University of Florida Thesis Or

ORIGIN AND STRATIGRAPHIC SIGNIFICANCE OF KAOLINITIC SEDIMENTS FROM THE CYPRESSHEAD FORMATION: A SEDIMENTOLOGICAL, MINERALOGICAL AND GEOCHEMICAL INVESTIGATION

KENDALL BRIAN FOUNTAIN

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2009

"The Ridge is the Florida Divide, the peninsular watershed, and, to hear Floridians describe it, the world's most stupendous mountain range after the Himalayas and the Andes. Soaring two hundred and forty feet into the sub-tropical sky, the Ridge is difficult to distinguish from the surrounding lowlands, but it differs more in soil conditions than in altitude, and citrus trees cover it like a long streamer, sometimes as little as a mile and never more than twenty-five miles wide, running south, from Leesburg to Sebring, for roughly a hundred miles."

John McPhee, 1967

ACKNOWLEDGMENTS

I would like to acknowledge the financial support and in-kind service of the Clay Minerals

Society, the Graduate School of the University of Florida, the Florida Geological Survey, E.I. du

Pont de Nemours & Company, and Edgar Minerals (formerly Feldspar Corporation). Without their assistance, this research would have never been possible. I would also like to thank my committee, but in particular, Dr. Guerry H. McClellan, my committee chairman, for constantly challenging me during my time at the University of Florida and serving as my mentor both academically and professionally. Additionally, I would like to thank some of the other sources of academic, and sometimes personal, advice that impacted this research and my life as a whole.

They include, but are not limited to, Dr. J.L. Eades, Dr. T.M. Scott, Dr. V.J. Hurst, Dr. F.N.

Blanchard and Dr. W.G. Harris.

There were, of course, many people that assisted with the acquisition of data necessary to complete this research, or simply were a great source of conversation over a beer (or two). They include Dr. Michael (Mike) Rosenmeier, Dr. Phillip (Phil) Neuhoff, Dr. George Kamenov,

William Kenney, Dr. Jehangir (Jango) Bhadha, René Bohren, Dr. Richard (Rich) Hisert, Dr.

Todd Kincaid, Brickman (Bricky) Way, Christian George and Dr. Craig Oyen. Special recognition is afforded to George Kamenov for his assistance with the collection of neodymium isotopic data. Without his help, this portion of the study would have never been completed.

Additionally, a debt of gratitude is also owed to the staff of the Major Analytical Instrumentation

Center at the University of Florida (Wayne Acree in particular) for their assistance with collecting SEM data for this study, often at reduced rates and sometimes as bartered services.

Lastly, I would like to thank my family, for they are the ones that have always been there for me through this extended process, even when it seemed doubtful that I would ever finish. My geologist father has been a huge influence in my career, having introduced me to the wonder that

is this science at an early age, and helping to raise me with a strong work ethic and a sense of professionalism. On the other hand, it is my mother and brother who have been my “rocks” for many years, and have kept faith when I would fall off the path to finishing. They have also been

the ones that have offered moral and emotional support when it has been needed the most during trying times. Without them, I would not be the man I am today.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES...... 10

LIST OF FIGURES ...... 12

ABSTRACT...... 17

CHAPTER

1 INTRODUCTION ...... 20

Hypothesis ...... 22 Purpose and Scope...... 22 Geologic Setting ...... 24 Neogene Stratigraphic and Structural Framework ...... 24 Regional Physiography...... 25 Regional Stratigraphy...... 30 Miocene series...... 30 Pliocene series...... 33 Pleistocene series...... 38

2 REVIEW OF LITERATURE...... 39

Cypresshead Origin and Age...... 39 Facies Associations ...... 39 Paleoenvironment...... 41 Age Estimates...... 42 Correlative Siliciclastics ...... 43 Citronelle Formation ...... 43 Miccosukee...... 45 Kaolin Origin and Provenance...... 46 Georgia-South Carolina Kaolin District...... 46 Cypresshead Formation...... 50

3 METHODS...... 53

Sample Localities...... 53 Cypresshead Formation (Florida)...... 53 Cypresshead Formation (Georgia) ...... 58 Middle Georgia Kaolin District...... 58 Sample Preparation...... 59 Analytical Procedures...... 60 Grain-Size Analysis...... 60

Hydrometer and sieve analysis...... 61 SediGraph analysis...... 62 X-ray Diffraction Analysis...... 64 Oriented samples ...... 65 Random samples ...... 66 Petrographic/Scanning Electron Microscope Analysis ...... 67 Geochemical Analysis...... 68 Major and trace element analysis ...... 68 Nd isotopic analysis ...... 68

4 SEDIMENTOLOGICAL AND MINERALOGICAL EVIDENCE FOR THE ORIGIN AND STRATIGRAPHIC SIGNIFICANCE OF THE CYPRESSHEAD FORMATION .....73

Introduction...... 73 Related Deposits and the Siliciclastic Conveyor...... 74 Age Constraints on Cypresshead Deposition ...... 77 Pliocene Paleoclimate...... 78 Results...... 80 Outcrop and Core Descriptions ...... 80 Sedimentary Framework ...... 80 Grain-size distributions ...... 81 Sedimentary structures ...... 84 Mineralogy and Petrography ...... 91 Sand size-fraction...... 92 Clay size-fraction ...... 94 Facies Architecture...... 97 North-central Florida facies ...... 101 Southeastern Georgia facies ...... 104 Discussion...... 105 Depositional Environment...... 105 Cypresshead Formation model...... 106 Model consistency and the siliciclastic conveyor ...... 111 Timing and Regional Stratigraphic Correlation ...... 115 Pre-Cypresshead siliciclastic flux ...... 115 Cypresshead deposition and reworking...... 119 Paleoclimate Forcing of Cypresshead Deposition...... 124 Middle to late Miocene climate and sediment supply...... 125 Pliocene climate and the transition toward Northern Hemisphere Glaciation (NHG)...... 126 Evidence for increased current and storm activity ...... 127 Conclusions...... 129

5 EVIDENCE FOR NEOFORMATION AND RECRYSTALLIZATION OF KAOLINITE IN THE CYPRESSHEAD FORMATION...... 131

Introduction...... 131 Results...... 132

Mineralogy ...... 132 Kaolinite Disorder and Crystallite Size...... 137 Kaolinite Microtexture ...... 149 Particle-Size Analysis...... 153 Geochemistry...... 154 Major element data...... 155 Rare earth element (REE) data...... 157 Discussion...... 164 Kaolinite Origin...... 164 Kaolinite neoformation ...... 165 Kaolinite recrystallization and disorder ...... 170 Accessory Phase Paragenesis ...... 172 Conclusions...... 175

6 TRACE ELEMENT AND Nd ISOTOPIC EVIDENCE FOR THE PROVENANCE OF CYPRESSHEAD FORMATION KAOLINITIC SANDS...... 177

Introduction...... 177 Southern Piedmont ...... 179 Georgia-South Carolina Kaolin District...... 180 Results...... 182 Trace Elements ...... 182 Neodymium (Nd) Isotopes ...... 185 Discussion...... 190 Trace Element Mobility and Enrichment ...... 191 Comparison to Georgia-South Carolina Kaolin Provenance...... 193 Cypresshead Provenance...... 194 Conclusions...... 199

7 SUMMARY AND CONCLUSIONS...... 201

Cypresshead Formation Stratigraphy ...... 201 Cypresshead Formation Mineralogy...... 202 Cypresshead Formation Provenance...... 203

APPENDIX

A MINE SITE MAPS...... 204

B ANALYTICAL PROCEDURES FOR ICP-AES, ICP-MS, AND MC-ICPMS ...... 208

C STRATIGRAPHIC SECTIONS...... 215

D GRAIN-SIZE DATA (HYDROMETER AND SIEVE) ...... 223

E X-RAY DIFFRACTION DATA (ORIENTED) ...... 233

F X-RAY DIFFRACTION DATA (RANDOM)...... 249

G MINUS-200 MESH PARTICLE-SIZE DATA (SEDIGRAPH)...... 266

H MAJOR AND TRACE ELEMENT DATA ...... 278

LIST OF REFERENCES...... 283

BIOGRAPHICAL SKETCH ...... 306

LIST OF TABLES

Table page

1-1 Southern Atlantic Coastal Plain terraces (Florida, Georgia, and South Carolina)...... 28

3-1 Sample list and corresponding analyses performed for this study...... 55

3-2 MC-ICP-MS analyses of the Ames Nd in-house standard...... 70

3-3 MC-ICP-MS analyses of the JNdi-1, LaJolla Nd, and BCR-1 standards...... 71

3-4 TIMS analyses of the Ames Nd in-house standard...... 71

4-1 Summary of the lithostratigraphic and sequence stratigraphic nomenclature applied to South Florida siliciclastics...... 75

4-2 Clay size-fraction mineralogy of Cypresshead Formation and reworked Cypresshead Formation sediments...... 95

4-3 Facies summary of the Cypresshead Formation in north-central peninsular Florida and southeastern Georgia...... 98

4-4 Correlation of siliciclastic depositional events on the Florida Platform with Haq et al. (1988) sequence boundaries, sequence boundaries of Eberli (2000), and sea-level falls identified by Miller et al. (2005)...... 120

5-1 Results of disorder calculations for north-central Florida Cypresshead and reworked Cypresshead kaolinite...... 138

5-2 Statistical summary of kaolinite order and crystallite size calculations...... 141

5-3 Results of crystallite size calculations for north-central Florida and Georgia Cypresshead and reworked Cypresshead kaolinite...... 145

5-4 Major element concentrations as oxides for Cypresshead Formation clay (< 2 µm) samples...... 156

5-5 Correlation matrix of major elements and ∑REE for both Florida and Georgia Cypresshead samples...... 157

5-6 REE concentrations of Cypresshead Formation and related samples...... 159

6-1 Trace element concentrations (ppm) and elemental ratios for Cypresshead Formation and comparison samples...... 183

6-2 Correlation matrix of select trace elements and elemental ratios for all Cypresshead Formation samples...... 186

6-3 Nd isotope data for Cypresshead Formation and comparison samples...... 187

D-1 Grain-size distributions and moment statistics for the Grandin and Goldhead sand mines in north-central Florida...... 224

D-2 Grain-size distributions and moment statistics for the EPK kaolin mine in north- central Florida...... 225

D-3 Grain-size distributions and moment statistics for the Davenport and Joshua sand mines in central Florida...... 225

D-4 Grain-size distributions and moment statistics for Cypresshead Formation sampling locations in southeastern Georgia...... 226

H-1 Raw major element concentrations for samples used in this study...... 279

H-2 Minor element concentrations for samples used in this study...... 281

LIST OF FIGURES

Figure page

1-1 Major structural features of Florida and southern Georgia influencing Neogene to Holocene sedimentation...... 26

1-2 Pleistocene marine terraces and shorelines of Florida and Georgia...... 27

1-3 Regional stratigraphic correlation chart for Florida and southeast Georgia...... 31

3-2 Sample processing and analysis flow chart...... 60

3-3 SediGraph concentration test results of sample FRL-1-9 at concentrations of 1.8 g, 2.0 g, 2.2 g and 2.4 g mixed with 70 ml of dispersant solution...... 63

3-4 SediGraph precision test results of sample EPK36-J-12 (56-59) comparing two replicate samples...... 64

4-1 Histogram illustrating the distribution of mean grain-size values for samples of the Cypresshead Formation and reworked Cypresshead sediments...... 82

4-2 Representative grain-size distribution curves for Cypresshead Formation sediments...... 83

4-3 Scatter plot illustrating the relationship between mean grain-size and skewness for samples of Cypresshead Formation and reworked Cypresshead sediments...... 84

4-4 Example sedimentary structures from the Cypresshead Formation...... 85

4-5 Paleocurent rose diagrams for representative Cypresshead Formation exposures...... 88

4-6 Examples of trace fossil and bivalve mold occurrences...... 90

4-7 Photomicrographs of accessory sand-sized phases from the Cypresshead Formation...... 93

4-8 Correlation of Cypresshead Formation facies in north-central Florida...... 99

4-9 Correlation of Cypresshead Formation facies in southeastern Georgia...... 100

4-10 Examples of Cypresshead Formation facies...... 103

4-11 Correlation chart for Late Miocene through Pliocene siliciclastic units evaluated in this study...... 116

4-12 Maps illustrating the aerial extent of successive siliciclastic deposition events impacting the Florida Platform from 8.6 Ma to 1.8 Ma...... 117

4-13 Relationship between the Cypresshead Formation, related siliciclastics in southern Florida, seismic sequence boundaries of Eberli (2000), sequence chronostratigraphy, and the sea-level curves of Haq et al. (1988) and Miller et al. (2005)...... 118

5-1 XRD patterns of example Cypresshead Formation clays...... 134

5-2 SEM photomicrographs illustrating characteristic secondary weathering phases and textures...... 135

5-3 Histograms illustrating the distribution of results for both kaolinite order (HI and R2) and crystallite size calculations...... 142

5-4 Box-and-whisker diagrams for kaolinite disorder (HI and R2) and crystallite size calculations...... 143

5-5 Scatterplots illustrating the positive correlation between the Hinkley Index (HI) and measured CSD values for north-central Florida Cypresshead (and reworked Cypresshead) samples...... 149

5-6 Measured CSD distribution curves and fitted theoretical lognormal curves (red) for select EPK36-J-12 samples...... 150

5-7 Example CSD distributions illustrating changing curve shape with increased depth from the surface...... 151

5-8 SEM photomicrographs illustrating the microtextural variation noted in Cypresshead and reworked Cypresshead Formation sediments...... 152

5-9 Example SediGraph particle-size distributions for select Cypresshead Formation samples...... 154

5-10 Scatterplots illustrating mixing trends related to the presence of crandallite-florencite series minerals in Cypresshead Formation clays...... 158

5-11 Chondrite-normalized REE distribution patterns for select Cypresshead Formation clay (< 2 µm) fraction and related samples...... 161

5-12 Histogram illustrating the distribution of Eu/Eu* values for both Georgia and Florida Cypresshead Formation clays...... 164

5-13 SEM photomicrographs of Cypresshead and reworked Cypresshead Formation kaolinite textures associated with the in situ weathering of muscovite...... 167

5-14 Correlation of Eu/Eu* values to CSD (volume-weighted mean thickness) calculations...... 173

6-1 Tectonostratigraphic terranes and granites proposed as potential source materials for Cypresshead Formation sediments...... 181

6-2 Multi-element normalized diagrams for Cypresshead and comparison samples, normalized against average continental crust...... 184

6-3 143Nd/144Nd versus 147Sm/144Nd isochron diagram for Cypresshead Formation samples...... 189

6-4 εNd(t) versus Nd concentration scatterplot for Cypresshead Formation samples...... 189

6-5 Histogram illustrating the distribution of Nd model age (TDM) results for both Cypresshead Formation clay samples and comparison formations...... 190

6-6 Trace element plots of Cypresshead Formation clays illustrating evidence of weathering, provenance, and sediment recycling processes...... 192

6-7 Comparison of Cypresshead Formation trace element concentrations to the results of Dombrowski (1992; 1993)...... 195

6-8 Plot of εNd(t) versus Th/Sc for Cypresshead Formation samples based on the model of McLennan et al. (1990; 1993)...... 197

6-9 Plot of εNd(t) versus stratigraphic age for the Cypresshead Formation compared to the Nd isotopic evolution of potential sources...... 197

A-1 Site map for the Edgar Minerals EPK Mine...... 205

A-2 Site map for the VMC Goldhead Sand Mine...... 205

A-3 Site map for the VMC Grandin Sand Mine...... 206

A-4 Site map for the CEMEX Davenport Sand Mine...... 206

A-5 Site map for the CEMEX Joshua Sand Mine...... 207

C-1 Stratigraphic section for Grandin Sand Mine section FRG-1...... 216

C-2 Stratigraphic section for Grandin Sand Mine section FRG-2...... 217

C-3 Stratigraphic section for Goldhead Sand Mine section FRL-1...... 218

C-4 Stratigraphic section for Joshua Sand Mine core SSJ-1...... 219

C-5 Stratigraphic section for Davenport Sand Mine core SSD-1...... 220

C-6 Stratigraphic section for Jesup section J-1...... 221

C-7 Stratigraphic section for Birds section B-1...... 221

C-8 Stratigraphic section for Linden Bluff section L-1...... 222

D-1 Grain-size distribution curves for EPK Mine core EPK36-J-12...... 227

D-2 Grain-size distribution curves for EPK Mine core EPK31-P-40...... 227

D-3 Grain-size distribution curves for EPK Mine core EPK30-V-6...... 228

D-4 Grain-size distribution curves for Grandin Sand Mine section FGR-1...... 228

D-5 Grain-size distribution curves for Grandin Sand Mine section FRG-2...... 229

D-6 Grain-size distribution curves for Goldhead Sand Mine section FRL-1...... 229

D-7 Grain-size distribution curves for Joshua Sand Mine core SSJ-1...... 230

D-8 Grain-size distribution curves for Davenport Sand Mine core SSD-1...... 230

D-9 Grain-size distribution curves for Jesup section J-1...... 231

D-10 Grain-size distribution curves for Linden Bluff section L-1...... 231

D-11 Grain-size distribution curves for Birds section B-1...... 232

G-1 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK36-J-12)...... 267

G-2 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK31-P-40)...... 268

G-3 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK30-V-6)...... 269

G-4 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Grandin Sand Mine samples (FRG-1)...... 270

G-5 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Grandin Sand Mine samples (FRG-2)...... 271

G-6 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Goldhead Sand Mine samples (FRL-1)...... 272

G-7 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Joshua Sand Mine samples (SSJ-1)...... 273

G-8 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Davenport Sand Mine samples (SSD-1)...... 274

G-9 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Jesup type locality samples (J-1)...... 275

G-10 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Linden Bluff reference locality samples (L-1)...... 276

G-11 SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Birds reference locality samples (B-1)...... 277

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

ORIGIN AND STRATIGRAPHIC SIGNIFICANCE OF KAOLINITIC SEDIMENTS FROM THE CYPRESSHEAD FORMATION: A SEDIMENTALOGICAL, MINERALOGICAL AND GEOCHEMICAL INVESTIGATION

Kendall Brian Fountain

December 2009

Chair: Guerry H. McClellan Major: Geology

Kaolinitic sediments of the Cypresshead Formation (3.4–2.3 Ma), deposited as two distinct shoreface-shelf parasequences in response to sea level falls at 3.3 Ma and 2.5 Ma, define the final stages of a period of siliciclastic deposition which dominated the Florida Platform between

8.6 Ma and 1.8 Ma. Beginning with deposition of the Late Miocene SS2 siliciclastics of

Cunningham et al. (2003) and ending with the Late Pliocene cessation of Cypresshead deposition and reworking at approximately 1.8 Ma, episodes of sediment accumulation correlate with two paleoclimatic transitions from; (1) arid conditions during the Late Miocene to continual El Niño conditions during the early Pliocene warm period (~4.5–3.0 Ma), and (2) continual El Niño conditions to global cooling and the onset of significant Northern Hemisphere Glaciation (NHG)

(~3.0–1.5 Ma). Responding to the interplay of sea-level, sediment supply and accommodation associated with climate instability, the siliciclastics deposited during this Late Miocene through

Pliocene interval define a retrogradational parasequence set episodically deposited on the Florida platform over a 6.8 Ma period.

Facies identified within the Cypresshead Formation define the depositional environment as nearshore marine, consistent with a wave-dominated coastline in north-central Florida and a

mixed energy coastline in southeastern Georgia. Composed of coarsening-upward sequences, both facies sets define a progradational response to the coastal delivery and subsequent coast- parallel transportation of a substantial siliciclastic flux, which, in turn, correlates with the nonequilibrium landscape conditions predicted by a transitional mid-Pliocene paleoclimate. A potential fluvial-deltaic component of the Cypresshead Formation in southeastern Georgia is proposed based on paleocurrent indicators and the similarities noted for the offshore transition

(OST) facies in Georgia and the prodeltaic Miccosukee Formation.

Mineralogically, Cypresshead Formation sediments are highly weathered, but particularly in north-central Florida where local hydrologic conditions coupled with the depositional fabric and microtexture of the sediments have resulted in the formation of at least three kaolinite fractions; (1) in situ kaolinite formed at the expense of feldspars and mica, (2) detrital kaolinite deposited as part of the original clay mineral suite, and (3) near surface recrystallized kaolinite.

In situ kaolinite crystallizes via the combined topotactic (transformation) and epitactic

(neoformation) weathering of muscovite mica under saturated conditions to produce an enrichment of vermicular kaolinite in Cypresshead sediments. Kaolinite formed in this way exhibits a high degree of order in response to the epitactic nucleation of crystallites on the structurally similar muscovite surface, explaining the correlation between low disorder and small

particle-size that has confounded researchers for decades. Additionally, the feldspar dissolution

sourcing of the Al and Si necessary for in situ kaolinite crystallization is confirmed by inherited

positive Eu anomalies reported for these kaolinites and the presence (and geochemistry) of

residual feldspars in basal Cypresshead sediments.

Weathering occurring under oxic, vadose or mixed vadose/saturated conditions results in

the degradation of neoformed kaolinite as evidenced by a decrease in kaolinite order and

coherent scattering domain (CSD) values consistent with the formation of the near surface

recrystallized kaolinite fraction. Forming under conditions of extreme leaching and recystallization, this kaolinite fraction possesses microtextural characteristics suggestive of a

pedogenic origin. The trace phases gibbsite, halloysite and crandallite-florencite also form under similar near surface, oxic weathering conditions. The origin of the crandallite-florencite phase is uncertain, but most likely originated from the decomposition of post-depositional pore water organics coupled with detrital mineral dissolution.

Both trace elements and Nd isotopes were analyzed in order to constrain the provenance and broader geochemical characteristics of the Cypresshead Formation and to compare these results with the studies by Dombrowski (1992; 1993) on the provenance of Cretaceous and

Tertiary kaolins from the Georgia-South Carolina kaolin district. Of the trace elements used by

Dombrowski, only Th and Sc appear to be resistant to significant mobilization and depletion in

Cypresshead sediments, with the Th/Sc ratio a robust indicator of provenance composition.

Additionally, a significant proportion of Florida Cypresshead clays possess Th/U ratios well below that of average upper crust, and appear to have undergone redox-driven U enrichment dictated by pore water organics. As for Nd results, Cypresshead Formation samples appear to originate from sources intermediate between those associated with Cretaceous “soft” kaolins and

Tertiary “hard” kaolins of Georgia and South Carolina. This is consistent with these sediments being a mixture of materials originating from Carolina terrane and Alleghanian granite sources.

Additionally, Nd model ages (TDM) for Florida and Georgia Cypresshead samples range between

1.4-0.9 Ga and 1.5-0.8 Ga respectively, with both possessing an average TDM value of 1.1 Ga

consistent with the age of Grenville crust.

CHAPTER 1 INTRODUCTION

Kaolin deposits in the southeastern United States are important economically as an industrial mineral and geologically as an indicator of past climatic and depositional conditions.

As a result, the Cretaceous and Tertiary deposits of the Georgia-South Carolina kaolin district, one of the largest and most valuable coastal plain kaolin deposits in the world (Patterson and

Murray, 1984), have been the focus of extensive research. In comparison, the kaolinitic sands of the Cypresshead Formation, actively exploited for kaolin at the Edgar Minerals EPK Mine in north-central Florida, have received only limited study due in part to their minor economic value and poorly known geologic history.

From an industrial perspective, kaolinite, the essential clay mineral component in kaolin

(or china clay), ball clay, fire clay (or refractory clay), and flint clay, possesses an assortment of unique properties which make it suitable for a variety of applications. In fact, the kaolin-bearing sediments of the Georgia-South Carolina kaolin district constitute one of the world’s leading sources of high quality kaolin, representing a nearly $1 billion/year industry. These high quality

Coastal Plain kaolins are exploited due to superior properties suitable to film formation necessary in paper coating, by far the most economically important application of kaolin. This use commands the most stringent specifications, defined by high brightness, suitable particle size, and low abrasion. Other kaolin applications include paint, ink, fiber extension, polymer extension and reinforcement, catalysts, fiber glass, and use as a carrier, adsorbent, or diluent

(Bundy, 1993). Kaolin associated with the Edgar Minerals EPK deposit lacks the properties suitable for use in the paper industry, but possesses unusually good forming characteristics and high green strength which make it a superior product for use in the ceramics industry.

Of particular significance to this study is the uncertainty that exists as to the origin of the kaolin within the Cypresshead Formation, the origin of the unit itself, and the importance of the unit within the context of Florida Platform and broader Coastal Plain deposition. In particular, confusion persists as to whether the kaolin is primary, having formed in situ through post- depositional alteration of detrital feldspars, micas, and/or primary clays, or secondary, having formed elsewhere only to be eroded, transported, and deposited as the kaolinite-rich siliciclastics now observed (Sellards, 1912; Bell, 1924; Pirkle, 1960). Of further interest is the significance of the general low disorder/fine particle size properties of kaolin from this unit. First noted by

Fountain and McClellan (1993), this characteristic is inconsistent with the existing models of

Georgia-South Carolina kaolin formation, for which low defect/coarse particle size and high defect/fine particle size are the rule (Pickering and Hurst, 1989; Pickering et al., 1997). In order to answer these questions, this study seeks to define the depositional framework, provenance, mineralogy, and weathering history of the kaolinitic sediments associated with the Cypresshead

Formation, and to constrain this deposit within a regional context of siliciclastic deposition.

This dissertation has been organized into three separate papers for future publication. The first of these papers, entitled “Sedimentological and mineralogical evidence for the origin and stratigraphic significance of the Cypresshead Formation” defines the regional stratigraphy and sedimentology of the Cypresshead Formation, and correlates the unit with genetically related siliciclastics in order to define a sequence stratigraphic model for Late Miocene to Pliocene deposition on the Florida Platform. The second paper, entitled “Evidence for neoformation and recrystallization of kaolinite in the Cypresshead Formation” defines the recrystallization processes impacting the mineralogy and industrial properties of Cypresshead Formation kaolinite, and answers the primary verses secondary origin issue associated with these clays.

Lastly, the third paper, entitled “Trace element and Nd isotopic evidence for the provenance of

Cypresshead Formation kaolinitic sands” identifies the ultimate provenance of Cypresshead

Formation kaolinite, addressing the role of sediment sources on kaolin physical properties.

Hypothesis

The hypothesis to be tested in this study is that the Cypresshead Formation in north-central

Florida and eastern Georgia fits into a regional model of episodic kaolin deposition on the southern Atlantic coastal plain, which began in the Cretaceous and continued through the

Pliocene under specific environmental conditions favorable for formation and subsequent preservation of kaolinitic sediments. As part of this model, it is proposed that the kaolinite contained within the Cypresshead Formation possesses both primary and secondary characteristics consistent with deposition as part of an episodic siliciclastic flux, followed by post-depositional weathering and recrystallization. It is hypothesized that: (1) precursor materials eroded from source areas located within the Piedmont-Blue Ridge region, the Georgia-South

Carolina kaolin district, and/or the southeastern Georgia Coastal Plain are responsible for the bulk of kaolinite accumulation in Cypresshead Formation sediments, (2) subsequent weathering is responsible for increased kaolinite content and a substantial modification of the original fabric and mineralogy of the kaolinitic sediments, and (3) the Cypresshead Formation is part of a genetically related sequence of siliciclastics deposited episodically during the Late Miocene to

Pliocene, which formed in response to a favorable interplay of paleoclimate, weathering and sea- level fluctuations.

Purpose and Scope

The ultimate industrial potential of a kaolin deposit is governed, in part, by the geological conditions operating both before and after deposition. As a consequence, this study seeks to define the geological factors influencing Pliocene siliciclastic deposition and subsequent

weathering on the Florida Platform and southeastern Georgia Coastal Plain using a combination of stratigraphic, mineralogical and geochemical techniques. As part of this effort, the stratigraphic framework for the Cypresshead Formation will be evaluated in light of recent studies which have added significant insight into the deposition of related Late Miocene through

Pliocene siliciclastics in southern Florida. To accomplish this, both sedimentological and mineralogical indicators of Cypresshead Formation timing and deposition were reassessed through field-based observation and measurement, and through a review of previous studies on the unit and its correlatives. Because siliciclastic sedimentation has dominated much of the southeastern United States episodically since the Cretaceous, it is vital to understand sedimentation patterns and the changing nature of clastic sediment sources. By defining the genetic characteristics of these deposits, the processes influencing both the origin and subsequent modification of kaolin-bearing units are better understood.

Additionally, the integration of geochemical techniques and detailed clay mineralogy of kaolin deposits as an effective method for provenance determination was evaluated. This approach has not been previously employed in the study of Neogene Coastal Plain sedimentation, and was undertaken in response to previous studies of Cypresshead Formation sediments using traditional methods of provenance analysis (paleocurrent indicators, heavy mineral analysis, and standard petrography), which have resulted in ambiguous results and conflicting interpretations (Sellards, 1912; Bell, 1924; Pirkle, 1960; Pirkle et al., 1964; Kane,

1984). To this end, trace elements, particularly rare earth elements (REEs), and neodymium (Nd) isotopic systematics were employed to constrain kaolinite provenance. In general, geochemical approaches to sedimentary provenance of fine-grained sediments are more useful than traditional petrographic techniques due to limitations associated with particle size (McLennan et al., 1993).

Geologic Setting

Neogene Stratigraphic and Structural Framework

The Florida Peninsula and portions of southeastern Georgia are underlain by as much as

1220 m of sedimentary rocks. This sedimentary sequence of predominately carbonate units

comprises the broad, relatively flat Florida Platform, of which, the Florida Peninsula, veneered

by a relatively thin sequence of predominately siliciclastic sediments, is the emergent portion

(Scott et al., 1980).

Beginning in the Oligocene, siliciclastics, including quartz sands, silts, and clays, began to supplant carbonate deposition on the Florida Platform in response to changes in both the structural controls on deposition and sediment sources (Scott, 1992a; Cunningham et al., 2003).

Prior to that time, from the Middle Cretaceous to the Late Paleogene, siliciclastic sediment supply from Appalachian sources to the north had been restricted by the Georgia Channel

System (Huddlestun, 1993; includes the Gulf Trough and Suwannee Straits), effectively isolating the carbonate platform from a siliciclastic influx (Scott, 1992a; Cunningham et al., 2003).

However, during the Late Eocene through Oligocene, uplift and/or crustal arching associated with rejuvenation of the Appalachians (Stuckey, 1965; Dennison and Stewart, 2001; Stewart and

Dennison, 2006), along with weakening of the current through the Georgia Channel System

(Huddlestun, 1993), led to a renewed supply of siliciclastics flooding the southeastern North

American coastline. As a result, a major influx of sediments eventually filled the channel, blocking the current from reoccupying the Georgia Channel System, and permitted encroachment of the siliciclastic wedge onto the Florida Platform, and by the Late Miocene, siliciclastic deposition dominated the Florida Platform, particularly in the north and central peninsular areas (Scott, 1992a; 1997; Cunningham et al., 2003). The present highlands of the northern peninsula and panhandle of Florida are the dissected remains of the fluvial, deltaic, and

shallow-water marine deposits associated with this Neogene siliciclastic influx (Schmidt, 1997).

Ultimately, these sediments were reworked and/or reshaped by subsequent sea-level fluctuations

and associated nearshore, coast-parallel currents into the elongate system of upland ridges seen

today (Schmidt, 1997).

The Neogene structural setting of the Florida Peninsula and the southeastern Georgia

Coastal Plain is shown in Figure 1-1. In the study area, Neogene to Holocene sediments unconformably overlie an eroded and karstified Paleogene surface characterized by a multitude of structural highs and basins (Scott, 1997). Among the highs are the Ocala Platform (or

“uplift”), the Chattahoochee Anticline, the Sanford High, the St. Johns Platform, and the Brevard

Platform. Among the basins are the Georgia Channel System (comprising the Apalachicola

Embayment, Gulf Trough, and Southeast Georgia Embayment), the Osceola Low, and the

Okeechobee Basin. These features variably affected Neogene sedimentation, in part, directing siliciclastic fluxes on the Florida Platform. The reader is referred to Huddlestun (1988) and Scott

(1988b) for a detailed review of the structural framework of Florida and southeastern Georgia.

Regional Physiography

Some of the most distinguishable geomorphic features of peninsular Florida and the

coastal plain of Georgia are the marine terraces and ancient shorelines generated by sea-level

fluctuations that drastically altered the physiography of the region. Six to eight coastal terraces of

Pleistocene age approximately parallel the Atlantic coast of Georgia and Florida (Fig. 1-2, Table

1-1). Historically, various authors have assigned different elevations and terminology to these

terraces, based mainly on physiographic evidence. These terraces were first identified by Cooke

(1930; 1945),who along with others (Hoyt and Hails, 1974), held that each terrace was, in fact, a

formation consisting of a barrier island facies on the seaward side and a marsh-lagoon facies on

the landward side. This was later deemed to be an incorrect observation by Huddlestun (1988),

Figure 1-1. Major structural features of Florida and southern Georgia influencing Neogene to Holocene sedimentation (modified after Scott, 1997).

who noted that instead of being stratigraphic units, the marine terraces are simply geomorphic features. As a result, the use of terrace terminology in a lithostratigraphic context has ceased by most authors because there is no genetic relationship between any single terrace surface and any single lithostratigraphic unit along the Florida and Georgia coastal plains.

Figure 1-2. Pleistocene marine terraces and shorelines of Florida and Georgia (modified after Healy, 1975; Huddlestun, 1988).

Table 1-1. Southern Atlantic Coastal Plain terraces (Florida, Georgia, and South Carolina) (modified after Pirkle et al., 1970; Healy, 1975).

Elevation above msl Terrace Age (ft) (m)

After Cooke (1939; 1945) Hazlehurst/Brandywine 270 82.3 Aftonian Coharie 215 65.5 Yarmouth Sunderland 170 51.8 Yarmouth Wicomico 100 30.5 Sangamon Penholoway 70 21.3 Sangamon Talbot 42 12.8 Sangamon Pamlico 25 7.6 Wisconsin After MacNeil (1950) High Pliocene terrace 280 85.4 Aftonian Okefenokee 150 45.7 Yarmouth Wicomico 100 30.5 Sangamon Pamlico 25-35 7.6-10.7 Wisconsin Silver Bluff 8-10 2.4-3 Recent Hoyt and Hails (1974) Wicomico 95-100 29-30.5 Pleistocene Penholoway 70-75 21.3-22.9 Pleistocene Talbot 40-45 12.2-13.7 Pleistocene Pamlico 24 7.3 Pleistocene Princess Anne 13 4 Pleistocene Silver Bluff 4.5 1.4 Pleistocene After Healy (1975) Hazlehurst 215-320 65.5-97.6 Miocene or Pliocene Coharie 170-215 51.8-65.5 Pleistocene Sunderland/Okefenokee 100-170 30.5-51.8 Pleistocene Wicomico 70-100 21.3-30.5 Pleistocene Penholoway 42-70 12.8-21.3 Pleistocene Talbot 25-42 7.6-12.8 Pleistocene Pamlico 8-25 2.4-7.6 Pleistocene Silver Bluff 1-10 0.3-3 Pleistocene

Regional correlation of the marine terraces has been the goal of many authors. However, as noted by Winker and Howard (1977a; b), the terraces in the Carolinas and northern Florida have been uplifted, confusing correlations made on the basis of elevation alone. Opdyke et al. (1984) have explained that the uplift in Florida is due to isostatic rebound accompanying solution of limestone in the karst regions of central Florida. Based on the measurement of dissolved solids in

Florida’s springs, Opdyke et al. (1984) estimated that the uplift in the north-central Florida

peninsula has been at least 36 m during the Pleistocene and Holocene, in agreement with observations of marine fossil occurrences associated with the terraces.

The distribution of marine terrace sediments in peninsular Florida and southeastern

Georgia is dominated by a ridge morphology. The highest of these ridges is Trail Ridge (Fig. 1-

2), located on the western flank of the Duval Upland, an area underlain by Cypresshead

Formation sediments. It is believed to be a relict barrier or spit corresponding to a Wicomico shoreline (29–31 m) that may have formed at the crest of a marine transgression from erosion of the Northern Highlands located further to the west (Pirkle et al., 1974). The age of Trail Ridge has long been debated, with the discovery of shallow marine fossils in Trail Ridge indicating an age no older than late Pliocene or Pleistocene (Pirkle and Czel, 1983).

The Penholoway shoreline marks the eastern limit of the Cypresshead Formation exposed at the surface in eastern Georgia and northern Florida, and is bounded by a 9–15 m scarp along its eastern edge. The scarp likely indicates a period of sea-level transgression responsible for the present distribution and ridge morphology of Cypresshead Formation sediments in north-central

Florida and eastern Georgia. Although eroded to form a younger shoreline than the Wicomico,

Cypresshead Formation sediments are stratigraphically older than Trail Ridge.

The Lake Wales Ridge (Fig. 1-2) is a distinct topographic feature running roughly north-

south for more than 100 mi (161 km) along the center of north-central peninsular Florida

(White, 1958; White, 1970; Scott, 1980). This feature contains some of the highest elevations in

Florida, rising over 320 ft (97.6 m) above sea-level at Iron Mountain in Polk County, and contains the thickest sections of Cypresshead sediments in the state. It is a well defined topographic high from the central part of Lake County southward to its terminus near Lake

Placid, in Highlands County, but is not well defined in its northern part, most likely as a result of

Pleistocene erosion and karstification (Pirkle, 1960; White, 1970). The Winter Haven and

Lakeland ridges in Polk County, Florida, are, like the Lake Wales Ridge, underlain by

Cypresshead Formation sediments, and interpreted to be erosional features produced during

Pleistocene sea-level highstands. The northern remnant of the Lake Wales Ridge, located in the area of the Interlachen Karstic Highlands of Arrington (1985), is flanked on the west and north by Trail Ridge, and on the east by the Baywood Promontory (Pirkle et al., 1970) (Fig. 1-2). As the base of Trail Ridge sediments are located at 150 ft (45.7 m) msl, it is likely that the seas which deposited the sediments of this ridge also inundated the northern extension of the Lake

Wales Ridge, thereby destroying, via wave planation, the ridge morphology observed in central peninsular Florida.

Regional Stratigraphy

The Neogene stratigraphy of southeastern Georgia and peninsular Florida records a major transition toward episodic siliciclastic deposition in response to fluctuating sea-level and sediment sources. The timing and nomenclature assigned to the various units that record this history have long been a source of controversy among earlier workers (Scott, 1992b). Within the region included with this study, the presently accepted stratigraphic nomenclature is outlined in

Figure 1-3.

Miocene series

Hawthorn Group: northern peninsular Florida and southeastern Georgia. From northern peninsular Florida to southeastern Georgia, Hawthorn Group sediments, in general, disconformably underlie the Cypresshead Formation. The Hawthorn Group is easily distinguished from the overlying Cypresshead Formation in being typically thick-bedded and massive, commonly phosphatic (except where it grades into the Altamaha Formation), argillaceous, and locally dolomitic, calcareous, and siliceous (Huddlestun, 1988). Three

Figure 1-3. Regional stratigraphic correlation chart for Florida and southeast Georgia (modified after Huddlestun, 1988; Scott, 1992a; 2001; Scott et al., 2001).

Hawthorn Group formations share disconformable contacts with the overlying Cypresshead

Formation in northern peninsular Florida and southeastern Georgia. These units are, from oldest to youngest, the Parachucla Formation, the Marks Head Formation, and the Coosawhatchie

Formation. The reader is referred to Huddlestun (1988) and Scott (1988b) for a more detailed review of these units, including type localities, lithologies, and stratigraphic relationships.

Hawthorn Group: central and southern peninsular Florida. In central and southern peninsular Florida, the Cypresshead Formation disconformably overlies either undifferentiated

Hawthorn Group sediments or those of the Peace River Formation of the Hawthorn Group.

Undifferentiated Hawthorn Group sediments are limited to the vicinity of Lake County in central

Florida where the transition between northern Florida Coosawhatchie Formation and central

Florida Peace River Formation lithologies occurs (Scott, 1988b). As for the middle Miocene to

early Pliocene Peace River Formation of Scott (1988b), lithologies consist of interbedded quartz sands, clays, and carbonates, with the dominant (two-thirds or more) siliciclastic component distinguishing this unit. The very fine- to medium-grained quartz sands are characteristically clayey, calcareous to dolomitic, and variably phosphatic, with phosphate concentrations greatest in the Bone Valley Member of the unit. Clay beds are common in the Peace River Formation, and tend to be dominated by smectite and palygorskite (Reynolds, 1962; McClellan and Van

Kauwenbergh, 1990; Scott, 1988b). Scott (1988b) correlates the lower part of the Peace River

Formation with the Coosawhatchie and Statenville formations in northern Florida, based on stratigraphic position, diatoms of middle Miocene age, and the occurrence of an early to middle

Barstovian age vertebrate fauna (Webb and Crissinger, 1983) at the base of the unit. The reader is referred to Scott (1988b) for a more detailed review of the Peace River Formation, including type locality, lithology, and stratigraphic relationships.

Altamaha Formation. The Altamaha Formation of Huddlestun (1988), the principle

Miocene fluvial lithofacies found in Georgia, is the most widely occurring outcropping formation in the Georgia Coastal Plain (Huddlestun, 1993). The blanket deposit of kaolinitic sands and sandy kaolins of this formation are consistent with deposition by braided streams disgorging their sediment loads from the Piedmont (Huddlestun, 1988; 1993). Lithofacies include both overbank and channel-fill deposits consisting of variably indurated to nonindurated, pebbly, feldspathic, kaolinitic sands and sandy kaolins that are interpreted to have been generated by braided streams depositing their sediment loads from the Piedmont onto the coastal plain (Huddlestun, 1988). As noted, the clay mineral suite of the Altamaha Formation is dominated by kaolinite with illite and smectite as minor constituents.

In some regions of Georgia, the Altamaha Formation is divisible into an upper and lower

part (Huddlestun, 1988). The lower part is typically thick bedded, massive, sandy clays and

argillaceous sands to claystones and sandstones, while the upper part is typically a prominently

cross-bedded, pebbly to gravelly sand with clay lenses. The latter of these units, assigned by

Huddlestun (1988) to the Screven Member, is particularly well developed in the Altamaha and

Satilla River area, and consists of a maze of fluvial channels and cut-and-fill structures with corresponding channel-fill lithologies. Given an average thickness of between 100 and 200 feet

(30-60 m) (Huddlestun, 1988), and its widespread occurrence in southeastern Georgia, the

Altamaha Formation represents a large reservoir of siliciclastics similar in character to the

Pliocene siliciclastics of the Cypresshead Formation. The reader is referred to Huddlestun (1988) for a more detailed review of the Altamaha Formation, including type locality, lithology, and stratigraphic relationships.

Pliocene series

Cypresshead Formation. The name Cypresshead Formation was first used by Huddlestun

(1988) to describe “a prominently thin- to thick-bedded and massive, planar- to cross-bedded, variably burrowed and bioturbated, fine-grained to pebbly, coarse-grained sand formation in the terrace region of eastern Georgia”. Subsequently, the name was extended into Florida by Scott

(1988a) to encompass sediments in peninsular Florida previously assigned to the Citronelle

Formation of Matson (1916).

As defined by Huddlestun (1988) and Scott (1988a), the Late Pliocene (early Piacenzian) to early Pleistocene (Calabrian) Cypresshead Formation is composed entirely of siliciclastics; predominately quartz and clay minerals, with quartz and/or quartzite pebbles locally abundant

(Pirkle et al., 1970). The unit is characteristically a mottled, fine- to very coarse-grained, often gravelly, variably clayey quartz sand, containing minor amounts of feldspar, mica and heavy

minerals (Scott, 1988a). Sediments vary from poorly- to well-sorted and angular to subrounded, with induration generally poor to nonindurated. The binding matrix or cementing agent is normally clay, although iron oxide cement is known to occur. In areas where the Cypresshead outcrops, the sediments are characteristically oxidized and mottled, exhibiting shades of red, orange, and white (Scott, 1988a).

As noted, clays are present throughout the Cypresshead Formation as a binding agent and occasionally as a primary lithology. Clay content of the sediments appears to decrease in a general north to south trend with higher average clay contents in southern Georgia than in northern Florida. The clay mineral present in the oxidized, mottled portion is characteristically kaolinite while in the downdip unoxidized portion illite and smectite are reported to dominate

(Scott, 1988a; 1992a). Although not yet recognized in Florida, shells are reported to occur very sporadically near the base of the Cypresshead Formation in southeastern Georgia (Huddlestun,

1988).

The Cypresshead Formation is known to extend as far north as Dorchester County, South

Carolina, and to be widespread in southeastern Georgia and in the Central Highlands of the

Florida peninsula, south to Highlands County, although the extent of the Cypresshead Formation has not been accurately mapped in this area (Scott, 1992a). In Florida, the unit thins toward the west onto the flanks of the Ocala Platform, and appears to extend into the subsurface south of

Highlands County. In Georgia, north of the Altamaha River, the western limit of the

Cypresshead Formation occurs at or a few kilometers west of the Orangeburg Escarpment, while south of the river, the formation occurs west of the escarpment in northern Wayne County, and immediately west of Trail Ridge (Huddlestun, 1988). In both Florida and Georgia, the eastern edge of the Cypresshead appears to be truncated or grades laterally into age equivalent sands and

marls (e.g. Nashua Formation and/or Raysor Formation). The Cypresshead is thickest in the

Central Highlands of Florida, where it crops out and may attain thicknesses in excess of 200 ft

(~60 m) in Lake County. The reader is referred to Huddlestun (1988) and Scott (1988a; 1992a;

1997) for a more detailed review of the Cypresshead Formation, including type locality, lithology, and stratigraphic relationships.

Tamiami Formation. The late Early Pliocene to Late Pliocene Tamiami Formation of

Mansfield (1939) and Missimer (1992; 1993) is a poorly defined unit containing a wide range of mixed carbonate/siliciclastic lithologies, including limestone, sandstone, quartz sand, marl, shell, and clay (Missimer, 1992). This unit disconformably overlies Hawthorn Group sediments along an erosional contact marking a major Early Pliocene (Zanclean) regression. Illustrating its lithological diversity, the Tamiami Formation contains a number of named and unnamed members, including, but not limited to, the Buckingham Limestone, the Ochopee Limestone, the

Bonita Springs Marl, the Golden Gate Reef, and the Pinecrest beds (Sand) of Olsson and Petit

(1964) which are, in fact, a series of biofacies and lithologies that are mappable only over limited areas (Missimer, 1992; Scott, 1992a). The complex relationships exhibited by these facies are due, in part, to the diverse depositional environments involved in formation of the unit (Scott,

1992a). The reader is referred to Missimer (1992; 1993; 1997; 2001a; b) for a more detailed review of the Tamiami Formation, including type locality, lithology, and stratigraphic relationships.

Raysor Formation and the unnamed Raysor-equivalent shelly sand. The early Late

Pliocene (early Piacenzian) Raysor Formation (Raysor Marl) of Cooke (1936) is a soft, variably shelly, slightly argillaceous, calcareous quartz sand in southeastern Georgia. The quartz sand is typically fine-grained and well-sorted, although coarse sands and pebbles (quartz and feldspar)

have been reported to occur in basal sediments of the unit (Huddlestun, 1988). The early Late

Pliocene (Piacenzian) unnamed Raysor-equivalent shelly sand of Huddlestun (1988) is similar

lithologically to the Raysor Formation, but is restricted to the coastal area of Georgia.

Huddlestun (1988) has described the Cypresshead Formation as disconformably or paraconformably overlying the Raysor Formation in both Effingham and Wayne Counties, and the unnamed Raysor-equivalent shelly sand along the Georgia coast. Planktonic foraminifera recovered by Huddlestun (1988) from the Raysor Formation and the unnamed Raysor-equivalent shelly sand in Georgia are consistent with Zone PL3 of Berggren (1973), or in the case of the unnamed Raysor-equivalent shelly sand, PL3 or PL4, with PL3 roughly equivalent to Zone N20 of Blow (1969). The reader is referred to Huddlestun (1988) for a more detailed review of the

Raysor Formation and the unnamed Raysor-equivalent shelly sand, including type locality, lithology, and stratigraphic relationships.

Nashua Formation. Formerly the Nashua Marl of Matson and Clapp (1909), the late

Pliocene (Piacenzian) to early Pleistocene (Calabrian) Nashua Formation of Huddlestun (1988) is a fossiliferous, variably calcareous, sometimes clayey, quartz sand, with mollusks as the dominant fossil type (Scott, 1992a). Quartz sand is the dominant lithic component of the formation, ranging in grain-size from medium to fine, and constitutes the bulk of the unit to the west where it seems to grade laterally into the Cypresshead Formation (Huddlestun, 1988). The

Nashua Formation appears to underlie the St. Johns River area at least as far south as Deland in

Volusia County, Florida (Huddlestun, 1988), and extends north into Georgia in the subsurface near Jacksonville (Scott, 1992a). Bedding within the unit is massive and appears to be devoid of primary sedimentary or biogenic structures, possessing a maximum estimated thickness of between 40-60 feet (12-18 m), although the type locality is only 6-8 feet (1.8-2.4 m) thick

(Huddlestun, 1988). Based on stratigraphic position and elevation, the Nashua Formation appears

to represent deposition in an open-marine, shallow-water, inner neritic continental shelf setting

consistent with interpretation as an offshore facies of the coastal marine Cypresshead Formation

(Huddlestun, 1988). Planktonic foraminifera recovered by Huddlestun (1988) from two Florida

Geological Survey core sites (W-8400 and W-13815) are consistent with Zone PL5 of Berggren

(1973) or the middle of N21 of Blow (1969), while a suite recovered from near the type locality

are consistent with an early Pleistocene (Calabrian) age, equivalent to Zone N22 of Blow (1969).

The reader is referred to Huddlestun (1988) and Scott (1992a) for a more detailed review of the

Nashua Formation, including type locality, lithology, and stratigraphic relationships.

Caloosahatchee Formation. The Plio-Pleistocene Caloosahatchee Formation (Missimer,

1993; Scott and Wingard, 1995), informally referred to as the Caloosahatchee Marl (Cooke and

Mossom, 1929) or the Bermont Formation (DuBar, 1974), consists of fossiliferous quartz sand with variable amounts of carbonate matrix interbedded with variably sandy, shelly limestone, some of which has a fresh water origin (Scott, 1992a). The unit is reported to extend on the west coast of Florida from north of Tampa south to Lee County, and then to extend eastward to the east coast then northward into northern Florida (DuBar, 1974; Scott, 1992a). However,

Huddlestun (1988) suggests that the Caloosahatchee Formation is neither a mappable lithostratigraphic unit of formation rank nor continuous in the subsurface. Because the unit has historically been recognized on the basis of fossil content rather than lithology, Scott (1992b) now includes it in the informal Okeechobee formation, along with the overlying, faunally- derived Bermont and Ft. Thompson formations. The reader is referred to Scott (1992b) and

Missimer (1993) for a more detailed review of the Caloosahatchee Formation, including type locality, lithology, and stratigraphic relationships.

Pleistocene series

The Cypresshead Formation is overlain in much of the study area by Pleistocene to

Holocene undifferentiated surficial sand of variable origin. Much of this sand is loose, generally

structureless and massive, and ranges in color from pale gray to buff to white (Huddlestun,

1988). These sands include those of marine, pedogenic, and windblown origin that occur at the top of local geologic sections, and underlie, or are part of, the local soil profile (White, 1958;

Huddlestun, 1988; Scott, 1988a). Trail Ridge sands are included with these sediments.

CHAPTER 2 REVIEW OF LITERATURE

Cypresshead Origin and Age

A review of the history of stratigraphic nomenclature applied to Cypresshead Formation

sediments in north-central Florida is given by Kane (1984) and for southeastern Georgia by

Huddlestun (1988). Some of the many stratigraphic designations applied to the Cypresshead

Formation in Florida include the terms Citronelle Formation first used by Doering (1960), Fort

Preston Formation used by Puri and Vernon (1964), and the Grandin Sands, a designation used

by Kane (1984). The first of these, Citronelle Formation, evolved from the belief of Cooke

(1945), Doering (1960), and Pirkle (1960) that the sediments of peninsular Florida correlated to the Citronelle Formation described by Matson (1916) as reddish-orange quartz clastics located in

Mobile County, Alabama. Otvos (1998b) has suggested including both the Cypresshead and

Miccosukee formations in the Citronelle Formation. However, use of that term in northern and

central peninsular Florida implies a direct correlation to southern Alabama and western Florida

that has not been clearly demonstrated. In Georgia, Cypresshead Formation sediments have

previously been included with the Okefenokee and Altamaha Formations by Veatch and

Stephenson (1911) and in various shoreline complexes, among others (Huddlestun, 1988).

Facies Associations

Pirkle (1960) was the first to describe Cypresshead Formation lithofacies, basing his

conclusions on the physical appearance of these sediments in Florida. This stratigraphic technique divided the unit into three zones from the surface downward; (1) loose surface sands

(now considered to be, in part, Pleistocene cover of marine, pedogenic, or windblown origin), (2) red and yellow clayey sands, and (3) white clayey sands, also referred to as the kaolin zone.

Subsequently, Kane (1984) defined four facies, distinct from those of Pirkle (1960), based on

lithologic and biogenic associations observed in north-central Florida. In ascending order, these facies are: bivalve and burrowed, burrowed and trough cross-bedded, burrowed and planar cross- bedded, and unstructured. This vertical facies progression was interpreted by Kane (1984) to indicate a prograding shoreline, consistent with deposition within a coastal or nearshore

environment.

In southeastern Georgia, Huddlestun (1988) described the Cypresshead Formation as

consisting of two gross lithofacies independent of those defined by Kane (1984). The first of these, the updip lithofacies, is described as coarse-grained and pebbly, with the sand-size fraction ranging from fine to coarse with scattered gravel stringers. Sorting in this facies ranges from well-sorted to poorly sorted with typically prominent bedding ranging from thick to thin in thickness. Crossbedding is also conspicuous in this lithofacies, with the largest scale cross- bedding associated with the coarsest and most poorly sorted sands. Ophiomorpha nodosa, a trace fossil, is locally common, and is especially common in the massive, structureless, medium to coarse sands. This lithofacies is interpreted by Huddlestun (1988) to be similar in appearance to

the Citronelle Formation in the panhandle of western Florida, and is described as being typically

developed in updip areas and near large rivers.

The second lithofacies of Huddlestun (1988), the downdip lithofacies, consists of fine- grained sand and clay. It is characterized by thinly-bedded, fine-grained, well-sorted sand with thin layers, laminae, or partings of clay dispersed through the sand. In some areas, the bulk of the formation consists of massive, argillaceous, fine-grained sand that is devoid of any primary sedimentary or biogenic structures. In such cases, the sediment is interpreted by Huddlestun

(1988) as being completely mixed and homogenized by burrowing organisms. Intermediate lithologies consist of bioturbated, poorly mixed sediments commonly associated with a

discontinuous, gray, thinly layered, silty, diatomaceous clay. This lithofacies is interpreted by

Huddlestun (1988) to resemble the Miccosukee Formation of southwestern Georgia and western

Florida, and is described as being typically developed in downdip areas and between large rivers.

Paleoenvironment

The environment of deposition for the Cypresshead Formation in Georgia and Florida has

been interpreted as flood-plain (Davis, 1916), coastal or nearshore marine (Bell, 1924; Martens,

1928; Kane, 1984; Huddlestun, 1988), and alluvial or fluvial-deltaic (Bishop, 1956; Pirkle, 1960;

Pirkle et al., 1964). The theory of an alluvial or fluvial-deltaic origin suggests deposition

associated with a large delta, with terrestrial sediments in some areas laterally continuous with

marine fossiliferous strata (Bishop, 1956). Pirkle (1960) supported the alluvial model based on

the intimate mixing of sediments ranging in size from coarse quartzite pebbles to very fine clay,

as well as the irregular stratification characterizing these sediments. Observations of vertical and

horizontal irregularities in sediment mixtures, extensive cross-bedding and cut-and-fill structures

were all considered compatible with an alluvial origin (Pirkle, 1960). However, this

interpretation was based, in part, on a lack of marine fossils; an observation proven since to be

incorrect (Kane, 1984; Huddlestun, 1988). Also, as noted by Alt (1974), the alluvial or fluvial-

deltaic model is inconsistent with the long, relatively straight geometry of the deposit and its

orientation along the crest of the Florida peninsula.

Although a coastal or nearshore marine model now appears most appropriate, it is unclear

whether the Cypresshead Formation was deposited in a large sound or lagoon, partially isolated

from the open ocean as suggested by Huddlestun (1988), or whether it was deposited in a nearshore marine setting as part of a barrier island complex as suggested by others (Kane, 1984).

Evidence is mixed, as the presence of abundant Ammonia beccarii and Elphidium spp. at the

base of the Cypresshead Formation in Effingham County, Georgia, indicate brackish water

conditions, while the presence of sparse planktonic foraminiferal assemblages in Georgia and

kaolinite molds of pelecypod shell morphologies resembling Mercenaria spp. and Ensis spp. in

Florida suggest that near normal salinities must have prevailed in some areas (Kane, 1984;

Huddlestun, 1988). Additionally, the presence of locally abundant Ophiomorpha spp. trace fossils in both Georgia and Florida (Kane, 1984; Huddlestun, 1988) suggest that the associated sediments were deposited in shallow water, near to sea-level. Thus, it seems most likely that

Cypresshead Formation sediments were deposited in a mix of shallow water marine and coastal environments which appear to have varied in their characteristics in a north-south trend, a condition similar to what is seen along the modern coasts of Georgia and Florida.

Age Estimates

Age estimates for sediments composing the Cypresshead Formation have varied from

Miocene to Pleistocene, with Cooke (1945) the first to assign a Pliocene age to the unit.

Additionally, Cooke (1945) made the observation at that time that the unit was essentially contemporaneous with other Pliocene deposits in Florida, including the Caloosahatchee

Formation, Nashua Formation and Tamiami Formation, among others, and merely represented a littoral lithofacies of the other units (Matson and Clapp, 1909; Sellards, 1914; Cooke and

Mossom, 1929).

The best estimate of the maximum age range for the Cypresshead Formation prior to this study, based on stratigraphic position, limited internal paleontology, and physical correlation, has been interpreted as late Pliocene (early Piacenzian) to early Pleistocene (Calabrian) (Huddlestun,

1988), although deposition in Georgia is most likely restricted to the late Pliocene (late

Piacenzian to late Gelasian). This age range is based, in part, on two microfossil (planktonic and benthic foraminifera) assemblages recovered from the Cypresshead Formation in Wayne and

Chatham Counties, Georgia, and evidence from a third assemblage recovered from the Nashua

Formation in northern Florida. Each of these assemblages is consistent with a late Pliocene age, but based on the contention of Huddlestun (1988) that the Cypresshead correlates laterally with the Nashua Formation, timing for potential Cypresshead Formation deposition was extended into the early Pleistocene (Calabrian).

Correlative Siliciclastics

Although the Cypresshead Formation is thought to be the same age as the Citronelle and

Miccosukee formations (Scott, 1988a; Otvos, 1998b), and is similar to these units lithologically

(Scott, 1988a), the application of the name Cypresshead to the siliciclastics which are the focus of this study is desirable since the units are not traceable into each other (Scott, 1988a). In fact, the Cypresshead in southeastern Georgia and peninsular Florida is separated from the Citronelle

Formation by both the Miccosukee Formation and by a large area where these sediments have either been removed by erosion or where nondeposition occurred.

Citronelle Formation

Named by Matson (1916) for a town in Mobile County, Alabama, the Citronelle

Formation, the most extensive northeastern Gulf of Mexico coastal plain unit, can be traced westward from Alabama across Mississippi and Louisiana into Texas. Eastward, the Citronelle can be traced through Florida to the Little River in central Gadsden County, east of the

Apalachicola River. Including sediments previously considered part of the Lafayette (Matson,

1916; Cooke, 1945) and Bristol (Sellards, 1918) formations by earlier workers, the Citronelle was extended by Cooke and Mossom (1929) to include the “red sand” of the lake region of peninsular Florida, from Clay County southward to Highlands County. In 1988, these sediments were redefined as Cypresshead Formation (Scott, 1988a). Cooke (1945) initially interpreted the

Citronelle Formation as a “littoral or near-shore accumulation of sand and clay brought down by rivers and distributed by waves along the shore of the Gulf”. He referred to that portion of the

Citronelle (Cypresshead Formation) in the peninsula as “likewise a littoral deposit but one

formed far from any large river”. He further suggested that most of the sediment comprising the

Cypresshead Formation likely drifted southward from Georgia along the Atlantic coast, envisioning the Cypresshead as nearshore and beach deposits of the same sea in which the shell marl of the Caloosahatchee Formation accumulated father out. As with the Cypresshead

Formation, the relative paucity of age-diagnostic fossils has hampered efforts to constrain the age of the Citronelle Formation. However, recent dating of the underlying and/or laterally correlative

Jackson Bluff Formation and a few enclosed Japanese umbrella pine pollen specimens, as well as constraints of global climate, sea-level history, and surface elevations now provide a Late

Pliocene (3.4-2.7 Ma) age for the Citronelle (Otvos, 1998b).

Considered to be time equivalent to the Cypresshead Formation (Scott, 1988a; Huddlestun,

1988; Otvos, 1998b), the Citronelle Formation consists mostly of angular to subangular, very poorly sorted, fine- to very coarse-grained quartz sand intercalated with lenses of gravel and clay

(Scott et al., 1980; Scott, 1992a). Coarse- to fine-grained, gravelly alluvial facies sands, rarely more than 20-30 m thick, were deposited in major and minor stream channels and floodplains, with cyclic interlayering of fine- and coarse-grained beds, and indications of channel bank erosion and reworking common (Otvos, 1998a). Pebbles and intraformational mudclasts are common, with pebbles consisting mainly of chert, quartz, flint, and jasper (Otvos, 1998b). Chert pebbles are known to reach >10 cm in length in inland exposures. Burrowed paralic-nearshore facies sediments are recognized in a semi-continuous band from Mobile Bay, Alabama, into the western panhandle of Florida (Otvos, 1998a; b). Burrow traces in these sediments include

Ophiomorpha sp. (Ophiomorpha nodosa) and Skolithos sp., believed to be associated with callianassid (ghost shrimp) and polychaete worm activity, respectively. Internal molds of coastal-

nearshore veneride and other shallow water bivalves also have been identified at several locations (Otvos, 1998b). Intense post-depositional alteration of the Citronelle Formation has resulted in iron and silica mobilization, producing the typical bright orange-red, reddish, and pale orange- to yellowish-brown coloration associated with the unit, as well as the tripolitization of chert pebbles (Otvos, 1998b). The reader is referred to Otvos (1998b) for a more detailed review of the Citronelle Formation, including type locality, lithology, and stratigraphic relationships.

Miccosukee

The Citronelle Formation grades to the east, through a broad facies transition, into the time equivalent Miccosukee Formation of Hendry and Yon (1967). This formation includes all clastic sediments of the Tallahassee Hills that occur above the Miocene Hawthorn Group but below the

Pleistocene sands in the Northern Highlands region of the central panhandle of Florida (Scott et al., 1980). Considered a time equivalent of the Cypresshead Formation (Huddlestun, 1988;

Otvos, 1998b), sediments of the Miccosukee Formation occur in the eastern panhandle of

Florida, extending east from the Little River in central Gadsden County to eastern Madison

County (Rupert, 1990; Scott, 1992a). Various authors (Otvos, 1998b; and others) have argued that the Miccosukee and Cypresshead Formations were once continuous, only to have the connection eroded during terrace construction west of the Okefenokee Swamp.

Lithologically, the Miccosukee is dominated by well-sorted, fine- to medium-grained sand with scattered layers or laminae of white to gray clay (Huddlestun, 1988; Scott, 1992a). More rarely, very fine- and coarse-grained sand beds occur in the unit, as do poorly-sorted, pebbly, cross-bedded coarse-grained sands and clays (Huddlestun, 1988). Limonite pebbles also are common in the unit (Scott, 1992a). Where moderately to deeply weathered, Miccosukee

Formation sands are typically mottled reddish-orange to reddish-brown in color, with associated

clay layers white (Huddlestun, 1988; Otvos, 1998b). This produces a characteristic appearance analogous to the downdip lithofacies of Huddlestun (1988) for the Cypresshead Formation.

Huddlestun (1988) has suggested a coastal marine, possibly bay-sound, depositional environment for the Miccosukee Formation based on the scattered occurrence of burrows

(Ophiomorpha nodosa), bioturbated sediments, and tidal channel scour-and-fill structures.

However, others have suggested a prodeltaic origin for the Miccosukee based on cross-bedding

(Yon, 1966; Scott, 2001). In summary, the Miccosukee appears to represent a down-dip facies of the Citronelle Formation as suggested by Otvos (1988b), and shares depositional affinity with the down-dip facies of the Cypresshead Formation in Georgia as defined by Huddlestun (1988). The reader is referred to Huddlestun (1988) for a more detailed review of the Miccosukee Formation, including type locality, lithology, and stratigraphic relationships.

Kaolin Origin and Provenance

Georgia-South Carolina Kaolin District

As a proxy for potential models of kaolin origin in the Cypresshead Formation, Georgia-

South Carolina kaolin district deposits are the most likely correlative. Disagreement over the origin of the Cretaceous "soft" kaolins and the Tertiary "hard" kaolins has centered on two major problems; (1) the location of kaolinization, and (2) the factors responsible for the differences between the two kaolin types (Dombrowski, 1992). Reviews of the detailed differences between the two types of kaolin are given by Patterson and Murray (1984) and Pickering and Hurst

(1989).

Disagreement as to the origin of the coarse-grained (up to 65% coarser than 2 μm) "soft" kaolins has centered on whether the kaolinite formed in situ through the alteration of crystalline rocks in the southern Piedmont-Blue Ridge region followed by subsequent transportation to the

Cretaceous shoreline, or whether the kaolinite formed by the alteration of arkosic sands at the

Cretaceous shoreline. Keller (1977) was one of the first to conclude that both mechanisms played a role in the origin of the "soft" kaolins, combining data on the geologic history of the area, the morphology of the kaolinite developed from arkosic sediments, and the stability of kaolinite crystals undergoing transportation. Recent research concerning sea-level fluctuations during the

Cretaceous to Early Eocene have further substantiated the idea that multiple mechanisms were responsible for the formation of these deposits (Dombrowski, 1992; 1993).

Pickering and Hurst (1989) summarized a theory for the origin of the "soft" kaolins, noting as follows:

"The initial source material was aluminous detritus from weathered feldspathic rocks in the adjacent Piedmont Upland. The detritus consisted mainly of kaolinite, quartz, and metahalloysite. Lesser constituents were mica, feldspar, smectite, ferric pigments, and anatase."

They further indicate that these sediments, exposed during regressions, would have been subjected to intense weathering and some erosion and redeposition. They conclude their theory, stating:

"The Cretaceous kaolins consist partly of remnants of kaolinite-rich zones of weathering profiles overprinting kaolinitic sediments, but they consist mainly of eroded and redeposited materials derived from such profiles."

Such a model is consistent with kaolin microtextures assigned by Pickering and Hurst (1989) and

Pickering et al. (1997) to post-depositional leaching, oxidation, diagenesis, and authigenesis that have subsequently impacted these deposits.

The origin of fine-grained (> 80% finer than 2μm) "hard" kaolins became a focus of study after that of the Cretaceous deposits. Stull and Boles (1926) and Smith (1929) were some of the first researchers to investigate the differences between the two types of kaolin. Their theories suggested that variations in deposition and diagenesis accounted for the observed differences in the deposits. Later work by Kesler (1956) suggested that the water salinity in which the kaolins

were deposited controlled many of the differences between the "hard" and "soft" kaolins. This observation was supported by Hinckley (1961), who observed face-to-face grain orientation in

"hard" kaolins and face-to-edge grain orientation in "soft" kaolins. He believed this was evidence to support the idea that "hard" kaolins were deposited in saline water, while "soft" kaolins were deposited in fresh water. However, as noted by Pickering and Hurst (1989), the latter observation relating to “soft” kaolins is a function of in situ recrystallization rather than original sedimentation. Subsequent microtextural observations of Huber Formation “hard” kaolins support the observation of Pickering and Hurst (1989), which states:

"In these kaolins, the fine particle size, good sorting and face-to-face association of the kaolinite platelets, as well as the presence of hystrichospherids and dinoflagellates indicate that the hard kaolins are a marine sediment."

Furthermore, the preservation of such microtextures indicates that the younger “hard” kaolins have undergone far less post-depositional alteration than older “soft” kaolins.

Although Georgia-South Carolina district kaolins have commonly been referred to as sedimentary in origin (Ladd, 1898; Veatch, 1909; Kesler, 1956; Murray and Keller, 1993), Hurst and Pickering (1997) and Pickering et al. (1997) argue against describing these deposits as such.

Rather, they suggest the term “Coastal Plain”, to account for the hydrogeologic controls and many post-depositional processes ultimately responsible for kaolin formation. As for regional hydrogeology, they note that Georgia-South Carolina kaolins are restricted to within the recharge area of the regional groundwater system, and are variably impacted by oxic and/or dysoxic weathering reactions as a function of saturated verses unsaturated conditions. As many as twelve distinct post-depositional processes have been identified by Hurst and Pickering (1997) as impacting the formation of Georgia kaolins under these variable groundwater conditions, including the following:

• Bacterially mediated stripping of Fe from organic matter, kaolinite-metahalloysite, illite, and other minerals by HS-, with sequestration of Fe as sulfide (pyrite) under dysoxic conditions

• Destruction or organic matter via bacterially mediated pyrite formation and oxic weathering

• Recrystallization of kaolinite-metahalloysite and coarsening of kaolinite under oxic weathering conditions (Ostwald ripening and/or open-system recrystallization)

• In situ weathering/kaolinization of feldspar, mica, illite, and smectite (montmorillonite) to kaolinite and/or halloysite, particularly under oxic leaching conditions

• Removal through leaching of alkalis, alkaline earths, silica, Fe, and Mn

• Diagenetic changes under long-maintained dysoxic conditions, including pyrite and kaolinite coarsening

• Fe oxidation under oxic weathering conditions, including bacterially mediated breakdown of pyrite under oxic vadose conditions and oxidation of octahedral Fe in kaolinite- metahalloysite

These mechanisms, along with other weathering and diagenetic processes outlined by Hurst and

Pickering (1997) and Pickering et al. (1997), are believed to be responsible for transforming

high-alumina, kaolinite- (and metahalloysite) bearing sedimentary detritus into kaolin deposits of

commercial quality.

Early research on the provenance of the "soft" and "hard" kaolins suggested that the

Tertiary kaolins were derived from the fine fraction of the Cretaceous kaolins by reworking, as

evidenced by rounded kaolin balls in Tertiary sediments (Smith, 1929; Murray, 1976). This

concept was first challenged by Hassanipak and Eslinger (1985), who studied the crystallinity

and oxygen isotopes in different size-fractions of the two kaolin types. They concluded from this

data that the Tertiary kaolins were not the same as the fine fraction of the Cretaceous kaolins, but

failed to elaborate on the causes of the differences. It was not until the work of Dombrowski

(1982; 1992; 1993) and Dombrowski and Murray (1984) that the differences in the two kaolins was attributed, in part, to differences in source provenance. These studies used trace element

geochemistry (La, Th, Co, Sc) to show that the signature of the Cretaceous "soft" kaolins is

identical to kaolinite derived from local granite and gneiss, while the signature of the Tertiary

"hard" kaolins from eastern Georgia and South Carolina are consistent with a mixture of

approximately 70% metavolcanic rocks (Little River Group) and 30% granite/gneiss source

rocks. Trace element analyses of Tertiary "hard" kaolins from central Georgia indicate that these

deposits are derived from predominately metavolcanic source rocks with distinct zones

dominated by detritus from granite/gneiss sources.

The differences in sources appear to be controlled by the availability of material exposed to

alteration and erosion during different time periods (Dombrowski, 1992). During the Cretaceous,

high sea-level stands would have submerged the metavolcanic rocks of the Little River Group,

leaving exposed granite and gneiss as sources for kaolinitic sediment. During the Early Tertiary,

the metavolcanic rocks were exposed to kaolinization processes, and contributed substantially to

the accumulation of Tertiary "hard" kaolins.

Cypresshead Formation

Kaolinitic clays present in the Cypresshead Formation occur as irregular thin beds or

lenses and stringers of clay, or as a binding matrix for sands and gravels, some of which are commonly cross-bedded. Clay content may vary from absent to >50 percent in sandy clay lithologies, although the average content of clay-rich lithologies is 10-20 percent (Pirkle, 1960;

Kane, 1984; Huddlestun, 1988; Scott, 1988a). Armored clay balls and clay rip-up clasts roughly

1-5 cm in diameter are also common (Pirkle, 1960; Kane, 1984), with Pirkle (1960) assigning their origin to the current fragmentation of desiccated kaolinite stringers and lenses following intervals of subaerial exposure. Observations by Pirkle (1960) of the flocculation and settling capacity of these clays during industrial processing suggest that both clay balls and rip-up clasts could have formed under subaqueous conditions as well, without the need for subaerial exposure.

Concentration of these intraformational clasts in the upper portion of the Cypresshead has been

suggested by Kane (1984).

The lone commercially exploited Cypresshead Formation kaolin deposit, located at the

Edgar Minerals EPK facility in north-central Florida, averages approximately 15-18% ≤ 2 μm

(pers. com., Edgar Minerals), and commonly has been referred to as a kaolinitic sand.

Mineralogy of the clay fraction has been assumed to be kaolinite, although recent evidence

suggests the likelihood of a minor halloysite and/or metahalloysite component (Fountain and

McClellan, 1993). According to Pirkle (1960), the highest percentages of kaolinite appear to be

associated with the lower part of the Cypresshead Formation in north-central Florida.

Prior to this study, the existing understanding as to the origin and provenance of

Cypresshead Formation kaolinitic sands was similar to that of Georgia-South Carolina district

kaolins during the 1970s and 1980s, with ongoing questions focused on; (1) the location of

kaolinization, and (2) the source and sedimentological significance of associated sediments. As is

the case with the kaolinitic sedimentary deposits in Georgia, several theories on the origin of

Cypresshead Formation kaolinite have been proposed, including in situ formation from arkosic sands, weathering of smectitic precursor clays, or direct deposition as a sedimentary kaolin.

Arguments supporting the in situ formation of the kaolinitic sediments were first introduced by Sellards (1912), who considered the kaolinite to have formed by the weathering of arkosic sands transported from Piedmont and Blue Ridge sources. Bell (1924) was the first to contradict this view, supporting the alternative model of a sedimentary origin for Cypresshead kaolinite based on the observation that:

"Feldspar would have become completely decomposed long before it could have been transported from the crystalline area on the north to the sedimentary kaolin region, several hundred miles southward."

Building on the observations of Bell (1924), Pirkle (1960) outlined various lines of evidence

supporting a sedimentary origin, including field relationships and sedimentary features

incompatible with in situ models. Among these arguments were the lack of undecomposed or

partially altered feldspar in the unit, the lack of secondary silica phases, the presence of clay balls

and rip-up clasts, and the occurrence of stringers and lenses of nearly pure kaolinite in the unit.

Although a sedimentary, or rather detrital, origin for much of the clay content in

Cypresshead Formation sediments is likely, Austin (1998) suggests an important role for

groundwater leaching of aluminous components (mica, feldspar) in the formation of an in situ

kaolinite fraction, with the associated removal or preservation of Fe and organic compounds

dependent on groundwater acidity, redox, and biological content. Such an observation correlates

with those made by Hurst and Pickering (1997) for the Georgia-South Carolina kaolins. A

comprehensive review of arguments against the in situ theory of Sellards (1912), as well as the

possibility for the in situ weathering of smectitic clays as a potential kaolinite source is given by

Pirkle (1960).

Provenance has only been addressed in the most general of terms relative to Cypresshead

Formation sediments. Observations have been limited to those of Pirkle (1960); Pirkle et al.

(1964), and Kane (1984), with both focusing on the belief that crystalline rocks of the Piedmont and Blue Ridge are the most likely source of kaolinitic sediments. Heavy minerals suites (Pirkle et al., 1964; Kane, 1984) and quartz grain textures (Kane, 1984) evaluated by these researchers support this contention, suggesting a mixed igneous/metamorphic sediment source. To date, no detailed provenance evaluation specific to the kaolinitic clay component of the unit has been attempted.

CHAPTER 3 METHODS

Sample Localities

Samples collected for this study include mine exposure, road/railroad outcrop, and drilling

samples from Peninsular Florida and southeastern Georgia which were collected in cooperation

with the Florida Geological Survey (FGS) and the following mining companies: Vulcan

Materials Company (formerly Florida Rock Industries), Edgar Minerals (formerly Feldspar

Corporation), CEMEX (formerly Standard Sand & Silica/Rinker Materials) and E.I. du Pont de

Nemours & Company (DuPont). Sampling locations are illustrated in Figure 3-1, with a

complete sample list and corresponding analyses performed for this study in Table 3-1.

Cypresshead Formation (Florida)

Sampling sites were selected along the length of the Lake Wales Ridge and its northern

extension, the Interlachen Karstic Highlands, to permit the thorough description of the

Cypresshead Formation in Florida. Both outcrop and drill core/auger samples acquired from

mine sites in Florida were used. Sampling sites in northern Florida included the Edgar Minerals

EPK Mine located off CR 20A near Edgar, Florida, the Vulcan Materials Company (VMC, formerly Florida Rock Industries) Grandin Sand Mine located off SR 100 between Grandin and

Putnam Hall, Florida, the Vulcan Materials Company (VMC, formerly Florida Rock Industries)

Goldhead Mine located off SR 21 north of Keystone Heights, Florida, and the DuPont Trail

Ridge, Highland and Maxville heavy mineral mining areas located northeast of Starke, Florida

(TRF2214, WEX164, WEX366 and MCB109). Sampling sites along the southern portion of the

ridge include the Joshua and Davenport mines operated by CEMEX. These mines are located

near Haines City and Davenport, Florida, respectively. Additionally, the Paran Church site

located across SR 100 from the Grandin Sand Mine and described by Pirkle (1960) and Kane

Figure 3-1. Sample locations evaluated in this study (modified after Huddlestun, 1988; Scott et al., 2001).

Table 3-1. Sample list and corresponding analyses performed for this study.

Particle-Size Analysis Microscopy XRD (< 2 µm) XRD (2-30 µm/Other) Geochemistry Sample Location/ Interval Sample ID Hydrometer/ Major/Trace/ Source (ID/ft) SediGraph SEM PLM/ RLM Oriented Random Oriented Random Nd Isotopes Sieve REEs Cypresshead Formation - FL EPK Mine EPK36-J-12 25–27 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 27–30 ∗ ∗ ∗ ∗ ∗ 35–40 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 40–44 ∗ ∗ ∗ ∗ 44–46 ∗ ∗ ∗ ∗ 46–48 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 48–50 ∗ ∗ ∗ ∗ 50–53 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 53–56 ∗ ∗ ∗ ∗ 56–59 ∗ ∗ ∗ ∗ ∗ ∗ 59–62 ∗ ∗ ∗ ∗ ∗ ∗ ∗ EPK31-P-40 27–35 ∗ ∗ ∗ ∗ ∗ 35–45 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 45–50 ∗ ∗ ∗ ∗ 50–62 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 62–65 ∗ ∗ ∗ ∗ ∗ ∗ ∗ EPK30-V-6 16–22 ∗ ∗ ∗ ∗ 22–24 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 24–27 ∗ ∗ ∗ ∗ 55 30–35 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 35–39 ∗ ∗ ∗ ∗ 39–43 ∗ ∗ ∗ ∗ ∗ 43–48 ∗ ∗ ∗ ∗ 48–53 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 53–58 ∗ ∗ ∗ ∗ ∗ 58–63 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 63–68 ∗ ∗ ∗ ∗ ∗ 68–73 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 73–78 ∗ ∗ ∗ ∗ ∗ EPK Vermiforms — ∗ ∗ ∗ ∗ EPK Mica — ∗ ∗ ∗ ∗ EPK Feldspar — ∗ ∗ ∗ ∗ Grandin Mine FRG-1 1 ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 7 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 8 ∗ ∗ ∗ ∗ ∗ 9 ∗ ∗ ∗ ∗

Table 3-1. – (continued).

Particle-Size Analysis Microscopy XRD (< 2 µm) XRD (2-30 µm/Other) Geochemistry Sample Location/ Interval Sample ID Hydrometer/ Major/Trace/ Source (ID/ft) SediGraph SEM PLM/ RLM Oriented Random Oriented Random Nd Isotopes Sieve REEs Grandin Mine FRG-1 10 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 11 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 12 ∗ ∗ ∗ ∗ ∗ ∗ 13 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 14 ∗ ∗ ∗ ∗ 15 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ FRG-2 1 ∗ ∗ ∗ 2 ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ 7 ∗ ∗ ∗ ∗ ∗ 8 ∗ ∗ ∗ 9 ∗ ∗ ∗ 10 ∗ ∗ ∗ ∗ ∗ 11 ∗ ∗ ∗ 12 ∗ ∗ ∗ ∗ ∗ 13 ∗ ∗ ∗ Goldhead Mine FRL-1 1 ∗ ∗ ∗ ∗ ∗ 56 2 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 7 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 8 ∗ ∗ ∗ ∗ 9 ∗ ∗ ∗ ∗ ∗ ∗ ∗ Joshua Mine SSJ-1 1 ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ 4 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ ∗ 7 ∗ ∗ ∗ ∗ 8 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 9 ∗ ∗ ∗ 10 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 11 ∗ ∗ ∗ ∗ DuPont TRF2214 60.0–62.5 ∗ ∗ ∗ ∗ ∗ ∗ ∗ WEX164 18.0–26.0 ∗ ∗ ∗ ∗ ∗ ∗ ∗ WEX366 9.0–10.0 ∗ ∗ ∗ ∗ ∗ ∗ ∗

Table 3-1. – (continued).

Particle-Size Analysis Microscopy XRD (< 2 µm) XRD (2-30 µm/Other) Geochemistry Sample Location/ Interval Sample ID Hydrometer/ Major/Trace/ Source (ID/ft) SediGraph SEM PLM/ RLM Oriented Random Oriented Random Nd Isotopes Sieve REEs Reworked Cypresshead Formation - FL Davenport Mine SSD-1 1 ∗ ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ 7 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 8 ∗ ∗ ∗ ∗ ∗ 9 ∗ ∗ ∗ 10 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ Cypresshead Formation - GA Jesup J-1 1 ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ Linden Bluff L-1 1 ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ 57 3 ∗ ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 6 ∗ ∗ ∗ ∗ ∗ ∗ ∗ 7 ∗ ∗ ∗ Birds B-1 1 ∗ ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ ∗ 3 ∗ ∗ ∗ ∗ ∗ ∗ 4 ∗ ∗ ∗ ∗ 5 ∗ ∗ ∗ ∗ ∗ ∗ ∗ Hawthorn Group, Coosawhatchie Formation - FL/GA DuPont MCB109 15.0–20.0 ∗ ∗ ∗ ∗ Jesup J-1 BC ∗ ∗ ∗ ∗ Huber Formation - GA CMS Standard KGa-2 — ∗ ∗ ∗ Congo Boone Mine ECCI-CB — ∗ ∗ ∗ ∗ Buffalo Creek Formation - GA CMS Standard KGa-1 — ∗ ∗ ∗ Buffalo-China Mine ECCI-BC — ∗ ∗ ∗ ∗ Ennis Avant Mine TKC-EA — ∗ ∗ ∗ ∗ Brooks 93 Mine DBK-B93 — ∗ ∗ ∗ ∗

(1984) was investigated to assess sedimentary features exposed in the pit for comparison to

sampling sites investigated in this study. For more detailed reference, site maps for the mines are

included in Appendix A.

Cypresshead Formation (Georgia)

In order to compare data from Florida samples with that of the Cypresshead Formation in

Georgia, samples were collected from the type section at the type locality designation of

Huddlestun (1988) located near the town of Jesup, in Wayne County, Georgia (Fig. 3-1). Two

reference localities designated by Huddlestun (1988) also were sampled as a part of this effort.

The first, Linden Bluff, is located on the south bank of the Altamaha River, east-southeast of the

type locality. The second reference locality is a CSX railroad cut along Ebenezer Creek, located

at Birds in Effingham County, Georgia. The reader is referred to Huddlestun (1988) for a

detailed description of the Cypresshead Formation type section/locality and the two reference

localities sampled for this study.

Middle Georgia Kaolin District

Kaolin samples from the Middle Georgia Kaolin District were collected from several mine

sites in order to compare both mineralogy and geochemistry to Cypresshead Formation samples.

Selected Cretaceous kaolin samples include those from the IMERYS (formerly ECCI) Buffalo-

China Mine (ECCI-BC), the IMERYS (formerly Dry Branch Kaolin) Brooks 93 Mine (DBK-

B93) and the Thiele Kaolin Ennis Avant Mine (TKC-EA). One Tertiary kaolin sample was collected from the Jeffersonville Member of the Huber Formation exposed in the IMERYS

(formerly ECCI) Congo Boone Mine (ECCI-CB) in Hancock County. Additionally, two Clay

Mineral Society (CMS) Source Clay standards representative of low defect, Cretaceous (KGa-1) and high defect, Tertiary (KGa-2) Georgia kaolins were evaluated (Note: KGa-1b has now replaced KGa-1 due to exhaustion of the original standard). These Source Clay standards are

carefully selected from large, commercial, reasonably homogeneous deposits, and have been

characterized for common physical and chemical properties to allow for their use as well-

characterized standards in clay research. The reader is referred to Volume 49, Number 5 (2001)

of Clays and Clay Minerals for a collection of baseline studies addressing the physical and chemical properties of the Clay Mineral Society Source Clays.

Sample Preparation

Following detailed field description and collection, bulk samples from the Cypresshead

Formation and the Middle Georgia Kaolin District were split into subsamples for processing on

return to the Industrial Minerals Laboratory at the Department of Geological Sciences,

University of Florida. A flow chart outlining sample processing for the various analytical

techniques used in this study is shown in Figure 3-2. Samples collected for analysis were initially

crushed, as needed, prior to additional pretreatment to facilitate dispersion and avoid

flocculation.

Samples for hydrometer- and sieve-based particle-size analysis were processed using

standard procedures outlined in ASTM D422-63, which use a 4% Na-metaphosphate solution for

sample dispersal. For SediGraph, mineralogical and geochemical analyses, sample splits were

processed using a modification of the ASTM D422-63 procedure, where concentrated, Optima-

grade ammonium hydroxide (NH4OH) was added after introduction of the sample to the settling

column to adjust the pH of the clay-water solution, thereby facilitating dispersion. This

modification of the ASTM D422-63 procedure prevents contamination of the clay samples by

Na-metaphosphate, which is difficult or impossible to completely remove from clays with washing, and would have interfered with isotopic and trace element analyses. The clay (< 2 µm) and fine silt (2-30 µm) size-fractions were then separated using standard gravity settling and centrifuge techniques (Hathaway, 1956), and washed repeatedly to remove excess NH4OH. Once

Figure 3-2. Sample processing and analysis flow chart.

washing was complete, samples were freeze-dried, as needed, for mineralogical and geochemical analysis.

Analytical Procedures

Grain-Size Analysis

Grain-size characteristics were evaluated using hydrometer-, sieve- and SediGraph-based techniques in order to evaluate general sedimentological parameters and to evaluate specific weathering-induced characteristics associated with the fine silt/clay distribution within select samples. As both weathering and illuviation were considered to have likely had a significant effect on the details of grain-size distributions, the degree of accuracy afforded by the pipette method of analysis was deemed unnecessary.

Hydrometer and sieve analysis

Grain-size analyses were performed in accordance with ASTM D422-63 Standard Method for Particle-size Analysis using the hydrometer method in accordance with Stokes Law. For this study, a representative sample of approximately 30 g was used for each analysis along with a sample split to determine moisture content. The sample for hydrometer analysis was first soaked in 125 ml of a 4% Na-metaphosphate solution for no less than 16 hours. Following that step, the sample was transferred to an ASTM specified high-speed mixer in a stainless steel cup with stationary baffles. Following agitation for one minute, the sample slurry was transferred to a

1000 ml glass cylinder and brought to volume with deionized (DI) water. Each sample was allowed to stand to check for flocculation or thixotropy.

Next, the sample was stirred for one minute using a brass stirring rod with a perforated disk at the base. The hydrometer was then placed in the cylinder prior to each reading, and measurements were recorded at 1, 2, 4, 8, 15, 30, 60, 120, 300, and 1440 minutes. A reference hydrometer consisting of a hydrometer placed in a 1000 ml cylinder filled with a blank solution of 125 ml Na-metaphosphate and DI water was used as a modification of the ASTM D422-63 procedure. This allows for direct determination of the effects of temperature and atmospheric pressure on the density of the solution. Temperature also was recorded during each hydrometer reading in order to monitor any changes.

After the 24 hour (1440 min.) reading, samples were washed over a U.S. Standard 200 mesh sieve to remove clay- and silt-size materials. The plus-200 mesh fraction was then dried at

105˚C. The dried sand fraction was then sieved over U.S. Standard 20, 40, 60, 80, 100, and 200 mesh sieves using a Ro-Tap shaker. Weights for each size fraction were then recorded and the total weight of the plus-200 fraction determined. Subsequently, the relative percentages of sand, silt and clay were calculated using standard graphical techniques in association with a plot of the

grain-size distribution of each sample. Additionally, moment statistics were calculated for each set of results in accordance with the method outlined by Balsillie (1995). In combining hydrometer- and sieve-based grain-size data, datasets were normalized as needed. As noted by

Coakley and Syvitski (1991), there is a noted breakdown in Stokes Law near the sand-silt size boundary at 63 µm, which often corresponds to a dip in the size frequency distribution of a merged sample. This observation is commonly caused by the distributions of the two techniques not being overlapped and renormalized, but rather being abutted.

SediGraph analysis

Grain-size distribution test procedures based on gravitational sedimentation are relatively inaccurate for clay particles below 1 µm in e.s.d., particularly if the sedimenting slurry is disturbed by insertion and removal of a hydrometer. As a result, supplemental grain-size analyses were performed using a Micromeritics SediGraph 5100 at the University of Florida in order to more thoroughly characterize the fine silt and clay size fraction of select samples. The aim of these analyses was to determine the presence or absence of fine silt- and clay-size populations correlated to weathering processes. For this purpose, the SediGraph 5100 uses an X- ray/sedimentation method of grain-size analysis that is accurate and reproducible for grain-sizes ranging from 1-70 µm e.s.d., with the lower limit for acceptably accurate results being 1 µm as noted by Hendrix and Orr (1972) (the manufacturer reports accuracy to 0.1 µm). Below 1-0.5

µm, particle settling behavior becomes governed more by Brownian motion rather than Stokes

Law. Stein (1985) has noted that montmorillonite-rich samples are difficult to analyze on the

SediGraph due the ability of the thixotropic properties of montmorillonite to change viscosity and hinder grain settling.

Following standard SediGraph procedures, samples were prepared as sediment suspensions on the order of 0.02-0.1 g/ml in concentration to avoid hindered settling effects (Stein, 1985;

Coakley and Syvitski, 1991). This was accomplished by combining 2.2 g of air-dried minus-200 mesh (< 75 µm) sample with 70 ml of a 0.05% Na-metaphosphate dispersant solution. This sample concentration was found to be the most suitable for SediGraph analysis following testing of a range of concentrations using sample FRL-1-9 (Fig. 3-3). Following overnight dispersion, samples were agitated ultrasonically prior to introduction to the SediGraph 5100 sample reservoir. Using standard test parameters for silicate materials, triplicate analyses were performed for the size range 0.5-64 µm, with periodic collection of baseline data on the 0.05%

Na-metaphosphate dispersant/rise solution. Subsequently, the resulting mean data were graphically and statistically evaluated, with test precision assessed via analysis of sample

EPK36-J-12 (56-59) duplicates (Fig. 3-4).

1.8 g 2.0 g 2.2 g 2.4 g

2.5

2.0

1.5

1.0 Mass Frequency (%) Frequency Mass 0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure 3-3. SediGraph concentration test results of sample FRL-1-9 at concentrations of 1.8 g, 2.0 g, 2.2 g and 2.4 g mixed with 70 ml of dispersant solution (curve for each concentration is the average of two analyses).

36-J-12 (56-59) Replicate 1 36-J-12 (56-59) Replicate 2

8.0

6.0

4.0 Mass Frequency (%) 2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure 3-4. SediGraph precision test results of sample EPK36-J-12 (56-59) comparing two replicate samples (curve for each replicate is the average of three analyses).

X-ray Diffraction Analysis

X-ray diffraction (XRD) analysis of clay (< 2 µm) and fine silt (2-30 µm) size-fractions separated by standard gravity settling and centrifugation techniques was accomplished using a

Philips APD 3600/XRG3100 diffractometer located at the Department of Geological Sciences,

University of Florida. The diffractometer was equipped with a monochrometer and scintillation detector, and employing CuKα radiation at 45 KV and 30 mA for all analyses. Software used to process the resulting data was Jade 3.1+ published by Materials Data, Inc. As accurate peak shape definition was essential for the extraction of information about crystallite size distributions and structural order from the XRD data, a fixed-slit goniometer configuration employing a 2˚ divergence slit and a 0.5˚ receiving slit was used. The use of a wide receiving slit aperture (0.5˚) has been shown by Madsen and Hill (1988) to provide increased peak intensities, slightly more

Gaussian peak shapes, and slightly wider peaks. Moreover, divergence slits ≤ 2º have been deemed satisfactory for pattern evaluation purposes, while slits > 2˚ have been shown to generate serious deficiencies in Voigt and pseudo-Voigt peak shape models for low angle peaks (Madsen and Hill, 1988). Step size and scanning speed also have been shown to have an influence on peak shape (Wang, 1994), with the parameters of 0.02º 2θ and 1 sec respectively, chosen for this study in order to minimize these effects and still permit collection of XRD data in a timely fashion.

Data was collected over a range of 4°-65° 2θ and 4°-30° 2θ for random and oriented samples, respectively, with external XRD reference standards (Arkansas novaculite quartz plate and SRM

675) periodically analyzed to correct for instrument drift.

Oriented samples

Oriented samples of the < 2 μm size fraction were prepared and analyzed using the

Millipore filter transfer method (Drever, 1973; Moore and Reynolds, 1997). Wet sample splits of the < 2 μm mineralogy/geochemistry concentrate outlined in Figure 3-2 were separated prior to freeze-drying the remainder, and resuspended/dispersed using approximately 15 ml of 4% Na- metaphosphate solution to 5 ml of the ~70% solids concentrate in a 50 ml centrifuge tube.

Samples were allowed to soak overnight prior to further disaggregation/dispersion by a Model

450 Branson Sonifier, and were then agitated and an aliquot of the dispersed suspension, ranging from 10 to 20 ml was transferred to a 0.45 μm Millipore filter under vacuum. The sample was then filtered rapidly, agitating as necessary, in order to prevent size segregation and keep the sample homogenous as the flow rate slows during filtration.

After collection as a filter cake on the Millipore filter, each sample was then Mg saturated using a 20 ml aliquot of 0.1 M MgCl2 solution as outlined by Moore and Reynolds (1997) and rinsed with DI water to remove any excess soluble salts, and the filter and clay removed from the apparatus. The filter then was inverted onto a clean glass slide, briefly dried, and the filter peeled

off following a modification of the Drever (1973) method similar to that outlined by Moore and

Reynolds (1997). Drever (1973) rolled the clay cake onto the glass slide using a glass cylinder.

However, Moore and Reynolds (1997) note that this can cause a non-Gaussian particle

orientation, which, in turn, produces an unknowable Lorentz factor.

Samples were analyzed air-dried, solvated with ethylene glycol, and heat-treated (as necessary). Pretreatment of oriented clay samples by ethylene glycol is useful in identifying expansive phases such as smectites. After XRD analysis in the air-dried state, samples were placed on a platform in a Pyrex dessicator containing 2–3 cm of ethylene glycol in the bottom, and exposed to the reagent at 60ºC for a minimum of 12 hours per the method outlined by Moore and Reynolds (1997). On removal of the samples from the dessicator, XRD analysis was performed within 1 hour to prevent evaporation of the ethylene glycol. Heat treatments to 105°C for a minimum of four hours were employed as needed to define the presence of halloysite in some of the samples examined in this study. On completion of data collection, XRD data were evaluated using the MudMaster computer program of Eberl et al. (1996) to calculate crystallite size distributions. The reader is referred to Eberl et al. (1996) for a detailed review of the

MudMaster program and data requirements.

Random samples

Samples of freeze-dried clay (< 2 µm) and fine silt (2-30 µm) were powdered to minus-200 mesh using an agate mortar and pestle prior to analysis as random clay mineral aggregates. Care was taken to minimize grinding in order to prevent the creation of grinding induced crystalline defects as illustrated by Kristóf et al. (1993). Random powder mounts were prepared using aluminum holders and a side-drifted technique as a means of producing results consistent with a

randomly oriented sample. On completion of data collection, the qualitative mineralogy of the

clay and fine silt size-fractions was determined, with kaolinite reflections evaluated for

crystalline disorder characteristics using the methods of Hinkley (1963), Liètard (1977) and the

“expert system” of Plançon and Zacharie (1990). Quantitative XRD analyses were not performed

due to the mineral suite consistency noted for most of the samples studied.

Petrographic/Scanning Electron Microscope Analysis

Standard petrographic and scanning electron microscope (SEM) analysis of kaolinitic

sediments were performed at the University of Florida using equipment made available at the

Department of Geological Sciences and the Major Analytical Instrumentation Center (MAIC).

Additionally, sample preparation facilities at the Department of Soil and Water Science were

also employed. Reflected light microscopy (RLM) photomicrographs were taken with an

Olympus binocular microscope equipped with a digital camera attachment.

Thin sections for petrographic analysis were prepared using standard epoxy impregnation techniques (Araldite epoxy resin), employing a 2% Automate Blue 8HF dye to indicate sample

porosity. Once impregnated chips were prepared, they were sent to Spectrum Petrographics, Inc.,

in Vancouver, Washington for final preparation. However, rather than finishing the thin sections

to a 30 µm thickness, sections used in this study were finished to 50 µm in order to more readily

delineate kaolinite. The thin sections were then analyzed via polarized light microscopy (PLM)

using a Nikon Eclipse E600W POL microscope equipped with a Nikon Digital Still Camera

DXM1200.

Samples taken for SEM analysis were embedded in a colloidal silver paste on the surface

of an aluminum SEM stub, then prepared using standard coating techniques (C and/or Au-Pd),

with analysis performed at the Major Analytical Instrumentation Center (MAIC) using a JEOL

JSM-6400 scanning electron microscope equipped with an EDS system with a thin window solid

state detector.

Geochemical Analysis

Freeze-dried clay fraction (<2 μm) samples for geochemical analysis were powdered to minus-200 mesh using an agate mortar and pestle. In between each sample, the equipment was thoroughly cleaned. Agate was selected as the grinding medium in order to minimize the potential for trace element contamination of the samples.

Major and trace element analysis

Major and trace element analyses were performed by XRAL Laboratories, in Toronto,

Ontario, Canada. Powdered clay-size fraction (< 2 μm) separates were analyzed using an ICP-

AES method (ICP40) developed for the USGS along with a complimentary ICP-MS (MS95) analyses as outlined by Methods A and B, Appendix B. Samples for ICP-AES were decomposed using a multi-acid “total” digestion procedure at low temperature as outlined by Crock et al.

(1983). Digested samples were then aspirated into the ICP-AES discharge of a Perkin Elmer

Optima 3000 ICP-AES where the elemental emission signal was measured simultaneously for 40 elements. Calibration was performed by standardizing with digested rock reference materials and a series of multi-element solution standards (Lichte et al., 1987). Data was deemed acceptable if recovery for all 40 elements was ± 15 % at five times the Lower Limit of Determination (LOD) and the calculated Relative Standard Deviation (RSD) of duplicate samples is no greater than 15

%. CRM SO-3 was analyzed by ICP-AES and ICP-MS as a means of evaluating laboratory accuracy.

Nd isotopic analysis

Clay fraction (<2 μm) separates were analyzed for neodymium (Nd) isotopic values using a “Nu-Plasma” multiple-collector magnetic-sector inductively coupled mass spectrometer (MC-

ICP-MS) located at the Department of Geological Sciences, University of Florida. All samples were prepared in a class 1000 clean lab, equipped with class 10 laminar flow hoods, using

standard cation exchange column chemistry as outlined by Method C, Appendix B, then diluted

to approximately 250 ppb in Optima-grade 2% HNO3. Both samples and standard solutions were

then aspirated into the plasma source via a Micromist nebulizer with GE cinnabar spray

chamber. The instrument settings were carefully tuned to maximize the signal intensities on a

daily basis. Preamplifier gain calibration was performed before each analytical session. Nd

isotope measurements were conducted for 60 ratios in static mode acquiring simultaneously for

142Nd on low-2, 143Nd on low-1, 143Nd on Axial, 145Nd on high-1, 146Nd on high-2, 147Sm on

high-3, 148Nd on high-4, and 150Nd on high-5 Faraday detectors. The measured 144Nd, 148Nd, and

150Nd beams were corrected for isobaric interference from Sm using 147Sm/144Sm = 4.88,

147Sm/148Sm = 1.33, and 147Sm/150Sm = 2.03. All measured ratios were normalized to 146Nd/144Nd

= 0.7219 using an exponential law for mass-bias correction (O’Nions et al., 1977; Belshaw et al.,

143 144 4 1998). Nd/ Nd values are cited in εNd notation as parts in 10 deviation from CHUR

(chondritic uniform reservoir) defined by present-day 143Nd/144Nd = 0.512638 and 147Sm/144Nd =

0.1967 (Jacobsen and Wasserberg, 1980).

The mean value of 143Nd/144Nd for the Ames Nd in-house standard based on 23 repeat

analyses during the time that study samples were measured is 0.512140 (2σ = 0.000011),

corresponding to reproducibility of better than ± 0.3-ε units at 2σ levels (95% confidence) (Table

3-2). Three repeat analyses each of the JNdi-1 and LaJolla Nd standards during the same time

interval produced mean values of 0.512106 (2σ = 0.000013) and 0.511856 (2σ = 0.000013),

respectively (Table 3-3). Three samples of USGS SRM BCR-1 (Columbia River basalt) were

prepared via Method C, Appendix B, and analyzed for Nd isotopes together with the samples in order to further evaluate the analytical protocol. As shown in Table 3-3, the mean value of

143Nd/144Nd for the analyses was 0.512645 (2σ = 0.000011), which is indistinguishable from the

Table 3-2. MC-ICP-MS analyses of the Ames Nd in-house standard. Nd Error x Date 143Nd/144Nd 142Nd/144Nd 145Nd/144Nd 148Nd/144Nd 150Nd/144Nd 146Nd/144Nd εNd (ppb) 10-6 raw MC-ICPMS 06/02/03 ~250 0.512145 5 1.141746 0.348424 0.241561 0.236390 0.738502 -9.6 06/02/03 ~250 0.512150 5 1.141771 0.348411 0.241548 0.236373 0.738353 -9.5 06/30/03 ~250 0.512132 6 1.141761 0.348410 0.241557 0.236411 0.740981 -9.9 06/30/03 ~250 0.512142 5 1.141775 0.348409 0.241550 0.236397 0.740617 -9.7 06/30/03 ~250 0.512148 6 1.141763 0.348403 0.241544 0.236387 0.740618 -9.6 07/01/03 ~250 0.512145 4 1.141785 0.348413 0.241543 0.236380 0.739758 -9.6 07/01/03 ~250 0.512139 5 1.141747 0.348412 0.241548 0.236389 0.739771 -9.7 07/01/03 ~250 0.512144 6 1.141781 0.348412 0.241539 0.236371 0.739859 -9.6 07/01/03 ~250 0.512145 4 1.141782 0.348418 0.241554 0.236375 0.739979 -9.6 07/01/03 ~250 0.512146 4 1.141769 0.348416 0.241547 0.236374 0.739947 -9.6 07/02/03 ~250 0.512145 6 1.141758 0.348653 0.241551 0.236386 0.739881 -9.6 07/02/03 ~250 0.512140 6 1.141779 0.348409 0.241558 0.236386 0.740251 -9.7 07/16/03 ~250 0.512134 3 1.141727 0.348407 0.241539 0.236376 0.739788 -9.8 07/16/03 ~250 0.512137 4 1.141715 0.348408 0.241551 0.236382 0.739614 -9.8 07/16/03 ~250 0.512140 6 1.141766 0.348417 0.241546 0.236360 0.739190 -9.7 07/18/03 ~250 0.512135 5 1.141751 0.348404 0.241552 0.236392 0.740067 -9.8 07/18/03 ~250 0.512140 4 1.141764 0.348419 0.241551 0.236370 0.739495 -9.7 08/08/03 ~250 0.512134 4 1.141760 0.348406 0.241539 0.236394 0.740858 -9.8 08/08/03 ~250 0.512131 6 1.141743 0.348411 0.241551 0.236397 0.740721 -9.9 08/08/03 ~250 0.512143 5 1.141770 0.348422 0.241543 0.236383 0.740813 -9.7 09/03/03 ~250 0.512140 4 1.141746 0.348405 0.241551 0.236388 0.739989 -9.7 09/03/03 ~250 0.512134 4 1.141747 0.348403 0.241550 0.236385 0.739873 -9.8 09/03/03 ~250 0.512131 5 1.141751 0.348410 0.241548 0.236379 0.739799 -9.9

Mean 0.512140 1.141759 0.348422 0.241549 0.236384 0.739944 -9.7 Std. Dev. 5.7E-06 1.8E-05 5.1E-05 5.8E-06 1.1E-05 6.7E-04 1.1E-01 143 144 Note: Errors on Nd/ Nd measurements are 2σ. Measured εNd values calculated assuming a present-day 143Nd/144Nd(CHUR) = 0.512638 (Jacobsen and Wasserberg, 1980).

published thermal ionization mass spectrometer (TIMS) measurement of 0.51264 (Gladney et al.,

1990). Given the agreement between measured and published values for these standards, no

further corrections were performed. Internal measurement errors of standards are reported in

Table 3-2 and Table 3-3.

During the time interval of MC-ICP-MS analyses, a Micromass Sector 54 TIMS located at the University of Florida was used to collect Nd isotopic values for the Ames Nd in-house standard, following methods described in Scher and Martin (2004). Operated in dynamic mode and equipped with seven Faraday collectors and one Daly collector, the mean value of

143Nd/144Nd for 18 repeat analyses run as NdO was 0.512140 (2σ = 0.000012), in agreement with

MC-ICP-MS results (Table 3-4). Internal measurement errors for the Ames Nd standard are

included in Table 3-4. Advantages of MC-ICP-MS over the standard TIMS method include

Table 3-3. MC-ICP-MS analyses of the JNdi-1, LaJolla Nd, and BCR-1 standards. Nd Error x Date 143Nd/144Nd 142Nd/144Nd 145Nd/144Nd 148Nd/144Nd 150Nd/144Nd 146Nd/144Nd εNd (ppb) 10-6 raw JNdi-1 0.512115 7 -10.2 ~250 0.512102 6 1.141744 0.348408 0.241554 0.236386 0.739760 -10.5 ~250 0.512103 5 1.141744 0.348403 0.241546 0.236388 0.739633 -10.4 ~250 0.512114 7 1.147770 0.348406 0.241547 0.236369 0.739793 -10.2

Mean 0.512106 1.143753 0.348405 0.241549 0.236381 0.739729 -10.4 Std. Dev. 6.5E-06 3.5E-03 2.4E-06 4.3E-06 1.1E-05 8.5E-05 1.3E-01

LaJolla Nd 0.511858 7 -15.2 ~250 0.511849 6 1.141778 0.348409 0.241552 0.236413 0.739819 -15.4 ~250 0.511856 4 1.141772 0.348417 0.241546 0.236435 0.739635 -15.3 ~250 0.511862 5 1.141807 0.348412 0.241546 0.236433 0.739839 -15.1

Mean 0.511856 1.141786 0.348413 0.241548 0.236427 0.739764 -15.3 Std. Dev. 6.6E-06 1.9E-05 3.8E-06 3.7E-06 1.2E-05 1.1E-04 1.3E-01 Nd Error x Date 143Nd/144Nd 142Nd/144Nd 145Nd/144Nd 148Nd/144Nd 150Nd/144Nd 146Nd/144Nd εNd (ppm) 0 10-6 raw BCR-1 28.80 0.51264 30 0.0 0.512648 5 1.244630 0.348429 0.241649 0.310536 0.740311 0.2 * 0.512649 (2.4) 6 1.367637 0.348438 0.241618 0.293322 0.740096 0.2 * 0.512639 (3.2) 6 1.186770 0.348429 0.241666 0.315206 0.740052 0.0

Mean 0.512645 1.266346 0.348432 0.241644 0.306355 0.740153 0.1 Std. Dev. 5.4E-06 9.2E-02 5.3E-06 2.4E-05 1.2E-02 1.4E-04 1.1E-01 143 144 Note: Errors on Nd/ Nd measurements are 2σ. Measured εNd values calculated assuming a present-day 143Nd/144Nd(CHUR) = 0.512638 (Jacobsen and Wasserberg, 1980). 143 144 143 144 JNdi-1: Nd isotopic reference standard of Tanaka et al. (2000). Nd/ NdJNdi-1 = 1.000503 × Nd/ NdLaJolla Nd. LaJolla Nd: Nd isotopic reference standard of Lugmair and Carlson (1978). BCR-1: Columbia River Basalt USGS SRM data source reference of Gladney et al. (1990). * Average of duplicate analyses (n=2). Standard deviations for 143Nd/144Nd shown in parenthesis (×10-6).

Table 3-4. TIMS analyses of the Ames Nd in-house standard. NdO Error x Date 143Nd/144Nd 142Nd/144Nd 145Nd/144Nd 148Nd/144Nd 150Nd/144Nd 162NdO/160NdO εNd (ng) 10-6 06/04/03 75 0.512127 8 1.141840 0.348382 0.241574 0.236523 0.720799 -10.0 06/04/03 75 0.512139 8 1.141869 0.348377 0.241586 0.236509 0.723820 -9.7 06/06/03 150 0.512144 8 1.141895 0.348379 0.241569 0.236516 0.720284 -9.6 06/09/03 75 0.512144 9 1.141889 0.348389 0.241574 0.236500 0.719983 -9.6 06/12/03 75 0.512138 8 1.141866 0.348395 0.241579 0.236484 0.720113 -9.8 06/12/03 75 0.512149 8 1.141880 0.348381 0.241558 0.236484 0.720506 -9.5 06/13/03 75 0.512146 8 1.141872 0.348397 0.241572 0.236505 0.720598 -9.6 06/20/03 75 0.512134 8 1.141869 0.348380 0.241563 0.236510 0.719821 -9.8 06/21/03 75 0.512136 9 1.141838 0.348387 0.241566 0.236502 0.719876 -9.8 06/21/03 75 0.512137 7 1.141855 0.348376 0.241562 0.236501 0.720107 -9.8 06/28/03 150 0.512134 8 1.141914 0.348384 0.241565 0.236488 0.720333 -9.8 06/29/03 150 0.512146 8 1.141895 0.348388 0.241576 0.236493 0.720795 -9.6 06/30/03 150 0.512133 8 1.141861 0.348381 0.241575 0.236495 0.720897 -9.8 07/09/03 75 0.512139 8 1.141881 0.348384 0.241569 0.236492 0.720537 -9.7 07/08/03 75 0.512139 9 1.141874 0.348387 0.241586 0.236500 0.723172 -9.7 07/07/03 75 0.512138 9 1.141873 0.348384 0.241580 0.236494 0.720884 -9.8 07/13/03 75 0.512144 8 1.141897 0.348388 0.241571 0.236501 0.720763 -9.6 07/15/03 75 0.512149 7 1.141901 0.348382 0.241559 0.236497 0.720503 -9.5

Mean 0.512140 1.141876 0.348384 0.241571 0.236500 0.720766 -9.7 Std. Dev. 5.9E-06 2.0E-05 5.7E-06 8.2E-06 1.0E-05 1.1E-03 1.2E-01 Note: Errors on 143Nd/144Nd measurements are 2σ. Measured εNd values calculated assuming a present-day 143Nd/144Nd(CHUR) = 0.512638 (Jacobsen and Wasserberg, 1980).

increased sample throughput without noticeable loss of precision, and simplification in separation procedures due to the chemical isolation of Nd from Sm no longer being necessary

(Luais et al., 1997).

CHAPTER 4 SEDIMENTOLOGICAL AND MINERALOGICAL EVIDENCE FOR THE ORIGIN AND STRATIGRAPHIC SIGNIFICANCE OF THE CYPRESSHEAD FORMATION

Introduction

The timing and mechanism of Late Miocene through Pliocene siliciclastic transport and deposition on the Florida Platform has been a focus of many studies (Cunningham et al., 1998;

Guertin et al., 1999; Cunningham et al., 2003), yet remain in dispute, particularly as relates to depositional model inconsistencies and the significance of this sediment flux in relation to changes in global climate and sea-level. Illustrating this dispute are the questions surrounding the

Cypresshead Formation of Huddlestun (1988), a siliciclastic unit in peninsular Florida and southeastern Georgia which is characteristically a mottled, fine- to coarse-grained, often gravelly, variably clayey quartz sand. The formation has been variously described as Miocene to

Pleistocene in age (Puri and Vernon, 1964; Scott, 1988a), and originating either through fluvial- deltaic processes (Bishop, 1956; Pirkle, 1960; Pirkle et al., 1964; Klein et al., 1964; Peacock,

1983; Cunningham et al., 2003) or via longshore current transport and deposition in a nearshore, marine to brackish environment (Bell, 1924; Martens, 1928; Alt, 1974; Winker and Howard,

1977b; Peck et al., 1979; Kane, 1984; Huddlestun, 1988; Scott, 1988a). Even the stratigraphic nomenclature applied to the unit has varied, with the term “coarse clastics” first applied by

Cooke (1945) shifting through several stratigraphic designations prior to Huddlestun’s work in southeastern Georgia and extension of the Cypresshead Formation name into Florida by Scott

(1988a).

The Cypresshead Formation and correlative siliciclastics in the Florida panhandle, including the Citronelle Formation of Matson (1916), were once viewed as representing a unique, isolated episode of siliciclastic deposition during the Pliocene (Pirkle; 1960; Pirkle et al.,

1964). That view, however, has been erased by the work of Cunningham et al. (1998; 2001;

2003) and others over the past decade, which indicates that siliciclastic flooding of the Florida peninsula, involving sediments which share an affinity with those of the Cypresshead Formation, extends from the Late Miocene through at least the end of the Pliocene. Thus, the Cypresshead

Formation represents just a component of an apparent 6 to 7 Ma interval of episodic siliciclastic deposition (and reworking), and as a consequence, may be viewed as an indicator of the sedimentary dynamics which drove deposition throughout this period. Through a renewed evaluation of sedimentological and mineralogical indicators, this study defines both the timing and mechanism of Cypresshead Formation deposition, and incorporates that understanding into a synthesis of recent studies on related siliciclastics of the Florida Platform. This approach provides a basis for constraining the regional correlation and stratigraphic significance of the episodic siliciclastic deposition that dominated the Florida peninsula during this time interval.

Related Deposits and the Siliciclastic Conveyor

The Cypresshead Formation has long been correlated with the Pliocene age (3.4 – 2.7 Ma) sediments of the Citronelle Formation (Otvos, 1998b). Composed of unconsolidated siliciclastics consistent with an alluvial to nearshore (or shoreface) marine origin, Citronelle sediments can be traced from the central Florida panhandle to as far west as eastern Texas (Otvos, 1998a; b), and exhibit many similarities (e.g., trace fossils, cross bedding, etc.) to those of the Cypresshead.

However, these two deposits also exhibit distinct differences in both sediment sourcing and sedimentary architecture which confirm the stratigraphic delineation accepted by most researchers (Kane, 1984; Scott, 1988a), although deposition of the units appears to have been concurrent.

It is the siliciclastics in southern Florida (Table 4-1), and particularly the Late Miocene through Pliocene sequence studied by Cunningham (1998; 2001; 2003) and others, which offer the greatest insight into the origin, timing and significance of Cypresshead deposition. The oldest

Table 4-1. Summary of the lithostratigraphic and sequence stratigraphic nomenclature applied to South Florida siliciclastics.

Member/Interval/ Study Formation Age/Time (Ma) Environment Sequence

Scott (1992b) Okeechobee Formationa --- Pleistocene --- Late Pliocene (Gelasian) Cunningham et al. (1998) Okeechobee Formationa ------to Pleistocene Late Miocene (Messinian) Long Key Formationb ------to Pliocene

Guertin et al. (1999) Long Key Formationb III ~2.0 inner shelf outer shelf to inner shelf II 4.5-3.5 transition outer shelf to inner shelf I 6.2-5.5 transition Caloosahatchee shallow shelf/ramp, open Missimer (2001a;b) --- 2.14-0.6 Formation bay, lagoonal/embayment Tamiami Formation Pinecrest 3.22-2.15 shallow coastal marine

Sand Facies 4.29-3.22 shallow marine

Cunningham et al. (2001) Tamiami Formation Unnamed Sand Late Pliocene inner shelf Ochopee Limestone late Early Pliocene carbonate ramp (DS4) Peace River Formation DS3 5.23-3.83 ---

Late Miocene (latest DS2 --- Tortonian and Messinian)

DS1 Late Miocene (Tortonian) outer shelf late Early to Late Cunningham et al. (2003) Tamiami Formation SS3 --- Pliocene Upper Peace River outer shelf to inner shelf SS2* 8.6-3.75 Formation transition/lagoonal Arcadia/Lower Peace aggradational, mixed SS1 24.31-8.6 River Formations carbonate-siliciclastic ramp a The Okeechobee Formation is an informal stratigraphic unit introduced by Scott (1992b), and includes the faunally derived Caloosahatchee, Bermont and Ft. Thompson Formations. b The Long Key Formation is a stratigraphic unit proposed by Cunningham et al. (1998). * SS2 siliciclastics can be subdivided into eastern (SS2-E: 8.6-5.04 Ma) and western (SS2-W: 5.04-3.75 Ma) components. of these sediments are assigned to the upper Peace River Formation of the Hawthorn Group by

Cunningham et al. (2003) and the proposed Long Key Formation of Cunningham et al. (1998), and extend as far south as the Florida Keys, achieving a thickness of greater than 145 m. The latter of these units, the Long Key Formation, has been proposed as a potential correlative of the

Cypresshead Formation, and perhaps a partial correlative of the Tamiami, Caloosahatchee and upper Peace River Formations (Scott and Wingard, 1995; Cunningham et al., 1998) based, in

part, on a late Late Miocene (Messinian) to Late Pliocene (Gelasian) age determination

(Cunningham et al., 1998; Guertin et al., 1999). Furthermore, the unit appears to represent a

shallowing-upward succession corresponding to progradation in response to a substantial

siliciclastic flux (Guertin et al., 1999). Lithologically, the Long Key Formation consists of a

single quartz sand facies consisting primarily of moderately to well sorted, angular to

subrounded, very fine to fine quartz sand, with carbonate grains (≤ 50%), minor feldspar (≤ 5%),

mica, and phosphate. Quartzite pebbles, often discoid in shape, are also common.

Of the Late Miocene siliciclastics associated with the upper Peace River and Long Key

Formations, the SS2 sequence (8.6 – 3.75 Ma) of Cunningham et al. (2003), along with the

Interval I siliciclastics (6.2 – 5.5 Ma) of Guertin et al. (1999), mark the first evidence for the

onset of siliciclastic deposition which shares compositional affinity with the Cypresshead

Formation. Consisting primarily of quartz sand with varying admixtures of terrigenous mudstone and diatomaceous mudstone, the initial pulse of SS2 deposition is widespread over a broad portion of southern Florida, and correlates with the DS2 siliciclastics from an earlier study by

Cunningham et al. (2001). Siliciclastic supply appears to have been restricted during the subsequent early Pliocene (Cunningham et al., 2001), and as a consequence, deposition during this interval is more spatially limited (SS2-W and DS3) or exhibits significant evidence of reworking from earlier deposited siliciclastics (Interval II) (Guertin et al., 1999; Cunningham et al. 2003). During the remainder of the Pliocene, siliciclastic accumulation continued episodically in southern Florida (SS3/Interval III), but with intervening periods of carbonate deposition as characterized by the Ochopee Limestone Member of the Tamiami Formation (DS4) and the lower Caloosahatchee Formation (Guertin et al., 1999; Cunningham et al., 2001; 2003).

The relative volume and architecture of south Florida siliciclastics have been used as the basis for a siliciclastic conveyor model first proposed by Cunningham et al. (2003). This model, based primarily on the occurrence of 50-100 m thick siliciclastic clinoforms from the SS2 sequence, employs a fluvial-deltaic system as the principle means by which the bulk of siliciclastics were transported and deposited on the southeastern Florida Platform during this

interval. Such an interpretation is in agreement with earlier held views related to the deposition

of Cypresshead Formation sediments (Bishop, 1956; Pirkle, 1960; Pirkle et al., 1964), but these views have since been shown to be inconsistent with field observations (Kane, 1984;

Huddlestun, 1988). Proposed alternatives to the fluvial-deltaic transport model proposed by

Cunningham et al. (2003) include channeled deposition by strong paleocurrents associated with a

southward prograding shoreline (Warzeski et al., 1996) or via longshore transport, potentially as a prograding spit (Winker and Howard, 1977b; Warzeski et al., 1996). Identification of the transport mechanism aside, the siliciclastic conveyor model constrains a significant and potentially unique period of deposition on the Florida peninsula, of which the Cypresshead

Formation is the most readily studied component.

Age Constraints on Cypresshead Deposition

Temporal control for the Cypresshead Formation is poor, resulting from a paucity of fossil material of known age, a lack of radiometrically dateable volcanic rocks, and the lack of a

sufficiently long, continuous stratigraphic section suitable to magnetostratigraphic study.

However, lithologic correlation can be used to constrain the age of the unit in both central

peninsular Florida and southeastern Georgia. Researchers have for some time correlated the unit

with the Citronelle and Miccosukee Formations of the Florida panhandle (Scott, 1988a; Otvos,

1998b), the Nashua Formation in northern Florida (Scott, 1988a) and southeastern Georgia

(Huddlestun, 1988), and at least in part, with the sediments of the Tamiami Formation (Bishop,

1956; Cunningham et al., 1998) and possibly the Caloosahatchee Formation (Scott and Wingard,

1995; Cunningham, 1998) in southern Florida. Recently, Scott (2001) also noted that the

Cypresshead Formation appears to occur in the subsurface south of Highlands County, Florida,

and suggests correlation, at least in part, with the Long Key Formation of Cunningham et al.

(1998). Based on these relationships alone, the timing of Cypresshead deposition can be

constrained to the Late Pliocene, although the possibility for an early Pleistocene component

must be considered if the unit is determined to fully correlate with the Nashua Formation, the type section of which is early Pleistocene (Calabrian) in age (Huddlestun, 1988).

Pliocene Paleoclimate

The Pliocene represents a transitional period of global cooling following the warmer, low

amplitude, low frequency climate fluctuations of the Miocene, and preceding the high amplitude,

high frequency climate fluctuations of the more recent Pleistocene. Although representative of a

general cooling trend, warm episodes did occur, particularly during the early to mid-Pliocene

(4.8-3.2 Ma), when climate is believed to have been warmer than at any other time since the

Miocene (Dowsett and Loubere, 1992; Kennett and Hodell, 1995; Crowley, 1996; Dowsett et al.,

1996; Haywood et al., 2000). During this interval, enhanced thermohaline circulation (heat

transport) has been proposed as the primary cause for warming, rather than increased CO2 concentrations (Crowley, 1996; Haywood et al., 2000), resulting in a general reduction in the latitudinal gradient of both atmospheric and sea-surface temperature (Willard, 1994; Dowsett et al., 1996). Pollen data collected by Willard (1994) show that temperatures were much warmer than present at high latitudes, yet remained similar to conditions today at lower latitudes.

Following this warm period, steady temperature decline is believed to have occurred between

2.8-2.4 Ma (McDougall, 1994), with a significant drop in temperatures at 2.6 Ma corresponding to the onset of major Northern Hemisphere ice sheet formation.

As for sea-level during episodes of mid-Pliocene warmth, some climate models predict increases of more than 25 m above present levels (Dowsett and Cronin, 1990; Dowsett et al.,

1994). However, Kennett and Hodell (1993; 1995) have suggested that relative climate stability in the early to mid-Pliocene would have limited the fluctuation in Antarctic ice sheet volume during this time such that maximum sea-level would have been <25 m above present levels.

Alternatively, Naish and Wilson (2009) suggest that as much as a 30 m rise in sea-level may have been expected during the mid-Pliocene in response to a proposed 30 percent loss of the present-day mass of the East Antarctic Ice Sheet (EAIS) and complete deglaciation of the West

Antarctic Ice Sheet (WAIS) and Greenland.

Lastly, increased aridity has long been suggested for portions of the Late Miocene and

Pliocene (Axelrod and Raven, 1985; Jiang et al., 2005). Evidence for such late Pliocene increased aridity, associated with the onset of regression after 3.3 Ma, has been recorded for tropical regions of the Southern Hemisphere (Stein and Sarnthein, 1984). Alt (1974) was the first to support the proposal for increased aridity impacting the southeastern United States during the

Miocene and subsequent Pliocene, suggesting that arid (or semi-arid) climatic conditions may have influenced deposition of the coarse upland siliciclastics associated with the Citronelle and

Cypresshead Formations. As part of this model, Alt (1974) suggested that blanketing siliciclastics were deposited in response to periodic flooding under arid conditions similar to that associated with modern central and northern Australia. Alt (1974) further speculated that the driving mechanism for these conditions was a southward displacement of prevailing wind patterns, bringing the southeastern United States into the path of the prevailing westerlies, thereby isolating the region from sources of moisture. However, more recent studies support the concept that continual El Niño conditions prevailed at least during the early Pliocene warm

period (~4.5–3.0 Ma) (Ravelo and Wara, 2004; Fedorov et al., 2006). Under such conditions, regional climate in the southeastern United States would have been much wetter than previously

thought.

Results

Outcrop and Core Descriptions

A total of six sections and five cores totaling approximately 139 m of Cypresshead

Formation and reworked Cypresshead Formation sediments were evaluated for this study from

the sites shown on Figure 3-1. Detailed cross-sections and descriptions of the six exposures from

Florida and Georgia, along with two cores collected from central Florida locations are included

in Appendix C. Core samples collected from the Edgar Minerals EPK Mine at surface elevations

of 27 – 28 m were not suitable for stratigraphic analysis.

Of the three sections studied in north-central Florida, two are located at the same mine site

(Grandin), with the third located approximately 12 km north-northwest near the town of

Keystone Heights. Other than the Paran Church site, located across the road from the Grandin

pit, these sites represent the most complete exposures of the Cypresshead Formation in north-

central Florida. The three sites evaluated in Georgia include the type locality at Jesup (J-1) and

two reference localities as defined by Huddlestun (1988). Additionally, three Cypresshead

Formation samples supplied by E.I. du Pont de Nemours & Company (DuPont) from the Trail

Ridge (1) and Highland (2) heavy mineral mining areas located northeast of Starke, Florida, were

also evaluated for clay mineralogy.

Sedimentary Framework

Standard sedimentological parameters for the Cypresshead Formation including grain-size

characteristics, the distribution and types of sedimentary structures, and information regarding

sediment transport mechanisms were determined for this study in order to assist in defining a

comprehensive depositional model for the unit. The data outlined here were supplemented by the previous observations of Kane (1984) and others.

Grain-size distributions

Cypresshead Formation sediments consist of poorly to well sorted, very fine to coarse, clayey sands, sandy clays, gravels and clays. Discoid quartzite pebbles up to 4 cm in diameter and clay rip-up clasts 1 cm to 5 cm in diameter are common as well. Mean grain-size for

Cypresshead sediments ranges from 0.8 to 4.6 Φ (coarse sand to silt), with percent of sand, silt, and clay ranging from 15 to 98, 0 to 41, and 1 to 63, respectively (Fig. 4-1). Mean grain-size for reworked Cypresshead sediments (Davenport Sand Mine: SSD-1) ranges from 1.3 to 3.3 Φ

(medium sand to very fine sand), with percent of sand, silt, and clay ranging from 88 to 93, 2 to

7, and 3 to 8, respectively. Summary tables and figures defining percent sand, silt and clay, along with moment statistics and grain-size distribution curves for each sample included in this study are included in Appendix D. Additionally, select grain-size distribution curves are shown in

Figure 4-2 for discussion purposes.

Pervasive weathering has modified the original grain-size distribution of Cypresshead sediments, preventing detailed environmental interpretation. In particular, processes which include illuviation and the in-situ formation of clays at the expense of the coarser mineral fraction (Fountain and McClellan, 1993; Heuberger, 1995) have modified the depositional signature of the deposit, imparting a fines tail to most grain-size distribution curves (Fig. 4-2) which is inconsistent with hydrodynamic sorting. Further evidence for a post-depositional increase in the fines content of most samples is illustrated by the high positive skewness values seen with Cypresshead sediments (Fig. 4-3). Not only do coarser sediments (low Φ value) exhibit the highest values, but differences are noted between reworked and non-reworked

16 Reworked Cypresshead 14 Cypresshead

Frequency 6

0 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 Mean Grain-Size (phi)

Figure 4-1. Histogram illustrating the distribution of mean grain-size values for samples of the Cypresshead Formation and reworked Cypresshead sediments.

sediments, with reworked sediments exhibiting an overall reduction in fines likely in response to winnowing activity during the reworking process.

Another feature noted for the grain-size distributions are the coarse tails associated with individual sand-size modes within total sample curves (Fig. 4-2). This feature is consistent with shoreline hydrodynamic sorting generated by swash and/or wave action, and appears to be best developed higher up in the sections evaluated for this study, except at the Birds reference locality

(Fig. 4-2F). The coarse tails are also greatest in coarse-grained samples and diminished in finer- grained materials, the latter of which are associated with the lower portion of each section.

Lastly, both unimodal and bimodal distributions are seen with the grain-size curves. In the case of the unimodal curves, they tend to either correspond to coarse grained sands near the top of the section or fine to very fine sands near the base. Samples with intermediate grain-size characteristics exhibit the greatest degree of bimodality, usually with a dominant coarse or

Figure 4-2. Representative grain-size distribution curves for Cypresshead Formation sediments. A) Grandin Sand Mine (FRG-1), B) Goldhead Sand Mine (FRL-1), C) EPK Mine (EPK30-V-6), D) Joshua Sand Mine (SSJ-1), E) Jesup type locality (J-1) and F) Birds reference locality (B-1).

medium sand mode paired with a secondary fine sand component. Bimodal distributions such as these are indicative of mixed energy environments during deposition.

14.0 Cypresshead

12.0 Reworked Cypresshead

10.0

8.0

6.0 Skewness

4.0

2.0

0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Mean Grain-Size (phi)

Figure 4-3. Scatter plot illustrating the relationship between mean grain-size and skewness for samples of Cypresshead Formation and reworked Cypresshead sediments.

Sedimentary structures

A variety of environment and process sensitive sedimentary structures are associated with

Cypresshead Formation sediments. Included among these are various types of stratification and bedforms, discontinuity surfaces and trace fossils.

Stratification and bedforms. Several forms of stratification indicative of depositional environment are well preserved in Cypresshead sediments. The first of these, cross-stratification, has been described in detail by Kane (1984), and consists of three principle types; tabular, trough and hummocky. Tabular cross-stratification is small to large in scale, ranging from 5 cm to 3 m in bed thickness, with foreset dips generally in the 20-25° range (Fig. 4-4). Cross-strata are generally symmetrical, with angular to tangential lower bounding surface contacts consistent with formation by the migration of straight-crested sand waves and dunes (megaripples). The smaller of these features (5-60 cm) are concentrated near the top of the sections where they are

Figure 4-4. Example sedimentary structures from the Cypresshead Formation. A) Tabular and trough cross-bedding associated with sample interval FRG-1-5, illustrating scour and fill and horizontal bedding surfaces (arrows) along with escape structures (circles), B) Tabular cross-bedding associated with sample intervals FRG-1-9 through FRG-1-11 containing 0.5-1 m long Ophiomorpha spp. burrows (arrows) inclined in response to current activity, C) Hummocky cross-stratification from sample interval FRG-1-13, D) Large-scale infilling of a nearshore channel at the Grandin Sand Mine, E) Clay lense nondepositional discontinuity (arrow) from sample interval L-1-6 (offset due to slumping along exposure face), F) .Clay bed nondepositional discontinuity (bracket) from sample interval J-1-4.

present, but may occur sporadically near the base as well. Larger features (0.5-3 m) were noted only in north-central Florida at the Grandin and Goldhead Sand Mines, and are consistent with the migration of dune or megaripple bedforms associated with nearshore bars (Fig. 4-4B).

Trough cross-stratification is small to medium in scale, ranging from 5 cm to 60 cm in bed

thickness, with foreset dips in the same range as seen with tabular cross-strata (Fig. 4-4). Cross-

strata are lenticular and asymmetrical, and exhibit tangential contacts with erosional lower

bounding surfaces consistent with formation by the migration of undulatory and lunate sand

waves. Graded bedding within individual bed sets is common. As was noted for tabular cross-

strata, these features are concentrated near the top of most sections, with a less common and

more sporadic distribution with increased depth. Scour and fill structures are commonly

associated with this type of cross-stratification (Fig. 4-4), suggesting significant flow velocities.

Large-scale scoured surfaces with up to 15 m in vertical relief, likely representative of nearshore

current or tidal channels, are also evident in limited Cypresshead exposures in north-central

Florida (Fig. 4-4D).

The last type of cross-stratification noted here, hummocky, was not previously noted by

Kane (1984), but was observed in the basal fine sands of both exposures studied at the Grandin

Sand Mine locality (Fig. 4-4). These cross-strata are medium in scale, ranging from 0.3 m to 1 m

in bed thickness, with foreset dips and truncation angles < 15°. As was noted for trough cross-

stratification, cross-strata form tangential contacts along erosional lower bounding surfaces.

Diagnostic traits which differentiate these beds from other forms of cross-stratification are the

antiformal hummocks and synformal swales which are defined by randomly oriented, even

lamination (Dott and Bourgeois, 1982). Hummocky cross-stratification is most commonly

associated with redeposition of fine sand below normal fair-weather wave base by large waves.

Paleocurrent information derived from the orientation of cross-stratification offers insight into the transportation direction of Cypresshead siliciclastics. Representative measurements collected for this study at the Grandin Sand Mine to compliment data accumulated by Kane

(1984) for other sites in north-central Florida support earlier observations, indicating a dominant southerly flow orientation with a secondary bimodal, bipolar flow component generally oriented east-west (Fig 4-5A). Composite measurements made of cross-stratification in southeastern

Georgia, however, exhibit substantially different results, with the dominant flow orientation

generally to the east, and a secondary component suggesting a subordinate flow orientation to the

south. Paleocurrent measurements at Linden Bluff (72°, n = 15) exhibit the most pronounced

eastern flow characteristics, with measurements made at the Jesup type locality (121°, n = 8)

more consistent with the secondary flow orientation to the south (Fig. 4-5B).

The second form of stratification noted in Cypresshead sediments is horizontal, and sometimes massive, bedding. Where noted, horizontal bedding is more concentrated at the base of exposures, with the Birds reference locality exhibiting horizontal bedding throughout the section. This appears to be consistent with deposition under low flow velocity conditions insufficient to develop ripple or larger bedforms. Graded bedding within horizontally bedded strata is also common. Massively bedded Cypresshead strata appear to result from the destruction of original sedimentary fabric through bioturbation, or in some cases, weathering. These beds tend to be concentrated near the top of exposures where weathering is strongest, or in association with other bioturbated beds which have retained evidence of original stratification.

Discontinuity surfaces. Two types of stratigraphic discontinuities are recognized in

Cypresshead siliciclastics; nondepositional and erosional. The first of these, nondepositional discontinuities, mark abrupt decreases in sediment accumulation rates and are commonly

Figure 4-5. Paleocurent rose diagrams for representative Cypresshead Formation exposures. A) FRG-1, B) Georgia composite based on Linden Bluff (L-1: n = 15) and Jesup type locality (J-1: n = 8) measurements, C) FRG-1 USF facies, D) FRG-1 LSF (pLSF + dLSF) facies.

associated with increased concentrations of burrowing activity (e.g. FRG-1-6) or the deposition

of discrete clay lenses or beds (Fig. 4-4E and F). Erosional discontinuities are more common in

Cypresshead sediments, and mark an abrupt increase in sediment accumulation, grain-size (e.g.

graded bedding) and corresponding erosive scouring, either by currents or waves. These features

are particularly pronounced at the base of medium to large-scale planar cross-stratification associated with dune (or megaripple) bedforms (Fig. 4-4B).

Trace fossils. Kane (1984) summarized the occurrence of Ophiomorpha spp. trace fossils, bivalve molds and fecal pellets as the sum fossil assemblage associated with the Cypresshead

Formation. However, a more careful evaluation of exposures in Florida and southeastern Georgia indicate a slightly more diverse assemblage than previously described. Ophiomorpha spp. burrows, normally consistent with Ophiomorpha nodosa, continue to be the most common trace fossil in both Florida and Georgia sections, and are typically 3-4 cm in diameter and can exceed

1 m in length (Figs. 4-4B and 4-6). Specimens retain a characteristic knobby exterior and may exhibit minimal branching oriented at an oblique angle to bedding. In most instances,

Ophiomorpha spp. burrows are vertically oriented and non-branching although occasionally burrows may be inclined in response to strong current conditions during coeval deposition and burrowing activity (Fig. 4-6B). Burrows stand out in relief, consisting of clay cemented sands more resistant than the surrounding sediments. Where sediment clay contents are low, escape structures are most common (Fig. 4-6A). Burrows are normally sand filled with rare instances of kaolinite replaced fecal remains preserved in the base. In highly bioturbated units, bedding structures are lacking due to high concentrations of burrows (Fig. 4-6A and B). Along the modern Florida and Georgia coastlines, Callianassa major burrows are considered to be a modern analog, and are accepted as shoreline indicators. Although they primarily occur in the sandy, open marine littoral to shallow neritic environment, they have been reported on protected beaches, sandy tidal flats, and shoals (Frey, 1970), tidal deltas (Warme, 1971), and offshore bars

(Weimer and Hoyt, 1964).

Thallasinoides spp. burrows observed in north-central Florida basal Cypresshead sediments (FRG-1, FRG-2 and FRL-1) are often similar in size and appearance to Ophiomorpha spp. burrows, but occur in units consisting of fine sand with greater clay contents, have greater

Figure 4-6. Examples of trace fossil and bivalve mold occurrences. A) Densely bioturbated sample interval FRG-1-6, B) Close-up of interval FRG-1-6 from (A) showing the morphology of Ophiomorpha spp. (left arrow) and Skolithos spp. (right arrow) traces, C) Thallasinoides spp. burrows from the Paran Church site exhibiting secondary burrowing of shaft walls and burrow interiors, D) Fossil bivalve molds from the Paran Church site.

developed horizontal branching, and lack a knobby exterior (Fig. 4-6C). Differences in structure likely reflect differences in the substrate in which the burrow was constructed. Thallasinoides spp. burrows examined at the Paran Church site commonly exhibit secondary boring of shaft walls and backfilled burrow interiors (Fig. 4-6C), producing a trace which is similar in appearance to the Skolithos spp. traces observed in the same exposure interval from the nearby

Grandin Sand Mine.

Skolithos spp. traces are commonly associated with both Ophiomorpha spp. and

Thallasinoides spp. burrows. These traces have been described by Chamberlain (1978) as any

simple, even width vertical tube varying in diameter from 2mm to 10 mm, with walls which are

usually smooth, but may be segmented or striated. In the study area, these tubes range in

diameter from 2mm to 5 mm and possess burrow walls formed from agglutinated sand grains.

Polychaete (annelid) worms are likely responsible for these structures, and are indicative of a

marginal marine facies (Seilacher, 1967). Along the Georgia Sea Isles coast, analogous tubes

associated with the polychaete species Onuphis microcephala often occur in association with

Callianassa major burrows (Curran, 1985). Environmentally, Skolithos spp. traces appear

consistent with burrowing activity during periods of quiescent to highly reduced sedimentation.

Additionally, the low biodiversity exhibited by this trace fossil assemblage is consistent with a

high stress environment associated with episodic high sedimentation rates dictated by

fluctuating, but often high, current flow velocities.

The last fossils of note which have been described in detail by Kane (1984) are clay

(kaolin) molds of marine bivalves similar in morphology to the modern surf clam Mercenaria spp. and the razor clam Ensis spp. Not seen in sections evaluated for this study, but readily

visible in horizontal exposures at the north-central Florida Paran Church site (Fig. 4-6D), molds

represent disarticulated valves which appear to have been transported by current and/or tidal

activity from a proximal estuarine source. Preservation is poor, with stratigraphic positioning of

the fossils in basal Cypresshead sediments potentially favored by the relatively high clay content

of the fine sands.

Mineralogy and Petrography

Details of the mineralogy and petrography of both the sand and clay size-fractions were

evaluated for this study for correlation to previous studies on the Cypresshead Formation and

comparison to complimentary data on south Florida siliciclastics. X-ray diffraction (XRD) data

for the samples (oriented and random) are included with Appendices E and F. A comprehensive

evaluation of the clay mineralogy of Cypresshead Formation sediments and the implications for

weathering and recrystallization processes is addressed in the Chapter 5.

Sand size-fraction

Evaluation of the sand size-fraction confirms previous observations related to the mineral

suite, indicating quartz as the dominant lithic component. Coarse to fine quartz grains are angular

to subrounded in shape, consistent with particle morphologies noted for Long Key and SS2

siliciclastics (Cunningham et al., 1998; 2003), with angularity decreasing with increased grain- size as a consequence of transportation mechanism (i.e. suspension vs. saltation or traction).

Accessory components of the sand size-fraction include feldspar, mica, kaolinite vermiforms and

heavy minerals (Fig. 4-7), the latter of which has been discussed in detail by Pirkle et al. (1964).

Feldspars, confirmed as K-feldspar (microcline) via energy dispersive X-ray spectroscopy

(EDS), petrographic examination, and XRD, are found at the base of all southeastern Georgia sampling locations (J-1-6, L-1-5, L-1-7 and B-1-5), and only in lower sections of the EPK Mine and Joshua Sand Mine cores collected in Florida. Since feldspars grains would correlate depositionally with quartz grains of similar size, feldspar occurrences are related to preservation as a function of weathering rather than hydrodynamic sorting. At the north-central Florida EPK

Mine site, feldspars are preserved in all cores approximately 16 m below the surface elevation of

27-28 m (11-12 m msl), while at the central Florida Joshua Sand Mine preservation is seen approximately 11.5 m below the surface elevation of 40 m (28.5 m msl). For the Jesup, Linden

Bluff, and Birds locations, feldspars are preserved at 23.7 m msl, 21.7 m msl and 15.2 m msl, respectively. Feldspar concentrations, where seen, range from 1% to 5% of the sand size- fraction.

The distribution of mica, confirmed as muscovite via EDS, petrographic examination, and

XRD, appears related to hydrodynamic sorting, as it is notable as a trace or minor component in

Figure 4-7. Photomicrographs of accessory sand-sized phases from the Cypresshead Formation. A) PLM image of K-feldspar (microcline) grain with twinning from SSJ-1-8, B) RLM image of K-feldspar concentrate from EPK Mine cores, C) PLM image of mica (muscovite) grains (arrows) from FRG-1-8, with the lower grain exhibiting expansion (exfoliation) in response to kaolinization, D) RLM image of mica concentrate from EPK Mine cores, E) SEM image of kaolinite vermiforms from FRL-1-4, F) RLM image of large vermiform (book) concentrate of from EPK Mine cores.

most Cypresshead Formation sediments. The exceptions to this observation are highly scoured

and/or cross-bedded units within the Cypresshead which appear to have been deposited under

relatively high flow velocities. Such depositional conditions appear to have preferentially

winnowed mica grains, preventing their incorporation into the sediments, with the highest

concentrations of mica seen to be associated with more quiescent deposition, particularly with

basal medium to fine clayey sands and the entirety of the EPK Mine cores and Birds section.

Kaolinite vermiforms (or books) occurring in both the sand and silt size-fractions (Fig. 4-

7E and F), are observed throughout Cypresshead sediments in Florida, but are relatively rare in

southeastern Georgia Cypresshead sediments. The distribution of vermiforms is similar to that of the feldspar grains in Florida, with the largest and highest concentrations seen toward the base of exposures, although the distribution appears, in part, correlated with intervals of mica concentration, a preferred mineralogical template on which vermiforms commonly form during weathering (Jeong, 1998b).

Clay size-fraction

Kaolinite is the dominant clay mineral found in both Cypresshead Formation and reworked

Cypresshead Formation sediments (Table 4-2). In Florida, the clay (< 2 µm) size-fraction also includes lesser amounts of quartz, gibbsite, halloysite, hydroxyl-interlayered vermiculite (HIV),

and crandallite-florencite, with occurrences of rutile, anatase, boehmite, and diaspore in discrete

samples. Furthermore, metahalloysite is inferred to be a part of the suite based on the occurrence

of halloysite, and the likelihood for halloysite dehydration in response to fluctuating water table

conditions. Both smectite and illite are indicated to occur in north-central Florida Cypresshead

sediments in association with the Trail Ridge and Highland mining area samples collected updip

from the principle sampling sites employed in this region. Overlain by Pleistocene sediments

associated with Trail Ridge, these clays may be indicative of a less weathered Cypresshead suite

Table 4-2. Clay size-fraction mineralogy of Cypresshead Formation and reworked Cypresshead Formation sediments.

Crandallite- Sample ID Interval % Clay Kaolinite Gibbsite Halloysite Illite HIV Quartz Other Florencite Cypresshead Formation - FL EPK36-J-12 25-27 13.1 *** * ** ** EPK36-J-12 27-30 7.5 *** * * ** EPK36-J-12 35-40 8.2 *** * * * EPK36-J-12 40-44 14.7 *** EPK36-J-12 44-46 12.6 *** EPK36-J-12 46-48 12.9 *** EPK36-J-12 48-50 13.7 *** EPK36-J-12 50-53 4.7 *** EPK36-J-12 53-56 9.7 *** EPK36-J-12 56-59 15.6 *** EPK36-J-12 59-62 20.3 *** EPK31-P-40 27-35 13.8 *** EPK31-P-40 35-45 14.7 *** EPK31-P-40 45-50 16.0 *** EPK31-P-40 50-62 18.2 *** EPK31-P-40 62-65 16.8 *** EPK30-V-6 16-22 5.0 ** *** EPK30-V-6 22-24 3.8 ** *** EPK30-V-6 24-27 3.0 **** EPK30-V-6 30-35 5.1 ** * * *** * EPK30-V-6 35-39 10.1 *** * ** * EPK30-V-6 39-43 6.9 *** ** * EPK30-V-6 43-48 4.8 *** ** EPK30-V-6 48-53 9.0 *** ** EPK30-V-6 53-58 8.9 *** EPK30-V-6 58-63 10.5 *** EPK30-V-6 63-68 13.8 *** EPK30-V-6 68-73 13.0 *** EPK30-V-6 73-78 11.9 *** FRG-1 1 14.0 *** * * FRG-1 2 10.4 *** ** * * FRG-1 3 9.3 *** * * * FRG-1 4 1.0 *** ** * * FRG-1 5 1.7 *** * * ** FRG-1 6 1.8 *** FRG-1 7 3.7 *** FRG-1 8 11.5 *** FRG-1 9 2.7 *** * FRG-1 10 34.2 *** * FRG-1 11 9.3 *** FRG-1 12 8.3 *** FRG-1 13 7.5 *** FRG-1 14 9.4 *** FRG-1 15 2.5 *** FRG-2 1 8.7 *** * * * FRG-2 2 4.7 *** * * * FRG-2 3 6.6 *** FRG-2 4 2.0 *** FRG-2 5 10.4 *** FRG-2 6 8.3 *** FRG-2 7 6.6 *** FRG-2 8 5.6 *** FRG-2 9 4.4 *** FRG-2 10 8.6 *** * FRG-2 11 5.4 *** * FRG-2 12 10.3 *** * FRG-2 13 10.2 *** * * R (*) Note: Mineral component designations correspond to the following semi-quantitative scheme: *** = major (> 50%), ** = minor (10-50%), and * = trace (< 10%). R = rutile, S = smectite, A = anatase, B = boehmite, D = diaspore, Gr = goethite

Table 4-2. – (continued).

Crandallite- Sample ID Interval % Clay Kaolinite Gibbsite Halloysite Illite HIV Quartz Other Florencite Cypresshead Formation - FL (cont.) FRL-1 1 19.5 *** FRL-1 2 10.6 *** FRL-1 3 9.7 *** FRL-1 4 10.6 *** FRL-1 5 1.7 *** FRL-1 6 9.2 *** FRL-1 7 13.9 *** FRL-1 8 9.5 *** FRL-1 9 10.5 *** SSJ-1 1 16.9 *** R (*) SSJ-1 2 17.5 *** * R (*) SSJ-1 3 10.6 *** * SSJ-1 4 11.0 *** * SSJ-1 5 2.0 *** SSJ-1 6 8.8 *** SSJ-1 7 10.0 *** SSJ-1 8 9.9 *** SSJ-1 9 8.5 *** SSJ-1 10 10.6 *** SSJ-1 11 12.8 *** TRF2214 60.0-62.5 --- *** ** S (**) WEX164 18.0-26.0 --- *** * * S (**) WEX366 9.0-10.0 --- *** * ** Reworked Cypresshead Formation - FL SSD-1 1 8.3 ** * * *** * A (**)/R (*) SSD-1 2 3.3 *** ** ** B (*)/D (*) SSD-1 3 3.0 *** * * SSD-1 4 3.0 *** ** * SSD-1 5 4.7 *** * * SSD-1 6 6.5 *** SSD-1 7 3.6 *** SSD-1 8 6.7 *** SSD-1 9 5.2 *** SSD-1 10 4.9 *** Cypresshead Formation - GA J-1 1 19.6 *** ** ** Gr (*)/R (*) J-1 2 27.8 *** * * Gr (*) J-1 3 19.1 *** * * Gr (*) J-1 4 63.0 *** * Gr (*) J-1 5 12.3 *** * Gr (*) J-1 6 6.3 *** * LB-1 1 20.4 *** * * LB-1 2 24.6 *** * * LB-1 3 15.6 *** * LB-1 4 12.4 *** LB-1 5 2.4 *** LB-1 6 63.0 *** LB-1 7 2.2 *** B-1 1 26.9 *** ** Gr (*) B-1 2 33.3 *** ** * Gr (*) B-1 3 27.6 *** ** * Gr (*) B-1 4 33.4 *** ** Gr (*) B-1 5 24.0 *** ** Note: Mineral component designations correspond to the following semi-quantitative scheme: *** = major (> 50%), ** = minor (10-50%), and * = trace (< 10%). R = rutile, S = smectite, A = anatase, B = boehmite, D = diaspore, Gr = goethite consistent with what would have been originally deposited with the unit prior to weathering, or the clays may have been derived locally through reworking of proximal Hawthorn Group

sediments which would have been exposed adjacent to these locations. The trace illite noted for

FRG-2-13 is most likely reworked from Hawthorn Group sediments immediately below the sampling interval.

The mineralogy of Georgia clay size-fractions differ slightly from those in Florida, with the absence of halloysite and crandallite, and the addition of significant illite concentrations at both the Jesup type locality and the Birds site. Based on evaluation of the 060 reflection, this illite is of the 2M1 polytype and therefore of detrital origin. Goethite is also a common trace component in southeastern Georgia Cypresshead sediments (Jesup and Birds), and reflects higher iron (Fe) concentrations associated with the deposition of these sediments. Additionally, vertical trends in the clay mineral suite suggest an in situ weathering origin for several clay size-fraction phases, favoring the formation of HIV and gibbsite in the unsaturated zone and halloysite under saturated conditions.

Facies Architecture

Pirkle (1960), Kane (1984) and Huddlestun (1988) have made past attempts at describing the lithofacies which compose the Cypresshead Formation in north-central Florida and Georgia based primarily on field observations. In this study, information from field observations were combined with a detailed evaluation of the sedimentary framework of the unit to arrive at the facies outlined in Table 4-3. The distribution of stratigraphic discontinuities was then used to define facies boundaries and corresponding facies architecture as shown in Figures 4-8 and 4-9.

A detailed evaluation of the facies associated with reworked Cypresshead sediments from the

Davenport Sand Mine core (SSD-1) was beyond the focus of this study, but appears, based on limited data, to mimic the characteristics of the Cypresshead Formation.

Table 4-3. Facies summary of the Cypresshead Formation in north-central peninsular Florida and southeastern Georgia.

Facies Lithology Sedimentary structures Bioturbation

North-central Florida Upper shoreface (USF) Quartz sand, medium to coarse, with minor gravel Tabular and trough cross-bedding (10-60 cm) dominant Dense to sparse, consisting of throughout. Sand is slightly clayey to clayey near the top except where bioturbation or weathering has destroyed Ophiomorpha spp., Skolithos spp. and due to illuviation primary sedimentary structure (massive bedding) undiff. traces with escape structures Proximal lower shoreface (pLSF) Quartz sand, medium to coarse, with gravel throughout. Trough cross-bedding (10-40 cm) transitions into 0.5-3 m Dense near top, consisting of Sand is slightly clayey, with clay more common as stringers tabular cross-bedding associated with the migration of Ophiomorpha spp. and Skolithos spp. and lenses. dune (megaripple) bedforms traces Distal lower shoreface (dLSF) Quartz sand, medium to fine, with interbedded coarse sand Horizontal bedding with 10-40 cm trough cross-bedding Minor bioturbation consisting of and gravel. Sand is slightly to moderately clayey. Minor interbedded with 0.3-1 m hummocky stratification and Thalassinoides spp. and Skolithos spp. feldspar (<5%) and mica storm derived coarse sand and gravel graded beds traces Offshore transition (OST) Quartz sand, medium to fine, is the dominant lithic Horizontal to massive bedding is dominant unknown component. Sand is clayey, with up to 20% or more clay. Minor feldspar (< 10%) and mica Offshore inner shelf (OSI)* Quartz sand, medium to fine, is the dominant lithic Massive bedding which appears to be devoid of primary unknown component. Sand is fossiliferous, variably calcareous, and sedimentary or biogenic structures sometimes clayey. Mollusks are the dominant fossil type Southeastern Georgia 98 Proximal lower shoreface (pLSF) Quartz sand, medium to coarse, with minor gravel Tabular and trough cross-bedding (10-60 cm) dominant Moderate to sparse near top, consisting throughout. Sand is slightly clayey to clayey near the top except where weathering has destroyed primary of undiff. traces due to illuviation. Clay common as stringers and lenses. sedimentary structure (massive bedding) Distal lower shoreface (dLSF) Quartz sand, medium to fine, with interbedded coarse sand Horizontal to massive bedding is dominant Minor bioturbation consisting of and gravel. Sand is moderately clayey. Minor mica Ophiomorpha spp. traces

Offshore transition (OST) Quartz sand, fine, is the dominant lithic component. Sand is Horizontal to massive bedding is dominant Moderate to sparse bioturbation, clayey, with up to 30% or more clay. Clay common as consisting of undiff. traces stringers and partings. Minor mica Offshore inner shelf (OSI)* Quartz sand, medium to fine, is the dominant lithic Massive bedding which appears to be devoid of primary unknown component. Sand is fossiliferous, variably calcareous, and sedimentary or biogenic structures sometimes clayey. Mollusks are the dominant fossil type * Corresponds to the Nashua Formation of Huddlestun (1988) and Scott (1992a), which is considered a facies of the Cypresshead Formation for this study.

Figure 4-8. Correlation of Cypresshead Formation facies in north-central Florida.

Figure 4-9. Correlation of Cypresshead Formation facies in southeastern Georgia.

100

North-central Florida facies

In north-central Florida, a total of five facies were identified based on sedimentological factors which define a single coarsening-upward cycle consistent with the progradational siliciclastic deposition of a shoreface-shelf parasequence. Included with these facies is the

Nashua Formation of Huddlestun (1988), which has been previously defined by both Huddlestun

(1988) and Scott (1992a) as an offshore facies of the Cypresshead Formation in both north- central Florida and southeastern Georgia. For the purpose of this study, the Nashua is assigned to the offshore inner shelf (OSI) facies and defines the basal facies unit of the Cypresshead

Formation for both north-central Florida and southeastern Georgia.

Continuing upsequence, the second facies identified in north-central Florida is an offshore transition (OST) facies for which only a small section may be exposed at the base of the Grandin

Sand Mine (FRG-2-12?). However, it is this facies, which appears to correlate to the kaolinitic sands associated with the EPK Mine orebody. Based on sedimentological factors such as total clay and mica content, apparent horizontal to massive bedding characteristics, and a downdip position of the kaolinitic sand orebody relative to the Grandin Sand Mine exposure, a transitional offshore environment is likely. Cores sampled from this location were sited at surface elevations of 27 m to 28 m above msl, with the kaolinitic orebody encountered an additional 7.5 m to 9 m below the surface. This places the top of the orebody at no greater than 20.5 m above msl, an elevation which would be consistent with the positioning of an OST facies relative to the elevation-facies depth characteristics noted for Grandin Sand Mine exposures (Fig. 4-8). General characteristics of the OST facies include quartz sand in the medium to fine mean grain-size range, with a minor coarse sand component and up to 20% or more clay (< 2 µm) content. In north-central Florida, the clay content of this facies occurs primarily as a binding matrix.

101

Additionally, as much as 10% of the sand fraction may consist of feldspar and/or mica, with both phases concentrated in the medium and fine sand fraction.

The third facies, identified as a distal lower shoreface (dLSF) facies, is characterized by medium- to fine-grained quartz sands which commonly exhibit hummocky cross-stratification, and are interbedded with graded and trough cross-bedded storm-derived coarse sands with minor gravel (Figs. 4-8 and 4-10A). Bioturbation associated with this facies is sparse, and is dominated by Thalassinoides spp. and Skolithos spp. traces. Clay content of the dLSF facies can be as much as 10% or more in the medium to fine sand component where it occurs as a binding matrix.

Additionally, minor mica and/or feldspar are common, and may represent up to 5% or more of the sand fraction, the mica content being primarily a function of hydrodynamic sorting while the feldspar content is most likely related to preservation relative to weathering. Lastly, grain-size distributions for this facies are mainly unimodal, consistent with deposition below normal wave base (Balsillie, 1995).

The proximal lower shoreface (pLSF) facies is well developed in north-central Florida mine exposures due in no small part to the relative concentration of coarse sands associated with the large- to medium-scale tabular cross-stratification generated by the migration of dune (or megaripple) bedforms in this environment (Fig. 4-8, Fig. 4-10B). This facies is characterized by an increase in quartz sand grain-size (medium to coarse) with commonly associated gravel, a reduction in the overall clay content except for the occurrence of discrete clay lenses and stringers, and the extensive development of cross-stratification. The base of the facies is defined by an erosional disconformity which indicates a significant increase in scouring and sediment transport velocity, which most likely corresponded to a strong southerly oriented longshore current (Fig. 4-5D). During periods of reduced current flow, clay stringers and lenses could have

102

Figure 4-10. Examples of Cypresshead Formation facies. A) Exposure (~ 2 m) in the Grandin Sand Mine illustrating the vertical juxtaposition of dLSF, pLSF and USF facies (large-scale tabular cross-stratification is not developed in this section), B) FRL-1 exposure (~ 13 m) exhibiting well developed tabular cross-stratification associated with the pLSF facies, C) Upper portion of the B-1 exposure (~ 2.5 m) illustrating the OST facies, D) J-1 exposure (~ 5.5 m) illustrating an example of the dLSF and pLSF facies, with the J-1-4 clay bed (arrow) indicated for reference.

developed in response to quiescent conditions, with reactivation of current activity corresponding to the incorporation of clay rip-up clasts in overlying sediments. Based on the bimodal character of grain-size distributions for the pLSF facies, energies associated with current sorting and transport were mixed with wave, and possibly tidal, influences. Additionally, the top of this facies commonly exhibits dense bioturbation associated with Ophiomorpha spp. and Skolithos spp. traces (Fig. 4-10A), marking a non-depositional discontinuity between the pLSF facies and the overlying facies.

103

The last facies, the upper shoreface (USF) facies, is characterized by medium to coarse quartz sands with minor gravel, which exhibit well developed small- to medium-scale tabular and trough cross-stratification consistent with a southerly directed wave and current flow orientation (Figs. 4-5C and 4-10A). Many coarse beds exhibit graded bedding, and bioturbation varies from dense to sparse, dominated by Ophiomorpha spp., Skolithos spp. and undifferentiated traces. Additionally, sands are generally slightly clayey, except near the top of exposures where illuviation has resulted in the post-depositional concentration of clays.

Southeastern Georgia facies

Cypresshead Formation exposures in southeastern Georgia are not as well developed as what is available in north-central Florida (i.e., sand mines). As a result, only four facies have been identified for the Cypresshead in this region (Table 4-3), with two of the exposures used in this study exhibiting only a single facies (Fig. 4-9). However, as is seen in north-central Florida, these facies define a single coarsening-upward cycle consistent with a progradational shoreface- shelf parasequence.

As with Florida, the basal facies of the Cypresshead Formation in southeastern Georgia, the OSI facies, corresponds to the Nashua Formation per the previous discussion. Upsequence of the OSI facies, the OST facies shares many similarities with the description of the facies in

Florida, including a high clay content, sometimes exceeding 30%. However, as exposed at Birds, the facies in Georgia appears to be generally finer grained, consisting of often thinly-bedded fine quartz sand with discrete layers or partings of clay, and a significant clay matrix component (Fig.

4-10C). Dominated by horizontal bedding, the OST facies exhibits evidence of sparse to moderate bioturbation, and a notable lack of current or wave generated scouring or bedforms.

In southeastern Georgia, approximately 1.5 m of the dLSF facies is exposed at the Jesup type locality (Figs. 4-9 and 4-10D). Sedimentological characteristics in Georgia are similar to 104

that seen in Florida, with the facies characterized by medium- to fine-grained quartz sands with

interbedded coarse sand and gravel, although the mean grain-size of the facies is coarser in

Georgia. This coarser substrate is reflected in the presence of Ophiomorpha spp. traces rather

than Thalassinoides spp. as is observed for the dLSF facies in Florida.

Lastly, similar to what is seen in Florida, the pLSF facies in southeastern Georgia is characterized by medium to coarse quartz sands with minor gravel throughout. Clay content in

the facies is concentrated at the top by illuviation, but exhibits a significant matrix component (5-

20%) and concentrated deposition of clays as stringers, lenses and beds (Fig. 4-10D).

Additionally, the facies possesses well developed tabular and trough cross-stratification indicating significant current and/or wave action consistent with a dominant southerly flow direction at the type locality and an easterly direction at Linden Bluff. Clay rip-up clasts are common in cross-bedded sands overlying clay lenses or beds (Fig. 4-9).

Discussion

Depositional Environment

It is probable that nearshore conditions not unlike those seen along the modern Georgia

and Florida east coasts are responsible for deposition of the Cypresshead Formation, except for

the fact that climate, and consequently its impact on sea-level and sedimentation, differed from what is experienced today. As a result, a review of modern coastal conditions offers the most suitable basis for evaluating Cypresshead depositional environments.

The present coastline of Georgia and extreme northern Florida in the vicinity of the

Georgia Bight is characterized by barrier islands broken by tidal inlets (Kussel and Jones, 1986).

Impacted by semidiurnal tides with an average range of 2 m, spring tidal ranges up to 3 m, and a southerly longshore current (Davis et al., 1992), this coastline is consistent with a mesotidal mixed energy system. Further south, along the remainder of eastern Florida, are marine

105

depositional coasts dominated by barrier beaches, barrier islands, splits, and overwash fans

(Schmidt, 1997). This modern wave-dominated to mixed energy east coast barrier system is bounded by a relatively narrow and steep sloping shelf, and is impacted by wave energies associated with a mean annual wave height of up to 0.7 m (Nummedal et al., 1977). Longshore transport is dominantly to the south as a result of strong northerly winds associated with winter frontal systems that dominate longshore currents, even though prevailing winds have a distinct southerly component (Davis et al., 1992).

Cypresshead Formation model

Based on the collective sedimentological and mineralogical characteristics outlined for this study of the Cypresshead Formation in both north-central Florida and southeastern Georgia, the unit was deposited in a nearshore marine setting as two distinct shoreface-shelf parasequences, one of which was in a wave-dominated environment (north-central Florida) and the other in a mixed energy environment (southeastern Georgia). In both instances, deposition took place in response to relative sea-level fall, resulting in a single coarsening-upward cycle of siliciclastics.

Internally within the unit, this model for deposition is expressed by the presence of clinoform surfaces within the unit that dip gently at 2-3° to the east in northern-central Florida and southeastern Georgia. The most likely nearshore marine environment for Cypresshead deposition appears consistent with a strand plain setting, lacking well-developed lagoons or marshes.

In north-central Florida, the Cypresshead Formation thins toward the west onto the flanks of the Ocala Platform where it is absent (Scott, 1988a). Thus, the Ocala Platform acts as a depositional basin divide between the Cypresshead Formation to the east and the time equivalent

Miccosuckee and Citronelle Formations to the west. Deposition of the Cypresshead would have commenced during a sea-level highstand, with the westernmost, updip extent of the unit in close proximity to the Trail Ridge and Highland sample locations discussed in this study. This 106

location, proximal to exposed uplands consisting of Hawthorn Group (i.e. smectite- and illite-

bearing) sediments, along with the subsequent burial of these sediments by early Pleistocene

Trail Ridge sediments helped preserve the clay mineral suite noted herein. Whether a similar clay mineral suite was initially deposited elsewhere with Cypresshead sediments in north-central

Florida is uncertain, but it is probable that the original suite varied considerably from the highly

weathered, kaolinite-dominated suite seen today.

Subsequent deposition in north-central Florida would have been dictated by relative

regressing sea-level, sediment supply, current positioning, and available accommodation. In

north-central Florida, the latter of these would have been related to antecedent topography

developed on top of Hawthorn Group (Coosawhatchee Formation) sediments. Dictating available

accommodation for Cypresshead siliciclastics, this surface is known to have been scoured by

pre-Cypresshead erosion and submarine current activity, creating irregularities which would

have acted as preferred nearshore conduits for siliciclastic transportation by longshore current

and storm activity, or topographic lows for the preferential deposition of hydrodynamically

sorted fines (e.g. clays and mica).

As indicated by the well developed cross-stratification and overall grain-size of

Cypresshead siliciclastics, strong longshore current activity, significantly greater than seen along

the modern Florida and southeastern Georgia coasts, is proposed as a major factor in the coast

parallel transportation of these sediments. Evidence for similar processes extending down the

Florida peninsula are noted from an outcrop described by Johnson (1989) from Lady Lake pit in

Lake County, Florida, which contains two cross-bedded, coarse-grained sand beds (Beds 3 and

4), four and three feet thick, which appear to be similar to the dune (or megaripple) bedforms

(Figs. 4-4B and 4-10B) described at in the Grandin and Goldhead exposures. In fact, estimates

107

by Kane (1984) for current velocities based on the maximum grain-sizes and sedimentary

structures noted in Cypresshead sediments indicate values of 40–70+ cm/sec, which are

consistent with strong longshore current activity capable of transporting even the largest fraction

(i.e., discoid quartzite pebbles) noted for the Cypresshead. As indicated by Dobkins and Folk

(1970), the development of a discoid pebble shape, as seen in Cypresshead sediments, is

consistent with a high wave-energy shoreface environment. Under such conditions, the discoid

(oblate) shape can be generated through abrasion as pebbles slide back and forth over sand or

smaller pebbles in the surf zone. Although Pirkle et al. (1964) and others have argued against

using the shape of quartzite pebbles solely as an indicator of depositional environment, noting

that some component of shape is likely influenced by the inherited textural and structural

characteristics, the proposed current velocities of Kane (1984) help to explain their occurrence in

Cypresshead sediments.

Based on facies characteristics discussed in this study (e.g., dLSF gravels and hummocky

cross-stratification), storm-induced sedimentation appears to have played a significant role in

deposition of Cypresshead sediments in north-central Florida. In fact, support for storm-induced

depositional process in Florida playing a significant role during the mid-Pliocene is found

elsewhere in the observations of Allmon (1992) and Missimer (2001a;b), as they correlate the

deposition of massive shell beds from the Tamiami Formation to storm processes occurring in

shallow coastal waters at that time. Hummocky cross-stratification in the lower portion of the

Grandin Sand Mine exposures is further support for significant levels of storm activity, as it occurs in repetitive successions separated by horizontally bedded clayey fine sands. The hummocky beds themselves are relatively devoid of burrows, with evidence of sparse

Thalassinoides spp. burrows limited to the interbedded horizontal clayey fine sands. This appears

108

consistent with multiple storm-generated depositional events being separated by periods of

quiescent deposition during which burrowing would have been favored.

Further supporting the view of a high-stress depositional environment for the Cypresshead

Formation is the low diversity macrobenthic infaunal trace fossil assemblage outlined in this

study. Characteristic of shallow-water facies with relatively coarse substrates, this assemblage

possesses relatively low diversity in comparison to the modern shelf off south-central Texas

(Hill, 1985), but is similar to that described by Kussel and Jones (1986) for the late Pleistocene

to Holocene Satilla Formation deposited seaward of the Cypresshead in southeastern Georgia

and northernmost Florida. Although Ophiomorpha spp. and Skolithos spp. traces are dense

within select beds correlated to nondepositional discontinuities, the overall density of traces appears to decrease significantly with increased water depth. This is inconsistent with the opinion of Howard and Reineck (1972) that under ideal conditions bioturbation should increase

with increasing water depth, due to fewer bottom disturbances such as storms. Possible causes

for this trend might include deeper water wave base impacts of storms or nutrient conditions

unsuitable for supporting a more dense and/or diverse assemblage.

For southeastern Georgia, the nearshore marine depositional model proposed by this study differs from that applied in Florida only with respect to differences in coastal environment

(wave-dominated vs. mixed energy) and timing. For Cypresshead sediments deposited proximal

to the riverine discharge of siliciclastics along the southeastern Georgia coast, detrital clay

content is expected to be greater than the current and wave winnowed sediments transported

down the Florida coastline. Additionally, the potential for a fluvial-deltaic component associated

with deposition is more likely as reflected in the flow orientation of cross-stratification in Linden

Bluff sediments (Fig. 4-5B). However, once siliciclastics reached the coastline, southerly

109

directed longshore currents would have driven coast parallel migration of sediments. Huddlestun

(1988) used a depositional model with a very broad sound, at least 80 km wide, to explain many of the sedimentological characteristics described in this study, but a mesotidal coast (2-4 m tidal range) characterized by the development of short (or stunted) barriers cut by common tidal inlets describes the spatial distribution of Cypresshead sediments in Georgia as well.

If significant sediment fluxes are necessary to explain the overall volume of Cypresshead and related siliciclastics deposited on the Florida Platform, then total sediment loads carried by rivers to the Georgia coastline would have been large during periodic storm and flooding related erosional events. As such, homopycnal or even hyperpycnal river plumes would have been possible (Addington et al., 2007; Summerfield and Wheatcroft, 2007), favoring nearshore deposition of bed and suspended loads, including muds, which storm, wave, and longshore currents would then have moved southward along the coastline. Under such conditions favorable to nearshore mixing, salinities proximal to river outflows were likely brackish, favoring the occurrence of Ammonia beccarii and Elphidium spp. seen in Cypresshead sediments from

Effingham County, Georgia (Huddlestun, 1988). Additionally, high physico-chemical stresses in such a setting would have resulted in a paucity of infaunal burrowing as is noted for the

Cypresshead Formation in southeastern Georgia.Furthermore. However, rather than indicating an enclosed sound environment as suggested by Huddlestun (1988), such conditions would have favored river outflow along a strand plain or mixed barrier-estuary coastline significantly different than the modern mesotidal Georgia coast of today.

Corresponding to the downdip facies of Huddlestun (1988), and exhibiting similarities to the Miccosukee Formation in southwestern Georgia and the eastern panhandle of Florida as noted by Huddlestun (1988), the OST facies in Georgia may possess, at least in part, a prodeltaic

110

origin. Argued by Scott (2001) as the depositional origin of the Miccosukee Formation, a prodeltaic origin for the OST facies would be consistent with the well-sorted sands, discrete occurrence of layers or laminae of white to gray clay, overall high clay content, lack of scoured

contacts, and sparse to moderate bioturbation observed with the facies. As such, it appears

reasonable to propose a prodeltaic origin for the OST facies seen at the Birds site. This would be

consistent with the general spatial distribution of this facies as described by Huddlestun (1988)

as being distal from major rivers presently exiting along the Georgia coastline. If the coarser

sand lithologies observed at the type locality (J-1) and the Linden Bluff site (L-1) are to be

considered as proximal to points of coastal sediment delivery, then it is consistent for the OST

facies at Birds to have a prodeltaic origin, with deposition occurring distally from river outflows.

Model consistency and the siliciclastic conveyor

Other than exhibiting evidence for far more post-depositional weathering, Cypresshead

sediments share significant mineralogical similarities with Late Miocene through Pliocene south

Florida siliciclastics, particularly with respect to the occurrence of feldspar and mica noted in

both. Cunningham et al. (1998) first noted minor feldspar (≤ 5%) and mica, with the particle

shape of the dominant quartz fraction angular to subrounded, which is also a shared

characteristic of Cypresshead sediments. Additionally, it is recognized as a consequence of this

study that the Cypresshead Formation contains two shoreface-shelf parasequences marking

major episodes of shoreline regression during the Pliocene. Other than for timing, this is similar

to the models that have been proposed for Late Miocene through Pliocene siliciclastics in

southern Florida. Based on such a correlation, the lack of Late Miocene through early Pliocene

preservation of siliciclastics in southeastern Georgia and north-central Florida point to the

preferred mobilization and coast-parallel transport of these sediments southward as the

“siliciclastic conveyor” of Cunningham et al. (2003). Throughout the Late Miocene to latest 111

Early Pliocene, siliciclastic deposition dominated the eastern Florida Platform at the expense of

carbonate deposition (Scott, 1988b; Cunningham et al., 1998; 2001; 2003). In response to a rise

in Pliocene eustatic sea-level and/or loading-based subsidence, mixed carbonate and siliciclastic

deposition returned to the southern peninsula throughout the remainder of the Pliocene

(Missimer, 1992; Cunningham et al., 2001; 2003), while siliciclastic deposition continued to dominate to the north in response to retrogradation. This resulted in the deposition of the

Cypresshead Formation in north-central Florida and southeastern Georgia.

Similar to what has been noted for the Cypresshead Formation, the spatial record of Late

Miocene through Pliocene siliciclastic deposition in southern Florida was also controlled by the interplay of sediment supply, relative sea-level, current activity, and antecedent accommodation

(Guertin et al., 1999). As illustrated by Cunningham et al. (1998), a shallow marine basin or low at the top of the Arcadia Formation existed from west of Lake Okeechobee to the Florida Keys, which acted as a preferred conduit of accommodation for siliciclastics transported southward by longshore currents and storms. During periods of reduced siliciclastic flux, sediments would have been subjected to current reworking and winnowing, imparting characteristic sorting and gradation to these sediments, consistent with the observations of both Guertin et al. (1999) and

Cunningham et al. (1998; 2003).

Overall, the Late Miocene to Pliocene siliciclastic deposition from south Florida to southeastern Georgia represents a retrogradational parasequence set with individual parasequences progradational in character as a result of regression-based deposition.

Retrogradational stacking results when the long-term rate of accommodation exceeds the long- term rate of sedimentation. In this way, accommodation space is created more rapidly than it is filled, water depth becomes deeper, and facies increasingly move farther landward. Although

112

each parasequence is shallowing-upward, the amount of deepening at the flooding surface exceeds the amount of shallowing in the following parasequence, producing a net overall deepening within the parasequence set. A consequence of this observation is that it points to a period of subsidence as a consequence of Paleogene through Early Neogene sediment loading of the Florida peninsula, prior to the Pleistocene uplift noted by Opdyke et al. (1984).

In considering consistency for the Late Miocene through Pliocene siliciclastics that compose the retrogradational sequence tract of which the Cypresshead Formation is a part, depositional processes should be considered. Of particular interest is the assigning of a fluvial- deltaic model to the deposition of the SS2 siliciclastics of Cunningham et al. (2003). Whereas interest in assigning an alluvial or fluvial-deltaic origin to Cypresshead sediments extends back to the work of Bishop (1956), Pirkle (1960), and Pirkle et al. (1964), Kane (1984) successfully challenged this perception on the basis of paleocurrent and facies analysis, deposit geometry, and the distribution of marine paleontological evidence. However, similarities between the

Cypresshead and the predominately alluvial Citronelle Formation to the west have kept the alluvial model for the Cypresshead and similar sediments alive as evidenced by the fluvial- deltaic interpretation for the SS2 siliciclastics (Cunningham et al., 2003).

The interpretation of the SS2 sequence as representing a fluvial-deltaic depositional system was based on seismic reflection geometries, thus suggesting the presence of a N-S oriented fluvial system along the Florida Platform responsible for deposition in a nearshore, marine-to- brackish environment. As noted, such a fluvial-deltaic model has been suggested by previous authors for other Florida siliciclastics (Bishop, 1956; Pirkle, 1960; Pirkle et al., 1964; Klein et al., 1964; Peacock, 1983; Missimer and Maliva, 2006), but was first challenged by Winker and

Howard (1977b), who noted the unlikelihood of a major river flowing the length of the Florida

113

Platform rather than seeking a more direct route to the sea. Furthermore, evidence reported by

Cunningham et al. (2003) does not seem to support fresh to brackish water conditions one would associate with a deltaic complex. Rather, foraminiferal data for SS2 sediments correspond to a shallowing upward sequence from outer-shelf to inner-shelf and possibly lagoon environments, which can be interpreted as favoring the prograding spit model proposed by Ginsburg et al.

(1989) and Warzeski et al. (1996). In fact, Ginsburg et al. (1989) sought to combine both fluvial- deltaic and spit models, suggesting that siliciclastics were transported southward by a combination of longshore and riverine processes, eventually to be redistributed by wave and current activity to form a giant progradational spit with sediments extending as far south as the

Florida Keys. Warzeski et al. (1996) proposed that a pathway of maximum paleocurrents and a coincident trend of coarse-grained siliciclastics trend down the peninsula. This trend, later refined by Cunningham et al. (1998), offers a means by which sea-level fluctuations, strong currents, and storms could have redistributed siliciclastics to the south and east (Guertin et al.,

1999).

Further difficulty with employing the fluvial-deltaic model to SS2 sediments is the failure of Cunningham et al. (2003) to provide evidence for a riverine source of sediments landward (to the north) of the SS2 sediments. In fact, the observation by Bishop (1956) of “deltaic beds”, which is used to support the hypothesis of Cunningham et al. (2003), actually corresponds to sediments of the Cypresshead Formation near the SSD-1 and SSJ-1 drill hole sites used in this study, and are known today to be cross-bedded sands similar to those of marine origin noted in this study at several exposures in north-central Florida (FRG and FRL). Also, as noted by

Cunningham et al. (2003), the discoid pebbles within SS2 sediments are consistent with deposition in a beach environment (Dobkins and Folk, 1970). These observations fit the

114

spit/nearshore marine depositional model more suitably than the fluvial-deltaic model of

Cunningham et al. (2003).

Timing and Regional Stratigraphic Correlation

Based on field observations, sequence stratigraphic analysis, and a review of the works of

Huddlestun (1988), Cunningham et al. (1998; 2001; 2003), Guertin et al. (1999), and others on the late Neogene siliciclastics of Florida and Georgia, the timing for deposition and reworking of the Cypresshead Formation was determined for this study. The results are summarized in Figure

4-11 relative to the Berggren et al. (1995a; b) time scale, and constrain the age of the

Cypresshead Formation deposition to about 3.4 to 2.3 Ma and significant reworking of the unit to about 2.8 to 1.8 Ma.

Pre-Cypresshead siliciclastic flux

As noted previously, the initial onset of the siliciclastic flux ultimately associated with the

Cypresshead Formation is first seen with deposition of the Late Miocene SS2 siliciclastics of

Cunningham et al. (2003). Initially deposited over a broad swath of southern Florida beginning at

8.6 Ma (Fig. 4-12), this sequence is equivalent, in part, to the Interval I siliciclastics (6.2 – 5.5

Ma) of Guertin et al. (1999). This latter unit has been included with the upper Peace River

Formation for this study (Fig. 4-11), and is tentatively correlated to bracket the TB3.3-TB3.4 boundary of supercycle TB3 of Haq et al. (1988) corresponding to the 5.7 Ma sea-level fall of

Miller et al. (2005) (Fig. 4-13). Increased siliciclastic sedimentation during this interval appears to be coincident with the late Miocene intensification of the Florida Current-Gulf Stream system

(Mullins et al., 1980; Eberli et al., 1997), which, in turn, is linked to a eustatic lowstand correlated to late Miocene glaciation on Antarctica. With lowering of sea-level under conditions of increased aridity in the southeastern United States (Alt, 1974), a significant volume of siliciclastics would have been generated in response to renewed incision and erosion of early to

115

Figure 4-11. Correlation chart for Late Miocene through Pliocene siliciclastic units evaluated in this study. Correlation is to the geomagnetic polarity time scale of Berggren et al. (1995a; b), with calcareous nannoplankton zones according to Okada and Bukry (1980) and Bukry (1991).

middle Miocene Coastal Plain sediments and Appalachian sources. Subsequent mobilization, transport and deposition is reflected in the accumulation of the more than 100 m of siliciclastics in southernmost Florida associated with the lower SS2 (SS2-E) and Interval I sequences.

The succeeding TB3.4 early Pliocene transgression, believed to have occurred between 5.3 to 4.9 Ma (Willard et al., 1993; McNeill et al., 2001; Miller et al., 2005), marks a transition toward a more restricted depositional pattern beginning at ~5 Ma. As illustrated in Figure 4-12,

116

Figure 4-12. Maps illustrating the aerial extent of successive siliciclastic deposition events impacting the Florida Platform from 8.6 Ma to 1.8 Ma. A) SS2-E and Interval I, B) SS2-W and Interval II, C) Cypresshead Formation, D) Cypresshead Formation (Georgia only), reworked Cypresshead sediments, and Interval III.

117

Figure 4-13. Relationship between the Cypresshead Formation, related siliciclastics in southern Florida, seismic sequence boundaries of Eberli (2000), sequence chronostratigraphy, and the sea-level curves of Haq et al. (1988) and Miller et al. (2005). The Haq et al (1988) eustatic curve is adjusted to the Berggren et al. (1995b) time scale (modified after Cunningham et al., 2003).

the overall reduction in sediment supply is expressed by the restriction of upper SS2 (SS2-W)

siliciclastics to southwestern Florida, with a corresponding absence in the southernmost

peninsula, and the sourcing of reworked Miocene sediments as a dominant component of Interval

II siliciclastics (Guertin et al., 1999; Cunningham et al., 2003). This observation appears not to be a response to a reduction in sediment flux being delivered to the Florida peninsula, but rather represents retrogradation to the north as a function of rising Pliocene sea-level or subsidence in response to sediment loading in southern Florida. Further supporting this view of a restriction in sediment supply is the aggradational accumulation of carbonate-rich sediments during this period, including the Ochopee Limestone Member of the Tamiami Formation in southwest

Florida (Cunningham et al., 2001).

118

During the TB3.4 transgression, which appears to have flooded the Florida Platform,

siliciclastics to the north would have continued to be mobilized as was the case during the Late

Miocene (Guertin et al., 1999). Subsequent regressions associated with sea-level falls at 4.9 Ma

and 4.0 Ma (Fig. 4-13; Table 4-4) would have then transported the sediments southward,

resulting in the initial deposition of SS2-W and Interval II siliciclastics bracketing the TB3.4 –

TB3.5 transition, with cessation of deposition occurring during the TB3.6 cycle at about 3.6 Ma

(Guertin et al., 1999). Confirmation for the timing of SS2-W deposition during this Early

Pliocene (Zanclean) interval is based on diatoms that indicate an age of less than 5.5 Ma, and coccoliths characteristic of Zones CN10c-11 (Fig. 4-12) (Cunningham et al., 2003).

Cypresshead deposition and reworking

Further retrogradation in association with a continued rise in sea-level or subsidence following deposition of SS2 and Interval II siliciclastics is required to deposit the updip sediments which correspond to the Cypresshead Formation. Therefore, given the regression- based depositional characteristics of the unit, it is likely that Cypresshead siliciclastics in central and north-central Florida correlate to the TB3.6 – TB3.7 boundary (Fig. 4-13), suggesting an early Late Pliocene (Piacenzian) age (3.4–2.8 Ma) consistent with the 3.3 Ma sea-level fall of

Miller et al. (2005) (Figs. 4-11 and 4-13; Table 4-4). This correlates closely with the most recent early Late Pliocene (3.4-2.7 Ma) age estimate for the Citronelle Formation (Otvos, 1998b), and also supports correlation of the unit, in part, with the timing of siliciclastic deposition associated

with the Tamiami Formation (SS3/Pinecrest Sands) in southwest Florida (Missimer, 2001a; b;

Cunningham et al., 2001; 2003) (Fig. 4-11). However, it should be noted that correlation of the

Cypresshead to the Haq et al. (1988) sea-level curve should be considered tentative due to poor

age resolution of the siliciclastics and uncertainty in the age resolution and amplitude of the curve. 119

Table 4-4. Correlation of siliciclastic depositional events on the Florida Platform with Haq et al. (1988) sequence boundaries, sequence boundaries of Eberli (2000), and sea-level falls identified by Miller et al. (2005).

Sequence Boundary Age (Ma) Miller et al. Third-order Florida Platform Wornardt et al. (2005) Sea-level cycles Sen et al. (1999)a Eberli (2000)b Stratigraphic Event (2001)a Falls (Ma) N/A ------2.2/2.1/1.9 CH reworking/III deposition TB 3.8 2.4 2.50 --- 2.5 CH deposition/reworking TB 3.7 3.1 3.21 3.1 3.3 CH/SS3 deposition TB 3.6 3.9 3.95 3.6 4.0 ↑ TB 3.5 4.1 4.37 ------SS2-W/II deposition N/A ------4.9 ↓ TB 3.4 5.6 5.73 5.4 5.7 SS2-E/I deposition TB 3.3 7.1 6.98 ------a Based on revised Haq et al. (1988) seismic sequence boundaries corrected to the Berggren et al. (1995b) time scale. b ODP Leg 166 seismic sequence boundaries identified from Great Bahama Bank drilling results.

Supporting evidence for the Cypresshead Formation being related to a continuation in retrogradation similar to that recorded in SS2 siliciclastics (SS2-E to SS2-W) of Cunningham et al. (2003) was first noted by Klein et al. (1964), who observed that sediments now associated with the Cypresshead in Highlands County, Florida, appear to extend southward into the subsurface of Glades County. This observation was later reinforced by the belief of Scott (2001) that the Cypresshead appears to grade downdip into siliciclastics of the Long Key Formation of

Guertin et al. (1999) in southern Florida. However, based on timing and the relative elevation of the base of the Cypresshead, it seems more likely that any contact with the Long Key Formation

(Interval II) or upper Peace River SS2 siliciclastics is disconformable or paraconformable in nature (Fig. 4-12), favoring a retrogradational hypothesis. This position is corroborated by the age results from Wingard et al. (1994) and Weedman et al. (1995), which support partial correlation of Long Key Formation sediments with the upper Peace River Formation in central

Florida. As the top of SS2 siliciclastics are located near present day sea-level, and little epeirogenic uplift is expected during the Quaternary near 27˚N latitude in the vicinity of the

Highlands-Glades County border (Winker and Howard, 1977; Opdyke et al., 1984), sea-level at the time of SS2 deposition was most likely slightly to moderately above present day levels

120

relative to the present vertical positioning of the Florida Platform, with Interval II siliciclastics of the Long Key Formation deposited under deeper marine conditions. This would require ongoing subsidence of central and southern Florida to permit retrogradation and deposition of

Cypresshead sediments in a nearshore marine setting through the development of updip accommodation, as the middle Pliocene sea-level rise (25–30 m) which culminated at 3.3 Ma was likely of a lower magnitude than that associated with deposition of SS2 and Interval II sediments based on the Haq et al. (1988) curve. Even with the accepted uncertainties in the absolute magnitudes of these sea-level highstands as seen in comparisons to the Miller et al.

(2005) curve (Fig. 4-13), retrogradation would remain a prerequisite for Cypresshead deposition.

The TB3.6 sea-level highstand associated with deposition of the Cypresshead Formation is well expressed along the southeastern Atlantic coastal plain. In particular, this highstand is

responsible, in part, for eroding the Orangeburg scarp in Georgia (Dowsett and Cronin, 1990)

concurrent with deposition of the Raysor Formation and unnamed Raysor-equivalent shelly sand

of Huddlestun (1988). Given the occurrence of planktonic foraminiferal assemblages in each of

these units which correlate to zone PL3 of Berggren (1973), or PL4 in the case of the unnamed

Raysor-equivalent shelly sand, these units are early Late Pliocene (early Piacenzian to

Piacenzian) in age, and limit the initial onset of Cypresshead deposition in southeastern Georgia

(Fig. 4-11). In the case of the Raysor Formation, which disconformably or paraconformably

underlies the Cypresshead Formation in Georgia, the unit most likely corresponds to a condensed

section deposited during the TB3.6 (and potentially TB3.7) transgression episode preceding the

regression-based deposition of the Cypresshead. The same is likely for the unnamed Raysor-

equivalent shelly sand, although the potential foraminiferal correlation to PL4 of Berggren

(1973) suggests the possibility that the unit is instead representative, in part, of an offshore (inner

121

to middle shelf) facies of the Cypresshead Formation, similar to the relationship proposed by

Huddlestun (1988) and Scott (1988a) for the Nashua Formation in northern Florida. This

contention that the Nashua correlates, in part, with the Cypresshead Formation in northern

Florida and may represent an offshore facies of the unit appears consistent with field

observations and depositional timing for both units. This idea was built upon the earlier

observation of Pirkle (1960) that Cypresshead sediments overlay post-Hawthorn, sandy shell-

bearing marls (Nashua Formation) along the eastern flank of the Lake Wales Ridge topographic

trend in northern Florida. However, timing of deposition for the actual type section of the

Nashua as early Pleistocene (Calabrian) is inconsistent with foraminiferal assemblages collected

from two Florida Geological Survey core sites (Cassidy 1: W-13815 and Baywood 1: W-8400)

which penetrate the Nashua in Nassau and Putnam Counties proximal to north-central Florida exposures of the Cypresshead. These assemblages, collected from core depths consistent with the

position of a potential downdip Cypresshead equivalent facies, correlate with Zone PL5 of

Berggren (1973) (Huddlestun, 1988). Although Huddlestun (1988) proposed a multideposit

origin of the Nashua to explain this contradiction, it appears more likely that the “Nashua”

sediments sampled from the two cores are simply representative of an offshore (inner shelf)

facies of the Cypresshead Formation, and should be included with the unit as discussed

previously, while the type section is a discrete deposit unrelated to the Cypresshead.

Palynological data reported by Hansen et al. (2001) from the Peace Creek site sinkhole in

Polk County, Florida, constrains the minimum age of the Cypresshead in central Florida. These

sediments correlate with reworked Cypresshead sediments that would have accumulated during

the late Pliocene TB3.7 – TB3.8 transition at 2.8–1.8 Ma (Haq et al., 1988) which correlates to

the 2.5 Ma sea-level fall of Miller et al. (2005) (Fig. 4-13; Table 4-4). Marking the onset of

122

significant Eurasian and North American glaciation (Hansen et al., 2001), deposition of these sediments would have likely taken place near or even below modern sea-level given the present

Peace Creek elevation of 34 m and the minimum estimated Quaternary uplift of 36 m calculated by Opdyke et al. (1984) for this region of Florida. Given that marine Cypresshead Formation sediments occur at elevations above approximately 52 m in central peninsular Florida, a sea-level highstand of at least 16 m above present msl would be required during deposition, a figure well within the magnitude predicted by both the Haq et al. (1988) and Miller et al. (2005) curves.

Thus, based on the constraints placed on depositional timing by the Raysor Formation, the unnamed Raysor-equivalent shelly sand, reworked Cypresshead sediments, and south Florida siliciclastics, along with strong evidence for correlation of the Cypresshead Formation with the

Citronelle Formation in the Florida panhandle, deposition of the Cypresshead in central and north-central Florida appears to have occurred between 3.4–2.8 Ma (Figs. 4-11 and 4-12).

In southeastern Georgia, deposition of the Cypresshead Formation appears to be restricted to the late Pliocene (late Piacenzian to early Gelasian) from about 3.15–2.3 Ma. This conclusion is based on a Raysor-equivalent benthic foraminiferal assemblage recovered by Huddlestun

(1988) from basal sediments of the unit in southeastern Georgia (Chatham County), and the noted correlation of north Florida and southeastern Georgia Cypresshead sediments to those of the “Nashua” sediments discussed previously. Such timing correlates with the TB3.7 – TB3.8 boundary (Haq et al., 1988) or the 2.5 Ma sea-level fall of Miller et al. (2005), which is associated with a period of Cypresshead reworking in central Florida. Although some deposition likely occurred in response to the regression following the sea-level fall at 3.3 Ma, Cypresshead siliciclastics in southeastern Georgia (and potentially portions of north-central Florida) appear to be primarily deposited following further retrogradation north in response to a continuation of

123

subsidence in central Florida and the infilling of available accommodation to the south. This

observation supports the contention that the Cypresshead Formation, as had been proposed by

Huddlestun (1988) for his Nashua Formation, is a multideposit unit (i.e., it was deposited during

more than one episode of deposition), a theme reflected in many of the Neogene deposits of

Florida (Tamiami, Caloosahatchee, etc.). As for subsequent reworking of the Cypresshead in southeastern Georgia, that likely continued throughout the remainder of the Pliocene in concert with reworking in Florida (Figs. 4-11 and 4-12). Evidence for this is seen in the timing of deposition for Caloosahatchee Formation siliciclastics and Interval III of Guertin et al. (1999), the latter of which began to accumulate during TB3.8 near the end of the Pliocene (~2 Ma) and ceased deposition in the area of the Florida Keys near the Pliocene-Pleistocene transition. As

Interval III appears to be sourced from Cypresshead reworking, the timing for the cessation of deposition at ~1.8 Ma likely correlates with the end of major Cypresshead reworking in both

Florida and Georgia (Fig. 4-11). Continued accumulation of siliciclastics seen in the

Caloosahatchee Formation are likely more restricted to local reworking and appear to cease during the early Pleistocene (Calabrian) (Fig. 4-11).

Paleoclimate Forcing of Cypresshead Deposition

Significant accumulations of siliciclastics into a given deposystem can be attributed to both

regional tectonic factors and paleoclimate (Molnar and England, 1990; Zhang et al., 2001; Harris

and Mix, 2002). In the case of the Florida Platform, paleoclimatic transitions toward periods of

instability and the corresponding changes in temperature, precipitation and vegetation appear to

be the dominant driving mechanism behind the anomalous accumulations of siliciclastics

associated with the conveyor model of Cunningham et al. (2003). Two such episodes are

believed to have impacted the depositional timing of Cypresshead and related Late Miocene

siliciclastics; (1) the transition from arid conditions during the Late Miocene to continual El Niño 124

conditions during the early Pliocene warm period (~4.5–3.0 Ma), and (2) the transition from continual El Niño conditions to a period of global cooling and the onset of significant Northern

Hemisphere Glaciation (NHG) (~3.0–1.5 Ma). Combined with the Late Eocene through

Oligocene rejuvenation of the Appalachians (Stuckey, 1965; Dennison and Stewart, 2001;

Stewart and Dennison, 2006) which ultimately provided the sediment source and potential energy for erosion, such periods of transitional paleoclimate would have resulted in landscape disequilibrium, thereby enhancing erosion and resultant sedimentation rates as has been noted to occur globally during the middle to late Pliocene (~4–2 Ma) (Zhang et al., 2001).

Middle to late Miocene climate and sediment supply

Two factors dictated by mid-Miocene paleoclimate were likely prerequisites for the conveyor model of Cunningham et al. (2003) to function; intense weathering and increased aridity. The first of these, intense, even lateritic, weathering, has long been proposed to have occurred in the southeastern United States (Cooke and MacNeil, 1952; Isphording, 1970; 1971), and appears to correlate worldwide with the development of deep weathering sequences during the middle to late Miocene (10-13 Ma) (Ashley and Silberman, 1976; Alpers and Brimhall, 1988;

Vasconcelos et al., 1994a; b). The second, a worldwide transition toward arid conditions during the late Miocene (Yemane et al., 1985; van Zinderen Bakker and Mercer, 1986; Vasconcelos et al., 1994b) is supported by evidence for the increased supply of terrigenous dust in oceanic sediments (Ruddiman et al., 1989; Rea et al., 1990) and by the expansion of arid climate flora and fauna (Tedford, 1985; Wolfe, 1985; Yomane et al., 1985; Axelrod and Raven, 1986). In fact, the Miocene ended with the driest climate of the Tertiary (both regional and global), and was accompanied by conversion of savanna to steppe or scrub desert, spread of C4 grasses, and the greatest mammal extinction of the Neogene (Chapman, 2008). Together with Appalachian uplift and/or crustal arching, Miocene weathering and aridity are responsible for producing the 125

sediment reservoir required to supply the siliciclastic conveyor model during the Late Miocene

and Pliocene.

Increased aridity likely facilitated erosion and subsequent deposition of late Miocene

through Pliocene siliciclastics in two ways; (1) through a reduction in overall vegetative cover,

and (2) via an increase in the seasonality of precipitation. Fluvial transport is known to be minimal in both extremely arid and peri-humid (everwet) climates under equilibrium, with recent studies indicating that climates characterized by strong seasonality of annual precipitation are

capable of generating a significant sediment flux (Edgar and Cecil, 2003). Under such

conditions, the effect of vegetation on erosion is minimized, leaving soil vulnerable to rainsplash

and sheetwash erosion, which favors highly competent runoff and high rates of denudation (Alt,

1974). Additionally, arid conditions may have increased the relative magnitude of rare floods, or

conversely, increased the frequency of large floods (Molnar, 2001). As a consequence, incision

rates would have been significant, despite a decrease in total precipitation and discharge. If the

climate instability associated with the transition toward early Pliocene warm period conditions is

then superimposed on an arid Late Miocene landscape, the siliciclastic flux directed toward the

Florida platform would have been further enhanced given that relatively modest shifts in climate

can have large impacts on the sediment yield of rivers (Syvitski, 2004). As noted by Goodbred

and Kuehl (2000) for the Ganges-Brahmaputra, a period of several thousand years of intensified

discharge can lead to significant continental margin deposition (e.g. 50 m along the Bengal

margin).

Pliocene climate and the transition toward Northern Hemisphere Glaciation (NHG)

Several authors have proposed that paleoclimate during the early Pliocene warm period

(~4.5–3.0 Ma) was consistent with continual El Niño conditions in the tropics rather than

intermittent as is seen today (Ravelo and Wara, 2004; Fedorov et al., 2006). Under such 126

conditions, regional climate in the southeastern U.S., including Florida and Georgia, would have been characterized by cooler temperatures, above average rainfall, a reduction in hurricane activity, and an increase in winter storm activity (cyclogenesis). The latter of these factors would have contributed to an overall increase in nearshore erosion and transport during this interval, but the relative climate stability during this time would have reduced the delivery of siliciclastics to the coast. With onset of the climatic instability associated with the transition toward significant

Northern Hemisphere Glaciation that occurred after ~3.0 Ma, the siliciclastic flux to the coast would have increased as part of the global increase in sedimentation rates and grain sizes noted by Zhang et al. (2001) for the middle to late Pliocene (~2–4 million years ago).

Zhang et al. (2001) and Molnar (2004) postulate that an overall increase in the amplitude and frequency of climate change is the root cause for the increase in both sedimentation rates and sediment coarsening observed during the Pliocene as a result of a persistent state of disequilibrium in sediment source regions. Since the response time of landscapes dominated by fluvial incision is longer than the periodicity of major climate fluctuations during this interval

(Hancock and Kirwan, 2007), upland sediment sources remain in disequilibrium, driving the increased siliciclastic flux toward the coast. Increased fluvial incision rates are likely to be magnified during regressive periods of base-level lowering (Mills, 2000), which coincide with overall late Miocene through Pliocene cooling. Along the eastern North America margin, sedimentation rates are known to have doubled between the late Miocene and the Quaternary in offshore basins (Poag and Sevon, 1989), with Matmon et al. (2003) indicating that basin- averaged erosion rates of ~30 m/m.y. are common in the southern Appalachians during this time.

Evidence for increased current and storm activity

Coincident with deposition of the Cypresshead Formation, multiple climate models suggest that higher annual sea-surface temperatures and reduced ice cover in the Northern Hemisphere 127

likely led to intensification of the Icelandic low-pressure system and the Azores high-pressure

system (Dowsett et al., 1994; Haywood et al., 2000; Jiang et al., 2005). As a consequence,

corresponding strengthening of westerly wind velocities and wind stress over the North Atlantic

Ocean would have enhanced the flow of the Gulf Stream, with evidence of enhanced oceanic

upwelling along the east coast of North America and southwestern Florida during this time

(Cronin and Dowsett, 1990; Cronin, 1991) consistent with this result. Additionally,

intensification of the Azores high likely resulted in a strengthening of easterly Trade Winds and

increased tropical storm activity, with the later resulting in the Florida coast experiencing more

hurricanes than present (Hobgood and Cerveny, 1988; Barron, 1989). Storm activity associated

with either hurricanes or winter storms, perhaps at increased levels of intensity and/or

occurrence, has been used to explain the concentration of Pliocene shell beds associated with the

Pinecrest Sand (Allmon, 1992), a time equivalent of the Cypresshead Formation in Florida.

With a proposed increase in the strength of prevailing winds impacting the east coast of

Florida, conditions favorable to a substantial increase in longshore transport likely persisted

during the Pliocene. This contrasts with modern conditions along the east coast of Florida which

experience a mean significant wave height during winter months of only 1.2 m and a relatively

low frequency of high-intensity storms consistent with a microtidal, wave-dominated coast

(Davis et al., 1992). Although these conditions are capable of a relatively high gross longshore

transport (~600,000 m3/yr), there is relatively low net transport (50,000-150,000 m3/yr) along the

east coast of Florida due to the seasonally bimodal wave climate that impacts the coastline

(Walton, 1976; Davis et al., 1992). During late Miocene through Pliocene warming events,

intensified wave, storm, and frontal activity may have acted as major forcing mechanisms for

increased coast parallel siliciclastic fluxes, with increased mean wave height and stronger

128

longshore transport acting to mobilized and transport significant volumes of siliciclastics concentrated by fluvial processes along the southeastern Georgia coast. Such conditions would

not only explain the quantity of siliciclastics deposited on the Florida Platform, but also the

relative coarseness of these sediments, including discoid pebbles, and the common occurrence of storm generated sedimentary features.

Conclusions

Focusing on the stratigraphy and sedimentology of the Cypresshead Formation, the results outlined in this chapter highlight significant observations which clarify the nature, timing and significance of siliciclastic deposition impacting the Florida Platform during the Late Miocene through Pliocene. Among the results are the following:

• Cypresshead sediments were deposited in a nearshore marine environment, most likely in a strand plain setting, as two distinct progradational shoreface-shelf parasequences.

• Cypresshead facies define coarsening-upward sequences consistent with a wave-dominated environment in north-central Florida and a mixed energy environment in southeastern Georgia.

• Deposition of the Cypresshead took place in response to sea-level falls at 3.3 Ma and 2.5 Ma as a consequence of the interplay of sea-level, sediment supply and accommodation.

• Deposition in Florida at 3.4–2.8 Ma with reworking at 2.8–1.8 Ma.

• Deposition in Georgia at 3.15–2.3 Ma with reworking at 2.3–1.8 Ma.

• Timing of Cypresshead deposition at 3.4–2.3 Ma during the Late Pliocene (Piacenzian to early Gelsian) correlates with age estimates of the Citronelle Formation (3.4–2.7 Ma) as defined by Otvos (1988b) and timing of siliciclastic deposition associated with the Tamiami Formation (SS3/Pinecrest Sands).

• Viewed collectively with the Late Miocene SS2 siliciclastics of Cunningham et al. (2003), Cypresshead and associated siliciclastics define a retrogradational parasequence which was deposited on the Florida Platform over a 6.8 Ma period.

• Cypresshead deposition correlates with a paleoclimate transition from continual El Niño conditions associated with the Pliocene warm period (~4.5–3.0 Ma) to conditions associated with the onset of significant Northern Hemisphere Glaciation (NHG) (~3.0–1.5 Ma).

129

• The driving mechanism behind the anomalous accumulation of siliciclastics associated with the Cypresshead and related Late Miocene siliciclastics (SS2) is the shift from periods of climate stability to periods of climate transition (instability) characterized by changes in temperature, precipitation and vegetation.

130

CHAPTER 5 EVIDENCE FOR NEOFORMATION AND RECRYSTALLIZATION OF KAOLINITE IN THE CYPRESSHEAD FORMATION

Introduction

As previously noted in this study, the kaolinitic clays present in the Cypresshead

Formation occur as irregular thin beds, lenses, and stringers of clay, or as a binding matrix for sands and gravels, with clay contents that may vary from absent to > 50% (Pirkle, 1960; Kane,

1984; Huddlestun, 1988; Scott, 1988a). Although much of the kaolinitic clay possesses obvious sedimentological characteristics indicative of a detrital or sedimentary origin, discrepancies in grain-size relationships (i.e. clays with coarse sands and gravels) and overall distribution of clay in the formation continue to raise questions regarding the potential in situ or weathering origin for some portion of the clay mineral assemblage. While a detrital origin for much of the clay content in Cypresshead Formation sediments is likely, Austin (1998) has suggested an important role for groundwater leaching of aluminous components (mica, feldspar) in the formation of an in situ kaolinite fraction, with the associated removal or preservation of Fe and organic compounds dependent on groundwater acidity, redox, and biological content. This view is similar to that proposed for Georgia-South Carolina kaolins by Hurst and Pickering (1997), and reflects the significant role that post-depositional alteration plays in maturing kaolin assemblages through a complex series of early diagenetic and weathering reactions that increase overall kaolinite content of sediments over time.

Arguments supporting the in situ formation of the kaolinitic sediments of the Cypresshead

Formation were first introduced by Sellards (1912), who considered the kaolinite to have formed by the weathering of arkosic sands. The main aim of this proposed mechanism for kaolinite origin was to explain the presence of fine clay mixed with coarse sand and gravel, evidence of which is noted by the high positive skewness values reported for Cypresshead sediments in

131

Chapter 4 of this study. However, Sellards’ (1912) model still does not fully explain the distribution of all fines mixed with coarse sand and gravel, particularly as relates to the occurrence of muscovite throughout these sediments, a considerable quantity of which has been identified in the 44 to 10 µm range (Pirkle, 1960). As a result, this study seeks to define the detailed role of weathering reactions in the overall modification of Cypresshead clays, and the likelihood for an in situ kaolinite fraction in response to both transformational and neoformational weathering processes in order to constrain observed grain-size anomalies. The reader is referred to Chapter 2 of this study for a more detailed review of the previous arguments associated with the detrital versus in situ origin of Cypresshead Formation clays.

Results

X-ray diffraction (XRD) analysis of clay (< 2 µm) separates obtained from the

Cypresshead Formation in both Florida and Georgia (Fig. 3-1) was performed in order to fully characterize the detailed clay mineralogy of the formation, something which has already been done for Georgia-South Carolina kaolin district kaolinites (Keller, 1977; 1978; Hurst et al., 1979;

Hassanipak and Eslinger, 1985; Pickering and Hurst, 1989; and others). Both oriented and randomly mounted clay mineral samples were prepared and analyzed in order to define qualitative sample mineralogy and to characterize kaolinite disorder and crystallite size characteristics. Scanning electron microscope (SEM) analysis was aimed at discerning microtextural characteristics of kaolinites and associated sediments in order to facilitate recognition of kaolinite micromorphologies of detrital and in situ origin. XRD data for the samples used in this study (oriented and random) are included with Appendices E and F.

Mineralogy

A table outlining the clay mineralogy of Cypresshead Formation samples is included in

Chapter 4 for reference (Table 4-2). Based on this data, kaolinite is the dominant clay mineral in

132

both Cypresshead Formation and reworked Cypresshead Formation sediments. In Florida, the

clay (< 2 µm) size-fraction also includes lessor amounts of quartz, gibbsite, halloysite,

metahalloysite (inferred), hydroxyl-interlayered vermiculite (HIV), and crandallite-florencite,

with additional occurrences of rutile, anatase, boehmite, and diaspore. Georgia clay mineralogy

differs slightly, with the absence of halloysite and crandallite, and the addition of significant illite

at both the Jesup type locality and the Birds site. Based on evaluation of the 060 reflection, this

illite is of the 2M1 polytype and therefore of detrital origin. Goethite is also a common trace

component in both Jesup and Birds samples, and reflects higher iron (Fe) concentrations

associated with the deposition of these sediments.

Clay mineralogical trends of note in north-central Florida include the overlapping occurrence of gibbsite and halloysite at the near surface. Forming in proximity to fluctuating water table conditions, gibbsite (above) and halloysite (below) likely crystallize at the expense of allophane and amorphous Al hydroxide gels. Gibbsite occurs as either a trace or minor constituent in near surface Cypresshead sediments, with the formation of gibbsite favored by well drained soil conditions consistent with what is observed with the Entisols and Ultisols of the

Cypresshead (Heuberger, 1995). Where present, the phase is commonly concentrated in the near surface zone of pedogenic clay accumulation via illuviation. For north-central Florida sampling locations, gibbsite is notably absent in the Goldhead Sand Mine and Joshua Sand Mine exposures. In Georgia, gibbsite is absent only from the Birds locality.

The occurrence of halloysite in Cypresshead sediments, noted as a trace clay mineral phase

in north-central Florida samples, was confirmed by XRD and SEM (Figs. 5-1 and 5-2). Based on

heat treatment testing for halloysite, loosely bound interlayer water is lost, resulting in structural

collapse of the basal (d001) spacing from 10 Å to 7 Å, consistent with the formation of

133

Figure 5-1. XRD patterns of example Cypresshead Formation clays. A) Sample FRG-1-5 containing halloysite which is partially collapsed when air dried, B) Sample EPK36- J-12 (25-27) containing halloysite which requires completed heat treatment to collapse structure to 7 Å basal spacing, C) Comparison of north-central Florida (EPJ36-J-12 (48-50)) and southeastern Georgia (L-1-6) kaolinite, illustrating significant differences in disorder, D) Vertical variation in kaolinite disorder (G = gibbsite; Q = quartz; C = crandallite-florencite).

metahalloysite (halloysite-(7 Å)) as defined by Hurst and Pickering (1997). Behavior related to dehydration varied, with some samples exhibiting water loss with air drying (Fig 5-1A), while others required full-cycle heating to fully collapse the structure (Fig. 5-1B). The distribution of halloysite was restricted to near surface samples in close proximity to the modern water table, with water table levels closer to surface elevation at the EPK Mine site than what is observed at the Grandin Sand Mine.

Halloysite morphology in Cypresshead sediments is similar with that observed in other southeastern U.S. kaolin deposits, possessing a tubular morphology consistent with formation 134

Figure 5-2. SEM photomicrographs illustrating characteristic secondary weathering phases and textures. A) halloysite (EPK30-V-6 (22-24)), B) crandallite-florencite (FRG-1-5), C) etched quartz (EPK30-V-6 (48-53)), D) skeletal K-feldspar grain (SSJ-1-11).

under low Fe conditions (Joussein et al., 2005). Tubes are generally 1 μm in length and 0.2 μm in diameter, and appear to be crystallized on the basal surface of weathered muscovite mica (Fig. 5-

2A), a common association for halloysite neoformation (Robertson and Eggleston, 1991; Singh and Gilkes, 1992).

Consistent with the ubiquitous character of HIV in sandy soils along the southern U.S. coastal plain (Harris et al., 1992), HIV occurrences in the Cypresshead Formation are restricted to the near surface, and apparently above recent water table levels. Additionally where noted as a minor rather than trace phase (J-1), HIV concentration is at the top of the stratigraphic section,

135

suggesting that the phase is preferentially concentrated in upper sections of the unit most effected

by pedogenic processes. As noted by Harris et al. (1992) and Heuberger (1995), HIV forms

preferentially in the soil profile from sand- and silt-sized mica precursor grains which retain their

position in the profile while colloidal components such as kaolinite illuviate to deeper

weathering zones. Thus, HIV appears to be the preferred vadose weathering product for the near

surface detrital mica component of the Cypresshead Formation.

The occurrence of a crandallite-florencite phase ((Ca,REE)Al3(PO4)2(OH)5·H2O) was limited to Cypresshead and reworked Cypresshead sediments in north-central Florida. Known to occur in near surface sediments of central Florida as a secondary weathering product of phosphatic (francolite-bearing) clayey sediments (Blanchard, 1972), crandallite-florencite is primarily associated with near surface Cypresshead samples, except for the Grandin Sand Mine exposures (FRG-1 and FRG-2), where crandallite is also noted at depth. Of note, crandallite- florencite was identified near both the top and bottom of FRG-2. Crandallite-florencite observed in Cypresshead samples possesses a lath or acicular morphology (Fig. 5-2B) commonly noted in weathered phosphorites (Flicoteaux and Lucas, 1984).

Other phases of interest, which are not necessarily included with the clay fraction, are quartz, mica, and K-feldspar. Sand-size quartz, the dominant lithic component associated with

Cypresshead sediments varies in grain shape from angular to subrounded, and exhibits significant evidence of etching when viewed via SEM (Fig. 5-2C). This indicates extreme leaching conditions in Cypresshead sediments in response to weathering, and corroborates the evidence for silica mobilization, a factor previously indicated by the occurrence of gibbsite in near surface samples. Clay-size quartz also occurs in Cypresshead sediments, but was limited to north-central Florida sites, with EPK30-V-6 exhibiting the most significant concentrations to a

136

depth of ~ 16 m (53 ft). In other instances, clay-size quartz was isolated to near surface

occurrences as has been noted previously for the Cypresshead in Florida (Heuberger, 1995).

Feldspars, confirmed as K-feldspar (microcline) via energy dispersive X-ray spectroscopy

(EDS), petrographic examination, and XRD, were found to occur at both Florida and Georgia

sampling locations. The distribution of K-feldspar is addressed in Chapter 4, and appears to be

most significantly impacted by preservation potential in relation to weathering. When evaluated by SEM, K-feldspars were found to possess skeletal microtextures consistent with significant dissolution activity (Fig. 5-2D), with no evidence for pseudomorphic replacement of feldspar grains by kaolinite.

Kaolinite Disorder and Crystallite Size

Random clay mineral aggregates were prepared and analyzed by XRD in order to define kaolinite disorder using computer-based methods to describe the disorder occurring in the kaolinite structure (i.e. enantiomorphic distortions and displacement of octahedral Al vacancies), and to interpret both qualitatively and quantitatively the total density of stacking faults in the crystallite ensemble (Hinckley, 1963; Bish and Von Dreele, 1989; Plançon and Zacharie, 1990;

Artioli et al., 1995). To accomplish this, the Hinkley Index (HI), the Liètard Index (R2), and the

“expert system” of Plançon and Zacharie (1990) were calculated (Table 5-1). The HI is the most widely used of these measurements, and evaluates changes in the 02l and 11l peaks of kaolinite

(20-30 °2θ using CuKα), which are sensitive to random and specific interlayer displacements of

type b/3 (Aparicio and Galán, 1999). Normal HI values range from <0.5 (disordered) to 1.5

(ordered). The R2, on the other hand, measures changes in the 13l peak sequence of kaolinite

(37-40 °2θ using CuKα), which are affected by random displacements (Cases et al., 1982;

Aparicio and Galán, 1999). Reported R2 values normally range from <0.7 (disordered) to 1.2

137

Table 5-1. Results of disorder calculations for north-central Florida Cypresshead and reworked Cypresshead kaolinite.

Mineral Hinkley Index Liètard Index Expert System Parameters Sample ID Interval Interferences (HI) (R2) # Phases M Wc d p %ldp Cypresshead Formation - FL EPK36-J-12 25-27 H/Q/G 0.31 0.78 1 28 0.00 0.04 0.35 EPK36-J-12 27-30 H/Q/G 0.86 0.90 2 30.66 EPK36-J-12 35-40 H/Q/G 0.69 0.85 2 21.94 EPK36-J-12 40-44 1.16 0.94 2 46.21 EPK36-J-12 44-46 1.17 0.91 2 47.22 EPK36-J-12 46-48 1.10 0.90 2 45.08 EPK36-J-12 48-50 1.17 0.90 2 42.78 EPK36-J-12 50-53 1.16 0.89 2 43.79 EPK36-J-12 53-56 1.13 0.85 2 42.97 EPK36-J-12 56-59 1.05 0.82 2 38.68 EPK36-J-12 59-62 1.07 0.81 2 38.54 EPK31-P-40 27-35 1.02 0.88 2 38.61 EPK31-P-40 35-45 1.20 0.93 2 47.20 EPK31-P-40 45-50 1.23 0.89 2 49.77 EPK31-P-40 50-62 0.97 0.78 2 38.30 EPK31-P-40 62-65 0.64 0.65 2 21.73 EPK30-V-6 16-22 H/Q ------EPK30-V-6 22-24 H/Q ------EPK30-V-6 24-27 Q ------EPK30-V-6 30-35 H/Q/G 0.55 ------EPK30-V-6 35-39 Q/G 0.34 0.68 1 30 0.00 0.04 0.35 EPK30-V-6 39-43 Q 0.51 0.73 2 10.79 EPK30-V-6 43-48 Q 0.56 0.78 1 38 0.00 0.03 0.24 EPK30-V-6 48-53 0.54 0.75 2 14.49 EPK30-V-6 53-58 1.24 0.78 2 47.19 EPK30-V-6 58-63 1.30 0.93 2 51.43 EPK30-V-6 63-68 1.36 0.94 2 55.16 EPK30-V-6 68-73 1.23 0.70 2 49.93 EPK30-V-6 73-78 0.91 0.71 2 42.42 FRG-1 1 G 0.54 0.89 2 14.67 FRG-1 2 HIV/G 0.56 --- 2 14.04 FRG-1 3 HIV/G 0.56 0.85 1 34 0.00 0.03 0.29 FRG-1 4 HIV/G 0.64 --- 1 27 0.00 0.04 0.27 FRG-1 5 H/G 0.23 0.74 ------FRG-1 6 1.19 0.99 1 45 0.00 0.01 0.08 FRG-1 7 1.13 0.89 2 43.06 FRG-1 8 0.62 0.77 2 19.79 FRG-1 9 1.27 1.02 2 54.48 FRG-1 10 1.19 0.87 2 47.21 FRG-1 11 1.45 1.04 2 65.46 FRG-1 12 1.51 1.09 2 69.11 FRG-1 13 1.56 1.07 2 67.25 FRG-1 14 1.56 1.06 2 69.71 FRG-1 15 1.49 1.05 2 62.04 FRG-2 1 HIV/G 0.52 0.73 2 10.57 FRG-2 2 HIV/G 1.03 1.01 1 41 0.00 0.02 0.12 FRG-2 3 1.39 0.97 2 56.33 FRG-2 4 1.34 0.95 2 56.60 FRG-2 5 1.49 1.03 2 62.96 FRG-2 6 1.49 1.02 2 65.25 FRG-2 7 1.49 1.04 2 65.46 FRG-2 8 1.51 1.10 2 67.11 FRG-2 9 1.43 1.00 2 62.68 FRG-2 10 1.32 0.97 2 54.86 FRG-2 11 1.20 0.88 2 48.23 FRG-2 12 0.99 0.84 2 34.97 FRG-2 13 0.92 0.79 2 10.96 Note: For expert system calculations, %ldp = percentage of low-defect kaolinite, M = average layer number in the coherent domains along the c-axis, Wc = percentage of layers with vacant octahedral position, δ = small random translations between adjacent layers, and p = great translation defects between adjacent layers. Mineral interferences include: quartz (Q), halloysite (H), hydroxyl-interlayered vermiculite (HIV), gibbsite (G), and anatase (A).

138

Table 5-1. – (continued).

Mineral Hinkley Index Liètard Index Expert System Parameters Sample ID Interval Interferences (HI) (R2) # Phases M Wc d p %ldp Cypresshead Formation - FL (cont.) FRL-1 1 1.01 0.82 2 36.55 FRL-1 2 1.10 0.85 2 41.56 FRL-1 3 1.11 0.86 2 38.37 FRL-1 4 1.08 0.86 2 40.07 FRL-1 5 1.23 0.96 2 44.46 FRL-1 6 1.14 0.88 2 46.60 FRL-1 7 1.15 0.85 2 44.93 FRL-1 8 1.11 0.82 2 44.57 FRL-1 9 1.00 0.82 2 36.52 SSJ-1 1 0.82 0.67 2 13.50 SSJ-1 2 0.86 0.75 2 26.25 SSJ-1 3 0.94 0.83 2 31.90 SSJ-1 4 1.13 0.93 2 43.27 SSJ-1 5 1.28 0.99 2 49.34 SSJ-1 6 1.20 0.84 2 49.22 SSJ-1 7 1.26 0.89 2 51.45 SSJ-1 8 1.24 0.89 2 50.74 SSJ-1 9 1.24 0.91 2 50.12 SSJ-1 10 1.13 0.89 2 45.58 SSJ-1 11 1.06 0.82 2 41.45 Reworked Cypresshead Formation - FL SSD-1 1 Q/HIV/G/A ------SSD-1 2 Q/G 0.49 0.88 ------SSD-1 3 G 0.68 0.88 2 33.24 SSD-1 4 G 0.69 0.91 2 34.32 SSD-1 5 G 1.13 0.88 2 46.42 SSD-1 6 1.08 0.74 2 41.18 SSD-1 7 1.06 0.67 2 40.87 SSD-1 8 0.86 0.63 2 30.00 SSD-1 9 0.61 0.55 2 14.01 SSD-1 10 0.46 0.50 1 21 0.00 0.05 0.31 Cypresshead Formation - GA J-1 1 HIV/G 0.69 ------J-1 2 HIV/G 0.42 ------J-1 3 HIV/G 0.50 ------J-1 4 I 0.29 0.51 ------J-1 5 I 0.43 0.53 ------J-1 6 I 0.24 0.46 ------LB-1 1 HIV/G 0.18 0.45 ------LB-1 2 HIV/G 0.15 0.45 ------LB-1 3 G 0.29 0.46 ------LB-1 4 0.35 0.46 1 21 0.00 0.05 0.40 LB-1 5 0.43 0.55 1 24 0.00 0.05 0.32 LB-1 6 0.28 0.46 ------LB-1 7 0.38 0.58 1 25 0.00 0.05 0.35 B-1 1 HIV/I ------B-1 2 HIV/I ------B-1 3 HIV/I ------B-1 4 I ------B-1 5 I ------Note: For expert system calculations, %ldp = percentage of low-defect kaolinite, M = average layer number in the coherent domains along the c-axis, Wc = percentage of layers with vacant octahedral position, δ = small random translations between adjacent layers, and p = great translation defects between adjacent layers. Mineral interferences include: quartz (Q), halloysite (H), hydroxyl-interlayered vermiculite (HIV), gibbsite (G), and anatase (A).

(ordered). An alternative to these methods, the expert system of Plançon and Zacharie (1990),

uses multiple measurements from the XRD pattern to describe kaolinite defects, and provides an

abundance estimate of translation defects and the potential for a two-phase kaolinite mixture.

139

Accessory phases encountered with Cypresshead samples which are capable of interfering with

disorder calculations include halloysite, quartz, gibbsite, HIV, and anatase (Table 5-1).

Based on XRD results, significant differences exist between the disorder characteristics of

north-central Florida and Georgia Cypresshead kaolinites (Fig. 5-1C; Table 5-2), with Florida

Cypresshead samples consistent with ordered (low-defect) kaolinite similar to what is seen with

the Cretaceous “soft” kaolins form Georgia, and Georgia Cypresshead samples exhibiting

disordered (high-defect) characteristics. For Florida samples, including SSD-1 reworked

Cypresshead sediments, HI and R2 values range from 0.23 to 1.56 and 0.50 to 1.10, respectively, with corresponding means of 1.03 and 0.86. For Georgia samples, HI and R2 values range from

0.15 to 0.69 and 0.45 to 0.58, respectively, with means of 0.36 and 0.49. In the case of the

distribution of HI values illustrated in Figure 5-3A, Florida samples exhibit a bimodal character,

with one group of HI values corresponding to disordered kaolinite values of <0.7, and a second group possessing moderately to well ordered HI values >0.8. Georgia samples do not exhibit this characteristic, but rather are restricted to disordered HI values <0.7. R2 results, believed to be sensitive to random defects only (Cases et al., 1982), do not exhibit a bimodal character for either the Florida or Georgia samples (Fig. 5-3B).

Box-and-whisker plots of both the HI (Fig. 5-4A) and R2 (Fig. 5-4B) results for Florida

and Georgia Cypresshead samples illustrate several notable trends. First, the Florida sample sites

exhibit consistency in the median values of both HI and R2, except for SSD-1 which possesses a lower median value than the others. Additionally, the HI and R2 results for the EPK, FRG, and

SSD-1 localities exhibit a greater spread in data than what is observed for FRL-1 and SSJ-1

samples. This greater variation in both HI and R2 correlates to a consistent vertical trend noted

for these sites, where kaolinite order increases with depth from the surface (Fig. 5-1D). In most

140

Table 5-2. Statistical summary of kaolinite order and crystallite size calculations.

Locality ID/ Hinkley Liètard Area-wt Mean Volume-wt Mean Statistic Index (HI) Index (R2) Thickness (nm) Thickness (nm) EPK Max 1.36 0.94 7.2 11.0 Min 0.31 0.65 2.9 4.7 Mean 0.94 0.83 5.5 8.9 FRG Max 1.56 1.10 8.2 11.4 Min 0.23 0.73 2.7 4.2 Mean 1.13 0.95 5.7 8.8 FRL-1 Max 1.23 0.96 10.0 13.6 Min 1.00 0.82 6.6 10.1 Mean 1.10 0.86 7.8 11.1 SSJ-1 Max 1.28 0.99 8.1 11.0 Min 0.82 0.67 6.3 9.4 Mean 1.10 0.86 7.4 10.5 SSD-1 Max 1.13 0.91 7.2 9.6 Min 0.46 0.50 2.8 4.7 Mean 0.78 0.74 5.0 7.5 J-1 Max 0.69 0.53 6.6 8.9 Min 0.24 0.46 5.4 7.5 Mean 0.43 0.50 6.2 8.4 L-1 Max 0.43 0.58 6.9 9.3 Min 0.15 0.45 4.5 7.1 Mean 0.30 0.49 6.2 8.5 B-1 Max ------4.4 6.4 Min ------4.0 5.8 Mean ------4.2 6.0 instances, however, kaolinites appear to reverse this trend near the base of each section studied, becoming more disordered prior to reaching the basal surface of the formation (Table 5-1). Even for the Florida sample sites which do not exhibit the near-surface occurrence of disordered kaolinite (FRL-1 and SSJ-1), the latter observation holds true, as in both cases HI and R2 values exhibit a notable decrease in value near the base of the section. Excepting the EPK data, the

Florida sample sites with the greatest variability in HI and R2 results correlate to a lower average

141

Figure 5-3. Histograms illustrating the distribution of results for both kaolinite order (HI and R2) and crystallite size calculations. A) Hinkley Index (HI), B) Liètard Index (R2), C) area-weighted thickness, D) volume-weighted thickness.

clay content (FRG = 7.7% and SSD-1 = 4.9%) than those with little variability in results (FRL-1

= 10.4% and SSJ-1 = 10.8%) (Appendix D). Whereas clay content at the EPK site is more

concentrated toward the base of the formation, clay content noted for both FRL-1 and SSJ-1 is

highest near the top of the formation, likely impacting vertical infiltration rates and consequently

leaching/weathering. Box-and-whisker plots for the Georgia sample sites, for which HI and R2

results were measurable, exhibit consistently low HI (Fig. 5-4A) and R2 (Fig. 5-4B) median

values, with little variation in results.

A review of the results obtained via the expert system of Plançon and Zacharie (1990)

indicate that Florida Cypresshead kaolinite, including SSD-1, is primarily a two-phase (at least)

mixture, achieving a high relative percentage (~38-69%) of well-ordered (low defect) kaolinite

142

Figure 5-4. Box-and-whisker diagrams for kaolinite disorder (HI and R2) and crystallite size calculations.

(%ldp) in the middle of the sections evaluated by this study. However, as with the HI and R2 results, the percentage of well-ordered kaolinite for two-phase mixtures decreases at both the top

(11-32%) and bottom (~11-42%) of most sections. For the Florida sections exhibiting the greatest variability in HI and R2 values (EPK and FRG), single-phase kaolinite is common at the top of the sections consistent with a disordered phase exhibiting a moderate concentration of translation defects of probability p (Table 5-1). Octahedral C-site vacancies with an abundance of Wc are absent. Additionally, δ (in fraction of unit-cell), which corresponds to stacking disorders, also appears to be a minimal component of the disorder in these kaolinites. One basal

143

sample (SSD-1-10) was also determined to be a single-phase kaolinite similar in character to

those described from the near surface of EPK and FRG sections. In general, results from the

expert system correlated well with HI results, indicating a general vertical trend of increased

kaolinite order with depth from the surface followed by degradation in order near the base of

most sections (Table 5-1). As for Cypresshead samples from southeastern Georgia, only three

collected from the Birds reference locality could be evaluated by the expert system due to the

highly disordered (high defect) character of the kaolinite. In each case, the results indicate a

single-phase kaolinite similar in defect characteristics to those noted from Florida sections where translation defects (p = 0.32-0.40) appear to be responsible for the high degree of disorder noted for these samples (Table 5-1).

Crystallite size, or rather the coherent scattering domain (CSD) size, determined for

Cypresshead Formation kaolinite yields geological information related to weathering processes which have impacted the unit (Table 5-3). For kaolinite, CSD values measured for the 001 peak identifies the crystallite dimension perpendicular to the ab plane of the unit cell such that crystallite size is equal to (N-1) * d(hkl), where N is the number of hkl planes responsible for the

reflection. CSD values were collected for all Georgia samples, as analysis are based on the

diffraction characteristics of the basal d(001) reflection and does not suffer from the same mineral

phase interferences encountered with disorder calculations. Best mean thickness values

determined graphically via the Warren-Averbach method (1953) are included for reference,

while the area-weighted mean thickness and volume-weighted mean thickness values calculated

using the modified Bertaut-Warren-Averbach method employed by the MudMaster computer

program of Eberl et al. (1996) were the focus of evaluation (Table 5-3). The values of α and β2 define the theoretical lognormal distribution for crystallite size fitted to the measured data

144

Table 5-3. Results of crystallite size calculations for north-central Florida and Georgia Cypresshead and reworked Cypresshead kaolinite.

Mineral Best Mean Area-wt Mean Volume-wt Mean Distribution Sample ID Interval α β2 Interferences Thickness (nm)a Thickness (nm)b Thickness (nm)c Limit Cypresshead Formation - FL EPK36-J-12 25-27 H 2.9 2.9 0.72 0.60 6.5 20 EPK36-J-12 27-30 H 4.5 4.5 1.08 0.82 9.2 23 EPK36-J-12 35-40 H 4.3 4.3 1.07 0.76 8.7 22 EPK36-J-12 40-44 5.6 5.7 1.34 0.82 10.3 23 EPK36-J-12 44-46 6.1 6.1 1.49 0.68 9.9 21 EPK36-J-12 46-48 5.6 5.8 1.41 0.74 9.9 22 EPK36-J-12 48-50 6.3 6.3 1.51 0.73 10.2 22 EPK36-J-12 50-53 6.5 6.4 1.55 0.69 10.0 20 EPK36-J-12 53-56 6.6 6.6 1.59 0.66 10.2 21 EPK36-J-12 56-59 6.8 6.9 1.66 0.61 10.1 20 EPK36-J-12 59-62 6.7 6.9 1.69 0.55 10.0 20 EPK31-P-40 27-35 6.0 5.9 1.49 0.62 9.3 20 EPK31-P-40 35-45 7.2 7.2 1.66 0.71 11.0 22 EPK31-P-40 45-50 6.4 6.6 1.61 0.62 9.8 20 EPK31-P-40 50-62 6.8 6.9 1.73 0.43 9.4 20 EPK31-P-40 62-65 5.6 6.3 1.64 0.43 9.0 22 EPK30-V-6 16-22 H 2.8 3.0 0.86 0.45 4.7 15 EPK30-V-6 22-24 H 2.9 2.9 0.80 0.52 4.9 15 EPK30-V-6 24-27 ------EPK30-V-6 30-35 H 4.3 4.6 1.24 0.58 7.2 17 EPK30-V-6 35-39 3.1 3.3 0.89 0.54 6.0 20 EPK30-V-6 39-43 3.6 3.8 1.03 0.59 7.1 20 EPK30-V-6 43-48 4.2 4.3 1.13 0.63 7.7 20 EPK30-V-6 48-53 4.2 4.3 1.15 0.60 7.3 19 EPK30-V-6 53-58 6.0 6.2 1.45 0.79 10.7 25 EPK30-V-6 58-63 6.6 6.7 1.59 0.72 10.5 22 EPK30-V-6 63-68 6.3 6.3 1.53 0.71 10.0 21 EPK30-V-6 68-73 6.9 6.9 1.67 0.59 9.9 20 EPK30-V-6 73-78 5.8 6.2 1.58 0.50 9.2 21 FRG-1 1 3.5 3.4 1.02 0.43 5.1 12 FRG-1 2 HIV 3.6 3.8 1.11 0.42 5.5 13 FRG-1 3 HIV 4.3 4.5 1.26 0.49 6.7 15 FRG-1 4 HIV 3.8 4.1 1.17 0.49 6.4 16 FRG-1 5 H 2.5 2.7 0.78 0.40 4.2 14 FRG-1 6 4.5 4.3 1.13 0.63 7.5 17 FRG-1 7 4.8 4.3 1.15 0.61 7.3 16 FRG-1 8 3.7 3.9 1.09 0.51 6.9 20 FRG-1 9 5.2 5.1 1.32 0.62 8.4 20 FRG-1 10 6.3 6.3 1.60 0.54 9.1 18 FRG-1 11 7.1 6.7 1.56 0.78 10.7 21 FRG-1 12 7.6 6.9 1.59 0.80 10.7 20 FRG-1 13 6.2 6.0 1.39 0.83 10.6 23 FRG-1 14 6.5 5.8 1.38 0.80 9.9 20 FRG-1 15 6.3 5.8 1.39 0.77 9.7 20 FRG-2 1 HIV 3.8 3.8 1.13 0.44 5.8 13 FRG-2 2 HIV 4.2 4.1 1.13 0.56 7.0 17 FRG-2 3 6.8 6.6 1.55 0.76 10.4 20 FRG-2 4 5.8 5.6 1.37 0.74 9.5 20 FRG-2 5 7.7 7.0 1.62 0.75 10.5 19 FRG-2 6 8.7 8.2 1.82 0.68 11.4 20 FRG-2 7 7.9 7.2 1.63 0.79 10.9 20 FRG-2 8 7.5 6.9 1.61 0.75 10.6 20 FRG-2 9 7.6 7.5 1.71 0.72 11.2 22 FRG-2 10 8.4 8.1 1.87 0.50 10.8 20 FRG-2 11 7.8 7.8 1.81 0.57 11.1 22 FRG-2 12 7.5 7.5 1.81 0.45 10.1 20 FRG-2 13 5.1 5.6 1.48 0.51 8.5 20 Note: Crystallite size analyses based on the Bertaut-Warren-Averbach method. The parameters α and β2 define the theoretical lognormal distribution for crystallite size fitted to the measured data. Mineral interferences include: halloysite (H), hydroxy-interlayed vermiculite (HIV), and illite (I). a Determined by extrapolation of a line along the steepest slope of a plot of the Fourier coefficient of the Kα interference function H(n) verses n, where n = the number of data points divided by 2. b Calculated as the area-weighted mean of a plot of crystallite size normalized to the frequency of occurrence. c Calculated as the volume-weighted mean of a plot of crystallite size normalized to the frequency of occurrence. 145

Table 5-3. – (continued).

Mineral Best Mean Area-wt Mean Volume-wt Mean Distribution Sample ID Interval α β2 Interference Thickness (nm)a Thickness (nm)b Thickness (nm)c Limit Cypresshead Formation - FL (cont.) FRL-1 1 10.1 10.0 2.09 0.48 13.6 30 FRL-1 2 7.1 6.8 1.63 0.64 10.1 20 FRL-1 3 7.5 7.3 1.73 0.61 10.6 20 FRL-1 4 9.0 9.0 1.97 0.53 12.4 24 FRL-1 5 6.7 6.6 1.54 0.76 10.7 22 FRL-1 6 7.2 7.3 1.75 0.53 10.3 20 FRL-1 7 7.4 7.5 1.79 0.51 10.3 20 FRL-1 8 6.8 7.4 1.73 0.61 11.0 23 FRL-1 9 7.8 8.0 1.87 0.50 10.9 21 SSJ-1 1 8.0 8.0 1.88 0.44 10.7 22 SSJ-1 2 7.5 7.6 1.82 0.44 10.3 21 SSJ-1 3 6.5 6.6 1.66 0.49 9.4 20 SSJ-1 4 7.0 7.0 1.68 0.61 10.4 22 SSJ-1 5 6.4 6.3 1.52 0.69 9.8 20 SSJ-1 6 7.1 7.3 1.71 0.64 10.9 22 SSJ-1 7 7.4 7.5 1.75 0.63 10.9 21 SSJ-1 8 7.2 7.4 1.73 0.61 10.8 22 SSJ-1 9 7.2 7.3 1.73 0.58 10.6 21 SSJ-1 10 7.6 7.9 1.84 0.50 11.0 22 SSJ-1 11 8.0 8.1 1.89 0.45 10.9 22 Reworked Cypresshead Formation - FL SSD-1 1 2.9 2.8 0.80 0.47 4.7 15 SSD-1 2 3.1 3.1 0.85 0.53 5.6 17 SSD-1 3 3.4 3.4 0.93 0.58 6.2 16 SSD-1 4 4.0 3.9 1.05 0.64 7.0 17 SSD-1 5 4.3 4.3 1.14 0.65 7.4 17 SSD-1 6 6.4 6.9 1.75 0.42 9.2 18 SSD-1 7 7.0 7.2 1.78 0.43 9.6 19 SSD-1 8 6.6 7.0 1.78 0.40 9.6 23 SSD-1 9 5.5 6.0 1.61 0.37 8.1 19 SSD-1 10 5.3 5.8 1.59 0.33 7.6 18 Cypresshead Formation - GA J-1 1 4.9 5.4 1.47 0.44 7.5 16 J-1 2 6.1 6.3 1.68 0.37 8.3 17 J-1 3 5.9 6.4 1.67 0.39 8.5 18 J-1 4 I 5.7 6.3 1.63 0.45 8.9 20 J-1 5 I 5.6 6.1 1.63 0.39 8.4 19 J-1 6 I 6.4 6.6 1.71 0.38 8.9 20 LB-1 1 6.1 6.4 1.69 0.36 8.5 18 LB-1 2 6.6 6.8 1.76 0.34 8.9 20 LB-1 3 6.5 6.7 1.73 0.37 8.9 20 LB-1 4 6.7 6.9 1.77 0.36 9.3 22 LB-1 5 4.8 5.3 1.42 0.51 8.0 18 LB-1 6 6.3 6.5 1.69 0.38 8.8 20 LB-1 7 4.1 4.5 1.26 0.50 7.1 18 B-1 1 I 3.5 4.1 1.24 0.37 5.8 17 B-1 2 I 3.4 4.0 1.20 0.40 5.8 16 B-1 3 I 3.4 4.0 1.15 0.46 5.9 15 B-1 4 I 3.9 4.4 1.28 0.43 6.4 16 B-1 5 I 3.7 4.3 1.26 0.41 6.3 17 Note: Crystallite size analyses based on the Bertaut-Warren-Averbach method. The parameters α and β2 define the theoretical lognormal distribution for crystallite size fitted to the measured data. Mineral interferences include: halloysite (H), hydroxy-interlayed vermiculite (HIV), and illite (I). a Determined by extrapolation of a line along the steepest slope of a plot of the Fourier coefficient of the Kα interference function H(n) verses n, where n = the number of data points divided by 2. b Calculated as the area-weighted mean of a plot of crystallite size normalized to the frequency of occurrence. c Calculated as the volume-weighted mean of a plot of crystallite size normalized to the frequency of occurrence.

146

corresponding to the area-weighted mean thickness curve. Values related to instrumental broadening and the LpG2 factor used in these calculations were derived from standard kaolinite values included with the MudMaster program.

For CSD calculations of Florida samples, including SSD-1 reworked Cypresshead sediments, the area-weighted mean and volume-weighted mean values range from 2.7 nm to 10.0 nm and 4.2 nm to 13.6 nm, respectively, with corresponding means of 6.0 nm and 9.1 nm. For

Georgia samples, the area-weighted mean and volume-weighted mean values range from 4.0 nm to 6.9 nm and 5.8 nm to 9.3 nm, respectively, with means of 5.6 nm and 7.8 nm. Unlike what was noted with disorder indices (HI and R2), CSD distributions for Florida and Georgia samples exhibit significant overlap for both area-weighted (Fig. 5-3C) and volume-weighted (Fig. 5-3D) thickness calculations, resulting in no differentiation between the two sample populations based on this parameter. This is further reflected in the lack of differentiation between individual sample localities, although the spread within the data for individual sites is similar that observed for the disorder indices (Table 5-2). This latter observation is further supported by box-and- whisker plots of both the area-weighted (Fig. 5-4C) and volume-weighted (Fig. 5-4D) thickness calculations for Florida and Georgia Cypresshead samples. Additionally, several other trends are notable, including the relative coarseness of the FRL-1 and SSJ-1 samples and the fine crystallite thicknesses measured for the Birds locality (B-1).

In general, site specific trends related to the CSD results are consistent with those noted for disorder indices (HI and R2) and the expert system of Plançon and Zacharie (1990). For localities where significant trends in disorder were noted (EPK, FRG, and SSD-1), CSD values increase with depth from the surface. Additionally, where disorder indices indicate a reversal near the base of select localities, CSD results exhibit a similar trend, recording a decrease in

147

value prior to reaching the basal contact of the Cypresshead Formation. Contrasting with what is observed for the Florida localities, none of the Georgia localities exhibit significant vertical trends in CSD values. Correlation of the area-weighted mean thickness and volume-weighted mean thickness calculations to HI values for Florida Cypresshead samples is illustrated in Figure

5-5A and B.

With an increase in depth from the surface in north-central Florida, CSD distributions diverge significantly from the theoretical lognormal distribution curves defined by α and β2 (Fig.

5-6). In near surface samples, curves are most similar to a lognormal distribution, but tend to be narrower than predicted, often with a circumflex (or hat) that extends above the mode of the theoretical lognormal curve. This feature indicates that the curves may have started out with an asymptotic shape, a feature which is commonly attributed to the early stages of crystallization

(Eberl et al., 1998; Simić and Uhlík, 2006), with loss of the low end of the asymptotic curve potentially related to dissolution of a fine crystallite fraction. With depth, the curves take on a broader shape toward greater crystallite thicknesses, sometimes toward distribution curves possessing a bimodal or polymodal shape (Figs. 5-6 and 5-7). For north-central Florida

Cypresshead (and reworked Cypresshead) sediments, this shift in CSD thickness distributions is associated with β2 values remaining constant with increasing α, which correlates to an open system supply-controlled growth method according to the Law of Proportionate Effect (Eberl et al., 1998). Georgia samples exhibit no significant change in α values and more closely correlate to a lognormal distribution shape than north-central Florida kaolinites, suggesting minimal evidence for recrystallization (or neoformation). Examples of vertical trends in CSD distribution curves for select north-central Florida localities characterized by significant spreads in CSD and disorder values (EPK, FRG, and SSD) are included in Figure 5-7 for reference.

148

Figure 5-5. Scatterplots illustrating the positive correlation between the Hinkley Index (HI) and measured CSD values for north-central Florida Cypresshead (and reworked Cypresshead) samples. A) HI plotted against the area-weighted mean thickness, B) HI plotted against the volume-weighted mean thickness.

Kaolinite Microtexture

Microtextures within a kaolin record evidence of both deposit origin and post-depositional alteration processes. Furthermore, microtextures allow in situ crystallization to be differentiated from sedimentary fabrics (Hurst and Rigsby, 1984). Kaolins with a high concentration of parallel intergrowths, random intergrowths, coarse booklets, stacks, and vermiforms are indicative of in situ crystallization (Keller, 1977; Hurst and Pickering, 1989), while those possessing a

149

Figure 5-6. Measured CSD distribution curves and fitted theoretical lognormal curves (red) for select EPK36-J-12 samples. A) narrow curve possessing a circumflex characteristic of an initial asymptotic shape, B) curve exhibits initial broadening and development of a secondary mode at ~8 nm, C) continued development of a polymodal shape, D) same as (C).

sedimentary fabric of crystallites in face-to-face association are indicative of little post- depositional recrystallization (Hurst and Pickering, 1989). In Georgia, Cretaceous kaolins possess the textures consistent with in situ recrystallization, with Tertiary kaolins possessing a notably fine-grained and face-to-face oriented fabric (Keller, 1977; Hurst and Pickering, 1989), thus characteristic of little recrystallization.

SEM data from Cypresshead Formation samples indicate variations in kaolinite morphology for sediments both vertically and between sampling areas, particularly for north- central Florida sites (Fig. 5-8). Near surface north-central Florida kaolinite consists of small (≤1

150

Figure 5-7. Example CSD distributions illustrating changing curve shape with increased depth from the surface. A) EPK36-J-12, B) EPK30-V-6, C) FRG-1, D) SSD-1.

µm) rounded aggregates and anhedral to subhedral crystallites in an open microtexture associated with combined face-to-face and face-to-edge floccules (Fig. 5-8A). Where present, vermicular (or vermiform) kaolinite is highly degraded, or potentially recrystallized, and commonly appears coated by either secondary overgrowths of fine (<0.5 µm) anhedral to subhedral kaolinite or translocated (illuviated) kaolinite. For both FRL-1 and SSJ-1 sample localities, this microtexture is less developed, and correlates with a reduction in the apparent porosity of near surface samples. Elsewhere, this kaolinite morphology tends to transition down- section into significant accumulations of vermicular kaolinite composed of slightly coarser (1-2

µm) subhedral crystallites retaining a similar open microtexture (Fig. 5-8B) and samples indicating the development of overgrowth crystallization textures consisting of subhedral to

151

Figure 5-8. SEM photomicrographs illustrating the microtextural variation noted in Cypresshead and reworked Cypresshead Formation sediments. A) Near surface rounded aggregates of kaolinite (FRG-1-6), B) Degraded vermicular kaolinite (FRG-1-13), C) Fine secondary overgrowths of kaolinite (SSD-1-10), D) Well formed kaolinite vermiform (FRG-1-11), E) Matrix image of sample EPK36-J-12 (59-62), illustrating the dominance of vermicular kaolinite, F) Dense microtextue inconsistent with recrystallization fabric (L-1-6). 152

euhedral crystallites (Fig. 5-8C). Where overgrowths are developed on preexisting vermicular or discrete kaolinite, crystallization appears to be favored along grain edges, oriented face-to-edge or edge-to-edge at an oblique angle. At depth, north-central Florida kaolinite is dominated by the accumulation of vermicular kaolinite (Fig. 5-8D), which can compose the bulk of the clay matrix, particularly at the EPK Mine site (Fig. 5-8E). Evidence of this microtexture trend is well documented in the EPK36-J-12 core, where vermicular kaolinite appears degraded

(recrystallized) or coated with overgrowths or translocated clay down to 48 ft, while matrix clay retains an aggregate morphology observed at the top of the core. Below 50 ft, a more characteristic open microtexture with large vermiforms is dominant, but with individual crystallites still fine in particle-size (<2 µm). Cypresshead samples from southeastern Georgia differ significantly, consisting of small (≤1 µm) subhedral kaolinite crystallites in a closed, face- to-face flocculated microtexture inconsistent with significant recrystallizaton (Fig. 5-8F). Where a more open face-to-edge microtexture is developed in Georgia sediments (L-1), it tends to be restricted to sandy lithologies possessing minimal clay content. The general difference in sediment microtexture (microfabric) between north-central Florida and southern Georgia

Cypresshead Formation sediments indicates significant differences in weathering susceptibility as a function of porosity (and likely permeability), which in turn, appears to be related to spatial variations in the initial clay content of the unit at the time of deposition.

Particle-Size Analysis

SediGraph analyses were performed in order to characterize silt and clay particle-size distributions as a means of confirming XRD and SEM based observations related to general trends of increased structural order, crystallite (CSD) size, and clay particle-size with increased depth of sampling. Figure 5-9 is included for reference, and indicates a general trend of

153

Figure 5-9. Example SediGraph particle-size distributions for select Cypresshead Formation samples. A) EPK36-J-12, B) FRG-1, C) SSJ-1, D) J-1.

increasing coarsening of particle-size with depth. This is agrees with the previously noted observation related to both disorder and crystallite size. Additional data is included in Appendix

G for reference.

Geochemistry

Major element and rare earth element (REE) data were collected using a combination of

ICP-AES and ICP-MS methodologies to assist in defining weathering reactions impacting the

Cypresshead Formation. Of particular interest to this study is the source of Al and Si necessary for the neoformation of kaolinite, and coeval weathering reactions involving secondary accessory phases (e.g., crandallite-florencite, gibbsite, halloysite). The latter of these has the potential of

154

impacting the dissolution and/or crystallization kinetics of kaolinite and the chemical signature

(i.e., REE distributions and Nd isotopes) of the clay fraction.

Major element data

Major element data for the clay (< 2 µm) size-fraction of both north-central Florida and southeastern Georgia Cypresshead Formation and reworked Cypresshead sediments converted to oxides (%) is reported in Table 5-4. Original major element data is included in Appendix H.

Additionally, a correlation matrix indicating the correlation coefficients (r) between major

elements and ∑REE is shown in Table 5-5. Due to the complexity of weathering reactions

occurring within the unit, correlation coefficient values (r > 0.6) were considered significant.

However, due to the low concentrations encountered with Na2O for the total data set and CaO for

the Georgia samples, no correlations for these elements were considered significant.

Correlations among major elements for Cypresshead samples are related to variations in

sample mineralogy and weathering reactions impacting Cypresshead sediments. The most

significant of these as it relates to the overall geochemical signature of Cypresshead clays is the

occurrence of a crandallite-florencite phase in some samples. P2O5 concentrations approach

almost 2% in several samples (SSD-1), with significant enrichment evident in others (EPK and

FRG). P2O5 exhibits a range of 0.046% to 1.978% and 0.034% to 0.213% in north-central

Florida and southeastern Georgia samples, respectively, with means of 0.228% and 0.092%. It is

significant to note that P2O5 and CaO show no correlation for the Florida samples (Table 5-5), as

both crandallite and florencite appear to be present in these sediments (Fig. 5-10A).

The only major element correlations of significance are noted for the Georgia samples,

with MgO, Fe2O3, and K2O exhibiting positive correlations in response to the concentration of

both illite and hydroxyl-interlayered vermiculite (HIV) in these sediments. Negative correlations

155

Table 5-4. Major element concentrations as oxides for Cypresshead Formation clay (< 2 µm) samples (concentrations in %).

Sample ID Interval Na2O MgO Al2O3 P2O5 K2O CaO TiO2 Fe2O3 Cypresshead Formation - FL* EPK36-J-12 25-27 0.011 0.076 36.697 0.137 0.04 0.048 0.250 0.27 EPK36-J-12 35-40 0.013 0.050 33.806 0.149 0.05 0.049 0.234 0.36 EPK36-J-12 46-48 0.015 0.073 37.604 0.071 0.05 0.035 0.282 0.49 EPK36-J-12 50-53 0.020 0.050 32.200 0.046 0.08 0.028 0.317 0.51 EPK36-J-12 59-62 0.020 0.099 34.184 0.069 0.10 0.077 0.550 0.77 EPK31-P-40 35-45 0.018 0.083 37.623 0.062 0.06 0.045 0.347 0.66 EPK31-P-40 50-62 0.020 0.158 32.445 0.069 0.08 0.147 0.309 0.84 EPK31-P-40 62-65 0.018 0.393 34.543 0.172 0.23 0.288 0.285 1.24 EPK30-V-6 22-24 0.020 2.429 9.511 0.222 0.28 0.641 0.592 2.55 EPK30-V-6 30-35 0.034 1.277 25.113 1.180 0.29 0.840 0.300 1.53 EPK30-V-6 48-53 0.034 0.224 32.672 0.630 0.16 0.497 0.417 0.89 EPK30-V-6 58-63 0.020 0.060 37.245 0.050 0.08 0.022 0.252 0.41 EPK30-V-6 68-73 0.020 0.116 32.464 0.057 0.14 0.070 0.434 0.67 FRG-1 3 0.012 0.083 35.639 0.280 0.05 0.042 0.312 2.99 FRG-1 6 0.007 0.025 33.730 0.160 0.02 0.021 0.067 0.99 FRG-1 7 n.a. 0.025 35.072 0.092 0.02 n.a. 0.075 0.64 FRG-1 10 0.013 0.091 34.486 0.241 0.10 0.042 0.217 0.61 FRG-1 11 0.007 0.033 35.166 0.069 0.04 0.014 0.108 0.23 FRG-1 13 0.007 0.025 35.771 0.046 0.04 0.007 0.117 0.21 FRG-1 15 0.007 0.025 33.352 0.080 0.05 0.007 0.225 0.23 FRG-2 3 0.067 0.041 34.864 0.057 0.04 0.014 0.108 0.27 FRG-2 5 n.a. 0.025 34.108 0.055 0.04 n.a. 0.107 0.36 FRG-2 7 0.061 0.033 35.488 0.046 0.05 0.014 0.150 0.27 FRG-2 10 0.013 0.058 34.826 0.126 0.08 0.021 0.267 0.63 FRG-2 12 0.011 0.138 33.541 0.144 0.11 0.014 0.360 1.40 FRL-1 2 0.007 0.055 33.201 0.057 0.06 n.a. 0.309 0.83 FRL-1 4 0.007 0.056 33.296 0.064 0.07 n.a. 0.315 0.94 FRL-1 6 0.007 0.075 33.881 0.069 0.05 0.014 0.250 0.61 FRL-1 7 0.007 0.083 32.237 0.046 0.05 0.014 0.217 0.59 SSJ-1 2 0.019 0.151 34.278 0.124 0.10 0.038 1.154 3.22 SSJ-1 4 0.016 0.162 33.560 0.140 0.08 0.024 0.582 2.14 SSJ-1 6 0.020 0.091 35.053 0.057 0.08 0.028 0.158 0.47 SSJ-1 8 0.022 0.109 34.788 0.053 0.20 0.041 0.254 0.40 SSJ-1 10 0.027 0.158 34.656 0.057 0.16 0.070 0.409 0.51 SSD-1 1 0.081 0.240 28.458 1.978 0.41 0.101 7.970 2.12 SSD-1 3 0.013 0.070 37.831 1.668 0.08 0.070 0.724 0.56 SSD-1 6 0.011 0.101 33.050 0.108 0.05 0.008 0.229 0.66 SSD-1 7 0.020 0.108 33.768 0.080 0.07 0.014 0.259 0.73 SSD-1 10 0.020 0.290 33.636 0.080 0.11 0.098 0.225 1.04 Max 0.081 2.429 37.831 1.978 0.410 0.840 7.970 3.217 Min 0.007 0.025 9.511 0.046 0.024 0.007 0.067 0.214 Mean 0.020 0.191 33.432 0.228 0.099 0.100 0.506 0.894 Cypresshead Formation - GA J-1 2 0.023 0.247 15.246 0.105 0.19 0.013 1.401 7.65 J-1 4 0.034 0.207 34.240 0.080 0.52 0.014 1.026 4.48 J-1 6 0.027 0.182 33.617 0.034 0.37 0.007 0.767 1.90 L-1 3 0.019 0.131 36.357 0.112 0.18 n.a. 1.028 1.72 L-1 5 n.a. 0.050 35.355 0.044 0.05 n.a. 0.374 0.84 L-1 6 0.020 0.124 33.806 0.057 0.27 0.007 0.851 1.62 B-1 2 0.051 0.630 28.269 0.087 1.20 0.014 0.666 5.62 B-1 3 0.063 0.703 29.327 0.213 1.54 0.013 0.717 10.91 B-1 5 0.054 0.705 33.296 0.092 1.07 0.091 0.742 3.63 Max 0.063 0.705 36.357 0.213 1.542 0.091 1.401 10.909 Min 0.019 0.050 15.246 0.034 0.048 0.007 0.374 0.844 Mean 0.036 0.331 31.057 0.092 0.600 0.023 0.841 4.262 Note: n.a., element concentration was below the detection limit. * Florida Cypresshead Formation samples include reworked Cypresshead sediments collected at the SSD-1 locality.

156

Table 5-5. Correlation matrix of major elements and ∑REE for both Florida and Georgia Cypresshead samples.

Na2O MgO Al2O3 P2O5 K2O CaO TiO2 Fe2O3 ΣREE Cypresshead - FL* Na2O 1.00 MgO 0.10 1.00 Al2O3 -0.16 -0.92 1.00 P2O5 0.46 0.21 -0.20 1.00 K2O 0.51 0.59 -0.60 0.62 1.00 CaO 0.18 0.82 -0.71 0.37 0.63 1.00 TiO2 0.60 0.06 -0.21 0.70 0.65 0.03 1.00 Fe2O3 0.110.46-0.460.280.440.340.361.00 ΣREE 0.50 0.29 -0.35 0.86 0.74 0.46 0.76 0.34 1.00 Cypresshead - GA Na2O 1.00 MgO 0.96 1.00 Al2O3 -0.01 -0.21 1.00 P2O5 0.59 0.61 -0.29 1.00 K2O 0.99 0.95 -0.08 0.66 1.00 CaO 0.44 0.55 0.20 0.02 0.32 1.00 TiO2 -0.62 -0.17 -0.60 0.18 -0.26 -0.22 1.00 Fe2O3 0.63 0.66 -0.64 0.84 0.68 -0.13 0.29 1.00 ΣREE -0.20 0.02 0.46 0.17 -0.03 0.76 0.15 -0.30 1.00 * Florida Cypresshead Formation samples include reworked Cypresshead sediments collected at the SSD-1 locality.

noted between Al2O3 and MgO, K2O and CaO for Florida samples are insignificant because they are related to the high clay fraction quartz content at the top of EPK30-V-6 diluting the Al2O3 content. Consequently, the correlations among CaO, MgO and K2O are similarly related.

Rare earth element (REE) data

Although rare earth elements (REEs) can be reliable geochemical tracers of sedimentary

provenance because of their characteristic immobility and resistance to elemental fractionation in

the supracrustal environment (Wildeman and Condie, 1973; Piper, 1974; Nesbitt, 1979;

Chaudhrui and Cullers, 1979), studies suggest that mobilization and local fractionation of the

REEs can occur under extreme weathering and diagenesis (McLennan, 1989). In fact, relative

concentrations of REEs in neoformed and recrystallized phases will reflect, to varying degrees,

the geochemical conditions and redox potential within local pore waters at the time of formation

(Henderson et al., 1983; McLennan, 1989).

157

Figure 5-10. Scatterplots illustrating mixing trends related to the presence of crandallite- florencite series minerals in Cypresshead Formation clays. A) relationship between P2O5 and CaO for crandallite and florencite crystallization trends, B) correlation between P2O5 and ∑REE (r = 0.86).

REE concentrations for the clay (< 2 µm) size-fraction of both north-central Florida and southeastern Georgia Cypresshead Formation and reworked Cypresshead sediments along with relevant REE ratios are reported in Table 5-6. Additionally, REE data for three samples representative of updip, less weathered Cypresshead clays (TRF2214, WEX164, and WEX366), two samples of underlying Hawthorn Group clays (MCB109 and J-1-BC), and three EPK

158

Table 5-6. REE concentrations of Cypresshead Formation and related samples (concentrations in ppm). ΣLREE/ Sample ID Interval La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu ΣREE Eu/Eu* (La/Yb) (La/Sm) (Gd/Yb) n n n ΣHREE Cypresshead Formation - FL* EPK36-J-12 25-27 18.4 31.9 3.90 13.7 2.2 0.47 1.88 0.29 1.24 0.24 0.50 0.07 0.4 0.05 75.24 0.71 30.76 5.16 3.75 15.01 EPK36-J-12 35-40 29.6 53.4 6.66 24.2 4.3 0.97 4.34 0.59 2.90 0.50 1.26 0.15 0.8 0.09 129.76 0.69 24.74 4.25 4.32 11.12 EPK36-J-12 46-48 22.6 39.2 4.80 18.3 2.7 0.77 2.36 0.29 1.36 0.27 0.64 0.09 0.5 0.07 93.95 0.94 30.22 5.16 3.76 15.70 EPK36-J-12 50-53 20.0 32.5 3.82 13.3 2.2 0.58 1.63 0.19 0.94 0.16 0.42 0.06 0.4 0.06 76.26 0.94 33.43 5.61 3.25 18.61 EPK36-J-12 59-62 34.4 76.6 9.32 36.0 7.3 2.95 5.50 0.65 2.76 0.38 0.96 0.13 0.8 0.12 177.87 1.43 28.75 2.91 5.48 14.48 EPK31-P-40 35-45 20.2 38.0 4.11 15.3 2.5 1.04 2.03 0.25 1.12 0.19 0.47 0.07 0.5 0.08 85.86 1.42 27.02 4.99 3.24 17.01 EPK31-P-40 50-62 28.3 67.4 7.62 28.7 5.6 2.95 5.05 0.71 3.30 0.60 1.49 0.19 1.3 0.20 153.41 1.71 14.56 3.12 3.10 10.72 EPK31-P-40 62-65 45.9 115.0 14.50 64.1 13.5 6.23 13.50 1.99 9.91 1.69 3.91 0.47 2.9 0.42 294.02 1.42 10.58 2.10 3.71 7.27 EPK30-V-6 22-24 54.4 112.0 14.00 55.5 10.3 2.00 10.10 1.43 7.31 1.35 3.35 0.44 2.6 0.37 275.15 0.60 13.99 3.26 3.10 9.14 EPK30-V-6 30-35 243.0 516.0 72.20 306.0 60.2 12.70 60.70 8.38 43.30 8.01 20.40 2.35 12.7 1.60 1367.54 0.65 12.79 2.49 3.81 7.61 EPK30-V-6 48-53 89.9 169.0 22.50 93.4 19.5 4.69 18.90 2.50 13.00 2.45 6.04 0.73 4.1 0.48 447.19 0.75 14.66 2.84 3.67 8.18 EPK30-V-6 58-63 30.4 51.4 5.36 19.0 3.4 1.41 2.98 0.39 1.74 0.29 0.70 0.09 0.5 0.09 117.75 1.36 40.66 5.52 4.75 16.16 EPK30-V-6 68-73 24.8 47.0 6.10 21.8 4.3 1.75 3.71 0.53 2.73 0.48 1.31 0.19 1.3 0.17 116.17 1.35 12.76 3.56 2.27 9.98 FRG-1 3 44.3 86.7 10.90 43.9 8.3 1.66 7.85 1.11 5.42 0.99 2.48 0.29 1.7 0.26 215.86 0.63 17.43 3.29 3.68 9.66 FRG-1 6 36.0 76.3 9.86 36.3 6.4 1.69 5.33 0.69 3.01 0.48 1.07 0.14 0.6 0.08 177.95 0.89 40.12 3.47 7.08 14.46 FRG-1 7 31.2 60.9 7.23 26.6 4.4 1.20 3.68 0.48 1.94 0.28 0.55 0.07 0.4 0.06 138.99 0.92 52.16 4.38 7.33 17.47 FRG-1 10 139.0 281.0 35.60 139.0 23.1 6.85 16.80 2.02 7.22 0.97 1.84 0.19 1.0 0.12 654.71 1.07 92.95 3.71 13.39 20.48 FRG-1 11 40.6 71.4 8.01 30.3 4.5 1.48 3.51 0.45 1.69 0.23 0.45 n.a. 0.3 n.a. 162.92 1.14 90.50 5.57 9.33 23.35 FRG-1 13 20.3 33.8 4.00 14.2 1.9 0.78 1.49 0.20 0.88 0.12 0.24 n.a. 0.2 n.a. 78.11 1.43 67.87 6.59 5.94 23.71 FRG-1 15 55.1 122.0 14.30 47.3 7.7 2.76 5.24 0.71 2.58 0.33 0.64 0.07 0.4 n.a. 259.13 1.34 92.11 4.42 10.44 24.71 FRG-2 3 14.3 25.1 2.94 11.7 2.9 0.86 2.01 0.25 0.97 0.15 0.34 n.a. 0.2 n.a. 61.72 1.09 47.81 3.04 8.01 14.53 FRG-2 5 21.6 43.3 4.72 17.2 2.7 1.04 2.02 0.26 0.93 0.15 0.31 n.a. 0.2 n.a. 94.43 1.37 72.22 4.94 8.05 23.13 FRG-2 7 17.1 35.2 4.01 14.9 2.6 1.00 1.87 0.24 0.94 0.13 0.26 n.a. 0.2 n.a. 78.45 1.39 57.17 4.06 7.45 20.28 FRG-2 10 52.4 112.0 13.20 52.6 10.6 4.38 9.13 1.22 5.47 0.82 1.89 0.20 1.0 0.12 265.03 1.37 35.04 3.05 7.28 12.13 FRG-2 12 53.3 120.0 15.00 63.7 13.5 4.51 12.80 1.80 8.40 1.36 2.85 0.28 1.6 0.19 299.29 1.05 22.28 2.44 6.38 9.07

159 FRL-1 2 30.1 52.2 5.30 17.8 2.9 0.92 2.08 0.31 1.26 0.20 0.40 0.06 0.4 0.07 114.00 1.15 50.32 6.40 4.14 22.66 FRL-1 4 45.5 88.7 9.34 30.9 4.6 1.61 3.30 0.46 1.88 0.26 0.59 0.07 0.5 0.07 187.78 1.27 60.85 6.10 5.26 25.11 FRL-1 6 45.0 108.0 13.30 50.1 8.7 2.72 5.85 0.74 2.71 0.37 0.84 0.11 0.6 0.08 239.12 1.17 50.15 3.19 7.77 19.92 FRL-1 7 33.5 78.2 9.89 37.4 6.0 1.86 4.55 0.60 2.48 0.36 0.87 0.09 0.6 0.07 176.47 1.09 37.34 3.45 6.04 17.15 SSD-1 1 493.0 850.0 85.00 286.0 44.9 7.25 36.90 5.54 28.90 5.96 17.80 2.74 19.6 3.21 1886.80 0.55 16.82 6.77 1.50 14.58 SSD-1 3 123.0 229.0 23.50 80.5 13.3 1.98 9.61 1.30 4.96 0.76 1.75 0.24 1.6 0.24 491.74 0.54 51.41 5.71 4.79 22.94 SSD-1 6 43.0 93.5 11.40 48.6 9.2 2.70 7.06 0.88 3.29 0.45 0.88 0.10 0.6 0.09 221.75 1.03 47.92 2.88 9.38 15.41 SSD-1 7 50.1 123.0 14.50 58.5 11.1 3.38 8.40 1.04 4.24 0.53 1.09 0.11 0.6 0.07 276.66 1.08 55.84 2.78 11.16 16.00 SSD-1 10 22.8 57.7 5.79 22.4 4.6 1.69 3.70 0.49 2.09 0.34 0.76 0.09 0.5 0.05 123.00 1.26 30.49 3.06 5.90 14.13 SSJ-1 2 53.7 86.8 7.84 23.2 3.3 0.63 2.97 0.47 2.80 0.62 1.84 0.30 2.1 0.33 186.90 0.62 17.10 10.04 1.13 15.30 SSJ-1 4 32.4 56.1 5.45 18.0 2.9 0.54 2.46 0.36 1.97 0.39 1.19 0.17 1.2 0.19 123.32 0.62 18.05 6.89 1.63 14.48 SSJ-1 6 10.7 18.3 2.17 6.9 1.2 0.49 0.89 0.15 0.62 0.11 0.22 n.a. 0.2 n.a. 41.95 1.46 35.78 5.50 3.55 17.93 SSJ-1 8 9.6 16.3 1.75 5.9 0.9 0.43 0.78 0.13 0.55 0.11 0.26 n.a. 0.2 n.a. 36.91 1.58 32.10 6.58 3.11 16.97 SSJ-1 10 16.4 29.8 3.45 11.1 1.8 0.77 1.51 0.23 1.05 0.17 0.49 0.06 0.4 0.06 67.29 1.44 27.42 5.62 3.01 15.76 TRF2214 60.0-62.5 29.5 43.3 4.17 14.2 2.4 0.52 2.54 0.42 2.62 0.54 1.53 0.24 1.6 0.24 103.82 0.65 12.33 7.58 1.27 9.62 WEX164 18.0-26.0 76.1 131.0 15.00 58.9 10.9 2.55 11.80 1.80 10.10 2.20 6.05 0.85 5.1 0.72 333.07 0.69 9.98 4.31 1.84 7.56 WEX366 9.0-10.0 46.1 68.0 6.93 23.9 4.0 0.85 4.42 0.68 3.86 0.90 2.71 0.41 2.8 0.41 165.97 0.62 11.01 7.11 1.26 9.20 EPK Vermiforms --- 4.7 7.7 0.80 3.4 0.7 0.31 1.04 0.19 1.33 0.41 1.40 0.24 1.7 0.33 24.25 1.12 1.85 4.14 0.49 2.61 EPK Mica --- 7.2 14.7 1.64 6.8 1.2 0.48 1.38 0.21 1.38 0.29 0.92 0.14 1.0 0.17 37.51 1.15 4.81 3.70 1.10 5.74 EPK Feldspar --- 3.6 5.0 0.59 2.5 0.4 0.73 0.50 0.07 0.38 0.07 0.21 n.a. 0.2 n.a. 14.25 5.02 12.04 5.55 1.99 8.45 Cypresshead Formation - GA J-1 2 49.2 83.3 7.28 21.5 2.9 0.56 2.71 0.51 3.19 0.77 2.52 0.39 2.6 0.43 177.86 0.61 12.65 10.47 0.83 12.51 J-1 4 62.2 121.0 11.20 32.3 4.5 0.84 3.25 0.44 1.80 0.31 0.91 0.13 0.9 0.12 239.90 0.68 46.21 8.53 2.88 29.41 J-1 6 51.5 134.0 12.50 41.5 5.9 1.17 4.23 0.53 1.98 0.30 0.74 0.10 0.7 0.10 255.25 0.72 49.20 5.39 4.82 28.27 L-1 3 158.0 367.0 28.80 73.7 8.5 1.57 5.25 0.72 2.74 0.47 1.26 0.17 1.2 0.19 649.57 0.72 88.04 11.47 3.49 53.00 L-1 5 47.2 88.4 6.43 17.0 2.1 0.39 1.49 0.21 0.92 0.17 0.41 0.07 0.4 0.07 165.26 0.68 78.91 13.87 2.97 43.08 L-1 6 112.0 189.0 15.30 36.1 4.4 0.83 3.37 0.46 1.67 0.28 0.77 0.10 0.8 0.10 365.18 0.66 93.62 15.71 3.36 47.26 B-1 2 53.9 98.2 9.61 31.8 4.8 0.94 3.84 0.54 2.61 0.51 1.43 0.24 1.7 0.29 210.41 0.67 21.20 6.93 1.80 17.77 B-1 3 63.6 131.0 12.90 46.9 7.6 1.59 6.34 0.89 4.58 0.87 2.54 0.39 2.6 0.40 282.20 0.70 16.36 5.16 1.94 14.08 B-1 5 89.3 198.0 21.20 88.2 19.3 4.55 16.80 2.28 11.30 2.11 5.96 0.86 5.5 0.72 466.08 0.78 10.86 2.85 2.43 9.14 Hawthorn Group, Coosawhatchie Formation - FL/GA MCB109 15.0-20.0 56.3 94.5 9.85 36.0 6.3 1.41 6.51 0.95 5.47 1.23 3.42 0.52 3.6 0.55 226.61 0.68 10.46 5.51 1.44 9.12 J-1 BC 93.8 187.0 23.10 95.6 19.2 5.05 21.30 3.36 19.00 4.10 11.80 1.62 10.4 1.52 496.85 0.77 6.03 3.01 1.63 5.73 * Florida Cypresshead Formation samples include reworked Cypresshead sediments collected at the SSD-1 locality.

composite samples consisting of vermicular (vermiform) kaolinite, muscovite mica, and K-

feldspar concentrates are also included in Table 5-6. All sample localities are shown in Figure 3-

1. REE data for Cypresshead clays and accessory samples have been normalized to the chondrite

values of Nakamura (1974) for plotting purposes.

Data for ∑REE concentrations of Florida Cypresshead sediments (including SSD-1)

correlate strongly with P2O5 (r = 0.86) due to the occurrence of crandallite-florencite series

minerals ((Ca,REE)Al3(PO4)2(OH)5·H2O) in the clay fraction (Fig. 5-10B). ∑REE values for

Florida samples vary significantly, ranging between 36.91 ppm and 1886.80 ppm, whereas

Georgia samples are more constrained, ranging between 165.26 ppm and 649.57 ppm, with means of 207.96 ppm and 312.41 ppm, respectively. The distribution in ∑REE values correlated to secondary crandallite-florencite supports mobilization and reconcentration of REEs in north- central Florida sediments as a consequence of weathering. Furthermore, given the high concentration of REEs in the crandallite-florencite phase, it has the potential to impart a change in shape to chondrite-normalized REE distribution patterns should the phase preferentially enrich light (L) REEs as has been indicated in other studies (Dill et al., 1995; Rasmussen et al., 1998) or retain the REE signature of a precursor marine phosphate phase. As such, Florida clay fraction samples exhibiting evidence of crandallite-florencite via XRD or anomalous P2O5 and/or

anomalous ∑REE concentrations may be considered to possess a mixed REE signature if either

of these conditions are observed.

Chondrite-normalized REE distribution patterns for Cypresshead Formation and associated

samples are shown in Figure 5-11. Both Florida and Georgia Cypresshead samples (including

SSD-1) exhibit a wide range in the slope of REE distributions, with (La/Yb)n values varying

from 10.58 to 92.95 and 10.86 to 93.62, respectively. For Florida samples, LREE enrichment

160

Figure 5-11. Chondrite-normalized REE distribution patterns for select Cypresshead Formation clay (< 2 µm) fraction and related samples. A) EPK36-J-12, B) EPK30-V-6, C) FRG- 1, D) SSJ-1, E) SSD-1, F) J-1 and B-1, G) EPK vermiform, mica, and feldspar composite concentrates, H) updip Cypresshead and Hawthorn Group clays.

161

relative to chondrite (La/Sm)n is moderate (2.10-10.04) as is heavy (H) REE depletion ((Gd/Yb)n

= 1.13 to 13.39). No significant difference in LREE enrichment is apparent between phosphate

and non-phosphate bearing samples, thereby excluding any evidence of a fractionation or REE

mixing effect associated with the crandallite-florencite phase. For Florida samples, one of the

most notable REE trends corresponds to the variable HREE depletion values seen between some

near surface samples and samples collected at depth. For shallow samples associated with FRG-

1, SSJ-1 and SSD-1 sampling locations, HREE trends are only slightly depleted to flat, whereas

all other Florida Cypresshead samples exhibit moderate HREE depletion. The reason for this

observation is most likely related to the preferential adsorption of HREEs under reducing (anoxic

to dysoxic) conditions by pore water organics (humic acid) (Sonke and Salters, 2006; Wan and

Liu, 2006). Below the water table, in situ derived clays would exhibit HREE depletion in

response to the preferential uptake of the HREEs by the organics. With exposure to oxic

weathering conditions above the water table favoring decomposition of organics, clays would

uptake liberated HREEs, resulting in the flat HREE trends noted for some near surface samples.

For sampling locations where the water table is at or near the surface (EPK), this trend is not

observed.

For Georgia samples, LREE enrichment is slightly greater than that seen in Florida

((La/Sm)n = 2.85 – 15.71), likely as a consequence of LREE enrichment by goethite, while

HREEs are flat to only slightly depleted ((Gd/Yb)n = 0.83 to 4.82), reflecting the overall reduced

severity of leaching (and potential for in situ clay formation) impacting these sediments. As for the less weathered updip Cypresshead Formation and Hawthorn Group clays (Fig. 5-11H), their

REE patterns are consistent, possessing moderate LREE enrichment, flat HREE, and moderate negative Eu anomalies similar to what is noted for Georgia Cypresshead clays. Furthermore, the

162

similarity in the REE distribution of these clays suggests a reworked origin for the updip

Cypresshead clays in agreement with observations outlined in Chapter 4.

Whereas Eu anomalies (Eu/Eu*) are most commonly used to discern intracrustal igneous differentiation processes (McLennan et al., 1993), the vertical distribution of Eu anomalies in

Florida Cypresshead sediments are more significantly an indicator of weathering processes and kaolinite origin. For Florida samples, Eu/Eu* values range between 0.54 and 1.71, reflecting both positive and negative Eu anomalies of moderate magnitude, with the frequency distribution of values suggesting the occurrence of three modes, a positive Eu/Eu* mode at ~1.5, a flat to slightly positive Eu/Eu* mode at ~1.1, and a negative Eu/Eu* mode at ~0.7 (Figure. 5-12). The most positive Eu anomalies occur at of near the base of the Cypresshead Formation, where residual feldspar has been noted (Chapter 4), and where microtextural evaluation of these sediments suggests the alteration of mica into vermicular kaolinite. Given these associations, and the positive Eu anomaly determined for the EPK feldspar concentrate (Fig. 5-11G), it appears that the dissolution of feldspars possessing an original plagioclase component (EPK Feldspar

CaO = 0.432 %) imparted the observed Eu anomaly to the clay fraction. Occurring below the water table under dysoxic to anoxic conditions due to pore water organics, positive Eu anomalies would be preserved by clays formed in situ from the dissolution and/or weathering of feldspar and micas. Further evidence for this process is seen in the REE distributions of both the mica and vermicular kaolin (vermiform) concentrates (Fig. 5-11G), which show evidence of slight Eu enrichment, supporting their weathering (mica) or formation (vermiform) in proximity to feldspar dissolution. With decreasing depth from the surface, Eu anomalies decrease in magnitude, and ultimately transition into negative values in near surface clays (Fig. 5-11). This transition is likely in response to two near surface factors impacting the clay fraction and pore

163

12 Cypresshead - GA Cypresshead - FL* 10

Frequency 4

0 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 Eu/Eu*

Figure 5-12. Histogram illustrating the distribution of Eu/Eu* values for both Georgia and Florida Cypresshead Formation clays.

water organics; exposure to oxic weathering conditions and kaolinite recrystallization under

vadose or mixed vadose/saturated conditions near the water table. Contrasting with Florida

observations, Eu/Eu* values for Georgia Cypresshead samples are fairly constant (Eu/Eu* =

0.61-0.78), consistent with a moderate negative Eu anomaly.

Discussion

Hurst and Pickering (1997) have noted that “sedimentary kaolins” do not originate exclusively through sedimentation processes, but rather through a complicated assortment of post-depositional alteration processes acting on sediments. This is true, although original sedimentation characteristics can have a significant impact on post-depositional reaction pathways and the relative susceptibility of sediments to alteration processes.

Kaolinite Origin

Kaolinite in the Cypresshead Formation consists of at least three separate components; detrital kaolinite associated with discrete clay beds, stringers, and lenses that were deposited

164

during the initial sedimentation of the unit, neoformed kaolinite associated with the topotactic replacement and/or epitactic overgrowth of muscovite mica (and kaolinite) at the expense of dissolved K-feldspar, mica, and clays, and near surface recrystallized kaolinite which formed at the expense of precursor clays, including detrital and neoformed kaolinite, and other labile components under pedogenic (vadose) and fluctuating water table (vadose/saturated) conditions.

Although a significant portion of the kaolinite fraction found in Cypresshead Formation sediments is most certainly detrital in origin, the original mineralogical maturity of those precursor clays remains a question, as even the kaolins of the Georgia-South Carolina kaolin district were modified mineralogically following deposition as a less mature clay mineral suite.

Based on REE data evaluated for less weathered updip Cypresshead clay samples in north- central Florida and Georgia Cypresshead clays, it appears likely that the detrital clay component of the formation was derived from local sources, with little coast parallel reworking of the clay fraction from Georgia southward. However, limiting the origin of Cypresshead Formation clays to a detrital source is not corroborated by the mineralogical and geochemical observations compiled in this study. Among the possible pathways for the formation of an in situ kaolinite fraction are transformation (topotaxy), neoformation (epitaxy), and recrystallization weathering processes.

Kaolinite neoformation

Although mica is rather resistant to weathering under normal conditions, the kaolinization of micas, including muscovite, by topotactic replacement and epitactic overgrowths has been well documented in the literature as outlined by Jeong (1998b) and Arostegui et al. (2001). Mica, and particularly a muscovite mica substrate, fits the requirements for topotactic replacement by kaolinite in possessing a high degree of three-dimensional structural accord between both reactant and product. Additionally, mica grain surfaces are well suited to the epitactic

165

nonrandom overgrowth of kaolinite given that the mica cell dimensions in the a-b plane match

those of kaolinite to within 4% (Bailey, 1980). As such, the basal surface is an excellent template

for epitactic neoformation.

Based on SEM analysis, kaolinization of muscovite in Cypresshead sediments, and the

corresponding formation of vermicular kaolinite, appears to take place under reducing (dysoxic

to anoxic), saturated conditions according to the model proposed by Jeong (1998b). For

Cypresshead sediments, initial weathering of the muscovite takes place along grain edges leading

to the development of the characteristic splaying noted by both SEM and standard petrographic

microscopy (Figs. 5-13A and B). This first phase of the weathering process appears to involve

the topotactic replacement of muscovite grain edges, which generates tensional stresses within

the grain interior in response edge splaying. These stresses, in turn, result in the formation of

lenticular voids along interior basal cleavage surfaces and smaller voids along grain edges in

response parting, accessing basal grain surface area suitable for the crystallization of kaolinite

via an epitactic overgrowth model. Dissolution studies suggest that during this initial phase of weathering the muscovite surface is modified by the precipitation on an Al-hydroxide or kaolinite monolayer when surface K+ is exchanging with H+ in solution, an early stage alteration

process consistent with what is observed in soil when mica transforms to HIV (Harris et al.,

1992; Nagy and Pevear, 1993). With continued replacement of grain edges inward, splaying and

void development continues, resulting in the significant expansion of the muscovite flake along

the c-axis (Fig. 5-13C). Eventually exfoliation of discrete vermiforms takes place as topotactic

replacement continues to migrate toward the core of the grain, exposing additional grain edges

and basal surfaces to continued kaolinization (Fig. 5-13D).

166

Figure 5-13. SEM photomicrographs of Cypresshead and reworked Cypresshead Formation kaolinite textures associated with the in situ weathering of muscovite. A) Neoformed kaolinite developing along the edges (arrow) and on the basal surface (arrow) of muscovite (FRG-1-10), B) Broader view of same sample described in (A), C) Characteristic splayed edge of muscovite flake with resulting lenticular void development (arrow) and formation of pre-exfoliated vermiform, D) Detached vermicular kaolinite with residual muscovite core as confirmed by EDS.

This weathering process also provides an answer to the question surrounding the distribution of fine mica in coarse- to medium-sands and gravel. Not adequately explained by hydrodynamic sorting during sedimentation, the actual cause of this association is due to the previously described weathering process. As mica flakes are constantly exfoliated during kaolinization and vermicular kaolinite formation, the grains will constantly decrease in size.

Thus, mica flakes which may have been of a grain-size suitable to co-deposition with coarse

167

sediments will degrade through weathering, eventually decomposing into fine sand and silt sizes

consistent with earlier noted observations (Pirkle, 1960). Observations related to a decrease in

kaolinite order and CSD values in basal north-central Florida clays are also explained by this mica weathering model. If topotactic replacement is favored during the early stages of mica weathering, with epitactic overgrowth becoming the dominant neoformation processes over time, then early stage kaolinites associated with a high percentage of topotactic replacement are likely to possess greater disorder and smaller CSD values consistent with observations. With continued weathering, and an increased role for epitactic neoformation of kaolinite, kaolinite populations

should exhibit increased structural order and CSD values under saturated groundwater

conditions. Once exposed to fluctuating water table and/or vadose conditions, this weathering

model is likely to be replaced by dissolution and/or recrystallization reactions.

Kaolinizing muscovite grains altered according to this model exhibit a substantial increase

in grain volume. Since conservation of Al during the weathering process would dictate a volume

reduction following a topotactic only replacement model for muscovite, a substantial import of

dissolved Al (and Si) from an external weathering source is required. Several studies have

described the weathering of muscovite to kaolinite via dissolution-recrystallization (Banfield and

Eggleton, 1990; Jiang and Peacor, 1991; Singh and Gilkes, 1991), thus, participation of an

imported Al and Si component during epitactic neoformation is expected. Based on the

likelihood that Cypresshead Formations sands contained up to 5% or more K-feldspar during

initial deposition, the presence of an inherited Eu anomaly in basal Cypresshead clays, and given

the highly weathered character of skeletal feldspar grains noted by SEM, there should have been

sufficient Al and Si delivered into the aqueous environment to drive neoformation of kaolinite.

168

The dissolution rate constant of K-feldspar is one order of magnitude greater than that for

muscovite (Arostegui et al., 2001), supporting the view that K-feldspars were the most likely

source of Al and Si necessary to crystallize neoformed kaolinite via epitaxy onto muscovite

surfaces. Al and Si which would have been mobilized during muscovite weathering would have

been consumed by the topotactic replacement process. In order for this crystallization reaction to

work, H+ would need to be supplied, and pore water circulation would need to be continuous to

facilitate the removal of K+ and support a low aK+/aH+ ratio promoting kaolinite crystallization

(Arostegui et al., 2001). As a major aquifer recharge area, hydrologic conditions in north-central

Florida would be well suited to satisfy these requirements as evidenced by the open system

supply-controlled growth characteristics of Cypresshead clays.

Jeong (1998b) suggests up to a 9-fold increase in volume from primary micas to resultant

kaolinite through the vermiform crystallization process. Not only does this indicate that the

epitactic method of kaolinite neoformation is dominant with respect to topotactic replacement,

but this also represents a significant method by which total kaolinite (clay) content can increase

within a given sedimentary sequence. In order to achieve the 20% clay content as seen in the

EPK deposit, this would require the complete weathering of only a 2.2% mica fraction. The only

limiting factor for this being of significant importance in the formation of Cypresshead kaolinite

is the availability of dissolved Al and Si. Although dissolved Al is likely in Cypresshead

Formation pore waters, freshwater concentration in equilibrium with kaolinite is typically on the

order of 0.1 µg/l to 1 µg/l (Drever, 1982), with the dissolved concentration of silica commonly up to five orders of magnitude higher (1-100 mg/l; Hem, 1985). However, the concentration of dissolved aluminum in pore waters may be greatly increased by complexing with organic compounds, particularly monofunctional and difunctional carboxylic acids (acetic and oxalic

169

acids). Studies by Fein (1991) and Fein and Hestrin (1994) have demonstrated that aluminum-

oxalate complexation can dramatically increase aqueous aluminum concentrations. Subsequent

destabilization of these aluminum complexes may result in the formation of kaolinite in

siliciclastic units (Surdam et al., 1984; Small and Manning, 1994). With the Cypresshead

Formation known to concentrate secondary organic compounds in basal pore waters as is seen at

the EPK Mine site (Pirkle and Yoho, 1961), bacterial degradation of organic compounds may be a likely mechanism by which aluminum is released (Maliva et al., 1999).

Kaolinite recrystallization and disorder

Other than the origin of kaolinitic clays associated with the Cypresshead Formation, the

second most common question regarding clays in this unit is the unusually high degree of

structural order exhibited by the kaolinite fraction relative to the small particle-size of these

clays. As first noted by Pirkle (1960), it is the small particle size of Cypresshead clays which contributes to the overall greater strength, plasticity, surface area and base-exchange capacity of these clays as an industrial mineral product. In fact, variations in the structural order of kaolinite have been known for some time to correlate to industrial properties, including plasticity, brightness and viscosity (Murray and Lyons, 1956; Velho and Gomes, 1991; Chávez and Johns,

1995; Galán et al., 1998; Aparicio and Galán, 1999)). Based on the results of this study, it

appears that the in situ neoformed origin of kaolinite in the Cypresshead is the answer to this

question.

If a major component of the kaolinite fraction in north-central Florida sediments is derived

via the combined topotactic and epitactic replacement of mica at the expense of dissolved

feldspar, then a high degree of crystalline order is to be expected, particularly if the greatest

volume of kaolinite originates through epitaxial overgrowth on mica template surface. Such a

reaction pathway would favor the formation of a well ordered nucleated phase, particularly when

170

the highly leached nature of Cypresshead sediments below the water table is considered. With

few competing cations (e.g., Fe2+) capable of incorporation in the kaolinite structure at depth

within the unit, high purity and resultant, well-ordered kaolinite is predicted.

As evidenced by trends in CSD values, disorder, and microtextural characteristics of

Cypresshead kaolinite, recrystallization processes appear superimposed on kaolinite

neoformation in near surface sediments. Under vadose or mixed vadose/saturated conditions,

oxic weathering would prevail, favoring leaching of labile components, including clays, and the

recrystallization of precursor phases (Hurst and Pickering, 1997). Through dissolution of Si- and

Al-bearing phases, mobilized Si and Al would be allowed to participate in vadose crystallization

or could migrate vertically to below the existing water table. Evidence for strong leaching conditions persisting in near surface sediments is confirmed by the presence of gibbsite, which is evidence for laterization/feralization processes impacting Cypresshead sediments. Halloysite at

the near surface, along with disordered kaolinite consistent with pedogenic formation, is further

evidence for the extreme leaching, particularly at the north-central Florida localities which

exhibited the greatest spread in disorder and CSD values (EPK, FRG, and SSD-1).

The concentration of fine aggregates (or microaggregates) of anhedral to subhedral

kaolinite noted for near surface sediments in north-central Florida is further evidence for oxic

weathering, and likely a pedogenic-related origin. Consisting primarily of disordered and often single-phase kaolinite, these clays possess negative Eu anomalies consistent with formation and/or recrystallization from precursor phases, including kaolinite, under oxic conditions. This shift toward negative Eu anomalies in near surface sediments correlates well with trends in CSD values and disorder (Fig. 5-14), with the near surface decrease in Eu/Eu* values corresponding to the oxidation of Eu2+ to Eu3+ during dissolution, inhibiting the incorporation of Eu into

171

recrystallized phases. For near surface sediments associated with the Rio Jari kaolin deposit in

Brazil, disordered kaolinite sharing the same rounded aggregate morphology corresponds with an

increase in the amount of structural Fe substituted for Al in the kaolinite (Montes et al., 2002).

Although no measurements of Fe in Cypresshead kaolinites have been performed, incorporation

of trace Fe could be possible due to the weathering of accessory heavy minerals. However, as

noted previously, the halloysite morphology observed in near surface Cypresshead sediments is

inconsistent with elevated Fe concentrations.

Of additional significance to observed trends in disorder and CSD values are the local

hydrologic conditions existing at the various Cypresshead localities evaluated for this study. As

noted by Hurst and Pickering (1997), location of kaolins within the groundwater system can have

a significant impact on the compositional variations of the deposit. This seems particularly true

for north-central Florida Cypresshead localities, where increased concentrations of detrital or

residual clays in near surface sediments (FRL-1 and SSJ-1) appear to impede vadose and mixed

vadose/saturated leaching and recrystallization, resulting in a lack of significant spread in near

surface disorder and CSD, and the absence of gibbsite, halloysite, and crandallite-florencite

phases which appear to indicate strong, even lateritic/feralitic, leaching conditions.

Accessory Phase Paragenesis

Significant weathering reactions driving the post-depositional alteration of Cypresshead

sediments occur under both saturated and unsaturated groundwater conditions. Below or at the

water table, the formation of halloysite as a kinetically favored metastable precursor phase to

kaolinite with subsequent recrystallization to disordered kaolinite as the thermodynamically

stable phase is likely favored. The main factors favoring kinetic control of halloysite formation

are intense, but short wet periods followed by prolonged extremely dry seasons, and

microenvironmental conditions leading to immediate uptake of released Al by the halloysite-

172

14.0

2 12.0 R = 0.47

10.0

8.0

6.0

4.0 Cypresshead - FL* Cypresshead - GA

Volume-Weighted Mean Thickness (nm) Mean Volume-Weighted Cypresshead - FL* 2.0 0.40.60.81.01.21.41.61.8 Eu/Eu*

Figure 5-14. Correlation of Eu/Eu* values to CSD (volume-weighted mean thickness) calculations (* Florida Cypresshead Formation samples include SSD-1).

precursor mineral allophane. Halloysite is known to be a common metastable precursor to

kaolinite (Jeong, 1988a; Joussein et al., 2005), with the crystallization of metastable phases often

favored in low-temperature geological environments. Although, nucleation of more soluble

halloysite is kinetically easier due to a high surface free energy (Morse and Casey, 1988; Stumm,

1992), metastable phases such as halloysite will ultimately transform to stable phase such as

kaolinite with time. In response to eventual dehydration and recrystallization, the halloysite

crystallized in Cypresshead sediments with ultimately alter into kaolinite via the halloysite →

metahalloysite → kaolinite reaction pathway described by Hurst and Pickering (1997). This is

consistent with thermodynamic considerations as the Gibbs free energy of formation of

halloysite is approximately 4 kcal/mol higher than that of kaolinite (Anovitz et al., 1991).

The crystallization of halloysite under near surface water table conditions is most likely

+ related to the interplay of pH (H+ activity), and the activities of Al species, K , and H4SiO4 under conditions favorable to repeated wetting and drying (Ziegler et al., 2003). As indicated by

173

the curved or tubular morphology of Cypresshead halloysite, the relative uptake of tetrahedrally

coordinated Al as a function of pH may also be of importance to the preferred crystallization of

this phase, with the change in coordination state from an octahedral state taking place between pH 5.5 and 6.5 at 25°C (Merino et al., 1989). With a shifting of this pH range to lower values at higher temperature, one would predict tetrahedrally coordinated Al under near surface conditions in north-central Florida based on the pH range of 4.5 to 5.5 noted by Heuberger (1995) for soils developed on the Cypresshead Formation in the Ocala National Forest.

Two likely mechanisms exist that could potentially explain the origin of crandallite- florencite phases ((Ca,REE)Al3(PO4)2(OH)6) in Cypresshead Formation sediments; (1)

diagenetic precipitation in response to organic decomposition, or (2) crystallization in response

to the weathering of a precursor phosphate phase (francolite). As originally proposed by

Rasmussen (1996) for sandstones, the first model suggests that authigenic precipitation could

occur in response to organic matter decomposition coupled with detrital mineral dissolution

shortly after sediment burial. This seems unlikely as the sedimentary carbonate-fluorapatite,

francolite, is kinetically favored for early diagenetic crystallization during marine sedimentation

(Flicoteaux and Lucas, 1984). Additionally, trace fossil assemblages found in Cypresshead

sediments (Chapter 4) do not suggest a high organic matter accumulation rate in these sediments

during their initial deposition. However, an alternative source of organics is the post-depositional

concentration of pore water organics noted at the EPK Mine and other sites. Continuously

delivered to the groundwater system via the surface decomposition and infiltration of organic

detritus, pore water organics would periodically decomposed in response to water table

fluctuations and corresponding redox conditions. As suggested by the vertical distribution of

P2O5 concentrations in most sampled sections (Table 5-4), this would liberate phosphorous (P) to

174

surrounding pore waters near the water table interface where crandallite-florencite appears to be concentrated. Additionally, the dissolution of plagioclase feldspars, postulated as a source of Al,

Si and the inherited positive Eu anomalies associated with the formation of in situ kaolinite could potentially be the source of calcium (Ca) required for crandallite-florencite crystallization.

As for the weathering of trace reworked phosphate (francolite), dissolution of francolite would supply both Ca and P to surrounding pore waters, with the addition of Al necessary for crandallite-florencite crystallization supplied by coeval clay, mica, and/or feldspar weathering

(Flicoteaux and Lucas, 1984). Proposed original deposition of minor francolite with Cypresshead sediments would be consistent with the minor accumulations of phosphate grains noted for southern Florida Late Miocene through Pliocene siliciclastics, where reworking of Hawthorn

Group phosphate is the most likely source (Cunningham et al., 1998). However, geochemical evidence is lacking to support this model given no negative Ce anomalies associated with the chondrite-normalized REE patterns of Florida Cypresshead clays.

Conclusions

Weathering processes are known have a profound impact on original sediments which host commercial kaolin deposits. Among the effects attributed to the intense post-depositional alteration of such deposits are; (1) the strong compositional and textural modification of the original sediments, (2) whitening of the sediments by partial removal of organic matter, Fe and

Mn, and (3) recrystallization of kaolinite (Hurst and Pickering, 1997). In the case of Cypresshead

Formation, complex hydrogeologic controls and post-depositional processes are ultimately responsible for the formation and recrystallization of the kaolinites. Focusing on the clay mineralogy and microtexture of Cypresshead sediments, the results outlined in this chapter highlight evidence for the origin and weathering history of Cypresshead kaolinites in order to

175

define processes impacting their mineralogy and industrial properties, and answers the primary

verses secondary origin issue associated with these clays. Among the results are the following:

• Cypresshead kaolinite consists of at least three fractions; in situ kaolinite formed at the expense of feldspars and mica, detrital kaolinite deposited as part of the original clay mineral suite, and near surface recrystallized kaolinite.

• In situ kaolinite crystallizes via the combined topotactic (transformation) and epitactic (neoformation) weathering of muscovite mica consistent with the model of Jeong (1988b) and confirmed by vertical trends in kaolinite disorder and coherent scattering domain (CSD) values.

• The muscovite weathering model of Jeong (1988b) explains the unusual low disorder/small particle-size characteristics of Florida Cypresshead kaolinite, the distribution of fine mica in coarse- to medium-sands and gravels, and the relatively high (> 20%) total clay content observed in some basal Cypresshead sediments.

• Feldspar dissolution as a source of Al and Si necessary for the crystallization of in situ kaolinite and post-depositional increase in Cypresshead clay content is confirmed by the presence of residual feldspars in basal Cypresshead sediments and the inherited positive Eu anomalies likely resulting from an original plagioclase component to the feldspar suite.

• Pore water organics are likely important in concentrating dissolved Al in order to crystallize in situ kaolinite and in controlling the pore water redox environment of Cypresshead sediments as indicated by the vertical trends in Eu anomalies and HREE depletion.

• Near surface recrystallized kaolinite formed under oxic, vadose or mixed vadose/saturated conditions are disordered and possess microtextural characteristics consistent with a pedogenic origin.

• Occurrences of near surface formed halloysite and gibbsite are consistent with Cypresshead sampling locations characterized by extensive evidence of vadose and mixed vadose/saturated leaching and recrystallization under oxic conditions.

• Trace crandallite-florencite minerals likely originated from the decomposition of post- depositional pore water organics coupled with detrital mineral dissolution.

176

CHAPTER 6 TRACE ELEMENT AND ND ISOTOPIC EVIDENCE FOR THE PROVENANCE OF CYPRESSHEAD FORMATION KAOLINITIC SANDS

Introduction

The geological significance of kaolinitic sands occurring in the Pliocene Cypresshead

Formation (3.4–2.3 Ma) of north-central Florida and southeastern Georgia has long been a

question confounding Coastal Plain geologists. Although the formation contains economic kaolin

concentrations in north-central Florida, little is known concerning these deposits, particularly the

provenance of the clay fraction. Overshadowed by the large economic kaolin deposits of the

Georgia-South Carolina kaolin district (Fig. 3-1), the Cypresshead has received limited study,

and remains poorly understood within the overall context of Coastal Plain deposition. As is the

case with the kaolin deposits in Georgia and South Carolina (Keller, 1978; Patterson and Murray,

1984; Dombrowski, 1992; Hurst and Pickering, 1997), several theories on the origin of these

kaolinitic sands have been proposed, but have failed to fully define provenance. Existing data on

the provenance of Cypresshead sediments prior to this study has been limited to information on

monocrystalline and polycrystalline quartz grain textures identified by Kane (1984), which

suggest sources consistent with granitoids, quartzo-feldspathic gneisses and /or schists, and

heavy mineral suites collected from the Cypresshead (Pirkle et al., 1964; Kane, 1984) which

have supported the previous view of a mixed igneous and metamorphic source terrane for these

sediments. In this study, both trace element and neodymium (Nd) isotopic data were used to

answer the question of Cypresshead provenance with respect to possible crystalline

(metamorphic and/or plutonic) source rocks in the southern Piedmont and reworked kaolinite

from the Georgia-South Carolina kaolin district.

In the past two decades, attention has focused on the use of trace elements, particularly the

high-field-strength elements (HFSE) such as the rare earth elements (REEs), Zr, Th, and U, and

177

Nd isotopic systematics as constraints on provenance (Tabbutt, 1990; Linn et al., 1991;

Dombrowski, 1992; Toulkeridis et al., 1994; Condie et al., 1995; Gleason et al., 1995). The

REEs are often reliable geochemical tracers of sedimentary provenance because of their characteristic immobility and resistance to elemental fractionation in the supracrustal environment (Wildeman and Condie, 1973; Piper, 1974; Nesbitt, 1979; Chaudhrui and Cullers,

1979). However, evidence to the contrary of immobility has been given by Nesbitt (1979),

Nesbitt and Markovics (1997), and Wade (2002). As shown by Nesbitt and Markovics (1997), careful monitoring of REE distributions in a weathering profile reveals a tendency toward light

(L) REE enrichment (relative to heavy (H) REEs) in the most extremely weathered materials of the profile. Therefore, profiles showing slight enrichment in LREEs can be explained by this mechanism. Additionally, as outlined in Chapter 5 of this study, variations in the redox potential of pore waters can have a significant impact on the REE distribution of fine-grained sediments.

In general though, certain trace elements, such as the REEs with a valence of +3, Th, and Sc, and to a lesser extent Co and Cr, are believed to be transported quantitatively from source rocks into sediments (McLennan, 1989). These elements also are the least prone to diagenetic redistribution, thus providing useful information on the composition of sediment sources (Taylor and McLennan, 1985; McLennan, 1989). REE distributions appear to be sensitive to tectonic setting (Cullers et al., 1987; Cullers, 1988; McLennan, 1989; McLennan et al., 1993; and others), and are particularly useful when evaluating fine-grained sediments.

Although many studies have used the distribution of REEs and other trace elements to constrain provenance (Dypvik and Brunfelt, 1979; Bhatia, 1985; Tabbutt, 1990), it is the Nd isotopic character of sediments which has received the most attention recently in identifying potential source terrains for clastic sediments (Frost and O'Nions, 1984; Miller and O'Nions,

178

1984; 1985; Frost and Winston, 1987; Nelson and DePaolo, 1988; Ghosh and Lambert, 1989;

Bouquillon et al., 1990; Tabbutt, 1990; Linn et al., 1991; Gleason et al., 1994; 1995). Because

weathering and diagenesis can affect Sm-Nd isotopic ratios in sediments (Ohr et al., 1991;

McDaniel et al., 1994; Bock et al., 1994), spurious Nd model ages (TDM) may result. Therefore,

initial εNd values (εNd(t)) calculated for the stratigraphic age of a sediment are often emphasized.

This approach assumes that any diagenetic effects resulting in the disturbance of Sm-Nd isotopic ratios would have occurred near the time of deposition; thus, calculated εNd(t) values should

reflect true εNd of the material at that time (Gleason et al., 1995). Furthermore, using εNd notation

permits the comparison of data sets that are normalized to different Nd isotopic ratios without

accounting for normalization differences.

Southern Piedmont

The deeply weathered rocks of the Piedmont, stretching from Alabama in the southwest to

southernmost New York in the northeast, have long been considered a major source for

sediments deposited along the Atlantic Coastal Plain, and as the ultimate source of sediments

which are now the Cretaceous “soft” and Tertiary “hard” kaolins in central and eastern Georgia

(Murray, 1976; Keller, 1977; Austin, 1978; Hurst and Pickering, 1989). Weathering of these

crystalline rocks produced kaolinite-metahalloysite-rich saprolites, which on erosion produced

detritus delivered to the ancient Georgia coastline via fluvial transport. Subsequent laterization of

these sediments played an important role in kaolin formation (Austin, 1978), with periods of sea

level regression favoring intense weathering, particularly in updip sediments (Lowe, 1991).

The southern Piedmont can be divided into tectonostratigraphic terranes on the basis of

age, origin, and shared affinity, and further divided into metamorphic belts based mainly on

lithological characteristics, metamorphic grade, and in some areas, structural boundaries. For this

study, the terranes of interest include the Carolina terrane, the Inner Piedmont, and Eastern slate

179

belt (Fig. 6-1). Possible source rocks for kaolinitic sediments include both metamorphic and igneous rocks within these terranes. Of particular interest are the Carboniferous-Permian

Alleghanian (285-340 Ma) granites found throughout the southern Piedmont (Fig. 6-1), which are but one of three groups of plutons emplaced in the area. The other groups include a Late

Proterozoic-Cambrian (495-734 Ma) mafic to felsic calc-alkaline series and a Silurian-Devonian

(436-375 Ma) mafic-felsic bimodal series with alkalic tendencies, both occurring mainly in the

Charlotte belt portion of the Carolina terrane (McSween et al., 1991).

Georgia-South Carolina Kaolin District

The Cretaceous and Tertiary kaolin-bearing sedimentary rocks of the Georgia-South

Carolina kaolin district lie to the southeast of the Fall Line, the contact between the crystalline rocks of the southern Piedmont and the sedimentary assemblage of the Atlantic Coastal Plain. A review of the stratigraphy associated with the Cretaceous and Tertiary kaolins of the Georgia-

South Carolina kaolin district is given by Pickering and Hurst (1989) and Dombrowski (1992).

Dombrowski (1992; 1993) identified La, Th, Co, and Sc as effective provenance indicators for

Cretaceous and Tertiary kaolins based on the concentration of Th and La in felsic (acidic) source rocks, and Sc and Co in mafic (basic) source rocks, as well as the relative immobility of these elements during weathering processes. As such, results from the work of Dombrowski (1992;

1993) are addressed in the discussion section of this study.

The Cretaceous-Tertiary boundary is a major unconformity in the Georgia-South Carolina kaolin district (Buie and Fountain, 1967; Murray, 1976; Patterson and Murray, 1984; Nystrom et al., 1986; Pickering and Hurst, 1989), marking the boundary between Cretaceous "soft" kaolins and Tertiary "hard" kaolins. Above the unconformity in middle and eastern Georgia and South

Carolina are Paleocene through Middle Eocene sediments assigned to the Huber Formation

(Buie, 1978). Nystrom et al. (1986) restricts this term to Early to Middle Eocene sediments.

180

Figure 6-1. Tectonostratigraphic terranes and granites proposed as potential source materials for Cypresshead Formation sediments (modified after Horton et al., 1989; Samson et al., 1995).

Commercial kaolin deposits of Tertiary age are concentrated near the top of the Huber Formation and are elongated from northeast to southwest, parallel to the orientation of the paleoshoreline

(Dombrowski, 1992). Below the unconformity in Georgia and South Carolina lie Late

Cretaceous sediments composed of continental to near-shore marginal marine deposits adjacent to the Fall Line, with marine deposits in the remainder of the region (Kesler, 1956; Herrick,

1961; Herrick and Vorhis, 1963; Applin and Applin, 1964; Gohn et al., 1979; Nystrom et al.,

1986; and others). Those sediments located adjacent to the Fall Line are referred to as the

Buffalo Creek Formation west of the Ocmulgee River in western Georgia (Pickering and Hurst,

1989) and as "Cretaceous Undifferentiated" or "Unnamed Cretaceous Sediments" east of the river in South Carolina (Eargle, 1955; Nystrom et al., 1986; Pickering and Hurst, 1989).

181

Results

Provenance investigations of the kaolin deposits of the Georgia-South Carolina kaolin

district have focused on both plutonic and metamorphic source rocks in the southern Piedmont as

primary sources for kaolinitic sediments delivered to these locations (Dombrowski, 1982; 1992;

Dombrowski and Murray, 1984). A similar provenance is possible for the kaolinitic sands of the

Cypresshead Formation in southeastern Georgia and peninsular Florida, or that kaolinite

accumulated in response to reworking of kaolin district sediments. In order to answer that

question, the trace elemental and Nd isotopic characteristics of possible source rocks in the

southern Piedmont and Georgia-South Carolina kaolin district are used in conjunction with

similar data from Cypresshead clays to decipher characteristics of sediment provenance.

Trace Elements

Trace element concentrations of Cypresshead Formation clay (< 2 µm) separates and

comparison samples are reported in Table 6-1 along with select major elements, ∑REE, and applicable elemental ratios. A complete summary of all trace elements analyzed for this study is included in Appendix H. In comparison with average continental crust (ACC), the concentrations of most of the trace elements for Florida Cypresshead samples are relatively low, except for Pb,

Th, and U (Fig. 6-2A). Rb and Co are particularly depleted, averaging a relative concentration ratio of ~ 0.1 or greater. Zr is also relatively depleted in most samples other than the Davenport

Mine composite. This sample possesses an anomalously high Zr concentration in response to zircon enrichment in several of the near surface samples used to calculate the average used in

Figure 6-2A. Trace element concentrations for Georgia Cypresshead samples are slightly more enriched than Florida samples (Fig. 6-2B). As with the Florida samples, Pb, Th, and U remain enriched relative to ACC while Cr, V, and Rb, although depleted to equivalent with ACC, are more enriched relative to Florida samples. Sr, however, is more depleted in Georgia samples

182

Table 6-1. Trace element concentrations (ppm) and elemental ratios for Cypresshead Formation and comparison samples.

Sample ID Interval Rb Sr Th U Zr Sc V Cr Co Th/Sc Th/U Zr/Sc La/Sc La/Th P 2O5 (%) TiO 2 (%) Fe 2O3 (%) ΣREE Eu/Eu* (La/Yb) n (La/Sm) n (Gd/Yb) n Cypresshead Formation - FL EPK36-J-12 25-27 2.7 48.8 5.1 2.51 45.2 3 24 43 3.1 1.7 2.0 15.1 6.1 3.6 0.14 0.25 0.27 75.24 0.71 30.76 5.16 3.75 EPK36-J-12 35-40 3 121 7 4.79 73.7 5 18 16 3.7 1.4 1.5 14.7 5.9 4.2 0.15 0.23 0.36 129.76 0.69 24.74 4.25 4.32 EPK36-J-12 46-48 4.2 35.9 5.7 2.06 32 10 36 43 3.1 0.6 2.8 3.2 2.3 4.0 0.07 0.28 0.49 93.95 0.94 30.22 5.16 3.76 EPK36-J-12 50-53 5.2 21.5 3.5 1.93 24.9 10 32 16 2.4 0.4 1.8 2.5 2.0 5.7 0.05 0.32 0.51 76.26 0.94 33.43 5.61 3.25 EPK36-J-12 59-62 6.1 37.2 6.9 4.86 38 16 60 45 2.6 0.4 1.4 2.4 2.2 5.0 0.07 0.55 0.77 177.87 1.43 28.75 2.91 5.48 EPK31-P-40 35-45 4.5 21.3 4.7 17.7 22.2 12 52 45 1.8 0.4 0.3 1.9 1.7 4.3 0.06 0.35 0.66 85.86 1.42 27.02 4.99 3.24 EPK31-P-40 50-62 5.2 34.8 4.4 103 18.5 21 47 21 2.3 0.2 0.0 0.9 1.3 6.4 0.07 0.31 0.84 153.41 1.71 14.56 3.12 3.10 EPK31-P-40 62-65 17.6 123 7.8 92.9 30.5 20 77 67 3.4 0.4 0.1 1.5 2.3 5.9 0.17 0.29 1.24 294.02 1.42 10.58 2.10 3.71 EPK30-V-6 22-24 25.7 348 15.1 5.58 177 6 63 65 5.2 2.5 2.7 29.5 9.1 3.6 0.22 0.59 2.55 275.15 0.60 13.99 3.26 3.10 EPK30-V-6 30-35 23.1 2040 56.9 46.3 255 13 79 42 5.2 4.4 1.2 19.6 18.7 4.3 1.18 0.30 1.53 1367.54 0.65 12.79 2.49 3.81 EPK30-V-6 48-53 12.7 549 23 29.3 151 12 79 38 4.6 1.9 0.8 12.6 7.5 3.9 0.63 0.42 0.89 447.19 0.75 14.66 2.84 3.67 EPK30-V-6 58-63 6 19.8 3.8 18.9 23.1 14 84 46 1.7 0.3 0.2 1.7 2.2 8.0 0.05 0.25 0.41 117.75 1.36 40.66 5.52 4.75 EPK30-V-6 68-73 8.9 25.9 4.3 16.8 23 22 113 20 2.5 0.2 0.3 1.0 1.1 5.8 0.06 0.43 0.67 116.17 1.35 12.76 3.56 2.27 FRG-1 3 4.6 229 17.7 5.81 84.9 7 61 59 4.6 2.5 3.0 12.1 6.3 2.5 0.28 0.31 2.99 215.86 0.63 17.43 3.29 3.68 FRG-1 6 0.8 125 6 1.62 19.1 5 16 17 5.5 1.2 3.7 3.8 7.2 6.0 0.16 0.07 0.99 177.95 0.89 40.12 3.47 7.08 FRG-1 7 1.8 76 7.4 1.16 16.4 4 18 41 3.5 1.9 6.4 4.1 7.8 4.2 0.09 0.08 0.64 138.99 0.92 52.16 4.38 7.33 FRG-1 10 6.7 454 11.7 1.57 24.9 6 21 10 3.5 2.0 7.5 4.2 23.2 11.9 0.24 0.22 0.61 654.71 1.07 92.95 3.71 13.39 FRG-1 11 1.7 83.5 7.1 0.91 17.7 5 16 6 2.1 1.4 7.8 3.5 8.1 5.7 0.07 0.11 0.23 162.92 1.14 90.50 5.57 9.33 FRG-1 13 1.7 33.1 5.6 1.04 21.5 5 18 8 1.9 1.1 5.4 4.3 4.1 3.6 0.05 0.12 0.21 78.11 1.43 67.87 6.59 5.94 FRG-1 15 2.3 35.4 6.5 1.38 18.2 11 27 10 1.8 0.6 4.7 1.7 5.0 8.5 0.08 0.23 0.23 259.13 1.34 92.11 4.42 10.44 FRG-2 3 1.7 46.5 5.3 0.83 22.7 5 11 13 2.1 1.1 6.4 4.5 2.9 2.7 0.06 0.11 0.27 61.72 1.09 47.81 3.04 8.01 FRG-2 5 2.4 28.1 5.9 1.01 16.3 5 20 33 1.5 1.2 5.8 3.3 4.3 3.7 0.05 0.11 0.36 94.43 1.37 72.22 4.94 8.05 FRG-2 7 2.3 16.4 3.9 1.19 15.6 7 16 7 1.3 0.6 3.3 2.2 2.4 4.4 0.05 0.15 0.27 78.45 1.39 57.17 4.06 7.45 FRG-2 10 5.8 66.9 10 2.27 23.3 11 39 13 1.9 0.9 4.4 2.1 4.8 5.2 0.13 0.27 0.63 265.03 1.37 35.04 3.05 7.28 FRG-2 12 10 99.2 9 4.62 29.9 15 57 47 2.6 0.6 1.9 2.0 3.6 5.9 0.14 0.36 1.40 299.29 1.05 22.28 2.44 6.38 FRL-1 2 5.1 21.5 7.9 1.28 46 14 50 50 3.7 0.6 6.2 3.3 2.2 3.8 0.06 0.31 0.83 114.00 1.15 50.32 6.40 4.14 FRL-1 4 4.7 26.8 10.6 1.16 49.6 14 61 53 3.8 0.8 9.1 3.5 3.3 4.3 0.06 0.32 0.94 187.78 1.27 60.85 6.10 5.26 183 FRL-1 6 3.4 32.1 8.3 1.13 43.3 15 35 13 2.4 0.6 7.3 2.9 3.0 5.4 0.07 0.25 0.61 239.12 1.17 50.15 3.19 7.77 FRL-1 7 3.6 22.8 6.1 1.27 32.1 13 30 14 2.5 0.5 4.8 2.5 2.6 5.5 0.05 0.22 0.59 176.47 1.09 37.34 3.45 6.04 SSJ-1 2 5.8 87 29.4 3.11 216 5 78 92 4.6 5.9 9.5 43.2 10.7 1.8 0.12 1.15 3.22 186.90 0.62 17.10 10.04 1.13 SSJ-1 4 8.3 70.1 18.3 2.67 117 6 61 58 4.3 3.1 6.9 19.5 5.4 1.8 0.14 0.58 2.14 123.32 0.62 18.05 6.89 1.63 SSJ-1 6 5.9 49.2 2.2 3.03 18.8 6 27 11 2.2 0.4 0.7 3.1 1.8 4.9 0.06 0.16 0.47 41.95 1.46 35.78 5.50 3.55 SSJ-1 8 10.8 47.3 2.4 7.93 23.9 12 56 42 1.4 0.2 0.3 2.0 0.8 4.0 0.05 0.25 0.40 36.91 1.58 32.10 6.58 3.11 SSJ-1 10 11 43.2 4.6 14.1 33.7 18 85 37 1.4 0.3 0.3 1.9 0.9 3.6 0.06 0.41 0.51 67.29 1.44 27.42 5.62 3.01 SSD-1 1 21.3 777 145 31.3 1630 22 245 160 5.5 6.6 4.6 74.1 22.4 3.4 1.98 7.97 2.12 1886.80 0.55 16.82 6.77 1.50 SSD-1 3 5.2 144 56.8 7.54 180 7 41 77 4.5 8.1 7.5 25.7 17.6 2.2 1.67 0.72 0.56 491.74 0.54 51.41 5.71 4.79 SSD-1 6 4.1 41.2 10 1.64 34.8 8 44 50 3.4 1.3 6.1 4.4 5.4 4.3 0.11 0.23 0.66 221.75 1.03 47.92 2.88 9.38 SSD-1 7 4.5 41 6.8 2.71 28.8 10 45 16 2 0.7 2.5 2.9 5.0 7.4 0.08 0.26 0.73 276.66 1.08 55.84 2.78 11.16 SSD-1 10 7.8 38 4.4 2.04 23 9 33 18 1.7 0.5 2.2 2.6 2.5 5.2 0.08 0.23 1.04 123.00 1.26 30.49 3.06 5.90 TRF2214 60.0-62.5 29.2 66 10.6 2.43 127 15 119 60 3.6 0.7 4.4 8.5 2.0 2.8 0.07 0.84 1.79 103.82 0.65 12.33 7.58 1.27 WEX164 18.0-26.0 87.7 320 22.3 13.3 63.9 26 189 169 10.2 0.9 1.7 2.5 2.9 3.4 0.41 0.84 4.35 333.07 0.69 9.98 4.31 1.84 WEX366 9.0-10.0 124 118 26.6 9.9 238 23 183 122 5.9 1.2 2.7 10.3 2.0 1.7 0.13 1.21 9.42 165.97 0.62 11.01 7.11 1.26 Cypresshead Formation - GA J-1 2 23.3 35.9 36.6 5.21 231 21 195 110 6.3 1.7 7.0 11.0 2.3 1.3 0.11 1.40 7.65 177.86 0.61 12.65 10.47 0.83 J-1 4 39.4 34.8 21.2 3.04 74.7 24 190 77 5.4 0.9 7.0 3.1 2.6 2.9 0.08 1.03 4.48 239.90 0.68 46.21 8.53 2.88 J-1 6 28.2 23.9 14.4 2.03 57.5 17 67 45 6 0.8 7.1 3.4 3.0 3.6 0.03 0.77 1.90 255.25 0.72 49.20 5.39 4.82 L-1 3 17.4 40.5 34.3 2.87 104 19 100 80 5.8 1.8 12.0 5.5 8.3 4.6 0.11 1.03 1.72 649.57 0.72 88.04 11.47 3.49 L-1 5 6.4 10.7 10.8 1.16 38.7 10 68 59 7.1 1.1 9.3 3.9 4.7 4.4 0.04 0.37 0.84 165.26 0.68 78.91 13.87 2.97 L-1 6 21.5 28.8 16.6 1.85 60.3 16 80 24 6.1 1.0 9.0 3.8 7.0 6.7 0.06 0.85 1.62 365.18 0.66 93.62 15.71 3.36 B-1 2 97.4 52.6 15.9 4.01 82.3 25 149 84 7.3 0.6 4.0 3.3 2.2 3.4 0.09 0.67 5.62 210.41 0.67 21.20 6.93 1.80 B-1 3 116 77.4 20.4 5.55 83.6 31 198 91 10.4 0.7 3.7 2.7 2.1 3.1 0.21 0.72 10.91 282.20 0.70 16.36 5.16 1.94 B-1 5 88.5 112 18.5 4.03 74.4 36 168 30 8 0.5 4.6 2.1 2.5 4.8 0.09 0.74 3.63 466.08 0.78 10.86 2.85 2.43 Hawthorn Group, Coosawhatchie Formation - FL/GA MCB109 15.0-20.0 139 112 12.7 4.38 84.9 21 177 127 11.5 0.6 2.9 4.0 2.7 4.4 0.08 0.69 5.98 226.61 0.68 10.46 5.51 1.44 J-1 BC 42.9 60.8 11.3 1.72 37.6 22 139 65 44.4 0.5 6.6 1.7 4.3 8.3 0.03 0.62 6.03 496.85 0.77 6.03 3.01 1.63 Huber Formation - GA KGa-2 — 1.2 51.2 13 3.52 78.6 15 105 29 9.4 0.9 3.7 5.2 3.1 3.6 0.05 1.43 0.94 173.53 0.77 17.35 6.70 1.86 ECCI-CB — 2.6 45.2 14.7 4.39 51.9 20 76 16 9.9 0.7 3.3 2.6 2.2 2.9 0.06 1.11 0.90 198.54 0.72 32.02 2.83 5.81 Buffalo Creek Formation - GA Note:KGa-1 Detailed — major — element 43.2 36.3and REE 2.27 data 98.1 (∑REE 19 in 223ppm) are 38 included 3.3 1.9in Tables 16.0 5-4 5.2 and 1.85-6 of 0.9Chapter 0.06 5. 1.57 0.19 194.95 0.74 37.56 1.78 14.75 ECCI-BC — — 50.5 31.4 1.53 115 14 127 16 3.1 2.2 20.5 8.2 4.2 1.9 0.08 1.25 0.16 434.50 0.56 65.87 1.52 18.60 TKC-EA — 0.6 36.7 31.1 8.72 111 21 102 26 4.4 1.5 3.6 5.3 0.9 0.6 0.06 1.45 0.17 87.86 0.59 25.68 2.76 10.04 DBK-B93 — 0.3 41.2 31.9 6.52 149 22 249 90 1 1.5 4.9 6.8 1.7 1.1 0.07 2.21 0.24 262.40 0.63 24.47 1.36 10.20

Figure 6-2. Multi-element normalized diagrams for Cypresshead and comparison samples, normalized against average continental crust (Wedepohl, 1995). A) Florida Cypresshead composite samples, B) Georgia Cypresshead composite samples, C) Cypresshead Formation (FL and GA) composite and comparison units.

184

relative to Florida. In comparing a composite of all Cypresshead Formation samples to Hawthorn

Group, Huber Formation, and Buffalo Creek Formation composites (Fig. 6-2C), the most notable

difference is related to the significant depletion of Rb for both the Huber and Buffalo Creek

composites. For comparison to the work of Dombrowski (1992; 1993), the concentration ranges

of Sc, Co, and Th for Florida Cypresshead samples are 3–26 ppm, 1.3–10.2 ppm, and 2.2–145.0 ppm respectively. Georgia Cypresshead samples possess concentration ranges of 10–36 ppm,

5.4–10.4 ppm, and 10.8–36.6 ppm. A detailed summary of REE distributions in Cypresshead

clays is discussed in Chapter 5.

A correlation matrix based on the trace elements, major elements, and elemental ratios of

all Cypresshead Formation samples reported in Table 6-1 is included in Table 6-2. Significant

correlations (r > 0.6) highlight the effect of accessory mineral phases on the trace element

concentrations of Cypresshead clays, and include correlations between P2O5 and Th (r = 0.92),

Zr (r = 0.76), and Sr (r = 0.63). TiO2 also correlates with both Th (r = 0.90) and Zr (r = 0.98), but

can be differentiated from P2O5 by additional correlations with V (r = 0.72) and Cr (r = 0.63).

Fe2O3 exhibits significant correlations with Rb (r = 0.90), V (r = 0.66), Cr (r = 0.69), and Co (r =

0.64). The correlation between P2O5 and ∑REE was previously addressed in Chapter 5.

Additionally, the geochemical results for CMS Source Clays (KGa-1 and KGa-2) are considered

to be of questionable quality due to uncertainty surrounding the preparation of these standards

which were originally generated for comparative X-ray diffraction (XRD) studies.

Neodymium (Nd) Isotopes

Research has shown that the Nd isotopic composition of sedimentary rocks reflects the

integrated composition of contributing source terranes (Nelson and DePaolo, 1993; McLennan et al., 1990; Linn et al., 1991). For this study, Nd isotope data from the Cypresshead Formation and comparison clay-bearing strata (Hawthorn Group, Huber Formation, and Buffalo Creek

185

Table 6-2. Correlation matrix of select trace elements and elemental ratios for all Cypresshead Formation samples. Rb Sr Th U Zr Sc V Cr Co Th/Sc Th/U Rb 1.00 Sr 0.18 1.00 Th 0.22 0.58 1.00 U 0.10 0.30 0.19 1.00 Zr 0.19 0.42 0.94 0.16 1.00 Sc 0.57 0.13 0.26 0.50 0.29 1.00 V 0.71 0.31 0.65 0.26 0.68 0.76 1.00 Cr 0.68 0.24 0.64 0.16 0.60 0.51 0.86 1.00 Co 0.64 0.42 0.46 0.09 0.35 0.29 0.60 0.77 1.00 Th/Sc 0.02 0.46 0.78 0.02 0.61 -0.18 0.28 0.47 0.44 1.00 Th/U -0.18 -0.12 0.15 -0.45 0.08 -0.41 -0.20 0.01 0.08 0.41 1.00 Zr/Sc 0.10 0.42 0.87 0.06 0.87 0.00 0.52 0.59 0.43 0.83 0.24 La/Sc -0.04 0.65 0.71 0.03 0.57 -0.20 0.17 0.25 0.37 0.80 0.37 La/Th -0.29 -0.02 -0.27 0.12 -0.23 0.03 -0.29 -0.48 -0.34 -0.38 -0.13 P2O5 (%) 0.13 0.63 0.92 0.22 0.76 0.17 0.48 0.52 0.44 0.82 0.08 TiO2 (%) 0.21 0.29 0.90 0.15 0.98 0.36 0.72 0.63 0.32 0.53 0.07 Fe2O3 (%) 0.90 0.15 0.27 0.05 0.25 0.41 0.66 0.69 0.64 0.19 -0.02 ΣREE 0.13 0.79 0.90 0.29 0.82 0.27 0.53 0.44 0.40 0.63 0.05 Eu/Eu* -0.36 -0.40 -0.52 0.21 -0.41 0.13 -0.37 -0.54 -0.72 -0.66 -0.26 (La/Yb)n -0.40 -0.24 -0.23 -0.41 -0.25 -0.47 -0.56 -0.50 -0.42 -0.11 0.55 (La/Sm)n 0.18 -0.19 0.22 -0.27 0.26 -0.06 0.28 0.36 0.10 0.32 0.35 (Gd/Yb)n -0.40 -0.12 -0.29 -0.29 -0.31 -0.42 -0.59 -0.57 -0.39 -0.20 0.37 * Cypresshead Formation samples include reworked Cypresshead sediments collected at the SSD-1 locality.

Table 6-2. – (continued).

Zr/Sc La/Sc La/Th P2O5 (%) TiO2 (%) Fe2O3 (%) ΣREE Eu/Eu* (La/Yb)n (La/Sm)n (Gd/Yb)n Rb Sr Th U Zr Sc V Cr Co Th/Sc Th/U Zr/Sc 1.00 La/Sc 0.67 1.00 La/Th -0.39 0.12 1.00

P2O5 (%) 0.73 0.76 -0.22 1.00

TiO2 (%) 0.82 0.49 -0.20 0.69 1.00

Fe2O3 (%) 0.28 0.03 -0.40 0.11 0.26 1.00 ΣREE 0.70 0.80 0.05 0.86 0.75 0.13 1.00 Eu/Eu* -0.63 -0.53 0.45 -0.51 -0.34 -0.49 -0.40 1.00

(La/Yb)n -0.30 0.16 0.48 -0.18 -0.25 -0.46 -0.14 0.33 1.00

(La/Sm)n 0.39 0.04 -0.44 0.05 0.30 0.28 -0.07 -0.22 0.01 1.00 (Gd/Yb)n -0.39 0.18 0.65 -0.21 -0.32 -0.45 -0.06 0.30 0.83 -0.45 1.00 * Cypresshead Formation samples include reworked Cypresshead sediments collected at the SSD-1 locality.

Formation) were corrected to their stratigraphic age as shown in Table 6-3 to calculate for εNd(t).

For the Cypresshead Formation, stratigraphic ages of 3.3 Ma and 2.5 Ma were used for Florida and Georgia samples, respectively.

186

Table 6-3. Nd isotope data for Cypresshead Formation and comparison samples.

147 144 143 144 Error x 143 144 Sample ID Interval Sm* (ppm) Nd* (ppm) Sm/Nd Sm/ Nd Nd/ Nd εNd(0) Nd/ Nd εNd(t) TDM (Ga) 0 10-6 t Cypresshead Formation - FL (3.4–2.8 Ma: Fountain, this study) EPK36-J-12 25-27 2.2 13.7 0.16 0.0967 0.512291 12 -6.8 0.512288 -6.7 1.0 EPK36-J-12 35-40 4.3 24.2 0.18 0.1070 0.512249 6 -7.6 0.512247 -7.6 1.1 EPK36-J-12 46-48 2.7 18.3 0.15 0.0888 0.512266 6 -7.3 0.512264 -7.2 1.0 EPK36-J-12 50-53 2.2 13.3 0.17 0.0996 0.512293 10 -6.7 0.512290 -6.7 1.0 EPK36-J-12 59-62 7.3 36.0 0.20 0.1221 0.512356 4 -5.5 0.512353 -5.5 1.2 EPK30-V-6 22-24 10.3 55.5 0.19 0.1117 0.512215 4 -8.2 0.512213 -8.2 1.2 EPK30-V-6 30-35 60.2 306.0 0.20 0.1185 0.512246 5 -7.6 0.512243 -7.6 1.3 EPK30-V-6 48-53 19.5 93.4 0.21 0.1257 0.512259 5 -7.4 0.512256 -7.4 1.4 EPK30-V-6 58-63 3.4 19.0 0.18 0.1077 0.512372 5 -5.2 0.512369 -5.2 1.0 EPK30-V-6 68-73 4.3 21.8 0.20 0.1188 0.512379 4 -5.1 0.512376 -5.0 1.1 FRG-1 3 8.3 43.9 0.19 0.1138 0.512183 5 -8.9 0.512180 -8.8 1.3 FRG-1 6 6.4 36.3 0.18 0.1062 0.512252 6 -7.5 0.512250 -7.5 1.1 FRG-1 7 4.4 26.6 0.17 0.0996 0.512246 5 -7.7 0.512243 -7.6 1.1 FRG-1 10 23.1 139.0 0.17 0.1001 0.512260 5 -7.4 0.512258 -7.3 1.1 FRG-1 11 4.5 30.3 0.15 0.0894 0.512274 4 -7.1 0.512272 -7.0 1.0 FRG-1 13 1.9 14.2 0.13 0.0806 0.512288 5 -6.8 0.512286 -6.8 0.9 FRG-1 15 7.7 47.3 0.16 0.0980 0.512345 6 -5.7 0.512343 -5.7 0.9 FRL-1 4 4.6 30.9 0.15 0.0896 0.512309 6 -6.4 0.512307 -6.4 0.9 FRL-1 6 8.7 50.1 0.17 0.1046 0.512356 6 -5.5 0.512353 -5.5 1.0 FRL-1 7 6.0 37.4 0.16 0.0966 0.512332 5 -6.0 0.512330 -5.9 1.0 SSD-1 3 13.3 80.5 0.17 0.0995 0.512191 5 -8.7 0.512189 -8.7 1.1 SSD-1 7 11.1 58.5 0.19 0.1142 0.512338 3 -5.8 0.512336 -5.8 1.1 SSD-1 10 4.6 22.4 0.21 0.1236 0.512343 6 -5.8 0.512340 -5.7 1.2 SSJ-1 4 2.9 18.0 0.16 0.0970 0.512171 7 -9.1 0.512169 -9.1 1.1 SSJ-1 6 1.2 6.9 0.17 0.1047 0.512225 7 -8.1 0.512222 -8.0 1.2 SSJ-1 10 1.8 11.1 0.16 0.0976 0.512315 6 -6.3 0.512313 -6.3 1.0 TRF2214 60.0-62.5 2.4 14.2 0.17 0.1018 0.512140 21 -9.7 0.512138 -9.7 1.2 WEX164 18.0-26.0 10.9 58.9 0.19 0.1114 0.512238 14 -7.8 0.512235 -7.8 1.2 WEX366 9.0-10.0 4.0 23.9 0.17 0.1008 0.512160 11 -9.3 0.512158 -9.3 1.2 Cypresshead Formation - GA (3.15-2.3 Ma: Fountain, this study) J-1 4 4.5 32.3 0.14 0.0839 0.512253 5 -7.5 0.512252 -7.5 1.0 J-1 6 5.9 41.5 0.14 0.0856 0.512280 5 -7.0 0.512279 -6.9 0.9 L-1 5 2.1 17.0 0.12 0.0744 0.512299 5 -6.6 0.512298 -6.6 0.8 L-1 6 4.4 36.1 0.12 0.0734 0.512275 4 -7.1 0.512273 -7.0 0.9 B-1 3 7.6 46.9 0.16 0.0976 0.512157 19 -9.4 0.512155 -9.4 1.2 B-1 5 19.3 88.2 0.22 0.1318 0.512207 4 -8.4 0.512205 -8.4 1.5 Hawthorn Group, Coosawhatchie Formation - FL/GA (13.5-15 Ma: Huddlestun, 1988) MCB109 15.0-20.0 6.3 36.0 0.18 0.1054 0.512177 16 -9.0 0.512167 -8.8 1.2 J-1 BC 19.2 95.6 0.20 0.1209 0.512347 4 -5.7 0.512336 -5.5 1.2 Huber Formation - GA (40-50 Ma: Al-Sanabani, 1991; Dombrowski, 1992; Pickering et al., 1997) a KGa-2 --- 4.3 25.7 0.17 0.1007 0.512311 5 -6.4 0.512281 -5.8 1.0 ECCI-CBa --- 9.4 41.0 0.23 0.1380 0.512493 5 -2.8 0.512453 -2.5 1.1 Buffalo Creek Formation - GA (70-80 Ma: Tschudy and Patterson, 1975; Nystrom et al., 1986) KGa-1a --- 11.7 41.7 0.28 0.1689 0.512282 4 -6.9 0.512199 -6.7 2.4 ECCI-BC --- 24.0 131.0 0.18 0.1103 0.512216 3 -8.2 0.512162 -7.4 1.2 TKC-EA --- 4.3 12.6 0.34 0.2055 0.512252 5 -7.5 0.512151 -7.6 7.8 DBK-B93 --- 16.6 72.1 0.23 0.1386 0.512253 3 -7.5 0.512185 -6.9 1.6 Note: Initial Nd isotopic values have been normalized to 146Nd/144Nd = 0.7219. Errors on 143Nd/144Nd measurements 143 144 are 2σ. Measured εNd values calculated assuming a present-day Nd/ Nd(CHUR) = 0.512638 (Jacobsen and Wasserberg, 1980). * Analysis by LiBO2 fusion ICP-MS. a Average of duplicate analyses (KGa-1 and KGa-2, n=2; ECCI-CB, n=3).

Calculated εNd(t) values for both Florida and Georgia Cypresshead Formation samples

exhibit moderate variation, ranging between εNd(3.3 Ma) = -5.0 and -9.7 and εNd(2.5 Ma) = -6.6

and -9.4 respectively. Corresponding εNd(t) average values are -7.1 and -7.6, and show

consistency between the sample groups. The variation noted for εNd(t) values in Florida samples

187

appears to exhibit a vertical trend, with near surface εNd(t) more negative than samples collected from increasing depth. Sm/Nd ratios range from 0.13 to 0.21 and 0.12 to 0.22 for Florida and

Georgia samples respectively. Cypresshead Formation samples plotted on a 143Nd/144Nd versus

147Sm/144Nd isochron diagram (Fig. 6-3) exhibit no linear trends suggestive of diagenetic

resetting. Additionally, the lack of variation noted for εNd(t) when plotted against Nd

concentration (Fig. 6-4) indicates there is no evidence for a multiple component Nd isotopic

signature in Cypresshead samples. Rather, Cypresshead samples appear to originate from a

single, well mixed source. Nd model ages based on εNd(t) results exhibit a high degree of

consistency (Fig. 6-5) for Florida and Georgia samples as well. TDM values for Florida and

Georgia Cypresshead samples range between 1.4-0.9 Ga and 1.5-0.8 Ga respectively, with both

possessing an average TDM value of 1.1 Ga.

Comparison clay samples from the Hawthorn Group in Florida and Georgia, and samples

of kaolin from the Huber and Buffalo Creek formations in Georgia possess similar Nd isotopic

characteristics to Cypresshead samples. Hawthorn and Huber εNd(t) data corrected for

stratigraphic age exhibit moderate variation as noted with Cypresshead samples (Table 6-3). For the two Hawthorn samples evaluated, εNd(14.25 Ma) values area -5.5 and -8.8, and for the two

Huber samples, εNd(45 Ma) values are -2.5 and -5.8. As for the four Buffalo Creek samples

evaluated, εNd(75 Ma) values are less varied, ranging from -6.7 to -7.4. However, two of the

Buffalo Creek samples exhibit anomalous Sm/Nd ratios of 0.28 and 0.34. Correspondingly, these

two samples also exhibit spurious TDM ages of 2.4 Ga and 7.8 Ga. The remainder of the

comparison samples exhibit TDM ages consistent with results from the Cypresshead, with the remaining TDM ages ranging between 1.0 Ga and 1.6 Ga.

188

Figure 6-3. 143Nd/144Nd versus 147Sm/144Nd isochron diagram for Cypresshead Formation samples.

Figure 6-4. εNd(t) versus Nd concentration scatterplot for Cypresshead Formation samples.

189

12 Buffalo Creek Fm.* Huber Fm. 10 Hawthorn Gr. Cypresshead - GA 8 Cypresshead - FL

Frequency 4

0 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 TDM(Ga)

Figure 6-5. Histogram illustrating the distribution of Nd model age (TDM) results for both Cypresshead Formation clay samples and comparison formations (* spurious TDM values associated with samples KGa-1 and TKC-EA not included with Buffalo Creek Formation results).

Discussion

Trace elements are the primary geochemical tools used for paleoenvironmental interpretation, yet many are mobile under variable redox conditions accompanying diagenesis and weathering. Thus, trace element and Nd isotopic results for Cypresshead Formation clays are interpreted in light of the apparent mobility of many trace elements (including REEs) in weathering profiles developed on Cypresshead Formation sediments and materials similar to proposed sources (Cullers, 1988; Condie et al., 1995; Lev et al., 1998). Although many trace element systems previously assumed to be sensitive to source composition and immobile during weathering have proven to be unreliable for provenance studies (Condie et al., 1995), several do offer a means of interpreting kaolinite provenance or weathering and/or diagenetic pathways.

190

Trace Element Mobility and Enrichment

HFSEs such as Zr, Th, and U are preferentially partitioned into melts during crystallization

(Feng and Kerrich, 1990). As such, these elements are enriched in felsic rather than mafic

sources, and are thought to reflect provenance compositions as a consequence of their relatively

immobile character (Taylor and McLennan, 1985). For most upper crustal rocks, Th/U is

typically 3.5-4.0 (McLennan et al., 1993), with weathering and sedimentary recycling under

oxidizing conditions responsible for mobilization and loss of U (U6+). As such, weathering

elevates Th/U ratios (Fig. 6-6A), although the addition of heavy minerals (zircon) or secondary phosphates can complicate this trend. In the case of Florida Cypresshead clays, a significant proportion of the samples analyzed possess Th/U ratios well below that of average upper crust

(Fig. 6-6A). This appears most likely a consequence of redox-driven U enrichment dictated by pore water organics associated with basal Cypresshead sediments. As seen in Table 6-1, low

Th/U ratios are concentrated at the base of most sample locations where pore water organics and corresponding reducing conditions would be most favored. Additionally, there appears to be no correlation between Th/U values and either Zr or P2O5 concentrations, ruling out a role for U

enrichment associated with either detrital zircons or secondary crandallite-florencite. Georgia

Cypresshead samples correspond to the weathering trend in Figure 6-6A as would be predicted

given the highly oxidized appearance of these sample localities.

Since Zr is preferentially concentrated in the detrital heavy mineral zircon, which is known

to occur in Cypresshead sediments (Chapter 4), it is possible to evaluate the role of heavy

mineral concentration during sedimentary sorting of the unit (McLennan et al., 1993). For this

purpose, the Zr/Sc ratio is plotted against Th/Sc (Fig. 6-6B). Zr/Sc is an index for zircon

enrichment given the concentration of Zr in zircon and the corresponding behavior of Sc, which

191

Figure 6-6. Trace element plots of Cypresshead Formation clays illustrating evidence of weathering, provenance, and sediment recycling processes. A) Th/U versus Th (ppm), and B) Th/Sc versus Zr/Sc.

is not enriched but tends to preserve the provenance signature of its source. In contrast, the Th/Sc ratio is thought to be particularly sensitive to average provenance due to the incompatibility of

Th and the relative compatibility of Sc, and the fact that both elements tend to be transferred

192

quantitatively into terrigenous sediments during sedimentation processes (Taylor and McLennan,

1985; McLennan et al., 1990), although enrichments can also be indicative of a relatively severe weathering regime (McLennan et al., 1993). The trend illustrated in Figure 6-6B supports a mix in provenance compositions supplying Cypresshead sediments based on the sympathetic relationship between Zr/Sc and Th/Sc at low Zr/Sc values. For higher Zr/Sc values above 10,

Th/Sc ratios increase at a much lower rate consistent with zircon enrichment. Such a conclusion is in agreement with X-ray diffraction and REE data presented in Chapter 4 and Chapter 5 which indicate near surface enrichment of many Cypresshead sample locations by detrital heavy minerals.

Comparison to Georgia-South Carolina Kaolin Provenance

Based on the assumption of relative immobility of Co, La, Sc, and Th, Dombrowski (1992;

1993) focused on the use of these elements as provenance indicators for the commercial kaolin deposits located in the Georgia-South Carolina kaolin district. REE (La) and Th abundances are believed to be higher in felsic source rocks and their weathering products, whereas Sc and Co are more concentrated in mafic sources. As a result, Dombrowski was able to discriminate between felsic and mafic sources based on preferential enrichment. For kaolins derived from granite and gneiss source rocks adjacent to the kaolin district, Th (25.5–86.6 ppm) and La (46.0–256 ppm) concentrations we found to be high relative to Co (7.5–25.2 ppm) and Sc (13.8–28.3 ppm). In contrast, kaolins derived from metavolcanic sources (e.g., Little River Group) exhibited the opposite characteristics, with Th (5.6–20.7 ppm) and La (17.5–122 ppm) possessing relatively low concentrations as compared to Co (18.6–55.5 ppm) and Sc (13.8–48.9 ppm). This allowed

Dombrowski to propose that felsic lithologies (granites/gneiss) are the principle source rocks for

Cretaceous “soft” kaolins in both central Georgia and South Carolina, with eastern Georgia and

South Carolina Tertiary “hard” kaolins originating from a mixed metavolcanic/felsic (70/30)

193

source, and middle Georgia Tertiary kaolins derived almost exclusively from metavolcanic

sources.

Comparing the results of Dombrowski (1992; 1993) to those in this study, it is apparent

that this approach is not suitable as a basis for provenance determination father south on the

Florida peninsula. Th, Sc, and Co results for Cypresshead samples illustrate this problem (Fig. 6-

7). For Cypresshead samples, the fields corresponding to Sc and Th values for felsic (”soft”

kaolin) and metavolcanic (“hard” kaolin) sources overlap convincingly with the results from this

study (Fig. 6-7A). However, the range in values for Sc noted with the Dombrowski (1992; 1993)

study far exceeds values noted for either Florida (max = 26 ppm) or Georgia (max = 36 ppm),

even if the bulk of Cypresshead clays were derived from metavolcanic sources. Complicating

this interpretation is the occurrence of both heavy minerals (zircon and rutile) and crandallite-

florencite in Cypresshead sediments. Both phases correlate with Th, potentially impacting

relative concentration of the element beyond what might be indicative of provenance (e.g., SSD-

1-1). The plot of Th versus Co shows a far worse fit to Cypresshead data (Fig. 6-7B). Whereas

Th concentrations do fall within the range of values reported by Dombrowski (1992; 1993) for

both felsic and metavolcanic sources, Co values for Cypresshead samples are extremely

depleted. In fact, if a component of Cypresshead clays originated from a metavolcanic or reworked “hard” kaolin source, then Co has been significantly mobilized and removed from the resultant deposit. As such, Co appears to be an unsatisfactory indicator of Cypresshead provenance. La exhibits the same characteristics as Co, and is highly depleted in Cypresshead samples relative to concentrations reported by Dombrowski (1992; 1993).

Cypresshead Provenance

Of the trace elements reported in the Dombrowski studies, Th and Sc appear to be the most resistant to depletion relative to concentrations consistent with potential provenance sources.

194

Figure 6-7. Comparison of Cypresshead Formation trace element concentrations to the results of Dombrowski (1992; 1993). A) Th versus Sc plot illustrating the fields for metavolcanic and felsic sources relative to Cypresshead results (arrow indicates the field for metavolcanics extends beyond the concentration range shown on the plot), B) Th versus Co plot illustrating the fields for metavolcanic and felsic sources relative to Cypresshead results (arrows indicates the fields for both metavolcanics and felsic sources extend beyond the concentration range shown on the plot).

195

Based on this observation, and the coherent mixing trend illustrated in Figure 6-6B for Th/Sc,

this ratio appears to be a robust indicator of provenance composition. As such, Th/Sc was plotted

against εNd(t) using the model of McLennan et al. (1990) in order to clarify the bulk chemical

composition of Cypresshead clays at the time of deposition (Figure 6-8). Based on this plot,

Cypresshead provenance appears consistent with a dominantly felsic composition originating from an upper crust source, although some of the high Th/Sc values are likely due to heavy

mineral concentration rather than provenance bulk chemistry (Fig. 6-6B). This conclusion differs

somewhat from the model suggested by Figure 6-7A, and may be due to a slight depletion of

both Th and Sc dictating the high degree of correlation to metavolcanics in Figure 6-7A.

In comparing granites and metavolcanics associated with the Carolina terrane to

Cypresshead Nd isotope results, initial 143Nd/144Nd ratios of both granites and Carolina terrane

metavolcanics reported in the literature are considered to be robust (Kozuch, 1994; Samson et

al., 1995a; 1995b; Fullagar et al., 1997). Additionally, even large uncertainties in granite ages

have little effect on the calculated εNd(t) because of the low Sm/Nd ratio of the granites and the

147 106 Ga half-life of Sm. The εNd(t) values of Alleghanian granites range from -8.2 to +3.0, with

more restricted ranges for individual terranes (Samson et al., 1995a; Fullagar et al., 1997).

Carolina terrane granites range from -6.7 to +1.9, Eastern slate belt granites from -2.7 to +2.4,

Kiokee belt granites from -2.3 to +2.0, Raleigh belt granites from -4.3 to +3.0, and Inner

Piedmont granites from -8.2 to -3.4. The bulk of the εNd(t) values for the Alleghanian granites

range between -3 and +3, indicating that a significant proportion of their source material is not

old, evolved crust, but rather likely from partial melting of depleted mantle or wholesale melting

of terrane rocks (Samson et al., 1995a). The Nd isotopic evolution trend for Alleghanian granites

is plotted in Figure 6-9 for comparison to both Cypresshead and comparison samples.

196

Figure 6-8. Plot of εNd(t) versus Th/Sc for Cypresshead Formation samples based on the model of McLennan et al. (1990; 1993).

Figure 6-9. Plot of εNd(t) versus stratigraphic age for the Cypresshead Formation compared to the Nd isotopic evolution of potential sources (data for Grenville crust, Alleghanian granites, and the Carolina terrane are from Kozuch, 1994; Samson et al., 1995a; 1995b; Fullagar et al., 1997).

197

Carolina terrane rocks have εNd(t) values (most from -2 to +3) indistinguishable from those

of most Alleghanian granites which intrude the Carolina terrane (most from -2 to +2), suggesting

them as a likely source for the granites (Kozuch, 1994; Samson et al., 1995b; Fullagar et al.,

1997). Supporting this idea are the Nd isotopic compositions of Eastern slate belt volcanic rocks

which, although lower than Carolina terrane values, are similar to granites intruding the terrane

(Samson et al., 1995a). Some granites possess εNd(t) values suggesting a more evolved source,

possibly Grenville basement. The chemical and isotopic data for granites intruding the Kiokee

belt, Eastern slate belt, and Carolina terrane are similar, suggesting that the terranes are similar as

well (Samson et al., 1995a). As with the available Nd isotopic data for Alleghanian granites, the

Nd isotopic evolution trend for Carolina terrane rocks is also plotted in Figure 6-9.

Based on a comparison of Cypresshead, Hawthorn, Huber, and Buffalo Creek εNd(t) values to the Nd isotopic evolution trends of Alleghanian granites, Carolina terrane rocks, and Grenville crust, Cypresshead samples appear to originate from sources intermediate between those associated with Cretaceous “soft” kaolins and Tertiary “hard” kaolins (Fig. 6-9). Samples from the Cretaceous Buffalo Creek Formation appear consistent with the conclusions of Dombrowski

(1992; 1993), likely originating from Alleghanian granite sources, but apparently from granites having a significant portion of their source material originating from old, evolved Grenville crust. Tertiary “hard” kaolin samples from the Huber Formation are more consistent with a mixed sourcing from both Alleghanian and Carolina terrane lithologies (Fig. 6-9). Again, this conclusion is consistent with the work of Dombrowski (1992; 1993).

Cypresshead and Hawthorn clays exhibit similar provenance characteristics, and are apparently a mixture of Carolina terrane and Alleghanian granite sources (Fig. 6-9).

Additionally, both units appear to possess a significant affinity to an evolved crustal source, an

198

observation consistent with the distribution of TDM values similar in age to Grenville crust (Fig.

6-5). This conclusion is further supported by the low εNd(t) values associated with Carolina terrane (min = -6.7) and Inner Piedmont (min = -8.2) granites, the most likely source for both

Buffalo Creek clays and a felsic component of Cypresshead clays. The similarity in provenance between Cypresshead Formation and Hawthorn Group clays is further evidence supporting the detrital sourcing of some Cypresshead clays via reworking of Hawthorn sediments as discussed in Chapter 5. Furthermore, the similarity in provenance is also a predictable outcome of the significant sediment mixing expected for deposits located distally from their source.

Conclusions

The results outlined in this chapter combined trace element and Nd isotopic analyses to constrain the provenance and broader geochemical characteristics of the Cypresshead Formation and associated units. Among the results are the following:

• Trace element concentrations of Florida Cypresshead clays relative to average continental crust are relatively low, except for P, Th, and U, with Georgia Cypresshead clays slightly more enriched than Florida samples.

• Significant geochemical correlations exist for P2O5 (Th, Zr, and Sr), TiO2 (Th, Zr, V, and Cr), and Fe2O3 (Rb, V, Cr, and Co), the most significant of which are correlated to accessory phase (zircon and crandallite-florencite) enrichment.

• Cypresshead samples exhibit moderate variation in εNd(t) values, ranging between εNd(3.3 Ma) = -5.0 and -9.7 and εNd(2.5 Ma) = -6.6 and -9.4 for Florida and Georgia samples respectively

• Nd model ages (TDM) for Florida and Georgia Cypresshead samples range between 1.4-0.9 Ga and 1.5-0.8 Ga respectively, with both possessing an average TDM value of 1.1 Ga consistent with the age of Grenville crust.

• A significant proportion of Florida Cypresshead clays possess Th/U ratios well below that of average upper crust, and appear to have undergone redox-driven U enrichment dictated by pore water organics in basal Cypresshead sediments.

• Of the trace elements reported in the Dombrowski (1992; 1993) studies, Th and Sc appear to be the most resistant to depletion, with the Th/Sc ratio a robust indicator of provenance composition.

199

• Based on Nd isotopic results, the Buffalo Creek Formation appears to have originated from Alleghanian granite sources having a significant portion of their source material originating from old, evolved Grenville crust, while the Huber Formation is consistent with a mixed sourcing from both Alleghanian and Carolina terrane lithologies.

• Cypresshead Formation (and Hawthorn Group) samples appear to originate from sources intermediate between those associated with Cretaceous “soft” kaolins and Tertiary “hard” kaolins, possessing provenance characteristics consistent with a mixture of Carolina terrane and Alleghanian granite sources.

200

CHAPTER 7 SUMMARY AND CONCLUSIONS

Cypresshead Formation Stratigraphy

Focusing on the stratigraphy and sedimentology of the Cypresshead Formation, the results

outlined in Chapter 4 highlight significant observations which clarify the nature, timing and

significance of siliciclastic deposition impacting the Florida Platform during the Late Miocene

through Pliocene. Among the results are the following:

• Cypresshead sediments were deposited in a nearshore marine environment, most likely in a strand plain setting, as two distinct progradational shoreface-shelf parasequences.

• Cypresshead facies define coarsening-upward sequences consistent with a wave-dominated environment in north-central Florida and a mixed energy environment in southeastern Georgia.

• Deposition of the Cypresshead took place in response to sea-level falls at 3.3 Ma and 2.5 Ma as a consequence of the interplay of sea-level, sediment supply and accommodation.

• Deposition in Florida at 3.4–2.8 Ma with reworking at 2.8–1.8 Ma.

• Deposition in Georgia at 3.15–2.3 Ma with reworking at 2.3–1.8 Ma.

201

Cypresshead Formation Mineralogy

Weathering processes are known have a profound impact on original sediments which host

commercial kaolin deposits. Focusing on the clay mineralogy and microtexture of Cypresshead

sediments, the results outlined in Chapter 5 highlight evidence for the origin and weathering

history of Cypresshead kaolinites in order to define processes impacting their mineralogy and

industrial properties, and answers the primary verses secondary origin issue associated with these

clays. Among the results are the following:

• Near surface recrystallized kaolinite formed under oxic, vadose or mixed vadose/saturated conditions are disordered and possess microtextural characteristics consistent with a pedogenic origin.

202

• Trace crandallite-florencite minerals likely originated from the decomposition of post- depositional pore water organics coupled with detrital mineral dissolution.

Cypresshead Formation Provenance

The results outlined in Chapter 6 combined trace element and Nd isotopic analyses to

constrain the provenance and broader geochemical characteristics of the Cypresshead Formation and associated units. Among the results are the following:

• Cypresshead samples exhibit moderate variation in εNd(t) values, ranging between εNd(3.3 Ma) = -5.0 and -9.7 and εNd(2.5 Ma) = -6.6 and -9.4 for Florida and Georgia samples respectively

• Of the trace elements reported in the Dombrowski (1992; 1993) studies, Th and Sc appear to be the most resistant to depletion, with the Th/Sc ratio a robust indicator of provenance composition.

203

APPENDIX A MINE SITE MAPS

204

Figure A-1. Site map for the Edgar Minerals EPK Mine (DOQQ Source: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm. Last accessed October, 2007).

Figure A-2. Site map for the VMC Goldhead Sand Mine (DOQQ Source: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm. Last accessed October, 2007).

205

Figure A-3. Site map for the VMC Grandin Sand Mine (DOQQ Source: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm. Last accessed October, 2007).

Figure A-4. Site map for the CEMEX Davenport Sand Mine (DOQQ Source: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm. Last accessed October, 2007).

206

Figure A-5. Site map for the CEMEX Joshua Sand Mine (DOQQ Source: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm. Last accessed October, 2007).

207

APPENDIX B ANALYTICAL PROCEDURES FOR ICP-AES, ICP-MS, AND MC-ICPMS

208

METHOD A: ICP-AES MULTI-ACID ANALYSIS (SGS/XRAL: ICP40)

SUMMARY: This method involves the determination of 40 major, minor, and trace elements in geologic materials by inductively coupled plasma-atomic emission spectrometry (ICP-AES). The sample is decomposed using a mixture of hydrochloric, nitric, perchloric, and hydrofluoric acids at low temperature (Crock et al., 1983). The digested sample is aspirated into the ICP-AES discharge where the elemental emission signal is measured simultaneously for the 40 elements. Calibration is performed by standardizing with digested rock reference materials (BCR-2) and a series of multi-element solution standards (Lichte et al., 1987).

APPARATUS:

1) Sample preparation • 35 ml Teflon tube with screw cap. • Temperature controlled aluminum block machined to accept 1 inch diameter Teflon vessels with screw caps. 2) Sample analysis • Perkin-Elmer Optima 300 ICP-AES. • Perkin-Elmer Cross-flow nebulizer. • Perkin-Elmer peristaltic pump. • CETAC autosampler.

REAGEANTS:

• Hydrochloric acid (HCl) – 36-38%. • Nitric acid (HNO3) – 69-71%. • Perchloric acid (HClO4) – 62-70%. • Hydrofluoric acid (HF) – 48-51%. • Lu internal standard – 500 ppm.

PROCEDURE:

1) Weigh 0.200 g (± 2 mg) sample into Teflon dish. 2) Add 100 ml of 500 μg/ml Lu internal standard to each dish with repeating pipet. 3) Rinse side walls of Teflon dish with a minimum amount of deionized (DI) water (~ 2 ml). 4) In the fume hood, slowly add 3 ml HCl and allow any reaction to subside. 5) Add 2 ml HNO3, 1 ml HClO4, and 2 ml HF. 6) Place sample solution vessel on hot plate with aluminum heat block at a controlled temperature of 110°C in a perchloric acid fume hood. 7) Evaporate sample solution to hard dryness on hot plate (usually overnight). 8) Remove from hot plate, cool to touch and add 1 ml HClO4 and 2-3 ml DI water. 9) Return to hot plate and evaporate to hard dryness. The temperature of the hot plate is increased to 160°C. This step usually takes a few hours. 10) Remove dried sample from hot plate and cool. 11) Add 1 ml aqua regia with repeating pipet and let react for 15 minutes.

209

12) Add 9 ml 1% HNO3 and heat for a few minutes. 13) Cool, transfer solution into labeled disposable polypropylene test tube and dilute to 10 ml. 14) Analyze sample solution by ICP-AES.

ANALYTICAL PERFORMANCE: Data is deemed acceptable if recovery is ± 15% at five times the Lower Limit of Determination (LOD), and the calculated Relative Standard Deviation (RSD) of duplicate samples is ≤15%. CRM SO-3 was analyzed as a standard to evaluate accuracy.

REPORTING LIMITS:

Element Concentration Range Element Concentration Range Aluminum, Al 0.005 – 50% Gallium, Ga 4 – 50,000 ppm Calcium, Ca 0.005 – 50% Holmium, Ho 4 – 5,000 ppm Iron, Fe 0.02 – 25% Lanthanum, La 2 – 50,000 ppm Potassium, K 0.01 – 50% Lithium, Li 2 – 50,000 ppm Magnesium, Mg 0.005 – 5% Manganese, Mn 4 – 50,000 ppm Sodium, Na 0.005 – 50% Molybdenum, 2 – 50,000 ppm Mo Phosphorous, P 0.005 – 50% Niobium, Nb 4 – 50,000 ppm Titanium, Ti 0.005 – 25% Neodymium, Nd 9 – 50,000 ppm Silver, Ag 2 – 10,000 ppm Nickel, Ni 3 – 50,000 ppm Arsenic, As 10 – 50,000 ppm Lead, Pb 4 – 50,000 ppm Gold, Au 8 – 50,000 ppm Scandium, Sc 2 – 50,000 ppm Barium, Ba 1 – 35,000 ppm Tin, Sn 5 – 50,000 ppm Beryllium, Be 1 – 5,000 ppm Strontium, Sr 2 – 15,000 ppm Bismuth, Bi 10 – 50,000 ppm Tantalum, Ta 40 – 50,000 ppm Cadmium, Cd 2 – 25,000 ppm Thorium, Th 6 – 50,000 ppm Cerium, Ce 5 – 50,000 ppm Uranium, U 100 – 100,000 ppm Cobalt, Co 2 – 25,000 ppm Vanadium, V 2 – 30,000 ppm Chromium, Cr 2 – 50,000 ppm Yttrium, Y 2 – 25,000 ppm Copper, Cu 2 – 15,000 ppm Ytterbium, Yb 1 – 5,000 ppm Europium, Eu 2 – 5,000 ppm Zinc, Zn 2 – 15,000 ppm

REFERENCES:

Crock, J.G., Lichte, F.E., and Briggs, P.H., 1983, Determination of elements in National Bureau of Standard’s geological reference materials SRM278 obsidian and SRM688 basalt by inductively coupled argon plasma-atomic emission spectrometry: Geostandards Newsletter, v. 7, p. 335–340.

Lichte, F.E., Golightly, D.W., and Lamothe, P.J., 1987, Inductively coupled plasma-atomic emission spectrometry, in Baedecker, P.A., ed., Methods for Geochemical Analysis, USGS Bulletin, Report: B1770, USGS, Reston, VA, p. B1–B10.

210

METHOD B: ICP-MS LiBO2 FUSION (SGS/XRAL: MS95)

SUMMARY: This method involves the determination of 37 trace and rare earth elements in geologic materials by inductively coupled plasma-mass spectrometry (ICP-MS). The sample is fused with LiBO2 at 950ºC for twenty minutes, and then the cake is dissolved in dilute nitric acid prior to aspiration into the ICP-MS.

APPARATUS:

1) Sample preparation • Graphite crucible. • Muffle furnace. • Teflon vials. 2) Sample analysis • Perkin-Elmer Elan 6100 ICP-MS, VG ThermoElemental PlasmaQuad 2 ICP-MS, or VG ThermoElemental PlasmaQuad 3 ICP-MS.

REAGEANTS:

• Lithium borate (LiBO2) • Nitric acid (HNO3) – 10%. • Tartaric acid – 2%. • Lu internal standard – 1000 ppm. • Re and Rh internal standards – 50 ppb and 10 ppb, respectively. • Calibration standards diluted with 2% HNO3 and 4 ml of LiBO2 solution (2.8 g flux in 10% HNO3) into a 200 ml volumetric flask o #1 blank (all elements at 0 ppb) o #2 10.0 ppb all elements except Ag (1 ppb) o #3 25.0 ppb all elements except Ag (2.5 ppb) o #4 50.0 ppb all elements except Ag (5 ppb)

PROCEDURE:

1) Weigh 0.1 g sample into a graphite crucible. 2) Add 0.70 g of LiBO2 and mix. 3) Put crucibles in muffle furnace (preheated at 950ºC) for twenty (20) minutes. 4) Take out the crucibles and pour into vial containing 50 ml of 10% HNO3, 2% tartaric acid. 5) Shake for ten (10) minutes. 6) Add 0.25 ml of 1000 ppm Lu internal standard and mix. 7) Transfer 15 ml into a plastic test tube. 8) Transfer the rack to the ICP room. 9) Pipette 0.2 ml sample digest + 0.8 ml DDW into centrifuge tube. 10) Add 9 ml internal standard solution (Re + Rh) to correct for matrix effects. 11) Cover sample with parafilm and shake. 12) Load the sample(s) into a rack and set into the autosampler station.

211

13) Analyze sample solution by ICP-MS.

ANALYTICAL PERFORMANCE: CRM SO-3 was weighed, fused, and analyzed as a standard to evaluate accuracy. Calibration standard #2 is run after every 16 samples as a QC check.

REPORTING LIMITS:

Element Detection Limits Element Detection Limits Barium, Ba 0.5 ppm – 1% Praseodymium, Pr 0.05 ppm – 0.1% Cerium, Ce 0.1 ppm – 1% Rubidium, Rb 0.2 ppm – 1% Cesium, Cs 0.1 ppm – 1% Samarium, Sm 0.1 ppm – 0.1% Cobalt, Co 0.5 ppm – 1% Silver, Ag 1 ppm – 0.1% Copper, Cu 5 ppm – 1% Strontium, Sr 0.1 ppm – 1% Dysprosium, Dy 0.05 ppm – 0.1% Tantalum, Ta 0.5 ppm – 1% Erbium, Er 0.05 ppm – 0.1% Terbium, Tb 0.05 ppm – 0.1% Europium, Eu 0.05 ppm – 0.1% Thallium, Tl 0.5 ppm – 0.1% Gadolinium, Gd 0.05 ppm – 0.1% Thorium, Th 0.1 ppm – 0.1% Gallium, Ga 1 ppm – 0.1% Thulium, Tm 0.05 ppm – 0.1% Hafnium, Hf 1 ppm – 1% Tin, Sn 1 ppm – 1% Holmium, Ho 0.05 ppm – 0.1% Tungsten, W 1 ppm – 1% Lanthanum, La 0.1 ppm – 1% Uranium, U 0.05 ppm – 0.1% Lead, Pb 5 ppm – 1% Vanadium, V 5 ppm – 1% Lutetium, Lu 0.05 ppm – 0.1% Ytterbium, Yb 0.1 ppm – 0.1% Molybdenum, Mo 2 ppm – 1% Yttrium, Y 0.5 ppm – 0.1% Neodymium, Nd 0.1 ppm – 1% Zinc, Zn 5 ppm – 1% Nickel, Ni 5 ppm – 1% Zirconium, Zr 0.5 ppm – 1% Niobium, Nb 1 ppm – 1%

212

METHOD C: Nd ISOTOPIC ANALYSIS: SAMPLE PREPARATION

SUMMARY: This method involves procedures followed in the dissolution of silicate whole-rock powders and Nd collection via column chemistry.

PROCEDURE:

1) Clean tall Savillex Teflon hex-cap beakers: Put a little 6N trace metal (TM) grade HCl in the bottom of the beakers, cap them, and leave them on a warm hot plate for at least one hour. Rinse 3 times with 4× H2O. Dry by opening them up and leaving them on the hot plate for up to an hour. The caps must be left concave side up in this situation.

2) Add sample: Put the beaker on the balance in the outer clean lab room. Allow to settle, then push the tare button. Remove the beaker from the balance and put it on a clean, lint free wiper on the balance table. (No rock powders are allowed in the laminar flow hoods.) Open it up, putting the cap in a safe clean place. Using a clean spatula, add a small amount of rock powder to the beaker. Cap the beaker, and put it in the balance to check the weight. Keep adding rock powder and checking the weight until you get approximately 50 mg (0.050 g). Record the weight on the weigh sheet.

3) Begin dissolving the sample: Take the capped beakers over to the exhaust hood. Open them up and add 1-2 drops of Optima HNO3 from the dropper bottle. Then add 3 ml of 1x TM HF (once-distilled trace metal grade HF).

4) Heat the sample: Put the cap on the beaker and tighten it with the green plastic “wrenches”. Do not overtighten or you will strip the threads. Put the beaker in the small oven in the anteroom of the clean lab. Turn the temperature to 100˚C and leave it there for 2-3 days.

5) Convert to chloride salts: Remove the sample from the oven and let it cool. Put it in the exhaust hood. Taking all safety precautions, remove the cap with the green “wrenches”. Dry the sample by putting the beaker on the hotplate at 250˚F (setting 3). The cap should be put convex side down on a piece of parafilm. It will take a day or so to dry. Then, add 2 ml of 6N TM grade HCl. Replace the cap and tighten it down. Put the sample back into the oven overnight. Then remove it, let it cool, and dry it on the hotplate again.

6) Cation columns – separation of bulk rare earth elements (REEs): This step separates the REEs from the rest of the rock using chromatographic methods on Dowex 50 X12 cation exchange resin (see below). Before the sample is loaded on the columns, it must be centrifuged to remove any solid material. After dissolving the sample in 500 μl of 3.5N HCl, pour it into a clean centrifuge tube. Centrifuge the sample for 2 minutes, then pipet out just the clean solution off the top and load it. Once the REE are separated, they should be dried down on the hotplate at 200-250˚F.

1 – Equilibrate 5 ml 3.5N HCl 2 – Load sample dissolved in 250 μl 3.5N HCl

213

3 – Wash 500 μl 3.5N HCl 4 – Wash 500 μl 3.5N HCl 5 – Wash 500 μl 3.5N HCl 6 – Wash 23 ml 3.5N HCl 7 – Wash 4 ml 6N HCl 8 – Collect 10 ml 6N HCl (REE) 9 – Clean by washing with approximately 50 ml 6N HCl 10 – Re-equilibrate with 10 ml 4× H2O 11 – Store columns in 4× H2O

7) REE columns – separation of Nd: This step separates Nd from the rest of the REE using HCl elution on quartz columns packed with Teflon beads coated with bisethylhexyl phosphoric acid (after Richard et al., 1976) (see below). Normally, the sample will not have to be centrifuged before loading. Once the Nd is separated, it should be dried on the hotplate.

1 – Equilibrate 5 ml 0.18N HCl 2 – Equilibrate 5 ml 0.18N HCl 3 – Load sample dissolved in 200 μl 0.18N HCl 4 – Wash 200 μl 0.18N HCl 5 – Wash 200 μl 0.18N HCl 6 – Wash 200 μl 0.18N HCl 7 – Wash 8 ml 0.18N HCl 8 – Collect 5 ml 0.18N HCl (Nd) 9 – Clean by washing with approximately 50 ml 6N HCl 10 – Re-equilibrate with 10 ml ~0.2N HCl 11 – Store columns in ~0.2N HCl

REFERENCES:

Richard, P., Shimazu, N., and Allegre, C.J., 1976, 143Nd/144Nd, a natural tracer: An application to oceanic basalt: Earth and Planetary Science Letters, v. 31, p. 269–278.

214

APPENDIX C STRATIGRAPHIC SECTIONS

215

Figure C-1. Stratigraphic section for Grandin Sand Mine section FRG-1.

216

Figure C-2. Stratigraphic section for Grandin Sand Mine section FRG-2.

217

Figure C-3. Stratigraphic section for Goldhead Sand Mine section FRL-1.

218

Figure C-4. Stratigraphic section for Joshua Sand Mine core SSJ-1.

219

Figure C-5. Stratigraphic section for Davenport Sand Mine core SSD-1.

220

Figure C-6. Stratigraphic section for Jesup section J-1.

Figure C-7. Stratigraphic section for Birds section B-1.

221

Figure C-8. Stratigraphic section for Linden Bluff section L-1.

222

APPENDIX D GRAIN-SIZE DATA (HYDROMETER AND SIEVE)

223

Table D-1. Grain-size distributions and moment statistics for the Grandin and Goldhead sand mines in north-central Florida. Sample ID Moment Statistics % Sand % Silt % Clay Core/Section Interval/Sample Mean (Φ) Std. Dev. Skewness Kurtosis FRG-1 1 85.8 0.2 14.0 1.0 1.9 12.2 98.3 2 87.7 2.0 10.4 1.1 1.7 9.1 67.5 3 89.2 1.5 9.3 1.0 1.2 4.4 31.2 4 98.0 0.9 1.0 1.5 1.2 4.4 32.0 5 96.0 2.3 1.7 1.3 1.5 7.2 51.5 6 94.6 3.7 1.8 1.3 1.4 6.1 40.1 7 92.9 3.4 3.7 1.9 1.5 4.1 29.5 8 82.4 6.1 11.5 1.9 2.2 10.5 73.2 9 96.5 0.8 2.7 1.3 1.1 1.1 7.3 10 47.6 18.2 34.2 3.6 2.9 5.2 46.4 11 49.7 41.0 9.3 3.3 1.8 3.0 30.2 12 82.6 9.1 8.3 3.5 1.8 6.2 33.4 13 86.7 5.8 7.5 2.0 2.1 6.0 45.8 14 74.7 15.8 9.4 2.7 1.5 3.1 17.9 15 96.0 1.6 2.5 1.2 1.4 5.2 37.9 FRG-2 1 90.5 0.8 8.7 1.5 1.4 6.1 44.9 2 95.2 0.1 4.7 1.1 1.2 6.4 50.8 3 90.5 2.9 6.6 1.9 1.8 6.3 47.1 4 95.3 2.7 2.0 1.2 1.3 3.3 17.0 5 82.1 7.5 10.4 2.6 2.1 5.6 44.8 6 87.7 4.0 8.3 2.2 2.1 5.9 48.1 7 90.6 2.8 6.6 1.8 1.7 2.3 18.7 8 89.4 5.0 5.6 2.8 1.6 5.1 31.4 9 89.9 5.7 4.4 2.7 1.4 3.0 17.5 10 87.7 3.8 8.6 2.8 1.4 3.3 26.1 11 94.4 0.3 5.4 1.6 1.8 5.9 44.6 12 86.1 3.6 10.3 2.9 1.6 3.5 28.9 13 69.4 20.5 10.2 2.6 1.6 3.5 27.9 FGL-1 1 78.8 1.6 19.5 2.3 2.1 10.1 68.7 2 84.7 4.7 10.6 2.3 1.9 7.2 49.1 3 85.1 5.2 9.7 2.2 1.8 5.8 41.7 4A 82.9 4.7 12.4 2.4 1.8 6.8 45.6 4B 90.9 0.3 8.8 1.5 1.7 9.0 65.9 5 97.0 1.3 1.7 1.2 1.3 7.3 49.6 6A 90.9 0.4 8.7 1.4 1.5 8.7 64.4 6B 90.4 0.0 9.6 0.8 1.6 10.4 82.7 7 86.7 -0.6 13.9 1.7 1.3 6.8 48.7 8 90.2 0.3 9.5 1.0 1.9 12.0 93.4 9 73.7 15.8 10.5 3.7 1.6 4.5 22.5

224

Table D-2. Grain-size distributions and moment statistics for the EPK kaolin mine in north- central Florida. Sample ID Moment Statistics % Sand % Silt % Clay Core/Section Interval/Sample Mean (Φ) Std. Dev. Skewness Kurtosis EPK36-J-12 25-27 81.3 5.6 13.1 2.1 1.6 6.6 45.1 27-30 86.8 5.7 7.5 1.6 1.5 5.7 40.0 35-40 81.8 10.0 8.2 1.7 1.9 6.8 45.6 40-44 84.0 1.3 14.7 1.7 1.2 4.4 37.4 44-46 83.9 3.5 12.6 2.3 2.0 8.4 57.8 46-48 83.9 3.2 12.9 1.9 1.9 9.7 69.8 48-50 80.8 5.4 13.7 2.3 1.8 6.7 45.2 50-53 76.1 19.1 4.7 3.3 1.8 6.1 41.6 53-56 84.4 5.9 9.7 1.7 2.1 8.2 56.5 56-59 76.6 7.8 15.6 3.5 2.2 5.3 41.0 59-62 69.0 10.7 20.3 3.8 1.8 4.9 24.7 EPK31-P-40 27-35 82.7 3.5 13.8 1.8 1.9 9.5 67.4 35-45 75.0 10.3 14.7 3.5 2.2 6.3 40.8 45-50 77.8 6.2 16.0 3.1 1.8 6.3 40.7 50-62 80.3 1.5 18.2 3.0 1.7 6.7 44.3 62-65 80.0 3.2 16.8 2.5 1.2 2.4 25.1 EPK30-V-6 16-22 86.7 8.3 5.0 2.5 1.4 4.6 27.1 22-24 89.6 6.6 3.8 2.3 1.0 2.5 15.3 24-27 85.6 11.4 3.0 2.6 1.3 3.3 18.0 30-35 92.3 2.6 5.1 2.2 0.8 1.1 5.4 35-39 87.3 2.7 10.1 2.0 1.4 6.8 43.4 39-43 84.4 8.7 6.9 2.0 1.4 4.1 26.1 43-48 93.9 1.3 4.8 1.6 1.0 2.0 10.5 48-53 87.6 3.4 9.0 1.6 1.9 8.9 60.6 53-58 85.2 5.9 8.9 2.3 1.8 4.9 39.1 58-63 77.0 12.5 10.5 3.3 1.9 4.2 31.6 63-68 78.4 7.9 13.8 3.1 2.0 4.9 39.6 68-73 79.0 8.0 13.0 3.1 1.8 4.4 34.3 73-78 83.0 5.2 11.9 2.4 1.9 3.8 36.2

Table D-3. Grain-size distributions and moment statistics for the Davenport and Joshua sand mines in central Florida. Sample ID Moment Statistics % Sand % Silt % Clay Core/Section Interval/Sample Mean (Φ) Std. Dev. Skewness Kurtosis SSD-1 1 89.9 1.8 8.3 1.7 1.2 3.1 22.5 2 93.2 3.5 3.3 2.3 1.4 5.0 32.6 3 92.7 4.4 3.0 1.3 1.5 3.7 20.1 4 89.7 7.3 3.0 2.6 1.5 3.4 21.1 5 93.3 2.0 4.7 1.4 1.2 1.8 8.0 6 89.4 4.1 6.5 2.7 1.7 3.9 31.7 7 89.6 6.7 3.6 3.3 1.4 4.0 24.9 8 87.9 5.4 6.7 3.1 1.0 1.5 10.3 9 90.1 4.7 5.2 3.1 1.1 2.8 15.2 10 89.5 5.6 4.9 3.1 1.1 3.0 14.9 SSJ-1 1 80.6 2.5 16.9 1.0 1.1 3.4 18.3 2 81.3 1.3 17.5 1.2 1.2 3.8 25.1 3 86.7 2.6 10.6 1.3 1.6 7.8 50.8 4 85.4 3.6 11.0 1.9 1.5 5.7 35.8 5 97.1 0.8 2.0 1.4 1.5 6.2 42.0 6 90.7 0.5 8.8 1.2 1.3 4.1 29.2 7 86.5 3.5 10.0 2.1 1.5 3.8 28.1 8 86.2 3.9 9.9 2.3 1.5 3.5 28.6 9 87.5 4.0 8.5 2.3 1.5 2.9 23.1 10 87.7 1.7 10.6 2.4 1.4 4.2 29.9 11 85.2 1.9 12.8 2.9 1.3 4.0 27.6

225

Table D-4. Grain-size distributions and moment statistics for Cypresshead Formation sampling locations in southeastern Georgia. Sample ID Moment Statistics % Sand % Silt % Clay Core/Section Interval/Sample Mean (Φ) Std. Dev. Skewness Kurtosis J-1 1 69.3 11.1 19.6 2.4 1.8 4.9 32.8 2 68.9 3.3 27.8 2.1 1.4 5.0 33.1 3 73.4 7.6 19.1 1.7 2.0 7.2 47.5 4 14.9 22.1 63.0 4.6 1.9 4.2 23.9 5 83.3 4.4 12.3 2.4 1.3 1.5 11.8 6 90.9 2.8 6.3 1.7 1.3 4.4 29.2 L-1 1 76.4 3.2 20.4 2.0 1.3 3.0 20.0 2 72.6 2.9 24.6 1.9 1.5 4.8 32.3 3 80.5 3.8 15.6 2.1 1.4 4.7 30.9 4 84.5 3.1 12.4 1.8 1.5 4.8 35.5 5 94.9 2.6 2.4 1.9 1.3 3.9 24.9 6 18.6 18.4 63.0 4.6 2.9 0.6 39.5 7 95.2 2.6 2.2 1.9 1.2 4.2 24.7 B-1 1 63.7 5.1 31.2 3.4 1.1 3.9 18.8 2 56.1 10.6 33.3 3.8 1.7 5.4 27.7 3 55.5 16.9 27.6 3.7 1.8 5.0 23.6 4 56.0 10.6 33.4 3.3 2.5 6.5 48.9 5 67.9 8.1 24.0 3.7 2.0 6.2 38.7

226

(25-27) (27-30) (35-40) (40-44) (44-46) (46-48) (48-50) (50-53) (53-56) (56-59) (59-62)

20 Mass Frequency (%) Frequency Mass

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-1. Grain-size distribution curves for EPK Mine core EPK36-J-12.

(27-35) (35-45) (45-50) (50-62) (62-65)

20 Mass Frequency (%)

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-2. Grain-size distribution curves for EPK Mine core EPK31-P-40.

227

(16-22) (22-24) (24-27) (30-35) (35-39) (39-43) (43-48)

(48-53) (53-58) (58-63) (63-68) (68-73) (73-78)

15 Mass Frequency (%) Frequency Mass 10

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-3. Grain-size distribution curves for EPK Mine core EPK30-V-6.

FRG-1-1 FRG-1-2 FRG-1-3 FRG-1-4 FRG-1-5 FRG-1-6 FRG-1-7 FRG-1-8 FRG-1-9 FRG-1-10 FRG-1-11 FRG-1-12 FRG-1-13 FRG-1-14 FRG-1-15

20 Mass FrequencyMass (%) 10

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-4. Grain-size distribution curves for Grandin Sand Mine section FGR-1.

228

FRG-2-1 FRG-2-2 FRG-2-3 FRG-2-4 FRG-2-5 FRG-2-6 FRG-2-7 FRG-2-8 FRG-2-9 FRG-2-10 FRG-2-11 FRG-2-12 FRG-2-13

20 Mass Frequency (%) Frequency Mass 10

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-5. Grain-size distribution curves for Grandin Sand Mine section FRG-2.

FRL-1-1 FRL-1-2 FRL-1-3 FRL-1-4 FRL-1-5 FRL-1-6 FRL-1-7 FRL-1-8 FRL-1-9

Mass Frequency (%) 20

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-6. Grain-size distribution curves for Goldhead Sand Mine section FRL-1.

229

SSJ-1-1 SSJ-1-2 SSJ-1-3 SSJ-1-4 SSJ-1-5 SSJ-1-6 SSJ-1-7 SSJ-1-8 SSJ-1-9 SSJ-1-10 SSJ-1-11

20 Mass Frequency (%) 10

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-7. Grain-size distribution curves for Joshua Sand Mine core SSJ-1.

SSD-1-1 SSD-1-2 SSD-1-3 SSD-1-4 SSD-1-5 SSD-1-6 SSD-1-7 SSD-1-8 SSD-1-9 SSD-1-10

20 Mass Frequency (%)

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-8. Grain-size distribution curves for Davenport Sand Mine core SSD-1.

230

J-1-1 J-1-2 J-1-3 J-1-4 J-1-5 J-1-6

Mass FrequencyMass (%) 10

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-9. Grain-size distribution curves for Jesup section J-1.

L-1-1 L-1-2 L-1-3 L-1-4 L-1-5 L-1-6 L-1-7

20 Mass Frequency (%)

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-10. Grain-size distribution curves for Linden Bluff section L-1.

231

B-1-1 B-1-2 B-1-3 B-1-4 B-1-5

20 Mass Frequency (%)

0 1 0.1 0.01 0.001 Particle Size (mm)

Figure D-11. Grain-size distribution curves for Birds section B-1.

232

APPENDIX E X-RAY DIFFRACTION DATA (ORIENTED)

Note: If not marked, diffraction peaks are assigned to kaolinite. The legend for other minerals is as follows:

• Q – quartz • G – gibbsite • C – crandallite-florencite • H – halloysite • HIV – hydroxyl-interlayered vermiculite • I – illite • S – smectite • B – boehmite • D – diaspore • A – anatase

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

APPENDIX F X-RAY DIFFRACTION DATA (RANDOM)

Note: If not marked, diffraction peaks are assigned to kaolinite. The legend for other minerals is as follows:

• Q – quartz • G – gibbsite • C – crandallite-florencite • H – halloysite • HIV – hydroxyl-interlayered vermiculite • I – illite • S – smectite • R – rutile • B – boehmite • D – diaspore • A – anatase • Gr - goethite

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

APPENDIX G MINUS-200 MESH PARTICLE-SIZE DATA (SEDIGRAPH)

266

36-J-12 (25-27) 36-J-12 (35-40) 36-J-12 (46-48) 36-J-12 (50-53) 36-J-12 (56-59)

100

20 Cumulative Mass (%) Cumulative Finer

0 100 10 1 0.1 Particle Diameter (um)

36-J-12 (25-27) 36-J-12 (35-40) 36-J-12 (46-48) 36-J-12 (50-53) 36-J-12 (56-59)

8.0

6.0

4.0

Mass Frequency (%) Frequency Mass 2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-1. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK36-J-12).

267

31-P-40 (35-45) 31-P-40 (50-62) 31-P-40 (62-65)

100

Cumulative Mass Finer (%) Finer Mass Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

31-P-40 (35-45) 31-P-40 (50-62) 31-P-40 (62-65)

2.5

2.0

1.5

1.0 Mass Frequency (%) Frequency Mass

0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-2. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK31-P-40).

268

30-V-6 (22-24) 30-V-6 (30-35) 30-V-6 (48-53) 30-V-6 (58-63) 30-V-6 (68-73)

100

20 Cumulative Mass Finer (%) Finer Mass Cumulative

0 100 10 1 0.1 Particle Diameter (um)

30-V-6 (22-24) 30-V-6 (30-35) 30-V-6 (48-53) 30-V-6 (58-63) 30-V-6 (68-73)

2.5

2.0

1.5

1.0 Mass Frequency (%) Frequency Mass 0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-3. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select EPK Mine samples (EPK30-V-6).

269

FRG-1-3 FRG-1-6 FRG-1-7 FRG-1-10 FRG-1-11 FRG-1-13 FRG-1-16

100

20 Cumulative Mass Finer (%) Mass Finer Cumulative

0 100 10 1 0.1 Particle Diameter (um)

FRG-1-3 FRG-1-6 FRG-1-7 FRG-1-10 FRG-1-11 FRG-1-13 FRG-1-16

3.0

2.5

2.0

1.5

1.0 Mass Frequency (%) Frequency Mass

0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-4. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Grandin Sand Mine samples (FRG-1).

270

FRG-2-3 FRG-2-5 FRG-2-7 FRG-2-10 FRG-2-12

100

Cumulative Mass Finer (%) Mass Finer Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

FRG-2-3 FRG-2-5 FRG-2-7 FRG-2-10 FRG-2-12

2.5

2.0

1.5

1.0 Mass Frequency (%)

0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-5. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Grandin Sand Mine samples (FRG-2).

271

FRL-1-2 FRL-1-4 FRL-1-6 FRL-1-7 FRL-1-9

100

Cumulative Mass Finer (%) Mass Finer Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

FRL-1-2 FRL-1-4 FRL-1-6 FRL-1-7 FRL-1-9

2.5

2.0

1.5

1.0 Mass Frequency (%)

0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-6. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Goldhead Sand Mine samples (FRL-1).

272

SSJ-1-2 SSJ-1-4 SSJ-1-6 SSJ-1-8 SSJ-1-10

100

Cumulative Mass Finer (%) Mass Finer Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

SSJ-1-2 SSJ-1-4 SSJ-1-6 SSJ-1-8 SSJ-1-10

10.0

8.0

6.0

4.0 Mass Frequency (%) Frequency Mass

2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-7. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Joshua Sand Mine samples (SSJ-1).

273

SSD-1-1 SSD-1-3 SSD-1-6 SSD-1-7 SSD-1-10

100

Cumulative Mass Finer (%) Finer Mass Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

SSD-1-1 SSD-1-3 SSD-1-6 SSD-1-7 SSD-1-10

10.0

8.0

6.0

4.0 Mass Frequency (%) Frequency Mass

2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-8. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Davenport Sand Mine samples (SSD-1).

274

J-1-2 J-1-4 J-1-6

100

Cumulative Mass Finer (%) Finer Mass Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

J-1-2 J-1-4 J-1-6

10.0

8.0

6.0

4.0 Mass Frequency (%) Frequency Mass

2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-9. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Jesup type locality samples (J-1).

275

LB-1-3 LB-1-5 LB-1-6

100

Cumulative Mass Finer (%) 20

0 100 10 1 0.1 Particle Diameter (um)

LB-1-3 LB-1-5 LB-1-6

10.0

8.0

6.0

4.0 Mass Frequency (%)

2.0

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-10. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Linden Bluff reference locality samples (L-1).

276

B-1-2 B-1-3 B-1-5

100

Cumulative Mass Finer (%) Mass Finer Cumulative 20

0 100 10 1 0.1 Particle Diameter (um)

B-1-2 B-1-3 B-1-5

2.5

2.0

1.5

1.0 Mass Frequency (%) Frequency Mass

0.5

0.0 100 10 1 0.1 Particle Diameter (um)

Figure G-11. SediGraph particle-size plots for the minus-200 mesh (< 75 µm) fraction of select Birds reference locality samples (B-1).

277

APPENDIX H MAJOR AND TRACE ELEMENT DATA

278

Table H-1. Raw major element concentrations for samples used in this study.

Sample ID Interval Na (%) Mg (%) Al (%) P (%) K (%) Ca (%) Ti (%) Fe (%) Cypresshead Formation - FL EPK36-J-12 25-27 0.008 0.046 19.42 0.06 0.03 0.034 0.15 0.19 EPK36-J-12 35-40 0.01 0.03 17.89 0.065 0.04 0.035 0.14 0.25 EPK36-J-12 46-48 0.011 0.044 19.9 0.031 0.04 0.025 0.169 0.34 EPK36-J-12 50-53 0.015 0.03 17.04 0.02 0.07 0.02 0.19 0.36 EPK36-J-12 59-62 0.015 0.06 18.09 0.03 0.08 0.055 0.33 0.54 EPK31-P-40 35-45 0.013 0.05 19.91 0.027 0.05 0.032 0.208 0.46 EPK31-P-40 50-62 0.015 0.095 17.17 0.03 0.07 0.105 0.185 0.59 EPK31-P-40 62-65 0.013 0.237 18.28 0.075 0.19 0.206 0.171 0.87 EPK30-V-6 22-24 0.015 1.465 5.033 0.097 0.23 0.458 0.355 1.78 EPK30-V-6 30-35 0.025 0.77 13.29 0.515 0.24 0.6 0.18 1.07 EPK30-V-6 48-53 0.025 0.135 17.29 0.275 0.13 0.355 0.25 0.62 EPK30-V-6 58-63 0.015 0.036 19.71 0.022 0.07 0.016 0.151 0.29 EPK30-V-6 68-73 0.015 0.07 17.18 0.025 0.12 0.05 0.26 0.47 FRG-1 3 0.009 0.05 18.86 0.122 0.04 0.03 0.187 2.09 FRG-1 6 0.005 0.015 17.85 0.07 0.02 0.015 0.04 0.69 FRG-1 7 n.a. 0.015 18.56 0.04 0.02 n.a. 0.045 0.45 FRG-1 10 0.01 0.055 18.25 0.105 0.08 0.03 0.13 0.43 FRG-1 11 0.005 0.02 18.61 0.03 0.03 0.01 0.065 0.16 FRG-1 13 0.005 0.015 18.93 0.02 0.03 0.005 0.07 0.15 FRG-1 15 0.005 0.015 17.65 0.035 0.04 0.005 0.135 0.16 FRG-2 3 0.05 0.025 18.45 0.025 0.03 0.01 0.065 0.19 FRG-2 5 n.a. 0.015 18.05 0.024 0.03 n.a. 0.064 0.25 FRG-2 7 0.045 0.02 18.78 0.02 0.04 0.01 0.09 0.19 FRG-2 10 0.01 0.035 18.43 0.055 0.07 0.015 0.16 0.44 FRG-2 12 0.008 0.083 17.75 0.063 0.09 0.01 0.216 0.98 FRL-1 2 0.005 0.033 17.57 0.025 0.05 n.a. 0.185 0.58 FRL-1 4 0.005 0.034 17.62 0.028 0.06 n.a. 0.189 0.66 FRL-1 6 0.005 0.045 17.93 0.03 0.04 0.01 0.15 0.43 FRL-1 7 0.005 0.05 17.06 0.02 0.04 0.01 0.13 0.41 SSJ-1 2 0.014 0.091 18.14 0.054 0.08 0.027 0.692 2.25 SSJ-1 4 0.012 0.098 17.76 0.061 0.07 0.017 0.349 1.5 SSJ-1 6 0.015 0.055 18.55 0.025 0.07 0.02 0.095 0.33 SSJ-1 8 0.016 0.066 18.41 0.023 0.17 0.029 0.152 0.28 SSJ-1 10 0.02 0.095 18.34 0.025 0.13 0.05 0.245 0.36 TRF2214 60.0-62.5 0.075 0.446 14.89 0.03 0.19 0.354 0.501 1.25 WEX164 18.0-26.0 0.131 0.704 14.05 0.178 0.68 0.217 0.506 3.04 WEX366 9.0-10.0 0.167 0.378 14.38 0.056 1.11 0.047 0.724 6.59 EPK Vermiforms — 0.01 0.011 18.19 0.016 0.01 0.026 0.019 0.3 EPK Mica — 0.349 0.44 17.19 0.008 5.75 0.013 0.393 1.62 EPK Feldspar — 0.305 0.017 5.703 0.118 6.16 0.309 0.01 0.07 Reworked Cypresshead Formation - FL SSD-1 1 0.06 0.145 15.06 0.863 0.34 0.072 4.778 1.48 SSD-1 3 0.01 0.042 20.02 0.728 0.07 0.05 0.434 0.39 SSD-1 6 0.008 0.061 17.49 0.047 0.04 0.006 0.137 0.46 SSD-1 7 0.015 0.065 17.87 0.035 0.06 0.01 0.155 0.51 SSD-1 10 0.015 0.175 17.8 0.035 0.09 0.07 0.135 0.73 Note: n.a., element concentration was below the detection limit.

279

Table H-1. – (continued).

Sample ID Interval Na (%) Mg (%) Al (%) P (%) K (%) Ca (%) Ti (%) Fe (%) Cypresshead Formation - GA J-1 2 0.017 0.149 8.068 0.046 0.16 0.009 0.84 5.35 J-1 4 0.025 0.125 18.12 0.035 0.43 0.01 0.615 3.13 J-1 6 0.02 0.11 17.79 0.015 0.31 0.005 0.46 1.33 L-1 3 0.014 0.079 19.24 0.049 0.15 n.a. 0.616 1.2 L-1 5 n.a. 0.03 18.71 0.019 0.04 n.a. 0.224 0.59 L-1 6 0.015 0.075 17.89 0.025 0.22 0.005 0.51 1.13 B-1 2 0.038 0.38 14.96 0.038 1 0.01 0.399 3.93 B-1 3 0.047 0.424 15.52 0.093 1.28 0.009 0.43 7.63 B-1 5 0.04 0.425 17.62 0.04 0.89 0.065 0.445 2.54 Hawthorn Group, Coosawhatchie Formation - FL/GA MCB109 15.0-20.0 0.066 1.015 11.69 0.037 1.25 0.743 0.416 4.18 J-1 BC 0.057 1.071 11.77 0.013 0.3 0.312 0.371 4.22 Huber Formation - GA KGa-2 — 0.01 0.02 17.96 0.02 0.03 0.01 0.855 0.66 ECCI-CB — 0.02 0.025 17.56 0.025 0.07 0.005 0.665 0.63 Buffalo Creek Formation - GA KGa-1 — 0.02 0.015 18.95 0.025 n.a. 0.02 0.94 0.13 ECCI-BC — 0.015 0.01 18.65 0.035 n.a. 0.015 0.75 0.11 TKC-EA — 0.01 0.01 18.97 0.025 0.02 0.005 0.87 0.12 DBK-B93 — 0.015 0.015 17.2 0.03 n.a. 0.015 1.327 0.17 Note: n.a., element concentration was below the detection limit.

280

Table H-2. Minor element concentrations for samples used in this study (concentration are in ppm).

Sample ID Interval Rb Sr Cs Ba Sc V Cr Mn Co Ni Cu Zn Ga Y Zr Nb Pb Th U Cypresshead Formation - FL EPK36-J-12 25-27 2.7 48.8 0.3 129 3 24 43 3.1 45 81 42 28 6 45.2 6 62 5.1 2.51 EPK36-J-12 35-40 3 121 0.3 72.8 5 18 16 10 3.7 39 93 47 28 14.2 73.7 5 63 7 4.79 EPK36-J-12 46-48 4.2 35.9 0.5 53.1 10 36 43 6 3.1 22 76 46 36 7 32 6 117 5.7 2.06 EPK36-J-12 50-53 5.2 21.5 0.4 31.1 10 32 16 7 2.4 18 21 21 35 4.2 24.9 6 69 3.5 1.93 EPK36-J-12 59-62 6.1 37.2 0.4 78.3 16 60 45 9 2.6 20 49 33 39 8.7 38 11 116 6.9 4.86 EPK31-P-40 35-45 4.5 21.3 0.3 37.2 12 52 45 6 1.8 14 23 20 36 4.6 22.2 6 119 4.7 17.7 EPK31-P-40 50-62 5.2 34.8 0.4 87.7 21 47 21 10 2.3 18 36 24 34 16.1 18.5 6 106 4.4 103 EPK31-P-40 62-65 17.6 123 1.4 277 20 77 67 12 3.4 28 37 45 33 41.2 30.5 6 115 7.8 92.9 EPK30-V-6 22-24 25.7 348 2.3 127 6 63 65 178 5.2 32 47 166 15 41.6 177 16 113 15.1 5.58 EPK30-V-6 30-35 23.1 2040 2.5 419 13 79 42 91 5.2 51 53 168 33 257 255 22 98 56.9 46.3 EPK30-V-6 48-53 12.7 549 1.8 189 12 79 38 26 4.6 50 43 46 32 73.2 151 12 103 23 29.3 EPK30-V-6 58-63 6 19.8 0.3 44.4 14 84 46 1.7 13 71 32 35 8.2 23.1 5 115 3.8 18.9 EPK30-V-6 68-73 8.9 25.9 0.5 71.3 22 113 20 11 2.5 17 15 15 37 12.6 23 8 76 4.3 16.8 FRG-1 3 4.6 229 0.8 114 7 61 59 11 4.6 30 67 45 23 26.9 84.9 9 59 17.7 5.81 FRG-1 6 0.8 125 0.1 80.8 5 16 17 11 5.5 39 153 76 19 8.5 19.1 2 108 6 1.62 FRG-1 7 1.8 76 0.2 44.8 4 18 41 3.5 19 58 37 22 4.8 16.4 2 130 7.4 1.16 FRG-1 10 6.7 454 0.6 151 6 21 10 7 3.5 17 44 29 30 13.8 24.9 5 367 11.7 1.57 FRG-1 11 1.7 83.5 0.1 38.9 5 16 6 2.1 18 55 30 26 3.6 17.7 2 125 7.1 0.91 FRG-1 13 1.7 33.1 20.7 5 18 8 4 1.9 16 122 55 27 1.7 21.5 2 63 5.6 1.04 FRG-1 15 2.3 35.4 0.1 46.6 11 27 10 5 1.8 20 291 130 34 5.5 18.2 5 255 6.5 1.38 281 FRG-2 3 1.7 46.5 0.1 28.5 5 11 13 5 2.1 18 14 11 29 2.6 22.7 2 40 5.3 0.83 FRG-2 5 2.4 28.1 0.2 24.4 5 20 33 1.5 14 43 24 31 2.4 16.3 2 84 5.9 1.01 FRG-2 7 2.3 16.4 0.1 19.9 7 16 7 4 1.3 14 51 25 31 2.1 15.6 3 75 3.9 1.19 FRG-2 10 5.8 66.9 0.5 178 11 39 13 8 1.9 18 50 27 36 16 23.3 6 151 10 2.27 FRG-2 12 10 99.2 0.9 228 15 57 47 10 2.6 20 40 30 34 25.9 29.9 7 108 9 4.62 FRL-1 2 5.1 21.5 0.4 27.6 14 50 50 5 3.7 22 74 42 35 4 46 6 83 7.9 1.28 FRL-1 4 4.7 26.8 0.4 32.9 14 61 53 5 3.8 25 57 41 35 5.3 49.6 6 152 10.6 1.16 FRL-1 6 3.4 32.1 0.4 35.8 15 35 13 8 2.4 22 76 39 34 7.8 43.3 5 127 8.3 1.13 FRL-1 7 3.6 22.8 0.4 26 13 30 14 6 2.5 16 51 37 33 8.4 32.1 5 65 6.1 1.27 SSJ-1 2 5.8 87 1 60.6 5 78 92 15 4.6 42 47 28 48 17.6 216 30 75 29.4 3.11 SSJ-1 4 8.3 70.1 1.5 48.3 6 61 58 14 4.3 34 56 36 42 10.4 117 16 62 18.3 2.67 SSJ-1 6 5.9 49.2 0.6 29.5 6 27 11 9 2.2 41 46 20 33 2.6 18.8 4 44 2.2 3.03 SSJ-1 8 10.8 47.3 0.7 92.1 12 56 42 1.4 10 44 22 36 2.4 23.9 5 50 2.4 7.93 SSJ-1 10 11 43.2 0.8 70.2 18 85 37 10 1.4 11 53 20 37 4.1 33.7 9 48 4.6 14.1 SSD-1 1 21.3 777 2.4 505 22 245 160 289 5.5 33 84 54 183 167 1630 253 514 145 31.3 SSD-1 3 5.2 144 0.6 103 7 41 77 21 4.5 47 103 60 37 16.4 180 18 136 56.8 7.54 SSD-1 6 4.1 41.2 0.4 61.7 8 44 50 3.4 18 61 37 29 8 34.8 5 99 10 1.64 SSD-1 7 4.5 41 0.4 66.9 10 45 16 9 2 22 54 31 33 8.9 28.8 5 117 6.8 2.71 SSD-1 10 7.8 38 0.7 82.2 9 33 18 22 1.7 20 27 20 32 7.3 23 5 87 4.4 2.04 TRF2214 60.0-62.5 29.2 66 3.7 184 15 119 60 32 3.6 8 23 28 37 16.8 127 26 44 10.6 2.43 WEX164 18.0-26.0 87.7 320 9.6 305 26 189 169 51 10.2 23 62 87 35 105 63.9 19 42 22.3 13.3 WEX366 9.0-10.0 124 118 10.9 322 23 183 122 50 5.9 18 107 100 42 32.6 238 29 43 26.6 9.9 EPK Vermiforms — 1.1 7.3 0.1 13.5 11 39 17 1.8 8 40 31 39 20.7 7.9 25 1 1.21 EPK Mica — 231 38.5 2.6 1610 38 153 65 87 2.4 7 84 53 52 9.3 130 31 31 2.9 1.46 EPK Feldspar — 191 265 1.3 1870 6 12 8 0.5 12 17 17 10 2.9 21.5 58 0.5 1.63

Table H-2. – (continued). Sample ID Interval Rb Sr Cs Ba Sc V Cr Mn Co Ni Cu Zn Ga Y Zr Nb Pb Th U Cypresshead Formation - GA J-1 2 23.3 35.9 4.7 90.2 21 195 110 35 6.3 47 59 51 49 22.3 231 37 51 36.6 5.21 J-1 4 39.4 34.8 3.5 143 24 190 77 35 5.4 22 322 164 41 7.9 74.7 22 36 21.2 3.04 J-1 6 28.2 23.9 2.7 102 17 67 45 33 6 52 879 397 37 6.6 57.5 16 43 14.4 2.03 L-1 3 17.4 40.5 2.4 116 19 100 80 20 5.8 38 65 47 40 10.7 104 22 113 34.3 2.87 L-1 5 6.4 10.7 0.7 34.7 10 68 59 8 7.1 56 173 74 29 3.9 38.7 7 65 10.8 1.16 L-1 6 21.5 28.8 2.1 87.4 16 80 24 21 6.1 39 75 47 31 6.8 60.3 19 87 16.6 1.85 B-1 2 97.4 52.6 9.2 227 25 149 84 54 7.3 29 66 77 32 14.4 82.3 15 57 15.9 4.01 B-1 3 116 77.4 10.8 270 31 198 91 108 10.4 33 68 96 34 24.3 83.6 15 64 20.4 5.55 B-1 5 88.5 112 8.1 201 36 168 30 51 8 34 80 75 38 50.2 74.4 15 60 18.5 4.03 Hawthorn Group, Coosawhatchie Formation - FL/GA MCB109 15.0-20.0 139 112 10.9 304 21 177 127 56 11.5 24 31 86 29 41 84.9 15 25 12.7 4.38 J-1 BC 42.9 60.8 2.6 180 22 139 65 136 44.4 46 42 83 26 147 37.6 12 29 11.3 1.72 Huber Formation - GA KGa-2 — 1.2 51.2 0.2 70.3 15 105 29 5 9.4 39 72 70 70 20.1 78.6 37 43 13 3.52 ECCI-CB — 2.6 45.2 0.3 79.4 20 76 16 6 9.9 35 69 59 48 10.1 51.9 24 29 14.7 4.39 Buffalo Creek Formation - GA KGa-1 — n.a. 43.2 n.a. 41.8 19 223 38 n.a. 3.3 17 63 24 53 8.4 98.1 39 18 36.3 2.27 ECCI-BC — n.a. 50.5 n.a. 64.6 14 127 16 n.a. 3.1 18 116 55 53 5.8 115 30 n.a. 31.4 1.53 TKC-EA — 0.6 36.7 n.a. 64.6 21 102 26 n.a. 4.4 20 81 34 46 8.1 111 37 13 31.1 8.72 DBK-B93 — 0.3 41.2 n.a. 155 22 249 90 n.a. 1 5 37 10 49 10.3 149 51 n.a. 31.9 6.52 282

LIST OF REFERENCES

Addington, L.D., Kuehl, S.A., and McNinch, J.E., 2007, Contrasting modes of shelf sediment dispersal off a high-yield river: Waiapu River, New Zealand: Marine Geology, v. 243, p. 18–30.

Allmon, W.D., 1992, Whence southern Florida’s Plio-Pleistocene shell beds? in Scott, T.M., and Allmon, W.D., eds., Plio-Pleistocene stratigraphy and paleontology of southern Florida: Florida Geological Survey Special Publication 36, p. 1–20.

Alpers, C.N., and Brimhall, G.H., 1988, Middle Miocene climatic change in the Atacama Desert, northern Chile; Evidence from supergene mineralization at La Escondida. Geological Society of America, Bulletin, v. 100, p. 1640–1656.

Alt, D., 1974, Arid climate control of Miocene sedimentation and origin of modern drainage, southeastern United States, in Oaks, R.Q., Jr., and Dunbar, J.R., eds., Post-Miocene Stratigraphy, Central and Southern Atlantic Coastal Plain: Logan, Utah, Utah State University Press, p. 21–29.

Anovitz, L.M., Perkins, D., and Essene, E.J., 1991, Metastability in near-surface rocks of minerals in the system Al2O3-SiO2-H2O: Clays and Clay Minerals, v. 39, p. 225–233.

Aparicio, P., and Galán, E., 1999, Mineralogical interference on kaolinite crystallinity index measurements: Clays and Clay Minerals, v. 47, p. 12–27.

Applin, E.R., and Applin, P.L., 1964, Logs of selected wells in the coastal plain of Georgia: Georgia Geological Survey, Bulletin 74, 229 p.

Arostegui, J., Irabien, M.I., Nieto, F., Sangüesa, J., and Zuluaga, M.C., 2001, Microtextures and the origin of muscovite-kaolinite intergrowths in sandstones of the Utrillas Formation, Basque Cantabrian Basin, Spain: Clays and Clay Minerals, v. 49, p. 529–539.

Arrington, D.V., 1985, Geology of the Interlachen Karstic Highlands, Florida [unpublished M.S. thesis]: University of Florida, Gainesville, Florida, 130 p.

Artioli, G., Bellotto, M., Gualtieri, A., and Pavese, A., 1955, Nature of structural disorder in natural kaolinites: a new model based on computer simulation of powder diffraction data and electrostatic energy calculation: Clays and Clay Minerals, v. 43, p. 438–445.

Ashley, R.P., and Silberman, M.L., 1976, Direct dating of mineralization at Goldfield, Nevada, by potassium-argon and fission-track methods: Economic Geology, v. 71, p. 904–924.

Austin, R.S., 1978, The origin of Georgia’s kaolin deposits: Georgia Geological Survey, Information Circular 49, p. 10–15.

Austin, R.S., 1998, Origin of kaolin of the southeastern U.S.: Mining Engineering, v. 50, p. 52– 57.

283

Axelrod, D.I., and Raven, P.H., 1985, Origins of the Cordilleran flora: Journal of Biogeography, v. 12, p. 21–47.

Axelrod, D.I., and Raven, P.H., 1986, Late Cretaceous and Tertiary vegetation history of Africa, in Werger, M.J., and Bruggen, A.C., eds., Biogeography and Ecology of Southern Africa: The Hague, Junk, p. 77–130.

Bailey, S.W., 1980, Structure of layer silicates, in Brindley, G.W., and Goodman, G., eds., Crystal structures of clay minerals and their X-ray identification: Mineralogical Society of London, p. 1–124.

Balsillie, J.H., 1995, William F. Tanner on environmental clastic granulometry: Florida Geological Survey, Special Publication 40, 144 p.

Banfield, J.F., and Eggleton, R.A., 1990, Analytical transmission electron microscope studies of plagioclase, muscovite, and K-feldspar weathering: Clays and Clay Minerals, v. 38, p. 77– 89.

Barron, E.J., 1989, Severe storms during earth history: Geological Society of America, Bulletin, v. 101, p. 601–612.

Bell O.G., 1924, A preliminary report on the clays of Florida (exclusive of Fuller's earth): Florida Geological Survey, 15th Annual Report, p. 53–266.

Belshaw, N.S., Freedman, P.A., O’Nions, R.K., Frank, M., and Guo, Y., 1998, A new variable dispersion double-focusing plasma mass spectrometer with performance illustrated for Pb isotopes: International Journal of Mass Spectrometry, v. 181, p. 51–58.

Berggren, W.A., 1973, The Pliocene time-scale: calibration of planktonic foraminiferal and calcareous nannoplankton zones: Nature, v. 243, p. 391–397.

Berggren, W.A., Hilgen, F.J., Langereis, C.G., Kent, D.V., Obradovich, J.D., Raffi, I., Raymo, M.E., and Shackleton, N.J., 1995a, Late Neogene chronology: new perspectives in high- resolution stratigraphy: Geological Society of America, Bulletin, v. 107, p. 1272–1287.

Berggren, W.A., Kent, D.V., Swisher, C.C., III, and Aubry, M.-P., 1995b, A revised Cenozoic geochronology and chronostratigraphy, in Berggren, W.A., Kent, D.V., Aubry, M.-P., and Hardenbol, J., eds., Geochronology, Time Scales and Global Stratigraphic Correlation: SEPM, Special Publication 54, p. 129–212.

Bertaut, F., 1950, Rais de Debye-Scherrer et repartition des dimensions des domains de Bragg dans les poudres polycristallines: Acta Crystallographica, v. 3, p. 14–18.

Bhatia M.R., 1985, Rare earth element geochemistry of Australian Paleozoic graywackes and mudrocks: provenance and tectonic control: Sedimentary Geology, v. 45, p. 97–113.

Bish, D.L., and Von Dreele, R.B., 1989, Rietveld refinement of non-hydrogen atomic positions in kaolinite: Clays and Clay Minerals, v. 37, p. 289–296.

284

Bishop E.W., 1956, Geology and ground water resources of Highlands County, Florida: Florida Geological Survey, Report of Investigation 15.

Blanchard, F.N., 1972, Physical and chemical data for crandallite from Alachua County, Florida: American Mineralogist, v. 57, p. 473–484.

Blow, W., 1969, Late middle Eocene to Recent planktonic foraminiferal biostratigraphy, in Brönnimann, P., and Renz, H.H., eds., Proceedings of the First International Conference on Planktonic Microfossils, v. 1, p. 199–422.

Bock, B., McLennan, S.M., and Hanson, G.N., 1994, Rare earth element redistribution and its effects on the neodymium isotope system in the Austin Glen Member of the Normanskill Formation, New York, USA: Geochimica et Cosmochimica Acta, v. 58, p. 5245–5253.

Bouquillon, A., France-Lanord, C., Michard, A., and Tiercelin, J-.J., 1990, Sedimentology and isotopic chemistry of the Bengal fan sediments: the denudation of the Himalaya, in Cochran, J.R., et al., eds., Proceedings of the Ocean Drilling Program, Scientific Results 116, p. 43–58.

Buie B.F., 1978, The Huber Formation of eastern Georgia, in Short Contributions to the Geology of Georgia: Georgia Geological Survey, Bulletin 93, p. 1–7.

Buie B.F., and Fountain R.C., 1967, Tertiary and Cretaceous age of kaolin deposits in Georgia and South Carolina (abstract): Geological Society of America, Program with Abstracts, v. 19, p. 465.

Bukry, D., 1991, Paleoecological transect of western Pacific Ocean late Pliocene coccolith flora: Part I – Tropical Ontong-Java Plateau at ODP 806B: U.S. Geological Survey, Open File Report 91-552, 35 p.

Bundy, W.M., 1993, The diverse industrial applications of kaolin, in Murray, H.H., et al., eds., Kaolin Genesis and Utilization: The Clay Minerals Society, Special Publication No. 1, p. 43–73.

Cases, J.M. Liètard, O., Yvon, J., and Delon, J.F., 1982, Étude des propiértés cristallochimiques, morphologiques, superficielles de kaolinites désordonnées: Bulletin of Minéralogico, v. 105, p. 439–455.

Chamberlain, C.K., 1978, Recognition of trace fossils in cores, in Basan, P.B., ed., Trace Fossil Concepts (SEPM Short Course no. 5), SEPM, Tulsa, Oklahoma, p. 119–166.

Chapman, C.E., 2008, Interplay of oceanographic and paleoclimate events with tectonism during middle to late Miocene sedimentation across the southwestern USA: Geosphere, v. 4, p. 976–991.

Chaudhrui, S., and Cullers, R.L., 1979, The distribution of rare-earth elements in deeply buried Gulf coast sediments: Chemical Geology, v. 24, p. 327–338.

285

Chávez, C.L., and Johns, W., 1995, Mineralogical and ceramic properties of refractory clays from central Missouri (USA): Applied Clay Science, v. 9, p. 407–424.

Coakley, J.P., and Syvitski, J.P.M., 1991, SediGraph technique, in Syvitski, J.P.M., ed., Principles, Methods, and Application of Particle Size Analysis: Cambridge University Press, New York, p. 129–142.

Condie, K.C., Dengate, J., and Cullers, R.L., 1995, Behavior of rare earth elements in a paleoweathering profile on granodiorite in the Front Range, Colorado, USA: Geochimica et Cosmochimica Acta, v. 59, p. 279–294.

Cooke, C.W., 1930, Correlation of coastal terraces: Journal of Geology, v. 38, p. 577–589.

Cooke, C.W., 1936, Geology of the Coastal Plain of South Carolina: Geological Survey, Bulletin 867, p. 132–133.

Cooke, C.W., 1945, Geology of Florida: Florida Geological Survey, Bulletin 29, 339 p.

Cooke, C.W., and MacNeil, F.S., 1952, Tertiary stratigraphy of South Carolina: U.S. Geological Survey, Professional Paper 243-B, 29 p.

Cooke, C.W., and Mossom, S., 1929, Geology of Florida: Florida Geological Survey, Annual Report 20, p. 29–227.

Cronin, T.M., 1991, Pliocene shallow water paleoceanography of the North Atlantic Ocean based on marine ostracods: Quaternary Science Reviews, v. 10, p. 75–188.

Cronin, T.M., and Dowsett, H.J., 1990, A quantitative micropaleontological method for shallow marine paleoclimatology: application to Pliocene deposits of the western North Atlantic ocean: Marine Micropaleontology, v. 16, p. 117–147.

Crowley, T.J., 1996, Pliocene climates: the nature of the problem: Marine Micropaleontology: v. 27, p. 3–12.

Cullers, R.L., 1988, Mineralogical and chemical changes of soil and stream sediment formed by intense weathering of the Danburg granite, Georgia, U.S.A.: Lithos, v. 21, p. 301–314.

Cullers, R.L., Barrett, T., Carlson, R., and Robinson, R., 1987, Rare-earth distributions in size fractions of Holocene soil and stream sediment, Wet Mountains region, Colorado, U.S.A.: Chemical Geology, v. 63, p. 275–297.

Cunningham, K.J., Bukry, D., Sato, T., Barron, J.A., Guertin, L.A., and Reese, R.S., 2001, Lithostratigraphy, sequence stratigraphy and biostratigraphy of a carbonate ramp and bounding siliciclastics (Late Miocene-Pliocene), southern Florida: Florida Geological Survey, Special Publication 49, p. 35–66.

286

Cunningham, K.J., Locker, S.D., Hine, A.C., Bukry, D., Barron, J.A., and Guertin, L.A., 2003, Interplay of Late Cenozoic siliciclastic supply and carbonate response on the southeast Florida Platform: Journal of Sedimentary Research, v. 73, p. 31–46.

Cunningham, K.J., McNeill, D.F., Guertin, L.A., Ciesielski, P.F., Scott, T.M., and de Verteuil, L., 1998, New Tertiary stratigraphy for the Florida Keys and southern peninsula of Florida: Geological Society of America, Bulletin, v. 110, p. 231–258.

Curran, H.A., 1985, The trace fossil assemblage of a Cretaceous nearshore environment: Englishtown Formation of Delaware, U.S.A., in Curran, H.A., ed., Biogenic Structures: Their use in Interpreting Depositional Environments: SEPM, Special Publication 35, Tulsa, Oklahoma, p. 261–276.

Davis N.B., 1916, The plasticity of clay and its relation to mode of origin. American Institute of Mining, Metallurgical, and Petroleum Engineers, Transactions, v. 51, p. 451–480.

Davis, R.A., Hine, A.C., and Shinn, E.A., 1992, Holocene coastal development on the Florida peninsula, in Fletcher, C.H., III, and Wehmiller, J.F., eds., Quaternary Coasts of the United States: Marine and Lacustrine Systems, SEPM, Special Publication 48, p. 193–212.

Dennison, J.M., and Stewart, K.G., 2001, Regional structural and stratigraphic evidence for dating Cenozoic uplift of Southern Appalachian highlands (abstract): Abstracts with Programs, 50th Annual Meeting, Southeastern Section of the Geological Society of America, v. 33, p. 6.

Dill, H.G., Fricke, A., and Henning, K.-H., 1995, The origin of Ba- and REE-bearing aluminum- phosphate-sulphate minerals from the Lohrheim kaolinitic clay deposit (Rheinisches Schiefergebirge, Germany): Applied Clay Science, v. 10, p. 231–245.

Dobkins, J.E., Jr., and Folk, R.L., 1970, Shape development on Tahiti-Nui: Journal of Sedimentary Petrology, v. 40, p. 1167–1203.

Doering, J.A., 1960, Quaternary surface formations of southern part of Atlantic Coastal Plain: Journal of Geology, v. 68, p. 182–202.

Dombrowski T., 1982, Abundance, distribution and origin of thorium in the Georgia kaolins [unpublished M.S. thesis]: Indiana University, South Bend, Indiana, 85 p.

Dombrowski T., 1992, The use of trace elements to determine provenance relations among different types of Georgia kaolins [unpublished Ph.D. dissertation]: Indiana University, South Bend, Indiana, 212 p.

Dombrowski T., 1993, Theories of origin for the Georgia kaolins: a review, in Murray, H.H., et al., eds., Kaolin Genesis and Utilization: The Clay Minerals Society, Special Publication 1, p. 75–97.

287

Dombrowski T., and Murray H.H., 1984, Thorium - a key element in differentiating Cretaceous and Tertiary kaolins in Georgia and South Carolina: Proceedings of the 27th International Geological Congress, v. 15, p. 305–317.

Dott, R.H., Jr., and Bourgeois, J., 1982, Hummocky stratification: significance of its variable bedding sequences: Geological Society of America, Bulletin, v. 93, p. 663–680.

Dowsett, H.J., Barron, J., and Poore, R., 1996, Middle Pliocene sea surface temperatures: a global reconstruction: Marine Micropaleontology, v. 27, p. 13–35.

Dowsett, H.J., and Cronin, T., 1990, High eustatic sea level during the middle Pliocene: evidence from the Southeastern U.S. Atlantic Coastal Plain: Geology, v. 18, p. 435–438.

Dowsett, H.J., and Laubere, P., 1992, High resolution Late Pliocene sea-surface temperature record from the Northwest Atlantic Ocean: Marine Micropaleontology, v. 20, p. 91–105.

Dowsett, H., Thompson, R., Barron, J., Cronin, T., Fleming, F., Ishman, S, Poore, R., Willard, D., and Holtz, T., Jr., 1994, Joint investigations of the Middle Pliocene climate I: PRISM paleoenvironmental reconstructions: Global and Planetary Change, v. 9, p. 169–195.

Drever, J.I., 1973, The preparation of oriented clay mineral specimens for X-ray diffraction analysis by a filter-membrane peel technique: American Mineralogist, v. 58, p. 553–554.

Drever, J.I., 1982, The Geochemistry of Natural Waters: Prentice-Hall, Englewood Cliffs, New Jersey, 388 p.

DuBar, J.R., 1974, Summary of the Neogene stratigraphy of southern Florida, in Oaks, R.Q., Jr., and DuBar, J.R., eds., Post-Miocene Stratigraphy of the Central and Southern Atlantic Coastal Plain: Utah State University Press, Logan, Utah, 206 p.

Dypvik, H., and Brunfelt, A.O., 1979, Distribution of rare earth elements in some North Atlantic Kimmeridgian black shales: Nature, v. 278, p. 339–341.

Eargle P.H., 1955, Stratigraphy of the outcropping Cretaceous rocks of Georgia: U.S. Geological Survey, Bulletin 1014, 101 p.

Eberl, D.D., Drits, V.A., and Środoń, J., 1998, Deducing growth mechanisms for minerals from the shapes of crystal size distributions: American Journal of Science, v. 298, p. 499–533.

Eberl, D.D., Drits, V.A., Środoń, J., and Nüesch, R., 1996, MudMaster: a program for calculating crystallite size distributions and strain from the shapes of x-ray diffraction peaks: U.S. Geological Survey, Open-File Report 96-171, 46 p.

Eberli, G.P., 2000, The record of Neogene sea-level in the prograding carbonates along the Bahamas transect – Leg 166 synthesis, in Swart, P.K., Eberli, G.P., Malone, M.J., and Sarg, J.F., eds., Proceedings of the Ocean drilling Program, Scientific Results, v. 166, p. 167–177.

288

Eberli, G.P., Swart, P.K., Malone, M., and Leg 166 Shipboard Party, 1997, Proceedings of the Ocean Drilling Program, Initial Reports 166: College Station, Texas, Ocean Drilling Program.

Edgar, N.T., and Cecil, C.B., 2003, Influence of climate on deep-water clastic sedimentation: application of a modern model, Peru-Chile Trough, to an ancient system, Ouachita Trough, in Cecil, C.B., and Edgar, N.T., eds., Climate Controls on Stratigraphy: SEPM, Special Publication 77, p. 185–191.

Fedorov, A.V., Dekens, P.S., McCarthy, M., Ravelo, A.C., deMenocal, P.B., Barreiro, M., Pacanowski, R.C., and Philander, S.G., 2006, The Pliocene paradox (mechanisms for a permanent El Niño: Science, v. 312, p. 1485–1489.

Fein, J.B., 1991, Experimental study of aluminum-oxalate complexing at 80°C: implications for the formation of secondary porosity within sedimentary reservoirs: Geology, v. 19, p. 1037–1040.

Fein, J.B., and Hestrin, J.E., 1994, Experimental studies of oxalate complexation at 80°C: gibbsite, amorphous silica, and quartz solubilities in oxalate-bearing fluids: Geochimica et Cosmochimica Acta, v. 58, p. 4817–4829.

Feng, R, and Kerrich, R., 1990, Geochemistry of fine-grained clastic sediments in the Archean Abitibi greenstone belt, Canada: Implications for provenance and tectonic setting: Geochimica et Cosmochimica Acta, v. 54, p. 1061–1081.

Flicoteaux, R., and Lucas, J., 1984, Weathering of phosphate minerals, in Nriagu, J.O., and Moore, P.B., eds., Phosphate Minerals, Springer-Verlag, New York, p. 292–317.

Frey, R.W., 1970, Environmental significance of recent marine lebenspuren near Beaufort, North Carolina: Journal of Paleontology, v. 44, p. 507–519.

Frost, C.D., and O'Nions, R.K., 1984, Nd evidence for Proterozoic crustal development in the Belt-Purcell Supergroup: Nature, v. 312, p. 53–56.

Frost, C.D., and Winston, D., 1987, Nd isotope systematics of coarse- and fine-grained sediments: examples from the middle Proterozoic Belt-Purcell Supergroup: Journal of Geology, v. 95, p. 309–327.

Fountain, K.B., and McClellan, G.H., 1993, Origin of northcentral Florida kaolinite deposits in light of mineralogical and geochemical evidence (abstract): 1993 Annual Meeting of the Clay Minerals Society, San Diego, California, p. 132.

Fullagar, P.D., Goldberg, S.A., and Butler, J.R., 1997, Nd and Sr isotopic characterization of crystalline rocks from the southern Appalachian Piedmont and Blue Ridge, North and South Carolina, in Sinha, A.K., Whalen, J.B., and Hogan, J.P., eds., The Nature of Magmatism in the Appalachian Orogen: Geological Society of America Memoir 191, p. 165–179.

289

Galán E., Aparicio, P., González, I., and Miras, A., 1998, Contribution of multivariate analysis to the correlation of some properties of kaolin with its mineralogical and chemical composition: Clay Minerals, v. 33, p. 65–75.

Ghosh, D.K., and Lambert, R.StJ., 1989, Nd-Sr isotopic study of Proterozoic to Triassic sediments from southeastern British Columbia: Earth and Planetary Science Letters, v. 94, p. 29–44.

Ginsburg, R.N., Brown, K.M., and Chung, G.S., 1989, Siliciclastic foundations of South Florida’s Quaternary carbonates (abstract): Geological Society of America, Programs with Abstracts, v. 21, p. A-290.

Gladney, E.S., Jones, E.A., Nickell, E.J., and Roelandts, I., 1990, 1988 Compilation of elemental concentration data for USGS Basalt BCR-1: Geostandards Newsletter, v. 14, p. 209–359.

Gleason, J.D., Patchett, P.J., Dickinson, W.R., and Ruiz, J., 1994, Nd isotopes link Ouachita turbidites to Appalachian sources: Geology, v. 22, p. 347–350.

Gleason, J.D., Patchett, P.J., Dickinson, W.R., and Ruiz J., 1995, Nd isotopic constraints on sediment sources of the Ouachita-Marathon fold belt: Geological Society of America, Bulletin, v. 107, p. 1192–1210.

Gohn G.S., Bybell L.M., Christopher R.A., Owens J.P., and Smith C.C., 1979, Cretaceous and Paleogene stratigraphic framework of the South Carolina and Georgia coastal margins (abstract): Programs and Abstracts for the 2nd Symposium on the Geology of the Southeastern Coastal Plain, Americus, Georgia, p. 11.

Goodbred, S.L., Jr., and Kuehl, S.A., 2000, Enormous Ganges-Brahmaputra sediment discharge during strengthened early Holocene monsoon: Geology, v. 28, p. 1083–1086.

Guertin, L.A., McNeill, D.F., Lidz, B.H., and Cunningham, K.J., 1999, Chronologic model and transgressive-regressive signatures in the late Neogene siliciclastic foundation (Long Key Formation) of the Florida Keys: Journal of Sedimentary Research, v. 69, p. 653–666.

Hancock, G., and Kirwan, M., 2007, Summit erosion rates deduced from 10Be: implications for relief production in the central Appalachians: Geology, v. 35, p. 89–92.

Hansen, B.C.S., Grimm, E.C., and Watts, W.A., 2001, Palynology of the Peace Creek site, Polk County, Florida: Geological Society of America, Bulletin, v. 113, p. 682–692.

Haq, B.U., Hardenbol, J., and Vail, P.R., 1988, Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level change, in Wilgus, C.K., Hastings, B.S., Kendall, C.G.St.C., Posamentier, H.W., Ross, C.A., and Van Wagoner, J.C., eds., Sea Level Changes—An Integrated Approach: SEPM, Special Publication 42, p. 72–108.

Harris, S.E., and Mix, A.C., 2002, Climate and tectonic influences on continental erosion of tropical South America, 0-13 Ma: Geology, v. 30, p. 447–450.

290

Harris, W.G., Morrone, A.A., and Coleman, S.E., 1992, Occluded mica in hydroxyl-interlayered vermiculite grains from a highly-weathered soil: Clays and Clay Minerals, v. 40, p. 32–39.

Hassanipak, A.A., and Eslinger, E.V., 1985, Mineralogy, crystallinity, 18O/16O, and D/H of Georgia kaolins: Clays and Clay Minerals, v. 33, p. 99–106

Hathaway J.C., 1956, Procedure for clay mineral analysis used in the sedimentary petrology laboratory of the USGS: Clay Minerals Bulletin, v. 3, p. 8–13.

Haywood, A.M., Sellwood, B.W., and Valdes, P.J., 2000, Regional warming: Pliocene (3 Ma) paleoclimate of Europe and the Mediterranean: Geology, v. 28, p. 1063–1066.

Healy, H.G., 1975, Terraces and shorelines of Florida: Florida Geological Survey, Map Series 71.

Hem, J.D., 1985, Study and interpretation of the chemical characteristics of natural water, 3rd Edition: U.S. Geological Survey, Water-Supply Paper 2254, 263 p.

Henderson, P., Marlow, C.A., Molleson, T.I., and Williams, C.T., 1983, Patterns of chemical change during bone fossilization: Nature, v. 306, p. 358–360.

Hendrix, W.P., and Orr, C., 1972, Automatic sedimentation size analysis instrument, in Groves, M.J., and Wyatt-Sargent, J.L., eds., Particle Size Analysis 1970: Society of Analytical Chemistry, London, p. 133–146.

Hendry, C.W., Jr., and Yon, J.W., Jr., 1967, Stratigraphy of Upper Miocene Miccosukee Formation, Jefferson and Leon counties, Florida: American Association of Petroleum Geologists, Bulletin, v. 51, p. 250–256.

Herrick S.M., 1961, Well logs of the coastal plain of Georgia. Georgia Geological Survey, Bulletin 70, 461 p.

Herrick, S.M., and Vorhis, R.C., 1963, Subsurface geology of the Georgia Coastal Plain: Georgia Geological Survey, Information Circular 25, 78 p.

Heuberger, D., 1995, The influence of post-Miocene sediments on soil formation in Big Scrub, Ocala National Forest [unpublished M.S. thesis]: University of Florida, Gainesville, Florida, 195 p.

Hill, G.W., 1985, Ichnofacies of a modern size-graded shelf, northwestern Gulf of Mexico, in Curran, H.A., ed., Biogenic Structures: Their use in Interpreting Depositional Environments: SEPM, Special Publication 35, Tulsa, Oklahoma, p. 195–210.

Hinkley, D.N., 1961, Mineralogical and chemical variations in the kaolin deposits of the coastal plain of Georgia and South Carolina: N.S.F. Technical Report, Pennsylvania State University, University Park, Pennsylvania, 180 p.

291

Hinkley, D.N., 1963, Variability in “crystallinity” values among the kaolin deposits of the coastal plain of Georgia and South Carolina: Clays and Clay Minerals, v. 13, p. 229–235.

Hobgood, J.S., and Cerveny, R.S., 1988, Ice-age hurricanes and tropical storms: Nature, v. 333, p. 243–245.

Horton, J.W., Jr., Drake, A.A., Jr., and Rankin, D.W., 1989, Techtonostratigraphic terranes and their Paleozoic boundaries in the central and southern Appalachians: Geological Society of America Special Paper 230, p. 213–245.

Howard, J.D., and Reineck, H.E., 1972, Georgia coastal region, Sapelo Island, U.S.A.: sedimentology and biology. IV. Physical and biogenic sedimentary structures of the nearshore shelf: Senckenbergiana Maritima, v. 4, p. 81–123.

Hoyt, J.H., and Hails, J.R., 1974, Pleistocene stratigraphy of southeastern Georgia, in Oaks, R.Q., Jr., and DuBar, J.R., eds., Post-Miocene Stratigraphy Central and Southern Atlantic Coastal Plain: Utah State University Press, Logan, Utah, p. 191–205.

Huddlestun, P.F., 1988, A revision of the lithostratigraphic units of the coastal plain of Georgia: The Miocene through Holocene: Georgia Geological Survey, Bulletin 104, 162 p.

Huddlestun, P.F., 1993, An overview of the Neogene of Georgia, in Zullo, V.A., Harris, W.B., Scott, T.M., and Portell, R.W., eds., The Neogene of Florida and Adjacent Regions, Proceedings of the Third Bald Head Island Conference on Coastal Plain Geology: Florida Geological Survey, Special Publication 37, p. 91–96.

Hurst, V.J., and Pickering, S.M., Jr., 1989, Significance of crystalline size and habit in determining origin and industrial applications of Georgia kaolins: Geological Society Southeast Section, Abstracts with Programs, p. 22.

Hurst, V.J., and Pickering, S.M., Jr., 1997, Origin and classification of coastal plain kaolins, southeastern USA, and the role of groundwater and microbial action: Clays and Clay Minerals, v. 45, p. 274–285.

Hurst, V.J., and Rigsby, W.E., 1984, Intergrowths and aggregates of natural kaolinite, in Bailey, G.W., ed., Proceedings of the 42nd Annual meeting of the Electron Microscopy Society of America, p. 22–23.

Isphording, W.C., 1970, Late Tertiary paleoclimate of eastern United States: American Association of Petroleum Geologists, Bulletin, v. 54, p. 334–343.

Isphording, W.C., 1971, Provenance and petrography of Gulf Coast Miocene sediments, in Geological Review of Some North Florida Mineral Resources: Southeastern Geological Society, Fifteenth Field Conference Guidebook, p. 43–54.

Jacobsen, S.B., and Wasserberg, G.J., 1980, Sm – Nd isotopic evolution of chondrites: Earth and Planetary Science Letters, v. 41, p. 139–155.

292

Jeong, G.Y., 1998a, Formation of vermicular kaolinite from halloysite aggregates in the weathering of plagioclase: Clays and Clay Minerals, v. 46, p 270–279.

Jeong, G.Y., 1998b, Vermicular kaolinite epitactic on primary phyllosilicates in the weathering profiles of anorthosite: Clays and Clay Minerals, v. 46, p. 509–520.

Jiang, W-.T., and Peacor, D.R., 1991, Transmission electron microscopic study of the kaolinization of muscovite: Clays and Clay Minerals, v. 39, 1–13.

Jiang, D., Wang, H., Ding, Z., Lang, X., and Drange, H., 2005, Modeling the middle Pliocene climate with a global atmospheric general circulation model: Journal of Geophysical Research, v. 110, p 14.

Johnson, R.A., 1989, Geologic descriptions of selected exposures in Florida: Florida Geological Survey, Special Publication 30, 175 p.

Joussein, E., Petit, S., Churchman, J., Theng, B., Righi, D., and Delvaux, B., 2005, Halloysite clay minerals – a review: Clay Minerals, v. 40, p. 383–426.

Kane B.C., 1984, Origin of the Grandin (Plio-Pleistocene) sands, western Putnam County, Florida [unpublished M.S. thesis]: University of Florida, Gainesville, Florida, 85 p.

Keller W.D., 1977, Scan electron micrographs of kaolins collected from diverse environments of origin-IV Georgia kaolins and kaolinizing source rocks: Clays and Clay Minerals, v. 25, p. 311–346.

Keller W.D., 1978, Classification of kaolins exemplified by their textures in scanning electron micrographs: Clays and Clay Minerals, v. 26, p. 1–20.

Kennett, J.P., and Hodell, D.A., 1993, Evidence for relative climatic stability of Antarctica during the early Pliocene: a marine perspective: Geografiska Annaler., v.75A, p. 205–220.

Kennett, J.P., and Hodell, D.A., 1995, Stability or instability of Antarctic ice sheets during warm climates of the Pliocene?: GSA Today, v. 5, p. 10–13.

Kesler T.L., 1956, Environment and origin of the Cretaceous kaolin deposits of Georgia and South Carolina: Economic Geology, v. 51, p. 541–554.

Ketner, K.B., and McGreevy, L.J., 1959, Stratigraphy of the area between Hernando and Hardee counties, Florida: U.S. Geological Survey, Bulletin 1074-C, p. 49–124.

Klein, H.B., Schroeder, M.C., and Lichtler, W.F., 1964, Geology and ground water resources of Glades and Hendry Counties: Florida Geological Survey, Report of Investigations 37, 132 p.

Kozuch, M., 1994, Age, isotopic, and geochemical characterization of the Carolina slate and Charlotte belts: Implications for stratigraphy and petrogenesis [unpublished M.S. thesis]: University of Florida, Gainesville, Florida, 115 p.

293

Kristóf, E., Juhász, A.Z., and Vassanyi, I., 1993, The effect of mechanical treatment on the crystal structure and thermal behavior of kaolinite, v. 41, p. 608–612.

Kussel, C.M., and Jones, D.S., 1986, Depositional history of three Pleistocene bluffs in northeastern Florida: Florida Scientist, v. 49, p. 242–254.

Ladd G.E., 1898, A preliminary report on a part of the clays of Georgia: Georgia Geological Survey, Bulletin 6A, 204 p.

Land Boundary Information System [doqq_04_utm.cfm], 2004, Tallahassee, Florida: Florida Department of Environmental Protection. Available FTP: http://data.labins.org/2003/MappingData/DOQQ/doqq_04_utm.cfm [October 1, 2007].

Lev, S.M., McLennan, S.M., Meyers, W.J., and Hanson, G.N., 1998, A petrographic approach for evaluating trace-element mobility in black shale: Journal of Sedimentary Research, v. 68, p. 970–980.

Lichte, F.E., Golightly, D.W., and Lamothe, P.J., 1987, Inductively coupled plasma-atomic emission spectrometry, in Baedecker, P.A., ed., Methods for Geochemical Analysis, U.S. Geological Survey, Report B1770, Reston, Virginia, p. B1–B10.

Liètard, O., 1977, Contribution à l’étude des propiétés phisicochimiques, cristallographiques et morphologiques des kaolins [unpublished Ph.D. thesis]: Nancy, France, 345 p.

Linn, A.M., DePaolo, D.J., and Ingersoll, R.V., 1991, Nd-Sr isotopic provenance analysis of Upper Cretaceous Great Valley fore-arc sandstones: Geology, v. 19, p. 803–806.

Lowe, R.A., 1991, Microtextural and mineralogical differences of Georgia’s kaolins and the search for Ostwald ripening [unpublished M.S. thesis]: University of Georgia, Athens, Georgia, 81 p.

Luais, B., Telouk, P., and Albarede, F., 1997, Precise and accurate neodymium isotopic measurements by plasma-source mass spectrometry: Geochimica et Cosmochimica Acta, v. 61, p. 4847–4854.

Lugmair, G.W., and Carlson, R.W., 1978, The Sm-Nd history of KREEP: Proceedings 9th Lunar and Planetary Science Conference, p. 689–704.

Madsen, I.C., and Hill, R.J., 1988, Effect of divergence and receiving slit dimensions on peak profile parameters in Rietveld analysis of X-ray diffractometer data: Journal of Applied Crystallography, v. 21, p. 398–405.

Maliva, R.G., Dickson, J.A.D., and Fallick, A.E., 1999, Kaolin cements in limestones: potential indicators or organic-rich pore waters during diagenesis: Journal of Sedimentary Research, v. 69, p. 158–163.

Mansfield, W.C., 1939, Notes on the upper Tertiary and Pleistocene mollusks of peninsular Florida: Florida Geological Survey, Bulletin 18, 75 p.

294

Martins, J.H.C., 1928, Sand and gravel deposits of Florida. Florida Geological Survey, 19th Annual Report, p. 33–123.

Matmon, A., Bierman, P.R., Larsen, J., Southworth, S., Pavich, M.J., and Caffe, M.W., 2003, Temporally and spatially uniform rates of erosion in the southern Appalachian Great Smoky Mountains: Geology, v. 31, p. 155–158.

Matson, G.C., and Clapp, F.G., 1909, A preliminary report on the geology of Florida with special reference to the stratigraphy: Florida Geological Survey, Annual Report 2, p. 25–173.

Matson, G.C., 1916, The Pliocene Citronelle formation of the Gulf Coastal Plain: U.S. Geological Survey, Professional Paper 98, p. 167–192.

McDaniel, D.K., Hemming, S.R., McLennan, S.M., and Hanson, G.N., 1994, Resetting of neodymium isotopes and redistribution of REEs during sedimentary processes: the Early Proterozoic Chelmsford Formation, Sudbury Basin, Ontario, Canada: Geochimica et Cosmochimica Acta, v. 58, p. 931–941.

McDougall, K., 1994, Late Cenozoic benthic foraminifers of HLA Borehole Series, Beaufort Sea Shelf, Alaska: U.S. Geological Survey, Bulletin 2055, 100 p.

McClellan, G.H., and Van Kauwenbergh, S.J., 1990, Clay mineralogy of the phosphorites of the southeastern United States, in Burnett, W.C., and Riggs, S.R., eds., Phosphate Deposits of the World, v. 3: Neogene to Modern phosphorites, p. 337–351.

McLennan S.M., 1989, Rare earth elements in sedimentary rocks: influence of provenance and sedimentary processes, in Lipin, B.R., and McKay, G.A., eds., Geochemistry and Mineralogy of Rare Earth Elements: Mineralogical Society of America Reviews in Mineralogy, v.21, p. 169–200.

McLennan, S.M., Taylor, S.R., McCulloch, M.T., and Maynard, J.B., 1990, Geochemical and Nd-Sr isotopic composition of deep-sea turbidites: crustal evolution and plate tectonic associations: Geochimica et Cosmochimica Acta, v. 54, p. 2015–2050.

McLennan, S.M., Hemming, S., McDaniel, D.K., and Hanson, G.N., 1993, Geochemical approaches to sedimentation, provenance, and tectonics, in Johnsson, M.J., and Basu, A., eds., Processes Controlling the Composition of Clastic Sediments: Geological Society of America, Special Paper 284, p. 21–40.

McNeill, D.F., Eberli, G.P., Lidz, B.H., Swart, P.K., and Kenter, J.A.M., 2001, Chronostratigraphy of prograding carbonate platform margins: a record of dynamic slope sedimentation, western Great Bahama Bank, in Ginsburg, R.N., ed., Subsurface Geology of a Prograding Carbonate Platform Margin, Great Bahama Bank: Results of the Bahamas Drilling Project: SEPM, Special Publication 70, p. 101–136.

McSween, H.Y., Jr., Speer, A., and Fullagar, P.D., 1991, Plutonic rocks, in Horton, J.W., Jr., and Zullo, V.A., eds., Geology of the Carolinas: Carolina Geological Society, p. 109–126.

295

Merino, E., Harvey, C., and Murray, H.H., 1989, Aqueous-chemical control of the tetrahedral- aluminum content of quartz, halloysite, and other low-temperature silicates: Clays and Clay Minerals, v. 37, p. 135–142.

Miller, K.G., Kominz, M.A., Browning, J.V., Wright, J.D., Mountain, G.S., Katz, M.E., Sugarman, P.J., Cramer, B.S., Christie-Blick, N., and Pekar, S.F., 2005, The Phanerozoic record of global sea-level change: Science, v. 310, p. 1293–1298.

Miller, R.G., and O'Nions, R.K., 1984, The provenance and crustal residence ages of British sediments in relation to palaeogeographic reconstructions: Earth and Planetary Science Letters, v. 68, p. 459–470.

Miller, R.G., and O'Nions, R.K., 1985, Source of Precambrian chemical and clastic sediments: Nature, v. 314, p. 325–330.

Mills, H.H., 2000, Apparent increasing rates of stream incision in the eastern United States during the late Cenozoic: Geology, v. 28, p. 955–957.

Missimer, T.M., 1992, Stratigraphic relationships of sediment facies within the Tamiami Formation of southwest Florida: proposed intraformational correlations, in Scott, T.M., and Allmon, W.D., eds., Plio-Pleistocene Stratigraphy and Paleontology of Southern Florida: Florida Geological Survey, Special Publication 36, p. 63–92.

Missimer, T.M., 1993, Pliocene stratigraphy of southern Florida: unresolved issues of facies correlation in time, in Zullo, V.A., Harris, W.B., Scott, T.M., and Portell, R.W., eds., The Neogene of Florida and Adjacent Regions, Proceedings of the Third Bald Head Island Conference on Coastal Plains Geology: Florida Geological Survey, Special Publication 37, p. 33–42.

Missimer, T.M., 1997, Late Oligocene to Pliocene evolution of the central portion of the South Florida Platform: Mixing of siliciclastic and carbonate sediments [unpublished Ph.D. dissertation]: University of Miami, Coral Gables, Florida, 2 volumes, 1001 p.

Missimer, T.M., 2001a, Late Neogene geology of northwestern Lee County, Florida, in Missimer, T.M., and Scott, T.M., eds., Geology and hydrology of Lee County, Florida: Florida Geological Survey, Special Publication 49, p. 21–34.

Missimer, T.M., 2001b, Late Paleogene and Neogene Chronostratigraphy of Lee County, Florida, in Missimer, T.M., and Scott, T.M., eds., Geology and hydrology of Lee County, Florida: Florida Geological Survey, Special Publication 49, p. 67–90.

Missimer, T.M., and Maliva, R.G., 2006, Late Miocene to early Pliocene fluvial transport of siliciclastic sediment onto the southern Florida Platform, in Lock, B.E., Willis, J.J., and Hammes, U., eds., Transactions - Gulf Coast Association of Geological Societies, v. 56, p. 605–612.

Molnar, P., 2001, Climate change, flooding in arid environments, and erosion rates: Geology, v. 29, p. 1071–1074.

296

Molnar, P., 2004, Late Cenozoic increase in accumulation rates of terrestrial sediment: How might climate change have affected erosion rates?: Annual Review of Earth and Planetary Sciences, v. 32, p. 67–89.

Molnar, P., and England, P., 1990, Late Cenozoic uplift of mountain ranges and global climatic change: chicken or egg?: Nature, v. 346, p. 29–34.

Montes, C.R., Melfi, A.J., Carvalho, A., Vieira-Coelho, A.C., and Formoso, M.L.L., 2002, Genesis, mineralogy and geochemistry of kaolin deposits of the Jari River, Amapá State, Brazil: Clays and Clay Minerals, v. 50, p. 494–503.

Moore, D.M., and Reynolds, R.C., Jr., 1997, X-ray diffraction and the identification and analysis of clay minerals, 2nd Ed., Oxford University Press, New York, 378 p.

Morse, J.W., and Casey, W.H., 1988, Ostwald processes and mineral paragenesis in sediments: American Journal of Science, v. 288, p. 537–560.

Mullins, H.T., Neumann, A.C., Wilber, R.J., Hine, A.C., and Chinburg, S.J., 1980, Carbonate sediment drifts in Northern Straits of Florida: American Association of Petroleum Geologists, Bulletin, v. 64, p. 1701–1717.

Murray H.H., 1976, The Georgia sedimentary kaolins, in Shimoda, S., ed., The 7th Symposium on Kaolin: Tokyo, p. 114–125.

Murray, H.H., and Keller, W.D., 1993, Kaolins, kaolins, and kaolins, in Murray, H.H., et al., eds., Kaolin Genesis and Utilization, Special Publication No. 1: The Clay Minerals Society, Boulder, Colorado, p. 1–24.

Murray, H.H., and Lyons, S.C., 1956, Correlation of paper-coating quality with degree of crystal perfection of kaolinite, in Swineford, A., ed., Clays and Clay Minerals, Proceedings of the 4th National Conference, p. 31–40.

Nagy, K.L., and Pevear, D.R., 1993, Kinetics of muscovite dissolution and kaolinite precipitation onto muscovite at 80°C and pH 3 (abstract): Annual Meeting of the Geological Society of America, Programs with Abstract, v. 25, p. A-437.

Naish, T.R., and Wilson, G.S., 2009, Constraints on the amplitude of Mid-Pliocene (3.6-2.4 Ma) eustatic sea-level fluctuations from the New Zealand shallow-marine sediment record: Philosophical Transaction of the Royal Society A: Mathematical, Physical and Engineering Sciences, v. 367, p. 169–187.

Nakamura, N., 1974, Determination of REE, Ba, Fe, Mg, Na, and K in carbonaceous and ordinary chondrites: Geochimica et Cosmochimica Acta, v. 38, p. 757–775.

Nelson, B.K., and DePaolo, D.J., 1988, Comparison of isotopic and petrographic provenance indicators in sediments from Tertiary continental basins of New Mexico: Journal of Sedimentary Petrology, v. 58, p. 348–357.

297

Nesbitt, H.W., 1979, Mobility and fractionation of rare earth elements during weathering of a granodiorite: Nature, v. 279, p. 206–210.

Nesbitt, H.W., and Markovics, G., 1997, Weathering of granodioritic crust, long-term storage of elements in weathering profiles, and petrogenesis of siliciclastic sediments: Geochimica et Cosmochimica Acta, v. 61, p. 1653–1670.

Nummedal, D., Oertel, G.F., Hubbard, D.K., and Hine, A.C., 1977, Tidal inlet variability – Cape Hatteras to Cape Canaveral: Proceedings, Coastal Sediments ’77, ASCE, Charleston, South Carolina, p. 543–562.

Nystrom P.G.Jr., Willoughby R.H., and Kite L.E., 1986, Cretaceous-Tertiary stratigraphy of the upper edge of the coastal plain between North Augusta and Lexington, South Carolina: South Carolina Geological Survey, Field Trip Guidebook, October 11-12, Columbia, South Carolina, 82 p.

Ohr, M., Halliday, A.N., and Peacor D.R., 1991, Sr and Nd isotopic evidence for punctuated clay diagenesis, Texas Gulf Coast: Earth and Planetary Science Letters, v. 105, p. 110–126.

Okada, H., and Buhry, D., 1980, Supplemental modification and introduction of code numbers to the low-latitude coccolith biostratigraphic zonation: Marine Micropaleontology, v. 5, p. 321–325.

Olssson, A.A., and Petit, R.E., 1964, Some Neogene Mollusca from Florida and the Carolinas: Bulletin of American Paleontology, v. 47, p. 509–575.

O’Nions, R.K., Hamilton, P.J., and Evenson, N.M., 1977, Variations in the 143Nd/144Nd and 87Sr/86Sr ratios in oceanic basalts: Earth and Planetary Science Letters, v. 34, p. 13–22.

Opdyke, N.D., Spangler, D.A., Smith, D.L., Jones, D.S., and Lindquist, R.C., 1984, Origin of the epeirogenic uplift of Plio-Pleistocene beach ridges in Florida and development of Florida karst: Geology, v. 12, p. 226–228.

Otvos, E.G., 1998a, Citronelle Formation, northeast Gulf Coastal Plain; stratigraphic and age issues (abstract): American Association of Petroleum Geologists, Bulletin, v. 82, p. 1789.

Otvos, E.G., 1998b, Citronelle Formation, northeastern Gulf Coastal Plain: Pliocene stratigraphic framework and age issues: Gulf Coast Association of Geological Societies Transactions, v. 48, p. 321–333.

Patterson S.H., and Murray H.H., 1984, Kaolin, refractory clay, ball clay and halloysite in North America, Hawaii, and the Caribbean region. U.S. Geological Survey, Professional Paper 1306, 56 p.

Peacock, R., 1983, The post-Eocene stratigraphy of southern Collier County, Florida: South Florida Water Management District, Technical Publication 83-5, 42 p.

298

Peck, D.M., Missimer, T.M., Slater, D.H., Wise, S.W., and O’Donnell, T.H., 1979, Late Miocene glacio-eustatic lowering of sea level: evidence from the Tamiami Formation of South Florida: Geology, v. 7, p. 285–288.

Pickering S.M.Jr., and Hurst V.J., 1989, Commercial kaolins in Georgia - occurrence, mineralogy, origin, use, in Fritz, W.J., ed., Excursions in Georgia Geology: Georgia Geological Society, Guidebook 9-1, p. 29–75.

Pickering, S.M., Jr., Hurst, V.J., and Elzea, J.M., 1997, Mineralogy , stratigraphy, and origin of Georgia kaolins: Field Excursion Guidebook, 11th International Clay Conference, Ottawa, Canada, 69 p.

Piper, D.Z., 1974, Rare earth elements in the sedimentary cycle: a summary: Chemical Geology, v. 14, p. 285–304.

Pirkle E.C., 1960, Kaolinitic sediments in peninsular Florida and origin of the kaolin. Economic Geology, v. 55, p. 1382-1405.

Pirkle, E.C., Pirkle, W.A., and Yoho, W.H., 1974, The Green Cove Springs and Boulougne heavy-mineral sand deposits of Florida: Economic Geology, v. 69, p. 1129–1137.

Pirkle E.C., and Yoho W.H., 1961, Folding or warping resulting from solution with associated joints and organic zones in clayey sands at Edgar, Florida: Florida Academy of Science, Quarterly Journal, v. 24, p. 247–266.

Pirkle E.C., Yoho W.H., and Allen A.T., 1964, Origin of the silica sand deposits of the Lake Wales Ridge area of Florida: Economic Geology, v. 59, p. 1107–1139.

Pirkle, E.C., Yoho, W.H., and Hendry, C.W., 1970, Ancient sea level stands in Florida: Florida Geological Survey, Bulletin 52, 61 p.

Pirkle, F.L., and Czel, L.J., 1983, Marine fossils from region of Trail Ridge, a Georgia-Florida landform: Southeastern Geology, v. 24, p. 31–38.

Plançon A., and Zacharie, C., 1990, An expert system for the structural characterization of kaolinites: Clays and Clay Minerals, v. 25, p. 249–260.

Poag, C.W., and Sevon, W.D., 1989, A record of Appalachian denudation in postrift Mesozoic and Cenozoic sedimentary deposits of the U.S. middle Atlantic continental margin: Geomorphology, v. 2, p. 119–157.

Puri H.S., and Vernon R.O., 1964, Summary of the geology of Florida and a guide to the classic exposures: Florida Geological Survey, Special Publication 5, 312 p.

Rasmussen, B., 1996, Early-diagenetic REE-phosphate minerals (florencite, gorceixite, crandallite, and xenotime) in marine sandstones: a major sink for oceanic phosphorous: American Journal of Science, v. 296, p. 601–632.

299

Rasmussen, B., Buick, R., and Taylor, W.R., 1998, Removal of oceanic REE by authigenic precipitation of phosphatic minerals: Earth and Planetary Science Letters, v. 164, p. 135– 149.

Ravelo, A.C., and Wara, M.W., 2004, The role of the tropical oceans on global climate during a warm period and a major climate transition: Oceanography¸ v. 17, p. 32–41.

Rea, D.K., Zachos, J.C., Owen, R.M., and Gingerich, P.D., 1990, Global change at the Paleocene-Eocene boundary: Climatic and evolutionary consequences of tectonic events. Palaeogeography Palaeoclimatology Palaeoecology, v. 79, p. 117–128.

Reynolds, W.R., 1962, The lithostratigraphy and clay mineralogy of the Tampa-Hawthorn sequence of peninsular Florida [unpublished M.S. thesis]: Florida State University, Tallahassee, Florida, 126 p.

Robertson, I.D.M., and Eggleston, R.A., 1991, Weathering of granites muscovite to kaolinite and halloysite and of plagioclase-derived kaolinite to halloysite: Clays and Clay Minerals, v. 39, p. 113–126.

Ruddiman W., Sarnthein, M., et al., 1989, Late Miocene to Pleistocene evolution of climate in Africa and the low-latitude Atlantic: overview of Leg 108 results: Proceedings ODP Scientific Results, v. 108, p. 463–484.

Rupert, F.R., 1990, Geology of Gadsden County, Florida: Florida Geological Survey, Bulletin 62, 61 p.

Samson, S.D., Coler, D.G., and Speer, J.A., 1995a, Geochemical and Nd-Sr-Pb isotopic composition of Alleghanian granites of the southern Appalachians: Origin, tectonic setting, and source characterization: Earth and Planetary Science Letters, v. 134, p. 359–376.

Samson, S.D., Hibbard, J.P., and Wortman, G.L., 1995b, Nd isotopic evidence for juvenile crust in the Carolina terrane, southern Appalachians: Contributions to Mineralogy and Petrology, v. 121, p. 171–184.

Scher, H.D., and Martin, E.E., 2004, Circulation in the Southern Ocean during the Paleogene inferred from Nd isotopes: Earth and Planetary Science Letters, v. 228, p. 391–405.

Schmidt, W., 1997, Geomorphology and physiography of Florida, in Randazzo, A.F., and Jones, D.S., eds., The Geology of Florida: University Press of Florida, Gainesville, p. 1–12.

Scott, T.M., 1980, The sand and gravel resources of Florida: Florida Geological Survey, Report of Investigations 90, 41 p.

Scott, T.M., 1988a, The Cypresshead Formation in northern peninsular Florida, in Pirkle, F.L., and Reynolds, J.G., eds., Southeastern Geological Society, Annual Field Trip Guidebook, Feb. 19-29, p. 70–72.

300

Scott, T.M., 1988b, The lithostratigraphy of the Hawthorn Group (Miocene) of Florida: Florida Geological Survey, Bulletin 59, 148 p.

Scott, T.M., 1992a, A geological overview of Florida. Florida Geological Survey, Open-File Report 50, 78 p.

Scott, T.M., 1992b, Coastal Plains stratigraphy: The dichotomy of biostratigraphy and lithostratigraphy – A philosophical approach to an old problem, in Scott, T.M., and Allmon, W.D., eds., Plio-Pleistocene stratigraphy and paleontology of southern Florida: Florida Geological Survey, Special Publication 36, p. 21–25.

Scott, T.M., 1997, Miocene to Holocene history of Florida, in Randazzo, A.F., and Jones, D.S., eds., The Geology of Florida: University Press of Florida, Gainesville, p. 57–67.

Scott, T.M., 2001, Text to accompany the geologic map of Florida: Florida Geological Survey, Open-File Report 80, 29 p.

Scott, T.M., Campbell, K.M., Rupert, F.R., Arthur, J.D., Missimer, T.M., Lloyd, J.M., Yon, J.W., and Duncan, J.G., 2001, Geologic map of the state of Florida: Florida Geological Survey, Map Series 146.

Scott, T.M., Hoestine, R.W., Knapp, M.S., Lane, E., Ogden, G.M., Jr., Deuerling, R., and Neel, H.E., 1980, The sand and gravel resources of Florida. Florida Geological Survey, Report of Investigation 90, 41 p.

Scott, T.M., and Wingard, G.L., 1995, Facies, fossils, and time – A discussion of the litho- and biostratigraphic problems in the Plio-Pleistocene sediments in southern Florida, in Scott, T.M., ed., Stratigraphy and paleontology of the Plio-Pleistocene shell beds, southwest Florida: Southeastern Geological Society, Guidebook 35.

Seilacher, A., 1967, Bathymetry of trace fossils. Marine Geology, v. 5, p. 413–428.

Sellards E.H., 1912, The soils and other surface residual materials of Florida. Florida Geological Survey, 4th Annual Report, p. 1–79.

Sellards, E.H., 1914, The relation between the Dunnellon formation and the Alachua clays of Florida: Florida Geological Survey, 6th Annual Report, p. 161–162.

Sellards, E.H., 1918, Geology between the Apalachicola and Ochlocknee Rivers in Florida: Florida Geological Survey, 10th Annual Report, p. 9–56.

Sen, A., Kendall, C.G.St.C., and Levine, P., 1999, Combining a computer simulation and eustatic events to date seismic sequence boundaries: a case study of the Neogene of the Bahamas: Sedimentary Geology, v. 125, p. 47–59.

Simić, V., and Uhlík, P., 2006, Crystallite size distribution of clay minerals from selected Serbian clay deposits: Annales Géologiques de la Péninsule Balkanique, v. 67, p. 109–116.

301

Singh, B., and Gilkes, R.J., 1992, an electron optical investigation of the alteration of kaolinite to halloysite: Clays and Clay Minerals, v. 40, p. 212–229.

Small, J.S., and Manning, D.A.C., 1994, On-line monitoring of clay mineral precipitation in sandstone pore space under flow conditions: Mineralogical Magazine, v. 58A, p. 852–853.

Smith R.W., 1929, Sedimentary kaolins of the coastal plain of Georgia. Georgia Geological Survey, Bulletin 44, 48 p.

Sommerfield, C.K., and Wheatcroft, R.A., 2007, Late Holocene sediment accumulation on the northern California shelf: Oceanic, fluvial, and anthropogenic influences: Geological Society of America, Bulletin, v. 119, p. 1120–1134.

Sonke, J.E., and Salters, V.J.M., 2006, Lanthanide-humic substances complexation. I. Experimental evidence for a lanthanide contraction effect: Geochimica et Cosmochimica Acta, v. 70, p. 1495–1506.

Stein, R., 1985, Rapid grain-size analyses of clay and silt fraction by SediGraph 5000D: comparison with Coulter Counter and Atterberg methods: Journal of Sedimentary Petrology, v. 55, p. 590–593.

Stein, R., and Sarnthein, M., 1984, Late Neogene events of atmospheric and oceanic circulation offshore northwest Africa: high resolution record from deep-sea sediments, in Coetzee, J.A. and van Zinderen Bakker, E.M., eds., Palaeoecology of Africa, v. 16: Balkema, Rotterdam, p. 9–36.

Stewart, K.G., and Dennison, J.M., 2006, Tertiary-to-Recent arching and the age and origin of fracture-controlled lineaments in the Southern Appalachians (abstract): Abstracts with Programs, 55th Annual Meeting, Southeastern Section of the Geological Society of America, v. 38, p. 27.

Stuckey, J.L., 1965, North Carolina: Its geology and mineral resources: North Carolina Department of Conservation and Development, 550 p.

Stull, R.T., and Boles, G.A., 1926, Beneficiation and utilization of Georgia clays. U.S. Bureau of Mines, Bulletin 252, 72 p.

Stumm, W., 1992, Chemistry of the Solid-Water Interface: Wiley & Sons, New York, 346 p.

Surdam, R.C., Crossey, L.J., and Boese, S.W., 1984, The chemistry of secondary porosity, in McDonald, D.A., and Surdam, R.C., eds., Clastic Diagenesis: American Association of Petroleum Geologists, Memoir 37, p. 127–150.

Syvitski, J.P.M., 2004, The influence of climate on the sediment load of rivers (abstract): Annual Meeting Expanded Abstracts, American Association of Petroleum Geologists Annual Meeting, v. 13, p. 135.

302

Tabbutt, K.D., 1990, Provenance of some Mesozoic terrigenous clastic strata in the western interior of North America: a mineralogical and geochemical approach [unpublished Ph.D. dissertation], Dartmouth College, 162 p.

Tanaka, T., Togashi, S., Kamioka, H., Amakawa, H., Kagami, H., Hamamoto, T., Yuhara, M., Orihashi, Y., Yoneda, S., Shimizu, H., Kunimaru, T., Takahashi, K., Yanagi, T., Nakano, T., Fujimaki, H., Shinjo, R., Asahara, Y., Tanimizu, M., and Dragusanu, C., 2000, JNdi-1: a neodymium isotopic reference in consistency with LaJolla neodymium: Chemical Geology, v. 168, p. 279–281.

Taylor S.R., and McLennan S.M., 1985, The Continental Crust: Its Composition and Evolution: Blackwell Scientific Publications, Oxford, 312 p.

Tedford, T.H., 1985, Late Miocene turnover of the Australian mammal fauna: South Africa Journal of Science, v. 81, p. 262–263.

Toulkeridis, T., Goldstein, S.L., Clauer, N., Kröner, A., and Lowe D.R., 1994, Sm-Nd dating of Fig Tree clay minerals of the Barberton greenstone belt, South Africa: Geology, v. 22, p. 199–202.

van Zinderen Bakker, E.M., and Mercer, J.H., 1986, Major late Cenozoic climatic events and palaeoenvironmental changes in Africa viewed in a world wide context: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 56, p. 217–235.

Vasconcelos, P.M., Becker, T.A., Renne, P.R., and Brimhall, G.H., 1994a, Direct dating of weathering phenomena by 40Ar/39Ar and K-Ar analysis of supergene K-Mn oxides. Geochimica et Cosmochimica Acta, v. 58, p. 1635–1665.

Vasconcelos, P.M., Brimhall, G.H., Becker, T.A., and Renne, P.R., 1994b, 40Ar/39Ar analysis of supergene jarosite and alunite: implications to the paleoweathering history of the western USA and West Africa: Geochimica et Cosmochimica Acta, v. 58, p. 401–420.

Veatch O., 1909, Second report on the clay deposits of Georgia. Georgia Geological Survey, Bulletin 18, p. 92–104.

Veatch, O., and Stephenson, L.W., 1911, Preliminary report on the geology of the Coastal Plain of Georgia: Georgia Geological Survey, Bulletin 26, 466 p.

Velho, J.A., and Gomes, C., 1991, Characterization of Portuguese kaolins for the paper industry: beneficiation through new delamination techniques: Applied Clay Science, v. 6, p. 155– 170.

Wade, J.A., 2002, Element mobility and secondary mineral formation during the early stages of alteration in rocks from the Tecuamburro volcanic complex, southeast Guatemala: Ph.D. Dissertation, Michigan State University.

Walton, T.L., 1976, Littoral drift estimates along the coastline of Florida: Florida Sea Grant Program, Report 13, 41 p.

303

Wan, Y., and Liu, C., 2006, The effect of humic acid on the adsorption of REEs on kaolin: Colloids and Surfaces A: Physicochemical and Engineering Aspects, v. 290, p. 112–117.

Wang, H., 1994, Step size, scanning speed and shape of X-ray diffraction peak: Journal of Applied Crystallography, v. 27, p. 716–721.

Warme, J.E., 1971, Paleoecological aspects of a modern coastal lagoon: University of California Publication in Geological Sciences, v. 87, p. 1–131.

Warren, B.E., and Averbach, B.L., 1953, The effect of cold-work distortion on X-ray patterns: Journal of Applied Physics, v. 21, p. 595–599.

Warzeski, E.R., Cunningham, K.J., Ginsburg, R.N., Anderson, J.B., and Ding, Z.-D., 1996, A Neogene mixed siliciclastic and carbonate foundation for the Quaternary carbonate shelf, Florida Keys: Journal of Sedimentary Research, v. 66, p. 788–800.

Webb, S.D., and Crissinger, D.B., 1983, Stratigraphy and vertebrate paleontology of the central and southern Phosphate District of Florida, in Central Florida Phosphate District: Geological Society of America, Southeast Section Field Trip Guidebook, p. 28–72.

Wedepohl, K.H., 1995, The composition of the continental crust: Geochimica et Cosmochimica Acta, v. 59, p. 1217–1232.

Weedman, S.D., Scott, T.M., Edwards, L.E., Brewster-Wingard, G.L., and Libarkin, J.C., 1995, Preliminary analysis of integrated stratigraphic data from the Phred #1 corehole, Indian River County, Florida: U.S. Geological Survey, Open-File Report 95-824, 63 p.

Weimer, R.J., and Hoyt, J.H., 1964, Burrows of Callianassa major Say, geological indicators of littoral and shallow neritic environments. Journal of Paleontology, v. 38, p. 761–767.

White, W.A., 1958, Some geomorphic features of central peninsular Florida: Florida Geological Survey, Bulletin 41, 92 p.

White, W.A., 1970, The geomorphology of the Florida peninsula: Florida Geological Survey, Bulletin 51, 164 p.

Wildeman, T.R., and Condie, K.C., 1973, Rare earths in Archean graywackes from Wyoming and from the Fig Tree Group, South Africa: Geochimica Cosmochimica et Acta, v. 37, p. 439–453.

Willard, D.A., 1994, Palynological record from the North Atlantic region at 3 Ma: vegetational distribution during a period of global warmth: Review of Palaeobotany and Palynology, v. 83, p. 275–297.

Willard, D.A., Cronin, T.M., Ishman, S.E., and Litwin, R.J., 1993, Terrestrial and marine records of climatic and environmental changes during the Pliocene in subtropical Florida: Geology, v. 21, p. 679–682.

304

Wingard, G.L., Weedman, S.D., Scott, T.M., Edwards, L.E., and Green, R.C., 1994, Preliminary analysis of integrated stratigraphic data from the South Venice corehole, Sarasota County, Florida: U.S. Geological Survey, Open-File Report 95-3, 129 p.

Winker, C.D., and Howard, J.D., 1977a, Correlation of tectonically deformed shorelines on the southern Atlantic coastal plain: Geology, v. 5, p. 123–127.

Winker, C.D., and Howard, J.D., 1977b, Plio-Pleistocene paleogeography of the Florida Gulf Coast interpreted from relict shorelines: Gulf Coast Association of Geological Societies, Transactions, v. 27, p. 409–420.

Wolfe, J.A., 1985, Distribution of major vegetation types during the Tertiary, in The Carbon Cycle and Atmospheric CO2: Natural Variations Archaen to Present, American Geophysical Union, p. 357–375.

Wornardt, W.W., Jr., Shaffer, B., and Vail, P.R., 2001, Revision of the Late Miocene, Pliocene and Pleistocene sequence cycles (abstract): American Association of Petroleum Geologists, Bulletin, v. 85, p. 1710.

Yemane, K., Bonnefille, R., and Faure, H., 1985, Palaeoclimatic and tectonic implications of Neogene microflora from the northwestern Ethiopian highlands: Nature, v. 318, p. 653– 656.

Yomane, K., Bonnefille, R., and Faure, H., 1985, Paleoclimatic and tectonic implications of Neogene microflora from the northwestern Ethiopian Highlands: Nature, v. 318, p. 653– 656.

Yon, J.W., Jr., 1966, Geology of Jefferson County, Florida: Florida Geological Survey, Bulletin 48, 115 p.

Zhang, P., Molnar, P., and Downs, W.R., 2001, Increased sedimentation rates and grain sizes 2– 4 Myr ago due to the influence of climate change on erosion rates: Nature, v. 410, p. 891– 897.

Ziegler, Z., Hsieh, J.C.C., Chadwick, O.A., Kelly, E.F., Hendricks, D.M., and Savin, S.M., 2003, Halloysite as a kinetically controlled end product of arid-zone basalt weathering: Chemical Geology, v. 202, p. 461–478.

305

BIOGRAPHICAL SKETCH

Kendall Fountain was born in Winter Haven, Florida, where he was exposed to the geology of the Central Florida Phosphate District at an early age via a geologist father. He earned a Bachelor of Science degree with honors at the University of Florida in May 1989, majoring in geology. He then began Graduate School at the University of Florida and received a

Master of Science degree in April 1994, with a major in geology. Kendall subsequently continued his education at the University of Florida in order to receive a Ph.D. in geology, with a

minor in environmental engineering, focusing on long held questions related to the origin and

significance of Cypresshead Formation sediments in Florida and Georgia. Presently employed by

Plum Creek Timber Company as their Senior Manager – Mineral Resources, Kendall’s responsibilities include identification and development of mineral resources associated with

Plum Creek’s land base across the United States, and coordination of group activities with Plum

Creek’s timber resource, real estate, and business development divisions. Prior to joining Plum

Creek, and while attending the University of Florida, Kendall started a sole proprietorship

consulting firm, Fountain Geological, which provided consulting expertise in the identification

and development of industrial mineral prospects and addressed mineral resource recovery and

quality issues through the application of advanced analytical techniques.

306