LATE HOLOCENE CLIMATE VARIABILITY AS PRESERVED IN
HIGH-RESOLUTION ESTUARINE AND LACUSTRINE
SEDIMENT ARCHIVES
BY
JEREMIAH BRADFORD HUBENY
A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
OCEANOGRAPHY
UNIVERSITY OF RHODE ISLAND
2006
DOCTOR OF PHILOSOPHY DISSERTATION
OF
JEREMIAH BRADFORD HUBENY
APPROVED:
Dissertation Committee:
Major Professor______
______
______
DEAN OF THE GRADUATE SCHOOL
UNIVERSITY OF RHODE ISLAND
2006 Abstract
Current concern regarding human-induced environmental and climate changes is becoming higher-profile, especially as officials become more outspoken on the topic. An important piece of the debate regarding anthropogenic change is the determination of to what degree humans are changing systems beyond natural variability. Since comprehensive instrumental records only go back 100 or fewer years, there is a demand for high-quality proxy records of environmental and climate conditions that extend past the instrumental period. One such archival source is aquatic sediment that is preserved at the bottom of a lake, estuary, or ocean. If the water column conditions are conducive to permanent stratification, then annually resolved sediment records can be preserved. This dissertation performs high- resolution proxy analyses from annually resolved sediment records in Rhode Island and central New York State to interpret the natural and anthropogenically induced changes that have occurred over the Late Holocene.
The Pettaquamscutt River Estuary’s Lower Basin was studied and the post- glacial stratigraphy was interpreted. High-productivity lacustrine sedimentation started at ca. 15,500 cal BP and density-stratified estuarine conditions began about
1000 year ago. Over the last four centuries, the estuary has experienced anthropogenic influence through land clearance associated with European settlers, input of organic and non-organic pollutants, and nitrogen loading associated with domestic septic systems. Land clearance (ca. 1700 A.D.) resulted in increased sediment input and soil erosion from the watershed and increased primary productivity in the estuary. These effects lasted approximately two decades. The input of pollutants did not have a significant effect on the estuarine primary producers. Finally, the nitrogen loading (ca. 1960 A.D.) has led to cultural eutrophication in the estuary that is observed in ecologic proxies.
High-resolution proxy data spanning the last millennium from the
Pettaquamscutt River Estuary were used to interpret climate variability in the region.
Mass accumulation rates of the photosynthetic fossil pigment bacteriochlorophyll e were linked to climate processes through water temperature limitation of bacterial production. Observed productivity/climatic cycles reveal linkages between the atmospheric-driven North Atlantic Oscillation phenomenon and the oceanic-driven
Atlantic Multidecadal Oscillation at subdecadal and multidecadal periodicities.
Clastic lamination thicknesses preserved in Pettaquamscutt River Estuary varved sediments record precipitation variability. This relationship was used to reconstruct precipitation variability in Rhode Island over the last millennium and to compare this record to teleconnection climate indices. A significant positive correlation was calculated between the lamination thicknesses and the Pacific/North
American climate pattern at both interannual and decadal time-scales. The mechanism responsible for this linkage is driven by storm tracking associated with variability in the amplitude of the polar front jet stream. During periods of zonal atmospheric flow the region is dry due to the influence of dry continental air masses.
As the amplitude of the jet stream increases, meridional atmospheric circulation becomes dominant thus increasing the influence of moist Gulf of Mexico and coastal air masses.
iii Finally, a varve record was reconstructed from the sediments of Fayetteville
Green Lake, New York in order to compare precipitation variability in Rhode Island and New York. The carbonate laminations were significantly correlated to precipitation conditions in the state. The positive correlation is likely caused by increased precipitation leading to increased groundwater flow. Since groundwater introduces calcium and carbonate ions to the lake, periods of higher precipitation will increase the concentrations of these ions, thus making it easier to precipitate calcium carbonate in the water column. The variability was significantly correlated to the
Pacific/North American climate pattern at decadal time scales. This observation along with significant cross spectral analysis with Pettaquamscutt River data, suggests that precipitation variability between Rhode Island and New York are in phase with each other and are partially driven by the Pacific/North American pattern.
iv Acknowledgements
First and foremost I would like to thank John King for the opportunities that he has provided me with over the last six years. I have had the freedom to find my own directions for research after helpful initial guidance and have also had the opportunity to pursue my passion for teaching, despite the fact that it took me away from lab activities at times. The experiences that I have had in the field and with the various projects that have gone on in the lab have made me a far better scientist than simply doing my work could have offered. In addition, the lack of funding concerns during my tenure at GSO has truly been a blessing.
My committee has been quite helpful over the last few years, offering help and advice when I was looking for it. Jon Boothroyd has taught me much about the
Quaternary geology of southern New England, and gave me perhaps the most valuable opportunity for my career as an educator by hiring me to teach in the URI
Geosciences Department during the Spring 2005 Semester. Jim Quinn has been very helpful for me in learning about chemical compounds, especially the organic ones.
His prompt and thoughtful reviews have been very helpful. Scott Rutherford, although smoking me more than once on the ice, has been instrumental in my learning about spectral analysis techniques and appropriate applications for the methods. Kate Moran has pushed me to learn geophysical concepts and techniques to a level that will continue to make me a more complete geologist.
As a multidisciplinary dissertation, the work presented here is in reality the result of contributions from many people. First, I would like to thank Chip Heil for all of the productive coring trips, both on and off the boat. Mark Cantwell and
v Antelmo Santos have been extremely helpful in analyzing samples at EPA. Carol
Gibson and Danielle Cares have been important players in getting much of this work completed in South Lab. Funding is acknowledged from the State of Rhode Island
(beach survey), the National Science Foundation (GK-12; Great Lakes grant), and the
NAC CESU program.
Perhaps above all, thanks are due to KC who has dealt with me over the years and kept me on track. I appreciate all that you have done and continue to do! In addition, Leigh and the rest of the family have been very supportive through the whole process. Finally, thanks are in order for Doug Allen, Jim Cullen, Lindley
Hanson, Jeanette Sablock, and Peter Sablock for giving me the ultimate reward for finishing the degree.
vi Preface
This dissertation consists of in-depth analyses of the sediment record from the
Pettaquamscutt River Estuary, Rhode Island in regards to natural and anthropogenic environmental and climate change. In addition, a high-resolution paleoclimate record from Fayetteville Green Lake, New York is analyzed and compared to the Rhode
Island data to examine potential regional correlation. The dissertation is written in manuscript format and consists of the following four manuscripts:
Chapter 1, Late Quaternary stratigraphy and environmental history of the
Pettaquamscutt River Lower Basin, evaluates the environmental conditions in the watershed since the inception of this water body after the glacial retreat. In addition, it uses a multi-proxy approach to evaluate the physical and ecological effects of anthropogenic changes in the watershed.
Chapter 2, Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments, uses a biennially resolved record of fossil photosynthetic pigments from the Pettaquamscutt River Estuary to reconstruct variability in climate cycles in the region and links these cycles to atmospheric and oceanic forcings. This manuscript is published in the July 2006 issue of Geology (volume 34, number 7, p. 569-572).
Chapter 3, Late Holocene precipitation variability in southern New England and the relationship to Northern Hemisphere teleconnection patterns, reconstructs precipitation and drought variability from annually resolved sediment laminations preserved in the Pettaquamscutt River Estuary. The study reconstructs the
vii Pacific/North American teleconnection pattern and highlights previously unresolved variability in this climate pattern.
Chapter 4, Paleoclimate correlations between central New York State and
Rhode Island over the last four centuries, reconstructs precipitation variability from central New York State using annually resolved sediment lamination thickness data from Fayetteville Green Lake, New York and compares the associated climate variability to the results discussed in Chapters 3 and 4 from Rhode Island.
viii Table of Contents
Abstract...... ii
Acknowledgements ...... v
Preface...... vii
Table of Contents...... ix
List of Tables...... xii
List of Figures ...... xiii
Chapter 1: Late Quaternary stratigraphy and environmental history of the Pettaquamscutt River Lower Basin...... 1
1.10 ABSTRACT ...... 1
1.20 INTRODUCTION...... 2
1.21 Scope of manuscript ...... 2
1.22 A note on ages reported in this manuscript...... 5
1.23 Late Quaternary deglacial history of southern New England...... 5
1.30 METHODS ...... 10
1.31 Sample Acquisition ...... 10
1.32 Core Logging ...... 12
1.33 Paleomagnetics and Mineral Magnetics ...... 13
1.34 Laminated Sediments and Chronology...... 14
1.35 Contaminant Analysis...... 17
1.36 Stable Isotopes...... 18
1.37 Aquatic Ecology Proxies ...... 20
1.38 Pollen...... 21
1.40 RESULTS ...... 22
1.41 Stratigraphy and chronology...... 22
ix 1.42 Physical and magnetic proxies...... 25
1.43 Chemical proxies...... 28
1.44 Biologic proxies ...... 36
1.50 DISCUSSION ...... 40
1.51 Geologic evolution of the Lower Basin...... 40
1.52 European settlement and environmental change...... 42
1.53 Estuarine response to anthropogenic metals and organics pollution...... 46
1.54 Estuarine response to cultural eutrophication ...... 47
1.60 CONCLUSIONS ...... 50
1.70 REFERENCES ...... 51
Chapter 2: Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments...... 105
2.10 ABSTRACT ...... 105
2.20 INTRODUCTION...... 106
2.30 METHODS ...... 108
2.40 RESULTS AND DISCUSSION ...... 111
2.41 Climate Cycles ...... 111
2.42 Variable Cycles over the Last Millennium...... 113
2.50 CONCLUSIONS ...... 117
2.60 REFERENCES ...... 118
Chapter 3: Late Holocene precipitation variability in southern New England and the relationship to Northern Hemisphere teleconnection patterns ...... 129
3.10 ABSTRACT ...... 129
3.20 INTRODUCTION...... 130
3.30 METHODS ...... 133
3.40 RESULTS AND DISCUSSION ...... 135
x 3.41 Precipitation and Drought Reconstruction...... 135
3.42 Linkage to Northern Hemisphere Teleconnections...... 140
3.50 CONCLUSIONS ...... 147
3.60 REFERENCES ...... 148
Chapter 4: Paleoclimate correlations between central New York State and Rhode Island over the last four centuries ...... 168
4.10 ABSTRACT ...... 168
4.20 INTRODUCTION...... 169
4.21 Regional Paleoclimatology ...... 169
4.22 Fayetteville Green Lake Limnology and Sediment Sources...... 170
4.30 METHODS ...... 171
4.40 RESULTS AND DISCUSSION ...... 173
4.41 Green Lake Age Model...... 173
4.42 Green Lake Climate Correlations...... 174
4.43 Green Lake Teleconnection Correlations ...... 179
4.44 Climate Correlations Between Central New York and Southern New England...... 183
4.50 CONCLUSIONS ...... 188
4.60 REFERENCES ...... 189
Appendix: Thin section preparation and analysis of frozen and soft laminated sediment cores...... 212
Bibliography...... 220
xi List of Tables
Table 1.1. List of sediment cores...... 59
Table 1.2. Fossil pigments analyzed...... 60
Table 1.3. Pettaquamscutt River Estuary freeze core radiocarbon data...... 61
Table 1.4. Pettaquamscutt River Estuary piston core radiocarbon data...... 62
Table 1.5. Effects range low (ERL) and effects range median (ERM) concentrations for pollutants analyzed in this study...... 63
Table 2.1. Pettaquamscutt River Radiocarbon Data...... 122
Table 3.1. Correlation tests between Pettaquamscutt River Estuary compaction- corrected clastic laminae thicknesses and regional climate variables...... 153
Table 3.2. Correlation tests between decadally smoothed Pettaquamscutt River Estuary compaction-corrected clastic laminae thicknesses and regional climate variables, before and after major residential development and storm drains in the watershed...... 154
Table 3.3. Correlation tests between Pettaquamscutt River Estuary compaction- corrected clastic laminae thicknesses and teleconnection patterns...... 155
Table 4.1. Correlation coefficients between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and New York State climate variables.....193
Table 4.2. Correlation coefficients between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and teleconnection indexes...... 194
Table 4.3. Proxy records and the climatic/ecological variable that they represent. ..195
Table 4.4. Correlation coefficients between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and Pettaquamscutt River Estuary varve and laminae thicknesses, and Bacteriochlorophyll e mass accumulation rates...... 196
Table 4.5. Correlation coefficients between decadally smoothed Fayetteville Green Lake compaction-corrected varve and laminae thicknesses and Pettaquamscutt River Estuary varve and laminae thicknesses...... 197
xii List of Figures
Figure 1.1: Locus map of the Pettaquamscutt River Estuary...... 64
Figure 1.2. Topography and bathymetry of the Lower Basin ...... 65
Figure 1.3. Generalized circulation of the upper Pettaquamscutt River Estuary...... 66
Figure 1.4. Major moraines of southern New England...... 67
Figure 1.5. General lithostratigraphy of Pettaquamscutt River Estuary Lower Basin sediments as represented by piston core NR03-4...... 68
Figure 1.6. ARM plots for composite surface freeze core (NR03-1 and NR04-2) and piston core NR03-4...... 69
Figure 1.7. GRAPE (bulk density; g/cc) plots for composite surface freeze core (NR03-1 and NR04-2) and piston cores NR03-4 and NR03-3 ...... 70
Figure 1.8. Magnetic susceptibility plots for composite surface freeze core (NR03-1 and NR04-2) and piston cores NR03-4 and NR03-3...... 71
Figure 1.9. Pettaquamscutt River upper laminated sediments as seen scanned under cross-polarized light...... 72
Figure 1.10. Age model for the varved portion of the Pettaquamscutt River Lower Basin sediment record from composite freeze cores ...... 73
Figure 1.11. Pettaquamscutt River piston core magnetic declination records as matched to dated Northeast regional PSV curve...... 74
Figure 1.12. Pettaquamscutt River piston core magnetic inclination records as matched to dated Northeast regional PSV curve...... 75
Figure 1.13. Zijderveld plots from Pettaquamscutt River piston cores ...... 76
Figure 1.14. Age model for piston core NR03-3...... 77
Figure 1.15. Age model for piston core NR03-4...... 78
Figure 1.16. Pettaquamscutt River declination data plotted on age scale ...... 79
Figure 1.17. Pettaquamscutt River inclination data plotted on age scale ...... 80
Figure 1.18. Physical properties of the Pettaquamscutt River Lower Basin sediments from reference core NR03-4...... 81
Figure 1.19. Same as Figure 1.18 except plotted on age scale...... 82
xiii Figure 1.20. Physical properties and magnetic grain size variations of the Pettaquamscutt River Lower Basin sediments from reference core NR03-4...... 83
Figure 1.21. Same as Figure 1.20 except plotted on age scale...... 84
Figure 1.22. Backscatter electron image (BSEI) of framboidal pyrite grains...... 85
Figure 1.23. Biplot used to help interpret the bulk magnetic minerals present in the Pettaquamscutt River Estuary...... 86
Figure 1.24. Compaction-corrected varve thickness time-series...... 87
Figure 1.25. Polycyclic aromatic hydrocarbon (PAH) concentrations from the Pettaquamscutt River Estuary...... 88
Figure 1.26. Total PAH and Retene concentrations in the Pettaquamscutt River Estuary...... 89
Figure 1.27. Total polychlorinated biphenyl (PCB) concentration in the Pettaquamscutt River Estuary sediment column...... 90
Figure 1.28. Total DDT concentration in the Pettaquamscutt River Estuary sediment column...... 91
Figure 1.29. Trace metal concentrations in the Pettaquamscutt River Estuary sediment column...... 92
Figure 1.30. Stable nitrogen (15N) and stable carbon (13C) isotopic values and carbon to nitrogen (C/N) ratio of the Pettaquamscutt River sediment column...... 93
Figure 1.31. Organic carbon concentrations over the late Quaternary Pettaquamscutt River Estuary record...... 94
Figure 1.32. Organic carbon concentrations in the Pettaquamscutt River Estuary over the last millennium (varved record)...... 95
Figure 1.33. Fossil pigment concentrations over the late Quaternary Pettaquamscutt River Estuary record...... 96
Figure 1.34. Fossil pigment mass accumulation rates in the Pettaquamscutt River Estuary over the last millennium (varved record)...... 97
Figure 1.35. Total percent of pollen grains Rumex and Ambrosia around the turn of the eighteenth century...... 98
Figure 1.36. Total organic carbon, magnetic susceptibility, and bacteriochlorophyll e concentration with associated interpreted depositional environments for the entire post-glacial Pettaquamscutt River history...... 99
xiv Figure 1.37. European land clearance and its effect on sedimentation in the Pettaquamscutt River Estuary...... 100
Figure 1.38. European land clearance and its effect on ecologic productivity in the Pettaquamscutt River Estuary...... 101
Figure 1.39. Compilation of chemical pollutants that exceed sediment quality control guidelines...... 102
Figure 1.40. Mean Effects Range Median Quotient for chemical pollutants in the Pettaquamscutt River record...... 103
Figure 1.41. Evidence for cultural eutrophication in the Pettaquamscutt River Lower Basin...... 104
Figure 2.1. Locus map of the Pettaquamscutt River Estuary, Rhode Island...... 123
Figure 2.2. North Atlantic region showing locations of the Pettaquamscutt River Estuary (circle) and the regional pressure centers associated with a positive NAO during (A) increased, and (B) decreased meridional oceanic heat flux (AMO)...... 124
Figure 2.3. Age model for the varved portion of the Pettaquamscutt River lower basin sediment record...... 125
Figure 2.4. Multi-taper spectral analysis of the Pettaquamscutt River Bchle MAR time-series (1058-2004 A.D.)...... 126
Figure 2.5. A: Pettaquamscutt River Bchle MAR time-series along with (B) the corresponding wavelet transform (Torrence and Compo, 1998). C: Sum of Bchle MAR time-series band-pass filtered at 38.5 and 95.9 years used to represent multidecadal variability. (D) Decadally-smoothed biologic lamination thicknesses, which represent total productivity and runoff...... 127
Figure 2.6. Cross spectral analysis of Pettaquamscutt River Bchle and WNAO time series (1824 – 1960)...... 128
Figure 3.1: Locus map of the Pettaquamscutt River Estuary, Rhode Island...... 156
Figure 3.2. Varve age model for the Pettaquamscutt River Lower Basin...... 157
Figure 3.3. Clastic lamina time-series from the Pettaquamscutt River Lower Basin along with decadally smoothed record...... 158
xv Figure 3.4. Correlation coefficients between decadally smoothed Pettaquamscutt River clastic laminae thicknesses and monthly and seasonal decadally smoothed Rhode Island precipitation data...... 159
Figure 3.5. Reconstructed PDSI values for the last millennium from varved sediments in the Pettaquamscutt River Estuary...... 160
Figure 3.6. Reconstructed PDSI time-series from the Pettaquamscutt River plotted with Northern Hemisphere temperature anomalies from the last millennium...... 161
Figure 3.7. General character of 700-mb flow over the United States during positive (+) and negative (-) states of the PNA ...... 162
Figure 3.8. Correlation coefficients between (A) the PNA index and 500mb height during the winter/ spring seasons and (B) Pettaquamscutt River clastic lamination thicknesses and 500mb height during the winter/ spring seasons ...... 163
Figure 3.9. Correlation coefficients between (A) the PNA index and sea surface temperatures (SST) during the winter/ spring seasons and (B) Pettaquamscutt River clastic lamination thicknesses and SST during the winter/ spring seasons...... 164
Figure 3.10. PNA reconstruction from Pettaquamscutt River clastic laminae thicknesses along with decadally smoothed record...... 165
Figure 3.11. Multi-taper method spectral analyses for the PNA reconstruction (A) before 1695 A.D., and (B) after 1715 A.D...... 166
Figure 3.12. Reconstructed PNA time-series with wavelet transform...... 167
Figure 4.1. Map of the Northeastern United States showing the locations of Fayetteville Green Lake and the Pettaquamscutt River Estuary...... 198
Figure 4.2. Detailed locus map for Fayetteville Green Lake, New York...... 199
Figure 4.3. Image of thin section (FG04FC3I) from Fayetteville Green Lake...... 200
Figure 4.4. Age model for core FG04FC2, Fayetteville Green Lake, NY...... 201
Figure 4.5. Age model for core FG04FC3, Fayetteville Green Lake, NY...... 202
Figure 4.6. Time series for organic lamina, carbonate lamina, and varve thickness from Fayetteville Green Lake...... 203
xvi Figure 4.7. Monthly, seasonal, and annual correlations between Fayetteville Green Lake carbonate laminae thicknesses and annual (top) and decadal (bottom) total precipitation for New York State...... 204
Figure 4.8. Monthly, seasonal, and annual correlations between Fayetteville Green Lake carbonate laminae thicknesses and annual (top) and decadal (bottom) Palmer Drought Severity Index for New York State...... 205
Figure 4.9. Monthly, seasonal, and annual correlations between Fayetteville Green Lake organic laminae thicknesses and annual (top) and decadal (bottom) Palmer Drought Severity Index for New York State...... 206
Figure 4.10. Organic and carbonate lamina thickness time-series from Fayetteville Green Lake and the associated climate variability preserved in each ...... 207
Figure 4.11. Multi-taper method spectral analysis of Fayetteville Green Lake carbonate lamina thickness time-series...... 208
Figure 4.12. Multi-taper method spectral analysis of Fayetteville Green Lake organic lamina thickness time-series...... 209
Figure 4.13. Spectral analysis of Green Lake organic laminae (top) and Pettaquamscutt River biogenic laminae (bottom), with 90% confidence levels shown by dashed lines. Bottom is cross-spectral analysis of the two time-series...... 210
Figure 4.14. Spectral analysis of Green Lake carbonate laminae (top) and Pettaquamscutt River clastic laminae (bottom), with 90% confidence levels shown by dashed lines. Bottom is cross-spectral analysis of the two time-series...... 211
xvii Chapter 1: Late Quaternary stratigraphy and environmental history of the
Pettaquamscutt River Lower Basin
1.10 Abstract
Sediment accumulations preserved in water bodies provide archives of environmental changes that are associated with both natural and anthropogenic forcings. In this manuscript, the stratigraphy of the Lower Basin of the
Pettaquamscutt River Estuary, Rhode Island is investigated to reconstruct the Late
Quaternary geologic history for this watershed, including a high-resolution examination of recent anthropogenic change. Ice retreat exposed the Pettaquamscutt
River valley between 19,200 and 18,790 cal BP. The Lower Basin was formed as a kettle from the melting of stagnant ice blocks, with an extrapolated sediment contact age suggesting a date of formation at ca. 19,170 ± 570 cal BP. The depression housed a low-productivity pond from the time of ice block melting to ca. 15,500 cal
BP, at which time the environment transitioned to a high-productivity lacustrine environment with mature terrestrial vegetation in the watershed. Marine incursion occurred ca. 1000 cal BP due to global sea level rise, and for the last millennium the water column has been density stratified and varved sediments have accumulated.
Over the last three and a half centuries anthropogenic influences have been observed.
First, European land clearance at the end of the seventeenth century led to about two decades of increased sediment transport through the watershed as well as increased primary productivity in the water column. The turbidity increase associated with increased biomass in the water column limited green sulfur bacteria that reside below
1 the oxycline, however. Although a number of organic and inorganic pollutants exceed sediment quality guidelines at times during the last century, no ecologic effects were detected in proxy records. Finally, since the late 1950s evidence of cultural eutrophication is apparent in both stable nitrogen isotope values as well as in productivity proxies. This effect is directly related to residential development in the watershed and the use of septic systems to treat human waste. Overall, the stratigraphy of the Pettaquamscutt River is quite revealing, especially over the last millennium when annual time-series resolution is possible. The work here shows promise toward using high-resolution proxies over the last millennium to address specific climate controlling patterns in the region.
1.20 Introduction
1.21 Scope of manuscript
The recent introduction of the term “Anthropocene” for the present human- dominated geologic epoch is indicative of the growing influence that mankind is having on the environment (Crutzen, 2002). This influence is most widely recognized by rising greenhouse gas concentrations in the atmosphere and the associated global warming that has been observed over the past century and a half by instrumental observations and the past two millennia with proxy reconstruction of temperature (IPCC, 2001; Mann and Jones, 2003). The effects of humans as geologic agents also can be observed through land use changes and pollution of water bodies in that these affect the ecological health and the dynamics of environmental systems.
2 Although it is sometimes attractive to assign blame for environmental change on humans, in reality environmental conditions can change in response to both natural
(i.e. climate, ecological succession, etc.) and anthropogenic (i.e. pollution, land use change, etc.) forcings. In order to argue for anthropogenic influence on a particular natural system, it must first be demonstrated that the system changed as a result of human impacts. Since few instrumental or documentary archives have sufficient detail over long enough temporal scales, it is often hard to accomplish this task.
Therefore, it is necessary to find additional archives that have recorded environmental conditions through time.
Geologic depositional environments can provide such records in the deposits that are preserved. Estuarine, lacustrine, and marine sediments can act as virtual timelines when they can be accurately dated and useful proxies for environmental conditions can be measured. Physical, chemical, and biological proxies provide information on the depositional processes, sediment source, water column conditions, and biologic productivity. These reconstructions can then be compared to known regional geologic changes (i.e. glacial ice retreat and sea level change) as well as measured anthropogenic influences to the water body through land use change and pollution. Such a multi-proxy approach yields records of environmental conditions through time, which can be used to track both natural, and anthropogenic changes to the system.
This manuscript examines the sediment record in the Lower Basin of the
Pettaquamscutt River Estuary, Rhode Island (Figure 1.1) and reconstructs the environmental history since the retreat of the Laurentide Ice Sheet. The
3 Pettaquamscutt River Estuary is a drowned river valley estuary underlain by late
Quaternary stratified sediments and till deposits (Schafer, 1961a, b). In the northern section of the estuary are two basins, which occupy ice-block depressions resulting from the last deglaciation of the area. The Lower Basin is the deeper of the two, with depths exceeding 20 meters in the deepest zone (Figure 1.2) (Gaines, 1975). Due to the excessive depth of these basins as compared to the rest of the estuary, a fjord-like circulation pattern exists, driven by density stratification of the water column (Figure
1.3). Such a circulation pattern leaves the bottom waters stagnant and allows for the oxygen to become depleted. Therefore, the sediments in these basins lack the benthic epifauna and infauna that commonly bioturbate estuarine sediments. The result is a high-fidelity sediment record that has not been naturally “smoothed” by biologic mixing. The Lower Basin was chosen for this work because it is deeper than the
Upper Basin, and therefore less prone to physical disturbance associated with upper- water annual turnover and wind-induced turnover events that can disturb the sediment record.
The main focus of this manuscript is to determine the environmental history of the Pettaquamscutt River Estuary since the retreat of glacial ice. This work will focus on both the natural changes that have occurred in response to climate and sea level changes over the Late Quaternary as well as more recent changes in ecological conditions that are effected by both natural and anthropogenic influences. In addition to this environmental reconstruction, the manuscript also provides the necessary background work for the following manuscripts by identifying climate-sensitive proxies in the Pettaquamscutt River sediment record.
4 1.22 A note on ages reported in this manuscript
This manuscript spans the time period between the last glacial maximum and present day. In order to maintain chronologic consistency between the proxy records presented here and previous work in the area, it is necessary to express all ages in the same temporal framework. The most common Late Quaternary dating technique employed has been radiocarbon dating of various organic samples. Unfortunately, radiocarbon years are not the same as calendar years due to the fact that the concentration of radiocarbon in the atmosphere has varied through time.
In order to keep chronologic consistency, all ages in this manuscript are expressed as absolute ages. For the upper annually laminated section, ages are given as year A.D. For the older non-varved material (older than ~1000 years) ages are reported as calibrated years before present (cal BP). In the process, the author has calibrated all radiocarbon dates measured using the Calib 5.0 program (Reimer et al.,
2004). Radiocarbon dates reported from previous studies have also been calibrated.
1.23 Late Quaternary deglacial history of southern New England
The Quaternary Period in North America has been characterized by alternating glacial advances and retreats (Ruddiman, 2001). The last glacial maximum (Wisconsinan) occurred as the Laurentide Ice Sheet reached its southernmost extent, which is represented by the Martha’s Vineyard/ Ronkonkoma moraine in the northeast United States. In addition, a number of recessional moraines in southern New England and Long Island show locations where the ice margin was
5 in one place long enough to form an end moraine during the ice retreat. The moraines of southern New England and Long Island, therefore, preserve a record of ice marginal positions during the last deglaciation of the region. Attempts to date these geomorphic features are ongoing, and a brief review of dating techniques will be presented in this section. In addition, the most current and robust deglacial chronology for southern New England (Balco and Schaefer, 2006) will be presented and used to interpret the deglacial history of the Pettaquamscutt River as can be interpreted from the literature.
Traditionally, moraines in New England have been dated by using radiocarbon dates of organic material on or near a given moraine (Boothroyd et al.,
1998; Stone and Borns, 1986). There are a number of limitations, however with this technique. First, since dated organic material did not grow contemporaneously with the glacial feature, these dates are restricted to bracketing ages (Brigham-Grette,
1996). Second, radiocarbon dating assumes a constant initial concentration of radiocarbon in the atmosphere, which is a false assumption (Pilcher, 2003). As a result, radiocarbon years are not the same as calendar years and need to be adjusted using calibration data sets, which themselves have errors and multiple calibration solutions due to plateaus in the calibration curve (Reimer et al., 2004). Finally, there is a problem with finding organic material from the time period of the last glacial maximum due to sparse vegetation at the time of deposition.
An alternative approach, which has been utilized more recently, has been to anchor the extensive New England varve chronology to absolute time (Ridge, 2003,
2004; Ridge and Larsen, 1990). This has been accomplished by radiocarbon dating
6 plant remains that have been identified in varve deposits. Since the organic material is associated with a given varve year, such dates allow the chronology to be anchored in such a way that the absolute age of a given varve can be determined. Minimum dates of deglaciation from sites have been determined by counting the absolute age for the oldest varve found at a given location. In addition, the date of a former ice margin position can be precisely determined in the case where the northern termination of a basal varve can be identified. The varve approach is superior over that of bracketing radiocarbon dates because it sidesteps that problem of sparse vegetation, and can count the varves to the actually year of deposition. In some cases, the approach also gives an actual date of ice-margin location, as opposed to a bracketing age. Since the anchoring dates are radiocarbon dates, this approach still faces the calibration deficiencies that were mentioned above.
Most recently, cosmogenic-nuclide exposure age dating (Gosse and Phillips,
2001) has been utilized in order to date southern New England moraines directly
(Balco and Schaefer, 2006; Balco et al., 2002). This method measures rare isotopes
(10Be and 26Al) that are produced in near-surface rocks by cosmic-ray bombardment.
By measuring the concentrations of these isotopes in large boulders that have not moved since the formation of the moraine, it is possible to determine the exposure age of that boulder face, which should be contemporaneous with the retreat of ice from that location. This approach is superior over bracketing approaches because it dates the actual geomorphic feature, as opposed to organic material found on or near the landform. In addition, the number of candidate boulders for dating is larger than that of organic deposits from the time of deposition. Although the technique
7 circumvents the need to calibrate radiocarbon dates, it has its own assumptions of past isotope production rates, and therefore has its own associated error in its link to the absolute time scale (Gosse and Phillips, 2001).
As might be expected, not all three of the above approaches have yielded identical deglacial chronologies for southern New England. In fact, a ~1700 year discrepancy has been identified between new exposure-age dates and the New
England varve chronology (Balco and Schaefer, 2006). In an attempt to reconcile this difference, Balco and Schaefer (2006) have critically evaluated the uncertainties involved in linking each method’s dates to the absolute time scale. By using the most recent radiocarbon calibration dataset (Reimer et al., 2004), reducing the length of the
Claremont Gap in the New England varve chronology, and adjusting the assumed production rate of 10Be, it is possible to align the New England varve chronology and exposure-age chronology to adapt a new, more robust deglacial chronology of southern New England (Balco and Schaefer, 2006). This chronology includes four adjusted exposure-age dates from southern New England moraines as well as minimum retreat dates from two locations tied into the New England varve chronology (Figure 1.4). The revised dates put the formation of the Martha’s
Vineyard/ Ronkonkoma terminal moraine at ca. 24,000 cal BP. The ice margin retreated to the position of the Buzzards Bay/ Charlestown/ Fishers Island/ Harbor
Hill recessional moraine by ca. 19,200 cal BP. The ice retreated fairly quickly after this time to the positions of the Old Saybrook/ Wolf Rock and Ledyard moraines at ca. 18,790 and ca. 18,740 cal BP, respectively. As the ice margin retreated to the northwest, varve formation began in the Lower Quinnipiac Valley at ca. 18,400 cal
8 BP and in Glacial Lake Hitchcock (at the position of Rocky Hill Dam) at ca. 17,800 cal BP. These two dates are minimum dates for ice margin location.
The composite chronologic framework for southern New England deglaciation can be used to bracket the glacial/post glacial depositional history of the
Pettaquamscutt River. Between 19,200 and 18,790 cal BP, the Narragansett Lobe of the Laurentide ice sheet was retreating through the Pettaquamscutt River valley
(Figure 1.4). The highlands of the valley were left with till deposits from the ice sheet, while the lower portion of the valley was filled with stratified sediment
(Schafer, 1961a, b). Pro-glacial delta and underlying lacustrine deposits have been observed in bore holes from the Pettaquamscutt River valley, which suggests that a pro-glacial lake occupied the valley during glacial retreat, bound to the south by a high portion of the Saugatucket morphosequence (Boothroyd, 1991; Schafer, 1961a, b). In addition to the stratified deposits, there are kettle holes that have been left behind by stagnant ice blocks. These stagnant ice blocks were overlaid by the delta, which was building out into the lake as interpreted by morphosequence interpretations (Gustavson and Boothroyd, 1987; Koteff and Pessl, 1981). Sometime between local ice retreat and 18,790 cal BP (Balco and Schaefer, 2006) the ice lobe retreated far enough North that the melt water began to drain through Hamilton and
Bissel Coves and into Glacial Lake Narragansett (Schafer, 1961a, b).
After melt-water was diverted away from the Pettaquamscutt River Valley, this area officially started its post-glacial phase. Over the next few centuries to millennia, the buried ice blocks melted, causing local collapse of deltaic sediments in areas with underlying blocks. These kettle holes accumulated some volume of water
9 and formed either small lakes or freshwater marshes by at least 13,216 ± 874 cal BP
(Orr and Gaines, 1973). The small lakes and/ or freshwater marshes remained until marine inundation transformed the environment into an estuary. Changes in sulfur concentration from sediment cores have previously been cited as indicating marine inundation, and have been dated at 1694 ± 642 cal BP from bulk organic carbon (Orr and Gaines, 1973).
1.30 Methods
1.31 Sample Acquisition
A total of two piston cores and eight freeze cores were collected between
1999 and 2004 for the work incorporated in this manuscript (Table 1.1). All cores were taken from the deepest hole in the Lower Basin of the Pettaquamscutt River
Estuary, where depth is approximately 19.5 meters (Figure 1.2). The piston corer was necessary in order to penetrate to the basal stratigraphic units and recover the entire post–glacial record. The freeze corer was employed to recover undisturbed upper sediment. These coring techniques complement each other, and result in a complete composite record of post-glacial sediment in the Pettaquamscutt River’s Lower Basin.
The two piston cores obtained from the Pettaquamscutt River were taken using a Kullenburg piston corer (Glew et al., 2001). Briefly, a steel barrel with polycarbonate liner is attached to a core head with lead weight. A piston is then fit into the liner with a piston rope that is designed to keep the piston just above the sediment/water interface during the drive. The apparatus is held via a trigger during descent through the water column, and is designed to release at approximately 0.5
10 meters above the sediment surface. As the trigger releases, the piston is held at the interface while the core barrel is driven past it and into the sediment. This technique was able to recover the entire postglacial sediment package and the upper portion of a sand lithofacies, which represents pro-glacial outwash sands. Due to the force of the initial weight release, however, it is not uncommon to have poor recovery of the upper flocculent sediment.
Freeze cores were taken in order to capture undisturbed sediments from the upper section of the record (Glew et al., 2001). These upper flocculent sediments can either be lost during the piston coring process or disturbed due to interstitial gas expansion as the sediment is removed from the overburden pressure. Freeze coring solves both of these problems by freezing the sediments in situ to the sides of a box that is filled with dry ice and methanol. Briefly, a hollow stainless steel box with tapered bottom is filled with crushed dry ice and methanol, bringing the temperature of the box faces to approximately -78ºC. The coring box is then capped and pushed into the sediment using poles from the surface. The corer is left in the sediment for approximately 15-20 minutes, during which time sediment and interstitial water is frozen to the outside of the coring box. Once the corer is recovered, the four side slabs are removed from the box, wrapped in aluminum foil, and kept on dry ice for transport. The cores are permanently stored in chest freezers in order to maintain the original fabric of the sediment.
Both piston cores and the pre-2004 freeze cores were taken from a pontoon boat platform using a hydraulic winch and A-Frame. In the winter of 2004, a freeze coring operation was conducted from the frozen surface of the Pettaquamscutt River
11 using a tripod and cable hoists. This operation was designed to overcome the limit in penetration with the freeze corer. An initial surface freeze core (NR04-1) was recovered in the same manner as the previous freeze-cores (Table 1.1). A second freeze core (NR04-2) was then taken down the same hole in the sediment in order to recover essentially a second drive, which penetrated further into the sediment. This core was recovered and archived in order to extend the freeze-core record further into the past.
1.32 Core Logging
Both of the piston cores were split in their liners into working and an archive halves. The archive half was cleaned and imaged using a GEOTEK® core logging system. The program generates digital BITMAP images in red/green/blue (RGB) color scheme. These RGB data can either be examined as an image, or quantitatively by extracting the individual pixel RGB values down the length of a core. Both methods can be used to assist in the identification of lithofacies and the boundaries between them.
The archive halves of the piston cores as well as a selection of freeze cores were logged for bulk physical properties using the GEOTEK® core logging system
(King and Peck, 2001; Nowaczyk, 2001; Zolitschka et al., 2001). Piston cores were run through the system in their half-round liners. A strip of frozen sediment was cut out of the center section of one face of selected freeze cores and placed into a plastic
U-Channel sleeve. The sediment was then allowed to thaw before running the cores on the logger. The GEOTEK® logger was used to measure p-wave velocity through
12 the sediment, bulk density (GRAPE), and magnetic susceptibility at one-centimeter resolution down each core. These data are useful in interpreting lithofacies as well as in correlating between different cores.
1.33 Paleomagnetics and Mineral Magnetics
Paleomagnetic directional data were measured from each piston core as well as the freeze cores that were U-Channeled (King and Peck, 2001). Measurements were conducted on a 2-G® cryogenic magnetometer. Each U-Channel was measured at one-centimeter intervals for its natural remnant magnetization (NRM) in each of three main axes. Magnetic overprints can occur associated with coring and subsampling disturbances as well as fields associated with storage location.
Therefore, progressive demagnetization was performed with a diminishing alternating field (AF). After each step, the remanent magnetization was measured at one- centimeter resolution. The resulting data at individual sampling depths was plotted on Zijderveld plots in order to determine the AF field step necessary to “clean” the sample in such a way that the original NRM is represented, free from the influence of post-depositional physical and chemical disturbance. The directional data from this demagnetization step was used to calculate the magnetic inclination, declination, and intensity for each core.
Inclination and declination curves were compared to regional radiocarbon- dated composite curves of inclination and declination (King and Peck, 2001). At regional scales of up to a few thousand miles, paleo-secular variations (PSV) in the magnetic field are correlative between locations. With this premise, curve matching
13 was performed under the constraints of radiocarbon dates from each core to refine the age models of the piston cores. All radiocarbon dates associated with the PSV curves were converted to calibrated ages using the Calib 5.0 program (Reimer et al., 2004).
After the paleomagnetic measurements were made, mineral magnetic properties were measured for each core (King and Channell, 1991; Verosub and
Roberts, 1995). The premise behind mineral magnetics is that laboratory-induced permanent magnetizations of a sample are controlled by the magnetic concentration, magnetic grain size, and magnetic mineralogy of the bulk sample. The standard measurements used are volume magnetic susceptibility (MS) as measured on the
GEOTEK® core logger, mass magnetic susceptibility (X) as computed by dividing
MS by bulk density, Anhysteretic Remanent Magnetization (ARM), and Isothermal
Remnant Magnetization (IRM). The ARM was imparted on each sample with a constant DC field of 1 Oe in the presence of a decreasing alternating magnetic field.
The IRM was imparted by subjecting the samples to a 1.2T DC magnetic field. Both remanences were subsequently measured on a 2-G® cryogenic magnetometer.
1.34 Laminated Sediments and Chronology
In the upper section of the sediment column the sediments are laminated.
Thin sections were made down each freeze core as well as selected sections of two piston cores in order to analyze the laminations in detail. The full procedure for impregnating the sediment with epoxy, fabricating thin sections, and analyzing the laminations can be found in the appendix. For the purposes of this section, it will suffice to describe the lamination thickness measurements. Each thin section was
14 scanned on a flat bed scanner with transparency capabilities under cross-polarizing light. Images were scanned as tagged image format (TIFF) RGB images with 1440 dots per square inch (dpi) resolutions. Each digital image was opened in Adobe
Photoshop® and all lamination boundaries were marked with the path tool. The paths were exported as post script files and converted to thickness measurements using a specialized algorithm (Francus et al., 2002). Lamination boundaries were correlated between cores and a number of recounts were performed in order to determine the errors associated with individual counts.
Since most processes that produce laminated sediments in kettle depressions
(lakes or estuaries) are annually driven, it is hypothesized that the laminations observed in the upper portion of the Pettaquamscutt River sediment column are annually laminated, or varved. The validity of this assumption will be addressed in
Section 1.41 of this manuscript. Assuming that each couplet represents one year, a composite record of couplet number was constructed from the lamination thickness data set. Using eight freeze cores, it was possible to bridge any hiatuses in individual cores due to diffuse laminations or coring breaks. The final composite couplet count was compared to a 210Pb/ 137Cs age model (Lima et al., 2005) and four calibrated radiocarbon terrestrial macrofossil dates to validate the varve hypothesis.
In addition to chronologic control, if the laminations are indeed varves then individual laminae may contain valuable proxy information. Since sediment compacts over time as it is buried by additional sediment, a correction must be made to varve thickness time-series if they are to be used as paleoclimate proxies, however
(Hughen et al., 2000; Nederbragt and Thurow, 2005). As sediment is compacted, the
15 porosity and water content decrease. A correction for water content, therefore, will remove the non-climatic trend associated with differential compaction (Nederbragt and Thurow, 2005). For the Pettaquamscutt River, water content was measured for 1 cubic centimeter samples from freeze cores. It was determined that the percent water content is 93% and the top of the cores, and decreases exponentially to 78% at the base of the varved section. An exponential equation (r2=0.9913) was fit to the data such that:
y = (0.000000002 * e 0.0114x) + 78%, where y is equal to the percent water and x is equal to the year of deposition. This correction was applied to laminae thicknesses in order to normalize all thickness values to 78% water content. All values greater than two standard deviations from the mean were removed in order to diminish the effects of episodic events on the time-series. Most of these thick layers were coarser grained (fine sand), suggesting that they are turbidites and therefore not directly related to precipitation in that given year.
The age model for the rest of the sediment record (i.e. from the piston cores) was constructed using magnetic PSV correlation to calibrated Northeast regional inclination and declination curves (King and Peck, 2001; Reimer et al., 2004) anchored by seven radiocarbon dates. All correlations were consistent between inclination and declination interpretations as well as with physical property correlations between cores. Linear sedimentation rates were assumed between correlated age control points. Age models were constructed based on calibrated
16 radiocarbon dates, correlated PSV ages, and linear sedimentation rates and were used to convert the piston core depth scales to calibrated years before present (cal BP).
1.35 Contaminant Analysis
Organic contaminant concentrations (polycyclic aromatic hydrocarbons
(PAHs), polychlorinated biphenyls (PCBs), and organochlorine pesticides (OCPs)) were analyzed from freeze core NR02-1 at varying resolution down-core. At each sample interval, between 20 and 50 grams of frozen sediment was subsampled with a band saw and placed in a solvent cleaned glass container. Each sample was hand- carried to the URI/GSO Organic Geochemistry Laboratory for analysis.
Organic compounds were extracted with acetonitrile/hexane and fractionated with a 70:30 ratio of hexane and methylene chloride using nitrogen pressure on an activated silica gel column (Hartmann, 2001). PCBs and OCBs were analyzed on a gas chromatograph with electron capture detector (GC-ECD), and PAHs were analyzed on a gas chromatograph with mass spectrometer (GC-MS) (Hartmann et al.,
2005). All organic compound concentrations were normalized to dry weight to yield organic contaminant concentrations relative to sediment dry weight (ng/g). Quality control was assured by analyzing blanks, spiked samples, duplicates, and one standard reference material (Hartmann et al., 2005). At least four quality control samples were analyzed for twenty field samples. The criterion for quality control was between 70% and 130% of expected recoveries.
Trace metal concentrations (Ag, Cd, Cu, As, Cr, Ni, and Pb) were analyzed from freeze core NR99-2 at approximately 2-centimeter intervals. At each sample
17 interval, approximately 1.0 gram of frozen sediment was subsampled with a stainless- steel sampling device, and transferred to an acid-washed polypropylene centrifuge tube. Samples were freeze-dried for at least 48 hours until dry.
Once freeze-dried, all samples were prepared using a total digestion technique
(Lacey, 1997). For each sample the freeze-dried sediment, still in centrifuge tubes, and treated with concentrated hydrochloric (1 mL), nitric (5 mL), and hydrofluoric (4 mL) acids, covered, and placed in a heated sonicator for 48 hours. After cooling, the samples were combined with 30 mL of 5% boric acid solution to neutralize the hydrofluoric acid, and then brought up to volume in 50mL volumetric flasks using deionized water. Samples were stored in acid-cleaned bottles until analysis. Metals concentrations for the total digestions were determined using a Perkin Elmer 4100ZL graphic furnace atomic absorption spectrometer (GFAA) with Zeeman background correction. All metals concentrations (g/L) were normalized to dry weight to yield metals concentrations relative to sediment dry weight (g/g).
Quality control was assured for the digestion process through the use of two standard reference materials, a blank, and one duplicate for each twenty samples measured. Quality control for GFAA performance was assured by measuring one blank and one mid-range standard for each ten samples measured. The criterion for quality control was between 75% and 125% of expected recoveries.
1.36 Stable Isotopes
Stable carbon (13C) and nitrogen (15N) isotope ratios were calculated for samples from freeze cores NR99-2 and NR02-1 in order to yield a ca. 300-year
18 record of these environmental indicators. For each sample (n=73) approximately 0.5 grams of wet sediment was subsampled and dried overnight at 100°C. The dry sediment was ground and homogenized using a mortar and pestle. Between 6 and 10 milligrams of homogenized sediment was weighed into tin capsules for analysis of nitrogen stable isotopes. After the nitrogen subsample was removed, the sediment was treated to remove any carbonates, which could bias the carbon isotope values.
This was accomplished by moistening each samples with 0.1 ml deionized water and acidifying with 0.2 ml of 5.8M hydrochloric acid. Samples were mixed and allowed to react for greater than 12 hours. The acid was diluted with 0.5 ml deionized water and evaporated overnight at 100ºC. Between 0.1 and 0.25 mg of dry sediment was weighed into tin capsules for stable carbon isotope analysis.
Samples were analyzed for stable carbon and nitrogen isotope ratios using continuous flow elemental analysis/isotope ratio mass spectrometry (EA/IRMS)
(Burgess et al., 1996; McKinney et al., 2002). Duplicates and standard reference samples were measured for quality control. All samples are expressed in standard format where:
15 13 N, C (‰) = [(Rsample – Rreference)/Rreference] x 1000,
15 14 13 12 where R is either N/ N or C/ C, and the reference is atmospheric N2 or Pee Dee
Belemnite (PDB) for nitrogen and carbon, respectively.
19 1.37 Aquatic Ecology Proxies
Total organic carbon was determined using the loss on ignition method (Dean,
1974). Briefly, 1 cubic centimeter of sediment was subsampled from a given core and placed in a pre-weighted ceramic crucible. The wet weight was determined for the sediment, and then the sediment was dried in a muffle furnace at 100°C for at least 12 hours. After cooling to room temperature in a dessicator, the samples were reweighed to yield a dry weight. Samples were finally heated at 550°C for one hour.
The mass of the cooled sampled was weighed. The difference between the 550°C mass and the dry weight was the loss-on-ignition (LOI) value, and was directly related to organic matter. Organic carbon concentrations were calculated by multiplying the LOI value by 0.44 and dividing by the dry weight value, which yields normalized organic carbon in mg/g.
Subsamples for fossil pigment analysis over the last millennium were scraped off of frozen archive sediment slabs produced during the thin-sectioning process.
This approach enabled us to precisely determine the age of each sample. A total of
495 samples were analyzed for fossil pigments over the 980-year record. In the case of the older samples, the oxidized surface of the archive half of piston core NR03-3 was scraped off, and a sample was removed under minimal light conditions. For each sample, approximately 0.25 grams of wet sediment was weighed into a glass scintillation vial and placed in a water bath sonicator at 4º C. The pigments were extracted by successive sonication (1 minute) in cold acetone until the extracts were colorless. The combined extracts were filtered through a 0.45 um Acrodisc 13 PTFE membrane filter. Care was taken throughout the extraction process to keep the
20 samples and acetone on ice and in the dark as much as possible in order to preserve the pigments from degrading.
Fossil pigments (Table1.2) were analyzed by high performance liquid chromatography (HPLC) (Bianchi et al., 1996; Wright et al., 1991). The HPLC system consisted of a Waters 2690 Alliance separation module with a 996-photodiode array detector (PDA) and a 474 scanning fluorescence detector with excitation set at
410 nm and emission at 660 nm. Pigments were identified and quantified by comparing retention times and PDA spectra to authentic standards. For the laminated section of the record, pigment concentrations (g/l) were converted to mass accumulation rates (MARs) (g/cm2yr) using dry density measurements and annual varve sedimentation rates. In the lower portion of the record, where the chronologic control is not robust enough to compute MARs, pigment concentrations (g/l) were normalized to dry weight, which yields pigment concentrations relative to sediment dry weight (g/g).
1.38 Pollen
Six subsamples were taken from core NR03-1 for pollen identification. As the goal of this work was to determine the timing of local European land clearance, the six samples were taken from the interval most likely to have this change as estimated from local history (late seventeenth century to eighteenth century). Each sample (1 cubic centimeter) was sieved and pollen was extracted with alternate steps of hydrofluoric, acetic, and sulfuric acids (Faegri and Iversen, 1989; Mecray, 1994).
21 Percent Ambrosia and Rumex were determined by identifying at least 300 total pollen grains per samples, and calculating the associated percentages.
1.40 Results
1.41 Stratigraphy and chronology
The stratigraphy of the Pettaquamscutt River Estuary’s Lower Basin consists of four dominant lithofacies (Figure 1.5). The oldest unit is a medium to coarse- grained sand facies with no organic matter preserved. Overlying the sand is organic- poor sandy silt. This grades upward to organic-rich, gassy, and partially laminated/banded mud. The uppermost unit is a well-laminated organic-rich mud.
Due to the tendency for Kullenburg cores to disturb the upper portion of sediment, it is necessary to compare core NR03-4 to upper freeze cores from the same location.
This was accomplished by examining three physical properties with depth: ARM
(Figure 1.6), magnetic susceptibility (Figure 1.7), and bulk density (Figure 1.8). Tie lines consistent between these physical properties and distinct lamination locations between cores were drawn on each figure. Examination of the three plots confirms that the piston cores did not recover approximately 1 meter of sediment and compressed the upper portion of core by ca. 40%. In addition, piston core NR03-3 represents an expanded section as compared to piston core NR03-4. This is most likely due to the extra weight used for the NR03-4 drive resulting in additional compression. The difference may also demonstrate slight differences in sedimentation rate throughout the basin.
22 The upper, laminated portion of the sediment record is hypothesized to be annually laminated, or varved (Figure 1.9). This hypothesis is tested by comparing the varve age model to independently measured age constraints. Figure 1.10 plots varve age versus composite core depth for the eight freeze cores analyzed. The average sedimentation rate for this varved section of sediment is 0.18cm/year. Three independent age controls are also provided. First, a 210Pb age model provides high- resolution age control for the past two centuries (Lima et al., 2005). Second, two peaks of anthropogenically introduced 137Cs concentrations are observed in the record
(Lima et al., 2005). The bottom peak is attributed to atmospheric bomb testing, which culminated in 1963. The upper peak represents fallout from the Chernobyl incident, which occurred in 1986. Finally, four calibrated radiocarbon ages (Table
1.3) are plotted. Each of these is from terrestrial macrofossils sampled from freeze cores. Terrestrial macrofossils were used in order to avoid reservoir corrections that would be necessary for aquatic matter (Björck and Wohlfarth, 2001).
The varve age model from the Pettaquamscutt River fits within all of the independent age constraints, with the exception of the radiocarbon date at 78cm. This macrofossil was obtained from the same horizon as the Ambrosia and Rumex pollen horizons (Section 1.44), which represents the timing of land clearance by European settlers. During land clearance, older carbon that had been sequestered in the water shed can be remobilized and transported to the water body. Therefore, it is likely that this anomalously old date is due to reworking of a terrestrial macrofossil, and can be dismissed in light of the weight of all other evidence.
23 The age model for the piston cores has been constructed by tying the upper portion into the varve age model for an upper age (top of NR03-4 is ca. 400 cal BP), and through radiocarbon-anchored paleomagnetic correlation for the remainder of the record. Seven calibrated radiocarbon dates were calculated for terrestrial macrofossils sampled from the two piston cores (Table 1.4). These dates were used as temporal constraints as declination (Figure 1.11) and inclination (Figure 1.12) records from piston cores NR03-3 and NR03-4 are matched to the Northeast regional
PSV curves (King and Peck, 2001). The paleomagnetic data reported were demagnetized at 150 Oe in order to remove magnetic overprints on the data (Figure
1.13). With the exception of the 8833 ± 200 cal BP date (Lab ID # 53489; Table 1.4), all PSV interpretations fit within the constraints of calibrated radiocarbon dates. The lack of consistency for this particular radiocarbon age is supported due to the strength of correlation to the regional declination peak at ca. 8350 cal BP. This feature is robust enough in the PSV record, that it is not reasonable to ignore it. The discrepancy is ca. 300 years outside the range of error. The most reasonable explanation is that the date is anomalously old due to sample contamination and/or an older macrofossil being washed into the pond well.
Final age models for cores NR03-3 and NR03-4 are shown in Figures 1.14 and 1.15, respectively. Basal ages were determined by assuming a constant sedimentation rate and extending the linear trend of the oldest six age control points to the base of each core. In the case of core NR03-4, the contact between the basal sand lithofacies and overlying organic-poor sandy silt is at 460 cm core depth (Figure
1.5). Using a 200 year uncertainty for the PSV dates (King and Peck, 2001) and a
24 linear extrapolation of the sedimentation rate for that core, the age of this contact is determined to be 19,170 ± 570 cal BP. The average sedimentation rate for the non- varved sediment is 0.025cm/year, an order of magnitude less than for the modern varved sediment. Correlations to the regional PSV records are supported by plotting piston core declination and inclination on age scales, and comparing to the original regional curves (Figures 1.16, 1.17). In each case, main features of the records are correlative, further supporting the PSV interpretations.
1.42 Physical and magnetic proxies
Bulk physical and magnetic properties of the Pettaquamscutt River Estuary provide more in-depth information on the major lithofacies identified in Section 1.41
(Figures 1.18 through 1.21). The basal unit contains light-colored sands with high bulk density values. Mineral magnetic concentration proxies (magnetic susceptibility
(MS), anhysteretic remanent magnetization (ARM), and saturation isothermal remanent magnetization (SIRM)) generally show high values during this time. ARM values do not show the relationship as well as MS and SIRM. Since ARM is biased toward fine magnetic mineral grains, this discrepancy can be explained by a higher proportion of coarse-grained magnetic minerals being deposited during this time.
This interpretation is confirmed by examining interparametric ratios, which act as magnetic grain size proxies (Figure 1.20, 1.21). The ratios of SIRM/ARM and
ARM/MS both exhibit coarser sediments in the basal section of the core than during other times. SIRM/MS ratios suggest coarse grains, however the magnitude is not dramatically different from younger times.
25 As finer-grained sedimentation commenced, the color darkened, bulk density decreased, and magnetic concentration of the sediment all decreased. This likely is indicative of the onset of lacustrine deposition, and the gradually increased organic influence on the sediment. Finer magnetic grain sizes confirm a shift to a lacustrine depositional environment. Further up in the core, there is a transition zone most obviously seen in reduced ARM values. At the same time, the color lightens, density increases, and magnetic grain size coarsens and becomes more variable. Finally, the upper unit exhibits lighter color, higher density, and mixed signals on magnetic concentration and grain size parameters. This zone is the same as the upper well- laminated lithofacies discussed in Section 1.42.
The conflicting behavior of some of the magnetic proxies (i.e. grain size and upper IRM values) suggests that the interpretation of the magnetic properties of the
Pettaquamscutt River Estuary is not as straight forward as some traditional locations
(Thompson and Oldfield, 1986). The observed behavior is most likely due to the presence of large concentrations of sulfides in the water column and sediments. The anoxic environment can reduce the basic iron oxides (magnetite, hematite) that traditionally have defined many of the magnetic proxy measurements. If fully reduced, the iron oxides will degrade to paramagnetic pyrite. Backscatter electron imaging (BSEI) of thin-sectioned varved estuarine sediments from the Pettaquamscutt
River revealed framboidal pyrite grains, confirming the significance of reduction diagenesis in these sediments (Figure 1.22). In addition to pyrite, different redox conditions and availabilities of sulfide can produce two intermediate magnetic minerals: pyrrhotite and greigite (Snowball and Torii, 1999). One method that can be
26 used to test for iron sulfides uses biplots to plot data and compare them to experimentally determined values for pure samples. Figure 1.23 plots the
ARM(400)/ARM(0) vs. SIRM/X from core NR03-4 overlaid on experimentally- derived values for different pure magnetic minerals (Peters and Thompson, 1998).
This analysis shows that pyrrhotite is not a significant component of the
Pettaquamscutt River sediments, but that the sediments are likely a mixture of greigite and magnetite. In addition to the total data, it is instructive to examine the locations of samples from different lithofacies/depositional environments. There is a general “softening” of magnetic minerals in the transition from lacustrine to estuarine sediments (Figure 1.23). This trend suggests that remanences are less stable in the estuarine unit, perhaps due to increased availability of sulfide associated with marine waters.
In the most recent thousand years of the record, the sediments are varved. The thicknesses of the individual laminae can be used as a high-resolution proxy during the time period. The time-series of total varve thickness, clastic laminae thickness, and biogenic laminae thickness are shown in Figure 1.24. In order to account for differential compaction of the sediment, all thickness values are normalized to 78% water content, which yields compaction-corrected time-series of varve and laminae thickness (Nederbragt and Thurow, 2005) (Section 1.34). All three of the time-series show variability on a number of time-scales. The clastic laminae are formed via runoff in the watershed transporting sediment derived from till and outwash deposits to the estuary. The biogenic laminae are deposited during the spring and summer months as biologic productivity increases. These layers are predominantly composed
27 of amorphic organic material, which derives from these annual productivity blooms.
Therefore, the normalized thicknesses of laminae thicknesses can be used as proxies for runoff and biologic productivity. Chapter 3 will address the use of the clastic laminae thicknesses as a climate proxy over the last millennium.
1.43 Chemical proxies
This section mostly addresses organic and inorganic pollutant records as recorded by the Pettaquamscutt River sediments. In addition to discussion of trends, potential ecologic effects will be mentioned here and discussed further in Section
1.53. For the purpose of this section, an introduction to sediment quality criteria is required. This thesis will use the NOAA system of ERL/ERM values to address this
(Long et al., 1995). Briefly, these are contaminant concentrations that have been found to have potential adverse ecologic effects in estuarine and marine environments. The ERL (effects range low) value is defined as the experimentally determined concentration at which 10% of the test organisms displayed adverse effects. The ERM value (effects range median) is the concentration at which 50% of the test organisms displayed adverse effects. All applicable ERL and ERM values are discussed and appear on appropriate graphs. All values are from Long et al. (1995) and are listed in Table 1.5.
The total PAH concentration started its rise in the middle of the nineteenth century as coal combustion was used for heating purposes, and exceeded the ERL of
4000 ng/g ca. 1890 (Figure 1.25). The concentrations follow a relatively common profile, with a maximum concentrations in the 1920s and 1930s associated with the
28 Great Depression (Lima et al., 2003) and at ca. 1973, coincident with the OPEC oil embargo of that year (Doniger et al., 2002). A general decrease in concentration has occurred since 1973 due to better environmental controls on the burning of fossil fuels. The PAH profile is similar in many regards to a higher-resolution record taken from the same location (Lima et al., 2003) (Figure 1.25). However, a consistent offset is observed between the two data sets, especially during the ca. 1973 peak, with this study’s values showing lower concentrations. Most of the other differing features are due to the lower resolution of this study’s record not picking up detail as well as the higher resolution record does. Some of the anomalous features in the current study (i.e. ~1929 A.D.) are likely due to analytical errors associated with the analyses since the age models are robust enough to eliminate chronologic error as the cause.
Beside the general decreasing PAH concentrations trend over the last three decades, a minor increase in PAH concentration is observed in 1999, consistent with previous work (Lima et al., 2003) (Figure 1.25). In that publication this increase was the subject of great interest due to the potential increase of motor vehicle pollution in the area. The data presented in this work cast doubt on its significance and suggest that the increase is merely the result of an anomalous year. First, the only PAH component that contributes to the increase in 1999 in this study is 2,6- demethylnaphthalene (DMN) (Figure 1.25), which has questionable accuracy from these samples (Quinn, 2006). Therefore, the reproducibility of the 1999 peak is cast into question. Second, if it is accepted that the small increase in DMN from 1999 is correlated to the Lima et al. peak in 1999, two additional years of deposition allow this study to note that in 2000 and again in 2001, the total PAH concentration
29 decreases. Lima et al. (2003) argue that they see the increase represented by two data points that, they argue, represent multiple years. These two samples come from the top centimeter of sediment, however, and subsequent varve age constraints suggest that the upper 1-1.5 centimeters of sediment in the Pettaquamscutt River represent uncompacted sediment from the current year (Figure 1.9). Therefore, the two upper measurements in the previous study likely represent concentration from the same year of deposition (1999). As the current data has shown that no trend is evident, and that the 1999 value is not significantly different from the trend, it is proposed here that the argument for increased energy consumption leading to increased PAH flux is not justified.
Retene is a PAH which does not follow the general trend expected for anthropogenically introduced PAH compounds as it has a subsurface peak at about
1840 A.D. (Figure 1.26). The figure displays this study’s data as well as the higher resolution data of Lima et al. (2003) so that the full shape of the subsurface peak can be realized. This PAH is not necessarily associated with combustion, rather it is also produced by the natural degradation of abietic acid present in coniferous resins
(Tavendale et al., 1997). The timing of the retene peak is coincident with the time period of John A. Saunders’ ship building operation in the Pettaquamscutt River
(1813-1854) (Figgins, 2004). Although the structural components of ships were usually constructed of hard woods, the planking was often composed of available soft wood, such as white pine. In addition, pine tar was commonly used as a preservative for both wood and standing rigging of traditionally rigged sailing vessels. Because retene is associated with coniferous resins (Tavendale et al., 1997), and because the
30 timing of shipbuilding in the estuary is concomitant, the probable origin of the
Pettaquamscutt River subsurface peak in retene is pine tar and/or sawdust that was introduced to the estuary during this time. This conclusion is contrary to the findings of Lima et al (2003), who dismissed the linkage to shipbuilding due to chronologic mismatches. Recently, however, the age model used by Lima et al. (2003) has been proven inaccurate with the application of varve counts (Lima et al., 2005). When plotted with the revised age model (Figure 1.26) it is evident that the peak is the same as that observed in this study. The Lima et al. (2003) data appear to rise before the onset of Saunders’ operation, however this is being driven by one data point (1808
A.D.). It is possible that smaller boat building operations predate Saunders’ operation, that a natural source of retene was available from pine trees, or that the data point is an insignificant point, however the data do not allow for the resolution of this point. It is also interesting to note that the concentration of retene did not reach background levels until about 1890 A.D. Since shipbuilding ceased in 1854, this means that it took approximately 35 years for the residual coniferous flux to stabilize.
Polychlorinated biphenyls (PCBs) are manmade organic compounds most often used in the manufacturing of electrical capacitors and transformers, and were analyzed in the Pettaquamscutt River sediment record over the last ca. 180 years
(Figure 1.27). The total PCB data fit the standard trend of increased concentration from levels below the detection limit (<0.5 ng/g) before 1929 A.D. associated with the start of PCB manufacturing, to a peak concentration in the mid 1970s associated with bans in the United States, and continued decrease in concentration since the ban.
The record shows that concentrations first exceeded the ERL (23ng/g) in the mid
31 1930s and only in 2000 dropped below the ERL again. The concentration of PCBs never exceeded the ERM value of 180 ng/g. First, this record demonstrates that the toxicity of PCBs in the Pettaquamscutt River never reached high levels as defined by the sediment quality criterion (Long et al., 1995). Second, it demonstrates the persistence of these compounds in the environment as it took over 25 years for the concentrations in the sediment to reach close to background concentrations due to the deposition of sediment with lower concentrations of PCBs. It should also be noted that the dominant component to the PCB record is hexachlorobiphenyl almost exclusively throughout the record (Figure 1.27). This compound consists of the congeners CB138, CB163, and CB164, but the three congeners could not be separated by the techniques used in this study.
Dichloro-diphenyl-trichloroethane (DDT) is a persistent organic pollutant manufactured as a pesticide in the United States from ca. 1940 to 1972. It’s toxicity is well known, and due to its persistent nature, it can still be found in depositional environments more than 3 decades after the national ban took effect in 1972. In the
Pettaquamscutt River sediments, concentrations rose above background levels in
1940 (Figure 1.28). The rapid increase in concentration is most likely due to local
DDT spraying, which would have increased the efficiency of transport to the estuary via atmospheric fallout. Due to its high toxicity, the ERL value is only 1.6 ng/g
(Long et al., 1995), and it therefore exceeded the ERL value immediately. In the early 1940s the concentration increased rapidly and surpassed the ERM value (46 ng/g) before 1950. The concentration did not fall below the ERM again until ca.
1980, almost a decade after the national ban. DDT concentrations have continued to
32 decline, however modern concentrations are still above the ERL value (Long et al.,
1995). The most likely modern source of this organic pollutant is through runoff in the watershed remobilizing previously deposited contaminated sediment.
Trace metal concentration time-series are plotted in Figure 1.29. The rise from metals background is most evident for lead and silver, and occurs at around
1860 A.D. This timing postdates the Saunders’ shipbuilding operation (Figgins,
2004), which suggests that that this operation did not introduce metals to the environment, contrary to the conclusions of Mecray et al. (1991). All of the metals increases are modest over the length of the record. Lead exhibits the largest increase, and the primary source is most likely atmospheric. It is interesting to note that the common decrease in lead concentrations associated with the lead gasoline ban of the
1970s is short-lived in this record. In the mid 1980s the concentration started an increasing trend that continues to the present. This trend suggests an additional local source; perhaps runoff contributions from soil that has been in contact with lead paint and/or lead pipes. The other points of interest are subsurface peaks in copper (ca.
1880-1905 A.D.) and chromium (ca. 1875-1900 A.D.). The chromium peak is joined by a small peak in nickel at the same time, which is insignificant compared to the background values for nickel. The source of these metals at the end of the nineteenth century is likely to have been local. Upstream on the Mattatuxet River is the former
Shady Lea mill complex. This property was purchased by Robert Rodman in 1870 and used to manufacture warps and jeans (Cole, 1889). In fact he expanded the operations that had been operating at this location and built a new large mill in 1877, which still stands today. Since metals such as chromium and copper were used to dye
33 textiles during the time (Mecray et al., 1991) it is reasonable to attribute the subsurface spikes of these two metals to Mr. Rodman’s textile mill upstream.
None of the seven metals measured in the Pettaquamscutt River core have had concentrations that have exceeded ERM values over the span of this record (Table
1.5) (Long et al., 1995). Cadmium, copper, chromium, and nickel have all had periods in which the concentration has exceeded the ERL value (Figure 1.29).
Cadmium and copper are the only metals currently exceeding the ERL, demonstrating that the metals pollution in the Pettaquamscutt River is minimal. The current source of these metals is likely coming from remobilization of sediment in the watershed that has experienced historic input of these metals. Even the subsurface peaks in copper and chromium are low in this system, with the copper maximum concentration of
86.3 g/g (ERM=270 g/g) and chromium maximum concentration of 172.6 g/g
(ERM=370 g/g).
Finally, stable nitrogen (15N) and stable carbon (13C) isotope values, coupled with the carbon to nitrogen ratios (C/N) provide biochemical information on organic matter in the estuary (Figure 1.30). First, the origin of the organic matter can be determined by comparing the 13C values to the corresponding C/N ratios. The
13C value for the Pettaquamscutt River record fluctuates between –26‰ and -25‰, indicating an aquatic algae and/or C3 land plant source due to the depleted values
(Meyers and Teranes, 2001). The C/N ratio helps to narrow the source further as aquatic algae generally yield values between 4 and 10, whereas C3 land plants usually have values above 20 due to their richness in cellulose (Meyers and Teranes, 2001), and can become more enriched as leaf litter decays (Taylor et al., 1989). The values
34 in the Pettaquamscutt River of 8-12 indicate aquatic algae as the dominant source of organic matter to the sediment. There must be some component of terrestrial organic material, however, to shift the C/N value above 10.
In addition, the 15N value can provide information on the source of this nutrient to the estuary (Talbot, 2001). Since estuarine ecosystems tend to be nitrogen-limited (Boynton et al., 1982), nitrogen inputs and dynamics are important in understanding ecologic variability as it may be associated to nutrient changes. For the early part of the Pettaquamscutt River record, the mean 15N was 3.7‰. As the major source of Pettaquamscutt River freshwater input is from groundwater (Kelly and Moran, 2002; Urish, 1991), this background value is within the range of values reported for natural groundwater sources (2-8‰) (Kreitler and Browning, 1983).
Between 1950 and 1960, there is a statistically significant (p<0.0001) increase in the mean 15N value to 5.4‰. Although this shift is within the range of reported average groundwater values, it is significantly different from the background Pettaquamscutt
River value of 3.7‰. This increase suggests that an isotopically heavy source of nitrogen became more influential during the 1950s. The most likely candidate for this shift is an increase of septic systems in the watershed. Previous work has demonstrated that denitrification and volatilization of ammonia within leaching fields can result in 15N values between 10 and 22‰ in the subsequent groundwater
(Kreitler and Browning, 1983). The timing of this shift lends further support to this mechanism as major residential development occurred in the watershed during this time, with primarily individual septic systems (Ernst et al., 1999). This discussion provides strong evidence for cultural eutrophication (Conley et al., 2002; Gorham and
35 Sanger, 1976) over the last 4-5 decades in the Pettaquamscutt River. The ecological impacts of this eutrophication are addressed in Section 1.54.
1.44 Biologic proxies
In addition to physical and chemical proxy records of environmental change, it is important to examine biologic records of change to understand the ecologic response to both natural and anthropogenic forcings. This section will focus on aquatic biologic proxies, but will also discuss terrestrial pollen changes in the region.
The most basic measure of biologic productivity is the total organic carbon (TOC) content of dated sediment (Meyers and Teranes, 2001). This proxy gives a good first approximation of productivity changes, however it can be biased by the organic carbon source. Since the C/N ratio of sediment is mostly of algae origin (Figure
1.30), the tendency is to interpret changes in TOC in the Pettaquamscutt River as indicative of changes in aquatic productivity. However, terrestrial influence may have changed over time.
Nonetheless, TOC can be used to assess changes in productivity since deglaciation (Figure 1.31). As might be expected, the organic carbon content of sands deposited after an ice block collapse is very low. As the sediment grades up to finer compositions and darker colors, the productivity and/or preservation rapidly increases to ca. 90 mg/g by 15,500 cal BP. Over the next 5200 years, the productivity in the water body increased steadily until it reached modern values of about 120mg/g at 10,300 cal BP. In addition to the trends, there are various low and high frequency components to the record. The trend components are interpreted as being driven by
36 ecologic succession and long-term climate change. The other components of the time-series likely have climate origins.
A more detailed examination of TOC is afforded over the last millennium through the varved sediment section (Figure 1.32). At this scale, TOC continues to exhibit variability at a number of different scales. Over the last millennium, however, the concentrations have remained between about 90 mg/g and 120 mg/g. These values have been the norm since ca. 10,200 cal BP (Figure 1.31). The general pattern over the last thousand years is defined by declining productivity from 1100 A.D. to about 1840 A.D. A rapid rise is observed in the middle of the nineteenth century, followed by a drop at ca. 1930. Productivity has increased from 1950 to present, with relatively high amplitude fluctuations superimposed on the trend.
Since organic carbon data do not differentiate between sources of organic matter, it is important to also analyze proxy data with known sources, called biomarkers. To this end, fossil pigments have been analyzed and reported either normalized to sediment dry weight for older sediments with low-resolution chronologic control, or as mass accumulation rates (MARs), which corrects for sedimentation rate and density changes. Reported in this manuscript are concentrations and MARs of seven identified pigments (Table 1.2). Each pigment has an affinity toward one or more taxonomic groups, and can therefore be used as biomarkers of productivity as associated with the specific taxonomic group (Leavitt and Hodgson, 2001).
The full record of fossil pigments is rather different from the TOC record
(Figure 1.33). The main difference is low values during most of the record. Since
37 these organic compounds are highly labile in oxic environments compared to other organic compounds such as the contaminants measured in this work (Leavitt, 1993) these low concentrations are most likely indicative of poor preservation associated with an oxygenated water column. It is possible to interpret that productivity was much lower, but then the organic carbon values would follow the same trend, which they do not (Figure 1.31). The environment did not become conducive to the preservation of most pigments until ca. 1000 years ago (core depth in NR03-4 of
~80cm). The environment needed for quality preservation is an anoxic one, and so this timing is indicative of the onset of modern density-stratified waters in the estuary
(Figure 1.3). The other point to note is that there is an unidentified carotenoid that has been identified in the older section of the record (Figure 1.32). This pigment was the only significant pigment associated with sediment deposited between 15,500 and
12,800 cal BP, and low concentrations persist to ca. 1800 cal BP. This unidentified pigment is unique in its preservation as well as its decrease in concentration from
12,800 cal BP to present. Most likely a cold-water species produced this stable carotenoid, and was then depleted in numbers through succession as the climate warmed coming out of the last glacial period.
A high temporal resolution analysis was performed to determine fossil pigment mass accumulation rates in the varved record, with average sampling resolution of 2 years (Figure 1.33). It is clear from the figure that Bacteriochlorophyll e is the major component of the Pettaquamscutt River fossil pigments. This is expected due to the high concentration of biomass just below the oxycline of the water column (Sieburth and Donaghay, 1993). The pigment concentrations exhibit
38 variability on a range of timescales, presumably associated with both climate and anthropogenic forcings. The climate influence on aquatic productivity will be addressed in Chapter 2 by examining the Bacteriochlorophyll e time-series in detail.
Anthropogenic influences will be examined in Sections 1.52, 1.53, and 1.54.
Finally, information can be obtained on terrestrial ecologic shifts by examining pollen grains preserved in the sediment column. For the purpose of this manuscript, two pollen types were identified: Ambrosia and Rumex (Figure 1.35).
The rise from background for Ambrosia (ragweed) and the first occurrence of Rumex
(sorrel) are commonly associated with European land clearance in a given region
(Francis and Foster, 2001). Although a native species, Ambrosia prefers clear fields to forests, so the concentrations increase quickly after land clearance. Rumex is not native to North America, and therefore the occurrence of this pollen type is closely tied to land use changes caused by European settlers. In the Pettaquamscutt River
Estuary, both pollen types suggest land clearance by European settlers starting at the very end of the seventeenth century (Figure 1.35). This postdates King Philip’s War
(1675-1676), and occurs at roughly the same time that Rhode Island merchants were permitted to participate in the slave trade (Davis, 2006). Assuming that slaves cleared the land surrounding the Pettaquamscutt River as plantations were developed in the region, this timing agrees with the pollen horizons.
39 1.50 Discussion
1.51 Geologic evolution of the Lower Basin
The Late Quaternary geologic history of The Pettaquamscutt River Lower
Basin has been interpreted and dated in order to refine this history since inception of an aquatic environment following ice block melting (Figure 1.36). The precursor to the Lower Basin of the Pettaquamscutt River was as an ice block buried in pro-glacial outwash sands at ca. 19,200 to 18,790 cal BP. This delta presumably prograded into
Glacial Lake Pettaquamscutt and buried stagnant ice blocks in the process
(Boothroyd, 1991). At this time, a pro-glacial delta was also prograding into Glacial
Lake Narragansett in the vicinity of the present Jamestown Verrazzano Bridge (Peck and McMaster, 1991). The elevation of this topset/foreset contact is at –5 meters, and represents the surface elevation of this proglacial lake. Adjusting the land for an isostatic uplift gradient of 0.9 meters per kilometer (Koteff and Larsen, 1989) yields an elevation of –14 meters for the water surface of Glacial Lake Narragansett at the current mouth of the Pettaquamscutt River Estuary. Although possible, it is unlikely that the sill at the mouth of the Pettaquamscutt River was this low. This elevation suggests that Glacial Lake Pettaquamscutt was separate from and upland of Glacial
Lake Narragansett.
As the buried ice blocks melted, kettle depressions formed and the inception of the Lower Basin kettle pond occurred (ca. 19,170 ± 570 cal BP). Although the calculated age is based on a constant sedimentation rate assumption (Section 1.41), the timing is consistent with local ice-recession constraints of between 19,200 and
18,790 cal BP (Balco and Schaefer, 2006) as well as a basal radiocarbon date from
40 the Pequot Cedar Swamp, just north of the Ledyard moraine, dated at 18,710 – 18,560 cal BP (McWeeney, 1995). During this initial stage of deposition, magnetic susceptibility is high and total organic carbon is low (Figure 1.36). These observations suggest a cold environment with limited terrestrial vegetation, which is consistent with previous studies from the region (Boothroyd et al., 1998). Modern levels of organic carbon and magnetic susceptibility were reached by ca. 15,500 cal
BP, suggesting a warming climate and inception of mature terrestrial vegetation during this period, which helped to stabilize the sediment in the watershed. This timing is old as compared with previous work in the Pettaquamscutt River (Gaines,
1975), however it agrees well with an inception date of lacustrine sedimentation at
Great Swamp, South Kingstown, RI at ca. 14,970 ± 2600 cal BP (Freedman, 1997) .
Over the next 14,500 years the kettle housed a well-mixed, oxygenated pond with relatively high productivity. Cyclic sedimentation occurred as evidenced by organic carbon concentration and the occurrence of banding and lamination in these sediments. The relatively high organic load led to high metabolism in the sediment, and the high gas content of this lithofacies.
As sea level rose in the late Holocene, marine waters eventually inundated the
Lower Basin. Although previous work suggests this transition happened at ca. 1694
± 642 cal BP (Orr and Gaines, 1973), the current study provides evidence that permanent stratification did not occur until 1026 A.D. (i.e. 975 cal BP). This evidence comes from the inception of varved sediments, which requires zero bioturbation, and preservation of photosynthetic pigments, which are labile in oxic environments (Leavitt, 1993). The age discrepancy between previous work and this
41 study is relatively small, and can be explained by one (or both) of two mechanisms.
First, there must have been a lag between the time that marine waters first entered the estuary, and when permanent density stratification commenced. This time lag, however, was most likely on the order of months to years, and therefore may not explain discrepancies of decades to centuries. Reconnaissance diatom analysis found only fresh water tests immediately below the varved sediments (76 cm) in core
NR03-4 (Hargraves, 29 September 2003), which confirms a small time lag.
Therefore, it is likely that marine waters did not inundate the Lower Basin of the
Pettaquamscutt River Estuary until the late tenth or early eleventh centuries (ca. 1000 cal BP). Secondly, the previous radiocarbon date was taken on bulk sediment, which can yield anomalously old dates due to reservoir effects associated with saline waters
(Pilcher, 2003). Therefore, there are questions as to the robustness of the previous estimation of marine inundation.
1.52 European settlement and environmental change
As discussed in Section 1.44, the timing of European settlement in the region is dated at ca. 1700 A.D. This is based upon identifying the rise from background for the two indicator pollen grains Ambrosia and Rumex. In this section, an analysis will be made to determine if any estuarine response to this environmental change is preserved in the record. Varve thickness, magnetic, and fossil pigment records provide high-resolution data bracketing this time period.
From a physical point of view, it might be expected that land clearance will result in increased runoff and sediment supply to the water body. This hypothesis is
42 supported by examining the varve thickness and magnetic data (Figure 1.37). The total varve thickness, as well as the individual laminae, shows an increase in thickness at ca. 1695 A.D., in line with the rise from background pollen concentrations. For the next two decades, all three varve thickness proxies exhibit increased values, with large amplitude variations. At ca. 1715 A.D. the thickness values decrease to pre-European values. In addition to visual identification of anthropogenic influence, t-tests provide statistical support demonstrating that the varve, clastic laminae, and biogenic laminae thicknesses are significantly higher during the period 1695-1715 A.D. than the typical variability over the last millennium
(p= 0.0013, p=0.0253, p=0.0420, respectively). In line with the shifts in lamination thicknesses are step increases in magnetic susceptibility and ARM, and a spike in the magnetic interparametric ratio ARM/MS (Figure 1.37). Magnetic susceptibility and
ARM are proxies for magnetic concentration, biased toward coarse and fine magnetic grains, respectively. The step increase in both of these proxies is further supporting evidence of soil erosion as soils typically have magnetic minerals associated with them (Thompson and Oldfield, 1986). Finally, the spike in ARM/MS is indicative of a spike of fine magnetic grains. This is also expected from soils, as single domain magnetic is commonly the mineral found in soils. The magnetic proxy signals postdate the initial signal in lamination thicknesses by about 15 years, however they are contemporaneous with the largest spikes in varve and clastic lamination thicknesses.
Of the lamination proxies, the clastic laminae are the most directly influenced by increased sediment supply since they are composed primarily of siliciclastic
43 material. Total varve thickness has the highest significance of the three, but the clastic layer may drive this. The biologic layer is an interesting case because the mechanism that is increasing the thicknesses is less straightforward. Since the biogenic layer consists of both amorphic organic material and smaller concentrations of clastic material, either an increase in biological productivity and/or increases runoff could account for the change. In order to address this, the pollen record is also compared to high-resolution fossil pigment data to test the increased productivity hypothesis.
Figure 1.38 plots the mass accumulation rates of the seven main classes of fossil pigments (Table 1.2) during the two centuries bracketing the time of European influence. With the exception of bacteriochlorophyll e, all pigments exhibit some sort of peak during the time period 1701-1711 A.D. Green sulfur bacteria are the only organisms represented that are restricted to being below the oxycline. Therefore, the drop in concentration for bacteriochlorophyll e is most likely due to increased turbidity associated with increased biomass and/or sediment in the water column.
The timing noted on the pigment response is shorter lived then that seen in the varve record. The difference in duration is perhaps due to a lag time associated with the momentum of the biologic system.
In order to test whether or not the watershed clearance had an effect on the estuarine productivity, t-tests were performed to compare the 1701-1711 A.D. period
MARs to the rest of the record. Significant increases in productivity were observed in beta-Carotene (p=0.018), Alloxantin (p=0.0022), and Lutein (p=0.0289).
Comparing these MARs to MARs during the seventeenth and eighteenth centuries
44 yields higher significances for the above three pigments (p=0.0002, p=0.0045, p=0.0005, respectively) as well as significance in the increased Zeaxanthin MAR
(p=0.0004). Beta-Carotene and Lutein are both biomarkers for green algae, while
Alloxantin is produced by cryptophytes, and Zeaxanthin is produced by cyanobacteria. Therefore, the productivity spike observed concomitant with
European watershed clearance appears to be represented across the primary producers of the Pettaquamscutt River. The exception to this is the green sulfur bacteria, which presumably become light-limited during this time. The lack of significant change in chlorophyll a (green algae) and steryl chlorin esters (zooplankton grazing) suggest that the anthropogenic effect was not felt throughout the ecosystem.
These analyses provide solid evidence that land clearance changes in small watersheds can have significant consequences to erosion and sediment transport in the watershed as well as productivity in the water column. The increased productivity is most likely due to increased nutrient transport to the water as soil is eroded. It is interesting to note that this anthropogenic signal only lasted between one and two decades, depending on the proxy used. This duration is important information for two reasons. First, this duration provides information that should help with forecasting of land clearance effects in small watersheds. In addition, it demonstrates that the proxy data that will be used to infer past climate conditions in subsequent chapters have not been tainted by anthropogenic influence past about two decades from European arrival and extensive land clearance.
45 1.53 Estuarine response to anthropogenic metals and organics pollution
Trace metals and organic pollutants are known to have adverse ecologic effects in estuaries if concentrations are high enough (Clark, 1997). A summary of
Pettaquamscutt River contaminants that exceed Effects Range guidelines (ERL,
ERM) is shown in Figure 1.39 (Long et al., 1995). It is immediately obvious that only DDT has exceeded the ERM value, and therefore most pollutants probably did not have an impact on the estuarine ecosystem based on individual Effects Range values. With such a range of stressors, however, it is difficult to test this hypothesis.
One recent method used to address the problem of multiple pollutants is the mean sediment quality guideline quotient (Long et al., 2006). In this method, a quotient is calculated for each chemical, which is the percentage of the ERM value that the concentration represents. Average quotients are calculated for each of three pollutant classes (metals, PAHs, chlorinated organics), and these three are then averaged to produce a mean ERM quotient (ERMQ) for each year. The ERMQ represents the magnitude of pollution for a given year and can be compared to productivity proxies from the sediment record (Figure 1.40). The main peak in
ERMQ occurs between about 1966 and 1974 and is driven by the DDT concentrations during this time (Figure 1.27). This peak ERMQ has a value of 0.8, only 80% of the value that would probably yield biologic effects. Examination of the productivity proxies does not yield any significant changes during this time period.
Therefore, it is concluded that even though the Pettaquamscutt River has had historical pollution that can be quantified through time, the ecological impacts are
46 undetectable. This conclusion is encouraging as the Pettaquamscutt River is a valuable natural resource.
1.54 Estuarine response to cultural eutrophication
One of the most common stressors to estuaries and lakes today is nutrient loading from human sources (Conley et al., 2002; Gorham and Sanger, 1976).
Known as cultural eutrophication, the phenomenon is driven by anthropogenic sources of nitrogen and/or phosphorous. Depending on the limiting nutrient in the water body, the addition of such nutrients can first increase productivity through fertilization. As this organic matter becomes decomposed, however, the overall health of the ecosystem can deteriorate as oxygen is consumed at great rates. It is common in estuaries for the limiting nutrient to be nitrogen (Boynton et al., 1982).
Long-term stable nitrogen fluctuations are assessed here and compared to ecological proxy data in order to understand potential responses to the forcing.
Figure 1.41 shows the stable nitrogen variability over the last ca. three centuries. It is apparent from the figure that there is a step-wise shift toward heavier values (3.7 to
5.4) between 1950 and 1960 A.D. The significance of this step is confirmed with a t- test which demonstrates that the post 1950 nitrogen values are significantly heavier than the background from 1720-1950 A.D. (p<0.0001). The timing of this shift corresponds to increased residential development in the watershed (Ernst et al., 1999).
The houses that were built at this time were not connected to town sewers, and relied on individual septic systems. The significant shift to more enriched values follows this change as human waste has heavier nitrogen isotopic values than natural systems.
47 Around 1990 the towns of Narragansett and South Kingstown initiated sewer projects that has connected these houses to off-site, public sewage treatment facilities (Ernst et al., 1999). Although there is a significant drop in 15N that occurs after 1990 as compared to the post 1960-eutrophication period (5.2‰ vs. 5.7‰; t-test: p=0.026), values are still significantly higher than background values (5.2‰ vs. 3.8‰; t-test: p<0.0001).
Examining proxy records of productivity can assess the ecological response to the nutrient loading in the last few decades. Figure 1.41 displays three such time- series from the location. First, the biological lamination thickness can be used as a rough measure of organic material reaching the sediment surface each year. This proxy shows a clear trend toward thicker laminations since the main development time in the watershed. Chlorophyll a is used here as a proxy for phytoplankton in the water column. Once again, a clear trend toward higher productivity is noted over the last few decades. The trend is more complicated here, however, with an initial step up at around 1960 with decreasing Chlorophyll a MARs for about two decades followed by an increasing trend since about 1980. Finally, Bacteriochlorophyll e is used as a proxy for anoxic phototrophic bacteria that are restricted to below the oxycline. This proxy initially increases as the water becomes loaded with nitrogen.
In the late 1960s, however, the trend reverses and recent values have been abnormally low compared to the last almost three centuries. Since these bacteria are restricted to being below the oxycline, increases in turbidity negatively affect them. Therefore, the eutrophication that is happening in the upper water column (increased phytoplankton productivity) has been detrimental to these bacteria.
48 In addition to the productivity proxies, the C/N ratio in the sediment can provide insight into eutrophication in an estuary (Figure 1.30). The average C/N ratio of marine algal matter is 6-9, whereas the value for terrestrial plants is 20 or greater
(Meyers and Teranes, 2001). Since estuarine sediments incorporate a mixture of these two sources, relative changes in the C/N ratio indicate changes in dominant organic carbon source to the estuary. In the Pettaquamscutt River core data, C/N ratios were measured for a number of depths, however the concentrations were too low to resolve this ratios at all stratigraphic depths. The available data however show a statistically significant decrease in C/N ratio between the samples deposited before
1950 and those deposited after 1950 (t-test: p<0.0001). This shift suggests that autochthonous carbon sources became more dominant during the time of cultural eutrophication in the estuary. This observation further supports the conclusion that anthropogenic nutrient loading over the last 5 decades has had a significant influence on the estuarine ecosystem of the Pettaquamscutt River.
It is evident from these data that cultural eutrophication has occurred in tandem with residential development in the Pettaquamscutt River Estuary. The nutrient (nitrogen) loading has increased primary productivity in the upper water column to levels that have not been observed in the previous 280 years. As phytoplankton biomass has increased, so must have the turbidity in the water column.
This increased turbidity has had an adverse effect on the unique bacterial population in the estuary (Figure 1.41). The sewage treatment project that was initiated in the
1990s has arguably lightened the nitrogen values in the estuary; however there has
49 been no response in the ecologic proxy data. In fact, productivity is, if anything, increasing at a faster rate since the sewage treatment project was initiated.
1.60 Conclusions
In conclusion, the Pettaquamscutt River Estuary has archived a complete post- glacial history for southern Rhode Island, including a varved record for the last millennium. The post glacial record helps in bracketing glacial retreat timing in a state that lacks direct dates from glacial land features. In addition, the recent varved record provides evidence of European influence on the estuary around 1700 A.D. due to land clearance in the watershed. There is also strong evidence of cultural eutrophication over the last few decades related to residential development in the watershed.
50 1.70 References
Balco, G., and Schaefer, J., 2006, Cosmogenic-nuclide and varve chronologies for the deglaciation of southern New England: Quaternary Geochronology, In Press.
Balco, G., Stone, J. O. H., Porter, S. C., and Caffee, M. W., 2002, Cosmogenic- nuclide ages for New England coastal moraines, Martha's Vineyard and Cape Cod, Massachusetts, USA: Quaternary Science Reviews, v. 21, p. 2127-2135.
Bianchi, T. S., Demetropoulos, A., Hadjichristophorou, M., Argyrou, M., Baskaran, M., and Lambert, C. D., 1996, Plant pigments as biomarkers of organic matter sources and coastal waters of Cyprus (eastern Mediterranean): Estuarine, Coastal, and Shelf Science, v. 42, p. 103-115.
Björck, S., and Wohlfarth, B., 2001, 14C chronostratigraphic techniques in paleolimnology, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 205-245.
Boothroyd, J. C., 1991, The geologic history of the Narrow River: Maritimes, v. 35, p. 3-5.
Boothroyd, J. C., Freedman, J. H., Brenner, H. B., and Stone, J. R., 1998, The glacial geology of southern Rhode Island, in Murray, D. P., ed., Guidebook to Field Trips in Rhode Island and Adjacent Regions of Connecticut and Massachusetts: 1998 New England Intercollegiate Geological Conference.
Boynton, W. R., Kemp, W. M., and Keefe, C. W., 1982, A comparative analysis of nutrients and other factors influencing estuarine phytoplankton productivity, in Kennedy, V. S., ed., Estuarine Comparisons: New York, Academic Press, p. 69-90.
Brigham-Grette, J., 1996, Geochronology of Glacial Deposits, Ch 14, in Menzies, J., ed., Past Glacial Environments - Sediments, Landforms, and Techniques: Oxford, Butterworth Heinemann, p. 377-410.
Burgess, R. M., McKinney, R. A., and Brown, W. A., 1996, Enrichment of marine sediment colloids with polychlorinated biphenyls: Trends resulting from PCB solubility and chlorination: Environmental Science and Technology, v. 30, no. 8, p. 2556-2566.
Chen, N., Bianchi, T. S., McKee, B. A., and Bland, J. M., 2001, Historical trends of hypoxia on the Louisiana shelf: application of pigments as biomarkers: Organic Geochemistry, v. 32, p. 543-561.
51 Clark, R. B., 1997, Marine Pollution: Oxford, Clarendon Press, 161 p.
Cole, J. R., 1889, History of Washington and Kent Counties, Rhode Island: New York, W.W. Preston & Co.
Conley, D. J., Markager, S., Andersen, J., Ellermann, T., and Svendsen, L. M., 2002, Coastal eutrophication and the Danish National Aquatic Monitoring and Assessment Program: Estuaries, v. 25, no. 4b, p. 848-861.
Crutzen, P. J., 2002, Geology of mankind: Nature, v. 415, p. 23.
Davis, P., 2006, Buying and selling the human species: Newport and the Slave Trade, The Providence Journal: Providence.
Dean, W. E., Jr., 1974, Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss-on-ignition: comparison with other methods: Journal of Sedimentary Petrology, v. 44, p. 242-248.
Doniger, D., Friedman, D., Hwang, R., Lashof, D., and Mark, J., 2002, Dangerous addictions-Ending America's oil dependence: Natural Resources Defense Council and Union of Concerned Scientists.
Ernst, L. M., Miguel, L. K., and Willis, J., 1999, The Narrow River special area management plan for the watershed of the Narrow River in the towns of North Kingstown, South Kingstown, and Narragansett: Rhode Island Coastal Resources Management Council.
Faegri, K., and Iversen, J., 1989, Textbook of Pollen Analysis: New York, NY, John Wiley and Sons, Ltd., 237 p.
Figgins, D., 2004, The shipbuilding legacy of John A. Saunders, South County Independent: Wakefield, RI, p. A11.
Francis, D. R., and Foster, D. R., 2001, Response of small New England ponds to historic land use: The Holocene, v. 11, no. 3, p. 301-312.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Freedman, J. M., 1997, Reconstruction of late Pleistocene to early Holocene environments and paleoclimates of the Great Swamp area, South Kingstown, Rhode Island [MS thesis]: University of Rhode Island, 98 p.
52 Gaines, A. G. J., 1975, Papers on the geomorphology, hydrography, and geochemistry of the Pettaquamscutt River Estuary [Ph.D. thesis]: University of Rhode Island.
Glew, J. R., Smol, J. P., and Last, W. M., 2001, Sediment core collection and extrusion, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 73-105.
Gorham, E., and Sanger, J. E., 1976, Fossilized pigments as stratigraphic indicators of cultural eutrophication in Shagawa Lake, northeastern Minnesota: Geological Society of America Bulletin, v. 87, p. 1638-1642.
Gosse, J. C., and Phillips, F. M., 2001, Terrestrial in situ cosmogenic nuclides: theory and application: Quaternary Science Reviews, v. 20, no. 14, p. 1475-1560.
Gustavson , T. C., and Boothroyd, J. C., 1987, A depositional model for outwash, sediment sources, and hydrologic characteristics, Malaspina Glacier, Alaska: a modern analog of the southeastern margin of the Laurentide Ice Sheet: Geological Society of America Bulletin, v. 99, p. 187-200.
Hargraves, P. E., 29 September 2003, Personal Communication.
Harradine, P. J., Harris, P. G., Head, R. N., Harris, R. P., and Maxwell, J. R., 1996, Steryl chlorin esters are formed by zooplankton grazing: Geochimica et Cosmochimica Acta, v. 60, no. 12, p. 2265-2270.
Hartmann, P., 2001, Sources, transport, and fate of organic contaminants in Narragansett Bay, RI [PhD thesis]: University of Rhode Island, 335 p.
Hartmann, P., Quinn, J. G., Cairns, R. W., and King, J. W., 2005, Depositional history of organic contaminants in Narragansett Bay, Rhode Island, USA: Marine Pollution Bulletin, v. 50, no. 4, p. 388-395.
Hughen, K. A., Overpeck, J. T., and Anderson, R. F., 2000, Recent warming in a 500- year palaeotemperature record from varved sediments, Upper Soper Lake, Baffin Island, Canada: The Holocene, v. 10, no. 1, p. 9-19.
IPCC, 2001, Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, 881 p.
53 Kelly, R. P., and Moran, S. B., 2002, Seasonal changes in groundwater input to a well-mixed estuary estimated using radium isotopes and implications for coastal nutrient budgets: Limnology and Oceanography, v. 47, no. 6, p. 1796- 1807.
King, J. W., and Channell, J. E. T., 1991, Sedimentary magnetism, environmental magnetism, and magnetostratigraphy: Reviews of Geophysics Supplement, p. 358-370.
King, J. W., and Peck, J., 2001, Use of paleomagnetism in studies of lake sediments, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediment. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 371-389.
Koteff, C., and Larsen, F. D., 1989, Postglacial uplift in western New England: Geologic evidence for delayed rebound, in Gregersen, S., and Basham, P. W., eds., Earthquakes at North-Atlantic Passive Margins: Neotectonics and postglacial rebound: Dordrecht, Netherlands, Kluwer Academic Publishers, p. 105-123.
Koteff, C., and Pessl, F. J., 1981, Systematic ice retreat in New England: U.S. Geological Survey Professional Paper 1179.
Kreitler, C. W., and Browning, L. A., 1983, Nitrogen-isotope analysis of groundwater nitrate in carbonate aquifers: Natural sources versus human pollution: Journal of Hydrology, v. 61, p. 285-301.
Lacey, E. M., 1997, Evaluation of simultaneously extracted metals-acid volatile sulfide (SEM-AVS) in lacustrine sediment (Lake Champlain), estuarine sediment (Narragansett Bay), and under laboratory conditions [PhD thesis]: University of Rhode Island, 226 p.
Lauenstein, G. G., Cantillo, A. Y., and Dolvin, S. S., 1993, NOAA National Status and Trends Program development and methods, in Lauenstein, G. G., and Cantillo, A. Y., eds., Sampling and analytical methods of the National Status and Trends Program, National Benthic Surveillance, and Mussel Watch Programs: 1984-1992: Silver Spring, MD, National Oceanic and Atmospheric Administration.
Leavitt, P. R., 1993, A review of factors that regulate carotenoid and chlorophyll deposition and fossil pigment abundance: Journal of Paleolimnology, v. 9, p. 109-127.
54 Leavitt, P. R., and Hodgson, D. A., 2001, Sedimentary pigments, in Last, W. M., and Smols, J. P., eds., Tracking environmental change using lake sediments. Volume 3: Terrestrial, algal, and siliceous indicators: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 295-325.
Lima, A. L., Hubeny, J. B., Reddy, C. M., King, J. W., Hughen, K. A., and Eglinton, T. I., 2005, High-resolution historical records from Pettaquamscutt River basin sediments: 1. 210Pb and varve chronologies validate record of 137Cs released by the Chernobyl accident: Geochimica et Cosmochimica Acta, v. 69, no. 7, p. 1803-1812.
Lima, A. L. C., Eglinton, T. I., and Reddy, C. M., 2003, High-resolution record of pyrogenic polycyclic aromatic hydrocarbon deposition during the 20th century: Environmental Science and Technology, v. 37, p. 53-61.
Long, E. R., Ingersoll, C. G., and MacDonald, D. D., 2006, Calculation and uses of mean sediment quality guideline quotients: A critical review: Environmental Science and Technology, v. 40, no. 6, p. 1726-1736.
Long, E. R., MacDonald, D. D., Smith, S. L., and Calder, F. D., 1995, Evidence of adverse biological effects within ranges of chemical concentrations in marine and estuarine sediments: Environmental Management, v. 19, p. 81-97.
Mann, M. E., and Jones, P. D., 2003, Global surface temperatures over the past two millennia: Geophysical Research Letters, v. 30, no. 15, p. 1820.
McKinney, R. A., Lake, J. L., Charpentier, M., and Ryba, S., 2002, Using mussel isotope ratios to assess anthropogenic nitrogen inputs to freshwater ecosystems: Environmental Monitoring and Assessment, v. 74, p. 167-192.
McWeeney, L. J., 1995, Revised vegetation history for the post-glacial period (15,200 - 10,000 14C years b.p.) in southern New England: GSA Abstracts with Programs, v. 27, p. 68.
Mecray, E. L., 1994, An assessment of historical trace metal accumulation in the Burlington region of Lake Champlain, Vermont [Masters thesis]: University of Rhode Island, 271 p.
Mecray, E. L., King, J. W., and Corbin, J. M., 1991, Metal pollution recorded in Pettaquamscutt River sediments: Maritimes, v. 35, no. 2, p. 10-11.
Meyers, P. A., and Teranes, J. L., 2001, Sediment organic matter, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 2: Physical and geochemical methods: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 239-269.
55 Nederbragt, A. J., and Thurow, J., 2005, Amplitude of ENSO cycles in the Santa Barbara Basin, off California, during the past 15000 years: Journal of Quaternary Science, v. 20, no. 5, p. 447-456.
Nowaczyk, N. R., 2001, Logging of magnetic susceptibility, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 155-170.
O'Conner, T. P., 1996, Trends in chemical concentrations in mussels and oysters collected along the US coast from 1986-1993: Marine Environmental Research, v. 41, p. 183-200.
Orr, W. L., and Gaines, A. G., Jr., 1973, Observations on rate of sulfate reduction and organic matter oxidation in the bottom waters of an estuarine basin: the upper basin of the Pettaquamscutt River (Rhode Island), in Tissot, B., and Bienner, F., eds., Advances in Organic Geochemistry, Proceedings of the 6th International Meeting on Organic Geochemistry: Paris, Editions technip, p. 791-812.
Peck, J. A., and McMaster, R. L., 1991, Stratigraphy and geologic history of Quaternary sediments in lower West Passage, Narragansett Bay, Rhode Island: Journal of Coastal Research, v. 11, p. 25-37.
Peters, C., and Thompson, R., 1998, Magnetic identification of selected natural iron oxides and sulphides: Journal of Magnetism and Magnetic Materials, v. 183, p. 365-374.
Pilcher, J. R., 2003, Radiocarbon dating and environmental radiocarbon studies, in Mackay, A., Battarbee, R., Birks, J., and Oldfield, F., eds., Global Change in the Holocene: London, Arnold, p. 63-74.
Quinn, J. G., 2006, Personal Communication.
Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W., Bertrand, C. J. H., Blackwell, P. G., Buck, C. E., Burr, G. S., Cutler, K. B., Damon, P. E., Edwards, R. L., Fairbanks, R. G., Friedrich, M., Guilderson, T. P., Hogg, A. G., Hughen, K. A., Kromer, B., McCormac, G., Manning, S., Ramsey, C. B., Reimer, R., Remmele, S., Southon, J. R., Stuiver, M., Talamo, S., Taylor, F. W., van der Plicht, J., and Weyhenmeyer, C. E., 2004, INTCAL04 terrestrial radiocarbon age calibration, 0-26 cal kyr bp: Radiocarbon, v. 46, p. 1029- 1058.
56 Ridge, J. C., 2003, The last deglaciation of the northeastern United States: a combined varve, paleomagnetic, and calibrated 14C chronology, in Cremeens, D. L., and Hart, J. P., eds., Geoarchaeology of landscapes in the glaciated Northeast: New York State Museum Bulletin 497: Albany, NY, University of the State of New York, p. 15-45.
-, 2004, The Quaternary glaciation of western New England with correlations to surrounding areas, in Ehlers, J., and Gibbard, P. L., eds., Quaternary glaciations - Extent and chronology, Elsevier, p. 169-199.
Ridge, J. C., and Larsen, F. D., 1990, Reevaluation of Antevs' New England varve chronology and new radiocarbon dates of sediments from glacial Lake Hitchcock: Geological Society of America Bulletin, v. 102, p. 889-899.
Ruddiman, W. F., 2001, Earth's Climate: Past and future: New York, W.H. Freeman and Company, 465 p.
Schafer, J. P., 1961a, Surficial geology of the Narragansett Pier quadrangle, Rhode Island: U.S. Geological Survey.
-, 1961b, Surficial geology of the Wickford quadrangle, Rhode Island: U.S. Geological Survey.
Sieburth, J. M., and Donaghay, P. L., 1993, Planktonic methane production and oxidation within the algal maximum of the pycnocline: seasonal fine-scale observations in an anoxic estuarine basin: Marine Ecology Progress Series, v. 100, p. 3-15.
Snowball, I., and Torii, M., 1999, Incidence and significance of magnetic iron sulphides in Quaternary sediments and soils, in Maher, B. A., and Thompson, R., eds., Quaternary climates, environments, and magnetism: Cambridge, UK, Cambridge University Press, p. 199-231.
Stone, B. D., and Borns, H. W., 1986, Pleistocene glacial and interglacial stratigraphy of New England, Long Island, and adjacent Georges Bank and Gulf of Maine, in Sibrava, V., Bowen, D. Q., and Richmond, G. M., eds., Quaternary Glaciations in the Northern Hemisphere, Quaternary Science Reviews 5: Oxford, Permagon Press, p. 39-52.
Talbot, M. R., 2001, Nitrogen isotopes in palaeolimnology, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments: Volume 2: Physical and geochemical methods: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 401-439.
57 Tavendale, M., H., McFarlane, P., N., Mackie, K. L., Wilkins, A. L., and Langdon, A. G., 1997, The fate of resin acids-1. The biotransformation and degradation of deuterium labeled dehydroabietic acid in anaerobic sediments: Chemosphere, v. 35, no. 10, p. 2137-2151.
Taylor, B. R., Parkinson, D., and Parsons, W. F. J., 1989, Nitrogen and lignin content a predictors of litter decay rates: A microcosm test: Ecology, v. 70, no. 1, p. 97-104.
Thompson, R., and Oldfield, F., 1986, Environmental Magnetism: London, Allen & Unwin Ltd., 227 p.
Urish, D. W., 1991, Freshwater inflow to the Narrow River: Maritimes, v. 35, p. 12- 14.
Verosub, K. L., and Roberts, A. P., 1995, Environmental magnetism: Past, present, and future: Journal of Geophysical Research, v. 100, no. B2, p. 2175-2192.
Wright, S. W., Jeffrey, S. W., Mantoura, R. F. C., Llewellyn, C. A., Bjornland, T., Repeta, D., and Welschmeyer, N., 1991, Improved HPLC method for the analysis of chlorophylls and carotenoids from marine phytoplankton: Marine Ecology Progress Series, v. 77, p. 183-196.
Zolitschka, B., Mingram, J., van der Gaast, S., Jansen, J. H. F., and Naumann, R., 2001, Sediment logging techniques, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments: Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 137-153.
58 Table 1.1. List of sediment cores collected and analyzed for this manuscript.
Core Name Core Type Core Length Notes (cm) NR03-3 Kullenburg Piston 474.8 Expanded section of NR03-4 Core NR03-4 Kullenburg Piston 527.7 Full record, including basal sand Core lithofacies NR99-2 Freeze Core 72.0 Upper laminated sediments NR99-3 Freeze Core 73.0 Upper laminated sediments NR99-5 Freeze Core 74.0 Upper laminated sediments NR99-7 Freeze Core 70.0 Upper laminated sediments NR02-1 Freeze Core 70.0 Upper laminated sediments NR03-1 Freeze Core 120.2 Upper laminated sediments NR04-1 Freeze Core 112.0 Upper laminated sediments NR04-2 Freeze Core 180.0 Bottom drive from core NR04-1
59 Table 1.2. Fossil pigments analyzed for this manuscript along with associated taxonomic affinities and references.
Pigment Name Affinity Reference
Bacteriochlorophyll e Brown-green sulfur photobacteria (Chen et al., 2001)
Chlorophyll a Plants, algae (Leavitt and Hodgson, 2001) beta-Carotene Plants, algae, some phototrophic (Leavitt and bacteria Hodgson, 2001) Steryl Chlorin Esters Zooplankton grazing activity (Harradine et al., 1996) Alloxantin Cryptophytes (Leavitt and Hodgson, 2001) Lutein Green algae (Leavitt and Hodgson, 2001) Zeaxanthin Cyanobacteria (Leavitt and Hodgson, 2001)
60 Table 1.3. Pettaquamscutt River Estuary freeze core radiocarbon data.
Sample Composite Material Lab 14C age 2 calibrated Relative Name Depth ID (yr BP) age ranges (yr area under (cm) AD) distribution NR03-1 64 Leaf OS- 225 ± 40 1523 - 1568 0.052 78 pieces 47531 1627 - 1692 0.390 1727 - 1812 0.460 1919 - 1949 0.098 NR03-1 78 Leaf OS- 345 ± 40 1461 - 1640 1.000 92 pieces 47616 NR04-2 124 Leaf OS- 435 ± 40 1411 - 1519 0.925 J5 pieces 47532 1593 - 1622 0.075 NR04-2 186 Leaf OS- 915 ± 110 900 - 920 0.021 T3 pieces 47749 940 - 1290 0.979
All analyses were conducted at the National Ocean Sciences AMS Facility.
Calibrations were calculated using the CALIB 5.0 program (Reimer et al., 2004).
61 Table 1.4. Pettaquamscutt River Estuary piston core radiocarbon data.
Core Core Material Lab 14C age 2 calibrated Relative Name Depth ID (yr BP) age ranges (cal area under (cm) BP) distribution NR03-4 139 Leaf OS 2830 ± 100 2754 - 3218 0.997 pieces 53486 3232 - 3238 0.003 NR03-4 212 Leaf OS 5870 ± 55 6508 - 6512 0.003 pieces 53488 6533 - 6799 0.978 6817 - 6834 0.014 6836 - 6843 0.005 NR03-4 252 Leaf OS 8010 ± 70 8633 - 9032 0.992 pieces 53489 9061 - 9075 0.008 NR03-4 287 Leaf OS 8770 ± 160 9523 - 10,221 1.000 pieces 53490 NR03-4 305 Small OS 9220 ± 50 10,248 - 10,514 1.000 twigs 53491 NR03-4 357 Leaf OS 11,300 ± 270 12,814 - 13,728 1.000 pieces 53492 NR03-3 113 Misc. OS 2380 ± 40 2335 - 2498 0.888 organic 53494 2597 - 2612 0.020 2638 - 2690 0.092
All analyses were conducted at the National Ocean Sciences AMS Facility.
Calibrations were calculated using the CALIB 5.0 program (Reimer et al., 2004).
62 Table 1.5. Effects range low (ERL) and effects range median (ERM) concentrations for pollutants analyzed in this study. Sediment quality criteria are from Long et al.
(1995). All concentrations are normalized to sediment dry weight in grams.
Contaminant ERL Concentration ERM Concentration Total PAH 4000 ng/g 45000 ng/g Total PCBs 23 ng/g 180 ng/g Total DDT 1.6 ng/g 46 ng/g Lead 46.7 g/g 218 g/g Silver 1 g/g 3.7 g/g Cadmium 1.2 g/g 9.6 g/g Copper 34 g/g 270 g/g Arsenic 8.2 g/g 70 g/g Chromium 81 g/g 370 g/g Nickel 20.9 g/g 51.6 g/g
63 Figure 1.1: Locus map of the Pettaquamscutt River Estuary, Rhode Island.
Watershed is shown in white along with all streams in the watershed.
64 Figure 1.2. Topography and bathymetry of the watershed and estuary in the vicinity of the Lower Basin, Pettaquamscutt River Estuary. Deepest hole is approximately
19.5 meters. Topographic data from the Rhode Island Geographic Information
System (RIGIS) database. Bathymetric data from Gaines (1975).
65 Figure 1.3. Generalized circulation of the upper Pettaquamscutt River Estuary (Orr and Gaines, 1973). Note the stagnant, anoxic deep waters associated with density stratification. Vertical exaggeration approximately 40:1.
66 Figure 1.4. Major moraines of southern New England along with composite ice- recession dates labeled (calibrated thousands of years before present) (Balco and
Schaefer, 2006). Dates at Lower Quinnipiac Valley and Rocky Hill Dam are minimum dates of ice position based on adjusted New England varve chronology
(Balco and Schaefer, 2006; Ridge, 2003, 2004). This figure assumes temporal synchronicity between major moraine systems (e.g. Buzzards Bay/ Charlestown moraine; Old Saybrook moraine and Wolf Rocks) (Balco and Schaefer, 2006). Note the location of the Pettaquamscutt River Estuary (O), which is interpreted to have been deglaciated between the 19,200 cal BP and 18,790 cal BP.
67 0
Dark brown, well-laminated, organic-rich mud
1
2 Dark gray to black, partially laminated, gassy, organic- rich mud
3 Core Depth (m)
Dark to medium gray, firm, 4 organic-poor sandy silt with tan and black banding
5 Medium to coarse sand
6
Figure 1.5. General lithostratigraphy of Pettaquamscutt River Estuary Lower Basin sediments as represented by piston core NR03-4. Image from core NR03-4 with associated lithofacies labeled. Core depths represents depth in the recovered core sample.
68 NR03-1/04-2 NR03-4 0 0.004 0.008 0 0.004 0.008 0
1
2
3
4 Adjusted Core Depth (m)
5
6
7
Figure 1.6. ARM plots for composite surface freeze core (NR03-1 and NR04-2) and piston core NR03-4. Tie lines indicate correlations of distinct features in the data used to correlate between cores. Composite freeze core constructed by varve correlation. Note that the piston core did not recover approximately 1 meter of sediment and compressed the upper portion of sediment by approximately 40%.
69 NR03-1/04-2 NR03-4 NR03-3 1 1.2 1.4 1.6 1.8 1 1.2 1.4 1.6 1.8 1 1.2 1.4 1.6 1.8 0
1
2
3
4 Adjusted Core Depth (m)
5
6
7
Figure 1.7. GRAPE (bulk density; g/cc) plots for composite surface freeze core
(NR03-1 and NR04-2) and piston cores NR03-4 and NR03-3. Tie lines indicate correlations of distinct features in the data used to correlate between cores. Composite freeze core constructed by varve correlation. Note that the piston core did not recover approximately 1 meter of sediment and compressed the upper portion of sediment by approximately 40%. In addition, piston core NR03-3 represents an expanded section as compared to piston core NR03-4.
70 NR03-1/04-2 NR03-4 NR03-3 0 10203040 0 10203040 0 10203040 0
1
2
3
4 Adjusted Core Depth (m)
5
6
7
Figure 1.8. Magnetic susceptibility plots for composite surface freeze core (NR03-1 and NR04-2) and piston cores NR03-4 and NR03-3. Tie lines indicate correlations of distinct features in the data used to correlate between cores. Composite freeze core constructed by varve correlation. Note that the piston core did not recover approximately 1 meter of sediment and compressed the upper portion of sediment by approximately 40%. In addition, piston core NR03-3 represents an expanded section as compared to piston core NR03-4.
71 Figure 1.9. Pettaquamscutt River upper laminated sediments as seen scanned under cross-polarized light. Left is composite image of the last century of varve years.
Right is close-up of thin section marking clastic (C) and biogenic (B) layers. Scale bars are 1cm in vertical direction.
72 Figure 1.10. Age model for the varved portion of the Pettaquamscutt River Lower
Basin sediment record from composite freeze cores. Composite depth scale was computed using mean varve thicknesses for each year, and adjusted to a common zero depth at 1999 AD. 210Pb data from Lima et al. (Lima et al., 2005). 137Cs peaks in
1986 and 1963 (Lima et al., 2005), help to confirm the upper chronology. The most significant calibrated radiocarbon date for each sample is plotted here with the associated error bars (Table 1.2). One radiocarbon sample (*) is anomalously old due to land clearance and remobilization of older macrofossils from watershed during
European settlement in the area.
73 Figure 1.11. Pettaquamscutt River piston core magnetic declination records as matched to dated Northeast regional PSV curve (King and Peck, 2001). Gray polygons show calibrated radiocarbon dates from terrestrial macrofossils and the associated error on the PSV curve (Table 1.4). Dashed tie lines connect the local declination variations to the regional dated record within the framework of radiocarbon dates from the individual cores. All tie lines are consistent with the inclination interpretation (Figure 1.12).
74 Figure 1.12. Pettaquamscutt River piston core magnetic inclination records as matched to dated Northeast regional PSV curve (King and Peck, 2001). Gray polygons show calibrated radiocarbon dates from terrestrial macrofossils and the associated error on the PSV curve (Table 1.4). Dashed tie lines connect the local inclination variations to the regional dated record within the framework of radiocarbon dates from the individual cores. All tie lines are consistent with the declination interpretation (Figure 1.11).
75 Figure 1.13. Representative Zijderveld plots from Pettaquamscutt River piston cores.
Top from core NR03-3, 40cm depth. Bottom from core NR03-4, 35cm depth. Stable behavior was found for the majority of samples at the 150Oe demagnetization step.
76 Age (cal BP) 0 2000 4000 6000 8000 10000 12000 14000 16000 0 •
50 • 100 •• • 150 • • 200 • •• •• 250 • •
Core Depth (cm) 300 • • •• 350 • •• 400 •• 450 • 500
Figure 1.14. Age model for piston core NR03-3. Ages based on tie to varve freeze core record (Figure 1.10) and correlation to regional PSV records within the constraints of radiocarbon data (Figures 1.11; 1.12). Gray symbol with error bars shows calibrated radiocarbon date from terrestrial macrofossil (Table 1.4). Lowest age determined from linear extrapolation of age trend and assumes constant sedimentation rate.
77 Age (cal BP) 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 •
50 • •• •• 150 • • •• ••• • 250 •• • ••• ••• Core Depth (cm) • 350 ••
450 •
? 550
Figure 1.15. Age model for piston core NR03-4. Ages based on tie to varve freeze core record (Figure 1.10) and correlation to regional PSV records within the constraints of radiocarbon data (Figures 1.11; 1.12). Gray symbols with error bars show calibrated radiocarbon dates from terrestrial macrofossils (Table 1.4). Lowest age determined from linear extrapolation of age trend and assumes constant sedimentation rate. Bottom section of core is sand lithofacies, which was likely deposited as the ice block collapsed.
78 Figure 1.16. Pettaquamscutt River piston core declination data plotted on age scale
(Figures 1.14, 1.15). Gray boxes highlight correlative features between piston core data and regional declination curve (King and Peck, 2001).
79 Figure 1.17. Pettaquamscutt River piston core inclination data plotted on age scale
(calibrated years before present) (Figures 1.14, 1.15). Gray boxes highlight correlative features between piston core data and regional declination curve (King and Peck, 2001).
80 RGB GRAPE (g/cc) MS (SI) ARM (emu) SIRM (emu) 100 200 300 1.2 1.6 2 0 200 400 0 0.008 0 0.02 0.04 0
100
200
300 Core Depth (cm)
400
500
Lightness Bulk Density Magnetic Concentration 600
Figure 1.18. Physical properties of the Pettaquamscutt River Lower Basin sediments from reference core NR03-4. RGB is the quantitative sum of the red, green, and blue intensity values as determined from image analysis. Gamma ray density (GRAPE) values track the bulk density of the sediment, while magnetic susceptibility (MS), anhysteretic remanent magnetization (AMS), and saturation isothermal remanent magnetization (SIRM) are indicative of magnetic concentration.
81 RGB GRAPE (g/cc) MS (SI) ARM (emu) SIRM (emu) 100 200 300 1.2 1.3 1.4 1.5 040800 0.008 0 0.02 0.04 0
2000
4000
6000
8000
10000 Age (cal BP) 12000
14000
16000
18000
20000 Lightness Bulk Density Magnetic Concentration
Figure 1.19. Same as Figure 1.18 except plotted on age scale (calibrated years before present).
82 RGB GRAPE (g/cc) SIRM/ARM ARM/MS (emu)SIRM/MS (emu) 100 200 300 1.2 1.6 2 15 10 5 0 0 1000 2000 0 2500 5000 0
1
2
3 Core Depth (m)
4
5
Lightness Bulk Density Magnetic Grain Size fine coarse 6
Figure 1.20. Physical properties and magnetic grain size variations of the
Pettaquamscutt River Lower Basin sediments from reference core NR03-4. RGB is the quantitative sum of the red, green, and blue intensity values. Gamma ray density
(GRAPE) values track the bulk density of the sediment, while ratios of saturation isothermal remanent magnetization: anhysteretic remanent magnetization
(SIRM/ARM), anhysteretic remanent magnetization: magnetic susceptibility
(ARM/MS), and saturation isothermal remanent magnetization: magnetic susceptibility (SIRM/MS) are indicative of magnetic grain size variations.
83 RGB GRAPE (g/cc) SIRM/ARM ARM/MS (emu) SIRM/MS (emu) 100 200 300 1.2 1.3 1.4 1.5 15 10 5 0 0 1000 2000 0 2500 5000 0
2000
4000
6000
8000
10000 Age (cal BP) 12000
14000
16000
18000
20000
Lightness Bulk Density Magnetic Grain Size fine coarse
Figure 1.21. Same as Figure 1.20 except plotted on age scale (calibrated years before present).
84 Figure 1.22. Backscatter electron image (BSEI) of framboidal pyrite grains observed in estuarine varved mud from the Pettaquamscutt River Estuary’s Lower Basin.
85 1000 ◊ † † † † † ◊ †† ◊ + ◊ † † ◊ †† ◊ ◊◊ 100 † ◊◊◊◊ †† ◊◊◊◊◊* ††† † ◊ *** ◊◊ *** † † †◊◊◊◊◊◊◊** ** † † † ◊◊◊◊◊◊**** † †† ◊◊◊◊◊◊*◊*** pyrrhotite + † ◊◊◊◊◊ *** + † ◊◊ **** + •• † • • • *◊*•*** +++ • • † • • ******* + + ••† •• • ••**•*◊•*•* 10 +++ • † • •• • • •••**•••** • Post Glac +++++ + • † • • ••• • • *•**•*•*• ++++++ ••••†• • • ••• •**•*•••*** +++++++++ • ••††• • •• • *•* SIRM/X +++++ • • • • * +++++++++ •• • • • • • * Lower Lacustrine +++++ • •• + ••• •• •• ◊ Lacustrine 1 † • magnetite † Transition
titanomagnetite • + Estuarine • greigite 0.1 0.01 0.1 1 ARM(400Oe)/ARM(0Oe)
Figure 1.23. Biplot used to help interpret the bulk magnetic minerals present in the
Pettaquamscutt River Estuary (Peters and Thompson, 1998). Most samples fall within the experimentally determined zone for magnetite, titanomagnetite and greigite
(lower shaded polygon). A general shift toward magnetically soft behavior is observed in the transition from lacustrine sediments to estuarine sediments.
86 Varve Thickness (cm) Biologic Laminae Thickness (cm) 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 2000
1900
1800
1700
1600
1500 Year (A.D.) 1400
1300
1200
1100
1000 0 0.1 0.2 0.3 0.4 Clastic Laminae Thickness (cm)
Figure 1.24. Compaction-corrected varve thickness time-series. Total varve thickness is on the left with clastic layer thickness in the middle and biologic layer thickness on the right.
87 Total PAH (ng/g) 2,6- dimethylnaphthalene (DMN) (ng/g) Benzo[a]pyrene (ng/g) 0 4000 8000 12000 0 900 1800 0 200 400 2010 • • • •• ••• •• • • • ••• • ••• • ••• 1990 • • • • • • • • • • • • • • • • • • 1970 • • •• • • • • •• • • • • • • • • • • • • • • 1950 • • • • • • • • • • • • • • • • • • 1930 • • • • • • •• • • ••
1910 • • • • • • Year (A.D.) 1890 • • • • • •
1870 • • • • • • • • • • • • 1850 • • • • • • • • • • • • 1830 • • • • • • ERL • • • • • • 1810 0 5000 10000 0 300 600 0 400 800 Lima PAH (ng/g) Phenanthrene (ng/g) Pyrene (ng/g)
Figure 1.25. Polycyclic aromatic hydrocarbon (PAH) concentrations (ng/g sediment dry weight) from the Pettaquamscutt River Estuary. Total PAH (n=22) as defined by
NOAA Status and Trends Program (Lauenstein et al., 1993). Lima PAH is total PAH minus Coronene (n=14), as defined by Lima et al. (2003). DMN concentrations are responsible for the small increase in PAH concentration in 1999. Phenanthrene,
Pyrene, and Benzo[a]pyrene are the three dominant PAH components. Black curves are data from this study, while gray curves are data from Lima et al. (2003) adjusted temporally with the age model of Lima et al. (2005). Effects range low (ERL) value
(4000 ng/g) (Long et al., 1995) for total PAH concentration (NOAA Status and
Trends Program) is included on the total PAH figure.
88 Total PAH (ng/g) Retene (ng/g) 0 6000 12000 0 100 200 2000 •••• • • • • • • •• •• • • 1950 • • • • • • • • • • • 1900 • • • •
Year (A.D.) • • 1850 • • •• • • • •
1800
1750
Figure 1.26. Total PAH and Retene concentrations in the Pettaquamscutt River
Estuary. Gray retene data are from Lima et al. (2003), adjusted temporally with the age model of Lima et al. (2005). Gray box represents the time period during which the Saunders’ boat building operation was active in the Pettaquamscutt River Estuary
(1813-1854 A.D.) (Figgins, 2004).
89 Total PCBs (ng/g) 2,2',3,4,4',5'-hexachlorobiphenyl (ng/g) 0 102030405060708090100 024681012 2000 • •• • 2000 • • • • • • • • 1980 • 1980 • •• • • • • 1960 • 1960 • • • 1940 • 1940 • • • • 1920 1920 • 1900 1900
Year (A.D.) • Year (A.D.) 1880 1880 • 1860 • 1860 • 1840 • 1840 • ERL
1820 • 1820
1800 1800
Figure 1.27. Total polychlorinated biphenyl (PCB) (n=24) (O'Conner, 1996) concentration in the Pettaquamscutt River Estuary sediment column. The compound
2,2’,3,4’,5’-hexachlorobiphenyl is the dominant component of the record, and consists of congeners CB138, CB163, and CB164. Effects range low (ERL) value
(23ng/g) (Long et al., 1995) for total PCB concentration is included on the PCB plot.
90 Total DDT (ng/g) 0 50 100 150 200 2000 ••• • • 1980 • •• • 1960 • • 1940 • • •• 1920 • 1900
Year (A.D.) • 1880 • 1860 • • 1840 • ERL • ERM
1820 •
1800
Figure 1.28. Total DDT concentration in the Pettaquamscutt River Estuary sediment column. ERL (1.6ng/g) and ERM (46ng/g) values are plotted for reference (Long et al., 1995). Total DDT is defined by the sum of o,p’-DDT, p,p’-DDT, o,p’-DDD, p,p’-DDD, o,p’-DDE, and p,p’-DDE.
91 Lead (ug/g) Cadmium (ug/g) Arsenic (ug/g) Nickel (ug/g) 0 10203040 0123 246 0102030 2000 • •• •• • • •• • • • •• • • • •• •• • • •• •• • • • • • • • • • • • • • • • • • • • • • • • • • • •• • • • •• 1950 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1900 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1850 • • • • • • • • • • • • • • Shady Lea Mill
Year (A.D.) • • • • • • • • • • • • • • Rodman Purchases • • • • • • • 1800 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1750 • • • • • • • • • • • • • • • • •• •• •• • • • • • • • • • • • • • • • • ERL ERL ERL 1700 ERL 0 0.4 0.8 0 50 100 0 75 150 Silver (ug/g) Copper (ug/g) Chromium (ug/g)
Figure 1.29. Trace metal concentrations (g/g) in the Pettaquamscutt River Estuary sediment column. Cadmium, copper, chromium, and nickel all have exceeded the
ERL (1.2 g/g, 34 g/g , 81 g/g , and 20.9 g/g , respectively) (Long et al., 1995) for some period of time during the record. Date of Robert Rodman’s purchase of
Shady Lea Mill (1870) to manufacture jeans and warps is indicated by horizontal line.
92 del N-15 del C-13 C/N 234567-27 -26 -25 -24 6 8 10 12 2000 • • ••• • • •• • • • • • •• • • • • • • •• ••• •• • • • • • • • • •• • • • • • • • • • • • 1950 •• • • • • • • • • • • • • • •• • • • • • • • • • • •• • • •• • • • • • • 1900 • • • • •• • • • • •• • • • •• •• • • •• • • • • • • • • • • 1850 •• • • • • • • • •• • • • • Year (A.D.) • • • • • • • • • • • •• • • • • • 1800 • • • • • • • • • • • • • • • •• • • • • • • • 1750 • • •• • • • • • • • • • • • • •
1700
Figure 1.30. Stable nitrogen (15N) and stable carbon (13C) isotopic values along with carbon to nitrogen (C/N) ratio of the Pettaquamscutt River sediment column over the last ca. three centuries. Gaps in C/N plot are due to small samples that inhibited the quantification of nitrogen concentrations.
93 Organic Carbon (mg/g) 0 20 40 60 80 100 120 140 160 0
2000
4000
6000
8000
10000 Age (cal BP) 12000
14000
16000
18000
20000
Figure 1.31. Organic carbon concentrations (mg/g) over the late Quaternary
Pettaquamscutt River Estuary record.
94 Organic Carbon (mg/g) 80 90 100 110 120 130 2000
1900
1800
1700
1600
1500 Year (A.D.) 1400
1300
1200
1100
1000
Figure 1.32. Organic carbon concentrations (mg/g) in the Pettaquamscutt River
Estuary over the last millennium (varved record).
95 Bacteriochlorophyll e beta-Carotene Alloxantin Unknown Carotenoid 0 300 600 036 0 2.5 5 036 0 0
2000 2000
4000 4000
6000 6000
8000 8000 Age (cal BP) Age (cal BP)
10000 10000
12000 12000
14000 14000
16000 16000 02040 0 80 160 0612 Chlorophyll a Steryl Chlorin Esters Zeaxanthin
Figure 1.33. Fossil pigment concentrations (g/g) over the late Quaternary
Pettaquamscutt River Estuary record.
96 Bacteriochlorophyll e beta-Carotene Alloxantin Zeaxanthin 0 200 400 048 0123 0510 2000 2000
1900 1900
1800 1800
1700 1700
1600 1600
1500 1500 Year (A.D.) Year (A.D.) 1400 1400
1300 1300
1200 1200
1100 1100
1000 1000 01020 04080 0510 Chlorophyll a Steryl Chlorin Esters Lutein
Figure 1.34. Fossil pigment mass accumulation rates (g/cm2yr) in the
Pettaquamscutt River Estuary over the last millennium (varved record).
97 Percent Pollen Grains 012345678910 1800 + 1780 •
1760 +•
1740 • +
1720 •+ + 1700 • +• Year (A.D.) 1680
1660
1640
1620 • Rumex 1600 + Ambrosia
Figure 1.35. Total percent of pollen grains Rumex and Ambrosia around the turn of the eighteenth century. Both pollen types indicate European influence in the watershed.
98 TOC MS Bchle 0 80 160 025500 250 500 Depositional Environment 0 0 Density stratified estuary
2000 2000
4000 4000
6000 6000
Lacustrine, mature 8000 terrestrial vegetation, 8000 oxygenated water column
10000 10000 Age (cal BP) Age (cal BP) 12000 12000
14000 14000
16000 16000 Lacustrine, cold climate, limited terrestrial vegetation 18000 18000
Glacial retreat, ice block 20000 20000 collapse
Figure 1.36. Total organic carbon (mg/g), magnetic susceptibility, and bacteriochlorophyll e concentration (g/g) with associated interpreted depositional environments for the entire post-glacial Pettaquamscutt River history.
99 Pollen Grains (%) Clastic Lam. Thickness MS ARM(0)/MS 0246810 0 0.2 0.4 0 5 10 15 0 0.002 0.004 1800 1800
1780 +• 1780
1760 +• 1760
1740 •+ 1740
1720 •+ 1720 ) A.D. 1700•+ 1700 ( +• Year Year (A.D.) 1680 1680
1660 1660
1640 1640
1620 1620
1600 1600 0 0.4 0.8 0 0.2 0.4 0.004 0.006 0.008 Varve Thickness Biologic Lam. Thickness ARM(0)
Figure 1.37. European land clearance and its effect on sedimentation in the
Pettaquamscutt River Estuary. Black cross pollen data are Ambrosia percentages and gray circles represent Rumex percentages. Lamination thicknesses in centimeters and adjusted for compaction. Magnetic susceptibility (MS), initial anhysteretic remanent magnetization (ARM(0)), and the interparametric ratio ARM(0)/MS are also included.
100 Pollen Grains (%) Chlorophyll a Steryl Chlorin Esters Lutein 0246810 0 5 10 15 03060 048 1800 1800 • • •• • • • • • • • • • • •• ••• • •• • •• • • •• • • • • • • • • • • •• • • • • • •• • •• •• • 1780 +• • • • • • • •• • • • • • • 1780 • •• •• • • •• • •• • •• ••• •• • • • • • ••• ••• • • • •• • •• ••• •• • • • • •• ••• •• • •• 1760 +• •• • • • ••• ••• •• • •• • •• 1760 • • • • • • • • • • • • • • •• •• •• •• •• • •• •• •• •• • • • •• •• • • • • • • • •• • •• • •• • • • • • 1740 •+ • •• ••• • • • • • • ••• •• • •• • 1740 •• • • • •• • •• • • • • •• • • • •• • • •• • • • •• • ••• •• • •• • 1720 +• • • • • •• • • • • • •• 1720 • • • ••• •• • • • • ••• •• • • • • • • • • • • • • • • • • • • • • •• • •• •+ •• • • • • • •• • • • • 1700 ••• ••• • • • •• • • • ••• •• 1700 • • •• ••• •• • •• • •• • •• • +• • • • • • • • • • • • • •• • • • • • • Year (A.D.) • • • • • • • Year (A.D.) 1680 • • •• •• • • •• • • • • 1680 • • • • •• ••• • • • • • • •• • • • •• • ••• •• • •• • • •• ••• • • • • • • • • • • • • • • •• • • • • 1660 •• •• •• • • • •• •• 1660 •• • •• • ••• • • • • ••• •• • • •• • • • •• ••• •• • ••• •• • 1640 • • • • • • • •• • • • • 1640 • •• ••• • • • • • •• • ••• ••• •• ••• • •• • •• ••• • • ••• • • •• •• ••• ••• ••• • • 1620 ••• ••• •• • •• ••• ••• •• 1620 • •• ••• ••• • • • • • • •• • • • • •• •• ••• ••• ••• •• ••• 1600 •• • ••• •• • • • •• • •• • ••• 1600 0 150 300 0123 012 048 Bacteriochlorophyll e beta-Carotene Alloxantin Zeaxanthin
Figure 1.38. European land clearance and its effect on ecologic productivity in the
Pettaquamscutt River Estuary. Black cross pollen data are Ambrosia percentages and gray circles represent Rumex percentages. All fossil pigment values are mass accumulation rates (g/cm2yr).
101 DDT (ng/g) PCBs (ng/g) Copper (ug/g) Nickel (ug/g) 0 100 200 0 50 100 0 90 180 01530 2000 2000
1980 1980
1960 1960
1940 1940
Year (A.D.) 1920 1920 Year (A.D.)
1900 1900
1880 1880
1860 1860 0 6000 12000 0 1.5 3 0 90 180 PAHs (ng/g) Cadmium (ug/g) Chromium (ug/g)
Figure 1.39. Compilation of chemical pollutants that exceed sediment quality control guidelines (Long et al., 1995). Light gray shading indicates values exceeding the
ERL concentration. Dark gray shading indicates concentrations that exceed the ERM value.
102 mean ERM Quotient Biogenic Lamination Thickness Chlorophyll a 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 02040 2000 2000
1990 1990
1980 1980
1970 1970
1960 1960
1950 1950 Year (A.D.) Year (A.D.) 1940 1940
1930 1930
1920 1920
1910 1910
1900 1900 80 110 140 0 200 400 0 50 100 TOC Bacteriochlorophyll e Steryl Chlorin Esters
Figure 1.40. Mean Effects Range Median Quotient for chemical pollutants in the
Pettaquamscutt River record (Long et al., 2006) plotted with proxies for estuarine productivity. Gray box represents the time period with the highest ERMQ values.
103 del 15-N Bio Lam Thickness (cm) Chl a MAR Bchl e MAR 2345670 0.1 0.2 0.3 0.4 0.5 0 102030400 250 500 2000 2000
Mean=5.4
1950 1950
1900 1900
Mean=3.7
1850 1850 Year (A.D.) Year (A.D.)
1800 1800
1750 1750
1700 1700
Figure 1.41. Evidence for cultural eutrophication in the Pettaquamscutt River Lower
Basin. Stable nitrogen isotopic values become significantly heavier during the period of residential development in the estuary (ca. 1950-1960). Total productivity shifts to greater values during this time and continues to present, as seen in biogenic laminae thicknesses and Chlorophyll a mass accumulation rates. Green sulfur bacteria initially increase productivity, then decrease since ca. 1970 due to increased turbidity in the water column. Vertical lines with 15N are calculated means of the data.
Upper gray box represents timing of Sewer Project, while lower gray box indicates timing of increased residential development in the watershed. See text for details.
104 Chapter 2: Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments
2.10 Abstract
The climate system in the North Atlantic region is complex and influenced by outside forcings as well as internal modes of the system. Modeling and observational work have suggested that a better understanding of the connections between oceanic- and atmospheric-driven variability could lead to predictive power for North American and European weather patterns. Here we present a new millennial-length proxy record of the estuarine fossil pigment Bacteriochlorophyll e and use this record to investigate cyclic components of North Atlantic climate through the effects on estuarine ecosystems. The time-series exhibits significant cyclic components that can be related to two of the dominant internal modes of climate variability in the region: the North Atlantic Oscillation (NAO) and the Atlantic Multidecadal Oscillation
(AMO). The NAO signal is associated with internal atmospheric variability, while the AMO has been linked to previously modeled and observed changes in thermohaline circulation/meridional heat flux. In our record, the dominant periodicity of the AMO has shifted over time, in concert with Medieval Warm Period – Little Ice
Age – Present Warm Period transitions. A relationship between an intermittent NAO cycle and the AMO signal suggests coupling of the ocean-atmosphere system at multidecadal time scales. Although the causal relationship is not resolved, predictive models of Northern Hemispheric interannual weather patterns and estuarine productivity may be improved by incorporating the results of this study.
105 2.20 Introduction
Although recent work (Jones and Mann, 2004) has greatly advanced our understanding of North Atlantic regional climate variability, its complexity has left questions unanswered regarding interactions between various forcings and internal modes of the climate system. Much of the variability in the region during boreal winter can be attributed to the atmospheric quasi-decadal North Atlantic Oscillation
(NAO) (Hurrell, 1995). In addition, multidecadal oceanic fluctuations associated with the Atlantic Multidecadal Oscillation (AMO) have significant influence on the region’s climate (Delworth and Mann, 2000). Both of these internal modes affect temperature, precipitation, and atmospheric circulation around the North Atlantic
Basin, and subsequently have large effects on ecosystems and society. Due to the short duration of instrumental records, preanthropogenic behavior and low frequency climate variability are difficult to assess in observational records. Therefore, a priority in paleoclimate science is to produce high-quality proxy reconstructions of
North Atlantic climate variability that preserve both decadal and multidecadal periodicities. Such records are crucial to validate general circulation models and to further our understanding of the natural dynamics present in regional climate.
In this study we present a single-proxy reconstruction of estuarine productivity in southern New England that exhibits various cyclic components associated with North Atlantic climate fluctuations. A unique core record of well- preserved photosynthetic pigments from the anoxic, annually laminated (varved) estuarine sediments of the Pettaquamscutt River Estuary, Rhode Island (Figures 2.1,
106 2.2) is used to produce a robustly dated, 980-year (1024-2004 A.D.) time-series with an average temporal resolution of 2 years.
The field site is a small north-south trending estuary immediately southwest of
Narragansett Bay, Rhode Island. In the upper reaches of the estuary are two ice-block depressions formed during the retreat of the Laurentide Ice Sheet at the close of the last glacial maximum (Orr and Gaines, 1973). These kettles are 4-5 times deeper than the rest of the estuary and produce a fjord-like circulation pattern, which isolates bottom waters in the two basins due to density-driven stratification. The isolated water is dominated by hydrogen sulfide and as a result, no bioturbation is evident in the sediment record. We have examined pigments from the deeper of the two basins in order to minimize overturn events, which could affect the sediment record by oxygenating the bottom waters. The primary productivity in this environment is dominated by brown-green anoxygenic photobacteria residing just below the oxycline
(~4 meters), with lesser quantities of oxygenic green phytoplanktonic algae, just above the oxycline (Sieburth and Donaghay, 1993).
We focus this study on the pigment Bacteriochlorophyll e (Bchle), which is produced just below the oxycline of the water column by brown-green sulfur photobacteria (Chen et al., 2001; Sieburth and Donaghay, 1993). Due to the anoxic habitat of these bacteria, Bchle is an ideal pigment to use for climate studies as it avoids zooplankton grazing complications often encountered with phytoplankton productivity studies (Keller et al., 1999). Since local air and water temperatures during the growing season (winter/spring) are significantly correlated to climate fluctuations such as the NAO (Hawk, 1998), we hypothesize that productivity will be
107 associated with regional climate. Therefore, mass accumulation rates (MARs) of
Bchle can be used as a proxy for North Atlantic climate variability over the last millennium.
2.30 Methods
A total of seven sediment freeze cores (Wright, 1980) were taken from the lower basin of the Pettaquamscutt River Estuary between 1999 and 2004 (Figures 2.1,
2.2). In each case, a stainless steel box measuring 165 cm x 20 cm x 8 cm was filled with crushed dry ice and methanol and sealed with a vented lid. This corer was lowered from a pontoon platform or frozen water surface into the sediment and left for 15-20 minutes. During this time, sediment and pore water froze to the outside of the corer box, maintaining the original sediment structure. The corer was raised back to deck and the four side slabs of sediment were removed from the box faces. Each slab was individually wrapped in aluminum foil, kept frozen, and restricted from light exposure during transport, storage, and subsampling.
The age model for the composite sedimentary record was constructed using varve counts, which were analyzed in thin section after impregnation of the sediment.
In preparation for thin section generation, the most intact face of each of the seven cores was cut into segments approximately 6 cm x 4 cm x 3 cm. These chunks were split perpendicular to the laminae in order to archive material that corresponds to the thin section produced. The working sub-chunks were freeze-dried and imbedded with
Spurr resin (Pike and Kemp, 1996; Spurr, 1969). The resulting slab was thin sectioned using standard petrographic technique.
108 Thin sections were scanned on a flat-bed scanner with transparency capabilities under cross-polarized films to produce tagged image file format (tiff) images with resolutions of 1440 dots per inch (dpi) (De Keyser, 1999). The images were analyzed using Adobe Photoshop® and lamination boundaries were marked with the path tool. Paths were exported and processed using an algorithm that counts and measures the thickness of each lamination (Francus et al., 2002). The seven resulting chronologies were crosschecked with each other and compiled in order to produce a master varve chronology. Counting errors were determined to be less than
1% on individual sections.
The varve chronology has been validated using radiometric age controls
(210Pb, 137Cs, 14C) (Lima et al., 2005) (Table 2.1) (Figure 2.3). All samples for radiocarbon analysis were terrestrial macrofossils (leaves) and were analyzed at the
National Ocean Sciences AMS Facility, Woods Hole, MA. Radiocarbon ages were converted to calendar years using the CALIB 4.3 program (Stuiver et al., 1998).
Although one of the radiocarbon dates does not fit the age model within the errors, the weight of evidence from all other age controls supports the varve age model.
Sample NR03-1 92 is anomalously old, however this is most likely due to remobilization of terrestrial macrofossils during European land clearing (Ambrosia and Rumex horizons are at the same core depth; see Chapter 1). Despite this radiocarbon sample issue, the terrestrial origin of the samples is preferable due to the lack of a reservoir correction.
Frozen sediment samples for fossil pigment analysis were scrapped off of the archive sediment slabs produced during the thin-sectioning process. This approach
109 enabled us to precisely determine the age of each sample. A total of 495 samples were analyzed over the 980-year record. For each sample, approximately 0.25 grams of wet sediment was weighed into a glass scintillation vial and placed in a water bath sonicator at 4º C. The pigments were extracted by successive sonication (1 minute) in cold acetone until the extracts were colorless. The combined extracts were filtered through a 0.45 um Acrodisc 13 PTFE membrane filter. Care was taken throughout the extraction process to keep the samples and acetone on ice and in the dark as much as possible in order to preserve the pigments from degrading.
Pigments were analyzed by high performance liquid chromatography (HPLC)
(Bianchi et al., 1996; Wright et al., 1991). The HPLC system consisted of a Waters
2690 Alliance separation module with a 996-photodiode array detector (PDA) and a
474 scanning fluorescence detector with excitation set at 410 nm and emission at 660 nm. Pigments were identified and quantified by comparing retention times and PDA spectra to authentic standards. Bchle concentrations (g/l) were converted to mass accumulation rates (MARs) (g/cm2yr) using dry bulk density measurements and annual varve sedimentation rates (Chapter 1).
In order to run a spectral analysis on the Bchle MAR time-series, the data were resampled at 2-years. In order to ensure a normal distribution of the data, a log transformation was applied to this resampled series. Spectral analysis was performed using the multi-taper method with three tapers. Significant peaks were identified with respect to a first order serially autocorrelated process (AR(1)) at the 90%, 95%, and
99% confidence levels (Mann and Lees, 1996).
110 The wavelet transform was computed with the entire Bchle MAR time-series, resampled at 2 years, using a Morlet wavelet function. The ends of the series padded with zeroes to prevent spurious data at the edges of the transform (Torrence and
Compo, 1998).
Cross spectral analysis was performed between the Bchle MAR and WNAO time series (1824-1960) (Jones et al., 1997) using the ARAND Crospec program.
Both series were first resampled at 2-years in Analyseries (Paillard et al., 1996), and then run on the Crospec program with 30 lags.
2.40 Results and Discussion
2.41 Climate Cycles
It is apparent that the reconstructed time-series of Bchle MAR exhibits variability at a number of different time scales (Figure 2.4, 2.5A). In order to examine this variability quantitatively, spectral analysis was performed on the portion of the time-series from 1058–2004 A.D. This time window was chosen in order to avoid the environmental contamination of the climate signal in the early part of the record due to water column stabilization after marine inundation and potentially questionable preservation of photosynthetic pigments during this time. A log transformation of the data was performed to obtain a more normal distribution, and the multi-taper method was performed. The spectral analysis exhibits five periodicities that are statistically significant at greater than 99% above the AR(1) red noise background (Figure 2.4) (Mann and Lees, 1996).
111 At the subdecadal end of the spectrum, there are two significant periodicities in the Bchle MAR time-series. First, there is a strong 8.0-year cycle, which has been shown to dominate NAO instrumental and proxy records (Cook, 2003; Hurrell, 1995;
Jones et al., 1997). Further support of the NAO origin of this cycle is gained through cross-spectral analysis between the Jones (1997) NAO index and the overlapping segment of the Bchle MAR time-series from 1824-1960 A.D. (Figure 2.6). This analysis demonstrates a significant coherence in the ca. 8-year spectral peak between the two records prior to local cultural eutrophication of the estuary after ca. 1960
A.D. (Hubeny and King, 2003). In addition to the 8-year cycle, there is significant power at the 5.5-year spectral peak. Although not as commonly cited as being associated with the NAO, similar significant periodicities have been found in NAO records (Luterbacher et al., 2002).
At the quasi-decadal range, the Bchle time-series has a significant spectral peak at 11.6-years. The origin of this cycle is unknown, but there are a number of possibilities. Commonly, ca. 11-year cycles in climate records are attributed to solar variability associated with the Schwabe Cycle due to coincidence in cycle length.
Recent studies have demonstrated potential communication of this rather small solar forcing to the atmosphere through upper tropospheric/lower stratospheric interactions
(Labitzke and van Loon, 1997). However, there has not been any definitive evidence in the literature linking the Schwabe Cycle to surface climate variability. Other potential candidates for driving ca. 11-year cycles have been proposed (Mann and
Park, 1996), and include potentially unstable extratropical ocean-atmosphere
112 interactions in the North Atlantic Ocean between the oceanic gyre system and atmospheric circulation through air-sea heat exchanges (Latif and Barnett, 1994).
At the multidecadal frequency ranges, significant power is observed in peaks centered at 38.5-years and 95.9-years. Both of these peaks can be attributed to the
AMO, which has been observed both in proxy (Delworth and Mann, 2000; Gray et al., 2004) and modeling studies (Delworth and Mann, 2000; Knight et al., 2005).
Although the AMO has commonly been cited as having a periodicity of 65-80 years, current proxy and modeling results exhibit a more complex picture with variable periodicities ranging from perhaps as low as 30-40 years (Gray et al., 2004; Knight et al., 2005) to 100 years and more (Delworth and Mann, 2000; Gray et al., 2004;
Knight et al., 2005). Despite this lack of a definitive period associated with the
AMO, there is general agreement that the phenomenon is internal and involves ocean- atmosphere coupling as well as variability in the strength of the thermohaline circulation (Delworth and Mann, 2000; Knight et al., 2005). In addition to the aforementioned AMO studies, various other proxy reconstructions of North Atlantic climate variability have found significant power in similar multidecadal spectral peaks from regional (Appenzeller et al., 1998; Proctor et al., 2002), and hemispheric
(Luterbacher et al., 2002; Mann et al., 1995) proxy and instrumental records.
2.42 Variable Cycles over the Last Millennium
Examination of a wavelet transform of the Bchle MAR time-series (Torrence and Compo, 1998) shows temporal variability in periodic components over the last millennium (Figure 2.5B). In the low-frequency sector, there is variable dominance
113 between the 38.5- and 95.9-year AMO cycles. In order to accentuate this phenomenon, we have band-pass filtered the time-series at the two AMO frequencies and summed the results to produce a record of total multidecadal variability in this record (Figure 2.5C). It is apparent from these analyses that both frequencies are present throughout the record, but there is a preference for the lower frequency component during the Little Ice Age compared to the Medieval or Present Warm
Periods. Further support for a distinct Little Ice Age in the record comes from the biologic lamination thickness associated with individual varves (Figure 2.5D). The decadally smoothed thicknesses of these laminations are statistically thinner during the Little Ice Age than either of the warm periods, showing that overall productivity and runoff were lower due to dominance of cool, dry climate during this time.
The variability found in the AMO periodicity is not unique to our record. For instance, in their tree ring reconstructed AMO time-series, Gray et al. (2004) demonstrate wide periodic dominance in the 40-128 year band from 1567 A.D. to the late nineteenth century. After which the dominance shifted to a narrower band centered at about 40 years. Although this may be an artifact of their technique, the timing matches our frequency transition (Figure 2.5B). Recent modeling studies provide additional support for variable AMO periodicities. It has been demonstrated using the HadCM3 climate model that the AMO can exhibit unforced variability between roughly centennial cyclicity and a higher-frequency (ca. 40 year) cyclic component (Knight et al., 2005).
The NAO component of the wavelet exhibits intermittent oscillatory behavior throughout the record (Figure 2.5B). This behavior is not unlike that observed in
114 Greenland ice records (Appenzeller et al., 1998), suggesting that it is not a unique feature to our proxy or geographic location. The intermittent nature of the NAO signal in the wavelet is related to the AMO signal in such a way that during phases of active NAO cyclicity, the AMO signal tends to be at a peak. This relationship can be seen in Figure 2.5 by comparing the 8-year NAO periodicity band in the wavelet to the vertical lines drawn through peaks in the data filtered at multidecadal periods.
This relationship is clearest during the Medieval Warm Period, when the amplitude of multidecadal variability is large. The amplitudes of the multidecadal variability are lower during the Present Warm Period, and the relationship between NAO and AMO is not as apparent. During the Little Ice Age the relationship between the AMO and the intermittent NAO behavior is similar to warmer periods, although there is a shift toward lower-frequency multidecadal variability (Figure 2.5; dashed vertical lines).
The relationship between the NAO and AMO in our time-series suggests a coupling between the atmosphere and ocean at multidecadal time scales (Figure 2.1).
The causal relationship, however, cannot be determined from these data. There is strong modeling evidence that the dominant influence on climate variability in the extratropical Northern Hemisphere is from the atmosphere, not from oceanic circulation variability (Jones and Mann, 2004). Recent modeling studies have argued for direct radiative forcing of the NAO (Shindell et al., 2001; Shindell et al., 2003).
The NAO has a known effect on sea surface temperature through atmospheric windstress changes, which can in turn affect thermohaline circulation via associated changes in the density of seed waters (Delworth and Dixon, 2000). Therefore, if the
NAO is being forced by solar variability, then the low frequency nature of the AMO
115 could be attributed to atmospheric changes. Observational studies support the interpretation that NAO is forcing the Gulf Stream (and in turn thermohaline circulation) by showing that the NAO leads changes in water properties associated with Gulf Stream migration with leads of 1.5 to 2 years (Rossby and Benway, 2000;
Taylor and Stephens, 1998).
Although the NAO has been established as a driver for sea surface temperatures over the North Atlantic, these affects may be restricted to decadal and subdecadal timescales (Czaja et al., 2003). If this is actually the case, then the possibility remains that multidecadal oscillations in the North Atlantic region could be forced by oceanic dynamics (Kushnir, 1994). The primary suspect for this driver is variability of the thermohaline circulation, which is related to the AMO. Modeling studies have supported this hypothesis by demonstrating that sea surface temperature variability can influence North Atlantic climate at multidecadal timescales (Sutton and Hodson, 2003), and that if the deep ocean dynamics are dampened, that the NAO can lose the red portion of its spectrum (Wu and Gordon, 2002). In addition, there is evidence that multidecadal variability in the NAO can be predicted with a knowledge of North Atlantic sea surface temperatures (Rodwell et al., 1999).
Either of the above scenarios is possible but cannot be ascertained from this time-series. The key to the problem is to determine which aspect of the system has enough of a memory to maintain periodic fluctuations of up to 100 years. The underlying assumptions as stated above are that the sun and/or the oceans provide this memory. Since studies suggesting atmospheric forcing of the oceans do not include dynamic oceanic processes (Shindell et al., 2001; Shindell et al., 2003) and studies
116 suggesting oceanic forcing of the atmosphere do not incorporate dynamic stratospheric/tropospheric interactions (Sutton and Hodson, 2003; Wu and Gordon,
2002), the ultimate driver of these periodicities remains unresolved.
2.50 Conclusions
In conclusion, we have utilized the photosynthetic estuarine pigment
Bacteriochlorophyll e to investigate North Atlantic climate dynamics over the last millennium through ecological responses. The pigments exhibit variable frequencies over the length of the time-series, which can be associated with the North Atlantic
Oscillation and the Atlantic Multidecadal Oscillation. The Atlantic Multidecadal
Oscillation has varied in dominant frequency, and the North Atlantic Oscillation cyclicity has been intermittent throughout the record. The intermittent nature of the
North Atlantic Oscillation signal is related to the Atlantic Multidecadal Oscillation, which suggests coupling between the atmosphere and ocean. The driver in this relationship, however, has not yet been determined.
117 2.60 References
Appenzeller, C., Stocker, T. F., and Anklin, M., 1998, North Atlantic Oscillation dynamics recorded in Greenland ice cores: Science, v. 282, p. 446-449.
Bianchi, T. S., Demetropoulos, A., Hadjichristophorou, M., Argyrou, M., Baskaran, M., and Lambert, C. D., 1996, Plant pigments as biomarkers of organic matter sources and coastal waters of Cyprus (eastern Mediterranean): Estuarine, Coastal, and Shelf Science, v. 42, p. 103-115.
Chen, N., Bianchi, T. S., McKee, B. A., and Bland, J. M., 2001, Historical trends of hypoxia on the Louisiana shelf: application of pigments as biomarkers: Organic Geochemistry, v. 32, p. 543-561.
Cook, E. R., 2003, Multi-proxy reconstructions of the North Atlantic Oscillation (NAO) Index: A critical review and a new well-verified Winter NAO Index reconstruction back to AD 1400, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 63-79.
Czaja, A., Robertson, A. W., and Huck, T., 2003, The role of Atlantic Ocean- atmosphere coupling in affecting North Atlantic Oscillation variability, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 147-172.
De Keyser, T. L., 1999, Digital scanning of thin sections and peels: Journal of Sedimentary Research, v. 69, no. 4, p. 962-964.
Delworth, T. L., and Dixon, K. W., 2000, Implications of the recent trend in the Arctic/North Atlantic Oscillation for the North Atlantic thermohaline circulation: Journal of Climate, v. 13, p. 3721-3727.
Delworth, T. L., and Mann, M. E., 2000, Observed and simulated multidecadal variability in the Northern Hemisphere: Climate Dynamics, v. 16, p. 661-676.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Gray, S. T., Graumlich, L. J., Betancourt, J. L., and Pederson, G. T., 2004, A tree-ring based reconstruction of the Atlantic Multidecadal Oscillation since 1567 A.D.: Geophysical Research Letters, v. 31, L12205, doi:10.1029/2004GL019932.
118 Hawk, J. D., 1998, The role of the North Atlantic Oscillation in winter climate variability as it relates to the winter-spring bloom in Narragansett Bay [Master of Science thesis]: University of Rhode Island, 148 p.
Hubeny, J. B., and King, J. W., 2003, Anthropogenic eutrophication as recorded by varved sediments in the Pettaquamscutt River Estuary, Rhode Island, USA: GSA Abstracts with Programs, v. 35, no. 6, p. 282.
Hurrell, J. W., 1995, Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation: Science, v. 269, p. 676-679.
Jones, P. D., Jonsson, T., and Wheeler, D., 1997, Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland: International Journal of Climatology, v. 17, p. 1433- 1450.
Jones, P. D., and Mann, M. E., 2004, Climate over past millennia: Reviews of Geophysics, v. 42, RG2002, doi:10.1029/2003RG000143.
Keller, A. A., Oviatt, C. A., Walker, H. A., and Hawk, J. D., 1999, Predicted impacts of elevated temperature on the magnitude of the winter-spring phytoplankton bloom in temperate coastal waters: A mesocosm study: Limnology and Oceanography, v. 44, no. 2, p. 344-356.
Knight, J. R., Allan, R. J., Folland, C. K., Vellinga, M., and Mann, M. E., 2005, A signature of persistent natural thermohaline circulation cycles in observed climate: Geophysical Research Letters, v. 32, L20708, doi:10.1029/2005GL024233.
Kushnir, Y., 1994, Interdecadal variations in North Atlantic sea surface temperature and associated atmospheric conditions: Journal of Climate, v. 7, p. 141-157.
Labitzke, K., and van Loon, H., 1997, The signal of the 11-year sunspot cycle in the upper troposphere-lower stratosphere: Space Science Reviews, v. 80, no. 3-4, p. 393-410.
Latif, M., and Barnett, T. P., 1994, Causes of decadal climate variability over the North Pacific and North America: Science, v. 266, p. 634-637.
Lima, A. L., Hubeny, J. B., Reddy, C. M., King, J. W., Hughen, K. A., and Eglinton, T. I., 2005, High-resolution historical records from Pettaquamscutt River basin sediments: 1. 210Pb and varve chronologies validate record of 137Cs released by the Chernobyl accident: Geochimica et Cosmochimica Acta, v. 69, no. 7, p. 1803-1812.
119 Luterbacher, J., Xoplaki, E., Dietrich, D., Jones, P. D., Davies, T. D., Portis, D., Gonzalez-Rouco, J. F., von Storch, H., Gyalistras, D., Casty, C., and Wanner, H., 2002, Extending North Atlantic Oscillation reconstructions back to 1500: Atmospheric Science Letters, v. 2, p. 114-124.
Mann, M. E., and Lees, J. M., 1996, Robust estimation of background noise and signal detection in climatic time series: Climatic Change, v. 33, p. 409-445.
Mann, M. E., and Park, J., 1996, Joint spatiotemporal modes of surface temperature and sea level pressure variability in the Northern Hemisphere during the last century: Journal of Climate, v. 9, no. 9, p. 2137-2162.
Mann, M. E., Park, J., and Bradley, R. S., 1995, Global interdecadal and century- scale climate oscillations during the past five centuries: Nature, v. 378, p. 266- 270.
Orr, W. L., and Gaines, A. G., Jr., 1973, Observations on rate of sulfate reduction and organic matter oxidation in the bottom waters of an estuarine basin: the upper basin of the Pettaquamscutt River (Rhode Island), in Tissot, B., and Bienner, F., eds., Advances in Organic Geochemistry, Proceedings of the 6th International Meeting on Organic Geochemistry: Paris, Editions technip, p. 791-812.
Paillard, D., Labeyrie, L., and Yiou, P., 1996, Macintosh program performs time- series analysis: Eos, v. 77, p. 379.
Pike, J., and Kemp, A. E. S., 1996, Preparation and analysis techniques for studies of laminated sediments, in Kemp, A. E. S., ed., Palaeoclimatology and Palaeoceanography from laminated sediments, Geological Society Special Publication No. 116, p. 37-48.
Proctor, C. J., Baker, A., and Barnes, W. L., 2002, A three thousand year record of North Atlantic climate: Climate Dynamics, v. 19, p. 449-454.
Rodwell, M. J., Rowell, D. P., and Folland, C. K., 1999, Oceanic forcing of the wintertime North Atlantic Oscillation and European climate: Nature, v. 398, p. 320-323.
Rossby, T., and Benway, R. L., 2000, Slow variations in mean path of the Gulf Stream east of Cape Hatteras: Geophysical Research Letters, v. 27, no. 1, p. 117-120.
Shindell, D. T., Schmidt, G. A., Mann, M. E., Rind, D., and Waple, A., 2001, Solar Forcing of Regional Climate Change During the Maunder Minimum: Science, v. 294, p. 2149-2152.
120 Shindell, D. T., Schmidt, G. A., Miller, R. L., and Mann, M. E., 2003, Volcanic and solar forcing of climate change during the Preindustrial Era: Journal of Climate, v. 16, p. 4094-4107.
Sieburth, J. M., and Donaghay, P. L., 1993, Planktonic methane production and oxidation within the algal maximum of the pycnocline: seasonal fine-scale observations in an anoxic estuarine basin: Marine Ecology Progress Series, v. 100, p. 3-15.
Spurr, A. R., 1969, A low-viscosity epoxy resin embedding medium for electron microscopy: Journal of Ultrastructure Research, v. 26, p. 31-43.
Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormac, G., van der Plicht, J., and Spurk, M., 1998, INTCAL98 radiocarbon age calibration, 24,000-0 cal BP: Radiocarbon, v. 40, no. 3, p. 1041-1083.
Sutton, R. T., and Hodson, D. L. R., 2003, Influence of the ocean on North Atlantic climate variability 1871-1999: Journal of Climate, v. 16, p. 3296-3313.
Taylor, A. H., and Stephens, J. A., 1998, The North Atlantic Oscillation and the latitude of the Gulf Stream: Tellus, v. 50A, p. 134-142.
Torrence, C., and Compo, G. P., 1998, A practical guide to wavelet analysis: Bulletin of the American Meteorological Society, v. 79, no. 1, p. 61-78.
Wright, H. E., Jr., 1980, Cores of soft lake sediments: Boreas, v. 9, p. 107-114.
Wright, S. W., Jeffrey, S. W., Mantoura, R. F. C., Llewellyn, C. A., Bjornland, T., Repeta, D., and Welschmeyer, N., 1991, Improved HPLC method for the analysis of chlorophylls and carotenoids from marine phytoplankton: Marine Ecology Progress Series, v. 77, p. 183-196.
Wu, P., and Gordon, C., 2002, Oceanic influence on North Atlantic climate variability: Journal of Climate, v. 15, p. 1911-1925.
121 Table 2.1. Pettaquamscutt River Radiocarbon Data
Sample Composite Material Lab 14C age 2 calibrated Relative Name Depth ID (yr BP) age ranges (yr area under (cm) AD) distribution NR03-1 64 Leaf OS- 225 ± 40 1523 - 1568 0.052 78 pieces 47531 1627 - 1692 0.390 1727 - 1812 0.460 1919 - 1949 0.098 NR03-1 78 Leaf OS- 345 ± 40 1461 - 1640 1.000 92 pieces 47616 NR04-2 124 Leaf OS- 435 ± 40 1411 - 1519 0.925 J5 pieces 47532 1593 - 1622 0.075 NR04-2 186 Leaf OS- 915 ± 110 900 - 920 0.021 T3 pieces 47749 940 - 1290 0.979
All analyses were conducted at the National Ocean Sciences AMS Facility.
Calibrations were calculated using the CALIB 4.3 program (Stuiver et al., 1998).
122 Figure 2.1. Detailed locus map of the Pettaquamscutt River Estuary, Rhode Island,
USA. Note coring location in the lower kettle basin of the upper portion of the estuary.
123 Figure 2.2. North Atlantic region showing locations of the Pettaquamscutt River
Estuary (circle) and the regional pressure centers associated with a positive NAO during (A) increased, and (B) decreased meridional oceanic heat flux (AMO). Vector symbolizes northward heat flux via Gulf Stream/North Atlantic Current and hatched areas symbolize zones of positive atmospheric temperature anomalies during a positive NAO. Note the southward shift of the atmospheric system coincident with reduced oceanic meridional heat flux.
124 Figure 2.3. Age model for the varved portion of the Pettaquamscutt River lower basin sediment record. Composite depth scale was computed using mean varve thicknesses for each year, and adjusted to a common zero depth at 1999 AD. 210Pb data from
Lima et al. (Lima et al., 2005). Additional age controls include 137Cs peaks in 1986 and 1963 (Lima et al., 2005), rise from background concentrations of organic pollutants (PCBs and DDT in the 1930s) and metals (Pb in the 1850s), as well as
European settlement and clear-cutting in the region (Ambrosia and Rumex rises from background around 1700). Radiocarbon data are presented in Table 2.1, and the most significant calibrated date for each sample is plotted here with the associated error bars. One radiocarbon sample (*) is anomalously old due to land clearance and remobilization of older macrofossils from watershed during European settlement in the area.
125 Figure 2.4. Multi-taper spectral analysis of the Pettaquamscutt River Bchle MAR time-series (1058-2004 A.D.). Dashed lines represent the 90%, 95%, and 99% confidence levels with respect to AR(1) red noise background. All periodicities significant above 99% are labeled.
126 Figure 2.5. A: Pettaquamscutt River Bchle MAR time-series along with (B) the corresponding wavelet transform (Torrence and Compo, 1998). Gray scale indicates power, which is scaled to percent total power, and hatched areas illustrate the cone of influence and hence edge effects of the transform. North Atlantic Oscillation (NAO) and multidecadal variability (MDV) periodicities are labeled. C: Sum of Bchle MAR time-series band-pass filtered at 38.5 and 95.9 years used to represent multidecadal variability. The Medieval Warm Period, Little Ice Age, Present Warm Period
(“PWP”), and local cultural eutrophication (“CE”) times are labeled at the top of the figure, and are supported by (D) decadally-smoothed biologic lamination thicknesses, which represent total productivity and runoff. Vertical lines indicate multidecadal amplitude peaks, which are related to the intermittent 8-year NAO periodicity. See text for full discussion of the data.
127 Figure 2.6. Cross spectral analysis of Pettaquamscutt River Bchle and WNAO (Jones et al., 1997) time series (1824 – 1960). Note the significant coherence between the series in the ca. 8-year NAO band.
128 Chapter 3: Late Holocene precipitation variability in southern New England and the relationship to Northern Hemisphere teleconnection patterns
3.10 Abstract
Hydroclimatology and associated water budget planning are wild cards in future global change scenarios due to the complex and dynamic nature of precipitation patterns and the human need for water. One way to address the uncertainties involved is to increase our understanding of past precipitation variability in order to determine patterns associated with specific regions. If these patterns can be correlated to teleconnection indices, then the applicability of the findings increases spatially. Along these lines, a time-series of clastic laminae thickness from the
Pettaquamscutt River Estuary’s varved record has been used to examine hydrologic variability in the region. Results indicate that laminae thicknesses are strongly linked to Rhode Island precipitation and drought state during years of instrumental overlap.
In addition to these regional climate variables, the time-series is also strongly correlated to the Pacific/North American (PNA) teleconnection pattern, which influences the shape of the sup-polar jet stream, and hence tropospheric circulation, over North America. This result suggests that meridional atmospheric flow leads to higher precipitation in New England due to the associated dominance of moist Gulf of Mexico and Atlantic coastal air masses. Dry periods are associated with dry continental air masses that become dominant during zonal flow. This relationship allows for the generation of a proxy-based reconstruction of the PNA over the last millennium. The reconstruction is applicable to many areas of the United States,
129 including the Southwest, and demonstrates that the last four to five decades have had anomalously high PNA values. In regions that are extremely drought sensitive, the recent dominance of high PNA values suggests that water resource planning that is based on historical weather data may misrepresent natural hydroclimate variability.
Finally, the dominant periodic components of the time-series are in the 8- and 3-year ranges, and amplitudes of these and other cycles were smaller during the Little Ice
Age in comparison with the rest of the series. The identification of these cycles in the data set helps in our understanding of PNA and New England precipitation variability.
3.20 Introduction
One of the most pressing concerns associated with future global change is water availability and its effects on human health, economic activity, and ecosystem functions (Mooney et al., 2005; Reitler et al., 2004; Shiklomanov and Rodda, 2003).
Instrumental records of precipitation have shown recent increases in Northern
Hemisphere mid- to high-latitude precipitation, but tropical trends are not as clear
(Folland et al., 2001). Although the increasing trends in precipitation are good, there is significant modeling evidence suggesting that large regional anomalies and increases in the variability of future precipitation may be more important than the globally averaged projected increase (Cubasch et al., 2001; Giorgi et al., 2001).
Therefore, it is in society’s best interest that we work to better understand natural precipitation variability at hemispheric to regional scales so that we might better predict future water availability. Recent modeling studies have made progress
130 predicting future changes to the hydrologic cycle (Allen and Ingram, 2002; Milly et al., 2005), however there are still gaps in our understanding of the underlying mechanisms of variability. In addition, a lack of high-resolution precipitation reconstructions over past millennia (Jones and Mann, 2004) makes it more difficult for modelers to test against known conditions from the past to validate processes represented in the models.
Although not traditionally classified as a drought sensitive region, New
England is prone to both flood hazards (Leathers et al., 1998) and soil, meteorological, and hydrological droughts (Leathers et al., 2000; Namias, 1966).
Since a large population resides in this area of the United States and there is active agriculture in the region, it is timely to increase our understanding of precipitation variability in the region to help in planning for water resources. In addition to local concerns, New England is located in a geographically sensitive region for storm tracking and hemispheric teleconnections (Bradbury et al., 2002a; Bradbury et al.,
2003; Yarnal and Leathers, 1988). Therefore, a more complete understanding of the processes that are controlling New England precipitation variability may be applicable in other regions that are affected by the same teleconnection patterns.
In this study, a varved sediment record from the Pettaquamscutt River
Estuary, Rhode Island (Figure 3.1) is used as a proxy for local precipitation variability. The Lower Basin of the estuary has unique bathymetry, as it is an ice block depression that has been inundated by marine waters. Therefore, a stable density stratification has been present for the last millennium (Chapter 1). The stratified water has cut off circulation to the bottom, thereby depleting the oxygen and
131 inhibiting benthic infauna and epifauna from living in the deep hole. The lack of bioturbation has allowed sediment to accumulate undisturbed, and since there is an annual productivity cycle in the water column, varved sediments are preserved. Each varve is composed of a summer biogenic layer deposited after the annual phytoplankton bloom and a clastic layer that is deposited during the remainder of the year. As runoff is the mechanism that introduces siliciclastic sediment to the deep water of the estuary, it is hypothesized that the thickness of the clastic laminae are correlated to local precipitation and runoff. The regional applicability of the hydrologic signal from southern New England includes all of New England, west to eastern Ohio, and south to Northern Virginia and Delaware (Lins, 1997).
Teleconnection influences have been proposed as having influence on New
England hydroclimatology through their influences on mid-tropospheric circulation patterns and storm tracking (Bradbury et al., 2002a; Bradbury et al., 2003; Yarnal and
Leathers, 1988). This hypothesis will be tested with Rhode Island precipitation and clastic laminae data in order to test the validity in southern New England.
Specifically three teleconnection indices will be investigated: the Pacific/North
American pattern (PNA) (Wallace and Gutzler, 1981), the North Atlantic Oscillation
(NAO) (Jones et al., 1997), and the Southern Oscillation Index (SOI) (Ropelewski and Jones, 1987), which is the atmospheric component of the El Niño/Southern
Oscillation (ENSO) climate phenomenon. Recent work has suggested that the PNA and NAO represent the two principal patterns of variability in the extratropical
Northern Hemisphere winter circulation (Quadrelli and Wallace, 2004), which justifies testing for their influence on New England hydroclimate. The SOI is
132 examined due to its global influence on atmospheric circulation and associated influence on climate (Jones and Mann, 2004).
3.30 Methods
A total of eight sediment freeze cores (Glew et al., 2001) were taken from the
Lower Basin of the Pettaquamscutt River between 1999 and 2004 (Figure 3.1). Thin sections were produced down each core in order to examine the laminated sediments in detail (Lamoureux, 2001; Pike and Kemp, 1996). Thin sections were scanned under cross-polarized light with a flat bed scanner to produce TIFF digital images.
These images were analyzed in Photoshop® by marking all lamina boundaries with the path tool. The exported paths were run through an algorithm designed to calculate widths in centimeters from digital pixel widths (Francus et al., 2002).
Counts from individual cores were crosschecked with each other to produce a master varve chronology. Recounts from individual sections yield counting errors of less than 1%. The chronology for the master varve age model was validated by comparing the curve to independent radiogenic age controls (Figure 3.2, Chapter 1).
The clastic laminae time-series was extracted from the master varve chronology to produce an annually resolved record of clastic input to the Lower
Basin. Since sediment compacts over time as it is buried by additional sediment, a correction must be made to varve thickness time-series if they are to be used as paleoclimate proxies (Hughen et al., 2000; Nederbragt and Thurow, 2005). As sediment is compacted, the porosity and water content decrease. A correction for water content, therefore, will remove the non-climatic trend associated with
133 differential compaction (Nederbragt and Thurow, 2005). For the Pettaquamscutt
River, water content was measured for 1 cubic centimeter samples from freeze cores.
It was determined that the percent water content is 93% and the top of the cores, and decreases exponentially to 78% at the base of the varved section. An exponential equation (r2=0.9913) was fit to the data such that:
y = (0.000000002 * e 0.0114x) + 78%, where y is equal to the percent water and x is equal to the year of deposition. This correction was applied to the clastic laminae thickness time-series in order to normalize all thickness values to 78% water content. All values greater than two standard deviations from the mean were removed in order to diminish the effects of episodic events on the time-series. Most of these thick layers were coarser grained
(fine sand), suggesting that they are turbidites resulting from turbidity currents down the slopes of the basin, and therefore are not directly related to average precipitation in that given year. Instead, they represent episode events, which may or may not be related to a single rain storm.
The compaction-corrected clastic laminae time-series was compared to instrumental records of climate variables and teleconnection indices using Pearson correlation tests. All analyses were performed between annually resolved clastic laminae thicknesses and monthly, seasonal, and annual climate variable and index values. Years with missing data were omitted from analyses. In addition, decadal correlations were calculated in order to examine low-frequency aspects of the climate reconstruction. Decadally-smoothed clastic laminae, climate variable, and index
134 time-series were generated with a 9-year running average smoothing function in
Excel®.
Significance (p) was determined for all correlation coefficients using
GraphPad© software. The requisite degree of freedom (DOF) was calculated as the population (N) minus 1 for raw time-series. For the smoothed series, it was necessary to calculate the DOF as:
DOF = N * (flow / fNyquist), where flow is the frequency of the low-frequency cutoff (1/9yr in this case), and fNyquist is the Nyquist frequency (0.5yr in this case). Correlations were determined significant if they were calculated as above the 95% confidence level (p<0.05).
3.40 Results and Discussion
3.41 Precipitation and Drought Reconstruction
The compaction-corrected clastic lamina thickness time-series, along with the decadally smoothed time-series, is plotted in Figure 3.3. It is apparent from the plot that the clastic layer thickness has exhibited a broad range of variability over the last millennium. Statistical tests were performed in order to verify climate linkages with the time-series. The three climate variable data sets most relevant to water availability studies are precipitation, drought, and runoff. Monthly Rhode Island precipitation and drought data were both obtained from NOAA’s National Climate
Data Center (Guttman and Quayle, 1996; Karl et al., 1986). Precipitation values are total monthly values and the Palmer Drought Severity Index (PDSI) is used for the drought index. State data are used in order to examine the state/regional patterns of
135 hydroclimate. There are not any instrumental records of runoff or stream flow in the
Pettaquamscutt River Estuary, so a nearby station needed to be located. Although the watershed is larger than the Pettaquamscutt’s, the proximity and length of record for the Hunt River, East Greenwich, Rhode Island was chosen as the best approximation for runoff in the area. These data were obtained from the United States Geological
Survey’s Water Resources webpage:
http://nwis.waterdata.usgs.gov/ri/nwis/discharge/?site_no=01117000 accessed on 12 January 2006. Data are compared at both the annual and decadal scales. Decadal correlations are perhaps the most realistic since the smoothing functions remove much of the stochastic nature inherent to the atmosphere and focus on the broader-scale climatic processes.
The correlations between the clastic laminae thicknesses and climate indices are listed in Table 3.1. The correlations to both precipitation and PDSI are both very strong for both the annual and decadally smoothed time-series (p<0.0001), suggesting that the clastic laminae of the Pettaquamscutt River preserve hydrologically forced variability for the region. The robustness of the correlation is evident when it is realized that at decadal time scales, the laminae thicknesses explain more than 70% of the precipitation variance (i.e. r2=0.71). The correlation to Hunt River discharge is statistically significant at the annual and decadal time scales (p=0.0021 and p=0.0087, respectively), however the correlations are not as strong as those to precipitation and
PDSI. This is presumably due to the different watershed conditions between the Hunt
River and Pettaquamscutt River Estuary, specifically the lack of major fluvial input to the Lower Basin and the steep west bank that favors runoff.
136 More information can be obtained regarding the mechanism of lamina formation by regressing the laminae thickness time-series to monthly and seasonal precipitation values (Figure 3.4). In this case, the strongest correlations are observed in the winter and spring months (December through May), suggesting that the underlying mechanism controlling much of the precipitation variability in Rhode
Island is occurring during these months. This result is consistent with the varve formation model proposed in Chapter 1, in which the biogenic layer is deposited during the summer months as organic matter settles through the water column during and after the annual primary-producer bloom. This mechanism explains the low correlation to summer precipitation values. A similar pattern is observed with the
PDSI, suggesting that precipitation and drought are closely linked in New England.
One potential concern with using variations in lamination thickness to infer past changes in precipitation is that anthropogenic changes may lead to different sediment transport dynamics than were operating during pre-European settlement. If this is the case, then a modern regression may be inappropriate for reconstructions.
The potential concern in the Pettaquamscutt River Estuary is the accelerated residential development with associated storm water drain installation that occurred ca. 1960 (Ernst et al., 1999). These drains could possible increase the efficiency of sediment transport during precipitation events, thereby altering the relationship between clastic input (lamina thickness) and precipitation. To address this concern, correlation coefficients were determined between decadally smoothed laminae thicknesses and regional hydrologic variables for the time periods pre-1960 and post-
1960 (Table 3.2). Although the r values change slightly from the original values
137 (Table 3.1), correlations to precipitation and drought are significant both before 1960 and after 1960. Hunt River flow is only significant after 1960 in this analysis. The low p value for the pre-1960 Hunt River flow, however, is largely a function of the low degree of freedom (DOF=3), and may therefore be interpreted with some caution.
These lines of evidence suggest that the construction of storm drains did not introduce a previously non-existent climate-sedimentological connection, and that it is appropriate to examine the geologic record of clastic laminae thicknesses as a proxy for moisture fluctuations in the region.
The true power of these correlations is the ability to extend instrumental records back to the beginning of the varve time-series (1026 A.D.). In order to do this, the decadally-smoothed clastic laminae thicknesses were regressed against decadally-smoothed PDSI values for the period 1899-1997 A.D. and the equation was calculated as:
PDSI = (20.12 * Clastic Lamina Thickness) – 3.21.
As was noted above, the correlation between the laminae thicknesses and PDSI is strong and robust (r=0.83, p<0.0001) and explains 69% of the variance (r2=0.69).
Therefore, the regression equation can be used to reconstruct decadal trends in the
PDSI for Rhode Island, and perhaps the northeastern United States (Lins, 1997).
Figure 3.5 illustrates the reconstructed PDSI values. The period associated with
European land clearance and increased erosion at the turn of the eighteenth century is highlighted with a gray box. This interval represents a time in which anthropogenic changes to the watershed produced anomalously high PDSI values by increasing the efficiency of sediment transport to deeper water. Since the effect of this clearance
138 was only observed for about two decades (Chapter 1.52), it can be interpreted as merely an anomalous spike in the PDSI reconstruction, unrelated to climate changes at this time.
Anthropogenic influences aside, the PDSI reconstruction is powerful in its realization of past hydroclimatology in the northeastern United States during winter and spring seasons. There has been much variability at a number of scales, but overall it is evident that the historical period has been relatively wet, with much wetter than normal conditions over the last half of a century (Figure 3.5). The late sixteenth and early seventeenth centuries were the driest on record, with little variability and PDSI values suggesting drought conditions. In an attempt to understand the broader connections of this PDSI reconstruction, a comparison can be made between the PDSI reconstruction and hemispheric climate reconstructions.
Perhaps the most robust hemispheric reconstruction in the literature is that of average
Northern Hemisphere temperature anomalies (Mann and Jones, 2003), and some similarities are evident between the two records (Figure 3.6). Although the underlying trends and amplitudes of the temperature record are not well reproduced by the PDSI reconstruction, the similarity between the records suggests that perhaps there are influences with larger spatial scales involved with this reconstruction.
These influences tend to have produced warm and wet conditions alternating with cold and dry periods in the northeastern United States.
139 3.42 Linkage to Northern Hemisphere Teleconnections
The clastic lamina thickness time-series, which is statistically representative of drought and precipitation patterns in the Northeast, is compared to teleconnection patterns in this section. Specifically, correlations will be tested between laminae and the PNA, NAO, and SOI. This indices were selected due to their influences on winter circulation patterns in the Northern Hemisphere (Jones and Mann, 2004; Quadrelli and Wallace, 2004). As with the regional climate data correlations, the tests were performed for both annual and decadally smoothed time-series (Table 3.3). The PNA pattern was defined as the average of December through May values for a given calendar year, as this interval covers the main season of climatic influence (Leathers and Palecki, 1992). NAO index was defined for the winter season (December through February) (Hurrell et al., 2003), and the SOI index was defined as the annual average value.
The PNA shows the best correlations to clastic laminae thickness, with over
30% of interannual and 49% of decadal variance of laminae thicknesses being explained by the PNA. Correlations to the NAO are not statistically significant at interannual or decadal timescales. The correlation to the SOI is not significant at interannual timescales. There is significant correlation (negative) at decadal time scales with the SOI explaining 23% of the laminae thickness variance. This correlation shows that persistent atmospheric patterns associated with El Niño
(negative SOI) favor storm tracks that deliver precipitation to New England. A t-test confirms this conclusion by demonstrating that during El Niño years (defined by SOI
140 < -1) clastic laminations in the Pettaquamscutt River are significantly thicker then they are during non El Niño years (SOI > -1) (p=0.017).
In order to put these correlations in context, it is necessary to discuss the mechanisms by which the above teleconnection patterns can influence precipitation in
New England. The three dominant air masses that enter New England are dry continental air, moist Gulf or Mexico air, and moist Atlantic coastal air masses
(Bryson and Hare, 1974; Burnett et al., 2004). Therefore, mid-tropospheric climate patterns that affect circulation and influence the movement of these air masses will be directly tied to regional climate. The most straightforward mechanism for altering the dominant air mass is by shifting from zonal to meridional tropospheric flow (Figure
3.7), thereby moving from dry continental dominance to moist Gulf of Mexico and/or moist Atlantic coastal air masses. The PNA has a strong influence on the nature of tropospheric circulation, with positive index years being defined by meridional flow, and negative index years being defined by zonal flow (Figure 3.7) (Leathers et al.,
1991; Wallace and Gutzler, 1981). In the case of ENSO, El Niño years (SOI < -1) result in above normal sea surface temperatures (SSTs) in the Eastern tropical Pacific
Ocean. This state of the Pacific Ocean is associated with positive PNA index years, and since SSTs in the Pacific partially drive the PNA (Leathers and Palecki, 1992), it is proposed that strongly negative SOI states will help initiate positive PNA conditions.
These patterns are supportive of the correlation analyses in that the PNA has a positive correlation (positive state, meridional flow, moist). SOI has a negative correlation with precipitation due to its negative correlation to the PNA. Since the
141 PNA correlation is the strongest, it has the most influence on New England hydroclimate and will be the subject of the rest of this section. It should be kept in mind, however, that on decadal time scales, ENSO patterns are also influential.
Although the temporal correlation between the PNA and clastic lamina thickness time-series is strong, it is also important to ensure that the relationship displays coherency in spatial patterns. This analysis has been done using the Climate
Explorer program (http://climexp.knmi.nl/) to analyze spatial climate series applicable in the discussion of the PNA. Figure 3.8 illustrates the correlations between the DJFMAM PNA/ annual clastic lamina thickness and 500mb heights in the Northern Hemisphere during the winter and spring seasons. Atmospheric data are from the National Centers for Environmental Prediction/National Center for
Atmospheric Research (NCEP/NCAR) Reanalysis Project, which can be found at:
http://www.cdc.noaa.gov/cdc/reanalysis/reanalysis.shtml.
The PNA index is defined by these atmospheric heights, and therefore is the most appropriate atmospheric parameter to use in this analysis (Wallace and Gutzler,
1981). This analysis yields significant correlations in the two main low-pressure regions that are found with the PNA index. In addition, the high-pressure region south of Greenland is also resolved. This pattern is the driver for mid-tropospheric circulation variability between zonal and meridional states of the PNA.
Global teleconnections as exhibited by sea surface temperature (SST) correlations are also quite similar between the PNA index and the clastic lamination thickness time-series (Figure 3.9). The large SST anomalies in the Pacific Ocean that partially drive the PNA (Leathers and Palecki, 1992) are well represented with the
142 high SST along the west coast of the Americas and in the Equatorial Pacific Ocean and low SST in the mid-North Pacific. In addition, nearly all of the major teleconnection patterns are realized with the proxy data. The notable exception is the lack of positive SST in the tropical North Atlantic and the larger negative SST in the high-latitude North Atlantic. The Arctic Oscillation affects both of these regions, and although there is not a significant annual correlation between the proxy data and this atmospheric phenomenon, SST regressions with AO indices produce negative SST anomalies in both of these regions. These anomalies can explain the shift in the proxy correlations toward negative SST anomalies in these two regions.
With confidence in both the temporal and spatial domains for the relationship between clastic lamina thickness and the PNA, it is now appropriate to use a regression to extend the PNA record back in time. The regression equation was calculated as:
PNA = (5.465 * Clastic Lamina Thickness) – 0.866.
The resulting 975-year PNA reconstruction is displayed in Figure 3.10 along with a 9- year running mean to highlight decadal variability. In addition, the period 1695-1715
A.D. has been removed due to anthropogenic influence that occurred in association with European land clearance (Chapter 1). Perhaps the most distinct portion of the reconstruction is the step increase in values at ca. 1960, which may be at least partially related to a step to lower SOI values about a decade later (Ropelewski and
Jones, 1987). This increase has been noted in instrumental PNA series (Leathers et al., 1991), but previously the context of this step was unrealized. In light of this reconstruction, it is apparent that the last 4 decades of the PNA pattern have been
143 anomalous compared with the last millennium. This observation suggests that a fundamental change in the climate system occurred in the early 1960s and that data used for resource planning, such as water budget data in the American Southwest
(Redmond and Koch, 1991), must include pre-1960 data to capture the full variability of this important climate teleconnection pattern.
In addition, our data show a steady increase in PNA value since ca. 1630 A.D.
(Figure 3.10). This observation is consistent with previous work that suggested an increasing trend over the last 300 years (Moore et al., 2002). The trend implies that low-pressure centers over the Western and Southeastern regions of the United States have generally become stronger over this time (Figure 3.8). As a result, atmospheric circulation over North America has steadily become more dominated by meridional patterns over the last four centuries. Perhaps not coincidently, the trend starts at a millennial low during the approximate timing of the North American sixteenth century mega drought (Stahle et al., 2000; Woodhouse and Overpeck, 1998) and associated dry spells in the Chesapeake Bay (Cronin et al., 2000; Willard et al.,
2003).
Prior to the sixteenth century megadrought, there was variability in the PNA index at a number of different timescales. The variability observed is within the range of pre 1960 values, with evidence of an additional period of extended low PNA values at the end of the thirteenth century. This timing is consistent with a dry spell reconstructed from New Mexican tree-tings (Grissino-Mayer, 1996). The results demonstrate that the PNA pattern has been in an anomalous state for the last 4-5 decades. If a shift to the pre-1960 behavior of the PNA is in our future, then
144 ramifications will be felt throughout the United States, mostly through changes in water budgets.
The analysis of past PNA behavior can be extended to the frequency domain by performing spectral analysis (Weedon, 2003). Due to the anomalous anthropogenically-induced period from 1695-1715 A.D., the multi-taper spectral analysis method was performed on the time periods before and after this event
(Figure 3.11). The common peaks between the two time periods that are significant above 95% confidence are at ca. 8, 3-3.3, and 2.2 years. Since there is currently a lack of proxy PNA reconstructions, the only spectral analysis to compare these data to is from instrumental records. Leathers and Palecki (1992) determined that the dominant peak in PNA variability over the period 1958-1987 was about 3.3 years.
This periodicity corresponds well with this reconstruction’s power between 3 and 3.3 years (Figure 3.11). The 2.2-year cycle was not identified in instrumental records
(Leathers and Palecki, 1992), and this period is extremely close to the Nyquist frequency (2-years). Therefore, the conservative approach is to not interpret this cycle due to sampling constraints.
The ca. 8-year cycle is intriguing because it is robust, and a new feature associated with PNA work. This cycle could have one of two origins. If it is associated directly with the PNA, then it is reasonable to conclude that instrumental spectral analysis did not observe it due to the short length of record used (29 years)
(Leathers and Palecki, 1992). On the other hand, this frequency is very similar to the dominant periodic component of the NAO (Cook, 2003), suggesting a possible connection between the PNA and NAO. Additional high-resolution reconstructions
145 of the PNA from locations affected by the PNA, but not the NAO will be needed to sort out this issue. Since precipitation in the American Southwest is also influenced by the PNA (Redmond and Koch, 1991), this may be a productive region to pursue such a reconstruction.
Wavelet analysis permits an examination of how the dominant periodic components of the PNA reconstruction have changed over time (Torrence and
Compo, 1998) (Figure 3.12). It is clear from the results that the ca. 3-year PNA cycle has not been a consistent component of the time-series for the last millennium. Its influence has in general been stronger during the early half of the millennium and again over the last few decades. These periods roughly coincide with the so-called
Medieval Warm Period and the present warm period. As the stability of the PNA over long time-scales has not previously been addressed, it is possible that this is indicative of decreased variability as the PNA persists in a negative (zonal) state.
This interpretation is speculative, however, and additional PNA reconstructions will need to be used to confirm the hypothesis. An alternate mechanism for reduced variability during the middle portion of the millennium is a southward shift of the subpolar Jet Stream. This scenario is consistent with the Little Ice Age temperature anomalies for the region and would perhaps decrease the PNA’s influence on New
England hydroclimate. Again, PNA reconstructions from places such as the
American Southwest are needed to test these hypotheses.
146 3.50 Conclusions
Clastic laminae thicknesses from the Pettaquamscutt River Estuary are significantly correlated to local hydroclimatology at interannual and decadal time- scales. The main driver for the variability observed is the Pacific/North America teleconnection pattern (PNA), demonstrating hemispheric control on climate variables, especially at decadal time scales. A reconstruction of the PNA has been generated and is the longest, most robust reconstruction to date. It is apparent that the last 4-5 decades have had anomalously high PNA values compared to the last millennium. This result is new and increases our understanding of this influential climate pattern.
147 3.60 References
Allen, M. R., and Ingram, W. J., 2002, Constraints on future changes in climate and the hydrologic cycle: Nature, v. 419, p. 224-232.
Bradbury, J. A., Dingman, S. L., and Keim, B. D., 2002, New England drought and relations with large scale atmospheric circulation patterns: Journal of the American Water Resources Association, v. 38, no. 5, p. 1287-1299.
Bradbury, J. A., Keim, B. D., and Wake, C. P., 2003, The influence of regional storm tracking and teleconnections on winter precipitation in the northeastern United States: Annals of the Association of American Geographers, v. 93, no. 3, p. 544-556.
Bryson, R. A., and Hare, F. K., 1974, Climates of North America: Volume 11 of World Survey of Climatology, Elsevier Publishing Co.
Burnett, A. W., Mullins, H. T., and Patterson, W. P., 2004, Relationship between atmospheric circulation and winter precipitation 18O in central New York State: Geophysical Research Letters, v. 31, L22209, doi:10.1029/20004GL021089.
Cook, E. R., 2003, Multi-proxy reconstructions of the North Atlantic Oscillation (NAO) Index: A critical review and a new well-verified Winter NAO Index reconstruction back to AD 1400, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 63-79.
Cronin, T., Willard, D., Karlsen, A., Ishman, S., Verardo, S., McGeehin, J., Kerhin, R., Holmes, C., Colman, S., and Zimmerman, A., 2000, Climatic variability in the eastern United States over the past millennium from Chesapeake bay sediments: Geology, v. 28, no. 1, p. 3-6.
Cubasch, U., Meehl, G. A., Boer, G. J., Stouffer, R. J., Dix, M., Noda, A., Senior, C. A., Raper, S., and Yap, K. S., 2001, Projections of future climate change, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 525-582.
148 Ernst, L. M., Miguel, L. K., and Willis, J., 1999, The Narrow River special area management plan for the watershed of the Narrow River in the towns of North Kingstown, South Kingstown, and Narragansett: Rhode Island Coastal Resources Management Council.
Folland, C. K., Karl, T. R., Christy, J. R., Clarke, R. A., Gruza, G. V., Jouzel, J., Mann, M. E., Oerlemans, J., Salinger, M. J., and Wang, S. W., 2001, Observed climate variability and change, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 99-182.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Giorgi, F., Hewitson, B., Christensen, J., Hulme, M., von Storch, H., Whetton, P., Jones, R., Mearns, L., and Fu, C., 2001, Regional climate information- Evaluation and projections, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 583-638.
Glew, J. R., Smol, J. P., and Last, W. M., 2001, Sediment core collection and extrusion, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 73-105.
Grissino-Mayer, H. D., 1996, A 2129-year reconstruction of precipitation for northwestern New Mexico, USA, in Dean, J. S., Meko, D. M., and Swetnam, T. W., eds., Tree Rings, Environment, and Humanity: Tucson, Radiocarbon, p. 191-204.
Guttman, N. B., and Quayle, R. G., 1996, A historical perspective of U.S. climate divisions: Bulletin of the American Meteorological Society, v. 77, no. 2, p. 293-303.
Hughen, K. A., Overpeck, J. T., and Anderson, R. F., 2000, Recent warming in a 500- year palaeotemperature record from varved sediments, Upper Soper Lake, Baffin Island, Canada: The Holocene, v. 10, no. 1, p. 9-19.
149 Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., 2003, An overview of the North Atlantic Oscillation, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 1-35.
Jones, P. D., Jonsson, T., and Wheeler, D., 1997, Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland: International Journal of Climatology, v. 17, p. 1433- 1450.
Jones, P. D., and Mann, M. E., 2004, Climate over past millennia: Reviews of Geophysics, v. 42, RG2002, doi:10.1029/2003RG000143.
Karl, T. R., Williams, C. N., Jr., Young, P. J., and Wendland, W. M., 1986, A model to estimate the time of observation bias associated with monthly mean maximum, minimum, and mean temperature for the United States: Journal of Climate and Applied Meteorology, v. 25, no. 1, p. 145-160.
Lamoureux, S., 2001, Varve chronology techniques, in Last, W. M., and Smols, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 247-260.
Leathers, D. J., Grundstein, A. J., and Ellis, A. W., 2000, Growing season moisture deficits across the northeastern United States: Climate Research, v. 14, p. 43- 55.
Leathers, D. J., Kluck, D., and Kroczynski, S., 1998, The severe flooding event of January 1996 across north-central Pennsylvania: Bulletin of the American Meteorological Society, v. 79, p. 785-797.
Leathers, D. J., and Palecki, M. A., 1992, The Pacific/North American teleconnection pattern and United States climate. Part II: Temporal characteristics and index specification: Journal of Climate, v. 5, p. 707-716.
Leathers, D. J., Yarnal, B., and Palecki, M. A., 1991, The Pacific/North American teleconnection pattern and United States climate. Part I: Regional temperature and precipitation associations: Journal of Climate, v. 4, p. 517-528.
Lima, A. L., Hubeny, J. B., Reddy, C. M., King, J. W., Hughen, K. A., and Eglinton, T. I., 2005, High-resolution historical records from Pettaquamscutt River basin sediments: 1. 210Pb and varve chronologies validate record of 137Cs released by the Chernobyl accident: Geochimica et Cosmochimica Acta, v. 69, no. 7, p. 1803-1812.
150 Lins, H. F., 1997, Regional streamflow regimes and hydroclimatology of the United States: Water Resources Research, v. 33, no. 7, p. 1655-1667.
Mann, M. E., and Jones, P. D., 2003, Global surface temperatures over the past two millennia: Geophysical Research Letters, v. 30, no. 15, p. 1820.
Mann, M. E., and Lees, J. M., 1996, Robust estimation of background noise and signal detection in climatic time series: Climatic Change, v. 33, p. 409-445.
Milly, P. C. D., Dunne, K. A., and Vecchia, A. V., 2005, Global pattern of trends in streamflow and water availability in a changing climate: Nature, v. 438, p. 347-350.
Mooney, H., Cropper, A., and Reid, W., 2005, Confronting the human dilemma: Nature, v. 434, p. 561-562.
Moore, G. W. K., Holdsworth, G., and Alverson, K., 2002, Climate change in the North Pacific region over the past three centuries: Nature, v. 420, p. 401-403.
Namias, J., 1966, Nature and possible causes of the northeastern United States drought during 1962-65: Monthly Weather Review, v. 94, no. 9, p. 543-554.
Nederbragt, A. J., and Thurow, J., 2005, Amplitude of ENSO cycles in the Santa Barbara Basin, off California, during the past 15000 years: Journal of Quaternary Science, v. 20, no. 5, p. 447-456.
Pike, J., and Kemp, A. E. S., 1996, Preparation and analysis techniques for studies of laminated sediments, in Kemp, A. E. S., ed., Palaeoclimatology and Palaeoceanography from laminated sediments, Geological Society Special Publication No. 116, p. 37-48.
Quadrelli, R., and Wallace, J. M., 2004, A simplified linear framework for interpreting patterns of Northern Hemisphere wintertime climate variability: Journal of Climate, v. 17, p. 3728-3744.
Redmond, K. T., and Koch, R. W., 1991, Surface climate and streamflow variability in the Western United States and their relationship to large-scale circulation indices: Water Resources Research, v. 27, no. 9, p. 2381-2399.
Reitler, L., Falk, H., Groat, C., and Coussens, C. M., 2004, From source water to drinking water: Workshop summary: Washington, D.C., National Academies Press, 126 p.
Ropelewski, C. F., and Jones, P. D., 1987, An extension of the Tahiti-Darwin Southern Oscillation Index: Monthly Weather Review, v. 115, p. 2161-2165.
151 Shiklomanov, I. A., and Rodda, J. C., 2003, World Water Resources at the Beginning of the 21st Century: Cambridge, Cambridge University Press, 435 p.
Smith, T. M., and Reynolds, R. W., 2004, Improved extended reconstruction of SST (1854-1997): Journal of Climate, v. 17, p. 2466-2477.
Stahle, D. W., Cook, E. R., Cleaveland, M. K., Therrell, M. D., Meko, D. M., Grissino-Mayer, H. D., Watson, E., and Luckman, B. H., 2000, Tree-ring data document 16th century Megadrought of North America: Eos, v. 81, no. 12, p. 121, 125.
Torrence, C., and Compo, G. P., 1998, A practical guide to wavelet analysis: Bulletin of the American Meteorological Society, v. 79, no. 1, p. 61-78.
Wallace, J. M., and Gutzler, D. S., 1981, Teleconnections in the geopotential height field during the Northern Hemisphere winter: Monthly Weather Review, v. 109, p. 784-812.
Weedon, G., 2003, Time-series analysis and cyclostratigraphy: Examining stratigraphic records of environmental cycles, Cambridge Univ. Press, 259 p.
Willard, D. A., Cronin, T. M., and Verardo, S., 2003, Late-Holocene climate and ecosystem history from Chesapeake Bay sediment cores, USA: The Holocene, v. 13, no. 2, p. 201-214.
Woodhouse, C. A., and Overpeck, J. T., 1998, 2000 years of drought variability in the central United States: Bulletin of the American Meteorological Society, v. 79, no. 12, p. 2693-2714.
Yarnal, B., and Leathers, D. J., 1988, Relationships between interdecadal and interannual climatic variations and their effect on Pennsylvania climate: Annals of the Association of American Geographers, v. 78, no. 4, p. 624-641.
152 Table 3.1. Correlation tests between Pettaquamscutt River Estuary compaction- corrected clastic laminae thicknesses and regional climate variables.
Climate Index Correlation Significance Degree of Coefficient (p) Freedom (r) (DOF) Rhode Island Precipitation (annual) 0.45 <0.0001 106 Rhode Island Precipitation (decadal) 0.84 <0.0001 23 Palmer Drought Severity Index (annual) 0.43 <0.0001 106 Palmer Drought Severity Index (decadal) 0.83 <0.0001 23 Hunt River Discharge (annual) 0.38 0.0021 61 Hunt River Discharge (decadal) 0.65 0.0087 13
153 Table 3.2. Correlation tests between decadally smoothed Pettaquamscutt River
Estuary compaction-corrected clastic laminae thicknesses and regional climate variables, before and after major residential development and storm drains in the watershed.
Climate Index Correlation Significance Degree Coefficient (p) of (r) Freedom (DOF) Rhode Island Precipitation (pre-1960) 0.57 0.0265 13 Rhode Island Precipitation (post-1960) 0.83 0.0030 8 Palmer Drought Severity Index (pre-1960) 0.76 0.0010 13 Palmer Drought Severity Index (post-1960) 0.68 0.0305 8 Hunt River Discharge (pre-1960) 0.78 0.1197 3 Hunt River Discharge (post-1960) 0.79 0.0065 8
154 Table 3.3. Correlation tests between Pettaquamscutt River Estuary compaction- corrected clastic laminae thicknesses and teleconnection patterns.
Teleconnection Index Correlation Significance Degree of Coefficient (p) Freedom (r) (DOF) PNA (DJFMAM) (annual) 0.55 <0.0001 53 PNA (DJFMAM) (decadal) 0.70 0.0077 11 NAO (DJF) (annual) -0.08 0.2912 174 NAO (DJF) (decadal) -0.27 0.0920 38 SOI (annual) -0.10 0.2450 135 SOI (decadal) -0.48 0.0054 30
155 Figure 3.1: Locus map of the Pettaquamscutt River Estuary, Rhode Island.
Watershed is shown in white along with all streams in the watershed.
156 Figure 3.2. Varve age model for the Pettaquamscutt River Lower Basin. Composite depth scale was computed using mean varve thicknesses for each year, and adjusted to a common zero depth at 1999 AD. 210Pb data from Lima et al. (Lima et al., 2005).
137Cs peaks in 1986 and 1963 (Lima et al., 2005), help to confirm the upper chronology. The most significant calibrated radiocarbon date for each sample is plotted here with the associated error bars (Table 1.2). One radiocarbon sample (*) is anomalously old due to land clearance and remobilization of older macrofossils from watershed during European settlement in the area.
157 0.3
0.25
0.2
0.15
0.1
0.05 Compaction-Corrected Clastic Lamina Thickness (cm) 0 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Year (A.D.)
Figure 3.3. Clastic lamina time-series from the Pettaquamscutt River Lower Basin along with decadally smoothed record. All thicknesses were corrected for compaction by normalizing to 78% water content.
158 0.9
0.8
0.7
0.6
0.5
0.4
0.3
Correlation Coefficient (r) 0.2
0.1
0
Fall April May June July March Winter Spring January August October Annual February Summer December September November
Figure 3.4. Correlation coefficients (r) between decadally smoothed Pettaquamscutt
River clastic laminae thicknesses and monthly and seasonal decadally smoothed
Rhode Island precipitation data. Black horizontal line indicates the 95% critical r value (p=0.05). Note the strong correlation during the winter and spring seasons.
159 1
0.5 Wet
0
-0.5
-1
-1.5 Reconstructed PDSI
-2 Dry -2.5 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Year (A.D.)
Figure 3.5. Reconstructed PDSI values for the last millennium from varved sediments in the Pettaquamscutt River Estuary. Regression based on period 1899-1997 A.D.
Gray box indicates anthropogenically-influenced period.
160 1 0 y
0.5 -0.1 0
-0.5 -0.2 -1
-1.5 Reconstructed PDSI -0.3 -2
-2.5 -0.4 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Northern Hemisphere Temperature Anomal Year (A.D.)
Figure 3.6. Reconstructed PDSI time-series from the Pettaquamscutt River plotted with Northern Hemisphere temperature anomalies from the last millennium (Mann and Jones, 2003). Although the reconstruction does not perfectly match the trend of amplitudes attributed to hemispheric temperature changes, many of the peaks and troughs are correlative suggesting that there is a positive relationship between hemispheric temperature fluctuations and New England drought cycles.
161 Figure 3.7. General character of 700-mb flow over the United States during positive
(+) and negative (-) states of the PNA. From Leathers et al. (1991).
162 Figure 3.8. Correlation coefficients (r) between (A) the PNA index and 500mb height
(NCEP/NCAR reanalysis) during the winter and spring seasons and (B)
Pettaquamscutt River clastic lamination thicknesses and 500mb height during the winter and spring seasons. The dates of analysis are common between the PNA index and clastic lamination time series (1949-2001). The heavy black line represents the
95% confidence level for correlation coefficients.
163 Figure 3.9. Correlation coefficients (r) between (A) the PNA index and sea surface temperatures (SST) (Smith and Reynolds, 2004) during the winter and spring seasons and (B) Pettaquamscutt River clastic lamination thicknesses and SST during the winter and spring seasons. The dates of analysis are common between the PNA index and clastic lamination time series (1949-1997). The heavy black line represents the
95% confidence level for correlation coefficients.
164 0.8
0.6
0.4
0.2
0
-0.2
Reconstructed PNA -0.4 Zonal Meridional -0.6
-0.8 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Age (Year A.D.)
Figure 3.10. PNA reconstruction from Pettaquamscutt River clastic laminae thicknesses along with decadally smoothed record. Higher values represent meridional atmospheric flow over the United States, while lower relative values are indicative of zonal atmospheric flow. Data from the period 1695-1715 A.D. were removed due to anthropogenic influence on sediment supply during this time.
165 Figure 3.11. Multi-taper method spectral analyses for the PNA reconstruction (A) before 1695 A.D., and (B) after 1715 A.D. Dashed lines indicate 90%, 95%, and 99% confidence levels with respect to the AR(1) red noise hypothesis (Mann and Lees,
1996). Gray peaks indicate harmonic components that are removed from the spectra.
Periodicities significant above 95% confidence are labeled.
166 Figure 3.12. Reconstructed PNA time-series with wavelet transform (Torrence and
Compo, 1998). Wavelet transform displays variations in dominant periodic components over time, with power of periodicity displayed in gray scale. The time- series was separated into pre-1695 and post-1715 components to avoid biased wavelet results associated with the large European sediment transport influence (Chapter 1).
Hashed areas represent cones of influence in which the wavelet data are biased due to edge affects.
167 Chapter 4: Paleoclimate correlations between central New York State and
Rhode Island over the last four centuries
4.10 Abstract
Regionally coherent climate patterns are important to understand in order for society to predict and plan for future climate change. In this manuscript, a 424-year high-resolution varve time-series is produced from Fayetteville Green Lake, NY in order to spatially expand the scope of the dissertation. The time-series of individual laminae are significantly correlated to local precipitation and drought cycles. These series are compared to those constructed from Pettaquamscutt River Estuary sediments in order to determine similarities and differences in climate cycles over the last 4 centuries between these two locations. Although there is strong correlation between proxy records from both locations, there is little coherence in the cyclic components of the records. The observed minimal coherence demonstrates the complexity of the climate system in that it is difficult to understand climate variability simply by understanding teleconnection indexes. The primarily atmospheric
Pacific/North American (PNA) pattern is well represented in both locations. This representation is not surprising due to the “downwind” locations of the sites as compared to the PNA, which is driven by Pacific processes and affects westerly atmospheric flow. Multidecadal components that have previously been identified as originating in the ocean via variability in thermohaline circulation are not coherent in the two locations. Most likely, the higher-frequency multidecadal variability found in central New York is driven by a different process, such as solar variability.
168 4.20 Introduction
4.21 Regional Paleoclimatology
Much progress has been made recently with regard to understanding past climate variability and potential future climate scenarios (IPCC, 2001). The results are based on instrumental data, climate models, and paleoclimate proxy data. The value of paleoclimate proxy data is high when it is realized that these are the only sources of information on climate variability over time periods longer than a century.
Proxy data come from a variety of archives (Battarbee et al., 2003). The most widely used source of regional and hemispheric climate data are tree rings due to their high resolution, climate sensitivity, and robust age models (Briffa, 2000; Mann et al.,
1998). The value of multiple proxies in hemispheric reconstructions is high due to the ability of different proxies to compensate for weaknesses in others (Mann, 2002).
Therefore, there is a priority to locate additional high-resolution proxy data that are linked to climate variables. One such archive is aquatic sediment that forms annual layers , otherwise known as limnologic varves.
This manuscript was written in order to expand our knowledge of regional climate connections in the northeast United States. Two locations were chosen with annually laminated (varved) sediment accumulation: the Pettaquamscutt River
Estuary, Rhode Island and Fayetteville Green Lake, New York (Figure 4.1). Both locations are relatively unique since modern varve formation in Northeastern lakes is rare. Sediments from the Pettaquamscutt River Estuary have been linked to local, regional, and hemispheric climate variability (Chapters 2 and 3). Previous work has
169 demonstrated that Fayetteville Green Lake is also sensitive to climate variability at a number of different time scales (Kirby et al., 2001, 2002a; Kirby et al., 2002b). For this work, varve thickness time-series have been constructed with 424 years of temporal overlap (1578-2002 A.D.). Since this is the first study to examine
Fayetteville Green Lake varves with regard to climate, an analysis was performed to understand which aspects of climate are related to laminae thicknesses. The second portion of this manuscript analyzes the varve time-series from both locations in order to determine similar and different climate processes that have occurred at the two locations. The result will be a better understanding of the spatial patterns of climate variability in the northeastern United States over the last four centuries.
4.22 Fayetteville Green Lake Limnology and Sediment Sources
Fayetteville Green Lake (hereafter Green Lake) has been extensively studied over the last few decades due to its unique physical and chemical conditions
(Brunskill and Ludlam, 1969; Hilfinger and Mullins, 1997). The lake is meromictic, meaning that there is permanently stratified water that only partially turns over each year. This permanent stratification is due partly to the small surface area to depth ratio (Deevey et al., 1963), but also to the influx of saline groundwater at depth
(Takahashi et al., 1968). Groundwater is saline due to the dissolution of gypsum from the underlying Vernon Shale (Muller, 1967), which increases the concentrations of Ca and SO4 in bottom waters. The associated density contrast between upper and lower waters ensures the stability of this system.
170 In addition to affecting the water column, the influx of calcium ions leads to episodes of supersaturation and precipitation of calcite in the water column during whiting events (Brunskill, 1969). Calcite precipitation is related to photosynthetic activity of the cyanobacterium Synechococcus in the mixolimnion of the lake because calcium ions that concentrate on the cell surfaces serve as nucleation sites for the precipitation of calcite (Thompson et al., 1990; Thompson et al., 1997). Because there is a direct relationship between these cyanobacteria and calcite precipitation, and because biologic activity is greatest in the summer months, there is an annual cycle to the whiting events in Green Lake. This annual sedimentation cycle coupled with a lack of bioturbation has resulted in the accumulation of annually-laminated
(varved) sediments (Ludlam, 1969, 1981, 1984; Wahlen and Lewis, 1980). The annual nature of the laminae has been confirmed both through seasonal observation
(Ludlam, 1969), and through radiometric dating of sediment cores (Wahlen and
Lewis, 1980). Each varve consists of two laminae, a light calcite layer and a darker organic layer, which is deposited during non-whiting months. In addition to the varves, there are also massive and graded laminae and beds which have been identified as turbidites throughout the lake (Ludlam, 1969). These turbidites consist primarily of reworked marl sediment and remains of an aquatic moss that is ubiquitous in the modern shallows of the lake.
4.30 Methods
Two freeze cores (Glew et al., 2001) were obtained from the deep section
(depth = 54.6 m) of Green Lake, New York, in September 2004 (Figure 4.2). The
171 cores measured 52cm and 57cm long, had excellent sediment/water interfaces, and were both well laminated. Both cores were impregnated with Spurr resin (Spurr,
1969) and thin-sectioned (Lamoureux, 2001; Pike and Kemp, 1996) for lamination analysis, as described in the appendix. The thin sections from each core were scanned under cross-polarized light with a flatbed scanner to produce TIFF images.
These images were analyzed in Photoshop® by marking all lamina boundaries with the path tool. A number of massive turbidites were identified in the thin sections.
These were interpreted as instantaneous events, and were therefore removed from the record. Since the coring location was removed from subaqueous slopes and since erosional sedimentary structures were absent, an assumption was made that sediment overlying an individual turbidite was conformable with sediment below the turbidite, and therefore that the laminated record is continuous. The paths were exported and run through an algorithm designed to calculate widths in centimeters from digital pixel widths (Francus et al., 2002). Counts from the two cores were crosschecked with each other to produce a master varve chronology. Recounts from individual sections yield counting errors of less than 1% throughout the record.
Lamina thickness data were extracted from the master varve chronology in order to use as paleoclimate proxy data. The thicknesses were corrected for differential compaction by normalizing to water content (Nederbragt and Thurow,
2005). For Green Lake, water content was measured for 1 cubic centimeter samples from laminated portions of one freeze core. Percent water decreased exponentially from 87% at the top of the section to 53% at the bottom of the core. An exponential equation (r2=0.6664) was fit to the data such that:
172 y = (0.0000000000003 * e 0.016x) + 52.8%, where y is equal to the percent water and x is equal to the year of deposition. All lamination and varve thicknesses were normalized to 53% water content using this relationship in order to remove the exponential trend associated with compaction.
The compaction-corrected Fayetteville Green Lake laminae time-series were compared to instrumental records of climate variables and teleconnection indexes using Pearson correlation tests. All analyses were performed on monthly, seasonal, annual, and decadal time scales. The decadal correlations were run after applying a
9-year running average smoothing function to all time-series studied.
Significance (p) was determined for all correlation coefficients using
GraphPad© software. The requisite degree of freedom (DOF) was calculated as the population (N) minus 1 for raw time-series. For the smoothed series, it was necessary to calculate the DOF as:
DOF = N * (flow / fNyquist), where flow is the frequency of the low-frequency cutoff (1/9yr in this case), and fNyquist is the Nyquist frequency (0.5yr in this case). Correlations were determined significant if they were calculated as above the 95% confidence level (p<0.05).
4.40 Results and Discussion
4.41 Green Lake Age Model
Thin sections of Green Lake sediments reveal alternating light (carbonate) and dark (organic) laminae with intermittent turbidites (Figure 4.3). Using previous work as supporting evidence (Ludlam, 1969, 1981, 1984; Wahlen and Lewis, 1980), it was
173 assumed here that each light/dark couplet represents one year of deposition, and therefore the sediment is varved. With this premise, age models were constructed for each of two freeze cores after removing turbidites (Figures 4.4, 4.5). The removal of these layers is justified in that they are formed instantaneously, and do not represent a portion of a varve. Due to the distal location of the cores and the lack of observed erosional features, it was assumed that sediment was conformable above and below these layers, however this assumption is difficult to test. Nevertheless, both age models fit the general trend of increased sedimentation rate for the sediments deposited more recently then ca. 1800, which has been observed in previous studies using radiometric age controls (Hilfinger et al., 2001). Therefore, the evidence supports the assumptions that the sediments are varved and that turbidites in this distal location are non-erosional. The average sedimentation rate of the varved sediments is 0.6mm/year over the record. The rates are 1.0 and 0.3mm/year before and after 1800 A.D., respectively.
4.42 Green Lake Climate Correlations
Time-series of lamina thickness and varve thickness were constructed from the varve chronology data in order to use as paleoclimate proxies. These thicknesses are normalized in order to take into account compaction of sediment over time, and show variability at a number of different time scales (Figure 4.6). All three time- series display step increases at ca. 1800 A.D. This change has been identified previously in a number of other proxies from the lake (Hilfinger et al., 2001), and is associated with European settlement and deforestation of the region at about this
174 time. In addition, the most recent years of the time series exhibit a spike that is a function of imperfect normalization of the sediment in the uppermost centimeters of sediment associated with the flocculent layer, and can therefore be discounted from a climate standpoint. Despite these non-climate influences, there is much variability in the time-series that presumably can be tied to climate variability.
The hypothesized climate linkages of the laminae time-series can be investigated by performing statistical Pearson correlation tests on the proxy series and overlapping series of climate variables. The tests have been performed on both annual and decadally smoothed time-scales in order to diminish stochastic effects inherent in climate time-series. The climate variables used in the tests were precipitation, drought (as quantified with the Palmer Drought Severity Index (PDSI)), and surface air temperature for the state of New York, available from the National
Climate Data Center (Table 4.1). State climate variables were used in order to examine regional patterns of climate variability and to allow for more regional interpretations of the data. Carbonate laminae correlate significantly and positively with both precipitation and PDSI at both annual and decadal time scales. Examination of monthly correlations reveals that the strongest linkage with precipitation occurs during the summer months (Figure 4.7). This corresponds with the timing of
Synechococcus blooms, which trigger the carbonate precipitation in the water column
(Thompson et al., 1990; Thompson et al., 1997). Monthly PDSI values show significant correlations to carbonate thicknesses in the summer, fall, and winter
(Figure 4.8). The PDSI correlations occur later in the year than the precipitation correlations would suggest, however there is undoubtedly a lag between low (high)
175 precipitation anomalies and low (high) PDSI values due to the inertia associated with hydrologic systems.
The correlations between hydrologic conditions and carbonate deposition are counterintuitive to many paleolimnological reconstructions, in that lower rainfall normally leads to reduced lake volume, increased calcium and carbonate concentrations in the water column, and increased carbonate precipitation due to supersaturation. The cause for the positive relationship at Green Lake must be caused by something other than changes in lake volume. A probable explanation is that since the lake is groundwater fed, calcium ions are being introduced to the lake via groundwater flow through gypsum-rich (CaSO4·2H2O) bedrock. If a region is experiencing negative precipitation anomalies and is in a drought state (negative
PDSI), then the groundwater recharge will decrease. As this occurs, groundwater supply to the lake diminishes, and the calcium supply declines in tandem. Carbonate precipitation will then be limited by the availability of calcium, and total precipitation
(and flux to sediment) is reduced. Recent work has demonstrated this phenomenon in other groundwater dominated lakes (Shapley et al., 2005), and this is the simplest explanation for the observed positive correlations between carbonate accumulation and regional moisture at Green Lake.
In addition to precipitation, surface air temperatures are significantly correlated to carbonate accumulation on decadal time-scales (Table 4.1). This correlation is more straightforward and can be explained in one (or both) of two ways. First, since higher temperatures lead to reduced solubility of carbonate, it is reasonable to attribute this correlation to a physical cause. In the case of Green Lake,
176 however, carbonate precipitation is driven by biologic processes (Thompson et al.,
1997). Therefore, it is also possible that increased temperatures would increase metabolic activity, and therefore increase the efficiency of carbonate precipitation.
Either (or both) mechanisms are plausible, and for the case of this climate analysis, it is sufficient to note that at decadal scales, the carbonate laminae thicknesses are a proxy for regional air temperature.
The organic laminae do not correlate significantly to regional climate variables (Table 4.1), however they should to some extent reflect biological activity in the water column and watershed unrelated to Synechococcus blooms. Monthly correlation analyses shed light on the connection (Figure 4.9). Although the correlations are modest, significant positive correlations are found during winter/early spring months for both annual and decadally smoothed PDSI values and organic laminae thicknesses. The winter correlation is found during a different season than the carbonate correlations (Figures 4.7, 4.8) and suggests a different process. Winter precipitation can influence the influx of organic matter in one of two ways. First, it is possible that increased nutrients that are washed into the lake might increase productivity by phytoplankton. More likely, however, is that more wet conditions during the winter will wash organic matter from the watershed into the lake and deposit this material in the months before Synechococcus blooms, thereby depositing a distinct lamina each year. Both of these processes can exhibit a significant lag
(months) if the precipitation falls as snow because the limnologic effect will not be realized until the spring thaw.
177 The varve thickness correlations with climate variables are clearly driven by correlations to the individual components of the varves (i.e. carbonate and organic laminae) (Table 4.1). This relationship is expected, and helps to support the correlations that are observed. In addition, it provides support for varve studies to examine laminae individually to better understand the processes involved. Studies that merely analyze the total varve thickness cannot fully understand the processes involved in a system like this, and correlations are bound to be lower. The only exception to this generalization is in an environment where both laminae are deposited by the same process, but are temporally separated. An example of such an environment is an arctic lake, where total varve thickness is related to runoff during the melt season, and the fine layer is merely distinct due to the longer settling time of finer grains.
Finally, the time-series of laminae thicknesses can be examined in light of the climate correlations that they contain (Figure 4.10). Both series are indicative of moisture budgets over the last 424 years, but are based on different seasons.
Therefore, it is possible to reconstruct both summer/fall precipitation variability
(carbonate laminae) and winter/spring precipitation variability (organic laminae)
(Figure 4.10). Both series display variability at a number of different time scales.
Spectral analysis was performed to quantitatively analyze the cyclic components preserved by the carbonate and organic laminae (Figures 4.11, 4.12). The carbonate series (summer/fall moisture) exhibits a broad range of periodic components ranging from multidecadal to interannual (Figure 4.11). The interannual periodic components have periodicities commonly associated with the El Niño/Southern Oscillation and
178 the Pacific/North American climate patterns (Ropelewski and Jones, 1987; Wallace and Gutzler, 1981), suggesting Pacific influence on the climate of central New York.
In addition, multidecadal periodicities are observed, most notably the ca. 25-year cycle as this has been identified from stable isotope work from Green Lake (Kirby et al., 2001). This cycle likely is rooted in solar variability and the associated communication to the climate system through global surface ocean temperatures
(White et al., 1997). The organic laminae have similar cycles with dominance in the interannual and multidecadal ranges (Figure 4.12). Therefore, many of the climate periodic influences on Green Lake appear to be similar during the summer/fall months (carbonate) as they are during the winter/spring months (organic).
4.43 Green Lake Teleconnection Correlations
The above discussion has addressed the linkages between local climate variables and laminae thicknesses, however it did not investigate the drivers of the observed climate variability. The common drivers to climate variability are called teleconnection patterns, and describe atmospheric circulation patterns that can affect large regions, hemispheres, or even the planet. Recent work has demonstrated that the two main influences on Northern Hemisphere climate are the Pacific/North
American Pattern (PNA) and the North Atlantic Oscillation (NAO) (Quadrelli and
Wallace, 2004). In addition, the global influence of the El Niño/Southern Oscillation pattern (ENSO) and the associated Southern Oscillation Index (SOI) makes this an important contributor to regional climate variability (Jones and Mann, 2004). Since time-series of SOI extend further into the past and since SOI is intimately related to
179 El Niño, the SOI index will be used to represent ENSO variability. In the northeastern United States, the resulting patterns of teleconnection influence during the cold seasons can be observed in the shape and longitudinal position of the polar front jet stream. The Trough Intensity Index (TII) and Trough Axis Index (TAI) are regional indexes that define the amplitude and position of the trough that sets up over this region (Bradbury et al., 2002b). Therefore, these are also appropriate indexes to discuss in this section.
Correlations between lamina and varve thicknesses and teleconnection patterns are summarized in Table 4.2. It is first interesting to note that there are no significant correlations to teleconnection patterns at annual time scales. This is discouraging in light of the absence of possible predictive power that global or hemispheric climate indices can provide in some regions. In other words, no such interannual predictive power appears feasible through this method using climate series from Green Lake.
Significant correlations are found for a number of teleconnection patterns when series are decadally smoothed. This procedure removes much of the stochastic variability in climate time-series as well as non-climate related lacustrine sedimentary processes, and focuses on the large-scale climatic patterns. The strongest correlation is between the carbonate laminae thicknesses and the PNA. The positive value indicates that more carbonate is deposited during times of meridional atmospheric flow as opposed to zonal flow. Since the carbonate thicknesses are directly correlated to regional moisture (Table 4.1), this implies that during positive phases of the PNA
(meridional flow) more precipitation reaches central New York State. In this
180 scenario, storms originating from the Gulf of Mexico or Atlantic Coast provide more moisture than dry continental air masses, similar to what was observed in Rhode
Island (Chapter 3). The influence of lake-effect precipitation associated with zonal flow, however, may cast doubt on this mechanism as it implies high precipitation during zonal flow periods. Recent work has confirmed the dominance of lake-effect precipitation in nearby Hamilton, NY by concluding that 42% of precipitating weather events over the period 1999-2003 (November to March) were associated with the lake-effect pattern (Burnett et al., 2004). Warm sector and coastal systems, which can be associated with meridional flow patterns (and a positive PNA) comprise 25% and 15% of precipitating weather events, respectively. Perhaps the most important factor of moisture equations, however, is the amount of precipitation, and not the number of events. Supporting this statement, average liquid precipitation amounts for lake-effect, warm sector, and coastal systems are 0.16 cm, 0.56 cm, and 0.86 cm, respectively (Burnett et al., 2004). Therefore, it is reasonable to conclude that positive PNA periods (dominant meridional flow) lead to above average precipitation, moisture budgets, and carbonate precipitation in Green Lake.
In addition to the PNA index, a significant positive relationship is observed between carbonate laminae and the NAO at decadal time scales (Table 4.2). This result does not agree with the PNA analysis if the linkage between the NAO pattern and Green Lake sediments is precipitation variability because a positive NAO state leads to zonal atmospheric flow and reduced precipitation due to the influence of dry continental air masses. It should be noted that the correlation with the NAO is weak, and likely not related to precipitation. In the northeastern United States positive
181 temperature anomalies are observed during positive NAO years. Since temperature is positively correlated to carbonate accumulation in Green Lake at decadal time scales
(Table 4.1), it is likely that the NAO signal is communicated to Green Lake through temperature solubility control on carbonate saturation. This scenario explains the seemingly incompatible associations between carbonate laminae and the PNA and
NAO.
The SOI has a significant negative relationship with carbonate laminae deposited in Green Lake (Table 4.2). The relationship indicates that during El Niño years central New York is wetter and more carbonate is precipitated in Green Lake.
El Niño years cause positive sea surface temperature (SST) anomalies in the Eastern
Equatorial Pacific Ocean. Since Pacific SST is a contributor in driving the PNA
(Leathers and Palecki, 1992), the most likely connection between ENSO and moisture budgets in New York is through PNA forcing.
Finally, significant correlations are found between the TAI (Bradbury et al.,
2002b) and carbonate laminae (negative) and between the TII (Bradbury et al.,
2002b) and organic laminae (positive). Despite the high correlations and significances (p<0.01), a note of caution is necessary with these correlations since the smoothing function reduced the degrees of freedom to 9. That being said, the linkages are intriguing. First, the negative correlation to the TAI indicates that during years when the subpolar trough is displaced to the west, higher amounts of carbonate are deposited. This is consistent with the PNA correlation as far as the air masses are concerned because the further west the trough, the more influence New York will feel from warm-sector and coastal air masses. The intensity of the trough (TII) only
182 affects the organic laminae significantly. This may be linked to the seasonality of the different laminae since the TII is defined by winter season data and organic laminae are most strongly correlated to the winter months (Figure 4.9).
This discussion on the influence of different teleconnection indexes on the moisture state of central New York state brings to the forefront the complications involved. For instance, some teleconnections appear to contradict others as in the cases of PNA and NAO, emphasizing the complex nature of associated weather patterns in this region. Perhaps this is the reason that correlations are only significant when the series are smoothed. At interannual timescales, atmospheric dynamics are more complex due to the stochastic nature of weather events. At decadal time scales in which high-frequency dynamics are filtered out, the underlying climate modulators are exposed.
4.44 Climate Correlations Between Central New York and Southern New
England
A current priority in Holocene climate research is to establish high-resolution records of climate variability with robust age control that can be correlated spatially in order to better understand regional connections (Jones and Mann, 2004). In addition, since most work has focused on high-resolution temperature variability over recent millennia, there is a relative gap in our knowledge of precipitation variability, especially at regional scales. This dissertation has produced high-resolution proxy records of climate variability over the last four centuries from two locations, Green
Lake, NY and the Pettaquamscutt River Estuary, RI. Not only are the records
183 annually resolved, the age models are robust due to the nature of annually laminated sediments. These proxy records exhibit primarily moisture variability over a transect from central New York to southern New England. In addition, both locations have yielded primary productivity information that gives information on ecosystem responses to climate change, and some climate variables that we do not have instrumental data for such as light availability. These proxy time-series will be compared and contrasted in this section in order to better understand the climate linkages that exist between these locations. The length of record is long enough to span the Little Ice Age and subsequent Present Warm Period, and as such, can shed light on perturbations that may have occurred in the region.
The proxies examined here are primarily indicative of moisture variability, however biologic productivity is also represented and can be forced by other factors such as temperature and light availability (Table 4.3). The thicknesses of carbonate and organic laminae from Green Lake are annually resolved proxy records of moisture availability in the region. They are biased toward the summer/fall and winter/spring seasons, respectively. In addition, both are intimately connected to primary productivity in the lake, with the carbonate layer specifically linked to cyanobacterial productivity. From the Pettaquamscutt River Estuary, the clastic and biogenic laminae represent annually resolved records of moisture availability and aquatic productivity/ temperature/ precipitation variability, respectively. Finally, the fossil pigment Bacteriochlorophyll e is indicative of estuarine productivity and temperature for the period prior to cultural eutrophication starting in 1960 (Chapter
2).
184 Tables 4.4 and 4.5 list the correlation coefficients between Green Lake proxies and Pettaquamscutt River proxies at annual and decadal time scales, respectively. All correlations are significant and positive, demonstrating the regional similarities of climate variability. Interannual correlations are strongest between the organic laminae in Green Lake and the biogenic laminae in the Pettaquamscutt River
(r=0.73). Both of these proxies are indicative of productivity in their respective watersheds, suggesting linkages to climate variables that would limit primary producers. Of the instrumental data sets used in this study, decadal precipitation correlations are the common linkage for both of these records. The correlation between productivity is stronger than would be suggested by correlations to precipitation, however. This inconsistency is most likely due to the influence of other variables not available for this study, such as light availability, ice-out dates, etc. In addition, the Bacteriochlorophyll e proxy data have a lower correlation to Green Lake organic laminae, although the correlation is still highly significant (Table 4.4).
The proxies that are most strongly related to precipitation and moisture availability in the region (Green Lake carbonate lamina thickness and Pettaquamscutt
River clastic lamina thickness) exhibit weak, yet highly significant correlations at interannual scales. The correlation suggests that although the locations have similar patterns of hydrologic variability, local effects such as lake-effect patterns in New
York and marine influences in Rhode Island impact these records. At decadal scales, however, much of the local effects are removed and the correlation increases to 0.66, which is quite robust (Table 4.5). Therefore, low-frequency variability in moisture budgets along this transect act in tandem.
185 In order to understand the cyclic components of the series and how these correlate to each other, cross-spectral analyses were performed between locations.
Specifically, a comparison was made between Green Lake organic laminae and
Pettaquamscutt River biogenic laminae (Figure 4.13) and between Green Lake carbonate laminae and Pettaquamscutt River clastic laminae (Figure 4.14). The productivity proxies show three peaks in spectral coherence, at 13.7, 7.7, and 2.4 years. Although these coherences are significant, only the 13.7 and 2.4 year cyclicities are significantly different from a red noise spectral background (Mann and
Lees, 1996). These cyclicities are interesting, but are not easily assigned to known climate patterns in the literature and may be attributed to regional phenomena affecting temperature, precipitation, and light availability (cloudiness) throughout the region.
Cross-spectral analysis on the two moisture proxies also shows coherency at a number of frequencies, however the only one that has corresponding significant periodicities in the spectral analyses is a peak at 3.2 years (Figure 4.14). This periodicity has been identified in PNA instrumental records (Leathers and Palecki,
1992). Since the PNA is significantly correlated to both of these proxies, this finding is not surprising. However, it lends more credence to the findings of Chapter 3 regarding regional significance of the PNA pattern in the northeastern United States.
It is interesting to note the lack of coherence among proxies at multidecadal timescales. Previous work has concluded that such periodicities are ocean-driven
(Hubeny et al., 2006; Kirby et al., 2001), however the coherence between locations is non-existent. For instance, the ca. 25-year periodicity identified by Kirby et al.
186 (2001) is found preserved in Green Lake carbonate laminae in this work (Figure
4.11), however a significant periodicity from the Pettaquamscutt River is lacking. On the other hand, periodicities of 95 and 39 years found in Pettaquamscutt River productivity series have not been found significant in Green Lake. Multidecadal periodicities have been identified by modeling studies and have been broadly defined as the Atlantic Multidecadal Oscillation (AMO) (Delworth and Mann, 2000; Knight et al., 2005), however it is unlikely that one phenomenon would have such a broad variety of periodic components over such a short time scale. Since previous work has discussed periodicities for the AMO closer to those found in the Pettaquamscutt
River, and since this location is proximal to the ocean, it is likely that these are the dominant AMO cycles. In addition, the 25-year Green Lake periodicity is of higher frequency than typically attributed to the AMO. Therefore, it is concluded that the
Green Lake multidecadal signals are more likely not driven by the AMO. A reasonable driver for this cycle may be solar variability as multidecadal cycles have been identified in solar irradiance proxy data (Lean and Rind, 1998). Although Kirby et al. (2001) did not find a significant correlation to solar irradiance, evidence suggests that solar variability is normally communicated to the climate system in complicated ways (Tinsley, 2000). Recent work has demonstrated that variability in solar irradiance forces fluctuating upper ocean temperatures with periods of 18-25 years (White et al., 1997). This mechanism is a strong candidate for the 25-year driver in the Green Lake time-series because the ocean effects have been identified worldwide, and because these sea surface temperature anomalies are connected to
187 teleconnection indexes that have been linked to Green Lake proxy data, such as the
PNA and ENSO.
4.50 Conclusions
In-depth varve analyses of two freeze cores from Green Lake, NY have produced time-series of carbonate and organic laminae over the past 424 years.
Statistical tests confirm climate linkages where the carbonate accumulation is positively correlated to moisture cycles biased toward the summer and fall seasons, whereas the organic layer has a weaker correlation to moisture cycles biased toward the winter and spring months. At decadal time scales, these correlations are linked to hemispheric teleconnection patterns, such as the PNA and ENSO. The linkage appears to be through regional storm tracking associated with the shape and location of the subpolar jet stream. Comparisons between these series and those from the
Pettaquamscutt River Estuary reveal few coherent cycles over the period of overlap.
This may be due to local effects, or the difference in oceanic proximity. A coherent
PNA periodicity (3.2 years) is found in moisture proxies from both locations, confirming the importance of this hemispheric climate pattern on the moisture availability of the northeastern United States.
188 4.60 References
Battarbee, R., Birks, J., Oldfield, F., and Mackay, A., 2003, Global Change in the Holocene, Hodden Arnold Publishers, p. 544.
Bradbury, J. A., Keim, B. D., and Wake, C. P., 2002, U.S. East Coast Trough Indexes at 500 hPa and New England winter climate variability: Journal of Climate, v. 15, p. 3509-3517.
Briffa, K. R., 2000, Annual climate variability in the Holocene: interpreting the message of ancient trees: Quaternary Science Reviews, v. 19, p. 87-105.
Brunskill, G. J., 1969, Fayetteville Green Lake, New York. II. Precipitation and sedimentation of calcite in a meromictic lake with laminated sediments: Limnology and Oceanography, v. 14, no. 6, p. 830-847.
Brunskill, G. J., and Ludlam, S. D., 1969, Fayetteville Green Lake, New York. I. Physical and chemical limnology: Limnology and Oceanography, v. 14, no. 6, p. 817-829.
Burnett, A. W., Mullins, H. T., and Patterson, W. P., 2004, Relationship between atmospheric circulation and winter precipitation 18O in central New York State: Geophysical Research Letters, v. 31, L22209, doi:10.1029/20004GL021089.
Deevey, E. S., Nakai, N., and Stuiver, M., 1963, Fractionation of sulfur and carbon isotopes in a meromictic lake: Science, v. 139, p. 407-408.
Delworth, T. L., and Mann, M. E., 2000, Observed and simulated multidecadal variability in the Northern Hemisphere: Climate Dynamics, v. 16, p. 661-676.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Glew, J. R., Smol, J. P., and Last, W. M., 2001, Sediment core collection and extrusion, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 73-105.
Hilfinger, M. F., IV, and Mullins, H. T., 1997, Geology, Limnology, and Paleoclimatology of Green Lakes State Park, New York, in Raynes, T. W., ed., 69th Annual New York State Geological Association Field Trip Guide: Clinton, NY, Hamilton College, p. 127-159.
189 Hilfinger, M. F., IV, Mullins, H. T., Burnett, A. W., and Kirby, M. E., 2001, A 2500 year sediment record from Fayetteville Green Lake, New York: Evidence for anthropogenic impacts and historic isotope shift: Journal of Paleolimnology, v. 26, p. 293-305.
Hubeny, J. B., King, J. W., and Santos, A., 2006, Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments: Geology, v. 34, no. 7, p. 569-572.
IPCC, 2001, Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, 881 p.
Jones, P. D., and Mann, M. E., 2004, Climate over past millennia: Reviews of Geophysics, v. 42, RG2002, doi:10.1029/2003RG000143.
Kirby, M. E., Mullins, H. T., Patterson, W. P., and Burnett, A. W., 2001, Lacustrine isotopic evidence for multidecadal natural climate variability related to the circumpolar vortex over the northeast United States during the past millennium: Geology, v. 29, no. 9, p. 807-810.
-, 2002a, Late glacial-Holocene atmospheric circulation and precipitation in the northeast United States inferred from modern calibrated stable oxygen and carbon isotopes: Geological Society of America Bulletin, v. 114, no. 10, p. 1326-1340.
Kirby, M. E., Patterson, W. P., Mullins, H. T., and Burnett, A. W., 2002b, Post- Younger Dryas climate interval linked to circumpolar vortex variability: isotopic evidence from Fayetteville Green Lake, New York: Climate Dynamics, v. 19, p. 321-330.
Knight, J. R., Allan, R. J., Folland, C. K., Vellinga, M., and Mann, M. E., 2005, A signature of persistent natural thermohaline circulation cycles in observed climate: Geophysical Research Letters, v. 32, L20708, doi:10.1029/2005GL024233.
Lamoureux, S., 2001, Varve chronology techniques, in Last, W. M., and Smols, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 247-260.
Lean, J., and Rind, D., 1998, Climate forcing by changing solar radiation: Journal of Climate, v. 11, p. 3069-3094.
190 Leathers, D. J., and Palecki, M. A., 1992, The Pacific/North American teleconnection pattern and United States climate. Part II: Temporal characteristics and index specification: Journal of Climate, v. 5, p. 707-716.
Ludlam, S. D., 1969, Fayetteville Green Lake, New York. III. The Laminated Sediments: Limnology and Oceanography, v. 14, no. 6, p. 848-857.
-, 1981, Sedimentation rates in Fayetteville Green Lake, New York, U.S.A.: Sedimentology, v. 28, p. 85-96.
-, 1984, Fayetteville Green Lake, New York, U.S.A. VII. Varve Chronology and Sediment Focusing: Chemical Geology, v. 44, p. 85-100.
Mann, M. E., 2002, The value of multiple proxies: Science, v. 297, p. 1481-1482.
Mann, M. E., Bradley, R. S., and Hughes, M. K., 1998, Global-scale temperature patterns and climate forcing over the past six centuries: Nature, v. 392, p. 779- 787.
Mann, M. E., and Lees, J. M., 1996, Robust estimation of background noise and signal detection in climatic time series: Climatic Change, v. 33, p. 409-445.
Muller, E. H., 1967, Geologic setting of Green and Round Lakes near Fayetteville, New York, in Jackson, D. F., ed., Some Aspects of Meromixis: Syracuse, NY, Department of Civil Engineering, Syracuse University, p. 96-121.
Nederbragt, A. J., and Thurow, J., 2005, Amplitude of ENSO cycles in the Santa Barbara Basin, off California, during the past 15000 years: Journal of Quaternary Science, v. 20, no. 5, p. 447-456.
Pike, J., and Kemp, A. E. S., 1996, Preparation and analysis techniques for studies of laminated sediments, in Kemp, A. E. S., ed., Palaeoclimatology and Palaeoceanography from laminated sediments, Geological Society Special Publication No. 116, p. 37-48.
Quadrelli, R., and Wallace, J. M., 2004, A simplified linear framework for interpreting patterns of Northern Hemisphere wintertime climate variability: Journal of Climate, v. 17, p. 3728-3744.
Ropelewski, C. F., and Jones, P. D., 1987, An extension of the Tahiti-Darwin Southern Oscillation Index: Monthly Weather Review, v. 115, p. 2161-2165.
Shapley, M. D., Ito, E., and Donovan, J. J., 2005, Authigenic calcium carbonate flux in groundwater-controlled lakes: Implications for lacustrine paleoclimate records: Geochimica et Cosmochimica Acta, v. 69, no. 10, p. 2517-2533.
191 Spurr, A. R., 1969, A low-viscosity epoxy resin embedding medium for electron microscopy: Journal of Ultrastructure Research, v. 26, p. 31-43.
Takahashi, T., Broecker, W., Li, Y. H., and Thurber, D., 1968, Chemical and isotopic balances for a meromictic lake: Limnology and Oceanography, v. 13, p. 272- 292.
Thompson, J. B., Ferris, F. G., and Smith, D. A., 1990, Geomicrobiology and sedimentology of the mixolimnion and chemocline in Fayetteville Green Lake, New York: Palaios, v. 5, p. 52-75.
Thompson, J. B., Schultze-Lam, S., Beveridge, T. J., and Des Marais, D. J., 1997, Whiting events: Biogenic origin due to the photosynthetic activity of cyanobacterial picoplankton: Limnology and Oceanography, v. 42, p. 133- 141.
Tinsley, B. A., 2000, Influence of solar wind on the global electric circuit, and inferred effects on cloud microphysics, temperature, and dynamics in the troposphere: Space Science Reviews, v. 94, no. 1-2, p. 231-258.
Wahlen, M., and Lewis, D. M., 1980, Green Lake: Dating of laminated sediments: EOS, v. 61, no. 46, p. 964.
Wallace, J. M., and Gutzler, D. S., 1981, Teleconnections in the geopotential height field during the Northern Hemisphere winter: Monthly Weather Review, v. 109, p. 784-812.
White, W. B., Lean, J., Cayan, D. R., and Dettinger, M. D., 1997, Response of global upper ocean temperature to changing solar irradiance: Journal of Geophysical Research, v. 102, no. C2, p. 3255-3266.
192 Table 4.1. Correlation coefficients (r) between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and New York State climate variables. All climate data from the National Climate Data Center (NCDC). Significance values (p) are shown with superscripts.
Climate Variable Varve Carbonate Organic (New York State Average) Laminae Laminae Precipitation (annual) 0.28† 0.37* 0.08 Precipitation (decadal) 0.42‡ 0.53† 0.12 Palmer Drought Severity Index (annual) 0.30† 0.37* 0.10 Palmer Drought Severity Index (decadal) 0.54† 0.58† 0.27 Surface Air Temperature (annual) 0.07 0.11 0.00 Surface Air Temperature (decadal) 0.39 0.40‡ 0.11
* p<0.0001
† p<0.01
‡ p<0.05
193 Table 4.2. Correlation coefficients (r) between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and teleconnection indexes. Significance values (p) are shown with superscripts.
Teleconnection Index Varve Carbonate Organic Laminae Laminae PNA (annual) 0.22 0.23 0.10 PNA (decadal) 0.80† 0.80† 0.52 NAO (DJFM) (annual) 0.09 0.10 0.02 NAO (DJFM) (decadal) 0.50† 0.38† 0.24 SOI (annual) -0.17 -0.20 -0.07 SOI (decadal) -0.49† -0.47† -0.04 Trough Axis Index (DJF) (annual) 0.17 0.18 0.07 Trough Axis Index (DJF) (decadal) -0.75† -0.74† -0.15 Trough Intensity Index (DJF) (annual) 0.01 0.00 -0.03 Trough Intensity Index (DJF) (decadal) 0.57 0.11 0.79†
† p<0.01
194 Table 4.3. Proxy records used in Section 4.44 and the climatic/ecological variable that they represent.
Proxy Primary Representation Secondary Representation Green Lake Annual Precipitation/ Biased to Summer/Fall Carbonate Laminae Hydrologic State Seasons Green Lake Aquatic Productivity Decadal Trends of Organic Laminae Winter/Spring Moisture Pettaquamscutt River Annual Precipitation/ Clastic Laminae Hydrologic State Pettaquamscutt River Aquatic Productivity Air Temperature and Biogenic Laminae Decadal Moisture Pettaquamscutt River Aquatic Productivity Water Column Temperature Bacteriochlorophyll e
195 Table 4.4. Correlation coefficients (r) between Fayetteville Green Lake compaction- corrected varve and laminae thicknesses and Pettaquamscutt River Estuary varve and laminae thicknesses, and Bacteriochlorophyll e mass accumulation rates.
Significance values (p) are shown with superscripts.
Green Lake Green Lake Green Lake Varve Carbonate Laminae Organic Laminae Pettaquamscutt River 0.55* 0.49* 0.52* Varve Pettaquamscutt River 0.28* 0.27* 0.26* Clastic Laminae Pettaquamscutt River 0.68* 0.51* 0.73* Organic Laminae Pettaquamscutt River 0.42* 0.36* 0.40* Bacteriochlorophyll e
* p<0.0001
196 Table 4.5. Correlation coefficients (r) between decadally smoothed Fayetteville
Green Lake compaction-corrected varve and laminae thicknesses and Pettaquamscutt
River Estuary varve and laminae thicknesses. Significance values (p) are shown with superscripts.
Green Lake Green Lake Green Lake Varve Carbonate Laminae Organic Laminae Pettaquamscutt River 0.76* 0.75* 0.63* Varve Pettaquamscutt River 0.66* 0.66* 0.53* Clastic Laminae Pettaquamscutt River 0.70* 0.66* 0.64* Organic Laminae
* p<0.0001
197 Figure 4.1. Map of the Northeastern United States showing the locations of
Fayetteville Green Lake, New York, and the Pettaquamscutt River Estuary, Rhode
Island.
198 Figure 4.2. Detailed locus map for Fayetteville Green Lake, New York. Coring location is noted with hexagon, and is in the deepest portion of the lake. Figure from modified from Hilfinger et al. (2001). “GLSP” stands for Green Lake State Park.
199 Figure 4.3. Image of thin section (FG04FC3I) from Fayetteville Green Lake, New
York. Light layers are composed of primarily carbonate, while the dark layers have a higher concentration of organic and siliciclastic material. Note turbidite near bottom of image. Dark staining throughout image is a result of fossil pigments presumably produced by primary producers in the water column. Scale bar is 1mm x 0.5mm.
200 Year (A.D.) 2000 1950 1900 1850 1800 1750 1700 1650 1600 1550 1500 0
5
10
Core Depth (cm) 15
20
25
Figure 4.4. Age model for core FG04FC2, Fayetteville Green Lake, NY. All turbidites have been removed as they are interpreted as instantaneous events (see text for discussion).
201 Year (A.D.) 2000 1950 1900 1850 1800 1750 1700 1650 1600 1550 1500 0
5
10
Core Depth (cm) 15
20
25
Figure 4.5. Age model for core FG04FC3, Fayetteville Green Lake, NY. All turbidites have been removed as they are interpreted as instantaneous events (see text for discussion).
202 0.1 Organic Laminae (cm) 0.01 0.1
Carbonate Laminae (cm) 0.01 0.1 Varve (cm)
0.01 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 Year (A.D.)
Figure 4.6. Time series for organic lamina, carbonate lamina, and varve thickness from Fayetteville Green Lake.
203 0.4
0.35
0.3
0.25
0.2
0.15
0.1
Correlation Coefficient (r) 0.05
0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
0.7
0.6
0.5
0.4
0.3
0.2
0.1 Correlation Coefficient (r)
0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
Figure 4.7. Monthly, seasonal, and annual correlations between Fayetteville Green
Lake carbonate laminae thicknesses and annual (top) and decadal (bottom) total precipitation for New York State. Negative correlations were omitted from graphs.
Black horizontal lines indicate the 95% critical r value (p=0.05).
204 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1
Correlation Coefficient (r) 0.05 0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
0.7
0.6
0.5
0.4
0.3
0.2
0.1 Correlation Coefficient (r)
0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
Figure 4.8. Monthly, seasonal, and annual correlations between Fayetteville Green
Lake carbonate laminae thicknesses and annual (top) and decadal (bottom) Palmer
Drought Severity Index for New York State. Black horizontal lines indicate the 95% critical r value (p=0.05).
205 0.3
0.25
0.2
0.15
0.1
0.05 Correlation Coefficient (r)
0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
0.6
0.5
0.4
0.3
0.2
0.1 Correlation Coefficient (r)
0
Fall April May June July March August WinterSpring Annual January October February Summer September NovemberDecember
Figure 4.9. Monthly, seasonal, and annual correlations between Fayetteville Green
Lake organic laminae thicknesses and annual (top) and decadal (bottom) Palmer
Drought Severity Index for New York State. Black horizontal lines indicate the 95% critical r value (p=0.05).
206 Wet 0.1 Dry Organic Laminae (cm) 0.01
0.1 Wet Summer, Fall Winter, Spring
0.01 Dry
Carbonate Laminae (cm) 0.005 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 Year (A.D.)
Figure 4.10. Organic and carbonate lamina thickness time-series from Fayetteville
Green Lake and the associated climate variability preserved in each. Note that each lamina is correlated to moisture, but the organic laminae are driven by processes occurring in the winter and early spring, while the carbonate laminae are driven by moisture in the summer and fall.
207 Figure 4.11. Multi-taper method spectral analysis of Fayetteville Green Lake carbonate lamina thickness time-series. Dashed lines indicate 90%, 95%, and 99% confidence levels with respect to the AR (1) red noise hypothesis (Mann and Lees,
1996).
208 Figure 4.12. Multi-taper method spectral analysis of Fayetteville Green Lake organic lamina thickness time-series. Dashed lines indicate 90%, 95%, and 99% confidence levels with respect to the AR (1) red noise hypothesis (Mann and Lees, 1996).
209 Figure 4.13. Spectral analysis of Green Lake organic laminae (top) and
Pettaquamscutt River biogenic laminae (bottom), with 90% confidence levels shown by dashed lines. Bottom is cross-spectral analysis of the two time-series (1578-2002) with 95% confidence interval shown with dashed line. Significant coherences are illustrated with gray boxes and labeled periodicities (years).
210 Figure 4.14. Spectral analysis of Green Lake carbonate laminae (top) and
Pettaquamscutt River clastic laminae (bottom), with 90% confidence levels shown by dashed lines. Bottom is cross-spectral analysis of the two time-series (1578-2002) with 80% confidence interval shown with dashed line. Significant coherences are illustrated with gray boxes and labeled periodicities (years).
211 Appendix: Thin section preparation and analysis of frozen and soft laminated sediment cores.
In order to perform high-resolution analyses of laminated and varved sediments it is preferable to make thin sections out of the sediment. By doing this, high-resolution images can be obtained using flat-bed scanners, scanning electron microscope work can be performed, and mineral identification can be accomplished using a petrographic microscope. The challenge with making thin sections from sediment is that the material is unconsolidated, and therefore cannot be affixed to a glass slide. Therefore, the sediment must be “lithified” in the laboratory before thin sections can be fabricated. The general principle of this process is to exchange a low- viscosity resin with pore water and harder the resin without disturbing the sediment fabric. This appendix will discuss how this is accomplished for both frozen samples
(ie. from a freeze core) and soft sediment (ie. from piston and gravity cores). The reader is also directed to useful references from the literature (Clark, 1988; Francus and Asikainen, 2001; Jim, 1985; Lotter and Lemcke, 1999; Pike and Kemp, 1996;
Smith and Anderson, 1995; Spurr, 1969).
212 Creating Thin Sections from Frozen Sediment Cores • Clean core surface and photograph • In order to keep core frozen, clean and photograph in a cold room • Block plane blades work well for cleaning the core surface • Be sure to include label and scale in each photograph; note breaks in core • Sediment Cutting (Band Saw) • To keep sediment hard, either put the band saw in a freezer room and/or cool the band saw platform with dry ice. It also helps to put the wrapped core on dry ice prior to the cutting. • The final cut slab will have to fit on a 2” x 3” slide, and the slabs must overlap each other via an approximately 30˚ angle:
The thickness of each slab should not exceed about 1”. • Another option that can work well for correlation between thin sections is to set up two rows of sections side by side and have the section breaks off set. This way you have an entire thin section to look at that corresponds to section breaks. Of course you don’t want to do this if you have a limited amount of sediment. • First, cut off side strip of the core. Then score a line 1.5” from the edge of the cut and cut off this 1.5” strip. Keep track of the top! • If thickness is greater than about 1”, cut strips to reduce depth. Also, cut off water layer above the sediment/water interface. • On face of sediment, score guidelines to cut individual sample chunks as illustrated above. A template is useful in order to maintain consistent angles between chunks. • Saw off the chunks and place in a tray maintaining direction of the sediment/water interface. Place tray in freezer and allow chunks to thoroughly solidify once again. • Prepare Chunks for Freeze Drying • Foil is used to give structure to the chunks during and after freeze drying. Foil wraps consist of strips of aluminum foil approximately 35 cm long. The width of these strips should be about 0.5” wider than the thickness of the sediment chunks. • Once the chunks are sufficiently solid again, wrap them in the aluminum strips. The foil should wrap around the top and the bottom of the chunk, but should not completely cover these sides of the chunks. • It is helpful to always start the wrapping on a given side of the chunk relative to the sediment/water interface. • Note that the edges common between chunks are the most valuable. Handle the chunks accordingly. • Wrapped chunks are then placed in labeled “boats” made out of 250 ml rectangular Nalgene bottle cut to be about 1.5” high. • Freeze Dry
213 • Sediment chunks are freeze-dried on a Labconco brand Freeze Dry System for at least three days. A working temperature of -40°C and working pressure of 5-10μg Hg is maintained throughout the process. • Once the sediment is freeze-dried it is VERY FRAGILE. Any movement of the boats during this part of the process should be limited and should be conducted with the utmost of care! • Resin Impregnation • Resin is mixed in a ratio of 10(VCD) : 4(DER) : 26(NSA) : 0.4(DMAE). • Chemicals are from Polysciences, Inc. in Warrington, PA (800-523- 2575); catalog numbers are from Polysciences: • (VCD) 4-vinycyclohexene dioxide (cat. # 01912) • (DER) D.E.R. 736 epoxy resin (cat. # 02923) • (NSA) nonenyl succinic anhydride (cat. # 01542) • (DMAE) 2-dimethylaminoethnol (cat. # 01458) • Chemicals are hazardous and material safety data sheets (MSDSs) should be consulted before using these chemicals • Dunkin Donuts ice coffee clear plastic cups work well for mixing. • Make about a 612g batch for 12 boats (51g/boat). Components must be accurate to 0.05g. Stir well after each ingredient is added and keep the chemicals in a fume hood at all times during mixing and subsequent pouring. • Use a disposable pipette to slowly add the resin to each boat by pouring the liquid down the corner of the boat. Allow the sediment to absorb the epoxy from the bottom; from the top, you should see the resin soak up through the sediment like a sponge. Once the sediment is impregnated (allow at least 30 minutes), epoxy can be added to completely cover the sediment sample. • If samples contain high concentrations of pigments, allowing the samples to soak in resin at room temperature for >12 hours is recommended to evenly distribute the pigment staining in tine sections. • Cure the epoxy at 50°C for three days in fume hood. • Cutting slabs • Once cured, the epoxy “pucks” are removed from the boats; sample name and orientation arrow are sketched in the epoxy on both the top and bottom. • All pucks are cut in half horizontally using a diamond bladed slab saw with a water lubrication system. Consistency is maintained in the half chosen for thin sectioning. • The surrounding epoxy is trimmed around the sediment half that will be thin sectioned resulting in a slab of resin-impregnated sediment. • Thin-Section Preparation • All sediment chunks are sent to Ray at Quality Thin Sections, Tucson, AZ for the final thin-sectioning process (520-884-9935). • QTS makes the thin section using standard thin sectioning practices to produce an oversized 30μm unpolished thin section for each sediment chunk.
214 Creating Thin Sections from Soft Sediment Cores • Clean core surface and photograph • Be sure to include label and scale in each photograph; note breaks in core • Sediment Cutting (Band Saw) • To maintain original sediment fabric, it is best to subsample with a rigid sampling device. A good sample to use is the u-channel once all magnetic analyses have been performed • The final cut slab will have to fit on a 2” x 3” slide, and in the case of u- channel samples, two samples can be affixed side to side on a slide. The slabs must overlap each other via an approximately 30˚ angle:
The thickness of each slab should not exceed about 1”. • Another option that can work well for correlation between thin sections is to set up two rows of sections side by side and have the section breaks off set. This way you have an entire thin section to look at that corresponds to section breaks. Of course you don’t want to do this if you have a limited amount of sediment. • All cuts are made on a band saw. Be sure to keep track of the top! • Since thick impregnation samples are not necessary, it is useful to cut down the height of samples by about 50% and to archive the remaining material for precise depth measurements on other proxy measurement. • A template is useful in order to maintain consistent angles between chunks. • Saw off the chunks and place in a tray maintaining direction of the sediment/water interface. • Prepare Chunks for Dehydration (Acetone Exchange) • Place samples on plastic mesh available at craft stores to allow for fluid flow beneath samples. Place two samples on the same piece of mesh • Place the samples into labeled “boats” made out of 250 ml rectangular Nalgene bottle cut to be about 1.5” high. • NOTE: This is an acetone exchange process and acetone will degrade the plastic that u-channels are made of. Therefore, if u-channels are used for subsampling, it is necessary to remove the sediment from the plastic at this point using small a spatula. • Dehydration (Acetone Exchange) • Samples are soaked in acetone (reagent grade) to chemically dehydrate the sediment while not disturbing the original fabric. This work should be done in a fume hood. • Use a disposable pipette to slowly add acetone to each boat by pouring the liquid down the corner of the boat.
215 • Allow the sediment to soak for ~12 hours in a dessicator to reduce evaporation. • Every ~ 12 hours remove the acetone from the boats using a disposable pipette keeping in mind that this is hazardous waste and should be handled accordingly. • Add fresh acetone and repeat the soaking process. • The number of exchanges will be dependant on the permeability of the sediment. Since we know that the specific gravity of acetone is 0.785, the best way to determine when enough exchanges have been done is to monitor the specific gravity of the effluent once removed. Once this value reaches 0.78-0.79, the sample is sufficiently dehydrated. • Resin Impregnation • Resin is mixed in a ratio of 10(VCD) : 4(DER) : 26(NSA) : 0.4(DMAE). • Chemicals are from Polysciences, Inc. in Warrington, PA (800-523- 2575); catalog numbers are from Polysciences: • (VCD) 4-vinycyclohexene dioxide (cat. # 01912) • (DER) D.E.R. 736 epoxy resin (cat. # 02923) • (NSA) nonenyl succinic anhydride (cat. # 01542) • (DMAE) 2-dimethylaminoethnol (cat. # 01458) • Chemicals are hazardous and material safety data sheets (MSDSs) should be consulted before using these chemicals • Dunkin Donuts ice coffee clear plastic cups work well for mixing. • Make about a 612g batch for 12 boats (51g/boat). Components must be accurate to 0.05g. Stir well after each ingredient is added and keep the chemicals in a fume hood at all times during mixing and subsequent pouring. • Remove the final acetone using a disposable pipette. • Use a disposable pipette to slowly add the resin to each boat by pouring the liquid down the corner of the boat. Resin should cover the entire sample. • Experimentation in exchanges is necessary for different sediment types. • A conservative approach is to perform 8 resin exchanges: • 75% acetone, 25% resin (lower viscosity) • 50% acetone, 50 % resin • 25% acetone, 75% resin • 5 exchanges with 100% resin • Resin is non-toxic once cured, so waste resin should be cured in an oven at 50°C before disposal • Cure the epoxy at 50°C for three days. Be sure that the oven is working in the fume hood!
216 • A cheaper approach that has been successful with sediments from the Pettaquamscutt River is to do one 100% resin exchange, leaving the top of the sediment uncovered. The theory is that the acetone will evaporate off the top of the sediment and resin will fill in the pore spaces. In practice, not all of the acetone is evaporated. Therefore, the resin cures, but it takes weeks instead of days at 50°C. • Cutting slabs • Once cured, the epoxy “pucks” are removed from the boats; sample name and orientation arrow are sketched in the epoxy on both the top and bottom. • All pucks are cut in half horizontally using a diamond bladed slab saw with a water lubrication system. Consistency is maintained in the half chosen for thin sectioning. • The surrounding epoxy is trimmed around the sediment half that will be thin sectioned resulting in a slab of resin-impregnated sediment. • Thin-Section Preparation • All sediment chunks are sent to Ray at Quality Thin Sections, Tucson, AZ for the final thin-sectioning process (520-884-9935). • QTS makes the thin section using standard thin sectioning practices to produce an oversized 30μm unpolished thin section for each sediment chunk.
Thin Section Varve Analysis
Analysis of laminations is done on digital images in Photoshop® with assistance from a varve counting algorithm (Francus et al., 2002). Thin sections are scanned using a flat-bed scanner with transparency capabilities to produce digital tiff images (De Keyser, 1999). Resolution of the scans is determined by the scale of the lamination, however 1440 dot per inch (dpi) is advisable as this is the assumed resolution for the varve counting algorithm (Francus et al., 2002). Thin sections can be scanned under cross-polarized films or using plain light. Depending on the sediment components associated with a particular location, one technique may be preferable to the other. In the case of this thesis’ work, cross-polarized light produced the best results.
217 Bring tiff images into Adobe Photoshop for lamination analysis. Use the path tool to manually mark lamination boundaries with horizontal lines, adding a vertical axis line that intersects horizontal lines. The required procedure for the algorithm is detailed by Francus et al. (2003). Export paths as post script files once finished. Use algorithm to count and measure the thickness of each lamination
(Francus et al., 2002). The code is available for Visual Basic language and is run through Microsoft Excel. The resulting data files can be overlaid to produce a continuous record of varve thickness, number, and depth.
It is preferable to analyze multiple cores from the same location in order to compensate for laminae that may have been disturbed in one or more cores
(Lamoureux, 2001). In general, three cores from the same location are advisable because statistics can be generated on varve thickness measurements and unresolved laminations can be interpreted. The overall accuracy of the varve chronology can be determined by comparing counts to radiometric age controls (210Pb, 137Cs, 14C) as well as known introduction horizons of various proxies (DDT, PCBs, Ambrosia,
Rumex, etc.). Precision of the counts can be determined by comparing counts from adjoining cores from the same location and by recounting thin sections.
218 References
Clark, J. S., 1988, Stratigraphic charcoal analysis on petrographic thin sections: application to fire history in Northwestern Minnesota: Quaternary Research, v. 30, p. 81-91.
De Keyser, T. L., 1999, Digital scanning of thin sections and peels: Journal of Sedimentary Research, v. 69, no. 4, p. 962-964.
Francus, P., and Asikainen, C. A., 2001, Sub-sampling unconsolidated sediments: a solution for the preparation of undisturbed thin-sections from clay-rich sediments: Journal of Paleolimnology, v. 26, p. 323-326.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Jim, C. Y., 1985, Impregnation of moist and dry unconsolidated clay samples using Spurr resin for microstructural studies: Journal of Sedimentary Petrology, v. 55, p. 597-599.
Lamoureux, S., 2001, Varve chronology techniques, in Last, W. M., J.P Smols, ed., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 247-260.
Lotter, A. F., and Lemcke, G., 1999, Methods for preparing and counting biochemical varves: Boreas, v. 28, p. 243-252.
Pike, J., and Kemp, A. E. S., 1996, Preparation and analysis techniques for studies of laminated sediments, in Kemp, A. E. S., ed., Palaeoclimatology and Palaeoceanography from laminated sediments, Geological Society Special Publication No. 116, p. 37-48.
Smith, S. J., and Anderson, R. S., 1995, A method for impregnating soft sediment cores for thin-section microscopy: Journal of Sedimentary Research, Section A: Sedimentary Petrology and Processes, v. 65, no. 3, p. 576-577.
Spurr, A. R., 1969, A low-viscosity epoxy resin embedding medium for electron microscopy: Journal of Ultrastructure Research, v. 26, p. 31-43.
219 Bibliography
Allen, M. R., and Ingram, W. J., 2002, Constraints on future changes in climate and the hydrologic cycle: Nature, v. 419, p. 224-232.
Appenzeller, C., Stocker, T. F., and Anklin, M., 1998, North Atlantic Oscillation dynamics recorded in Greenland ice cores: Science, v. 282, p. 446-449.
Balco, G., and Schaefer, J., 2006, Cosmogenic-nuclide and varve chronologies for the deglaciation of southern New England: Quaternary Geochronology, In Press.
Balco, G., Stone, J. O. H., Porter, S. C., and Caffee, M. W., 2002, Cosmogenic- nuclide ages for New England coastal moraines, Martha's Vineyard and Cape Cod, Massachusetts, USA: Quaternary Science Reviews, v. 21, p. 2127-2135.
Battarbee, R., Birks, J., Oldfield, F., and Mackay, A., 2003, Global Change in the Holocene, Hodden Arnold Publishers, p. 544.
Bianchi, T. S., Demetropoulos, A., Hadjichristophorou, M., Argyrou, M., Baskaran, M., and Lambert, C. D., 1996, Plant pigments as biomarkers of organic matter sources and coastal waters of Cyprus (eastern Mediterranean): Estuarine, Coastal, and Shelf Science, v. 42, p. 103-115.
Björck, S., and Wohlfarth, B., 2001, 14C chronostratigraphic techniques in paleolimnology, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 205-245.
Boothroyd, J. C., 1991, The geologic history of the Narrow River: Maritimes, v. 35, p. 3-5.
Boothroyd, J. C., Freedman, J. H., Brenner, H. B., and Stone, J. R., 1998, The glacial geology of southern Rhode Island, in Murray, D. P., ed., Guidebook to Field Trips in Rhode Island and Adjacent Regions of Connecticut and Massachusetts: 1998 New England Intercollegiate Geological Conference.
220 Boynton, W. R., Kemp, W. M., and Keefe, C. W., 1982, A comparative analysis of nutrients and other factors influencing estuarine phytoplankton productivity, in Kennedy, V. S., ed., Estuarine Comparisons: New York, Academic Press, p. 69-90.
Bradbury, J. A., Dingman, S. L., and Keim, B. D., 2002a, New England drought and relations with large scale atmospheric circulation patterns: Journal of the American Water Resources Association, v. 38, no. 5, p. 1287-1299.
Bradbury, J. A., Keim, B. D., and Wake, C. P., 2002b, U.S. East Coast Trough Indexes at 500 hPa and New England winter climate variability: Journal of Climate, v. 15, p. 3509-3517.
-, 2003, The influence of regional storm tracking and teleconnections on winter precipitation in the northeastern United States: Annals of the Association of American Geographers, v. 93, no. 3, p. 544-556.
Briffa, K. R., 2000, Annual climate variability in the Holocene: interpreting the message of ancient trees: Quaternary Science Reviews, v. 19, p. 87-105.
Brigham-Grette, J., 1996, Geochronology of Glacial Deposits, Ch 14, in Menzies, J., ed., Past Glacial Environments - Sediments, Landforms, and Techniques: Oxford, Butterworth Heinemann, p. 377-410.
Brunskill, G. J., 1969, Fayetteville Green Lake, New York. II. Precipitation and sedimentation of calcite in a meromictic lake with laminated sediments: Limnology and Oceanography, v. 14, no. 6, p. 830-847.
Brunskill, G. J., and Ludlam, S. D., 1969, Fayetteville Green Lake, New York. I. Physical and chemical limnology: Limnology and Oceanography, v. 14, no. 6, p. 817-829.
Bryson, R. A., and Hare, F. K., 1974, Climates of North America: Volume 11 of World Survey of Climatology, Elsevier Publishing Co.
Burgess, R. M., McKinney, R. A., and Brown, W. A., 1996, Enrichment of marine sediment colloids with polychlorinated biphenyls: Trends resulting from PCB solubility and chlorination: Environmental Science and Technology, v. 30, no. 8, p. 2556-2566.
221 Burnett, A. W., Mullins, H. T., and Patterson, W. P., 2004, Relationship between atmospheric circulation and winter precipitation 18O in central New York State: Geophysical Research Letters, v. 31, L22209, doi:10.1029/20004GL021089.
Chen, N., Bianchi, T. S., McKee, B. A., and Bland, J. M., 2001, Historical trends of hypoxia on the Louisiana shelf: application of pigments as biomarkers: Organic Geochemistry, v. 32, p. 543-561.
Clark, J. S., 1988, Stratigraphic charcoal analysis on petrographic thin sections: application to fire history in Northwestern Minnesota: Quaternary Research, v. 30, p. 81-91.
Clark, R. B., 1997, Marine Pollution: Oxford, Clarendon Press, 161 p.
Cole, J. R., 1889, History of Washington and Kent Counties, Rhode Island: New York, W.W. Preston & Co.
Conley, D. J., Markager, S., Andersen, J., Ellermann, T., and Svendsen, L. M., 2002, Coastal eutrophication and the Danish National Aquatic Monitoring and Assessment Program: Estuaries, v. 25, no. 4b, p. 848-861.
Cook, E. R., 2003, Multi-proxy reconstructions of the North Atlantic Oscillation (NAO) Index: A critical review and a new well-verified Winter NAO Index reconstruction back to AD 1400, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 63-79.
Cronin, T., Willard, D., Karlsen, A., Ishman, S., Verardo, S., McGeehin, J., Kerhin, R., Holmes, C., Colman, S., and Zimmerman, A., 2000, Climatic variability in the eastern United States over the past millennium from Chesapeake bay sediments: Geology, v. 28, no. 1, p. 3-6.
Crutzen, P. J., 2002, Geology of mankind: Nature, v. 415, p. 23.
222 Cubasch, U., Meehl, G. A., Boer, G. J., Stouffer, R. J., Dix, M., Noda, A., Senior, C. A., Raper, S., and Yap, K. S., 2001, Projections of future climate change, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 525-582.
Czaja, A., Robertson, A. W., and Huck, T., 2003, The role of Atlantic Ocean- atmosphere coupling in affecting North Atlantic Oscillation variability, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 147-172.
Davis, P., 2006, Buying and selling the human species: Newport and the Slave Trade, The Providence Journal: Providence.
De Keyser, T. L., 1999, Digital scanning of thin sections and peels: Journal of Sedimentary Research, v. 69, no. 4, p. 962-964.
Dean, W. E., Jr., 1974, Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss-on-ignition: comparison with other methods: Journal of Sedimentary Petrology, v. 44, p. 242-248.
Deevey, E. S., Nakai, N., and Stuiver, M., 1963, Fractionation of sulfur and carbon isotopes in a meromictic lake: Science, v. 139, p. 407-408.
Delworth, T. L., and Dixon, K. W., 2000, Implications of the recent trend in the Arctic/North Atlantic Oscillation for the North Atlantic thermohaline circulation: Journal of Climate, v. 13, p. 3721-3727.
Delworth, T. L., and Mann, M. E., 2000, Observed and simulated multidecadal variability in the Northern Hemisphere: Climate Dynamics, v. 16, p. 661-676.
Doniger, D., Friedman, D., Hwang, R., Lashof, D., and Mark, J., 2002, Dangerous addictions-Ending America's oil dependence: Natural Resources Defense Council and Union of Concerned Scientists.
223 Ernst, L. M., Miguel, L. K., and Willis, J., 1999, The Narrow River special area management plan for the watershed of the Narrow River in the towns of North Kingstown, South Kingstown, and Narragansett: Rhode Island Coastal Resources Management Council.
Faegri, K., and Iversen, J., 1989, Textbook of Pollen Analysis: New York, NY, John Wiley and Sons, Ltd., 237 p.
Figgins, D., 2004, The shipbuilding legacy of John A. Saunders, South County Independent: Wakefield, RI, p. A11.
Folland, C. K., Karl, T. R., Christy, J. R., Clarke, R. A., Gruza, G. V., Jouzel, J., Mann, M. E., Oerlemans, J., Salinger, M. J., and Wang, S. W., 2001, Observed climate variability and change, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 99-182.
Francis, D. R., and Foster, D. R., 2001, Response of small New England ponds to historic land use: The Holocene, v. 11, no. 3, p. 301-312.
Francus, P., and Asikainen, C. A., 2001, Sub-sampling unconsolidated sediments: a solution for the preparation of undisturbed thin-sections from clay-rich sediments: Journal of Paleolimnology, v. 26, p. 323-326.
Francus, P., Keimig, F., and Besonen, M., 2002, An algorithm to aid varve counting and measurement from thin-sections: Journal of Paleolimnology, v. 28, p. 283-286.
Freedman, J. M., 1997, Reconstruction of late Pleistocene to early Holocene environments and paleoclimates of the Great Swamp area, South Kingstown, Rhode Island [MS thesis]: University of Rhode Island, 98 p.
Gaines, A. G. J., 1975, Papers on the geomorphology, hydrography, and geochemistry of the Pettaquamscutt River Estuary [Ph.D. thesis]: University of Rhode Island.
224 Giorgi, F., Hewitson, B., Christensen, J., Hulme, M., von Storch, H., Whetton, P., Jones, R., Mearns, L., and Fu, C., 2001, Regional climate information- Evaluation and projections, in Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linder, P. J., Dai, X., Maskell, K., and Johnson, C. A., eds., Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, p. 583-638.
Glew, J. R., Smol, J. P., and Last, W. M., 2001, Sediment core collection and extrusion, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 73-105.
Gorham, E., and Sanger, J. E., 1976, Fossilized pigments as stratigraphic indicators of cultural eutrophication in Shagawa Lake, northeastern Minnesota: Geological Society of America Bulletin, v. 87, p. 1638-1642.
Gosse, J. C., and Phillips, F. M., 2001, Terrestrial in situ cosmogenic nuclides: theory and application: Quaternary Science Reviews, v. 20, no. 14, p. 1475-1560.
Gray, S. T., Graumlich, L. J., Betancourt, J. L., and Pederson, G. T., 2004, A tree-ring based reconstruction of the Atlantic Multidecadal Oscillation since 1567 A.D.: Geophysical Research Letters, v. 31, L12205, doi:10.1029/2004GL019932.
Grissino-Mayer, H. D., 1996, A 2129-year reconstruction of precipitation for northwestern New Mexico, USA, in Dean, J. S., Meko, D. M., and Swetnam, T. W., eds., Tree Rings, Environment, and Humanity: Tucson, Radiocarbon, p. 191-204.
Gustavson , T. C., and Boothroyd, J. C., 1987, A depositional model for outwash, sediment sources, and hydrologic characteristics, Malaspina Glacier, Alaska: a modern analog of the southeastern margin of the Laurentide Ice Sheet: Geological Society of America Bulletin, v. 99, p. 187-200.
Guttman, N. B., and Quayle, R. G., 1996, A historical perspective of U.S. climate divisions: Bulletin of the American Meteorological Society, v. 77, no. 2, p. 293-303.
225 Hargraves, P. E., 29 September 2003, Personal Communication.
Harradine, P. J., Harris, P. G., Head, R. N., Harris, R. P., and Maxwell, J. R., 1996, Steryl chlorin esters are formed by zooplankton grazing: Geochimica et Cosmochimica Acta, v. 60, no. 12, p. 2265-2270.
Hartmann, P., 2001, Sources, transport, and fate of organic contaminants in Narragansett Bay, RI [PhD thesis]: University of Rhode Island, 335 p.
Hartmann, P., Quinn, J. G., Cairns, R. W., and King, J. W., 2005, Depositional history of organic contaminants in Narragansett Bay, Rhode Island, USA: Marine Pollution Bulletin, v. 50, no. 4, p. 388-395.
Hawk, J. D., 1998, The role of the North Atlantic Oscillation in winter climate variability as it relates to the winter-spring bloom in Narragansett Bay [Master of Science thesis]: University of Rhode Island, 148 p.
Hilfinger, M. F., IV, and Mullins, H. T., 1997, Geology, Limnology, and Paleoclimatology of Green Lakes State Park, New York, in Raynes, T. W., ed., 69th Annual New York State Geological Association Field Trip Guide: Clinton, NY, Hamilton College, p. 127-159.
Hilfinger, M. F., IV, Mullins, H. T., Burnett, A. W., and Kirby, M. E., 2001, A 2500 year sediment record from Fayetteville Green Lake, New York: Evidence for anthropogenic impacts and historic isotope shift: Journal of Paleolimnology, v. 26, p. 293-305.
Hubeny, J. B., and King, J. W., 2003, Anthropogenic eutrophication as recorded by varved sediments in the Pettaquamscutt River Estuary, Rhode Island, USA: GSA Abstracts with Programs, v. 35, no. 6, p. 282.
Hubeny, J. B., King, J. W., and Santos, A., 2006, Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments: Geology, v. 34, no. 7, p. 569-572.
Hughen, K. A., Overpeck, J. T., and Anderson, R. F., 2000, Recent warming in a 500- year palaeotemperature record from varved sediments, Upper Soper Lake, Baffin Island, Canada: The Holocene, v. 10, no. 1, p. 9-19.
226 Hurrell, J. W., 1995, Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation: Science, v. 269, p. 676-679.
Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., 2003, An overview of the North Atlantic Oscillation, in Hurrell, J. W., Kushnir, Y., Ottersen, G., and Visbeck, M., eds., The North Atlantic Oscillation: climatic significance and environmental impact: Geophysical Monograph 134: Washington, D.C., American Geophysical Union, p. 1-35.
IPCC, 2001, Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press, 881 p.
Jim, C. Y., 1985, Impregnation of moist and dry unconsolidated clay samples using Spurr resin for microstructural studies: Journal of Sedimentary Petrology, v. 55, p. 597-599.
Jones, P. D., Jonsson, T., and Wheeler, D., 1997, Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland: International Journal of Climatology, v. 17, p. 1433- 1450.
Jones, P. D., and Mann, M. E., 2004, Climate over past millennia: Reviews of Geophysics, v. 42, RG2002, doi:10.1029/2003RG000143.
Karl, T. R., Williams, C. N., Jr., Young, P. J., and Wendland, W. M., 1986, A model to estimate the time of observation bias associated with monthly mean maximum, minimum, and mean temperature for the United States: Journal of Climate and Applied Meteorology, v. 25, no. 1, p. 145-160.
Keller, A. A., Oviatt, C. A., Walker, H. A., and Hawk, J. D., 1999, Predicted impacts of elevated temperature on the magnitude of the winter-spring phytoplankton bloom in temperate coastal waters: A mesocosm study: Limnology and Oceanography, v. 44, no. 2, p. 344-356.
Kelly, R. P., and Moran, S. B., 2002, Seasonal changes in groundwater input to a well-mixed estuary estimated using radium isotopes and implications for coastal nutrient budgets: Limnology and Oceanography, v. 47, no. 6, p. 1796- 1807.
227 King, J. W., and Channell, J. E. T., 1991, Sedimentary magnetism, environmental magnetism, and magnetostratigraphy: Reviews of Geophysics Supplement, p. 358-370.
King, J. W., and Peck, J., 2001, Use of paleomagnetism in studies of lake sediments, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediment. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 371-389.
Kirby, M. E., Mullins, H. T., Patterson, W. P., and Burnett, A. W., 2001, Lacustrine isotopic evidence for multidecadal natural climate variability related to the circumpolar vortex over the northeast United States during the past millennium: Geology, v. 29, no. 9, p. 807-810.
-, 2002a, Late glacial-Holocene atmospheric circulation and precipitation in the northeast United States inferred from modern calibrated stable oxygen and carbon isotopes: Geological Society of America Bulletin, v. 114, no. 10, p. 1326-1340.
Kirby, M. E., Patterson, W. P., Mullins, H. T., and Burnett, A. W., 2002b, Post- Younger Dryas climate interval linked to circumpolar vortex variability: isotopic evidence from Fayetteville Green Lake, New York: Climate Dynamics, v. 19, p. 321-330.
Knight, J. R., Allan, R. J., Folland, C. K., Vellinga, M., and Mann, M. E., 2005, A signature of persistent natural thermohaline circulation cycles in observed climate: Geophysical Research Letters, v. 32, L20708, doi:10.1029/2005GL024233.
Koteff, C., and Larsen, F. D., 1989, Postglacial uplift in western New England: Geologic evidence for delayed rebound, in Gregersen, S., and Basham, P. W., eds., Earthquakes at North-Atlantic Passive Margins: Neotectonics and postglacial rebound: Dordrecht, Netherlands, Kluwer Academic Publishers, p. 105-123.
Koteff, C., and Pessl, F. J., 1981, Systematic ice retreat in New England: U.S. Geological Survey Professional Paper 1179.
228 Kreitler, C. W., and Browning, L. A., 1983, Nitrogen-isotope analysis of groundwater nitrate in carbonate aquifers: Natural sources versus human pollution: Journal of Hydrology, v. 61, p. 285-301.
Kushnir, Y., 1994, Interdecadal variations in North Atlantic sea surface temperature and associated atmospheric conditions: Journal of Climate, v. 7, p. 141-157.
Labitzke, K., and van Loon, H., 1997, The signal of the 11-year sunspot cycle in the upper troposphere-lower stratosphere: Space Science Reviews, v. 80, no. 3-4, p. 393-410.
Lacey, E. M., 1997, Evaluation of simultaneously extracted metals-acid volatile sulfide (SEM-AVS) in lacustrine sediment (Lake Champlain), estuarine sediment (Narragansett Bay), and under laboratory conditions [PhD thesis]: University of Rhode Island, 226 p.
Lamoureux, S., 2001, Varve chronology techniques, in Last, W. M., J.P Smols, ed., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 247-260.
Latif, M., and Barnett, T. P., 1994, Causes of decadal climate variability over the North Pacific and North America: Science, v. 266, p. 634-637.
Lauenstein, G. G., Cantillo, A. Y., and Dolvin, S. S., 1993, NOAA National Status and Trends Program development and methods, in Lauenstein, G. G., and Cantillo, A. Y., eds., Sampling and analytical methods of the National Status and Trends Program, National Benthic Surveillance, and Mussel Watch Programs: 1984-1992: Silver Spring, MD, National Oceanic and Atmospheric Administration.
Lean, J., and Rind, D., 1998, Climate forcing by changing solar radiation: Journal of Climate, v. 11, p. 3069-3094.
Leathers, D. J., Grundstein, A. J., and Ellis, A. W., 2000, Growing season moisture deficits across the northeastern United States: Climate Research, v. 14, p. 43- 55.
229 Leathers, D. J., Kluck, D., and Kroczynski, S., 1998, The severe flooding event of January 1996 across north-central Pennsylvania: Bulletin of the American Meteorological Society, v. 79, p. 785-797.
Leathers, D. J., and Palecki, M. A., 1992, The Pacific/North American teleconnection pattern and United States climate. Part II: Temporal characteristics and index specification: Journal of Climate, v. 5, p. 707-716.
Leathers, D. J., Yarnal, B., and Palecki, M. A., 1991, The Pacific/North American teleconnection pattern and United States climate. Part I: Regional temperature and precipitation associations: Journal of Climate, v. 4, p. 517-528.
Leavitt, P. R., 1993, A review of factors that regulate carotenoid and chlorophyll deposition and fossil pigment abundance: Journal of Paleolimnology, v. 9, p. 109-127.
Leavitt, P. R., and Hodgson, D. A., 2001, Sedimentary pigments, in Last, W. M., and Smols, J. P., eds., Tracking environmental change using lake sediments. Volume 3: Terrestrial, algal, and siliceous indicators: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 295-325.
Lima, A. L., Hubeny, J. B., Reddy, C. M., King, J. W., Hughen, K. A., and Eglinton, T. I., 2005, High-resolution historical records from Pettaquamscutt River basin sediments: 1. 210Pb and varve chronologies validate record of 137Cs released by the Chernobyl accident: Geochimica et Cosmochimica Acta, v. 69, no. 7, p. 1803-1812.
Lima, A. L. C., Eglinton, T. I., and Reddy, C. M., 2003, High-resolution record of pyrogenic polycyclic aromatic hydrocarbon deposition during the 20th century: Environmental Science and Technology, v. 37, p. 53-61.
Lins, H. F., 1997, Regional streamflow regimes and hydroclimatology of the United States: Water Resources Research, v. 33, no. 7, p. 1655-1667.
Long, E. R., Ingersoll, C. G., and MacDonald, D. D., 2006, Calculation and uses of mean sediment quality guideline quotients: A critical review: Environmental Science and Technology, v. 40, no. 6, p. 1726-1736.
230 Long, E. R., MacDonald, D. D., Smith, S. L., and Calder, F. D., 1995, Evidence of adverse biological effects within ranges of chemical concentrations in marine and estuarine sediments: Environmental Management, v. 19, p. 81-97.
Lotter, A. F., and Lemcke, G., 1999, Methods for preparing and counting biochemical varves: Boreas, v. 28, p. 243-252.
Ludlam, S. D., 1969, Fayetteville Green Lake, New York. III. The Laminated Sediments: Limnology and Oceanography, v. 14, no. 6, p. 848-857.
-, 1981, Sedimentation rates in Fayetteville Green Lake, New York, U.S.A.: Sedimentology, v. 28, p. 85-96.
-, 1984, Fayetteville Green Lake, New York, U.S.A. VII. Varve Chronology and Sediment Focusing: Chemical Geology, v. 44, p. 85-100.
Luterbacher, J., Xoplaki, E., Dietrich, D., Jones, P. D., Davies, T. D., Portis, D., Gonzalez-Rouco, J. F., von Storch, H., Gyalistras, D., Casty, C., and Wanner, H., 2002, Extending North Atlantic Oscillation reconstructions back to 1500: Atmospheric Science Letters, v. 2, p. 114-124.
Mann, M. E., 2002, The value of multiple proxies: Science, v. 297, p. 1481-1482.
Mann, M. E., Bradley, R. S., and Hughes, M. K., 1998, Global-scale temperature patterns and climate forcing over the past six centuries: Nature, v. 392, p. 779- 787.
Mann, M. E., and Jones, P. D., 2003, Global surface temperatures over the past two millennia: Geophysical Research Letters, v. 30, no. 15, p. 1820.
Mann, M. E., and Lees, J. M., 1996, Robust estimation of background noise and signal detection in climatic time series: Climatic Change, v. 33, p. 409-445.
Mann, M. E., and Park, J., 1996, Joint spatiotemporal modes of surface temperature and sea level pressure variability in the Northern Hemisphere during the last century: Journal of Climate, v. 9, no. 9, p. 2137-2162.
231 Mann, M. E., Park, J., and Bradley, R. S., 1995, Global interdecadal and century- scale climate oscillations during the past five centuries: Nature, v. 378, p. 266- 270.
McKinney, R. A., Lake, J. L., Charpentier, M., and Ryba, S., 2002, Using mussel isotope ratios to assess anthropogenic nitrogen inputs to freshwater ecosystems: Environmental Monitoring and Assessment, v. 74, p. 167-192.
McWeeney, L. J., 1995, Revised vegetation history for the post-glacial period (15,200 - 10,000 14C years b.p.) in southern New England: GSA Abstracts with Programs, v. 27, p. 68.
Mecray, E. L., 1994, An assessment of historical trace metal accumulation in the Burlington region of Lake Champlain, Vermont [Masters thesis]: University of Rhode Island, 271 p.
Mecray, E. L., King, J. W., and Corbin, J. M., 1991, Metal pollution recorded in Pettaquamscutt River sediments: Maritimes, v. 35, no. 2, p. 10-11.
Meyers, P. A., and Teranes, J. L., 2001, Sediment organic matter, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 2: Physical and geochemical methods: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 239-269.
Milly, P. C. D., Dunne, K. A., and Vecchia, A. V., 2005, Global pattern of trends in streamflow and water availability in a changing climate: Nature, v. 438, p. 347-350.
Mooney, H., Cropper, A., and Reid, W., 2005, Confronting the human dilemma: Nature, v. 434, p. 561-562.
Moore, G. W. K., Holdsworth, G., and Alverson, K., 2002, Climate change in the North Pacific region over the past three centuries: Nature, v. 420, p. 401-403.
Muller, E. H., 1967, Geologic setting of Green and Round Lakes near Fayetteville, New York, in Jackson, D. F., ed., Some Aspects of Meromixis: Syracuse, NY, Department of Civil Engineering, Syracuse University, p. 96-121.
232 Namias, J., 1966, Nature and possible causes of the northeastern United States drought during 1962-65: Monthly Weather Review, v. 94, no. 9, p. 543-554.
Nederbragt, A. J., and Thurow, J., 2005, Amplitude of ENSO cycles in the Santa Barbara Basin, off California, during the past 15000 years: Journal of Quaternary Science, v. 20, no. 5, p. 447-456.
Nowaczyk, N. R., 2001, Logging of magnetic susceptibility, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments. Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 155-170.
O'Conner, T. P., 1996, Trends in chemical concentrations in mussels and oysters collected along the US coast from 1986-1993: Marine Environmental Research, v. 41, p. 183-200.
Orr, W. L., and Gaines, A. G., Jr., 1973, Observations on rate of sulfate reduction and organic matter oxidation in the bottom waters of an estuarine basin: the upper basin of the Pettaquamscutt River (Rhode Island), in Tissot, B., and Bienner, F., eds., Advances in Organic Geochemistry, Proceedings of the 6th International Meeting on Organic Geochemistry: Paris, Editions technip, p. 791-812.
Paillard, D., Labeyrie, L., and Yiou, P., 1996, Macintosh program performs time- series analysis: Eos, v. 77, p. 379.
Peck, J. A., and McMaster, R. L., 1991, Stratigraphy and geologic history of Quaternary sediments in lower West Passage, Narragansett Bay, Rhode Island: Journal of Coastal Research, v. 11, p. 25-37.
Peters, C., and Thompson, R., 1998, Magnetic identification of selected natural iron oxides and sulphides: Journal of Magnetism and Magnetic Materials, v. 183, p. 365-374.
Pike, J., and Kemp, A. E. S., 1996, Preparation and analysis techniques for studies of laminated sediments, in Kemp, A. E. S., ed., Palaeoclimatology and Palaeoceanography from laminated sediments, Geological Society Special Publication No. 116, p. 37-48.
233 Pilcher, J. R., 2003, Radiocarbon dating and environmental radiocarbon studies, in Mackay, A., Battarbee, R., Birks, J., and Oldfield, F., eds., Global Change in the Holocene: London, Arnold, p. 63-74.
Proctor, C. J., Baker, A., and Barnes, W. L., 2002, A three thousand year record of North Atlantic climate: Climate Dynamics, v. 19, p. 449-454.
Quadrelli, R., and Wallace, J. M., 2004, A simplified linear framework for interpreting patterns of Northern Hemisphere wintertime climate variability: Journal of Climate, v. 17, p. 3728-3744.
Quinn, J. G., 2006, Personal Communication.
Redmond, K. T., and Koch, R. W., 1991, Surface climate and streamflow variability in the Western United States and their relationship to large-scale circulation indices: Water Resources Research, v. 27, no. 9, p. 2381-2399.
Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W., Bertrand, C. J. H., Blackwell, P. G., Buck, C. E., Burr, G. S., Cutler, K. B., Damon, P. E., Edwards, R. L., Fairbanks, R. G., Friedrich, M., Guilderson, T. P., Hogg, A. G., Hughen, K. A., Kromer, B., McCormac, G., Manning, S., Ramsey, C. B., Reimer, R., Remmele, S., Southon, J. R., Stuiver, M., Talamo, S., Taylor, F. W., van der Plicht, J., and Weyhenmeyer, C. E., 2004, INTCAL04 terrestrial radiocarbon age calibration, 0-26 cal kyr bp: Radiocarbon, v. 46, p. 1029- 1058.
Reitler, L., Falk, H., Groat, C., and Coussens, C. M., 2004, From source water to drinking water: Workshop summary: Washington, D.C., National Academies Press, 126 p.
Ridge, J. C., 2003, The last deglaciation of the northeastern United States: a combined varve, paleomagnetic, and calibrated 14C chronology, in Cremeens, D. L., and Hart, J. P., eds., Geoarchaeology of landscapes in the glaciated Northeast: New York State Museum Bulletin 497: Albany, NY, University of the State of New York, p. 15-45.
-, 2004, The Quaternary glaciation of western New England with correlations to surrounding areas, in Ehlers, J., and Gibbard, P. L., eds., Quaternary glaciations - Extent and chronology, Elsevier, p. 169-199.
234 Ridge, J. C., and Larsen, F. D., 1990, Reevaluation of Antevs' New England varve chronology and new radiocarbon dates of sediments from glacial Lake Hitchcock: Geological Society of America Bulletin, v. 102, p. 889-899.
Rodwell, M. J., Rowell, D. P., and Folland, C. K., 1999, Oceanic forcing of the wintertime North Atlantic Oscillation and European climate: Nature, v. 398, p. 320-323.
Ropelewski, C. F., and Jones, P. D., 1987, An extension of the Tahiti-Darwin Southern Oscillation Index: Monthly Weather Review, v. 115, p. 2161-2165.
Rossby, T., and Benway, R. L., 2000, Slow variations in mean path of the Gulf Stream east of Cape Hatteras: Geophysical Research Letters, v. 27, no. 1, p. 117-120.
Ruddiman, W. F., 2001, Earth's Climate: Past and future: New York, W.H. Freeman and Company, 465 p.
Schafer, J. P., 1961a, Surficial geology of the Narragansett Pier quadrangle, Rhode Island: U.S. Geological Survey.
-, 1961b, Surficial geology of the Wickford quadrangle, Rhode Island: U.S. Geological Survey.
Shapley, M. D., Ito, E., and Donovan, J. J., 2005, Authigenic calcium carbonate flux in groundwater-controlled lakes: Implications for lacustrine paleoclimate records: Geochimica et Cosmochimica Acta, v. 69, no. 10, p. 2517-2533.
Shiklomanov, I. A., and Rodda, J. C., 2003, World Water Resources at the Beginning of the 21st Century: Cambridge, Cambridge University Press, 435 p.
Shindell, D. T., Schmidt, G. A., Mann, M. E., Rind, D., and Waple, A., 2001, Solar Forcing of Regional Climate Change During the Maunder Minimum: Science, v. 294, p. 2149-2152.
Shindell, D. T., Schmidt, G. A., Miller, R. L., and Mann, M. E., 2003, Volcanic and solar forcing of climate change during the Preindustrial Era: Journal of Climate, v. 16, p. 4094-4107.
235 Sieburth, J. M., and Donaghay, P. L., 1993, Planktonic methane production and oxidation within the algal maximum of the pycnocline: seasonal fine-scale observations in an anoxic estuarine basin: Marine Ecology Progress Series, v. 100, p. 3-15.
Smith, S. J., and Anderson, R. S., 1995, A method for impregnating soft sediment cores for thin-section microscopy: Journal of Sedimentary Research, Section A: Sedimentary Petrology and Processes, v. 65, no. 3, p. 576-577.
Smith, T. M., and Reynolds, R. W., 2004, Improved extended reconstruction of SST (1854-1997): Journal of Climate, v. 17, p. 2466-2477.
Snowball, I., and Torii, M., 1999, Incidence and significance of magnetic iron sulphides in Quaternary sediments and soils, in Maher, B. A., and Thompson, R., eds., Quaternary climates, environments, and magnetism: Cambridge, UK, Cambridge University Press, p. 199-231.
Spurr, A. R., 1969, A low-viscosity epoxy resin embedding medium for electron microscopy: Journal of Ultrastructure Research, v. 26, p. 31-43.
Stahle, D. W., Cook, E. R., Cleaveland, M. K., Therrell, M. D., Meko, D. M., Grissino-Mayer, H. D., Watson, E., and Luckman, B. H., 2000, Tree-ring data document 16th century Megadrought of North America: Eos, v. 81, no. 12, p. 121, 125.
Stone, B. D., and Borns, H. W., 1986, Pleistocene glacial and interglacial stratigraphy of New England, Long Island, and adjacent Georges Bank and Gulf of Maine, in Sibrava, V., Bowen, D. Q., and Richmond, G. M., eds., Quaternary Glaciations in the Northern Hemisphere, Quaternary Science Reviews 5: Oxford, Permagon Press, p. 39-52.
Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormac, G., van der Plicht, J., and Spurk, M., 1998, INTCAL98 radiocarbon age calibration, 24,000-0 cal BP: Radiocarbon, v. 40, no. 3, p. 1041-1083.
Sutton, R. T., and Hodson, D. L. R., 2003, Influence of the ocean on North Atlantic climate variability 1871-1999: Journal of Climate, v. 16, p. 3296-3313.
236 Takahashi, T., Broecker, W., Li, Y. H., and Thurber, D., 1968, Chemical and isotopic balances for a meromictic lake: Limnology and Oceanography, v. 13, p. 272- 292.
Talbot, M. R., 2001, Nitrogen isotopes in palaeolimnology, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments: Volume 2: Physical and geochemical methods: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 401-439.
Tavendale, M., H., McFarlane, P., N., Mackie, K. L., Wilkins, A. L., and Langdon, A. G., 1997, The fate of resin acids-1. The biotransformation and degradation of deuterium labeled dehydroabietic acid in anaerobic sediments: Chemosphere, v. 35, no. 10, p. 2137-2151.
Taylor, A. H., and Stephens, J. A., 1998, The North Atlantic Oscillation and the latitude of the Gulf Stream: Tellus, v. 50A, p. 134-142.
Taylor, B. R., Parkinson, D., and Parsons, W. F. J., 1989, Nitrogen and lignin content a predictors of litter decay rates: A microcosm test: Ecology, v. 70, no. 1, p. 97-104.
Thompson, J. B., Ferris, F. G., and Smith, D. A., 1990, Geomicrobiology and sedimentology of the mixolimnion and chemocline in Fayetteville Green Lake, New York: Palaios, v. 5, p. 52-75.
Thompson, J. B., Schultze-Lam, S., Beveridge, T. J., and Des Marais, D. J., 1997, Whiting events: Biogenic origin due to the photosynthetic activity of cyanobacterial picoplankton: Limnology and Oceanography, v. 42, p. 133- 141.
Thompson, R., and Oldfield, F., 1986, Environmental Magnetism: London, Allen & Unwin Ltd., 227 p.
Tinsley, B. A., 2000, Influence of solar wind on the global electric circuit, and inferred effects on cloud microphysics, temperature, and dynamics in the troposphere: Space Science Reviews, v. 94, no. 1-2, p. 231-258.
Torrence, C., and Compo, G. P., 1998, A practical guide to wavelet analysis: Bulletin of the American Meteorological Society, v. 79, no. 1, p. 61-78.
237 Urish, D. W., 1991, Freshwater inflow to the Narrow River: Maritimes, v. 35, p. 12- 14.
Verosub, K. L., and Roberts, A. P., 1995, Environmental magnetism: Past, present, and future: Journal of Geophysical Research, v. 100, no. B2, p. 2175-2192.
Wahlen, M., and Lewis, D. M., 1980, Green Lake: Dating of laminated sediments: EOS, v. 61, no. 46, p. 964.
Wallace, J. M., and Gutzler, D. S., 1981, Teleconnections in the geopotential height field during the Northern Hemisphere winter: Monthly Weather Review, v. 109, p. 784-812.
Weedon, G., 2003, Time-series analysis and cyclostratigraphy: Examining stratigraphic records of environmental cycles, Cambridge Univ. Press, 259 p.
White, W. B., Lean, J., Cayan, D. R., and Dettinger, M. D., 1997, Response of global upper ocean temperature to changing solar irradiance: Journal of Geophysical Research, v. 102, no. C2, p. 3255-3266.
Willard, D. A., Cronin, T. M., and Verardo, S., 2003, Late-Holocene climate and ecosystem history from Chesapeake Bay sediment cores, USA: The Holocene, v. 13, no. 2, p. 201-214.
Woodhouse, C. A., and Overpeck, J. T., 1998, 2000 years of drought variability in the central United States: Bulletin of the American Meteorological Society, v. 79, no. 12, p. 2693-2714.
Wright, H. E., Jr., 1980, Cores of soft lake sediments: Boreas, v. 9, p. 107-114.
Wright, S. W., Jeffrey, S. W., Mantoura, R. F. C., Llewellyn, C. A., Bjornland, T., Repeta, D., and Welschmeyer, N., 1991, Improved HPLC method for the analysis of chlorophylls and carotenoids from marine phytoplankton: Marine Ecology Progress Series, v. 77, p. 183-196.
Wu, P., and Gordon, C., 2002, Oceanic influence on North Atlantic climate variability: Journal of Climate, v. 15, p. 1911-1925.
238 Yarnal, B., and Leathers, D. J., 1988, Relationships between interdecadal and interannual climatic variations and their effect on Pennsylvania climate: Annals of the Association of American Geographers, v. 78, no. 4, p. 624-641.
Zolitschka, B., Mingram, J., van der Gaast, S., Jansen, J. H. F., and Naumann, R., 2001, Sediment logging techniques, in Last, W. M., and Smol, J. P., eds., Tracking environmental change using lake sediments: Volume 1: Basin analysis, coring, and chronological techniques: Dordrecht, The Netherlands, Kluwer Academic Publishers, p. 137-153.
239