1996. Parr, A. E., A contribution to the Hydrography of the Caribbean and Cayman Seas, based upon the observations made by the research ship "Atlantis", 1933-1934, Bulletin of the Bingham Oceanographic Collection 5, 110, 1937. Peterson, L. C., and G. H. Haug, Variability in the mean latitude of the Atlantic Intertropical Convergence Zone as recorded by riverine input of sediments to the Cariaco Basin (Venezuela), Palaeogeography Palaeoclimatology Palaeoecology 234, 97-113, 2006. Poveda, G., P. R. Waylen, and R. S. Pulwarty, Annual and inter-annual variability of the present climate in northern South America and southern Mesoamerica, Palaeogeography Palaeoclimatology Palaeoecology 234, 3-27, 2006. Quinn, T. M., and F. W. Taylor, SST artifacts in coral proxy records produced by early marine diagenesis in a modern coral from Rabaul, Papua New Guinea, Geophysical Research Letters 33, L04601, 2006. Ramaswamy, V., O. Boucher, J. Haigh, D. Hauglustaine, J. Haywood, G. Myhre, T. Nakajima, G. Y. Shi, and S. Solomon, Radiative forcing of climate change. in Climate change 2001: the scientific basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel of Climate Change. edited by Houghton, J. T., Y. Ding, D. J. Griggs, M. Nouger, P. J. van der Linden, X. Dai, et al., Cambridge University Press, 2001.

PAGE 144

129 Rasmusson, E. M., X. Wang, and C. F. Ropelewski, The biennial component of ENSO variability, Journal of Marine Systems 1, 71-96, 1990. Rayner, N. A., D. E. Parker, C. K. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, Global analyses of sea surface temperature, sea ice and night marine air temperature since the late nineteenth century, Journal of Geophysical Research 108, 4407, 2003. Reimer, P., M. G. L. Baillie, E. Bard, A. Bayliss, W. Beck, J., C. J. H. Bertrand, P. G. Blackwell, C. E. Buck, G. S. Burr, K. B. Cutler, P. E. Damon, R. L. Edwards, R. G. Fairbanks, M. Friedrich, T. P. Guilderson, A. G. Hogg, K. A. Hughen, B. Kromer, G. McCormac, S. Manning, C. B. Ramsey, R. W. Reimer, S. Remmele, J. R. Southon, M. Stuiver, S. Talamo, F. W. Taylor, J. van der Plicht, and C. E. Weyhenmeyer, INTCAL04 Terrestrial Radiocarbon Age Calibration, 0-26 Cal kyr bp, Radiocarbon 46, 1029-1058, 2004. Richardson, P. L., Caribbean Current and eddies as observed by surface drifters, DeepSea Research Part Ii-Topical Studies in Oceanography 52, 429-463, 2005. Richardson, P. L., G. E. Hufford, R. Limeburner, and W. S. Brown, North Brazil Current retroflection eddies, Journal of Geophysical Research 99, 5081-5093, 1994. Rozanski, K., L. Aragus-Aragus, and R. Gonfiantini, Isotopic Patterns in Modern Global Precipitation. in Climate Change in Continental Isotopic Records edited by Swart, P. K., K. C. Lohmann, J. McKenzie and S. Savin, pp. 1-36, American Geophysical Union, Washington, DC, 1993. Schlesinger, M. E., and N. Ramankutty, An oscillation in the global climate system of period 65-70 years, Nature 367, 723-726, 1994.

PAGE 145

130 Schmidt, G. A., Error analysis of paleo-salinity calculations, Paleoceanography 13, 422429, 1999. Schmidt, G. A., G. R. Bigg, and E. J. Rohling, Global Seawater Oxygen-18 Database, http://data.giss.nasa.gov/o18data/ 1999. Schmidt, M. W., H. J. Spero, and D. W. Lea, Links between salinity variation in the Caribbean and North Atlantic thermohaline circulation, Nature 428, 160-163, 2004. Schmitz, W. J., Jr., and M. S. MCCartney, On the North Atlantic Circulation, Reviews of Geophysics 31, 29-49, 1993. Schmitz, W. J., and P. L. Richardson, On the sources of the Florida Current, Deep Sea Research 38, S379-409, 1991. Schmitz, W. J., and W. S. Richardson, On the transport of the Florida Current, Deep Sea Research 15, 679-693, 1968. Schrag, D. P., Rapid analysis of high-precision Sr/Ca ratios in corals and other marine carbonates, Paleoceanography 14, 97-102, 1999. Smith, J. M., T. M. Quinn, K. P. Helmle, and R. B. Halley, Reproducibility of geochemical and climatic signals in the Atlantic coral Montastraea faviolata, Paleoceanography 21, PA1010, 2006. Smith, S. V., R. W. Buddemeier, R. C. Redalje, and J. E. Houck, Strontium-calcium thermometry in coral skeletons, Science 204, 404-407, 1979. Stair, K., Assessing the reproducibility of skeletal geochemistry records in Atlantic corals using Montastrea faveolata coral heads from the Dry Tortugas, Florida, M.S. Thesis, University of South Florida, St. Petersburg, Florida, 2006.

PAGE 146

131 Stephans, C. L., T. M. Quinn, F. W. Taylor, and T. Correge, Assessing the reproducibility of coral-based climate records, Geophysical Research Letters 31, 2004. Stuiver, M., and H. G. Ostlund, GEOSECS Atlantic radiocarbon, Radiocarbon 22, 2553, 1980. Stuiver, M., and H. A. Polach, Discussion and reporting of 14 C data, Radiocarbon 19, 355-363, 1977. Suess, H. E., Natural radiocarbon and the rate of exchange of carbon dioxide between the atmosphere and the sea. in Nuclear Processes in Geologic Settings edited by Science, N. R. C. C. o. N., pp. 52-56, National Academy of Sciences, Washington D.C., 1953. Sutton, R. T., and M. R. Allen, Decadal predictability of North Atlantic sea surface temperature and climate, Nature 388, 563-567, 1997. Sverdrup, H. U., Wind-driven currents in a baroclinic ocean: with application to the equatorial currents of the eastern Pacific, Proceedings of the National Academy of Sciences of the U.S.A. 33, 318-326, 1947. Swart, P. K., H. Elderfield, and M. J. Greaves, A high-resolution calibration of Sr/Ca thermometry using the Caribbean coral Montastraea annularis Geochemistry Geophysics and Geosystems 3, 8402, 2002. Swart, P. K., A. Saied, C. Schroeder, and L. Williams, Temporal and spatial variations in the oxygen isotopic compositions of surface waters from the Northern Caribbean, EOS, Transactions, American Geophysical Union 84, Fall Meeting Supp. Abstract OS31C-0218, 2003.

PAGE 147

132 Takahashi, T., T.-H. Peng, and S. Sutherland, Radiocarbon and CO2 measurements from Tropical Atlantic Study (TTO/TAS), Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tennessee, 1995. Tang, L. Q., J. Y. Sheng, B. G. Hatcher, and P. F. Sale, Numerical study of circulation, dispersion, and hydrodynamic connectivity of surface waters on the Belize shelf, Journal of Geophysical Research-Oceans 111, 2006. Taylor, M. A., D. B. Enfield, and A. A. Chen, Influence of the tropical Atlantic versus the tropical Pacific on Caribbean rainfall, Journal of Geophysical Research-Oceans 107, 2002. Thompson, L. G., and E. Mosley-Thompson, Evidence of abrupt climatic change during the last 1,500 years recorded in ice cores from the tropical Quelccaya ice cap, Peru, paper presented at Abrupt Climatic Change Evidence and Implications, NATO/NSF Workshop, 1987. Thompson, L. G., and E. Mosley-Thompson, One-half millennia of tropical climate variability as recorded in the Stragraphy of the Quelccaya Ice Cap, Peru. in Aspects of climate variability in the Pacific and the western Americas, Geophysical Monograph edited by Peterson, D. H., pp. 15-31, American Geophysical Union, Washington, DC, 1989. Thompson, L. G., E. Mosley-Thompson, and B. M. Arnao, Major El Nio/Southern Oscillation events recorded in stratigraphy of the tropical Quelccaya Ice Cap, Science 26, 50-52, 1984a.

PAGE 148

133 Thompson, L. G., E. Mosley-Thompson, J. F. Bolzan, and B. R. Kochi, A 1500-year record of tropical precipitation in ice cores from the Quelccaya ice cap, Peru, Science 229, 971-973, 1985. Thompson, L. G., E. Mosley-Thompson, and P. Grootes, The Little Ice Age as recorded in the stratigraphy of the Torpical Quelccaya Ice Core, Science 234, 361-364, 1986. Thompson, L. G., E. Mosley-Thompson, P. Grootes, and M. Pourchet, Tropical glaciers: potential for paleoclimatic reconstruction, Journal of Geophysical Research 89, 4638-4646, 1984b. Timmermann, A., M. Latif, R. Voss, and A. Grotzner, Northern Hemispheric interdecadal variability: A coupled air-sea mode, Journal of Climate 11, 1906-1931, 1998. Torrence, C., and G. P. Compo, A practical guide to wavelet analysis, Bulletin of the American Meteorological Society 79, 61-78, 1998. Tudhope, A. W., D. W. Lea, G. B. Shimmield, C. P. Chilcott, and S. Head, Monsoon climate and Arabian Sea coastal upwelling recorded in massive corals from southern Oman, Palaios 11, 347-361, 1996. Vellinga, M., and R. Wood, Global climatic impacts of a collapse of the Atlantic thermohaline circulation, Climatic Change 54, 251-267, 2002. Vellinga, M., and P. Wu, Low-latitude fresh water influence on centennial variability of the Atlantic thermohaline circulation, Journal of Climate 17, 4498-4511, 2004. Vogel, J. S., J. R. Southon, and D. E. Nelson, Catalyst and binder effects in the use of filamentous graphite for AMS., Nuclear Instruments and Methods in Physics Research B 29, 50-56, 1987.

PAGE 149

134 von Storch, H., E. Zorita, J. M. Jones, Y. Dimitriev, F. Gonzalez-Rouco, and S. F. B. Tett, Reconstructing past climate from noisy data, Science 306, 679-682, 2004. Waliser, D. E., and C. Gautier, A satellite-derived climatology of the ITCZ, Journal of Climate 6, 2162-2174, 1993. Wang, C., Atlantic climate variability and its associated atmospheric circulation cells, Journal of Climate 15, 1516-1536, 2001. Wang, C., An overlooked feature of tropical climate: Inter-Pacific-Atlantic variability, Geophysical Research Letters 33, L12702, 2006. Wang, C., and D. B. Enfield, The tropical Western Hemisphere warm pool, Geophysical Research Letters 28, 1635-1638, 2001. Wang, C., and D. B. Enfield, A further study of the Tropical Western Hemisphere Warm Pool, Journal of Climate 16, 1476-1493, 2003. Wang, X., A. S. Auler, R. L. Edwards, H. Cheng, P. S. Christalli, P. L. Smart, D. A. Richards, and C.-C. Shen, Wet periods in northeastern Brazil over the past 210 kyr linked to distant climate anomalies, Nature 432, 740-743, 2004. Watanabe, T., A. Winter, and T. Oba, Seasonal changes in sea surface temperature and salinity during the little Ice Age in the Caribbean Sea deduced from Mg/Ca and 18 O/ 16 O ratios in corals, Marine Geology 173, 21-35, 2001. Watanabe, T., A. Winter, T. Oba, R. Anzai, and H. Ishioroshi, Evaluation of the fidelity of isotope records as an environmental proxy in the coral Montastrea, Coral Reefs 21, 169-178, 2002.

PAGE 150

135 Webster, P. J., G. J. Holland, J. A. Curry, and H. R. Chang, Changes in tropical cyclone number, duration and intensity, in a warming environment, Science 309, 18441846, 2005. Widmann, M., and S. F. B. Tett, Simulating the climate of the last millennium, Pages News 11, 2003. Wilson, W. D., and W. E. Johns, Velocity structure and transport in the Windward Islands passages, Deep Sea Research 44, 487-520, 1997. Winter, A., R. S. Appeldoorn, A. Bruckner, E. H. J. Williams, and C. Goenaga, Sea surface temperatures and coral reef bleaching off La Parguera, Puerto Rico (northeastern Caribbean Sea). Coral Reefs 17, 377-382, 1998. Winter, A., H. Ishioroshi, T. Watanabe, and T. Oba, Caribbean sea surface temperatures: two-to-three degrees cooler than present during the Little Ice Age, Geophysical Research Letters 27, 3365-3368, 2000. Wunch, C., What is the thermohaline circulation?, Science 298, 1179-1180, 2002. Wust, G., Stratification and circulation in the Antillean-Caribbean basins Columbia University Press, New York, 1964. Zhang, D. X., M. J. McPhaden, and W. E. Johns, Observational evidence for flow between the subtropical and tropical Atlantic: The Atlantic subtropical cells, Journal of Physical Oceanography 33, 1783-1797, 2003.

PAGE 151

136 APPENDICIES

PAGE 152

137 Appendix 1: Annual Data ID Year Extension Rate Cumulative Depth 13 C 18 O Sr/Ca Mg/Ca 14 C 14 C error (mm/year) (cm) ( VPDB) ( VPDB) (mmol/mol) (mmol/mol) () () 04LPTA1 1 2004 8 0.8 -1.46 -4.57 9.024 11.697 65.1 3.0 04LPTA1 2 2003 7 1.5 -1.91 -4.52 9.048 8.624 70.5 3.2 04LPTA1 3 2002 8.8 2.38 -2.27 -4.72 8.997 13.397 04LPTA1 4 2001 10 3.38 -2.45 -4.69 9.021 7.582 68.8 3.5 04LPTA1 5 2000 7.4 4.12 -1.50 -4.36 9.103 9.434 77.1 3.6 04LPTA1 6 1999 7 4.82 -1.74 -4.44 9.091 7.687 79.6 3.3 04LPTA1 7 1998 9.1 5.73 -2.08 -4.56 9.112 6.048 70.4 3.2 04LPTA1 8 1997 9.9 6.72 -2.26 -4.62 9.049 6.225 81.0 3.3 04LPTA1 9 1996 11 7.82 -1.33 -4.43 9.074 5.990 91.1 3.5 04LPTA1 10 1995 7.2 8.54 -2.39 -4.81 8.970 8.329 04LPTA1 11 1994 10 9.54 -1.72 -4.49 9.028 7.550 84.3 3.2 04LPTA1 12 1993 10 10.54 -2.28 -4.66 9.028 6.660 88.4 3.3 04LPTA1 13 1992 13 11.84 -1.86 -4.50 8.975 7.334 04LPTA1 14 1991 12 13.04 -1.67 -4.50 9.087 6.720 90.1 3.3 04LPTA1 15 1990 9.5 13.99 -1.80 -4.45 9.106 6.154 100.3 3.3 04LPTA1 16 1989 9.2 14.91 -1.88 -4.42 9.003 7.550 87.8 3.1 04LPTA1 17 1988 9.2 15.83 -2.43 -4.54 9.004 8.065 108.6 3.4 04LPTA1 18 1987 8 16.63 -1.88 -4.59 9.043 5.910 106.0 4.1 04LPTA1 19 1986 12 17.83 -1.93 -4.52 9.049 6.761 108.1 3.4 04LPTA1 20 1985 13.9 19.22 -1.22 -4.36 9.170 6.360 118.0 3.4 04LPTA1 21 1984 8.6 20.08 -1.66 -4.44 9.087 6.090 04LPTA1 22 1983 7.6 20.84 -1.48 -4.39 9.098 5.556 111.4 3.4 04LPTA1 23 1982 10 21.84 -1.68 -4.46 9.062 6.464 115.7 3.4 04LPTA1 24 1981 5.8 22.42 -1.17 -4.20 9.149 5.846 04LPTA1 25 1980 12.8 23.7 -1.37 -4.46 9.175 5.780 129.1 3.8 04LPTA1 26 1979 7.5 24.45 -2.00 -4.50 9.032 6.627 114.9 3.4 04LPTA1 27 1978 10.4 25.49 -1.74 -4.47 9.079 6.543 127.5 3.4 04LPTA1 28 1977 10.6 26.55 -1.33 -4.37 9.146 5.646 129.0 3.4 04LPTA1 29 1976 10 27.55 -2.10 -4.48 9.044 6.370 147.7 3.8 04LPTA1 30 1975 8.8 28.43 -1.50 -4.17 9.107 6.403 145.2 3.5 04LPTA1 31 1974 9.1 29.34 -1.64 -4.24 9.087 7.116 130.7 5.4 04LPTA1 32 1973 10 30.34 -1.81 -4.26 9.055 7.667 132.7 3.4 04LPTA2 2 1972 7.2 31.06 -1.74 -4.18 9.089 9.670 135.5 3.8 04LPTA2 3 1971 6.5 31.71 -1.53 -4.14 9.075 7.221 120.4 3.5 04LPTA2 4 1970 6.8 32.39 -0.96 -4.07 9.194 6.696 127.5 4.1 04LPTA2 5 1969 8.8 33.27 -1.40 -4.13 9.190 6.438 116.8 3.9 04LPTA2 6 1968 8.7 34.14 -1.41 -4.27 9.173 5.986 110.5 3.9 04LPTA2 7 1967 8.7 35.01 -1.47 -4.37 9.092 8.074 87.9 3.4 04LPTA2 8 1966 8.3 35.84 -1.65 -4.31 9.078 7.349 62.3 3.3 04LPTA2 9 1965 7.1 36.55 9.138 9.044 79.1 3.3 04LPTA2 10 1964 9 37.45 -0.69 -4.21 9.178 6.870 22.9 3.4 04LPTA2 11 1963 8 38.25 -1.40 -4.23 9.196 5.709 1.3 3.5 04LPTA2 14 1962 9.2 39.17 -1.24 -4.25 9.156 6.156 -9.9 2.9 04LPTA2 15 1961 9.5 40.12 -1.55 -4.24 9.153 7.991 -25.7 3.0 04LPTA2 16 1960 9 41.02 -0.89 -4.19 9.173 5.665 -32.1 3.0

PAGE 153

138 Appendix 1 (Continued) 04LPTA2 17 1959 15 42.52 -1.74 -4.29 9.076 7.042 -45.8 3.1 04LPTA2 18 1958 14 43.92 -1.29 -4.25 9.156 7.829 -47.3 2.9 04LPTA2 19 1957 10.5 44.97 -1.33 -4.32 9.057 7.209 -49.1 3.2 04LPTA2 20 1956 8.5 45.82 -0.39 -4.15 9.194 5.819 -57.4 2.6 04LPTA2 21 1955 6.75 46.495 -0.83 -4.36 9.130 6.287 -61.9 2.9 04LPTA2 22 1954 5.5 47.045 -1.50 -4.41 9.042 6.753 -63.4 2.9 04LPTA2 23 1953 7.75 47.82 -1.40 -4.41 9.060 6.720 -64.9 3.1 04LPTA2 24 1952 6 48.42 -1.37 -4.39 9.020 6.202 -64.5 2.8 04LPTA2 25 1951 7.5 49.17 -1.69 -4.40 9.032 6.503 -63.3 3.0 04LPTA2 26 1950 8.5 50.02 -1.02 -4.37 9.119 5.604 -53.5 2.8 04LPTA2 27 1949 8.5 50.87 -1.13 -4.42 9.050 5.918 04LPTA2 28 1948 8 51.67 -1.56 -4.43 9.049 6.269 04LPTA2 29 1947 9 52.57 -0.77 -4.30 9.148 5.863 04LPTA2 30 1946 7.7 53.34 -1.33 -4.31 9.059 7.294 04LPTA2 31 1945 10.5 54.39 -0.83 -4.26 9.121 5.698 -56.7 3.1 04LPTA2 32 1944 7 55.09 -0.73 -4.17 9.169 5.830 04LPTA2 33 1943 7.9 55.88 -0.80 -4.36 9.049 6.032 04LPTA2 34 1942 10 56.88 -1.25 -4.38 9.086 6.356 04LPTA2 35 1941 8.25 57.705 -1.02 -4.31 9.057 7.115 04LPTA2 36 1940 10.5 58.755 -1.52 -4.18 9.056 6.883 -53.6 3.4 04LPTA2 37 1939 7 59.455 -1.20 -4.33 9.060 6.716 04LPTA2 38 1938 7.4 60.195 -1.08 -4.40 9.094 6.758 04LPTA2 39 1937 7.75 60.97 -0.92 -4.23 9.132 6.017 04LPTA2 40 1936 8 61.77 -0.88 -4.27 9.157 5.931 04LPTA2 44 1935 8.5 62.62 -1.20 -4.37 9.126 6.759 -54.4 2.8 04LPTA2 45 1934 8 63.42 -1.11 -4.25 9.129 7.251 04LPTA2 46 1933 7.2 64.14 -1.05 -4.41 9.113 5.650 04LPTA2 47 1932 8 64.94 -1.31 -4.58 9.116 6.138 04LPTA2 48 1931 7.3 65.67 -0.91 -4.20 9.170 6.222 04LPTA2 50 1930 6.6 66.33 -0.99 -4.29 9.157 7.692 -52.6 2.6 04LPTA2 51 1929 7.4 67.07 -0.96 -4.15 9.184 6.676 04LPTA2 52 1928 6.6 67.73 -1.42 -4.44 9.131 7.612 04LPTA2 53 1927 7 68.43 -1.18 -4.35 9.092 6.038 04LPTA2 54 1926 6.7 69.1 -1.99 -4.45 9.038 6.374 04LPTA2 55 1925 5.2 69.62 -1.71 -4.30 9.099 8.103 -53.1 3.6 04LPTA2 56 1924 7 70.32 -1.80 -4.31 9.086 8.902 04LPTA2 57 1923 6.7 70.99 -1.49 -4.09 9.079 7.892 04LPTA2 58 1922 7.8 71.77 -1.07 -4.03 9.138 5.842 04LPTA2 59 1921 5.5 72.32 -0.96 -4.15 9.132 5.909 04LPTA2 60 1920 8 73.12 -1.36 -4.17 -49.0 3.4 04LPTA2 61 1919 6.2 73.74 -0.92 -4.16 9.155 5.585 04LPTA2 62 1918 8 74.54 -0.77 -4.02 9.138 5.892 04LPTA2 63 1917 5.2 75.06 -1.27 -4.15 9.084 6.878 04LPTA3 1 1916 6 75.66 -1.39 -4.28 9.059 5.434 04LPTA3 2 1915 8 76.46 -1.00 -4.33 9.134 5.170 -43.1 3.1 04LPTA3 3 1914 8.6 77.32 -1.14 -4.31 9.117 5.433 04LPTA3 4 1913 7.7 78.09 -1.00 -4.19 9.153 5.075 04LPTA3 5 1912 8 78.89 -1.27 -4.33 9.116 5.042 04LPTA3 6 1911 7 79.59 -0.81 -4.18 9.111 5.515

PAGE 154

139 Appendix 1 (Continued) 04LPTA3 7 1910 8 80.39 -0.97 -4.33 9.114 5.582 -49.2 2.8 04LPTA3 9 1909 7.4 81.13 -1.02 -4.12 9.125 5.753 04LPTA3 10 1908 8.2 81.95 -1.16 -4.10 9.154 5.620 04LPTA3 11 1907 10 82.95 -1.23 -4.26 9.095 5.884 04LPTA3 12 1906 8 83.75 -1.13 -4.18 9.087 6.112 04LPTA3 13 1905 8.4 84.59 -1.04 -4.06 9.079 5.671 -49.2 2.8 04LPTA3 14 1904 7.8 85.37 -1.29 -4.09 9.032 6.546 04LPTA3 15 1903 7.1 86.08 -1.08 -4.22 9.070 6.116 04LPTA3 16 1902 9.4 87.02 -1.16 -4.00 9.101 6.079 04LPTA3 17 1901 10 88.02 -1.26 -4.30 9.091 5.873 04LPTA3 18 1900 6.9 88.71 -1.10 -4.21 9.087 5.699 -50.9 2.8 04LPTA3 19 1899 7.4 89.45 -0.74 -4.25 9.107 5.444 04LPTA3 20 1898 6 90.05 -0.74 -4.35 9.066 5.635 04LPTA3 21 1897 6.5 90.7 -0.46 -4.18 9.139 5.107 04LPTA3 22 1896 6.6 91.36 -0.31 -4.10 9.107 5.373 04LPTA3 23 1895 8 92.16 -0.60 -4.05 9.132 5.455 -49.1 2.8 04LPTA3 24 1894 7.6 92.92 -1.19 -4.24 9.074 5.981 04LPTA3 25 1893 6.5 93.57 -0.85 -4.20 9.100 5.304 04LPTA3 26 1892 7.3 94.3 -0.75 -4.16 9.129 5.600 04LPTA3 27 1891 7.3 95.03 -0.48 -4.09 9.166 5.553 04LPTA3 28 1890 9 95.93 -1.36 -4.40 9.072 6.029 -59.3 2.8 04LPTA3 29 1889 8 96.73 -1.38 -4.30 9.043 6.960 04LPTA3 30 1888 8.3 97.56 -1.28 -4.19 9.045 7.251 04LPTA3 32 1887 8 98.36 -0.93 -4.24 9.098 5.684 04LPTA3 33 1886 7 99.06 -0.59 -4.24 9.154 5.171 04LPTA3 34 1885 7.3 99.79 -1.02 -4.17 9.122 5.125 -56.7 2.8 04LPTA3 35 1884 8.8 100.67 -1.38 -4.53 9.086 5.424 04LPTA3 36 1883 8 101.47 -1.10 -4.40 9.119 5.559 04LPTA3 37 1882 8 102.27 -1.19 -4.33 9.107 5.564 04LPTA3 39 1881 7 102.97 -0.99 -4.18 9.093 6.101 04LPTA3 40 1880 6 103.57 -0.94 -4.31 9.070 6.134 -50.4 2.8 04LPTA3 41 1879 6 104.17 -0.97 -4.34 9.183 5.288 04LPTA3 42 1878 5.7 104.74 -1.68 -4.52 9.106 5.459 04LPTA3 43 1877 6.4 105.38 -0.49 -4.36 9.141 5.304 04LPTA3 44 1876 6.7 106.05 -0.61 -4.37 9.131 5.582 04LPTA3 45 1875 7.3 106.78 -1.09 -4.37 9.096 5.869 -51.2 3.2 04LPTA3 46 1874 9.3 107.71 -1.05 -4.32 9.132 5.665 04LPTA3 47 1873 8.3 108.54 -1.00 -4.28 9.112 5.911 04LPTA3 48 1872 7.2 109.26 -0.39 -4.10 9.177 5.226 04LPTA3 49 1871 7.6 110.02 -1.43 -4.28 9.040 6.804 04LPTA3 50 1870 6.2 110.64 -1.40 -4.39 9.070 5.738 -48.0 3.6 04LPTA3 51 1869 7 111.34 -1.05 -4.36 9.135 6.418 04LPTA3b 2 1868 8 112.14 -0.96 -4.06 9.126 6.932 04LPTA3b 3 1867 7.8 112.92 -0.41 -4.10 9.182 5.571 04LPTA3b 4 1866 8 113.72 -0.69 -4.25 9.164 5.000 04LPTA3b 5 1865 6.6 114.38 -0.93 -4.15 9.149 4.958 -49.2 2.8 04LPTA3b 6 1864 7.5 115.13 -1.06 -4.23 9.182 6.207 04LPTA3b 7 1863 7.7 115.9 -1.35 -4.21 9.118 6.955 04LPTA3b 9 1862 7.2 116.62 -0.91 -4.12 9.107 5.889 04LPTA3b 10 1861 8.6 117.48 -0.71 -4.04 9.121 6.050

PAGE 155

140 Appendix 1 (Continued) 04LPTA3b 11 1860 7.2 118.2 -1.41 -4.13 9.089 5.877 -47.3 2.8 04LPTA3b 12 1859 6.6 118.86 -0.72 -4.12 9.158 5.475 04LPTA3b 13 1858 8.6 119.72 -0.97 -4.04 9.141 5.627 04LPTA3b 14 1857 8.8 120.6 -1.40 -4.29 9.096 6.155 04LPTA3b 15 1856 7 121.3 -0.94 -4.16 9.165 5.154 04LPTA4 1 1855 7 122 -0.74 -4.31 9.180 4.410 -55.0 2.8 04LPTA4 2 1854 8 122.8 -0.51 -4.15 9.198 4.538 04LPTA4 3 1853 6.5 123.45 -0.38 -3.90 9.099 4.892 04LPTA4 4 1852 7.4 124.19 -1.13 -4.27 9.112 4.993 04LPTA4 5 1851 7 124.89 -1.27 -4.19 9.150 5.065 04LPTA4 6 1850 7.3 125.62 -0.53 -3.94 9.181 4.709 04LPTA4 8 1849 9.6 126.58 -1.28 -4.25 9.106 4.860 -48.3 3.2 04LPTA4 9 1848 7.5 127.33 -0.61 -4.11 9.206 4.665 04LPTA4 10 1847 7.3 128.06 -1.30 -4.21 9.126 4.940 04LPTA4 11 1846 6.7 128.73 -1.03 -4.25 9.144 4.790 04LPTA4 12 1845 7.2 129.45 -0.28 -3.98 9.217 4.550 -47.1 2.7 04LPTA4 13 1844 8.4 130.29 -1.04 -4.20 9.103 4.974 04LPTA4 14 1843 7.5 131.04 -0.34 -4.09 9.168 4.691 04LPTA4 15 1842 8 131.84 -1.43 -4.37 9.086 5.116 04LPTA4 16 1841 8.4 132.68 -1.52 -4.23 9.105 5.421 04LPTA4 17 1840 7.5 133.43 -1.75 -4.30 9.055 5.139 -53.7 2.8 04LPTA4 18 1839 8 134.23 -0.91 -4.11 9.141 5.291 04LPTA4 20 1838 6.6 134.89 -0.89 -4.09 9.142 4.803 04LPTA4 21 1837 6.9 135.58 -1.19 -4.17 9.100 4.839 04LPTA4 22 1836 6.9 136.27 -0.40 -4.08 9.151 4.657 04LPTA4 23 1835 8.6 137.13 -0.75 -4.16 9.133 4.620 -54.8 2.9 04LPTA4 24 1834 9 138.03 -0.86 -4.08 9.146 4.784 04LPTA4 25 1833 6 138.63 -0.89 -4.12 9.150 4.633 04LPTA4 26 1832 7.3 139.36 -0.87 -4.17 9.112 4.790 04LPTA4 27 1831 9 140.26 -1.28 -4.27 9.110 4.822 04LPTA4 28 1830 7.7 141.03 -1.63 -4.19 9.095 5.160 -51.5 2.9 04LPTA4 29 1829 7.5 141.78 -1.04 -4.10 9.104 4.830 04LPTA4 30 1828 8.9 142.67 -1.08 -4.44 9.089 4.889 04LPTA4 31 1827 7.6 143.43 -0.89 -4.08 9.149 5.105 04LPTA4 32 1826 8 144.23 -1.15 -4.12 9.132 4.806 04LPTA4 33 1825 8 145.03 -0.77 -3.97 9.184 4.810 -52.4 3.0 04LPTA4 34 1824 8 145.83 -1.10 -4.19 9.127 5.129 04LPTA4 35 1823 8 146.63 -1.11 -4.12 9.160 4.980 04LPTA4 36 1822 7.7 147.4 -1.02 -4.15 9.193 4.951 04LPTA4 37 1821 8.6 148.26 -0.89 -4.10 9.153 4.833 04LPTA5 1 1820 7.2 148.98 -0.49 -4.18 9.226 4.706 -51.8 2.9 04LPTA5 2 1819 7.2 149.7 -0.50 -4.21 9.229 4.458 04LPTA5 3 1818 7.2 150.42 -0.74 -4.32 9.203 4.484 04LPTA5 4 1817 7.2 151.14 -1.14 -4.16 9.140 4.751 04LPTA5 5 1816 9.3 152.07 -0.86 -4.28 9.131 4.803 04LPTA5 6 1815 7.5 152.82 -0.80 -4.27 9.141 5.251 -49.8 2.9 04LPTA5 7 1814 8.8 153.7 -1.12 -4.28 9.155 4.950 04LPTA5 8 1813 9.2 154.62 -0.90 -4.16 9.131 4.964 04LPTA5 9 1812 8.5 155.47 -0.82 -4.40 9.090 4.953 04LPTA5 10 1811 8.6 156.33 -0.72 -4.16 9.129 4.982

PAGE 156

141 Appendix 1 (Continued) 04LPTA5 11 1810 10 157.33 -1.25 -4.60 9.075 4.963 -59.0 2.9 04LPTA5 12 1809 8 158.13 -0.91 -4.23 9.106 4.977 04LPTA5 13 1808 8 158.93 -0.46 -4.26 9.163 4.597 04LPTA5 14 1807 8 159.73 -0.86 -4.16 9.142 4.714 04LPTA5 15 1806 8.6 160.59 -0.76 -4.15 9.116 4.761 04LPTA5 16 1805 7.7 161.36 -1.13 -4.46 9.081 4.852 -57.3 3.6 04LPTA5 17 1804 8 162.16 -1.04 -4.23 9.142 4.778 04LPTA5 19 1803 8.9 163.05 -0.76 -4.07 9.127 4.927 04LPTA5 20 1802 7.5 163.8 -1.16 -3.90 9.166 5.112 04LPTA5 22 1801 6.5 164.45 -1.34 -4.20 9.119 5.116 04LPTA5 23 1800 5.6 165.01 -0.15 -3.86 9.199 4.775 -49.4 2.9 04LPTA5 24 1799 6.2 165.63 -0.52 -3.92 9.156 4.814 04LPTA5 25 1798 8.8 166.51 -0.97 -3.90 9.150 5.046 04LPTA5 26 1797 7.7 167.28 -0.70 -3.89 9.208 4.761 04LPTA5 27 1796 5.4 167.82 -0.73 -3.83 9.175 4.794 04LPTA5 28 1795 6.4 168.46 -1.42 -4.14 9.121 4.905 -49.2 4.2 04LPTA5 29 1794 7.4 169.2 -1.64 -4.19 9.080 5.032 04LPTA6 1 1793 6.7 169.87 -1.08 -4.01 9.219 4.445 04LPTA6 2 1792 6.7 170.54 -1.12 -3.86 9.158 4.639 04LPTA6 3 1791 6.2 171.16 -0.24 -3.70 9.212 4.386 04LPTA6 4 1790 6.2 171.78 -1.16 -3.99 9.136 4.693 -54.3 2.8 04LPTA6 5 1789 8 172.58 -1.57 -3.95 9.150 4.871 04LPTA6 6 1788 8.5 173.43 -0.75 -3.73 9.204 4.503 04LPTA6 7 1787 6.5 174.08 -0.71 -3.79 9.226 4.345 04LPTA6 8 1786 8 174.88 -1.37 -3.83 9.187 4.573 04LPTA7 1 1785 6.3 175.51 -1.06 -3.94 9.187 4.451 -50.1 3.5 04LPTA7 2 1784 8 176.31 -0.49 -3.96 9.219 4.383 04LPTA7 3 1783 6 176.91 -1.01 -3.86 9.177 4.894 04LPTA7 4 1782 6 177.51 -1.22 -4.05 9.144 4.535 04LPTA7 5 1781 6 178.11 -1.02 -3.87 9.200 4.319 04LPTA7 7 1780 9.4 179.05 -2.18 -4.01 9.155 4.256 -44.3 2.8 04LPTA7 8 1779 6.9 179.74 -1.78 -3.62 9.158 4.155 04LPTA7 9 1778 6.7 180.41 -0.47 -3.65 9.261 3.726 04LPTA7 10 1777 7.2 181.13 -1.40 -4.04 9.151 4.026 04LPTA7 11 1776 7.2 181.85 -1.69 -4.12 9.123 4.143 04LPTA7 12 1775 6.5 182.5 -2.31 -4.15 9.106 4.272 -45.4 3.0 04LPTA7 13 1774 8 183.3 -1.75 -4.38 9.072 4.352 04LPTA7 14 1773 8 184.1 -1.26 -4.16 9.187 3.952 04LPTA7 15 1772 6.6 184.76 -0.83 -4.00 9.115 4.137 04LPTA7 16 1771 6 185.36 -0.73 -4.13 9.120 4.508 04LPTA7 17 1770 9.3 186.29 -1.53 -4.04 9.225 4.832 -50.0 2.8 04LPTA7 18 1769 6.8 186.97 -0.43 -3.74 9.189 4.336 04LPTA7 20 1768 5.1 187.48 -0.64 -3.91 9.163 4.470 04LPTA7 21 1767 5 187.98 -1.24 -4.34 9.141 4.416 04LPTA7 22 1766 5.6 188.54 -1.47 -4.38 9.065 4.652 04LPTA7 23 1765 6.6 189.2 -0.97 -4.35 9.159 4.155 -53.0 2.8 04LPTA7 24 1764 7 189.9 -1.38 -4.30 9.151 4.103 04LPTA7 25 1763 8 190.7 -1.83 -3.99 9.113 4.446 04LPTA7 27 1762 9 191.6 -1.92 -4.08 9.102 4.479

PAGE 157

142 Appendix 1 (Continued) 04LPTA7 28 1761 5.8 192.18 -1.62 -4.22 9.125 4.171 04LPTA7 29 1760 5.7 192.75 -1.45 -3.96 9.178 3.876 -53.4 2.7 04LPTA7 30 1759 8 193.55 -1.46 -3.83 9.136 4.193 04LPTA7 31 1758 10 194.55 -1.62 -4.01 9.167 4.107 04LPTA7 32 1757 6.6 195.21 -1.76 -3.95 9.100 4.340 04LPTA7 33 1756 8 196.01 -1.25 -3.95 9.145 4.024 04LPTA7 34 1755 9.3 196.94 -1.39 -3.86 9.166 4.123 -47.6 2.8 04LPTA7 35 1754 8.7 197.81 -1.69 -4.06 9.150 4.105 04LPTA7 36 1753 9.5 198.76 -2.32 -4.23 9.114 4.473 04LPTA7 37 1752 7.2 199.48 -1.56 -3.97 9.140 4.285 04LPTA7 38 1751 6 200.08 -1.81 -3.86 9.196 4.176 -50.4 2.8

PAGE 158

143 Appendix 2: High Resolution Isotope Data sample ident 1 sample ident 2 Date depth 13 C PDB 18 O PDB MM/DD/YY (cm) ( VPDB) ( VPDB) 04LPTA1-1998 1 07/09/99 3.966 -1.22 -4.52 04LPTA1-1998 2 07/02/99 3.986 04LPTA1-1998 3 06/25/99 4.006 04LPTA1 1998 4 06/18/99 4.026 -1.34 -4.68 04LPTA1-1998 5 06/11/99 4.046 -1.10 -4.27 04LPTA1-1998 6 06/04/99 4.066 -1.03 -4.32 04LPTA1-1998 7 05/28/99 4.086 -1.48 -4.55 04LPTA1-1998 8 05/21/99 4.106 -1.45 -4.54 04LPTA1-1998 9 05/14/99 4.126 0.07 -4.08 04LPTA1-1998 10 05/07/99 4.146 -1.52 -4.40 04LPTA1-1998 11 04/30/99 4.166 -1.61 -4.14 04LPTA1-1998 12 04/23/99 4.186 -1.19 -4.34 04LPTA1-1998 13 04/16/99 4.206 -1.72 -4.32 04LPTA1-1998 14 04/09/99 4.226 -0.69 -3.97 04LPTA1-1998 15 04/02/99 4.246 -1.56 -4.20 04LPTA1-1998 16 03/26/99 4.266 -1.47 -4.26 04LPTA1-1998 17 03/19/99 4.286 -1.45 -4.26 04LPTA1-1998 18 03/12/99 4.306 -1.16 -4.08 04LPTA1-1998 19 03/05/99 4.326 -2.09 -4.26 04LPTA1-1998 20 02/26/99 4.346 -1.36 -4.05 04LPTA1-1998 21 02/19/99 4.366 -1.28 -4.05 04LPTA1-1998 22 02/12/99 4.386 -1.72 -4.12 04LPTA1-1998 23 02/05/99 4.406 04LPTA1-1998 24 01/29/99 4.426 -2.09 -4.12 04LPTA1-1998 25 01/22/99 4.446 -1.45 -4.05 04LPTA1-1998 26 01/15/99 4.466 -1.69 -4.25 04LPTA1-1998 27 01/08/99 4.486 04LPTA1-1998 28 01/01/99 4.506 -1.42 -4.11 04LPTA1-1998 29 12/25/98 4.526 -1.68 -4.20 04LPTA1-1998 30 12/18/98 4.546 -1.81 -4.04 04LPTA1-1998 31 12/11/98 4.566 -1.35 -4.34 04LPTA1-1998 32 12/04/98 4.586 -1.96 -4.23 04LPTA1-1998 33 11/27/98 4.606 -1.84 -4.22 04LPTA1-1998 34 11/20/98 4.626 04LPTA1-1998 35 11/13/98 4.646 04LPTA1-1998 36 11/06/98 4.666 -1.58 -4.28 04LPTA1-1998 37 10/30/98 4.686 -0.91 -4.25 04LPTA1-1998 38 10/23/98 4.706 -0.86 -4.28 04LPTA1-1998 39 10/16/98 4.726 -1.63 -4.46 04LPTA1-1998 40 10/09/98 4.746 -1.01 -4.28 04LPTA1-1998 41 10/02/98 4.766 -0.67 -4.32 04LPTA1-1998 42 09/25/98 4.786 -1.76 -4.67 04LPTA1-1998 43 09/18/98 4.806 -0.02 -4.15

PAGE 159

144 Appendix 2 (Continued) 04LPTA1-1998 44 09/11/98 4.826 -0.98 -4.27 04LPTA1-1998 45 09/04/98 4.846 -0.62 -4.19 04LPTA1-1998 46 08/28/98 4.866 0.30 -3.90 04LPTA1-1998 47 08/21/98 4.886 -0.55 -4.40 04LPTA1-1998 48 08/14/98 4.906 0.07 -3.79 04LPTA1-1998 49 08/07/98 4.926 -0.76 -4.55 04LPTA1-1998 50 07/31/98 4.946 -0.66 -4.34 04LPTA1-1998 51 07/24/98 4.966 -0.62 -4.29 04LPTA1-1998 52 07/17/98 4.986 -0.85 -4.17 04LPTA1-1998 53 07/10/98 5.006 -1.24 -4.29 04LPTA1-1998 54 07/03/98 5.026 -1.22 -4.11 04LPTA1-1998 55 06/26/98 5.046 -1.90 -4.24 04LPTA1-1998 56 06/19/98 5.066 -1.41 -4.02 04LPTA1-1998 57 06/12/98 5.086 -0.92 -4.11 04LPTA1-1998 58 06/05/98 5.106 -1.88 -3.86 04LPTA1-1998 59 05/29/98 5.126 -1.87 -4.38 04LPTA1-1998 60 05/22/98 5.146 -1.34 -4.17 04LPTA1-1998 61 05/15/98 5.166 -1.48 -4.24 04LPTA1-1998 62 05/08/98 5.186 -2.27 -4.32 04LPTA1-1998 63 05/01/98 5.206 -2.40 -4.47 04LPTA1-1998 64 04/24/98 5.226 -2.58 -4.87 04LPTA1-1998 65 04/17/98 5.246 -1.84 -4.29 04LPTA1-1998 66 04/10/98 5.266 -1.94 -4.22 04LPTA1-1998 67 04/03/98 5.286 -1.71 -4.25

PAGE 160

145 Appendix 3: Modern Monthly Data 2004-1993 C.E. Core name sample # 04LPTA1 Date 04LPTA1 Depth 04LPTA1 13 C 04LPTA1 18 O 04LPTA1 Sr/Ca 04LPTA1 Mg/Ca (Years C.E.) (cm) ( VPDB) ( VPDB) (mmol/mol) (mmol/mol) 04LPTA1 1 2004.62 0.033 -2.00 -4.91 9.11 5.05 04LPTA1 2 2004.54 0.100 -1.39 -4.49 9.06 5.34 04LPTA1 3 2004.45 0.167 -1.24 -4.46 04LPTA1 4 2004.37 0.233 -0.68 -4.02 9.18 8.65 04LPTA1 5 2004.29 0.300 -1.37 -4.13 9.12 21.26 04LPTA1 6 2004.20 0.367 -1.56 -4.21 9.06 8.26 04LPTA1 7 2004.12 0.433 -1.96 -3.79 9.07 9.01 04LPTA1 8 2004.05 0.500 -1.76 -3.83 9.08 8.29 04LPTA1 9 2003.99 0.567 -1.35 -4.15 9.09 8.95 04LPTA1 10 2003.92 0.633 -1.32 -4.20 9.14 9.12 04LPTA1 11 2003.85 0.700 -1.37 -4.80 9.17 17.45 04LPTA1 12 2003.79 0.767 -2.16 -4.89 9.08 10.85 04LPTA1 13 2003.72 0.833 -1.87 -4.64 9.22 6.37 04LPTA1 14 2003.65 0.900 -2.46 -4.99 9.07 7.30 04LPTA1 15 2003.59 0.967 -1.99 -4.73 9.03 6.25 04LPTA1 16 2003.52 1.033 -2.58 -4.83 9.05 6.81 04LPTA1 17 2003.45 1.100 -2.82 -4.52 9.01 6.89 04LPTA1 18 2003.39 1.167 -1.90 -4.46 8.98 8.18 04LPTA1 19 2003.32 1.233 -1.57 -4.32 9.13 10.19 04LPTA1 20 2003.25 1.300 -1.57 -4.05 8.96 19.97 04LPTA1 21 2003.19 1.367 -1.94 -4.02 9.02 8.96 04LPTA1 22 2003.12 1.433 -1.92 -3.83 9.03 29.26 04LPTA1 23 2003.06 1.500 -1.84 -4.11 8.98 41.49 04LPTA1 24 2003.01 1.567 -1.98 -4.01 8.93 44.39 04LPTA1 25 2002.95 1.633 -1.43 -4.13 9.10 27.15 04LPTA1 26 2002.90 1.700 -2.15 -4.33 9.11 11.25 04LPTA1 27 2002.84 1.766 -1.65 -4.43 9.13 8.24 04LPTA1 28 2002.79 1.833 -2.15 -4.44 9.03 7.89 04LPTA1 29 2002.73 1.900 -2.89 -4.65 9.12 6.12 04LPTA1 30 2002.68 1.966 -2.78 -4.60 8.97 7.52 04LPTA1 31 2002.62 2.033 -2.34 -4.91 9.01 6.69 04LPTA1 32 2002.52 2.100 -1.64 -4.31 9.08 7.89 04LPTA1 33 2002.41 2.166 -2.00 -4.67 8.98 11.16 04LPTA1 34 2002.31 2.233 -1.80 -4.40 9.08 16.07 04LPTA1 35 2002.21 2.300 -1.4 -4.26 9.04 19.92 04LPTA1 36 2002.06 2.366 -1.9 -4.32 8.99 8.04 04LPTA1 37 2001.92 2.433 -2.2 -4.33 8.98 11.34 04LPTA1 38 2001.77 2.500 -1.96 -4.57 9.00 13.72 04LPTA1 39 2001.62 2.566 -1.27 -4.71 04LPTA1 40 2001.52 2.633 -1.48 -4.46 9.20 16.41 04LPTA1 41 2001.42 2.700 -2.12 -4.41 9.06 9.51 04LPTA1 42 2001.32 2.766 -3.07 -4.24 9.08 9.16 04LPTA1 43 2001.22 2.833 -4.08 -4.07 9.09 9.25 04LPTA1 44 2001.12 2.900 -3.65 -4.04 9.01 7.57 04LPTA1 45 2001.05 2.966 -3.67 -4.03 8.99 10.92 04LPTA1 46 2001.03 3.033

PAGE 161

146 Appendix 3 (Continued) 04LPTA1 47 2001.01 3.100 -2.83 -4.27 9.02 8.73 04LPTA1 48 2000.97 3.166 -3.03 -4.36 8.96 7.77 04LPTA1 49 2000.94 3.233 -2.04 -4.15 8.94 6.57 04LPTA1 50 2000.87 3.300 -2.37 -4.12 8.89 11.74 04LPTA1 51 2000.85 3.366 8.99 8.93 04LPTA1 52 2000.83 3.433 -1.61 -4.62 8.98 7.55 04LPTA1 53 2000.79 3.500 -1.42 -4.63 9.05 10.84 04LPTA1 54 2000.76 3.566 -1.61 -4.53 9.15 11.74 04LPTA1 55 2000.72 3.633 -0.44 -4.77 9.20 7.93 04LPTA1 56 2000.69 3.700 -1.67 -4.55 9.11 20.43 04LPTA1 57 2000.65 3.766 -1.73 -4.65 9.26 13.89 04LPTA1 58 2000.58 3.833 -1.72 -4.87 9.15 6.96 04LPTA1 59 2000.56 3.900 9.14 6.36 04LPTA1 60 2000.54 3.966 -1.31 -4.91 9.06 5.78 04LPTA1 61 2000.47 4.033 -1.53 -4.81 9.16 6.08 04LPTA1 62 2000.40 4.100 -0.71 -4.62 9.08 5.80 04LPTA1 63 2000.33 4.166 -1.16 -4.17 9.07 7.07 04LPTA1 64 2000.26 4.233 -0.73 -4.24 9.14 7.12 04LPTA1 65 2000.19 4.300 -1.86 -3.99 9.10 8.22 04LPTA1 66 2000.12 4.366 -1.67 -4.00 9.15 6.55 04LPTA1 67 2000.04 4.433 -2.17 -4.23 04LPTA1 68 1999.95 4.500 -2.34 -4.46 9.15 7.94 04LPTA1 69 1999.87 4.566 -1.99 -4.60 9.15 10.59 04LPTA1 70 1999.83 4.633 -1.80 -4.32 9.03 6.84 04LPTA1 71 1999.79 4.700 -1.94 -4.45 9.02 8.54 04LPTA1 72 1999.75 4.766 -1.68 -4.75 8.96 7.34 04LPTA1 73 1999.71 4.833 -1.67 -5.08 9.01 5.43 04LPTA1 74 1999.63 4.900 -1.21 -4.79 9.10 4.96 04LPTA1 75 1999.54 4.966 -1.42 -4.72 9.12 6.51 04LPTA1 76 1999.46 5.033 -2.02 -4.63 9.07 9.06 04LPTA1 77 1999.37 5.099 -2.49 -4.60 04LPTA1 78 1999.29 5.166 -2.61 -4.51 9.12 9.42 04LPTA1 79 1999.20 5.233 -2.64 -4.23 9.01 9.47 04LPTA1 80 1999.12 5.299 -2.86 -4.13 9.02 6.92 04LPTA1 81 1998.95 5.366 -2.73 -4.47 9.01 10.11 04LPTA1 82 1998.79 5.433 -3.18 -4.55 9.03 9.00 04LPTA1 83 1998.62 5.499 -2.67 -4.59 8.98 6.07 04LPTA1 84 1998.56 5.566 -1.64 -4.46 9.06 6.45 04LPTA1 85 1998.50 5.633 -1.00 -4.33 9.04 6.85 04LPTA1 86 1998.43 5.699 -1.92 -4.39 9.02 6.27 04LPTA1 87 1998.37 5.766 -1.74 -4.39 8.99 7.55 04LPTA1 88 1998.31 5.833 -1.86 -4.32 9.02 7.25 04LPTA1 89 1998.24 5.899 -1.70 -4.16 9.04 6.16 04LPTA1 90 1998.18 5.966 -1.45 -4.24 9.13 5.43 04LPTA1 91 1998.12 6.033 -1.83 -4.00 9.08 6.40 04LPTA1 92 1997.98 6.099 -2.37 -4.28 9.12 9.94 04LPTA1 93 1997.85 6.166 -3.02 -4.23 9.00 6.47 04LPTA1 94 1997.71 6.233 -3.41 -4.48 04LPTA1 95 1997.59 6.299 -3.39 -4.45 9.04 6.97 04LPTA1 96 1997.47 6.366 -3.34 -4.40 8.95 7.46

PAGE 162

147 Appendix 3 (Continued) 04LPTA1 97 1997.36 6.433 -2.50 -4.14 9.01 9.85 04LPTA1 98 1997.24 6.499 -2.15 -4.25 8.98 10.50 04LPTA1 99 1997.12 6.566 -2.00 -3.80 8.96 9.08 04LPTA1 100 1997.06 6.633 -2.03 -4.09 8.97 16.54 04LPTA1 101 1996.99 6.699 -2.00 -4.28 9.00 8.91 04LPTA1 102 1996.93 6.766 -1.92 -4.33 9.06 10.72 04LPTA1 103 1996.87 6.833 -0.40 -4.92 9.11 12.23 04LPTA1 104 1996.81 6.899 -1.10 -4.89 9.02 8.82 04LPTA1 105 1996.68 6.966 -1.65 -4.92 8.97 5.71 04LPTA1 106 1996.65 7.033 9.09 5.72 04LPTA1 107 1996.62 7.099 -1.81 -5.01 9.20 6.04 04LPTA1 108 1996.57 7.166 -0.75 -4.98 9.04 6.85 04LPTA1 109 1996.51 7.233 -1.30 -4.85 9.21 10.97 04LPTA1 110 1996.46 7.299 -0.07 -4.74 9.12 8.91 04LPTA1 111 1996.36 7.366 -1.61 -4.52 9.00 15.60 04LPTA1 112 1996.26 7.433 -1.50 -4.38 9.12 7.34 04LPTA1 113 1996.16 7.499 -1.69 -4.13 9.00 5.20 04LPTA1 114 1996.07 7.566 -2.00 -4.18 9.05 7.14 04LPTA1 115 1995.97 7.633 -1.52 -4.41 9.03 6.95 04LPTA1 116 1995.87 7.699 -1.20 -4.70 9.06 7.19 04LPTA1 117 1995.82 7.766 -1.66 -4.65 9.04 11.56 04LPTA1 118 1995.77 7.833 -1.66 -4.64 9.07 6.17 04LPTA1 119 1995.72 7.899 -2.13 -4.59 9.06 9.24 04LPTA1 120 1995.67 7.966 -2.65 -4.75 9.10 8.40 04LPTA1 121 1995.46 8.033 -1.63 -4.43 9.10 12.54 04LPTA1 122 1995.41 8.099 8.95 6.22 04LPTA1 123 1995.36 8.166 04LPTA1 124 1995.30 8.233 8.99 11.45 04LPTA1 125 1995.36 8.299 -3.21 -4.62 8.96 5.82 04LPTA1 126 1995.31 8.366 04LPTA1 127 1995.30 8.432 -1.70 -4.60 9.05 7.45 04LPTA1 128 1995.25 8.499 -1.53 -4.71 8.97 7.28 04LPTA1 129 1995.20 8.566 -1.40 -4.36 9.04 6.02 04LPTA1 130 1995.14 8.632 -1.23 -4.16 9.04 6.49 04LPTA1 131 1995.09 8.699 -1.19 -4.26 9.02 8.35 04LPTA1 132 1995.04 8.766 -0.75 -4.13 9.06 9.09 04LPTA1 133 1994.91 8.832 -0.84 -4.42 9.07 7.27 04LPTA1 134 1994.79 8.899 -0.30 -4.85 04LPTA1 135 1994.73 8.966 -0.70 -4.45 9.26 6.16 04LPTA1 136 1994.68 9.032 -1.41 -4.40 9.15 6.90 04LPTA1 137 1994.62 9.099 -1.34 -4.50 8.99 13.63 04LPTA1 138 1994.57 9.166 -1.83 -4.53 9.01 15.16 04LPTA1 139 1994.51 9.232 -1.94 -4.67 9.10 8.98 04LPTA1 140 1994.46 9.299 -2.07 -4.71 9.21 17.39 04LPTA1 141 1994.39 9.366 -2.39 -4.55 9.13 20.15 04LPTA1 142 1994.32 9.432 -1.80 -4.43 9.09 17.06 04LPTA1 143 1994.26 9.499 -2.64 -4.38 9.01 11.20 04LPTA1 144 1994.19 9.566 -1.92 -4.15 9.07 14.35 04LPTA1 145 1994.12 9.632 -2.36 -4.22 9.02 11.68 04LPTA1 146 1993.94 9.699 -2.14 -4.14 9.05 15.65

PAGE 163

148 Appendix 3 (Continued) 04LPTA1 147 1993.90 9.766 04LPTA1 148 1993.86 9.832 -1.43 -4.38 04LPTA1 149 1993.77 9.899 -0.24 -4.47 9.19 11.18 04LPTA1 150 1993.68 9.966 -0.45 -4.57 04LPTA1 151 1993.59 10.032 -0.56 -4.64 04LPTA1 152 1993.50 10.099 -1.60 -4.42 04LPTA1 153 1993.41 10.166 -2.08 -4.25 04LPTA1 154 1993.33 10.232 -2.27 -4.21 04LPTA1 155 1993.24 10.299 -2.66 -4.17 04LPTA1 156 1993.15 10.366 -2.27 -4.19 04LPTA1 157 1993.04 10.432 -2.91 -4.16

PAGE 164

149 Appendix 4: Monthly Data sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) 04LPTA3-1 0.033 1911.54 9.152 4.998 -1.64 -4.43 04LPTA3-2 0.100 1911.49 9.092 5.331 -0.61 -4.15 04LPTA3-3 0.166 1911.44 9.010 5.906 04LPTA3-4 0.233 1911.38 9.030 5.673 -0.94 -4.02 04LPTA3-5 0.300 1911.32 9.080 5.570 -0.87 -3.99 04LPTA3-6 0.366 1911.26 9.119 5.391 -0.89 -3.9 04LPTA3-7 0.433 1911.21 9.099 5.503 -0.49 -3.88 04LPTA3-8 0.500 1911.13 9.083 5.884 -0.74 -3.97 04LPTA3-9 0.566 1911.04 9.160 5.317 -0.98 -4 04LPTA3-10 0.633 1910.96 9.088 6.281 -1.67 -4.07 04LPTA3-11 0.700 1910.88 9.060 6.796 -1.83 -4.25 04LPTA3-12 0.766 1910.79 -1.99 -4.51 04LPTA3-13 0.833 1910.70 9.103 5.245 04LPTA3-14 0.900 1910.60 9.032 5.838 -1.31 -4.34 04LPTA3-15 0.966 1910.49 9.075 5.785 -1.31 -4.32 04LPTA3-16 1.033 1910.39 9.006 6.833 -0.78 -4.11 04LPTA3-17 1.100 1910.30 9.066 6.936 -0.62 -4.11 04LPTA3-18 1.166 1910.19 9.180 5.279 -0.78 -4.14 04LPTA3-19 1.233 1910.09 9.143 5.387 -1.18 -4.07 04LPTA3-20 1.300 1909.99 9.099 6.418 -1.41 -4.13 04LPTA3-21 1.366 1909.89 9.186 5.812 -2.14 -4.19 04LPTA3-22 1.433 1909.79 -1.91 -4.32 04LPTA3-23 1.500 1909.71 9.301 4.704 -1.05 -4.08 04LPTA3-24 1.566 1909.62 9.258 4.621 -0.84 -3.99 04LPTA3-25 1.633 1909.54 -1.27 -4.23 04LPTA3-26 1.700 1909.46 9.263 5.503 -0.93 -4.17 04LPTA3-27 1.766 1909.28 9.066 6.835 -0.71 -4.14 04LPTA3-28 1.833 1909.12 9.095 6.179 -0.93 -3.75 04LPTA3-29 1.899 1909.06 9.122 5.354 -1.17 -3.88 04LPTA3-30 1.966 1908.99 9.050 6.994 -1.26 -3.97 04LPTA3-31 2.033 1908.93 9.105 5.900 -1.81 -4.12 04LPTA3-32 2.099 1908.86 9.134 5.740 -1.7 -4.37 04LPTA3-33 2.166 1908.80 9.032 7.229 -1.96 -4.47 04LPTA3-34 2.233 1908.73 9.017 5.825 -2.59 -4.58 04LPTA3-35 2.299 1908.67 -2.64 -4.64 04LPTA3-36 2.366 1908.60 9.152 5.313 -0.87 -4.08 04LPTA3-37 2.433 1908.54 9.016 6.274 -0.63 -4 04LPTA3-38 2.499 1908.47 9.189 5.282 -0.93 -4.2 04LPTA3-39 2.566 1908.40 9.120 5.241 -0.61 -4.09 04LPTA3-40 2.633 1908.34 9.209 5.456 -0.52 -3.82 04LPTA3-41 2.699 1908.27 9.126 5.154 04LPTA3-42 2.766 1908.20 9.187 5.283 -1.36 -3.85 04LPTA3-43 2.833 1908.13 9.087 6.406 -0.55 -3.68 04LPTA3-44 2.899 1908.07 9.121 5.969 -2.08 -3.91 04LPTA3-45 2.966 1907.99 9.088 6.446 -2.19 -4.02 04LPTA3-46 3.033 1907.93 9.079 6.132 -1.87 -4.29 04LPTA3-47 3.099 1907.86 9.037 6.733 -2.02 -4.21

PAGE 165

150 Appendix 4 (Continued) 04LPTA3-48 3.166 1907.79 -1.27 -4.46 04LPTA3-49 3.233 1907.70 9.009 5.442 04LPTA3-50 3.299 1907.60 9.035 5.536 -0.97 -4.28 04LPTA3-51 3.366 1907.50 9.036 6.252 -1.01 -4.18 04LPTA3-52 3.433 1907.40 -1.07 -4.01 04LPTA3-53 3.499 1907.31 9.114 6.155 -0.89 -3.96 04LPTA3-54 3.566 1907.21 9.122 6.825 04LPTA3-55 3.633 1907.11 9.189 5.574 -1.57 -3.89 04LPTA3-56 3.699 1907.02 9.094 6.178 -0.18 -3.89 04LPTA3-57 3.766 1906.91 9.061 6.959 -2.16 -4.22 04LPTA3-58 3.833 1906.82 9.086 5.619 -2.36 -4.41 04LPTA3-59 3.899 1906.72 9.094 5.593 04LPTA3-60 3.966 1906.62 9.011 6.136 -2.16 -4.45 04LPTA3-61 4.033 1906.55 9.191 4.882 -0.93 -4.23 04LPTA3-62 4.099 1906.48 9.057 5.487 -0.71 -4.03 04LPTA3-63 4.166 1906.40 9.210 5.109 -0.48 -4.16 04LPTA3-64 4.233 1906.34 9.017 5.751 04LPTA3-65 4.299 1906.27 9.199 5.135 -0.21 -3.85 04LPTA3-66 4.366 1906.19 9.198 5.250 -0.48 -3.76 04LPTA3-67 4.433 1906.12 9.084 6.023 -0.41 -3.8 04LPTA3-68 4.499 1905.99 9.109 5.790 -1.23 -4.05 04LPTA3-69 4.566 1905.84 9.090 5.434 -1.67 -4.22 04LPTA3-70 4.633 1905.71 9.138 5.812 -2.09 -4.66 04LPTA3-71 4.699 1905.66 9.178 5.065 -2.3 -4.66 04LPTA3-72 4.766 1905.60 -1.35 -4.43 04LPTA3-73 4.833 1905.54 9.209 4.808 -1.82 -4.39 04LPTA3-74 4.899 1905.49 9.149 5.173 04LPTA3-75 4.966 1905.43 9.098 4.975 -0.67 -4.09 04LPTA3-76 5.033 1905.38 9.128 4.949 04LPTA3-77 5.099 1905.32 9.073 5.831 -0.3 -3.97 04LPTA3-78 5.166 1905.26 -0.53 -3.91 04LPTA3-79 5.232 1905.21 9.033 6.352 -0.62 -3.8 04LPTA3-80 5.299 1905.16 9.034 6.139 -1.8 -4.05 04LPTA3-81 5.366 1905.10 9.052 5.998 -0.79 -3.75 04LPTA3-82 5.432 1905.04 9.023 6.698 -0.75 -3.75 04LPTA3-83 5.499 1904.99 9.053 6.426 -0.75 -3.99 04LPTA3-84 5.566 1904.93 9.118 5.927 -1 -4.07 04LPTA3-85 5.632 1904.88 9.254 5.073 04LPTA3-86 5.699 1904.82 9.124 5.474 -0.74 -4.06 04LPTA3-87 5.766 1904.77 9.020 6.543 04LPTA3-88 5.832 1904.71 9.030 5.652 -1.24 -4.25 04LPTA3-89 5.899 1904.60 9.040 6.181 -1.05 -4.17 04LPTA3-90 5.966 1904.48 9.016 6.481 -0.22 -4.03 04LPTA3-91 6.032 1904.37 9.090 6.420 04LPTA3-92 6.099 1904.26 9.105 5.811 -0.69 -3.89 04LPTA3-93 6.166 1904.13 9.078 6.517 -0.98 -3.87 04LPTA3-94 6.232 1904.02 9.126 5.658 -1.38 -4.11 04LPTA3-95 6.299 1903.91 9.045 6.305 -1.72 -4.24 04LPTA3-96 6.366 1903.79 9.021 5.393 -2.04 -4.27 04LPTA3-97 6.432 1903.72 -1.46 -4.19

PAGE 166

151 Appendix 4 (Continued) 04LPTA3-98 6.499 1903.66 9.262 4.937 -0.86 -4.07 04LPTA3-99 6.566 1903.58 9.194 5.135 -0.78 -4.14 04LPTA3-100 6.632 1903.52 9.143 5.174 -1.2 -4.26 04LPTA3-101 6.699 1903.45 9.040 5.173 -0.75 -4.14 04LPTA3-102 6.766 1903.38 9.045 5.722 -0.94 -4.14 04LPTA3-103 6.832 1903.31 9.052 5.484 -0.47 -4.24 04LPTA3-104 6.899 1903.24 9.054 5.118 04LPTA3-105 6.966 1903.17 9.167 4.885 -1.08 -4.2 04LPTA3-106 7.032 1903.11 9.181 5.152 -0.99 -4.02 04LPTA3-107 7.099 1903.04 9.128 6.721 -1.64 -3.97 04LPTA3-108 7.166 1902.95 -2.04 -4.28 04LPTA3-109 7.232 1902.87 9.113 6.091 -1.25 -4.3 04LPTA3-110 7.299 1902.79 9.151 6.464 -0.7 -4.22 04LPTA3-111 7.366 1902.70 -1.68 -4.41 04LPTA3-112 7.432 1902.62 9.225 5.768 -0.95 -4.05 04LPTA3-113 7.499 1902.54 9.197 5.126 -1.63 -4.27 04LPTA3-114 7.566 1902.45 9.130 6.401 -0.48 -4.05 04LPTA3-115 7.632 1902.37 9.119 5.609 -0.55 -4.08 04LPTA3-116 7.699 1902.29 9.160 5.407 -0.75 -3.94 04LPTA3-117 7.766 1902.21 9.133 5.808 04LPTA3-118 7.832 1902.12 9.084 6.400 -0.68 -3.86 04LPTA3-119 7.899 1902.04 9.189 4.695 -1.08 -3.63 04LPTA3-120 7.966 1901.95 9.005 5.879 -1.15 -3.91 04LPTA3-121 8.032 1901.87 9.104 5.170 -1.62 -4.26 04LPTA3-122 8.099 1901.79 9.096 5.340 -1.6 -4.29 sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) 04LPTA4 1 0.033 1856.27 9.146 5.022 -0.41 -4.05 04LPTA4 2 0.099 1856.21 9.096 5.350 0.24 -3.94 04LPTA4 3 0.165 1856.15 9.222 4.781 1.31 -3.56 04LPTA4 4 0.231 1856.1 9.183 4.189 -0.69 -3.93 04LPTA4 5 0.297 1856.04 9.263 4.791 0.56 -3.51 04LPTA4 6 0.363 1855.96 -1.02 -4 04LPTA4 7 0.429 1855.9 -1.06 -3.89 04LPTA4 8 0.495 1855.83 9.084 5.368 -0.27 -3.8 04LPTA4 9 0.561 1855.77 9.186 5.031 -2.17 -4.27 04LPTA4 10 0.627 1855.71 -2.78 -4.55 04LPTA4 11 0.693 1855.61 9.111 5.100 -1.17 -4.22 04LPTA4 12 0.759 1855.5 -0.17 -4.05 04LPTA4 13 0.825 1855.39 9.198 4.340 -1.06 -4.18 04LPTA4 14 0.891 1855.29 9.014 5.188 -1.05 -4.36 04LPTA4 15 0.957 1855.12 9.060 4.858 0.16 -3.86 04LPTA4 16 1.023 1855.05 -0.77 -4.11 04LPTA4 17 1.089 1854.99 -1.10 -4 04LPTA4 18 1.155 1854.92 9.172 4.567 -0.67 -3.94 04LPTA4 19 1.221 1854.86 9.165 4.578 -1.40 -3.99 04LPTA4 20 1.287 1854.79 9.097 5.079 -2.05 -4.43 04LPTA4 21 1.353 1854.68 9.225 4.588 -1.75 -4.38

PAGE 167

152 Appendix 4 (Continued) 04LPTA4 22 1.419 1854.57 8.960 5.795 -1.66 -4.37 04LPTA4 23 1.485 1854.45 9.162 4.542 -0.43 -4.25 04LPTA4 24 1.551 1854.34 9.150 4.582 -0.09 -3.99 04LPTA4 25 1.617 1854.23 9.363 4.051 0.35 -3.81 04LPTA4 26 1.683 1854.12 9.169 4.823 0.61 -3.57 04LPTA4 27 1.749 1854.04 9.145 4.791 -0.75 -3.99 04LPTA4 28 1.815 1853.96 9.205 4.634 -0.10 -3.76 04LPTA4 29 1.881 1853.87 9.183 4.899 -1.27 -4.08 04LPTA4 30 1.947 1853.79 9.252 4.385 -1.40 -4.34 04LPTA4 31 2.013 1853.71 9.285 4.140 -1.97 -4.65 04LPTA4 32 2.079 1853.58 9.190 4.603 -1.92 -4.56 04LPTA4 33 2.145 1853.46 9.153 5.010 -0.27 -4.1 04LPTA4 34 2.211 1853.34 9.190 4.794 -0.51 -4.25 04LPTA4 35 2.277 1853.21 9.296 4.056 0.91 -3.78 04LPTA4 36 2.343 1853.13 9.227 4.303 0.26 -3.98 04LPTA4 37 2.409 1853.04 9.159 4.865 -0.35 -4.04 04LPTA4 38 2.475 1852.96 9.083 5.544 -0.88 -3.97 04LPTA4 39 2.541 1852.84 -1.60 -4.08 04LPTA4 40 2.607 1852.71 9.195 4.465 -2.63 -4.9 04LPTA4 41 2.673 1852.59 -1.74 -4.75 04LPTA4 42 2.739 1852.46 0.26 -4.24 04LPTA4 43 2.805 1852.33 9.076 5.305 -0.84 -4.31 04LPTA4 44 2.871 1852.21 9.231 4.618 1.81 -3.58 04LPTA4 45 2.937 1852.12 9.311 4.073 0.57 -3.68 04LPTA4 46 3.003 1852.04 8.988 4.760 0.88 -3.61 04LPTA4 47 3.069 1851.98 9.354 4.210 -0.63 -4 04LPTA4 48 3.135 1851.92 9.127 5.029 -0.16 -3.85 04LPTA4 49 3.201 9.143 5.067 04LPTA4 50 3.267 1851.79 9.195 4.674 -1.67 -4.22 sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) 04LPTA7 51 3.333 1760.04 9.243 4.159 1.03 -3.42 04LPTA7 52 3.399 1759.98 9.193 3.809 0.31 -3.77 04LPTA7 53 3.465 1759.93 9.306 3.156 0.2 -3.67 04LPTA7 54 3.531 1759.88 9.183 3.647 -0.77 -3.9 04LPTA7 55 3.597 1759.82 9.142 2.757 -1.44 -3.82 04LPTA7 56 3.663 1759.76 9.118 3.605 -1.12 -3.84 04LPTA7 57 3.729 1759.71 9.166 4.067 -1.96 -4.12 04LPTA7 58 3.795 1759.61 9.198 3.948 04LPTA7 59 3.861 1759.51 9.200 6.587 -1.35 -3.97 04LPTA7 60 3.927 1759.41 9.186 3.980 0.3 -3.75 04LPTA7 61 3.993 1759.31 9.339 3.932 -0.17 -3.76 04LPTA7 62 4.059 1759.21 9.168 4.282 -0.99 -3.62 04LPTA7 63 4.125 1759.12 9.218 4.108 0.1 -3.53 04LPTA7 64 4.191 1759.05 9.187 4.213 -0.76 -3.65 04LPTA7 65 4.257 1758.99 9.113 4.314 -1.54 -3.86 04LPTA7 66 4.323 1758.93 9.098 4.272 -1.49 -3.86 04LPTA7 67 4.389 1758.87 9.171 4.004 -1.85 -3.82

PAGE 168

153 Appendix 4 (Continued) 04LPTA7 68 4.455 1758.81 9.185 4.085 -2.04 -3.85 04LPTA7 69 4.521 1758.75 9.206 4.072 -0.85 -3.84 04LPTA7 70 4.587 1758.68 9.282 3.740 -2.1 -4.21 04LPTA7 71 4.653 1758.62 9.193 4.486 -2.15 -4.26 04LPTA7 72 4.719 1758.56 9.093 4.426 -1.21 -4.12 04LPTA7 73 4.785 1758.49 9.128 4.259 -1.07 -4.03 04LPTA7 74 4.851 1758.43 9.074 4.399 04LPTA7 75 4.917 1758.37 9.121 4.284 -1.43 -4.08 04LPTA7 76 4.983 1758.31 9.195 3.901 -1.39 -4.03 04LPTA7 77 5.049 1758.25 9.036 4.466 -1.23 -3.72 04LPTA7 78 5.115 1758.19 9.168 4.275 -1.63 -3.62 04LPTA7 79 5.181 1758.12 9.191 4.158 -2.28 -3.96 04LPTA7 80 5.247 1758.02 9.194 4.078 -2.83 -3.92 04LPTA7 81 5.313 1757.92 9.124 4.222 -2.36 -3.71 04LPTA7 82 5.379 1757.81 9.150 4.039 -2.65 -3.79 04LPTA7 83 5.445 1757.71 9.135 4.209 -3.28 -4.29 04LPTA7 84 5.511 1757.62 9.024 4.663 -3.36 -4.26 04LPTA7 85 5.577 1757.51 9.071 4.257 -1.8 -4.01 04LPTA7 86 5.643 1757.41 9.075 4.440 -1.73 -4.06 04LPTA7 87 5.709 1757.32 9.039 4.384 -1.46 -4.03 04LPTA7 88 5.775 1757.21 9.163 15.885 -1.26 -3.83 04LPTA7 89 5.841 1757.12 9.091 4.347 0.54 -3.28 04LPTA7 90 5.907 1757.03 9.348 3.879 -0.91 -3.5 04LPTA7 91 5.973 1756.94 9.275 3.703 -1.96 -3.62 04LPTA7 92 6.039 1756.85 9.145 4.347 -1.34 -3.66 04LPTA7 93 6.105 1756.76 9.188 4.092 -3.05 -4.07 04LPTA7 94 6.171 1756.67 9.194 4.395 04LPTA7 95 6.237 1756.58 9.032 4.624 -2.23 -4.08 04LPTA7 96 6.303 1756.49 9.346 3.508 -1.56 -3.93 04LPTA7 97 6.369 1756.39 9.136 3.925 -1.77 -4.04 04LPTA7 98 6.435 1756.3 9.134 4.308 -0.69 -3.94 04LPTA7 99 6.501 1756.22 9.134 4.468 -1.08 -3.86 04LPTA7 100 6.567 1756.12 9.209 4.022 -1.68 -3.79 04LPTA7 101 6.633 1756.02 9.146 4.162 -3.14 -4.16 04LPTA7 102 6.699 1755.93 9.062 4.092 -2.64 -3.95 04LPTA7 103 6.765 1755.82 -2.9 -4.15 04LPTA7 104 6.831 1755.72 9.100 3.990 -2.23 -3.93 04LPTA7 105 6.897 1755.63 9.224 3.612 -1.34 -3.86 04LPTA7 106 6.963 1755.52 -1.46 -4.07 04LPTA7 107 7.029 1755.42 9.010 4.426 -1.72 -4.09 04LPTA7 108 7.095 1755.32 9.044 4.201 -1.17 -4.01 04LPTA7 109 7.161 1755.22 9.067 4.522 -1.05 -3.93 04LPTA7 110 7.227 1755.12 9.226 4.133 -0.67 -3.56 04LPTA7 111 7.293 1754.93 9.261 3.969 -1.06 -3.6 04LPTA7 112 7.359 1754.71 9.135 4.310 -2.66 -4.06 04LPTA7 113 7.425 1754.58 9.230 4.249 -1.39 -3.92 04LPTA7 114 7.491 1754.45 9.138 4.063 -1.8 -3.91 04LPTA7 115 7.557 1754.32 9.076 4.326 -2.07 -3.86 04LPTA7 116 7.623 1754.18 9.116 4.288 04LPTA7 117 7.689 1754.04 9.033 4.427 -0.74 -3.6

PAGE 169

154 Appendix 4 (Continued) 04LPTA7 118 7.755 1753.88 9.104 4.040 -2.64 -4.39 04LPTA7 119 7.821 1753.71 9.124 3.811 -2.51 -4.26 04LPTA7 120 7.887 1753.63 9.136 3.985 -1.8 -4.26 04LPTA7 121 7.953 1753.55 9.131 3.956 -2.19 -4.27 04LPTA7 122 8.019 1753.45 9.214 3.162 -0.63 -3.97 04LPTA7 123 8.085 1753.37 9.052 4.217 -1.78 -4.17 04LPTA7 124 8.151 1753.31 8.971 4.330 0.31 -3.73 04LPTA7 125 8.217 1753.24 9.073 4.147 -0.55 -3.93 04LPTA7 126 8.283 1753.17 9.130 4.058 -0.39 -3.5 04LPTA7 127 8.349 1753.11 9.291 3.747 -0.17 -3.21 04LPTA7 128 8.415 1753.04 9.214 3.904 -0.16 -3.24 04LPTA7 129 8.481 1752.98 9.190 4.153 -2.01 -3.75 04LPTA7 130 8.547 1752.91 9.173 3.903 -2.5 -4.15 04LPTA7 131 8.613 1752.84 9.043 4.582 -2.81 -4.27 04LPTA7 132 8.679 1752.77 9.060 4.711 -2.44 -4.16 04LPTA7 133 8.745 1752.71 9.196 3.651 -2.83 -4.4 04LPTA7 134 8.811 1752.62 9.163 3.789 -2.65 -4.39 04LPTA7 135 8.877 1752.54 9.058 4.199 -0.25 -3.87 04LPTA7 136 8.943 1752.46 9.211 3.916 -1.51 -4.14 04LPTA7 137 9.009 1752.37 9.127 4.208 -1.81 -4.09 04LPTA7 138 9.075 1752.29 -1.11 -3.76 04LPTA7 139 9.141 1752.21 9.272 3.955 0.17 -3.41 04LPTA7 140 9.207 1752.12 9.163 4.398 0.19 -3.27 04LPTA7 141 9.273 1752.05 9.127 4.748 -1.72 -3.8 04LPTA7 142 9.339 1751.99 9.189 5.642 -1.76 -3.55 04LPTA7 143 9.405 1751.91 9.298 4.088 -2.14 -3.86 04LPTA7 144 9.471 1751.85 9.093 4.469 -2.67 -4.24 04LPTA7 145 9.537 1751.78 9.018 4.962 -2.89 -4.3 04LPTA7 146 9.603 1751.71 9.028 4.549 -2.55 -4.39 04LPTA7 147 9.669 1751.64 9.188 4.147 -2.5 -4.22 04LPTA7 148 9.735 1751.57 9.128 3.926 -1.87 -4.17 04LPTA7 149 9.801 1751.49 9.094 4.240 04LPTA7 150 9.867 1751.41 9.038 4.625 -0.73 -3.98 04LPTA7 151 9.933 1751.34 9.238 4.104 -1 -3.89 04LPTA7 152 9.999 1751.26 9.189 4.512 -1.79 -3.99 04LPTA7 153 10.065 1751.19 9.110 4.700 04LPTA7 154 10.131 1751.12 9.150 4.540 -2.18 -3.74 04LPTA7 155 10.197 1750.99 9.240 4.050 -2.54 -4.08 04LPTA7 156 10.263 1750.87 -3.34 -4.41 04LPTA7 157 10.329 1750.78 9.050 4.790 -1.9 -4.07 04LPTA7 158 10.395 1750.68 -2.39 -4.17 04LPTA7 159 10.461 1750.59 9.050 4.590 -2.22 -4.05 04LPTA7 160 10.527 1750.5 9.070 5.050 -1.99 -4.34 04LPTA7 161 10.593 1750.41 9.180 4.050 -1.3 -3.87 04LPTA7 162 10.659 1750.32 -1.95 -4.17 04LPTA7 163 10.725 1750.23 9.090 4.760 -1.73 -4.07 04LPTA7 164 10.791 1750.135 9.060 4.590 04LPTA7 165 10.857 1750.04 9.100 4.400 -1.39 -3.89

PAGE 170

155 Appendix 4 (Continued) sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP B1 1 0.033 1672.85 9.410 3.780 -0.37 -3.81 LP B1 2 0.099 1672.80 -0.46 -4.20 LP B1 3 0.165 1672.75 9.350 4.050 -0.04 -4.05 LP B1 4 0.231 1672.70 9.250 4.030 -0.33 -4.27 LP B1 5 0.297 1672.65 9.330 4.050 0.14 -3.71 LP B1 6 0.363 1672.60 9.190 5.520 0.18 -3.98 LP B1 7 0.429 1672.56 9.368 4.338 0.23 -3.83 LP B1 8 0.495 1672.51 9.381 4.065 -0.33 -4.03 LP B1 9 0.561 1672.46 0.31 -3.63 LP B1 10 0.627 1672.41 9.180 3.880 0.24 -3.85 LP B1 11 0.693 1672.36 9.160 4.080 0.22 -3.61 LP B1 12 0.759 1672.31 9.400 4.150 -0.05 -3.79 LP B1 13 0.825 1672.26 0.31 -3.72 LP B1 14 0.891 1672.22 0.32 -3.37 LP B1 15 0.957 1672.17 9.240 4.150 0.34 -3.53 LP B1 16 1.023 1672.12 9.300 3.640 1.21 -3.11 LP B1 17 1.089 1672.07 9.240 4.180 0.88 -3.34 LP B1 18 1.155 1672.01 LP B1 19 1.221 1671.95 9.230 4.250 0.01 -3.63 LP B1 20 1.287 1671.90 9.240 3.990 -0.56 -3.81 LP B1 21 1.353 1671.84 9.260 4.190 -0.72 -4.04 LP B1 22 1.419 1671.78 9.260 4.400 -1.07 -4.41 LP B1 23 1.485 1671.73 9.200 4.080 -1.17 -4.34 LP B1 24 1.551 1671.67 9.090 4.040 -1.35 -4.42 LP B1 25 1.617 1671.62 9.230 3.900 -0.51 -4.07 LP B1 26 1.683 1671.56 9.250 3.970 -0.15 -4.25 LP B1 27 1.749 1671.50 9.120 5.030 -0.20 -4.11 LP B1 28 1.815 1671.45 9.170 4.350 -0.30 -4.31 LP B1 29 1.881 1671.37 9.260 4.630 -0.10 -4.11 LP B1 30 1.947 1671.28 9.360 4.320 -0.34 -3.97 LP B1 31 2.013 1671.20 0.00 -3.78 LP B1 32 2.079 1671.12 9.330 4.450 -0.24 -3.49 LP B1 33 2.145 1670.95 9.190 4.640 -0.72 -4.01 LP B1 34 2.211 1670.82 9.280 4.580 -0.99 -3.94 LP B1 35 2.277 1670.69 9.270 4.040 -1.06 -3.96 LP B1 36 2.343 1670.54 9.120 5.610 -1.24 -4.30 LP B1 37 2.409 1670.37 9.240 4.220 -1.89 -4.31 LP B1 38 2.475 1670.12 9.490 3.770 0.62 -3.49 LP B1 39 2.541 1669.95 9.450 3.740 0.31 -3.70 LP B1 40 2.607 1669.81 9.350 3.920 0.19 -3.83 LP B1 41 2.673 1669.66 9.290 4.130 0.22 -3.83 LP B1 42 2.739 1669.49 9.270 4.100 0.12 -3.86 LP B1 43 2.805 1669.35 9.190 7.160 -0.19 -3.87 LP B1 44 2.871 1669.20 9.280 4.000 0.11 -3.66 LP B1 45 2.937 1669.12 9.280 4.350 -0.03 -3.70 LP B1 46 3.003 1669.04 9.290 4.050 -0.39 -3.69 LP B1 47 3.069 1668.93 9.270 4.190 -0.19 -4.00 LP B1 48 3.135 1668.81 9.187 4.611 -0.34 -4.04

PAGE 171

156 Appendix 4 (Continued) LP B1 49 3.201 1668.70 -0.78 -4.26 LPB1 50 3.267 1668.62 9.180 4.180 -0.14 -4.17 LP B1 51 3.333 9.140 4.320 LP B1 52 3.399 1668.46 9.210 4.170 -0.86 -4.19 LP B1 53 3.465 1668.38 9.170 4.280 -0.69 -4.16 LP B1 54 3.531 1668.29 9.230 4.440 -0.59 -4.09 LP B1 55 3.597 1668.12 9.310 4.280 -0.06 -3.81 LP B1 56 3.663 1668.04 9.230 4.170 -0.02 -3.72 LP B1 57 3.729 1667.97 9.340 4.380 -0.20 -3.79 LP B1 58 3.795 1667.91 9.320 4.170 -0.33 -3.90 LP B1 59 3.861 1667.86 9.449 3.833 -0.38 -4.02 LP B1 60 3.927 1667.79 9.370 3.870 -0.89 -4.27 LP B1 61 3.993 1667.71 9.270 4.440 -0.68 -4.27 LP B1 62 4.059 1667.63 9.430 3.870 -0.07 -4.15 LP B1 63 4.125 1667.54 9.221 4.063 -0.12 -4.31 LP B1 64 4.191 1667.46 9.400 3.870 -0.11 -4.24 LP B1 65 4.257 1667.38 9.190 4.370 0.43 -3.98 LP B1 66 4.323 1667.29 9.280 4.210 0.07 -3.93 LP B1 67 4.389 1667.12 9.260 4.220 0.48 -3.58 LP B1 68 4.455 1667.05 9.270 4.300 -0.18 -3.89 LP B1 69 4.521 1666.98 9.220 4.210 -0.51 -4.10 LP B1 70 4.587 1666.91 9.205 4.291 -0.37 -4.02 LP B1 71 4.653 1666.85 9.170 4.320 -0.65 -4.40 LP B1 72 4.719 1666.78 4.270 -0.90 -4.47 LP B1 73 4.785 1666.71 9.180 3.970 -0.69 -4.60 LP B1 74 4.851 1666.62 9.140 4.010 -0.29 -4.47 LP B1 75 4.917 1666.51 9.090 4.430 -0.29 -4.38 LP B1 76 4.983 1666.41 9.150 4.210 0.05 -4.23 LP B1 77 5.049 1666.32 9.180 4.280 0.02 -4.10 LP B1 78 5.115 1666.21 9.150 4.350 -0.18 -3.99 LP B1 79 5.181 1666.12 9.220 4.440 0.08 -3.71 LP B1 80 5.247 1666.05 9.220 4.410 -0.06 -3.84 LP B1 81 5.313 1665.98 9.250 4.450 -0.15 -3.97 LP B1 82 5.379 1665.91 9.180 4.260 0.50 -3.77 LP B1 83 5.445 1665.85 9.130 4.340 -0.35 -4.22 LP B1 84 5.511 1665.78 -1.09 -4.48 LP B1 85 5.577 1665.71 9.070 4.200 -1.09 -4.57 LP B1 86 5.643 1665.65 9.100 4.080 -0.53 -4.41 LP B1 87 5.709 1665.58 9.090 4.360 -1.38 -4.52 LP B1 88 5.775 1665.51 9.070 4.550 -1.19 -4.53 LP B1 89 5.841 1665.45 9.050 4.830 -0.77 -4.40 LP B1 90 5.907 1665.38 9.030 4.970 -0.57 -4.12 LP B1 91 5.973 1665.32 9.220 4.640 -0.38 -4.00 LP B1 92 6.039 1665.25 9.130 4.890 -0.55 -4.05 LP B1 93 6.105 1665.18 9.219 5.354 -0.21 -3.83 LP B1 94 6.171 1665.12 9.245 4.724 0.19 -3.62 LP B1 95 6.237 1665.04 9.271 4.448 0.03 -3.83 LP B1 96 6.303 1664.95 9.173 4.862 -0.41 -4.15 LP B1 97 6.369 1664.87 9.135 4.862 -0.37 -4.30

PAGE 172

157 Appendix 4 (Continued) sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP B3 1 0.033 1627.93 9.110 6.060 -0.77 -4.05 LP B3 2 0.099 1627.88 -0.61 -3.87 LP B3 3 0.165 1627.83 9.200 5.380 -0.38 -4.36 LP B3 4 0.231 1627.78 9.250 6.000 -0.59 -4.12 LP B3 5 0.297 1627.74 9.130 5.000 -0.75 -4.13 LP B3 6 0.363 1627.68 9.150 5.080 -0.40 -4.11 LP B3 7 0.429 1627.64 9.000 4.640 -0.71 -4.52 LP B3 8 0.495 1627.59 -0.65 -4.30 LP B3 9 0.561 1627.54 9.220 4.560 -1.33 -4.62 LP B3 10 0.627 1627.44 9.030 4.760 -0.43 -4.40 LP B3 11 0.693 1627.34 9.120 4.880 -0.87 -4.30 LP B3 12 0.759 1627.24 9.120 4.630 -0.57 -4.36 LP B3 13 0.825 1627.14 9.150 4.830 -0.72 -4.34 LP B3 14 0.891 1627.04 9.120 4.540 -0.53 -4.22 LP B3 15 0.957 1626.96 9.170 4.650 -0.50 -4.33 LP B3 16 1.023 1626.90 9.080 5.410 -1.10 -4.58 LP B3 17 1.089 1626.83 9.050 4.480 -0.76 -4.47 LP B3 18 1.155 1626.75 9.140 4.720 -0.74 -4.74 LP B3 19 1.221 1626.69 9.110 4.500 -0.69 -4.71 LP B3 20 1.287 1626.62 9.150 4.350 -1.17 -4.75 LP B3 21 1.353 1626.50 9.070 4.730 -1.67 -4.76 LP B3 22 1.419 1626.39 -1.53 -4.34 LP B3 23 1.485 1626.27 9.090 5.560 -1.00 -4.21 LP B3 24 1.551 1626.15 9.050 5.380 -0.66 -4.04 LP B3 25 1.617 1626.04 9.170 4.710 -0.51 -3.94 LP B3 26 1.683 1625.87 9.110 4.960 -0.86 -4.45 LP B3 27 1.749 1625.79 9.020 4.500 -1.21 -4.56 LP B3 28 1.815 1625.71 9.110 4.590 -0.99 -4.59 LP B3 29 1.881 1625.63 9.060 5.210 -1.08 -4.56 LP B3 30 1.947 1625.54 9.040 5.010 -0.78 -4.32 LP B3 31 2.013 1625.46 9.170 4.810 -0.98 -4.32 LP B3 32 2.079 1625.38 9.210 4.720 -0.88 -4.14 LP B3 33 2.145 1625.29 9.260 4.970 -1.20 -4.11 LP B3 34 2.211 1625.21 9.230 5.390 -0.27 -3.63 LP B3 35 2.277 1625.13 9.170 5.460 -0.60 -3.88 LP B3 36 2.343 1625.04 9.260 4.480 -0.13 -3.72 LP B3 37 2.409 1624.93 9.260 5.340 -0.44 -4.02 LP B3 38 2.475 1624.83 9.190 5.410 0.18 -4.05 LP B3 39 2.541 1624.71 9.070 4.840 -0.59 -4.24 LP B3 40 2.607 1624.62 9.090 4.470 0.04 -4.07 LP B3 41 2.673 9.280 4.750 LP B3 42LP B3 51 2.739 1624.41 9.175 5.115 -0.23 -4.04 LP B3 43LP B3 52 2.805 1624.32 9.215 5.105 -0.44 -3.92 LP B3 44LP B3 53 2.871 1624.23 9.215 5.465 -0.38 -3.77 LP B3 45LP B3 54 2.937 1624.12 9.295 6.130 0.10 -3.65 LP B3 46LP B3 55 3.003 1624.05 9.290 5.635 -0.45 -3.74 LP B3 47LP B3 56 3.069 1623.97 9.260 5.365 -0.75 -3.84 LP B3 48LP B3 57 3.135 1623.89 9.270 5.430 -0.79 -3.91

PAGE 173

158 Appendix 4 (Continued) LP B3 49LP B3 58 3.201 1623.81 9.190 5.195 -1.03 -4.18 LP B3 50LP B3 59 3.267 1623.74 9.130 4.740 -0.74 -4.27 LP B3 60 3.333 1623.65 9.220 4.990 -0.73 -4.09 LP B3 61 3.399 1623.58 9.210 4.990 -0.79 -4.06 LP B3 62 3.465 1623.51 9.190 4.760 -0.71 -4.14 LP B3 63 3.531 1623.42 9.130 5.120 -0.56 -4.01 LP B3 64 3.597 1623.35 9.190 5.560 0.00 -3.90 LP B3 65 3.663 1623.27 9.130 5.710 -0.40 -3.94 LP B3 66 3.729 1623.19 9.180 5.230 -0.51 -3.67 LP B3 67 3.795 1623.11 9.090 4.930 -0.66 -3.70 LP B3 68 3.861 1623.04 9.160 5.130 -0.18 -3.39 LP B3 69 3.927 1622.92 9.200 5.540 -0.95 -4.06 LP B3 70 3.993 1622.82 9.150 5.080 -1.18 -4.24 LPT B3 71 4.059 1622.71 9.150 5.090 -1.29 -4.51 LP B3 72 4.125 1622.63 9.170 4.930 -1.31 -4.28 LP B3 73 4.191 1622.57 9.240 4.290 -0.32 -4.08 LP B3 74 4.257 1622.50 9.090 5.310 -0.79 -4.34 LP B3 75 4.323 1622.42 9.090 7.550 -0.27 -3.83 LP B3 76 4.389 1622.36 9.160 5.720 -0.15 -3.72 LP B3 77 4.455 1622.29 9.180 5.120 -0.68 -3.94 LP B3 78 4.521 1622.16 9.360 4.880 -0.74 -3.64 LP B3 79 4.587 1622.04 9.220 5.440 0.47 -3.36 LP B3 80 4.653 1621.97 9.170 5.860 -0.85 -3.88 LP B3 81 4.719 1621.90 9.250 4.940 -0.97 -4.03 LP B3 82 4.785 1621.83 9.290 4.680 -0.67 -3.98 LP B3 83 4.851 1621.76 9.350 4.130 -0.95 -4.21 LP B3 84 4.917 1621.68 9.150 4.610 -0.36 -4.33 LP B3 85 4.983 1621.61 9.180 4.520 -0.93 -4.21 LP B3 86 5.049 1621.55 9.050 4.700 -0.65 -4.29 LP B3 87 5.115 1621.47 9.080 4.890 -0.96 -4.39 LP B3 88 5.181 1621.40 9.060 4.940 -0.98 -4.38 LP B3 89 5.247 1621.33 9.180 5.350 -0.80 -3.92 LP B3 90 5.313 1621.26 9.260 5.650 -0.57 -3.94 LP B3 91 5.379 9.140 6.040 LP B3 92 5.445 1621.19 9.250 5.060 -1.05 -3.62 LP B3 93 5.511 9.130 5.170 LP B3 94 5.577 1621.12 9.130 5.310 -0.99 -3.58 LP B3 95 5.643 1621.04 9.050 5.900 -1.28 -3.40 LP B3 96 5.709 1620.88 9.150 4.800 -0.78 -4.27 LP B3 97 5.775 1620.71 9.120 4.570 -0.59 -4.68 LP B3 98 5.841 1620.64 9.350 6.460 -0.89 -4.56 LP B3 99 5.907 1620.58 9.150 5.740 -0.49 -4.51 LP B3 100 5.973 1620.51 9.210 5.820 -0.76 -4.44 LP B3 101 6.039 1620.44 9.140 9.010 -0.48 -3.93 LP B3 102 6.105 1620.38 -0.87 -4.19 LP B3 103 6.171 1620.25 9.240 8.890 -0.37 -4.08 LP B3 104 6.237 1620.12 -0.29 -3.45 LP B3 105 6.303 1620.07 9.260 6.150 -0.24 -3.57 LP B3 106 6.369 1620.03 9.300 5.180 -0.31 -3.66 LP B3 107 6.435 1619.99 9.120 6.940 0.14 -3.62

PAGE 174

159 Appendix 4 (Continued) LP B3 108 6.501 1619.95 9.250 5.760 -0.87 -3.84 LP B3 109 6.567 1619.91 9.240 5.310 -0.73 -4.02 LP B3 110 6.633 1619.87 9.160 4.930 -0.86 -4.23 sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP C1 1 0.033 1567.87 9.117 4.951 -0.48 -4.27 LP C1 2 0.099 1567.76 9.159 4.600 -0.83 -4.34 LP C1 3 0.165 1567.64 9.132 4.627 -0.46 -4.39 LP C1 4 0.231 1567.52 9.134 4.963 -1.15 -4.5 LP C1 5 0.297 1567.41 9.174 4.721 -1.04 -4.39 LP C1 6 0.363 1567.29 9.166 5.141 -1.10 -4.37 LP C1 7 0.429 1567.12 9.128 5.201 -0.62 -4.05 LP C1 8 0.495 1567.04 9.108 5.875 -1.08 -4.16 LP C1 9 0.561 1566.95 9.132 5.088 -1.14 -4.32 LP C1 10 0.627 1566.87 9.094 6.128 -1.09 -4.31 LP C1 11 0.693 1566.79 9.093 5.411 -1.13 -4.51 LP C1 12 0.759 1566.73 9.242 5.132 -1.09 -4.54 LP C1 13 0.825 1566.68 9.204 5.209 -0.74 -4.42 LP C1 14 0.891 1566.62 9.131 4.853 -0.82 -4.47 LP C1 15 0.957 1566.57 9.098 4.885 -0.95 -4.35 LP C1 16 1.023 1566.51 9.096 4.969 -1.18 -4.49 LP C1 17 1.089 1566.46 9.052 4.698 -1.14 -4.57 LP C1 18 1.155 1566.31 9.170 4.675 -1.09 -4.22 LP C1 19 1.221 1566.17 9.143 4.766 -0.26 -4.23 LP C1 20 1.287 1566.04 9.160 4.963 0.68 -3.75 LP C1 21 1.353 1565.97 9.112 4.978 -0.47 -3.99 LP C1 22 1.419 1565.91 9.123 4.854 -0.92 -4.13 LP C1 23 1.485 1565.86 9.120 5.116 -1.30 -4.33 LP C1 24 1.551 1565.79 9.047 5.282 -0.93 -4.57 LP C1 25 1.617 1565.73 9.151 4.602 -1.29 -4.54 LP C1 26 1.683 1565.67 9.135 4.734 -0.82 -4.54 LP C1 27 1.749 1565.60 9.068 4.863 -1.66 -4.6 LP C1 28 1.815 1565.54 9.138 4.982 -1.29 -4.52 LP C1 29 1.881 1565.48 9.092 5.213 -1.23 -4.45 LP C1 30 1.947 1565.41 9.115 5.329 -0.72 -4.33 LP C1 31 2.013 1565.36 9.113 5.101 -0.87 -4.42 LP C1 32 2.079 1565.30 9.229 5.027 -0.98 -4.21 LP C1 33 2.145 1565.23 9.102 5.131 -0.55 -4.04 LP C1 34 2.211 1565.17 -0.74 -4.04 LP C1 35 2.277 1565.11 9.240 4.817 -0.96 -4.14 LP C1 36 2.343 1565.04 9.105 4.872 -0.17 -3.8 LP C1 37 2.409 1564.96 9.181 4.962 -0.88 -4.07 LP C1 38 2.475 1564.88 9.079 5.219 -1.08 -4.23 LP C1 39 2.541 1564.79 9.126 4.982 -1.03 -4.54 LP C1 40 2.607 1564.72 9.316 4.580 -0.55 -4.4 LP C1 41 2.673 1564.65 9.250 4.419 -0.21 -4.42 LP C1 42 2.739 1564.56 9.322 4.917 -1.29 -4.57 LP C1 43 2.805 1564.49 9.298 4.522 -1.06 -4.4 LP C1 44 2.871 1564.42 9.114 4.984 -0.99 -4.41

PAGE 175

160 Appendix 4 (Continued) LP C1 45 2.937 1564.34 9.208 5.178 -0.84 -4.35 LP C1 46 3.003 1564.27 9.259 5.229 -1.15 -4.14 LP C1 47 3.069 1564.19 9.110 5.286 -1.15 -4.15 LP C1 48 3.135 1564.11 9.188 5.185 -1.15 -4.08 LP C1 49 3.201 1564.04 9.081 5.896 -1.14 -3.95 LP C1 50 3.267 1563.93 9.091 5.015 -1.13 -4.15 LP C1 51 3.333 1563.82 8.983 5.189 -1.12 -4.23 LP C1 52 3.399 1563.71 -0.91 -4.46 LP C1 53 3.465 1563.60 9.140 4.818 -0.68 -4.35 LP C1 54 3.531 1563.47 9.404 4.481 LP C1 55 3.597 1563.36 9.412 4.516 -0.76 -4.23 LP C1 56 3.663 1563.25 9.210 5.290 -0.67 -3.98 LP C1 57 3.729 1563.13 9.271 5.108 0.68 -3.69 LP C1 58 3.795 1563.08 9.321 4.452 0.00 -3.74 LP C1 59 3.861 1563.03 9.209 5.304 -0.49 -3.97 LP C1 60 3.927 1562.97 9.326 4.569 -0.80 -4.19 LP C1 61 3.993 1562.92 9.057 4.736 -0.94 -4.22 LP C1 62 4.059 1562.87 9.044 5.263 -1.29 -4.46 LP C1 63 4.125 1562.81 9.395 4.108 -0.93 -4.62 LP C1 64 4.191 1562.76 9.019 4.967 -0.99 -4.64 LP C1 65 4.257 1562.71 9.140 5.151 -1.46 -4.76 LP C1 66 4.323 1562.59 8.993 5.267 -1.21 -4.47 LP C1 67 4.389 1562.48 9.141 4.594 -1.12 -4.6 LP C1 68 4.455 1562.36 9.022 5.085 -0.91 -4.53 LP C1 69 4.521 1562.24 9.065 5.606 -0.96 -4.29 LP C1 70 4.587 1562.13 9.326 5.080 -0.98 -4.08 LP C1 71 4.653 1562.01 9.058 4.784 -1.44 -4.39 LP C1 72 4.719 1561.88 8.938 4.833 -1.37 -4.64 LP C1 73 4.785 1561.80 9.180 4.889 -1.15 -4.66 LP C1 74 4.851 1561.72 9.064 4.651 -1.86 -4.87 LP C1 75 4.917 1561.63 9.022 5.252 -1.36 -4.58 LP C1 76 4.983 1561.55 8.985 4.516 -1.33 -4.7 LP C1 77 5.049 1561.47 -1.31 -4.8 LP C1 78 5.115 1561.38 9.026 4.909 -0.67 -4.56 LP C1 79 5.181 1561.30 9.152 4.397 -0.67 -4.39 LP C1 80 5.247 1561.22 9.157 5.149 -0.52 -4.23 LP C1 81 5.313 1561.13 9.203 5.072 0.35 -3.84 LP C1 82 5.379 1561.09 9.200 4.864 0.30 -3.9 LP C1 83 5.445 1561.04 9.155 5.197 0.02 -3.83 LP C1 84 5.511 1560.96 9.046 5.293 -0.62 -4.21 LP C1 85 5.577 1560.90 9.156 4.858 -0.53 -4.4 LP C1 86 5.643 1560.83 9.041 4.974 -1.14 -4.54 LP C1 87 5.709 1560.76 9.132 4.671 -0.82 -4.4 LP C1 88 5.775 1560.69 9.204 4.802 -0.77 -4.49 LP C1 89 5.841 1560.62 9.152 4.570 -1.11 -4.58 LP C1 90 5.907 1560.55 9.029 4.825 -1.22 -4.59 LP C1 91 5.973 1560.48 9.140 4.939 -0.53 -4.5 LP C1 92 6.039 1560.41 9.054 4.897 -0.67 -4.59 LP C1 93 6.105 1560.34 9.085 5.198 -0.20 -4.48 LP C1 94 6.171 1560.27 9.278 4.744 -0.24 -4.19

PAGE 176

161 Appendix 4 (Continued) LP C1 95 6.237 1560.21 9.205 4.753 -0.27 -4.15 LP C1 96 6.303 1560.13 9.208 5.488 -0.41 -4 LP C1 97 6.369 1560.07 9.110 4.988 -0.52 -4.22 LP C1 98 6.435 1560.01 9.056 5.455 -0.26 -4.11 LP C1 99 6.501 1559.94 9.086 4.882 -0.61 -4.28 LP C1 100 6.567 1559.88 9.034 4.866 -1.10 -4.63 sample ID Depth Date Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) Years C.E. (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP D2 1 0.033 1493.32 9.070 5.172 -0.53 -3.99 LP D2 2 0.099 1493.26 -0.85 -4.21 LP D2 3 0.165 9.141 5.048 LP D2 4 0.231 1493.12 9.418 4.713 -0.61 -3.83 LP D2 5 0.297 1493.05 9.185 5.050 -0.66 -3.86 LP D2 6 0.363 1492.98 9.098 5.438 -1.00 -4.11 LP D2 7 0.429 1492.91 9.087 6.896 -1.27 -4.02 LP D2 8 0.495 1492.85 8.988 7.612 -1.66 -4.52 LP D2 9 0.561 1492.78 9.118 7.623 -1.55 -4.49 LP D2 10 0.627 1492.71 9.139 6.798 -1.74 -4.68 LP D2 11 0.693 1492.65 -1.16 -4.52 LP D2 12 0.759 1492.59 9.083 5.570 0.05 -4.4 LP D2 13 0.825 LP D2 14 0.891 1492.46 9.276 6.120 -0.98 -4.56 LP D2 15 0.957 1492.40 9.017 6.481 -1.17 -4.41 LP D2 16 1.023 1492.34 9.091 8.145 -0.83 -4.23 LP D2 17 1.089 1492.28 9.172 5.730 -0.87 -4.29 LP D2 18 1.155 1492.22 9.131 7.749 -0.98 -4.2 LP D2 19 1.221 1492.16 9.113 5.511 -1.01 -4.26 LP D2 20 1.287 1492.10 9.164 4.603 -0.69 -4.26 LP D2 21 1.353 1492.04 -0.26 -4.1 LP D2 22 1.419 1491.95 9.110 7.886 -0.52 -4.26 LP D2 23 1.485 1491.85 9.056 5.819 0.53 -4.22 LP D2 24 1.551 1491.76 9.198 5.508 0.12 -4.39 LP D2 25 1.617 1491.66 9.253 5.883 0.33 -4.19 LP D2 26 1.683 1491.57 9.130 5.321 -0.42 -4.28 LP D2 27 1.749 1491.48 9.098 8.553 -0.85 -4.31 LP D2 28 1.815 1491.38 -0.95 -4.02 LP D2 29 1.881 1491.29 9.193 7.165 -0.62 -4.08 LP D2 30 1.947 1491.13 9.265 5.807 0.20 -3.65 LP D2 31 2.013 1491.09 9.319 5.358 -1.15 -4 LP D2 32 2.079 1491.04 9.177 6.412 0.18 -3.77 LP D2 33 2.145 1490.96 9.105 6.751 -0.97 -4.18 LP D2 34 2.211 1490.88 9.089 7.569 -1.05 -4.35 LP D2 35 2.277 1490.79 9.184 7.914 -0.78 -4.47 LP D2 36 2.343 1490.71 9.309 4.539 -1.27 -4.73 LP D2 37 2.409 1490.61 9.215 6.079 -0.99 -4.66 LP D2 38 2.475 1490.52 9.229 4.657 -0.39 -4.14 LP D2 39 2.541 1490.42 9.180 6.852 -0.25 -4.32 LP D2 40 2.607 1490.32 9.065 6.018 -0.25 -4.17 LP D2 41 2.673 1490.23 9.098 6.899 0.24 -4.04

PAGE 177

162 Appendix 4 (Continued) LP D2 42 2.739 1490.13 9.245 6.654 0.94 -3.52 LP D2 43 2.805 1490.07 9.261 4.756 -0.19 -3.76 LP D2 44 2.871 1490.01 9.208 6.543 -0.20 -3.83 LP D2 45 2.937 1489.95 9.225 7.660 -0.65 -4 LP D2 46 3.003 1489.89 9.237 4.660 0.28 -3.73 LP D2 47 3.069 1489.83 9.354 5.025 -1.21 -4.16 LP D2 48 3.135 1489.77 9.287 5.195 -1.30 -4.3 LP D2 49 3.201 1489.71 9.368 6.856 -1.62 -4.52 LP D2 50 3.267 1489.65 9.214 5.367 -0.79 -4.19 LP D2 51 3.317 1489.58 9.096 7.359 -1.72 -4.35 LP D2 52 3.383 1489.52 9.344 4.495 -0.69 -4.26 LP D2 53 3.449 1489.45 9.251 5.328 -1.13 -4.26 LP D2 54 3.515 1489.39 9.146 5.095 -1.25 -4.33 LP D2 55 3.581 1489.32 9.218 5.852 -0.50 -3.99 LP D2 56 3.647 1489.26 9.279 7.517 -0.76 -3.95 LP D2 57 3.713 1489.19 9.208 5.999 -0.90 -3.76 LP D2 58 3.779 1489.13 9.206 7.305 -0.34 -3.55 LP D2 59 3.845 1488.96 9.106 6.053 -1.15 -4.03 LP D2 60 3.911 1488.86 9.250 5.050 -0.39 -3.72 LP D2 61 3.977 1488.75 9.336 6.379 -1.39 -4.39 LP D2 62 4.043 1488.65 -0.60 -4.21 LP D2 63 4.109 1488.55 9.204 5.426 -0.42 -4.04 LP D2 64 4.175 1488.44 9.095 6.059 -0.57 -4.08 LP D2 65 4.241 1488.34 9.235 5.789 -1.05 -4.18 LP D2 66 4.307 1488.23 9.226 7.081 -0.57 -3.97 LP D2 67 4.373 1488.13 9.321 5.381 0.09 -3.77 LP D2 68 4.439 1488.05 9.288 5.517 -0.56 -3.76 LP D2 69 4.505 1487.96 9.302 7.053 -0.80 -3.93 LP D2 70 4.571 1487.88 9.358 5.054 -1.21 -4.1 LP D2 71 4.637 1487.80 9.214 7.682 -1.05 -4.03 LP D2 72 4.703 1487.71 9.214 7.600 -1.46 -4.34 LP D2 73 4.769 1487.63 9.371 6.082 -0.87 -4.27 LP D2 74 4.835 1487.55 9.220 8.678 -0.29 -4.19 LP D2 75 4.901 1487.46 9.111 9.796 -0.93 -4.16 LP D2 76 4.967 1487.38 9.098 8.894 -1.41 -4.32 LP D2 77 5.033 1487.30 9.281 5.904 -0.87 -4.17 LP D2 78 5.099 1487.21 9.254 7.406 -0.60 -4.01 LP D2 79 5.165 1487.13 9.284 5.953 -0.32 -3.74 LP D2 80 5.231 1487.05 0.06 -3.73 LP D2 81 5.297 1486.96 9.134 7.996 -1.09 -4.12 LP D2 82 5.363 1486.88 9.190 6.754 -1.91 -4.42 LP D2 83 5.429 1486.80 9.200 6.212 -1.43 -4.25 LP D2 84 5.495 1486.71 9.098 9.148 -1.91 -4.89 LP D2 85 5.561 1486.63 9.189 9.230 -1.39 -4.93 LP D2 86 5.627 1486.55 9.058 8.267 -0.94 -4.43 LP D2 87 5.693 1486.46 9.050 7.685 -1.86 -4.53 LP D2 88 5.759 1486.38 9.083 8.004 -1.69 -4.43 LP D2 89 5.825 1486.30 9.047 9.390 -1.17 -4.28 LP D2 90 5.891 1486.21 9.315 7.952 -0.95 -4.07 LP D2 91 5.957 1486.13 9.376 6.349 -0.71 -3.72

PAGE 178

163 Appendix 4 (Continued) LP D2 92 6.023 1486.10 9.029 8.228 -0.97 -3.83 LP D2 93 6.089 1486.06 9.286 6.707 -1.16 -3.94 LP D2 94 6.155 1486.03 9.169 7.680 -1.03 -3.82 LP D2 95 6.221 9.165 6.805 LP D2 96 6.287 1485.96 9.326 8.981 -0.24 -3.79 LP D2 97 6.353 1485.88 -0.05 -4.1 LP D2 98 6.419 1485.79 -1.30 -4.53 LP D2 99 6.485 1485.71 -0.82 -4.19 LP D2 100 6.551 1485.63 -0.03 -3.93 LP D2 101 6.617 1485.54 -0.76 -3.98 LP D2 102 6.683 1485.46 -0.96 -4.06 LP D2 103 6.749 1485.38 -0.68 -3.88 LP D2 104 6.815 1485.29 0.05 -3.8 LP D2 105 6.881 1485.21 -0.81 -3.89 LP D2 106 6.947 1485.12 -1.13 -3.87 LP D2 107 7.013 1485.04 0.06 -3.56 LP D2 108 7.079 1484.89 -0.95 -4.09 LP D2 109 7.145 1484.74 -1.78 -4.26 LP D2 110 7.211 1484.59 -0.81 -4.1 LP D2 111 7.277 1484.44 -0.83 -4.24 LP D2 112 7.343 1484.29 0.05 -3.69 LP D2 113 7.409 9.139 5.683 LP D2 114 7.475 9.119 8.473 LP D2 115 7.541 9.144 6.201 LP D2 116 7.607 9.298 6.190 LP D2 117 7.673 9.228 5.726 LP D2 118 7.739 9.311 5.311 LP D2 119 7.805 9.232 6.562 LP D2 120 7.871 9.181 8.547 LP D2 121 7.937 9.119 5.503 LP D2 122 8.003 9.114 8.717 LP D2 123 8.069 9.116 10.521 LP D2 124 8.135 9.158 8.077 LP D2 125 8.201 9.127 8.168 LP D2 126 8.267 9.000 7.662 LP D2 127 8.333 9.043 8.711 LP D2 128 8.399 9.028 14.868 LP D2 129 8.465 9.101 11.148 LP D2 130 8.531 9.193 11.411 LP D2 131 8.597 9.158 6.866 LP D2 132 8.663 9.153 5.860 LP D2 133 8.729 9.099 7.904 LP D2 134 8.795 9.090 6.491 LP D2 135 8.861 9.122 8.586 LP D2 136 8.927 9.405 6.514 LP D2 137 8.993 9.129 5.932 LP D2 138 9.059 9.065 6.378 LP D2 139 9.125 9.160 5.804 LP D2 140 9.191 9.282 3.677

PAGE 179

164 Appendix 4 (Continued) sample ID Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LPD4-1 9.242 6.444 -0.49 -4.08 LPD4-2 9.261 5.378 -0.24 -3.76 LPD4-3 9.027 15.722 -0.56 -3.95 LPD4-4 9.204 5.189 -0.22 -4.06 LPD4-5 9.228 4.945 0.17 -3.97 LPD4-6 9.003 14.194 -0.60 -4.21 LPD4-7 9.129 11.644 -0.77 -4.19 LPD4-8 9.244 4.972 -0.17 -3.87 LPD4-9 9.269 6.624 -0.83 -3.91 LPD4-10 9.106 15.785 -0.66 -3.96 LPD4-11 9.291 5.125 -0.67 -3.81 LPD4-12 4.420 -0.76 -3.88 LPD4-13 9.152 7.827 -0.91 -4.04 LPD4-14 9.152 5.565 -1.06 -4.62 LPD4-15 9.352 4.283 -1.39 -4.35 LPD4-16 9.303 5.094 -1.07 -4.13 LPD4-17 9.147 4.677 -0.50 -3.91 LPD4-18 9.260 4.884 -0.83 -4.09 LPD4-19 9.238 5.302 -0.65 -4.17 LPD4-20 9.223 4.511 -0.91 -4.37 LPD4-21 9.191 9.145 0.33 -3.74 LPD4-22 9.315 5.270 -0.67 -4.19 LPD4-23 9.267 5.457 -0.13 -4.10 LPD4-24 9.316 4.596 -0.41 -4.20 LPD4-25 9.320 4.647 -0.49 -4.04 LPD4-26 9.261 4.286 LPD4-27 9.088 8.505 -0.54 -4.03 LPD4-28 9.318 4.263 LPD4-29 9.233 4.682 0.13 -4.09 LPD4-30 4.444 -0.90 -4.64 LPD4-31 9.256 5.338 -0.05 -3.99 LPD4-32 9.299 6.104 0.43 -3.78 LPD4-33 9.073 9.439 -0.48 -4.21 LPD4-34 9.234 6.194 -0.37 -4.02 LPD4-35 17.550 0.50 -4.12 LPD4-36 9.127 4.507 -1.00 -4.42 LPD4-37 9.146 5.329 -0.37 -4.30 LPD4-38 9.254 4.912 0.03 -3.84 LPD4-39 9.195 7.079 0.02 -3.70 LPD4-40 9.070 10.923 -0.12 -3.76 sample ID Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP10.2-1 9.209 4.373 -0.88 -3.57 LP10.2-2 9.219 4.022 -0.24 -3.51 LP10.2-3 9.111 3.832 0.35 -3.83 LP10.2-4 9.054 4.129 0.17 -3.87 LP10.2-5 9.064 4.317 -2.31 -4.00

PAGE 180

165 Appendix 4 (Continued) LP10.2-6 9.099 4.006 -1.90 -3.90 LP10.2-7 9.078 4.099 -1.42 -3.89 LP10.2-8 9.148 3.835 0.58 -3.65 LP10.2-9 9.090 4.282 -1.33 -3.95 LP10.2-10 9.164 4.236 0.81 -3.37 LP10.2-11 9.165 4.088 -0.28 -3.79 LP10.2-12 9.268 3.546 -0.64 -3.45 LP10.2-13 9.141 3.805 0.77 -3.45 LP10.2-14 9.077 6.030 0.19 -3.86 LP10.2-15 9.067 4.474 -1.64 -3.90 LP10.2-16 9.134 4.353 -1.89 -3.95 LP10.2-17 9.343 3.667 -0.39 -3.43 LP10.2-18 9.261 3.851 0.56 -3.16 LP10.2-19 9.228 4.000 -0.64 -3.63 LP10.2-20 9.255 4.039 -0.35 -3.46 LP10.2-21 9.130 3.733 -2.09 -4.05 LP10.2-22 9.291 3.573 -1.90 -4.05 sample ID Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP7-1 9.105 4.829 -0.42 -4.07 LP7-2 9.240 4.062 LP7-3 9.122 4.807 -0.84 -4.23 LP7-4 9.114 4.683 -1.13 -4.29 LP7-5 9.080 5.116 -0.71 -3.93 LP7-6 9.106 4.985 -0.90 -3.96 LP7-7 9.095 5.154 -0.50 -4.12 LP7-8 9.107 4.866 -1.78 -4.20 LP7-9 9.100 4.947 -0.51 -4.19 LP7-10 9.124 5.467 -0.17 -3.85 LP7-11 9.110 5.213 -1.57 -3.79 LP7-12 9.101 5.208 -1.53 -4.05 LP7-13 9.114 5.081 -1.23 -3.98 LP7-14 9.082 5.597 -0.46 -3.78 LP7-15 9.099 4.635 -0.23 -3.82 LP7-16 9.111 10.110 -0.04 -3.70 LP7-17 9.100 4.554 -0.03 -3.71 LP7-18 9.099 4.379 0.22 -3.72 LP7-19 9.066 4.524 -1.07 -4.21 LP7-20 -1.17 -4.03 LP7-21 -0.89 -3.96 LP7-22 9.129 4.997 -0.65 -3.79 LP7-23 9.151 4.648 -0.36 -3.76 LP7-24 9.177 9.431 0.46 -3.37 LP7-25 9.171 4.531 0.25 -3.48 LP7-26 9.269 4.366 0.36 -3.47 LP7-27 9.187 5.386 -0.03 -3.59 LP7-28 9.238 4.738 -0.22 -3.57 LP7-29 9.015 5.245 -1.12 -3.92 LP7-30 9.127 4.663 0.15 -3.54

PAGE 181

166 Appendix 4 (Continued) LP7-31 9.034 4.917 0.38 -3.49 LP7-32 9.070 4.930 -1.35 -4.00 LP7-33 9.128 3.892 -0.84 -4.14 sample ID Depth Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (cm) (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP8.2-1 0.033 9.081 3.796 -0.42 -3.65 LP8.2-2 0.100 9.177 3.990 0.09 -3.83 LP8.2-3 0.166 9.125 4.144 -0.86 -3.88 LP8.2-4 0.233 9.120 4.081 -0.32 -3.74 LP8.2-5 0.299 9.117 3.839 0.14 -3.81 LP8.2-6 0.366 9.197 3.817 0.50 -3.25 LP8.2-7 0.433 9.181 3.788 0.39 -3.74 LP8.2-8 0.499 9.242 4.795 0.20 -3.66 LP8.2-9 0.566 9.202 3.859 0.12 -3.54 LP8.2-10 0.632 9.175 3.856 0.53 -3.47 LP8.2-11 0.699 9.237 3.609 0.91 -3.39 LP8.2-12 0.766 9.259 3.750 -0.33 -3.81 LP8.2-13 0.832 9.150 4.214 -0.35 -3.89 LP8.2-14 0.899 9.170 3.984 -0.08 -3.87 LP8.2-15 0.965 9.131 3.927 -0.13 -3.82 LP8.2-16 1.032 9.166 3.739 LP8.2-17 1.099 9.186 3.786 -0.17 -3.66 LP8.2-18 1.165 9.236 3.672 -0.23 -3.59 LP8.2-19 1.232 9.135 3.977 -1.14 -3.64 LP8.2-20 1.298 9.144 3.940 -0.81 -3.60 LP8.2-21 1.365 9.087 4.081 -0.98 -3.89 LP8.2-22 1.432 9.140 4.325 -0.97 -3.92 LP8.2-23 1.498 9.175 3.610 -0.15 -3.66 LP8.2-24 1.565 9.239 3.487 0.23 -3.54 LP8.2-25 1.631 9.272 3.594 LP8.2-26 1.698 9.216 3.587 -0.53 -3.71 LP8.2-27 1.765 9.252 3.617 -0.07 -3.16 LP8.2-28 1.831 9.181 3.670 0.05 -3.27 LP8.2-29 1.898 9.352 3.407 0.52 -3.29 LP8.2-30 1.964 9.286 3.369 0.13 -3.55 LP8.2-31 2.031 9.205 3.770 -0.76 -3.80 LP8.2-32 2.098 9.203 3.818 -0.43 -3.72 LP8.2-33 2.164 9.172 4.083 -0.25 -3.75 LP8.2-34 2.231 9.223 3.863 -0.43 -3.47 LP8.2-35 2.297 9.242 3.530 0.12 -3.57 LP8.2-36 2.364 9.233 3.752 -0.35 -3.49 LP8.2-37 2.431 9.280 3.914 -0.11 -3.36 LP8.2-38 2.497 9.271 3.944 -0.14 -3.26 LP8.2-39 2.564 9.364 3.536 -1.00 -3.36 LP8.2-40 2.630 9.226 3.231 -1.11 -3.52 LP8.2-41 2.697 9.062 4.690 LP8.2-42 2.764 9.123 4.492 -0.82 -3.50 LP8.2-43 2.830 9.102 4.235 -1.16 -3.73 LP8.2-44 2.897 9.115 4.007 -0.62 -3.70

PAGE 182

167 Appendix 4 (Continued) LP8.2-45 2.963 9.034 4.119 -1.45 -3.95 LP8.2-46 3.030 9.128 3.742 -1.07 -3.58 LP8.2-47 3.097 9.109 3.849 -2.04 -4.01 LP8.2-48 3.163 9.070 4.105 -0.33 -3.17 LP8.2-49 3.230 9.122 3.900 -1.92 -3.84 LP8.2-50 3.296 9.035 4.036 -1.80 -3.87 LP8.2-51 3.363 9.043 3.986 -0.66 -3.62 LP8.2-52 3.430 9.039 3.840 -1.26 -3.90 LP8.2-53 3.496 9.055 3.724 -1.08 -3.97 3.563 LP8.2-55 3.629 9.036 4.261 LP8.2-56 3.696 9.099 3.129 -1.71 -3.30 LP8.2-57 3.763 9.122 3.971 -1.58 -3.77 LP8.2-58 3.829 9.058 3.949 -1.58 -3.84 LP8.2-59 3.896 9.076 3.705 -1.48 -3.90 LP8.2-60 3.962 9.110 3.684 -1.47 -4.09 4.029 -1.97 -4.04 LP8.2-62 4.096 9.045 3.994 -2.77 -4.25 LP8.2-63 4.162 9.076 3.874 -2.19 -4.04 LP8.2-64 4.229 9.034 3.932 -1.90 -4.01 LP8.2-65 4.295 9.065 3.949 -1.48 -3.94 LP8.2-66 4.362 9.037 3.888 -1.42 -4.24 LP8.2-67 4.429 9.150 3.761 -0.99 -4.27 LP8.2-68 4.495 9.089 4.239 LP8.2-69 4.562 9.064 4.120 -2.17 -4.00 LP8.2-70 4.628 9.153 3.952 -1.96 -4.01 sample ID Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP11-1 9.206 5.830 0.51 -3.50 LP11-2 9.307 4.987 0.60 -3.56 LP11-3 9.193 5.368 0.14 -3.51 LP11-4 9.355 5.963 0.46 -2.94 LP11-5 9.364 6.275 0.01 -3.30 LP11-6 9.439 6.198 0.01 -3.10 LP11-7 9.299 5.237 -0.18 -3.25 LP11-8 9.131 4.785 -0.44 -3.70 LP11-9 9.295 4.536 -0.40 -3.88 LP11-10 9.327 4.265 0.19 -3.46 LP11-11 9.203 4.136 0.19 -3.71 LP11-12 9.452 3.645 1.11 -3.08 LP11-13 9.389 3.617 0.97 -3.25 LP11-14 9.197 3.910 0.02 -3.84 LP11-15 9.241 3.718 0.14 -3.77 LP11-16 9.193 3.652 -0.16 -3.87 LP11-17 9.295 3.605 0.26 -3.49 LP11-18 9.261 3.744 0.19 -3.65 LP11-19 9.209 3.853 1.17 -3.18 LP11-20 9.196 3.934 0.82 -3.55 LP11-21 9.166 3.790 0.22 -4.12

PAGE 183

168 Appendix 4 (Continued) LP11-22 9.195 3.949 -0.68 -4.20 LP11-23 9.073 4.532 -0.33 -4.00 LP11-24 4.152 0.15 -3.60 LP11-25 9.175 4.002 -0.06 -3.68 LP11-26 9.198 3.828 0.18 -3.73 LP11-27 9.212 4.034 0.07 -3.90 LP11-28 9.172 3.929 -0.28 -3.94 LP11-29 9.181 3.938 -0.39 -3.75 LP11-30 9.138 3.813 0.29 -3.75 LP11-31 3.558 0.15 -3.61 LP11-32 0.15 -3.83 LP11-33 9.097 4.029 -0.25 -3.84 LP11-34 9.059 4.135 -0.63 -3.77 LP11-35 9.166 3.732 -0.37 -3.69 LP11-36 9.191 3.895 -0.45 -3.72 LP11-37 9.184 4.051 -0.41 -3.70 LP11-38 9.116 4.041 0.07 -3.63 LP11-39 9.033 3.969 -0.23 -3.99 LP11-40 9.232 3.846 0.03 -3.73 LP11-41 9.313 4.507 0.30 -2.85 LP11-42 9.385 4.896 1.33 -3.02 LP11-43 9.204 8.444 2.19 -2.74 LP11-44 9.182 9.798 0.94 -3.43 LP11-45 9.222 8.060 0.61 -3.19 LP11-46 9.338 4.317 0.73 -3.35 LP11-47 9.337 3.880 0.82 -3.13 LP11-48 9.332 3.757 0.44 -3.34 LP11-49 9.226 3.725 -0.25 -3.65 LP11-50 9.132 5.827 -0.29 -3.49 LP11-51 9.160 3.936 -0.84 -3.82 LP11-52 9.095 3.942 -0.81 -3.83 LP11-53 9.141 4.070 -0.42 -3.96 LP11-54 9.227 3.661 -0.45 -4.05 LP11-55 9.189 3.770 -0.55 -3.87 LP11-56 9.290 3.701 -0.14 -3.66 LP11-57 9.223 3.701 0.08 -3.82 LP11-58 3.205 0.39 -3.66 LP11-59 9.244 3.423 -0.29 -3.95 LP11-60 9.232 3.355 -0.80 -3.91 LP11-61 9.253 3.407 -0.13 -3.45 LP11-62 9.121 6.339 -0.26 -3.79 LP11-63 9.176 3.729 -0.65 -4.24 LP11-64 9.245 3.437 0.02 -3.75 LP11-65 9.160 3.877 -0.98 -4.11 LP11-66 9.157 3.930 -0.78 -3.80 LP11-67 9.107 4.086 -0.63 -3.67 LP11-68 9.145 4.029 -1.01 -3.84 LP11-69 9.142 4.025 -0.88 -4.00 LP11-70 9.178 3.799 -0.68 -3.95 LP11-71 9.076 3.850 -1.18 -4.11

PAGE 184

169 Appendix 4 (Continued) LP11-72 9.134 4.011 -1.07 -3.88 LP11-73 9.151 4.177 -1.08 -3.72 LP11-74 9.079 4.420 -1.44 -4.02 LP11-75 9.167 4.095 -0.93 -3.90 LP11-76 4.366 -1.55 -4.18 LP11-77 4.048 sample ID Sr/Ca Mg/Ca 13 C PDB 18 O PDB # (mmol/mol) (mmol/mol) ( VPDB) ( VPDB) LP12-1 9.101 6.461 -0.85 -3.57 LP12-2 9.181 4.663 -0.28 -3.17 LP12-3 9.112 4.355 -0.55 -3.57 LP12-4 9.092 3.948 -0.64 -3.71 LP12-5 LP12-6 9.141 4.228 -0.13 -3.58 LP12-7 9.201 3.675 -0.65 -3.84

PAGE 185

170 Appendix 5: Sr/Ca Replication Data Depth 04LPPB1 Sr/Ca 04LPPB1 Mg/Ca 04LPPC1 Sr/Ca 04LPPC1 Mg/Ca 04LPPD1 Sr/Ca 04LPPD1 Mg/Ca 04LPPH1 Sr/Ca 04LPPH1 Mg/Ca (cm) (mmol/mol) (mmol/mol) (mmol/mol) (mmol/mol) (mmol/mol) (mmol/mol) (mmol/mol) (mmol/mol) 0.033 9.146 7.714 9.045 7.024 9.129 5.181 9.169 5.734 0.099 9.177 8.532 9.065 17.921 9.011 8.052 9.230 5.264 0.165 9.306 7.058 9.078 8.579 9.134 6.458 9.213 5.735 0.231 9.280 4.580 9.132 5.687 9.172 7.626 9.145 29.755 0.297 9.141 5.101 9.151 5.699 9.135 11.128 9.175 29.632 0.363 9.133 5.717 9.175 6.619 9.073 16.829 9.176 26.772 0.429 9.190 5.165 9.140 4.988 9.087 12.906 9.208 52.697 0.495 9.152 6.612 9.113 4.918 9.195 6.951 9.232 14.491 0.561 9.154 8.489 9.280 5.811 9.202 5.838 0.627 9.171 5.629 9.207 6.468 9.204 5.054 0.693 9.069 4.625 9.098 5.136 9.170 5.094 0.759 9.124 4.601 9.066 6.951 9.205 6.318 9.138 5.359 0.825 8.978 4.769 9.081 6.445 9.275 6.398 9.052 3.317 0.891 8.990 4.754 9.041 5.384 9.247 23.843 0.957 8.979 4.645 9.037 5.710 9.197 7.413 9.072 5.331 1.023 8.998 4.659 9.031 6.044 9.164 5.437 9.088 5.138 1.089 9.057 4.925 9.040 5.656 9.136 6.557 9.054 5.363 1.155 9.013 7.443 9.122 5.290 9.118 7.502 9.175 5.125 1.221 9.044 5.987 9.086 5.823 9.161 7.424 9.114 5.302 1.287 9.027 4.930 8.989 6.001 9.193 4.966 1.353 9.110 4.456 8.974 6.088 9.138 5.392 9.166 4.975 1.419 9.162 4.627 9.087 5.297 9.044 5.490 9.135 5.818 1.485 9.117 4.566 9.055 5.670 9.078 5.125 9.182 5.246 1.551 9.167 4.049 9.002 5.860 9.051 5.315 9.111 5.592 1.617 9.128 4.092 8.994 6.203 9.191 4.976 9.226 5.268 1.683 9.014 4.222 9.050 6.448 9.195 4.972 9.174 5.408 1.749 9.149 5.664 9.134 5.441 9.133 5.074 1.815 9.054 4.174 9.132 6.382 9.185 5.229 9.234 5.010 1.881 9.026 4.064 9.186 5.732 9.115 5.167 9.285 4.693 1.947 9.039 3.882 9.131 5.229 9.029 5.657 9.261 4.657 2.013 9.270 3.869 9.246 5.461 2.079 9.132 4.000 9.046 6.524 9.052 5.081 9.236 5.227 2.145 9.236 3.970 9.145 4.845 9.045 5.223 9.207 6.397 2.211 9.109 4.101 9.118 5.799 9.132 4.756 9.043 8.050 2.277 9.071 4.232 9.057 5.572 9.053 5.695 9.111 5.459 2.343 9.019 3.994 9.054 5.558 9.130 5.757 9.060 7.144 2.409 9.236 4.014 8.984 5.206 9.163 5.706 9.086 5.742 2.475 8.953 4.333 9.075 4.995 9.031 6.424 9.129 4.925 2.541 9.206 4.297 8.989 5.339 9.035 5.790 9.210 4.273 2.607 9.091 4.743 9.185 4.918 9.146 5.365 9.149 5.431 2.673 9.185 4.294 9.078 5.222 9.071 6.512 9.095 5.780 2.739 9.248 4.555 9.159 5.120 9.053 7.836 9.253 5.351 2.805 9.189 5.096 9.180 5.204 9.071 7.546 9.240 6.904 2.871 9.264 10.914 9.071 5.821 9.115 5.467 9.315 14.839 2.937 9.257 7.541 9.151 4.896 9.151 5.484 9.257 21.561 3.003 9.030 4.863 9.131 5.221 9.256 5.705 9.183 21.351 3.069 9.054 4.283 9.099 5.123 9.174 6.009 9.201 12.209

PAGE 186

171 Appendix 5: (Continued) 3.135 9.055 4.234 9.151 5.890 9.170 13.494 3.201 8.872 4.439 9.057 6.433 9.201 5.938 9.153 8.692 3.267 8.987 4.888 9.063 5.810 9.211 6.281 3.333 9.024 4.665 9.186 5.920 9.143 4.788 3.399 9.101 4.562 9.053 5.082 9.111 5.808 9.129 4.935 3.465 9.046 4.452 9.095 5.469 9.119 5.228 9.169 5.364 3.531 9.050 4.426 9.095 5.326 9.195 4.859 9.156 4.758 3.597 9.081 4.629 9.130 5.842 9.271 5.304 9.205 4.362 3.663 9.094 4.584 9.152 5.580 9.194 4.306 3.729 9.171 4.902 9.052 5.700 9.080 4.602 3.795 9.134 4.534 9.049 5.204 9.131 5.866 9.162 4.530 3.861 9.178 4.423 9.029 5.590 9.135 4.682 9.177 4.584 3.927 9.191 4.482 9.099 7.892 9.117 4.870 3.993 8.962 4.668 9.024 8.495 9.131 5.510 9.164 4.640 4.059 8.998 4.189 9.135 5.495 9.181 4.844 4.125 9.108 4.237 9.064 9.156 9.151 5.020 4.191 9.073 4.806 9.119 11.013 9.187 5.892 9.050 4.836 4.257 9.198 4.248 9.107 9.798 9.193 5.941 9.123 4.596 4.323 9.036 4.451 9.037 9.382 9.160 5.944 9.183 4.431 4.389 9.056 4.842 9.050 9.662 9.107 7.577 9.079 4.301 4.455 9.065 4.347 9.073 8.310 9.114 6.902 9.160 4.474 4.521 9.012 4.439 9.050 7.402 9.098 6.802 9.247 4.741 4.587 9.074 4.050 9.013 7.938 9.112 6.338 9.224 4.384 4.653 9.025 4.439 9.159 7.077 9.133 5.245 9.276 4.799 4.719 9.038 4.633 9.146 8.572 9.258 4.267 4.785 9.038 5.201 9.086 6.501 9.163 10.380 9.234 4.401 4.851 9.087 4.756 9.097 9.010 9.186 19.683 9.122 4.497 4.917 9.088 4.536 9.205 5.795 9.192 27.243 9.112 4.596 4.983 9.081 4.699 9.107 6.503 9.148 8.898 9.106 5.443 5.049 9.084 4.740 9.107 7.574 9.209 6.164 9.117 5.075 5.115 9.124 4.606 9.116 6.264 9.135 6.406 9.203 5.149 5.181 9.122 4.587 9.030 5.842 9.127 5.890 9.170 5.482 5.247 9.100 4.845 9.088 6.033 9.078 5.657

PAGE 187

1 ABOUT THE AUTHOR Kelly Halimeda Kilbourne grew up in Novato and Davis, California. She earned her bachelors degree from Smith College with High Honors from the Department of Geology in 1998. She worked for two years as an Earth Science Intern on the Coastal Marine Team at the U.S. Geological Survey in Menlo Park, California. In 2000, she began her graduate studies at the University of South Florida with Terrence M. Quinn, earning her Masters Degree in 2003. Hali began her doctoral work in the Fall of 2003. Her graduate work was funded by the Von Rosenstiel, Paul Getting, Gulf Oceanographic Charitable Trust and the Knight Oceanographic fellowships awarded to her by the College of Marine Science. Upon completion of her doctoral degree, Ms. Kilbourne will begin a National Research Council postdoctoral research position at the Earth System Research Laboratory with NOAA in Boulder, Colorado.