Miocene of the southeastern United States : proceedings of a symposium held December 4-5, 1980, Tallahassee, Florida ( FGS: Special publication 25 )

YONGt
LIB. OF
FLA. HIST.

STATE OF FLORIDA
DEPARTMENT OF NATURAL RESOURCES
Elton J. Gissendanner, Executive Director

DIVISION OF RESOURCE MANAGEMENT
Casey J. Gluckman, Division Director

BUREAU OF
C.W. Hendry,

GEOLOGY
Jr., Chief

SPECIAL PUBLICATION NO. 25

MIOCENE OF THE SOUTHEASTERN UNITED STATES

Proceedings of

a Symposium Held
Tallahassee, Flor

December 4-5, 1980

Edited by

THOMAS M. SCOTT

SAM B. UPCHURCH

Published by
Florida Department of Natural Resources
Division of Resource Management
Bureau of Geology

in cooperation with

The Southeastern Geological Society

Tallahassee
1982

&-3

DEPARTMENT

OF

NATURAL RESOURCES

BOB GRAHAM, Governor

GEORGE FIRESTONE, Secretary of State

JIM SMITH, Attorney General

GERALD A. LEWIS, Comptroller

BILL GUNTER, Treasurer

DOYLE CONNER, Commissioner of Agriculture

RALPH D. TURLINGTON, Commissioner of Education

ELTON J. GISSENDANNER, Executive Director

LETTER OF TRANSMITTAL

BUREAU OF GEOLOGY
TALLAHASSEE
OCTOBER 20, 1982

Governor Bob Graham, Chairman
Florida Department of Natural Resources
Tallahassee, Florida 32301

Dear Governor Graham:

The Bureau of Geology, Division of Resource Management, is
publishing as its Special Publication 25, "The Miocene of the
Southeastern United States, Proceedings of the Symposium."

Special Publication 25 discusses the Miocene stratigraphy,
paleontology, and economic geology of the southeast. It is
the proceedings of a symposium co-sponsored by the Bureau of
Geology and the Southeastern Geological Society and was held
December 4-5, 1980.

Respectfully yours,

Charles W. Hendry, Jr., Chief
Bureau of Geology

CWHJr/tsj

iii

Printed for the
Florida Department of Natural Resources
Division of Resource Management
Bureau of Geology

Tallahassee
1982

iv

CONTENTS

Page

INTRODUCTION........................ .... ............................... .. viii

Thomas M. Scott and Sam B. Upchurch

DEPOSITIONAL FRAMEWORK AND PALEOENVIRONMENTS OF MIOCENE STRATA FROM
NORTH CAROLINA TO MARYLAND........................................ 1

Thomas G. Gibson

DIATOM BIOSTRATIGRAPHY OF THE CHESAPEAKE GROUP, VIRGINIA AND MARYLAND.... 23

William H. Abbott

THE EFFECT OF PREDATION ON MIOCENE MOLLUSC POPULATIONS OF THE CHESAPEAKE
GROUP................................. ..... ......... ..... ... 1............ 35

Patricia H. Kelley

STRATIGRAPHY AND PETROLOGY OF THE PUNGO RIVER FORMATION, CENTRAL COASTAL
PLAIN OF NORTH CAROLINA..........................* ........ ........... 49

A. Kelly Scarborough, Stanley R. Riggs, and Scott W. Snyder

CYCLIC DEPOSITION OF THE UPPER TERTIARY PHOSPHORITES OF THE AURORA AREA,
NORTH CAROLINA, AND THEIR POSSIBLE RELATIONSHIP TO GLOBAL SEA LEVEL... 65

Stanley R. Riggs, Don W. Lewis, A. Kelly Scarborough, and Scott Snyder

SYNTHESIS OF PHOSPHATIC SEDIMENT FAUNAL RELATIONSHIPS WITHIN THE PUNGO
RIVER FORMATION: PALEOENVIRONMENTAL IMPLICATIONS..................... 85

Scott W. Snyder, Stanley R. Riggs, Mark R. Katrosh, Don W. Lewis, and
A. Kelly Scarborough

FORAMINIFERA OF THE PUNGO RIVER FORMATION, CENTRAL COASTAL PLAIN OF NORTH
CAROLINA ........................ .. .......... 100

Mark R. Katrosh and Scott W. Snyder

PRELIMINARY STRATIGRAPHIC REPORT ON THE PUNGO RIVER FORMATION IN ONSLOW
BAY, CONTINENTAL SHELF, NORTH CAROLINA............................... 122

Don W. Lewis, Stanley R. Riggs, Stephen W. Snyder, Albert C. Hine,
Scott W. Snyder, and Virginia J. Waters

MIOCENE SEISMIC STRATIGRAPHY, STRUCTURAL FRAMEWORK AND SEA LEVEL CYCLICITY:
NORTH CAROLINA CONTINENTAL SHELF...................................... 138

Stephen W. Snyder, Albert C. Hine, and Stanley R. Riggs

STRUCTURAL AND SEDIMENTARY SETTING OF PHOSPHORITE DEPOSITS IN NORTH
CAROLINA AND IN NORTHERN FLORIDA..................................... 162

James A. Miller /

THE STRATIGRAPHIC SUBDIVISION OF THE HAWTHORNE GROUP IN GEORGIA
(Abstract).................... .................................. 183

Paul F. Huddlestun

THE STRATIGRAPHIC DEFINITION OF THE LOWER PLIOCENE INDIAN RIVER BEDS OF
THE HAWTHORNE IN SOUTH CAROLINA, GEORGIA, AND FLORIDA (Abstract)...... 184

P.F. Huddlestun, R.W. Hoenstine, W.H. Abbott, and R. Woolsey

THE SUCCESSION OF MIOCENE (ARIKAREEAN THROUGH HEMPHILLIAN) TERRESTRIAL
MAMMALIAN LOCALITIES AND FAUNAS IN FLORIDA........................... 186

Bruce J. MacFadden and S. David Webb

HIGH MARINE TERRACES OF MIO-PLIOCENE AGE, FLORIDA PANHANDLE.............. 200

William F. Tanner

STRATIGRAPHIC REVISION OF THE TORREYA FORMATION OF FLORIDA (Abstract).... 210

Paul F. Huddlestun and Muriel E. Hunter

THE BIOSTRATIGRAPHY OF THE TORREYA FORMATION OF FLORIDA.................. 211

Muriel E. Hunter and Paul F. Huddlestun

NEOGENE CARBONATES OF THE FLORIDA PANHANDLE............................. 224

Walter Schmidt, Murlene Wiggs Clark, and Sharon Bolling

NEOGENE STRATIGRAPHY OF THE SOUTHWESTERN FLORIDA PANHANDLE (Abstract).... 235

Ramil Wright and Murlene Wiggs Clark

A COMPARISON OF THE COTYPE LOCALITIES AND CORES OF THE MIOCENE HAWTHORN
FORMATION IN FLORIDA ............................... ................ 237

Thomas M. Scott

CLAY MINERALOGY OF THE HAWTHORN FORMATION IN NORTHERN AND EASTERN
FLORIDA..................... ........... .................... ........ 247

B.A. Reik

SILICIFICATION OF MIOCENE ROCKS FROM CENTRAL FLORIDA..................... 251

Sam B. Upchurch, Richard N. Strom, and Mark G. Nuckels

MIOCENE CYCLIC SEDIMENTATION IN WESTERN LEE COUNTY, FLORIDA.............. 285

Thomas M. Missimer and Roland S. Banks

A DISCUSSION OF THE MIOCENE/PLIOCENE DIATOMS OF LEE COUNTY, FLORIDA
(Abstract) ........................... .. ............................... 300

Susan L. Klinzing

UPPER MIOCENE-PLIOCENE PLANKTIC AND BENTHIC FORAMINIFERA FROM LEE COUNTY,
FLORIDA (Abstract).................................................... 300

Douglas M. Peck and Sherwood W. Wise, Jr.

BIBLIOGRAPHY................ ......... .......... ............... 301

INTRODUCTION

Miocene strata in the southeastern United States have been studied for over
one hundred years. Early work dealt primarily with the abundant fauna and
flora of the Miocene, especially in the northern portion of the Atlantic
Coastal Plain. Interest has continued in the Miocene Series of the southeast
because of the important mineral resources that it contains. Phosphate is
mined from Miocene strata in North Carolina and Florida. Recent exploration
for phosphate indicates that large reserves exist in Miocene strata both on-
shore and offshore on the continental shelf. The nation's most important
deposits of Fuller's earth occur in South Georgia and Florida. Therefore,
interest in the Miocene Series has shifted to areas of economic potential
and extensive exploration continues.

In spite of this growing interest in the economic aspects of the Miocene
Series in the southeastern United States, the strata remain poorly understood.
The Miocene worldwide is characterized by sedimentary deposits that represent
unusual climatic conditions. These deposits include phosphorites, diatomites,
opaline cherts, and a wide suite of silica-rich clays. The unusual nature of
these deposits and the uncertainty regarding the environments they represent
have hindered interpretations of the Miocene Series. Many problems remain as
to Miocene lithostratigraphy, biostratigraphy, and chronostratigraphy on the
Coastal Plain of the United States.

Recent developments in mineral exploration and related stratigraphic studies
provide the first major insights into the details of the Miocene Series in the
southeast. Because these studies have not been brought together elsewhere and
many have not yet been published, it appeared to us that a major symposium on
the Miocene of the southeastern United States was needed.

This publication, the proceedings of the symposium, brings together many of
the most up-to-date studies on the Miocene of the southeast in a format where
regional and topical comparisons are possible. The Southeastern Geological
Society and the Florida Bureau of Geology agreed to host the symposium. The
subject areas to be included in the symposium were open. The single restric-
tion imposed was that the Miocene Series be the focal topic. Invitations for
contributed papers were mailed to all persons known to us to be working on the
Miocene in the Coastal Plain from Maryland to Florida. A call for papers was
also distributed to institutions of higher learning, state and federal agen-
cies, and industry. A screening committee consisting of the editors and
Felipe A. Pontigo, Jr. reviewed all abstracts. The symposium was held on
December 4-5, 1980 in Tallahassee, Florida. Eighteen papers and six
abstracts were submitted for publication in the proceedings volume. All
papers were reviewed by the editors for content and style. In general, we
made no effort to limit the ideas expressed by the authors in order for the
proceedings volume to reflect the nature of the symposium. The symposium was
open and ideas, many of which were spontaneous, were enthusiastically shared
by all.

We have organized this proceedings volume geographically. Thus, the papers
start with discussions of the Miocene in the Chesapeake Bay area. Then the
papers follow a progression southward, ending in southern Florida. Within
each geographic area we organized the papers so that structure and lithostra-
tigraphy are discussed first with biostratigraphy and petrology following.

viii

The Chesapeake Group of Maryland and adjacent areas has been studied for many
years. It is well known for its fossils and superb exposures along the
Atlantic and Chesapeake Bay shores. Gibson's paper discusses the depositional
framework of the Miocene in the Salisbury Embayment, which includes the
Maryland exposures, and compares it with the Albemarle, or Aurora, Embayment
in North Carolina. His paper provides a regional overview of the Miocene in
the northern half of the study area. Abbott's paper, which follows, discusses
the important diatom suite in the Chesapeake Group and provides chronostra-
tigraphic evidence for the strata using Blow's (1969) foraminiferal zones and
Abbott's (1978) diatom zones. Abbott also discusses the Miocene section in
the B-3 C.O.S.T. well drilled in the Baltimore Canyon on the continental
slope. Finally, Kelley's paper deals with one of the most recent research
aspects of Miocene molluscs from the Chesapeake Group. It is the mollusc
fauna that first attracted attention to Miocene strata in the Salisbury
Embayment and Miocene studies throughout the Coastal Plain compare faunas with
the well-known mollusc assemblage of the Chesapeake Group. Kelley's paper
provides a new and interesting dimension to this mollusc fauna by describing
dynamic interactions between predator and prey.

The phosphorite deposits of the Albermarle, or Aurora, Embayment of North
Carolina are being exploited now, with one mine presently in operation and
another under development. The Miocene in this economically important area is
poorly understood and few papers have been published on it. The papers
included here constitute the first reports of a series of studies on the
detailed sedimentology and biostratigraphy of the Pungo River Formation, which
is being mined and which constitutes much of the Miocene Series in North
Carolina. The paper by Scarborough and others details the stratigraphy and
petrology of the Pungo River. The paper by Riggs and others relates the depo-
sitional sequence of the Pungo River to cyclic, global sea-level fluctuations.
Finally, the papers by Snyder and others and Kotrosh and Snyder discuss the
biostratigraphy of the Pungo River in the Aurora Embayment.

Also included in this volume are two papers describing a newly explored area
in North Carolina south of the Aurora Embayment. This area may contain com-
mercial phosphorite deposits in an offshore environment. The new area, Onslow
Bay, lies offshore between Cape Fear and Cape Lookout. The paper by Lewis and
others is the first paper to describe the stratigraphy of this potentially
important deposit and the paper by Snyder and others relates the Onslow Bay
deposits to tectonic and sea-level variations.

Little is known about the Southeast Georgia Embayment. The paper by Miller
describes the tectonic and depositional conditions proposed for phosphorite
horizons in North Carolina and Florida. His study brackets the Southeast
Georgia Embayment and provides continuity to these proceedings. In addition,
Miller's paper provides new data on the stratigraphy of the Miocene phosphor-
ites of north Florida.

This area is just now being explored and the extensive phosphate deposits
found there have resulted in a bitter debate over mining in the area. The
abstracts by Huddlestun and Huddlestun and others briefly describe current
work on the Miocene Series in and around the Southeast Georgia Embayment.
These abstracts deal with the Hawthorn Formation, an extremely complex unit
that contains extensive beds of opaline chert, phosphorite, authigenic paly-
gorskite and sepiolite, dolosilt, and terrigenous clastics. The phosphate and

Fuller's earth deposits of the Hawthorn and related Torreya Formation, are
mined in south Georgia and north and central Florida. The Huddlestun
abstracts reflect on-going research aimed at defining the complex stra-
tigraphy of this unit in its northern extent.

North Florida is also described in the paper by MacFadden and Webb. The
Miocene of Florida contains one of the world's most notable vertebrate faunas.
Their paper describes this fauna and provides important data on the climate
during deposition of the Miocene strata in the deep south. Tanner's paper on
the Mio-Pliocene terraces of the Florida panhandle provides additional data on
sea levels in latest Miocene and Pliocene time in Florida.

The Torreya Formation is a poorly understood unit in northwest Florida. The
unit, as defined, is important because it includes the most extensively mined
Fuller's earth deposits in the United States. The Torreya is a recently
designated formation and the abstract by Huddllestun and Hunter and the paper
by Hunter and Huddlestun provide valuable data as to its age and distribution.

The Miocene Series lies in the subsurface in most of the panhandle of Florida,
so relatively little is known about its geometry and distribution. The papers
by Schmidt and others and Wright and Clark provide discussions of these strata
and their relationships to strata in adjacent areas. Several formations,
which have only recently been described, are included and new data on the
micropaleontology are presented.

The Hawthorn Formation extends from the panhandle of Florida eastward and from
central Georgia southward. It was first described in central Florida and, in
spite of its economic importance, it is still a controversial unit. The
Hawthorn comprises a complex sequence of terrigenous clays, silts and sands,
carbonates, phosphorites, and authigenic clays and silica. Since there has
been a great deal of difficulty in recognizing and correlating the Hawthorn
and in interpreting its origin, it is important that the type section be pre-
served and thoroughly described for future geologists. Scott's paper descri-
bes three cotype localities of the Hawthorn, including cores taken at
exposures that are partially lost to modern geologists. This paper provides
an important new standard for discussion of the Hawthorn. The paper by Reik
discusses the clay mineralogy of the Hawthorn and presents new data on the
association of authigenic clay (palygorskite) and dolomite. Finally, the
paper by Upchurch and others discusses the origin of the siliceous sediments
of peninsular Florida. This paper puts forth a model for the origin and
diagenesis of the opaline and quartz-rich sediments of the Hawthorn and, more
important, offers an explanation for some of the authigenic clays and dolo-
silts that complicate interpretation of Hawthorn depositional environments.

The Miocene sediments of central and south Florida contain more carbonate and
less plastic material than those to the north. In this region exposures of
the Miocene are rare and it has been difficult for geologists to work out the
stratigraphy. The Miocene Series, including the Hawthorn Formation and parts
of the Tamiami Formation, dip southward from a northeast-southwest line be-
tween Tampa and Cape Canaveral. In some areas portions of the Miocene consti-
tute important aquifers. Recent work delineating the aquifer system in Lee
County, Florida, has provided the first extensive data on the lithostra-
tigraphy and biostratigraphy of the Miocene. The paper by Missimer and Banks
describes a small area in which there are abundant well data. They are able

to demonstrate significant structure in the Miocene section and show that
cyclic sedimentation has occurred. It is interesting to compare their paper
with the papers on cyclic sedimentation in the Aurora Embayment and Onslow
Bay. The abstracts by Klinzing and Peck and Wise indicate some of the
biostratigraphic and chronostratigraphic relationships that are just now
emerging in the Miocene of South Florida.

There are still many unanswered questions regarding the Miocene in the Coastal
Plain of the southeastern United States. The complex depositional and geo-
chemical environments of the Hawthorn are largely unexplained. There is still
controversy as to the origin of phosphorite and Fuller's earth. Sea level
fluctuations are well documented in the Miocene Series of the area but details
in the stratigraphic record and regional effects are still not explained. The
paleontology of the Miocene Series is diverse and complicated. More work
needs to be done to reconcile the chronostratigraphy and explain the various
depositional environments represented. Finally, the Miocene of the southeast
contains abundant data concerning global Miocene events. The unusual mineral
and biotic assemblages of the Coastal Plain are found in the Miocene sediments
of other areas. Considering that the Miocene rocks of the southeast are rela-
tively undeformed and have not been subjected to significant burial-related
diagenesis, they should contain information as to global Miocene sedimentation
and environments.

We would like to thank the many people who worked on this symposium and pro-
ceedings. The officers of the Southeastern Geological Society were supportive
of the project. Paulette Bond reviewed many of the manuscripts. Sheila
Weissinger and Peggy Chaffin typed the manuscript. We also thank C.W. Hendry,
Jr., Chief of the Bureau, for his continuing support.

May 18, 1982

THOMAS M. SCOTT SAM B. UPCHURCH
Florida Bureau of Geology University of South Florida
Tallahassee, Florida Tampa, Florida

DEPOSITIONAL FRAMEWORK AND PALEOENVIRONMENTS OF MIOCENE
STRATA FROM NORTH CAROLINA TO MARYLAND

Thomas G. Gibson
U. S. Geological Survey
Washington, D. C.

ABSTRACT

Two large, adjacent, sedimentary basins, the Salisbury Embayment from New
Jersey to Virginia, and the Albemarle Embayment in North Carolina, received
widespread marine deposition in the Early Miocene and intermittently through
the remainder of the epoch. Lithofacies differ significantly between the two
basins in the lower and middle Miocene but are similar in the upper Miocene.

Siliciclastic material dominates the Miocene section in the Salisbury
Embayment. Initial Miocene deposition, however, consists of two sequences of
basal sand followed by diatomaceous clay, indicating two periods of biogenic
deposition. The initial strata belong to the Calvert Formation and probably
reflect widespread, shallow, open-marine conditions. Upwards, the Calvert
lithofacies become more fragmented near the lower-middle Miocene boundary
because of deltaic input into the northern part of the basin. Delta out-
building in southern New Jersey and its influence southward onto the present
Eastern Shore of Maryland led to restricted oceanic circulation in the western
part of the embayment. Several periods of renewed open circulation to the
western part of the basin in the upper part of the Calvert Formation and in
the overlying Choptank Formation resulted in shallow, open-marine deposition.
The overlying St. Marys Formation in the western part of the embayment was
largely deposited in very shallow-marine and in restricted marine environ-
ments. Later Miocene deposits reflect a southward migration of the locus of
deposition from southern Maryland into central Virginia where there were pri-
marily open, shallow-marine environments.

In contrast, the Albemarle Embayment to the south in North Carolina is domi-
nated by chemical and biochemical deposition, in the form of phosphate, car-
bonate, and diatomite lithofacies throughout the lower and middle Miocene.
Faunal information suggests three ages of deposition comprising the largely
biochemical strata. The lower rates of sedimentation in this basin are accom-
panied by depositional environments commonly deeper (to midshelf) than coeval
deposits in Virginia and Maryland. Siliciclastic units dominate the upper
Miocene section, which is limited to the northern part of the embayment.

INTRODUCTION

Strata of Miocene age are widespread over the Atlantic Coastal Plain from
North Carolina to New Jersey and have economic potential, including deposits
of phosphate and diatomite. These strata are found in two adjacent deposi-
tional basins, the Salisbury to the north and the Albemarle to the south. The
Salisbury Embayment contains many surface exposures of Miocene strata and a
limited amount of subsurface control. Many of the outcropping units are
richly fossiliferous and are well known for their diverse vertebrate and

invertebrate faunas. In the Albemarle Embayment, surface exposures of Miocene
strata are mainly found in the northern part of the basin; mostly subsurface
data were obtained for the central and southern part.

Many marine transgressions are represented by the strata, some having taken
place in both embayments and others apparently restricted to the northern, the
Salisbury Embayment. The deposits left by these transgressions extend a con-
siderable distance across the Coastal Plain, even in their present, partly-
eroded condition, and lie upon Cretaceous to upper Oligocene strata in most
areas. Miocene strata are even found resting upon Piedmont crystalline rocks
in southern Virginia and northern North Carolina.

The purpose of this paper is to summarize the times of transgression and the
patterns of deposition in the two embayments during the Miocene. From this
data base can be drawn inferences concerning the tectonic history and
paleogeography of the area.

Knowledge of the biostratigraphy of Miocene-age strata has progressed con-
siderably in the past decade, primarily because of the considerable amount of
work done on the oceanic sequences by the JOIDES sampling and other related
studies. The shallow-water environments of deposition represented by most
Miocene strata in these two embayments result in a limited number and variety
of the planktonic organisms upon which most intercontinental correlations
depend. Thus, the exact correlations of some Atlantic Coastal Plain strata
with the Miocene stages of Europe are uncertain. Some of those strata that
have yielded sufficient biostratigraphic data have changed the placements that
were generally accepted during the past years. The Yorktown Formation,
generally believed to occupy the entire Late Miocene by most authors (Cooke,
Gardner, and Woodring, 1943) was later considered to be at least partly of
Pliocene age (Akers, 1972; Hazel, 1971); it is now shown as being entirely of
Pliocene age (Gibson, in press b). Even some of the underlying strata, placed
in the St. Marys Formation by Mansfield and herein termed "Late Miocene", have
yielded some floral indications of being at least partial Early Pliocene in
age (Andrews, 1980).

The increasingly refined biostratigraphic determinations also show that
although many marine transgressions took place during the Miocene, a large
part of Miocene time is not represented by strata in the central Atlantic
Coastal Plain (figure 1). This "spottiness" is also characteristic of the
more southern sections as in Florida, as indicated by Hudlestun at this sym-
posium.

The correlations in this study used both Foraminifera and Mollusca for the
surface outcrops and largely Foraminifera for the subsurface samples. The
environmental reconstructions are based upon lithologic and faunal criteria.
The faunal interpretations are based largely upon knowledge of modern bentho-
nic foraminiferal ecology of individual species and groups of species and also
upon assemblage characteristics, particularly species diversity of the bentho-
nics and the relative abundance of planktonic specimens (see Phleger, 1960;
Walton, 1964; Gibson, 1968; Gibson and Buzas, 1973; Murray, 1973; and
Boltovskoy and Wright, 1976, for discussions and further references).

MIO-
CENE

~1fflli~

Figure 1. Correlation chart for strata of late Oligocene through early Pliocene
age in the central Atlantic Coastal Plain. Stages, time scale, and
planktonic foraminiferal zones are from Vail and Mitchum (1979).
Dashed lines indicate incompletely kribwn time limits.

iltUiWil

REGIONAL SETTING

The central Atlantic Coastal Plain generally has been considered to be domi-
nated by a series of structural arches and intervening basins (Murray, 1961).
These arches were believed to result from flexures in the basement rocks.
Although some suggestions of possible faulting had been made for the Coastal
Plain (Cederstrom, 1945; Ferenczi, 1959), faulting was not considered of major
significance in the area. Brown, Miller, and Swain (1972), however, proposed
that the faces distribution of Atlantic Coastal Plain Mesozoic and Cenozoic
strata was largely fault controlled, and they produced a fault model for the
area. An integral part of the model is the presence_ of hinge lines that
Miller (1971) proposed as largely controlling phosphate deposition in the
eastern part of the Albemarle Embayment and that Brown, Miller, and Swain
(1972) proposed for the entire Coastal Plain. Sheridan (1974) postulated a
block-fault origin for the entire Atlantic continental margin of North America
and considered the marginal basins as isolated, fault-bound troughs.

Figure 2 shows positive and negative features affecting this part of the
Coastal Plain. Although they have generally been called "arches," some of
these features will be termed "highs" in this paper as they show charac-
teristics of the structures that Brown, Miller, and Swain (1972) considered
fault bounded. These characteristics include the complex faces patterns and
the presence of rocks of some strata on the "arches" that are absent from the
neighboring "basins." The mobility of these features is seen in the amount of
the stratigraphic record preserved in the highs compared with that of the
basins. On the Norfolk high, strata of Paleocene and Early and Middle Eocene
age are not present, at least in the area of Norfolk (Brown, Miller, and
Swain, 1972), although strata of these ages are widespread and several hundred
feet thick both northwest and south of this feature. However, strata of Late
Eocene age, although restricted in their distribution to only a small part of
central Virginia, are found on the high at Norfolk. A similar situation is
found at the New Bern high. Here units of Oligocene age are found on the
high, but they are poorly represented in the Albemarle Embayment immediately
adjacent; this distribution indicates either large-scale erosion of Oligocene
strata in the embayment or, most likely, nondeposition. The succeeding
Miocene-age strata, however, are well represented in the Albemarle Embayment,
but they are largely or completely absent on the New Bern high and also show a
thinning trend as the high is approached from the north. These distributional
patterns suggest a structural mechanism, like faulting, which probably would
be more mobile than basement-controlled arching would be.

The major structure in the north is the Normandy high, proposed by Brown,
Miller and Swain (1972). This structure is north of the study area involved
here and is not discussed. South of the Normandy high is the Salisbury
Embayment. This large embayment contains a more complete record of the
Miocene than does the Albemarle Embayment to the south. The Salisbury
Embayment contains at least 1,000 feet of Miocene as shown in the Hammond well
on the Eastern Shore of Maryland (Anderson, 1948). At the southern end of the
Salisbury Embayment is the Norfolk high. This high also marks the northern
border of the Albemarle Embayment, the southern border of which is the New
Bern high.

DEPOSITIONAL ENVIRONMENTS

The depositional framework of the Miocene strata is largely controlled by the
tectonic setting and activity of the highs and embayments. As a result of
these factors, the water depth, distance from shore, and influx of terrigenous
plastic debris varied and caused the change in the nature of the lithology
seen in the Miocene deposits in these two embayments. The water depths repre-
sented by these strata were variable but were mainly from marginal-marine
environments to middle-shelf limits. None of the strata found in outcrop or
in subsurface sampling are characteristic of depths greater than middle shelf
or approximately 100-150 m. Strata approaching these greater depositional
depths are found only in the easternmost parts of the areas, except for the
Pungo River Formation in central North Carolina, where depositional depths of
100-150 m are found a moderate distance inland from the present coast.

Another important factor in the nature of the Miocene strata was the relative
influx of terrigenous plastic debris. The amount of incoming terrigenous
material was variable in these embayments during the Miocene; relatively low-
input plastic deposits including phosphatic sand, diatomaceous clay, and car-
bonate beds, were characteristic of parts of the record, and quartz sand and
gravels were found in other areas and times. Although climatic conditions may
have had an important bearing on the amount of plastic debris being carried
into the basin and cannot be discounted (little is presently known about
this), the major influence on rate of plastic supply in these embayments
appears to have been the result of changes in the altitude of the source area.
Information on the character of the source area is determined from the litho-
logy and thickness of the strata; it is generally assumed that thicker depo-
sits of quartz sand and clay indicate a considerable influx of plastic
materials, whereas diatomaceous and phosphatic strata are used as indicators
of lower terrigenous input and proportionately increased biogenic activity.
As lithologies vary with distance from shore and depth of water, the location
of the shoreline and depth of water involved is derived from paleontologic
information to determine the importance of the plastic or biogenic input.

STRATIGRAPHY

The Salisbury Embayment has a more complete record of the Miocene than that
found in the Albemarle Embayment, although there still are many time gaps
(figure 1). The late Early and early Middle Miocene is fairly well repre-
sented in both embayments, but the rest of the Miocene is best represented in
the Salisbury. Miocene strata in each embayment will be discussed from oldest
to youngest, as well as lithology, thickness, areal distribution, and paleoen-
vironments of each formation. Facies changes and times of deposition are
emphasized, especially to contrast the Salisbury and Albemarle Embayments.
Figure 3 shows the location of important outcrop and well sections.

SALISBURY EMBAYMENT

Strata of Miocene and Pliocene age in the Salisbury Embayment comprise the
Chesapeake Group, named for the good exposures along Chesapeake Bay and its
tributaries. The Chesapeake Group, as used by Shattuck (1904), Mansfield
(1943), and most authors, consisted of, in ascending order, the Calvert,

PENNSYLVANIA

S. r MARYLAND

W. VA. D C

VIRGINIA

NORTH CAROLINA

Normandy
arch

Salisbury
embayment

SCAC L
0 50 100 11is
0 80 16 KllIrUBo'f

ape Fear arch

SCALE
0 50 100 'llles
0 80 1 60 KIomclcrI

re 2. Map showing major embayments
and highs of the middle

Atlantic Coastal Plain

during the Miocene.

Hammond Well

i gravelly and

Upper Miocene

St. Marys Formation

limy and, median, Choptank Formation
grey, foaillferou.

very fine sand,
fo.llltfroua
gray silty clay,
fewfossils
fine sand

Calvert Formation

Figure 3. Map of localities

1.AMCOR core 6011 12.Tappahannock

2.Dover, Delaware 13.Richmond

3.Well 231 14.Petersburg

4.Type Choptank 15.Moores Bridge

5.Hammond well 16.Gatesville

6.Lyons Wharf 17.Murfreesboro

7.Calvert Cliffs 18.Belhaven

8.St. Marys City 19.Bonnerton

9.Popes Creek 20.Lee Creek mine

10.Nomini Cliffs 21.Great Lake

11.Oak Grove

Well 231
Deptcht)

--- -

50-

100 -
_. e-l
-* - *T
r^T^r_

gray and brown
sand and gravel
green sandy clay
green sand,
abundat shael
green sandy cay

brown day,
ray sand
interbeds

green
diatomaceous
clay

green and brown
interbedded days

glauconitic snd

Plum Point
Marl Member

Alloway
Clay Member

Fairhaven
Member

Calvert Formation(?)

Calvert Formation

Kirkwood Formation

Calvert Formation

Eocene

Figure 4. Geologic column for Larry

G. Hammond No. 1 well, Dorchester

County, Maryland.

Figure 5. Geologic column for well

231, near Sudlersville, Queen

Annes County, Maryland.

Figu

Depth (f)
0-

200-

400-

o

o
a

grevlly very coare
eind

gravelly into
very coarM nd

gravelly and,
.hell fragments

gray yiltyclay,
few foI ila

brownish-gray
silty clay,
fewofoassls
glauconitic sand Eoc
Eocene

Choptank, St. Marys, and Yorktown Formations. Marine units of the lowest
three formations are well exposed in Maryland and southward into northern
Virginia, but marine units of the Yorktown are restricted primarily to
Virginia and North Carolina.

Several changes in the composition and age of the Chesapeake Group have been
made in the past decade. A new formation, the Eastover, was proposed by Ward
and Blackwelder (1980) for the beds that Mansfield (1943) placed in the upper
part of the St. Marys Formation in Virginia. The Yorktown Formation, the
uppermost formation in the group, formerly placed in the upper Miocene (Cooke,
Gardner, and Woodring, 1943; Mansfield, 1943), is now considered entirely of
Pliocene age on the basis of the calcareous nannofossils (Akers, 1972) and
planktonic Foraminifera (Gibson, in press b). Because of its Pliocene age, it
will not be treated in this report.
CALVERT FORMATION

Shattuck (1902) named the Calvert Formation from the famous exposures in the
Calvert Cliffs along the western shore of Chesapeake Bay in Calvert County,
Maryland (figure 3). A continuous core hole drilled at the Calvert Cliffs
site of the Baltimore Gas and Electric Company's (BG&E) nuclear power plant
found the Calvert Formation to be 200 feet thick, compared with thicknesses of
150 to 170 feet proposed by Shattuck (1904) from integrated sections in the
updip outcrop area. The formation thickens to 500 feet on the Eastern Shore
of Maryland, as seen in the Hammond well (figures 3 and 4) (Anderson, 1948).
An isopach map of the Calvert Formation and generally equivalent strata
(figure 7) shows the wide distribution and relatively thick section.

Shattuck (1904) divided the Chesapeake Group in the Calvert Cliffs area into
24 lithologic units, the lowermost 15 of which were in the Calvert Formation.
Although Shattuck termed them "zones," they were not biostratigraphic, but
lithic units characterized by the sediment type and nature or abundance of
shells. Some of the zones can be traced over considerable distances, as
Gernant (1970) did in the Choptank Formation. Because of confusion over
Shattuck's use of the term "zones," most recent authors have substituted a new
terminology. Gernant (1970) gave member status to some of the "zones" in the
Choptank Formation and replaced the zonal numbers with names; Gibson (1971)
used the term "units"; and Andrews (1978) used the term "Miocene Lithologic
Units". As most of the zones can be recognized in the Calvert Cliffs and
followed over shore to medium distances away from the cliffs with varying
degrees of success, some manner of retention of the "zones" seems desirable.
In this paper, they will be referred to as "beds" of Shattuck (1904).

Shattuck (1904) placed the three lowest beds of the Calvert Formation into the
Fairhaven Diatomaceous Earth Member, so named because of the thickness of
diatomaceous beds near Fairhaven, Maryland, at the northern end of the Calvert
Cliffs (figure 3). The lowest two beds (1 and 2) of this member, as defined
by him, are along the Patuxent River near Lyons Wharf (figure 3); they consist
of 7 feet of sand containing molluscan shells. At Lyons Wharf, the sand of
beds 1 and 2 was found by Shattuck to be overlain by 5 feet of diatomaceous
clay, which extended to the top of the exposed section. A thick section of
diatomaceous silty clay exposed at Fairhaven along the Chesapeake Bay was used

by Shattuck as the type area for his bed 3. Shattuck assumed that this diato-
maceous clay was a continuation of the top of the section at Lyons Wharf. Bed
3 was 55 feet thick in a well section at Fairhaven (Shattuck, 1904), and the
diatomaceous bed thickens to about 100 feet in the BG&E core hole (Gibson, in
press a). In this core hole, sandy strata are found below the diatomaceous
bed; this same sequence also is found at Popes Creek on the northern bank of
the Potomac River (figure 3). These sandy strata are 10 to 15 feet thick and
contain pebbles near the base. Most workers considered all the sand sequences
below the diatomaceous clay to be of similar age. Gibson (in press a)
separated the sandy beds below the diatomaceous strata at Popes Creek into a
new member of the Calvert Formation, the Popes Creek Sand Member.

Recent work by Andrews (1978) on diatoms from the Miocene strata of Maryland
indicated that the diatomaceous strata on the Patuxent River near Dunkirk,
Maryland (south of Lyons Wharf), belonged to his Zone I, an older zone than
Zone II, to which he assigned the diatomaceous beds found near Fairhaven, in
the BG&E corehole, and at Popes Creek. A hiatus between these two diatoma-
ceous intervals also was suggested in Abbott (1978) by the noncoincidence of
Abbott's diatom zones I and II in this area. Previous field work by the
author at the Kaylorite mine on the Patuxent River west of Dunkirk showed the
lowest part of the Calvert formation, the sand beds 1 and 2 of Shattuck
(1904), sitting upon greensand assigned to the Nanjemoy Formation of Eocene
age; here, beds 1 and 2 are overlain by a diatomaceous clay, the situation
found a few miles to the north by Shattuck at Lyons Wharf. This diatomaceous
clay yielded the fossils that Andrews determined to be older than those found
in diatomaceous beds at Fairhaven and Popes Creek. Resting upon this diatoma-
ceous clay unit is a 13-foot bed of brownish fine sand, similar lithologically
to that found at the base of the Calvert Formation at Popes Creek and the BG&E
core hole. Overlying this sand in the BG&E core hole and at Popes Creek are
diatomaceous clays belonging to Andrews Zone II (Andrews, 1978; Andrews, oral
commun., 1980); Zone II species are also found in the lowest part of the
diatomaceous section at Fairhaven. Thus, there appear to be two diatomaceous
clay beds, each underlain by a bed or beds of sand. As complete sections
showing both sequences of sand overlain by diatomaceous clay are not known
(except for the partial section in the Kaylorite mine), the common correlation
of fragmented sections by the lithological similarity of sand overlain by
diatomaceous clay equated them wherever found. The dating by Andrews (1978),
however, showed the age differences, and the stratigraphic relations are shown
by the section at the Kaylorite mine, where the lower sequence of sand and
diatomaceous clay is overlain by the sand of the upper sequence.

Figure 6 shows the author's arrangement of the lower strata of the Calvert
Formation. The amount of time between the two sequences is somewhat uncer-
tain, but an early Burdigalian age for his Zone I was suggested by Andrews
(1978) on the basis of the presence of a characteristic silicoflagellate spe-
cies. The uppermost of the two sand sequences is called the Popes Creek Sand
Member by Gibson (in press a). The lower sand and diatomaceous clay sequence
will be herein informally called the "Dunkirk beds." Bed 3 of Shattuck
(1904), by far the thickest part of his lower member of the Calvert Formation,
is retained as the Fairhaven Member. The occurrence of the older diatom Zone
I of Andrews (1978) only at Dunkirk and not in the other localities in
southern Maryland where the Popes Creek Sand Member is found indicates that

Shattuck (1904)
Composite Section

bed no.
15

6 6

-0
ft3

0E

t-e---;e-l
5 `H-

4

eO

K -* **
!!c.-..;

2 e
E ~ o cene
CVOerooe

> 9e - -
el r **
* *
', A *
-a *
< - *
2 __**- y* ~

Eoc-eneT
*'~ *r v
3 *- **r
*-**- *
*-** y a-
"* ** **-*
********"
! *~v *
l **-- r
*~ V ~V
r* ** **
'* '"*
> T Fl)
r V *
**-* -* *-*
* ~* *-
*I * *
*-* T *
****-***-*
f~B* *" *
-w* **v *"
* r **-*
y-w -y *
* ~r <"
r *T *-

T - *-
2 -C

1 ; ; :

Eocene"

Patuxent River
near Dunkirk

bed no.

sandy clay,
shells, greenish

sandy clay,
shells, greenish

silty and sandy
clay,
diatomaceous,
gray-green

sandstone, white

sand, shells,
brown

This paper
Popes Creek,
Potomac River

bed no.

Pleistocen

3 -...

n -^-. .-

Pleistocene

1--

**

2 .^

Eocene

sand, fine, gray

clay,
diatomaceous,
light gray-green

sand, medium,
shelly

Eocene

e

silty clay,
diatomaceous, brown

silty clay, shell
molds, olive-brown

clayey sand, fine,
olive-brown

sand, fine, few
pebbles, olive-brown
silty clay, olive-brown
sand, fine, pebbly,
lignitic, brown

Chart showing rearrangement of beds and members of Calvert Formation

as revised in this paper; beds 1 and 2 (Dunkirk beds) are separated

from bed 3 (Fairhaven Member) by sand and clay of Popes Creek Sand

Member, this member probably being disconformable upon the Dunkirk

beds (see Fig. 1).

0
-D

0

Em
OE

c
CLU

CO
LL

oo
0)
CIL

3 .0

1- 0

Figure 6.

the "Dunkirk beds" probably were eroded over much of the area during the
hiatus between Andrews' Zones I and II and presently are found only in iso-
lated pockets. This knowledge that deposition of diatomaceous sediments took
place at more than one time interval in this embayment is important for a
better understanding of the cause and significance of the three periods of
phosphate deposition in the Albemarle Embayment.

The overlying Plum Point Marl Member named by Shattuck (1904) consists of
olive-green to olive-brown silty and clayey fine sand that generally contains
scattered to highly abundant molluscan shells and a considerable number of
vertebrate remains, mostly of marine mammals. Beds 4 to 15 of Shattuck (1904)
comprise this member, which reaches a thickness of 90 to 100 feet in the
Calvert Cliffs area.

A general summary (from base to top) for the type Calvert in the Calvert
Cliffs and surrounding area is as follows: a basal sequence of a
transgressive sand (Shattuck's beds 1 and 2) overlain by diatomaceous clays,
followed by a depositional hiatus, then another transgressive sand (Popes
Creek Sand Member) overlain by diatomaceous clay of the Fairhaven Member
(Shattuck's bed 3), which has an upper scour surface as described by Dryden
(1936), This diatomaceous clay was followed by deposition of the generally
fossiliferous clay and sand of the Plum Point Marl Member (Shattuck's beds
4-15).

FACIES RELATIONSHIPS

Individual beds change somewhat in sedimentary nature and thickness throughout
the Calvert Cliffs-Patuxent River area, but the general faces and thickness
are similar. Toward the northeast, east, and south, more significant
changes in faces composition and/or thickness take place.

Equivalent to the lowermost beds of the Calvert Formation, the Dunkirk beds,
have been found in other areas in the Salisbury Embayment, including the
Dover, Delaware, well and the Atlantic Margin Coring Project (AMCOR) core
6011, off the coast of southern New Jersey (figure 3) (Abbott, 1978).
Although these beds are apparently thicker to the east (26 feet in the Dover
well), no significant change in faces occurs.

A major depositional feature affecting the younger strata of the Calvert, a
delta that was built southward from New Jersey, affected the northern part
of the Salisbury Embayment during Miocene. The diatomaceous clay of the
Fairhaven Member continues across the embayment from the Calvert Cliffs to the
northern part of the Eastern Shore of Maryland (well 231, figures 3 and 5),
and also southward into Virginia (Oak Grove corehole, Gibson et al., 1980)
(figure 3). The Fairhaven Member represented widespread, shallow, open-marine
conditions, as evidenced by common, though low, frequencies of planktonic
foraminifers in the BG&E core hole (figure 15). The overlying lower part of
the Plum Point Marl Member (beds 4-9) shows fragmentation of the environment
in the Salisbury Embayment (figure 8) as equivalent strata represent prodelta
deposition on the Eastern Shore of Maryland (Gibson, in press a), deltaic
deposits farther north in New Jersey (Isphording, 1970), and the protected
embayment environment along the Calvert Cliffs area. The prodelta deposits in

Maryland consist of thinly interbedded, brown, carbonaceous clays and gray,
well-sorted sands (well 231, figure 5). A protected embayment that formed
west of the deltaic area apparently had restricted circulation, as seen by the
almost complete absence of planktonic foraminifers in the BG&E corehole
(figure 15). The upper part of the Plum Point Marl Member of the Calvert on
the Eastern Shore, presumably equivalent to beds 10-15, contains fossilif-
erous, green, sandy clay and suggests inter-shelf deposition for this part of
the Calvert in this area. Apparently, the strength of the deltaic outbuilding
was episodic and had less influence at this time, allowing inner-shelf
environments to predominate.

Farther south of the Eastern Shore, as seen in the Hammond well near
Salisbury, Maryland (figure 4), the Calvert Formation is much thicker (500
feet). The absence of diatomaceous clay at the base of the Calvert in the
Hammond well indicates either that strata of "Dunkirk" and Fairhaven age are
not found here or that the brownish-gray silty clay has replaced the diatoma-
ceous beds. The rest of the Calvert section is mainly slightly fossiliferous,
silty clay, which suggests that this area was sufficiently removed from the
delta to the north so that the prodelta deposits were replaced by shallow
inner-shelf environments.

In northeastern Virginia, south and southeast of the Calvert Cliffs area, the
Calvert Formation is similar lithologically to that seen in the BG&E core hole
and the Calvert Cliffs areas. The Fairhaven diatomite thickens to as much as
180 feet (Teifke, 1973) and overlies a sand unit, the probable equivalent of
the Popes Creek Sand Member. The basal sand becomes more phosphatic toward
the south. The diatomaceous clay is overlain by silty clay containing shelly
horizons.

Westward and southwestward in Virginia, the Calvert thins, as seen in the Oak
Grove core hole (Gibson et al., 1980), although the composition of the units
remain similar. In the Oak Grove core hole, the Fairhaven diatomite is only
17 feet thick; this unit shows more thinning than do the overlying Plum Point
Marl Member equivalents. In the southwestern part of the outcrop area, near
Richmond (figure 3), only beds correlated with the upper part of the Calvert
have been identified (Abbott, 1978); the latest part of the Calvert-age sea
probably was considerably more transgressive in this area than was the early
part.

In the most southeastern onshore complete section in the Salisbury Embayment,
the U. S. Geologic Survey (U.S.G.S.) test well at Moores Bridge near Norfolk
(figure 3), the Calvert is 60 feet thick (Gibson, unpub. data). The lowermost
8 feet are phosphatic sand, overlain by shelly, green, silty clays. The basal
phosphatic sand indicates a zone of overlap between the deposits of the
Calvert and Pungo River Formations in the southeastern part of the embayment.

Figure 8 summarizes the paleogeography and paleoenvironments during the
Calvert, particularly for the later part of Calvert time. The deltaic
influence on the Eastern Shore resulted in a protected embayment in the
northern Chesapeake Bay area during deposition of beds 4-9 of the Calvert
Formation. Pulses of open-ocean circulation were felt in this area during the
later part of the Calvert Formation. In the southern part of the embayment,

4- I -.--

W.VA. '

1 .
"*'' ' r

-0 10
SS

s c \ ^ _,

N

0 50 100 MI
o 80 160 Km

Figure 7. Isopaeh map of upper lower and

lower middle Miocene strata. Contours
in feet. 1 is Kirkwood Formation, 2 is
Calvert, and 3 is Pmnge River.

PENNSYLVANIA

'"' -' I--,,
I MARYLAND
W VA. I
.*/ I '.

VIRGINIA ..
r-,

. . . . . ..

NORTH CAROLINA ci

80 )CR

0
NEW
JERSEY

- 00

0 50 10 M.
0 80 160 Km

80 75"
Figure 8. Paleoenvironments postulated
for lower half of Plum Point Marl

Member of Calvert Formation and

equivalent units.

Figure 9. Isopach map of Choptank For-

mation and equivalent units. Contours
are in feet.

Figure 10. Paleoenvironments postulated

for Choptank Formation and equivalent

units.

35 -

351

inner-shelf environments predominated; some environments approached middle-
shelf depths in the southeastern part of the embayment.

CHOPTANK FORMATION

Overlying the Calvert Formation is the Choptank Formation, named by Shattuck
(1902) from exposures along the Choptank River on the Eastern Shore of
Maryland (figure 3). Most subsequent work has been more concerned with the
extensive exposures along the Calvert Cliffs on the western shore of
Chesapeake Bay, although Gernant (1970) studied strata from both places.
Shattuck (1904) gave a thickness of 45 to 55 feet in outcrops along the
western shore, and this thickness is similar to the 60 feet found in the BG&E
core hole. The formation thickens to the east in the subsurface (figure 9),
reaching 125 feet in the Hammond well on the Eastern Shore (figure 4)
(Anderson, 1948); a maximum of 150 feet was found in a well southwest of the
Hammond well (Gibson, in press a).

The dominant lithology of the formation in the Calvert Cliffs area is quartz
sand and silt, containing lesser clayey intervals and indurated limestone
layers. Two intervals here contain abundant molluscan shells; shells are
sparse in the rest of the formation.

Shattuck (1904) divided the Choptank Formation into a series of zones, num-
bered 16 through 20. As in Shattuck's division of the Calvert Formation,
these are not biostratigraphic zones, but zones locally distinguishable by
lithologic characteristics. Gernant (1970) mapped these "zones" in the
Choptank Formation and redefined them as lithologic members.

The distribution and thickness of the Choptank Formation are considerably
restricted compared with those of the underlying Calvert Formation (figure 9).
However, significant changes in faces are present over this limited area. In
the type area along the Choptank River on the Eastern Shore (figure 3), the
faces are similar to those found on the western shore of Chesapeake Bay
(Gernant, 1970); deposition in shallow-marine environments is suggested by the
relatively low species diversity and low percentage of planktonic specimens
(figure 15). In wells northeast of the type area, subsurface sections con-
tain considerable thickness of thinly laminated, micaceous, brown clay and
gray sand which may be interbedded with shell hash (Gibson, in press a).
These lithologies suggest a deltaic influence in the northeastern part of the
Choptank distributional area (figure 10). The remainder of the formation in
this area consists of blue, shelly, clayey sand, suggestive of more open
marine access at times. Southeast of the type area, Choptank strata in the
Hammond well (figure 4) are composed of shelly sand containing some limy
intervals and small amounts of glauconite; these strata indicate inner-shelf
environments of deposition. The strata in the Hammond well and in other
nearby wells (Gibson, in press a) suggest that the deltaic influence seen in
the north did not extend this far south and that this area was dominated by
open marine, inner-shelf environments.

The Choptank Formation also extends southward into Virginia. In the Oak Grove
core hole (figure 3), slightly diatomaceous clay and silt were correlated by
diatoms with beds 18 and 19 of the Choptank Formation (Gibson et al., 1980).

Beds 16 and 17, however, are missing here, which indicates a probable south-
ward pinching out of the lower part of the formation. Abbott (oral commun.,
1977) indicated the presence southward into Virginia as far as Richmond, of
diatomaceous strata that correlate with the Choptank Formation. Near
Richmond, the lithology is slightly diatomaceous silt and clay, finer than
generally found in the Choptank in the Chesapeake Bay area. In the Richmond
area, the similar lithologies of the Choptank and Calvert make it difficult to
distinguish between the two formations.

Diagnostic planktonic Foraminifera have not been found to date in the Choptank
Formation. The tentative age placement of the base of the Choptank (figure 1)
is-based upon the absence of Choptank diagnostic benthonic foraminiferal spe-
cies in uppermost strata of the Pungo River Formation in North Carolina; these
strata are dated as N.11 in age, which would place the Choptank as post N.11.
The age of the top of the Choptank is also tentative; the only control here is
a K-Ar date of 12.5 m.y. for the overlying St. Marys Formation (Blackwelder
and Ward, 1976). Some indication of a time break between the Calvert and
Choptank Formations is suggested by the appearance of several species of
benthonic Foraminifera in the basal part of the Choptank Formation that are
not present in the assemblagaes in the uppermost part of the Calvert (Gibson,
1962). Shattuck (1904) suggested an unconformity between the two formations;
Gernant (1970) discussed the boundary as a possible unconformity.

ST. MARYS FORMATION

The St. Marys Formation was named by Shattuck (1902) for exposures in St.
Marys County, Maryland, particularly those along the St. Marys River
(figure 3). The formation is probably the least extensive of any of the
Miocene units in the Salisbury Embayment (figure 11). The northernmost report
of this formation or its equivalent was on the basis of subsurface fossils in
southwestern New Jersey (Richards and Harbison, 1942). A few outcrops in
north-central Virginia near Tappahannock on the Rappahannock River (figure 3)
(L.W. Ward, oral commun., 1971) show that the formation extends this far south
(figure 3).

Shattuck (1904) divided the St. Marys into four lithologic zones or beds as
herein used, numbered 21 through 24. Combining thicknesses from the four beds
in the scattered exposures described by Shattuck gives a composite outcrop
thickness of 73 feet for the formation on the western shore of Chesapeake Bay.
The thickness is slightly greater than 100 feet on the Eastern Shore in the
Hammond well (figure 4) and surrounding wells (figure 11).

The beds are dominantly clay, sandy clay, and clayey sand in the type area in
southern Maryland and generally are finer grained and more clayey than the
underlying formations. Many of the clay units are massive and do not contain
megafossils; some clay units have slightly fossiliferous sandy lenses as much
as 1 to 2 feet thick interbedded with the nonfossiliferous clays. Commonly
these shell beds are dominantly composed of a few molluscan species of
Turritella, Nassarius, and Mercenaria. Sand and clayey sand units are locally
abundant, and burrowing, channeling, and crossbedding within the units are
common. In the Hammond well on the Eastern Shore, the strata placed in the
St. Marys (Anderson, 1948) are coarse sand and fine gravel containing shell
fragments and some glauconite (figure 4).

Figure 11. Isopach map of St. Marys Figure 12. Paleoenvironments postulated
Formation and equivalent units, for St. Marys Formation and equivalent
Contours are in feet. units.

Figure 13. Isopach map of Eastover
Formation and equivalent units.
Contours are in feet.

Figure 14. Paleoenvironments postulated
for Eastover Formation and equivalent
units.

The St. Marys strata in the BG&E core hole constitute the lower and middle
parts of the formation, and samples from the corehole are characterized by
either the total absence of foraminifers or low benthonic foraminiferal diver-
sities of 13 species or less and a total absence of planktonic specimens
(figure 15). The BG&E core hole appears to be near the northwestern limit of
marine deposits in the St. Marys, many of the beds being of marginal marine
environments and totally lacking foraminifers.

To the south, near St. Marys City, in the upper part of the formation, the
foraminiferal diversity increases to 15 to 20 species, and planktonic speci-
mens begin to appear in low frequencies. These occurrences suggest more open
shallow-marine environments for part of the strata here, but some of the non-
fossiliferous clays appear to be of marginal marine origin.

The more open marine conditions represented by the shelly strata extend into
northern Virginia, placing the maximum axis of the basin in the southernmost
part of Maryland, farther south than in Calvert and Choptank time. The coarse
sand and fine gravel in the Hammond well suggest that plastic debris was being
carried in by the same delta system that had been building in the area since
middle Calvert time (figure 12).

In the embayment west of the rapidly prograding delta sequence, an area also
undergoing periodic shallowing because of uplift, open ocean circulation was
cut off for significant periods of time, and deposits of restricted marine to
brackish environments were formed, as seen in beds 21, 23, and part of 22.

"UPPER MIOCENE STRATA"

Strata slightly younger in age than those of the St. Marys Formation as
exposed in St. Marys County, Maryland, as well as some of a similar age to
those of the type area were recognized in Virginia by Mansfield (1943).
Mansfield placed all the strata involved into an expanded St. Marys
Formation and divided the formation into three units.

The lowermost unit, called "stratum A" by Mansfield, was questionably corre-
lated by him with the lowest part of the St. Marys in Maryland, beds 21 and
22. Stratum A is an unfossiliferous silty clay found only in northernmost
Virginia, and presumably represents a marginal-marine or nonmarine environment
at the southern extent of the early St. Marys sea.

The overlying unit in Virginia, termed zone 1 by Mansfield, was correlated
with zones 23 and 24 of the St. Marys in Maryland. Zone 1 extends farther
south than Stratum A, reaching approximately the Rappahannock River near
Tappahonnock (figure 3).

Mansfield believed that his highest unit, zone 2, was younger than any fossi-
liferous St. Marys in Maryland, and that zone 2 extended from the Rappahannock
River southward to the James River in southern Virginia. Subsequent work by
the present author has shown a distribution of fossiliferous strata of this
apparent age from the Nomini Cliffs in northern Virginia southward into
northeastern North Carolina along the Meherrin River in the vicinity of
Murfreesboro (figure 3). As the strata at the Nomini Cliffs indicate a

Percent Planktonic
Sample Foraminifera Number of Species of Benthonic Foraminifera H ( S )
Number 10 20 30 40 50 60 70 80 0 5 10 15 20 25 30 35 40 45 50 0.5 1.0 1.5 2.0 2.5 3,0

St. Marya
Formation

Choptank
Formation

Plum Point
Marl Member

Fairhaven
Member

Pungo River
Formation

i

Figure 15. Percentage of planktonic Foraminifera, number of species of

benthonic Foraminifera, and diversity measure of benthonic species
in samples from the Baltimore Gas & Electric Company corehole at
Calvert Cliffs, Maryland (Calvert, Choptank, and St. Marys Forma-
tions) and the Lee Creek mine of Texasgulf, Inc. (Pungo River
Formation).

m NO
mmmm

mmm

I
I

I

shallow-marine environment of deposition, shallower marine or marginal-marine
environments probably continued into southern Maryland and could be repre-
sented by outliers in Maryland, as suggested for Yorktown strata by Stephenson
and MacNeil (1954).

As Mansfield's zone 2 of the St. Marys was considered younger than the type
St. Marys in Maryland, and as it occurs largely in Virginia, the informal term
"Virginia St. Marys beds" was used (Gibson, 1971) to differentiate it from the
type St. Marys. Because the geographic distribution and lithology of the
"Virginia St. Marys beds" are similar to those of the immediately overlying
Yorktown and because they occupy a similar depositional basis within the
Salisbury Embayment, in contrast to the underlying formations, which predomi-
nantly have a more northerly locus of deposition, the "Virginia St. Marys
beds" were placed in the basal part of the Yorktown Formation (Gibson, 1971).
Recently, Ward and Blackwelder (1980) named the "Virginia St. Marys beds" the
Eastover Formation.

The strata of the Eastover Formation consist mainly of greenish-blue clayey
sand and sandy clay, commonly containing abundant mollusk shells. The
thickness as pieced together from outcrop sections is 25 to 75 feet; downdip,
the strata thicken to more than 200 feet in the Norfolk Moores Bridge well
(figures 3 and 13). The lithology remains similar over southern and central
Virginia and even southward into North Carolina and represents deposition in
shallow inner-shelf environments. East and northeast, however, as seen in the
Hammond well, gravelly sand rests on top of the St. Marys Foundation
(figure 4). These strata were considered by Owens and Minard (1979) to be of
Late Miocene age and probably represent a major time of fluvial and deltaic
outbuilding southward into a restricted Salisbury Embayment (figure 14).

The exact age of the Eastover Formation is uncertain because of the absence of
diagnostic species that can be correlated with the various intercontinental
zonations. Planktonic Foraminifera are rare in these beds, and those that are
present belong to long-ranging species. In southeastern Virginia, three
transgressions within the Eastover Formation are indicated by the presence of
three consecutive subspecies of Chesapecten in strata separated by disconform-
able contacts. The amount of time both within and between the transgressive
pulses is uncertain. A radiometric date of 6.46 1 0.15 m.y. for the upper
part of the Eastover was given by Blackwelder and Ward (1976), as well as a
K/Ar date of 4.4 0.2 m.y. on the lower beds of the Yorktown (Blackwelder and
Ward, 1976). Another general upper age limit to the Eastover is given by the
dating of othe lowermost part of the overlying Yorktown Formation as belonging
to planktonic foraminiferal Zone N19 of Blow (Gibson, in press b). This zone
was placed by Berggren and van Couvering (1974) in the lower Pliocene,
although now it has been given a longer range by Vail and Mitchum (1979).

Andrews (1980) placed the uppermost part of the Eastover Formation at
Petersburg, Virginia (figure 3), in the earliest Pliocene on the basis of the
contained diatoms. This correlation is based upon the Pacific diatom
zonation; if this zonation proves to be usable for the Atlantic as well during
this part of the Cenozoic, at least part of the "Upper Miocene" or Eastover
Formation is of early Pliocene age.

ALBEMARLE EMBAYMENT

The lithologies and ages represented in this embayment differ significantly
during part of the Miocene from those found in the Salisbury Embayment to the
north. Although equivalents of the Calvert and upper Miocene strata are found
over at least part of the Albemarle Embayment, equivalents of the middle part
of the Chesapeake Group are not represented in any of the onshore strata. The
Calvert equivalent in the Albemarle Embayment is similar in being largely of
biogenic origin, but it differs in being dominantly phosphatic in composition
rather than diatomaceous and in being of a biogenic composition through a
greater part of Miocene time. The upper Miocene units in both embayments are
composed of plastic sediments, largely clayey sand and sandy clay. The units
in the Albemarle Empayment will be discussed from oldest to youngest.

PUNGO RIVER FORMATION

The Pungo River Formation was named by Kimrey (1964) for the phosphatic units
found only in the subsurface in eastern North Carolina. An artificial expo-
sure is found in the Lee Creek mine of Texasgulf, Inc., near Aurora
(figure 3); pictures of the exposure were shown in Gibson (1967). The
thickness of the formation is commonly 40 to 60 feet in the western part of
its area and increases to more than 100 feet in the eastern (figure 7). The
middle to inner-shelf environments of deposition represented by the formation
at its western limits indicate that it originally was more widespread.

The dominant lithologies of the Pungo River Formation are phosphatic and
diatomaceous clay, fine to medium grained phosphatic sand, phosphatic
limestone, and coquina composed largely of bryozoan and barnacle fragments.
The beds of clay are commonly bright yellow-green, although they may range
from light to dark green. The phosphatic content generally is lowest in the
clay units and highest in the sand units. The sand units are fine to medium
grained, and are composed of quartz and phosphate grains. The phosphatic,
sand-sized grains are cellophane, and, although brown, smooth, and ovate
grains are most abundant, irregularly shaped pieces of bone and teeth also are
common. Phosphate content of the bulk sands commonly ranges from 10 to 21
percent (Kimrey, 1965). Indurated sandy limestone is most common in the upper
part of the formation, whereas thin beds of dolomitic siltstone are found at
scattered intervals throughout the formation.

At the Lee Creek mine, the lower and middle parts of the Pungo River Formation
consist of two phosphatic sand sequences separated by a dolomitic siltstone,
followed upward by interbedded phosphatic sand and carbonate layers, and then
by sandy carbonate and coquina. In this area, the carbonate units are 12 feet
thick and compose approximately 25 percent of the section. South of the mine,
approaching the New Bern high (figure 2), the Pungo River Formation changes to
a largely calcareous faces. The lithologies include calcareous clay,
limestone, phosphatic limestone, and bioclastic debris. This largely car-
bonate upper faces was named the Bonnerton Member of the Pungo River
Formation by the author (Gibson, in press a). As shown by Kimrey (1965,
figure 6), the Bonnerton Member reaches northwestward as far as Bonnerton
(figure 3), apparently being absent west of there because of erosion. Along
the New Bern high at the southern part of its distributional area, the member

extends at least as far east as Great Lake (figure 3), where it was found in a
core hole (Gibson, in press a).

In the Lee Creek mine and general vicinity, the Pungo River strata below the
Bonnerton Member are dominantly composed of phosphatic sand, sometimes clayey,
and lesser thicknesses of moderately phosphatic to nonphosphatic clay; in
addition, thin beds of diatomaceous clay, dolomitic silt, and phosphatic
limestone are found. These beds compose the major part of the Pungo River
Formation at the Lee Creek mine and their dominance increases to the northeast
as seen in well ACCO-8-64 (figures 3 and 16). These generally high phosphatic
sand units were named the Belhaven Member of the Pungo River Formation by the
author (Gibson, in press a) because of their abundance northeast of the Lee
Creek mine. This member overlies strata of Oligocene age in the southern part
of the embayment and commonly overlies older beds such as the Castle Hayne
Formation of Eocene age in the central part and various units of Paleocene and
Eocene age in the northern part.

The upper part of the Pungo River Formation in the Lee Creek mine was dated by
means of planktonic Foraminifera as belonging to Blow's Zones N.8 or early N.9
(Gibson, 1967, and in press a), making them latest Early Miocene or earliest
Middle Miocene in age (Berggren and van Couvering, 1974; Vail and Mitchum,
1979). In the central part of the distributional area of the Pungo River
Formation, northeast of the Lee Creek mine, Abbott and Ernissee (in press), by
the use of diatoms from a core hole, were able to correlate the lower part of
the Pungo River Formation with planktonic foraminiferal Zones N.8-N.9, and the
upper part with Zone 11. Gibson (in press a) found that the Pungo River beds
in the northern part of the Albemarle Embayment near Gatesville, North
Carolina (figure 3), correlated with planktonic foraminiferal Zone N.11; these
beds rest upon strata of Paleocene age. These occurrences suggest that strata
of at least two ages make up the Pungo River, strata of both ages being super-
posed in the central part of the embayment but only beds of the younger, Zone
N.11, age being found in the northwestern part of the embayment. Northeast of
Gatesville, in the Moore Bridge well near Norfolk at the southeastern end of
the Salisbury Embayment, the Pungo River strata can be dated only as of Zone
N.8-N.9 age. The two transgressions either had differing areas of coverage or
deposits have been selectively removed by erosion. A third transgression
covering the southern part of the embayment is likely.

East and northeast of the Lee Creek mine, the lowermost strata of the Pungo
River Formation in several core holes contain benthonic Foraminifera, which
are characteristic of the "Silverdale beds" of latest Oligocene age in North
Carolina (Gibson, unpub. data). The upper age limits of these benthonic spe-
cies are not clearly known, as these species occur in the uppermost Oligocene
strata but not in the Pungo River strata of latest Early Miocene (Zone
N.8-N.9) age; therefore, they could have become extinct anytime during the
latest Ologocene or Early Miocene. The author observed one of the most abun-
dant and characteristic species in samples collected by Paul Huddlestun from
the Torreya Formation in Florida which is dated by M. E. Hunter (oral commun.,
1980) as probably being of early Miocene (Zone N.5) age. In addition, a pecten
recovered from a core hole in the lowermost part of the Pungo River Formation
northeast of Lee Creek is similar to a species characteristic of the
"Silverdale beds". Although not the same species, its affinites are with the

earlier species and not to any found in the Pungo River and Calvert
Formations. The presence of the benthonic Foraminifera and pecten suggests
that the lowermost beds of the Pungo River Formation in the southern part of
its range are Early Miocene in age, somewhere between the latest Oligocene age
and the "Silverdale beds" and the late Early Miocene age typical of the upper
part of the formation in these sections. These strata are provisionally
placed in the middle Early Miocene (figure 1) at a level similar to that of
the Torreya Formation. These biostratigraphic data indicate that probably at
least three periods of phosphate deposition took place within the time repre-
sented by the Pungo River Formation.

Deposition of the phosphatic beds of the Pungo River Formation appears to have
been in deeper water than those in which the younger Miocene deposits in the
Albemarle Embayment and the exposed Miocene strata in the Salisbury Embayment
were formed. Depths of approximately 100 m for the deeper units are suggested
by the high planktonic percentages and high species diversity of the bentho-
nics (figure 15) (see also Gibson, 1967, 1968). Postulated water depths are
considerably shallower for the upper carbonate beds of the Bonnerton Member
(sample H, figure 15) as indicated by considerably lower planktonic percentage
and species diversity.

UPPER MIOCENE STRATA

Beds of Middle and Late Miocene age either
have a more restricted distribution in the
Albemarle Embayment than in the Salisbury
or are not found at all in the Albemarle.
Equivalents of the Choptank or St. Marys
Formations are not known in the onshore
part of the Albemarle Embayment.
Equivalent strata to the upper part of the
"upper Miocene" Eastover Formation in the
Salisbury Embayment are found in the
northern part of the Albemarle Embayment
and were placed by Ward and Blackwelder
(1980) in the Eastover Formation.
Although these strata crop out in the
vicinity of Murfreesboro on the Meherrin
River and are found in the subsurface to
the east, they do not occur in the Lee
Creek mine and have not been identified in
the subsurface in the southern part of the
empayment.

Well Acco-8-64
Depth (ft)

194 a

ET--~-=--I
O a
B

,Oe
LI

P~(LI.
ttl"'

~~~~-~-~-
.9\
%,a:b o .o'.
''''
ri

Yorktown Formation

gre nirnmyC.y,
.bund.nt ha.ll.
green hlly and
undy lim-ton.
lirnetons and
phosphatic sand
ghr.philo ..nd
phosphatic "n

grarnih-black
phoaphatic shd

Pungo River Formation

dark green clayey
phosph.ticaand

sandy do-toot
greenish-!bdOk
phoaphat'c snd, hall.
greaniah- "ak pub
phO.phatio ..nd .hofii
nlayly phosphatro Snd
dakpiahtcsnd

Castle Hayne Limestone

Figure 16. Geologic column for
well Acco-8-64, near Belhaven,
Beaufort County, North Carolina.

In contrast to the lower Miocene deposits, the upper Miocene strata are litho-
logically similar in both embayments. In the Albemarle Embayment, the beds
are largely clayey sand together with a lesser amount of sandy clay and are
similar to the clayey sand of the Eastover to the north. Clastic sediments
compose the entire section; no chemical or biochemical sediments, so typical
of the underlying units, occur in the upper Miocene strata. Molluscan shells
are common throughout most of the beds.

2 1

112

Agewise, the strata are placed by the ages determined in the Salisbury
Embayment as no diagnostic planktonic Foraminifera have been found in the
beds. Only the uppermost zone of the three Chesapecten zones of the Eastover
Formation has been identified at Murfreesboro, and this zone is the same as
the one found in the strata tentatively placed in the early Pliocene by
Andrews (1980). Therefore, all the Eastover Formation in the Albemarle
Embayment may prove to be of Pliocene age.

SUMMARY

Miocene strata in the Albemarle and Salisbury Embayments show an episodic
depositional pattern and significant periods of nondeposition. Although much
of the initial deposition in both embayments was of chemical or biogenic ori-
gin, the entire later column in the Salisbury Embayment and the strata of late
Miocene age in the Albemarle Embayment are of plastic origin. The overall
increase in clastic material in both embayments through the Miocene and
Pliocene indicates an increase in the amount of plastic debris reaching the
basin, particularly in comparison with the earlier deposits in the Cenozoic.
The Paleocene, Eocene, and Oligocene strata are dominantly composed of highly
glauconitic sands in the Salisbury Embayment and glauconitic containing signi-
ficant amounts of carbonate material in the Albemarle Embayment. The reason
proposed for this increase in the amount of plastic debris and also for the
deltaic outbuilding in the northern part of the Salisbury Embayment is an
uplift of the Appalachians. The increase in plastic debris is seen earlier in
the Salsibury Embayment and effectively ended the two periods of diatomaceous
deposition in the Early Miocene. Phosphatic deposition by chemical and bio-
genic means continued later in the Miocene in the Albemarle Embayment and indi-
cates that the northern source areas were uplifted earlier and that the uplift
later moved south. Superimposed on this uplift pattern are the changes in the
basin shape and extent, presumably by fault movement, that cause the southward
movement during the Miocene of the locus of deposition in the Salisbury
Embayment and the exclusion of marine transgressions from the onshore part of
the Albemarle Embayment during the Middle Miocene.

ACKNOWLEDGMENTS

The author benefited from stimulating discussions with James Miller, U.S.G.S.
Elizabeth Funk, U.S.G.S., did the illustrations and many other tasks that
greatly aided in completion of the manuscript.

DIATOM BIOSTRATIGRAPHY OF THE CHESAPEAKE GROUP, VIRGINIA AND MARYLAND

William H. Abbott
Mobile Exploration and Producing Services, Inc.
Dallas, Texas

ABSTRACT

Examination of classical Miocene units of the Chesapeake Group, Virginia and
Maryland, for diatoms produced excellent results in the Calvert and Choptank
Formations, but nothing in the St. Marys Formation. Sections along both
Chesapeake Bay and Maryland and Virginia rivers correlated well with Abbott's
(1978) Miocene diatom zonation. Correlation of the lower Fairhaven Member of
the Calvert to Blow's N6/N7, the Plum Point Member of the Calvert to Blow's
foraminifera zones N8/N1O and the Choptank to N11/N12 are confirmed. The
Chesapeake Group along Chesapeake Bay can be correlated to subsurface sections
to the north in Delaware and out on the shelf. These transects can then be
tied into the B-3 C.O.S.T. well on the Atlantic slope.

INTRODUCTION

The Chesapeake Cliffs of Maryland represent some of the thickest and best
exposed sections of Miocene of the U.S. Atlantic coast. These sections have
been studied by paleontologists for hundreds of years. Of the three for-
mations that make up this section, two, the Calvert and Choptank, are very
diatomaceous. The St. Marys, the youngest of the three units, seems to be
devoid of diatoms. These three formations were placed in the Chesapeake Group
by Dall and Harris (1892) and consist of a series of sands, silts and clays
containing abundant shell and bone materials, as well as diatoms.

W.B. Rogers was among the first to find diatoms in the Chesapeake Group in
1841, and he sent samples to J.W. Bailey, who was then a professor at the U.S.
Military Academy. Bailey published in 1842 on diatoms from the Richmond sec-
tion and from sections along the Rappahannock River in Virginia. He also sent
samples to C.G. Ehrenberg, and Ehrenberg published a manuscript in 1843 on
North and South American microfossils including diatoms from the Chesapeake
area. Ehrenberg later published a paper (1844) on microfossils, including the
diatoms, from the Richmond, Virginia area. Grunow in Van Heruck (1885)
described taxa from the Chesapeake Group of both Maryland and Virginia. Boyer
(1904) did a systematic study of the diatoms from the Calvert of Maryland.
This was later followed by Lohman (1948), who looked at Chesapeake Group
diatoms from the Hammond Water Well. Following Lohman, very little was done
with the diatoms until a thesis by Cavallero (1974) in which he attempted a
zonation of the Chesapeake Group based on diatoms. This work was followed by
Andrews (1976), who did a taxonomic study of the diatoms in the Choptank.
Andrews (1978) also set up a regional zonation based on diatoms from the
Calvert and Choptank Formations. Abbott (1978), in his Atlantic diatom zona-
tion, included a breakdown of the Calvert and Choptank along Chesapeake Bay.

The lithologies of the Chesapeake Group have been the subject of a number of
descriptions and revisions throughout the years. Within this century Shattuck
(1904) did the first real lithological breakdown of the units by separating
out in the section twenty-four individual lithologic units. Gernant (1970)

formally assigned names to these units in the Choptank Formation. Andrews
(1978) preferred to relate to Shattuck's old lithological units and used the
term Miocene Lithologic Units for the section. Andrews' Miocene Lithologic
Units or M.L.U. can be related to the formations in the following manner:
within the Calvert, M.L.U. one through three represents the Fairhaven
Diatomaceous Earth Member; M.L.U. four through fifteen represents the Plum
Point Marl Member; M.L.U. twenty through twenty-four represents the St. Marys
Formation. Both Abbott (1978) and Andrews (1978) suggested that the lower
portion of Shattuck's zone three, or the lower portion of the Fairhaven
Diatomaceous Earth Member, was actually far older than the bulk of the
Fairhaven Diatomaceous Earth Member. This older section is Early Burdigalian
in age instead of the Early Langhian age found in the upper portion of this
unit. A hiatus therefore occurs with M.L.U. three. This lower section repre-
sents the oldest diatoms found in the Chesapeake Group. Andrews (1978) also
suggested that the Early/Middle Miocene boundary occurs in the Chesapeake
Group at or near M.L.U. twelve, and that the Langhian/Serravillian boundary or
foraminifera zones N9/N1O occurs approximately at the M.L.U. fourteen/fifteen
boundary. Gibson (1967) suggests that the M.L.U. zone ten is approximately
foraminifera zone N8/N9 in age. J.E. Hazel gives a foraminifera zone N17
correlation for the top of the St. Marys in Andrews (1978).

In this paper I will relate in a more detailed way than has been done pre-
viously Abbott's (1978) Miocene Zonation (Table 1) to the Chesapeake Group,
not only in Maryland but also in Virginia. Since the embayment within which
these sediments are deposited, the Salisbury Embayment, is part of the margin
of the Baltimore Canyon Trough, I will also attempt to relate these zonations
or correlations to other wells and cores in the area, as well as to the B-3
C.O.S.T. well out on the Atlantic Shelf, within the Baltimore Canyon trough.
This comparison should help in the correlation of the Miocene throughout the
area.

RELATIONSHIP OF THE CALVERT/CHOPTANK FORMATIONS
TO THE BLOW'S FORAMINIFERA ZONATION

The diatom zones used in this study have been correlated with Blow's (1969)
foraminifera zones based on data from Abbott (1978) (figure 1). Diatom Zone
I, Actinoptychus helipelta zone, is equivalent to Blow's Zones N6 and N7; Zone
II, Delphineis ovata partial range zone, is equivalent to Blow's N8 to N9;
Zone III, DelphiTeis ovata/Delphineis penelliptica concurrent range zone is
equivalent to approximately the top of Blow's N9 to the base of N10; Zone IV,
Delphineis penelliptica partial range zone, is equivalent to Blow's N10 and
possibly the base of N11; Zones V and VI, Delphineis penelliptica/
Coscinodiscus plicatus concurrent range zone and Coscinodiscus plicatus
partial range zone, are equivalent to Blow's foraminifera zone N11 and N12.
When the diatom correlations of the Calvert and Choptank are then related to
Blow's zonations we find that the Fairhaven member, or Miocene Lithologic
Units one through lower three, of the Calvert is equivalent to approximately
N6/N7 and upper unit three or Upper Fairhaven member is equivalent to N8/N9
with an apparent unconformity within the Fairhaven. The Plum Point Marl
member of the Calvert or units four through fifteen, are equivalent to Blow's
N9 to possibly N11. The Choptank Formation or units sixteen through twenty,
is equivalent to Blow's N11/N12. Figure 2 shows the range of stratigraphi-
cally important diatoms and silicoflagellates from the Baltimore Gas and
Electric core and adjacent outcrops along the Chesapeake Bay as determined by

TABLE 1

ABBOTT'S (1978) ATLANTIC MIOCENE DIATOM ZONES

Zone 1, Actinoptychus heliopelta Concurrent Range Zone
Definition. The base is not defined. The top is defined
by the extinction of the diatom Actinoptychus helio-
pe/ta Grunow.

Zone II. Delphineis ovata Partial Range Zone
Definition: The base is defined by the first appearance
of the diatom Delphineis ovata Andrews and the
absence of Actinoptychus heliope/ta. The top is defined
by the first appearance of the diatom Delphineis
pene/hptica Andrews.

Zone III, Delphineis ovatalDelphineis penelliptica Concurrent Range
Zone
Definition: The base is defined by the first appearance of
Delphineis pene//iptica and the top by the last appear-
ance of Delphineis ovata.

Zone IV, Delphineis penelliptica Partial Range Zone
Definition: The base is defined by the last appearance of
Delphineis ovata: the top is defined by the first appear-
ance of the diatom Coscinodiscus plicatus Grunow.

Zone V, Delphineis penelliptica/Coscinodiscus plicatus Concurrent
Range Zone
Definition. This zone is defined at the base by the first
appearance of Coscinodiscus plicatus and is defined at
the top by the last appearance of Delphineis penellip-
tica.

Zone VI. Coscinodiscus plicatus Partial Range Zone
Definition: The base is defined by the last appearance
of Delphineis pene//iptica and at the top by the last
appearance of the silicoflagellate Distephanus stau-
racanthus Ehrenberg.

MIOCENE O

EARLY MIDDLE

C ,cn
os os <
"I >zo
) I a1
I o> z
0Z
I

.I IH (P II 't II -
o o .c
> "0 0g"

SI 500

I I r I ZZ

I Z rNz :
I OZ '
v I 00
-. wo io

0 (/

N 0

ALVERTCHOPTANK
Z Z 9 2 z -
I-M I
I III

WO m

I Z 2

N 0 n

CALVERT CHOPTANK

Correlation of zones used in this study.

Figure 1.

S $ S S

'a 1 I" I 1 a S -

SDEPTH I" I I t t DE T

'a "a '-a '-a '-a a ~ ~ ('a 2a a ~ o N Na Na P "a Na(a) 0I

* FOUND IN BG 8 E
O --FOUND IN OUTCROP
REPORTED BY ANDREWS 1979
O - CAVALLERO 1974

Figure 2. Ranges of selective diatom species found in B.G.& E. core and

outcrops in Maryland and Virginia.

ZONE 21

ZONE20

ZONE19

ZONE 18

ZONE 17

ZONE 16

ZONE 15

ZONE14

ZONE13

ZONE 12

54 ZONE II

60
m' ZONE 5

ZONE4
66

72

UPPER
3
7e ZONE

84

9.0

96
ZONE I

EOCENE

24

36

42

48

0 D54

60

66

72

S7

84

90

96

ZONE
IV

ZONE
III

ZONE
I I

Cavallero (1974), Andrews (1978) and the author. Of the species listed,
Annellus californica's range is restricted to Blow's N8/N9 zones (Burckle,
1975). Coscinodiscus plicatus and C. yabei have their base in N11 (Ernissee,
Abbott and Huddleston7, 1977T ; De~ticular hustedtii has a base in N11,
Delphineis ovata ranges within Zones N8/N9, D. penelliptica within Zones
N9/N11 (Abbott, 1978). These ranges tend to confirm the correlation of the
Calvert and Choptank Formation already discussed and also agree with the dates
proposed by Andrews (1978) and Gibson (1967) for specific sections of these
two formations.

DIATOM ZONES IN VIRGINIA AND MARYLAND

Abbott's (1978) Atlantic Margin Zones correlate well in the lower two for-
mations of the Chesapeake Group. The Baltimore Gas and Electric core proved
to be a good base for this correlation in that it contains all of the Atlantic
Margin Zones, except Zone 1. If we examine a section along Chesapeake Bay
(figures 3 and 4) we find a good correlation back to the B.G. and E. core.
In this section we can see that Zone 1 is missing. Zone 1 is only found in
the upper portions of the bay where it appears to be pinching out. As one
travels along the bay from north to south the units become younger so that the
entire section can be sampled horizontally. This is most fortuitous in that
most of the cliffs are nearly vertical. The diatom zones dip approximately
1.3 km between Randle Cliff and the Baltimore Gas and Electric core or 37 m
over an area of 28.5 km. Looking at zonal correlations along the Patuxent
River to the west of the Chesapeake Bay Cliffs, the seaward dip is also
apparent (figure 5).

Data points along rivers in Maryland and Virginia (figure 5) are generally
from samples taken at river level unless otherwise indicated in the figure.
The distribution of the Abbott (1978) zones may in some cases be suggestive of
small tectonic displacement but closer sampling and more transits are needed
to substantiate that possibility.

The zonal determination at Richmond (Abbott's Zone VI) is of interest in that
this section has always been considered part of the Calvert Formation
(Fairhaven Diatomaceous Member). It is apparently Calvert lithology (Buck
Ward, personal communication, 1980), but has one of the youngest Miocene
diatom assemblages found in the Salisbury Embayment. This assemblage is above
the last appearance of the silicoflagllate Distephanus stauracanthus and is
thus late Middle Miocene rather than the late Early Miocene date found in the
Fairhaven Member of the Calvert Formation.

CONTINUATION OF THE DIATOM ZONES ACROSS BALTIMORE CANYON TROUGH

The Baltimore Canyon Trough extends over an area more than 500 kilometers,
subparallel with the U.S., Middle Atlantic Region between Long Island and Cape
Hatteras, with sediments up to 14 kilometers thick (Poag, 1979). The
Salisbury Embayment beneath the Coastal Plain of Virginia, Maryland, Delaware
and New Jersey is a western extension of the Baltimore Trough. Poag (1978,
1979) suggests that much of the Miocene deposited in the trough (figure 6) is
deltaic based upon foraminifera and geophysical profiles. In examining the
Miocene diatoms within this trough I have attempted to continue the correla-
tion of the Abbott (1978) diatom zones (figure 7) from the Baltimore Gas and
Electric core in Maryland, northeast in the long direction of the trough,

Figure 3. Location of points along Chesapeake Bay, Maryland, from which
points were taken for stratigraphic classification.

CALAMS POINT OF B.G. E. N. of N.CALVERT KENWOOD
RUN ROCKS CORE B.G.&E. BEACH BEACH

GOVERNOR'S SCIENTIST PARKER
RUN CLIFFS CREEK

CAMP S.of CAMP RANDLE
KAUFFMAN ROOSEVELT CLIFF

2k.m 0.Skm 3km 05km 7km 5 5km

Figure 4. Stratigraphic cross section along the Chesapeake Bay cliffs.

c

c

SCAL E
---~-U------
STATUTE
MILES

Figure 5. Abbott's (1978) diatom zones applied to outcrop sections in

Maryland and Virginia. All samples taken at river or bay water

level unless indicated by footage.

MASS.

d

N.Y.

A.M.C.O.R. 6011

N.J.

S B3
/ OC.O.S.T.

/ INDEX MAP
ATLANTIC /
OCEAN N/ 0 o0 40

I KILOMETERS
e6 ---WATER
o DEPTH (200m)
CI\

Figure 6. Location of cross section between B.G.& E. well, Dover well, and
A.M.C.O.R. 6011 as well as location of the B-3 C.O.S.T. well.

B.G.a

METERS
50-

40-

30-

20

10

SEA LEVEL-

-10-

-20-

-30-

-40-

-50-

-60-

-70-

-80-

-90-

-100-

-110-

-120-

-130-

-140-

-150-

-160-

-170I

-180-

-190-

-200-

'1lo

-220

-230-

-240-

-250-

-260-

-270-

BALTIMORE
ELECTRIC

DOVER
WELL

PLEIST.

LOWER
MIOCENE
?

EOCENE

A.M.C.O.R.
6011

\\

\\
\\ PLEIST.

\\

\ \
\ \
\
\\

\\ \ PLOCENE
\ \ \\
\\ \\

\\ \
\\ \\
\ \\
\ \ \\
\\ \\ I
\\ \\
\ \ \\
\\ \\

\\ \\
\\ \ I I
\\ i
\
\\

\\

LOWER
MIOCENE
7

METERS
BELOW
SEALEVE
1207

-1250-

-1300-

-1350-

-1400-

-1450-

-1500

-1550-

-1600-

-1650-

B-3
C.O.S.T.
WELL

VI

V

IV

III

I I

LOWER
MIOCENE

OLI

OLIGO.

Figure 7. Cross section between B.G.& E. core and A.M.C.O.R. 6011 along
with section from B-3 C.O.S.T. well.

through Dover, Delaware and then to AMCOR 6011. A correlation was then
attempted utilizing the diatom zones across the trough to the B-3 C.O.S.T.
well.

In going from the Baltimore Gas and Electric core to the Dover Air Base water
well (figure 7) it can be seen that the trend of the southern dip that was
noted in Chesapeake Bay continues into Delaware with the younger diatom zones
or diatom Zones IV through VI missing. The youngest of the diatom zones
encountered in the Dover well is diatom Zone III. There are sediments above
this zone that are potentially Miocene but no diatoms were encountered. Also,
it can be noted that in the Dover well that the Zone I diatom zone thickens
from its feather edge on the upper Patuxent River in Maryland. There is also
a possible Lower Miocene section beneath Zone I, between that zone and the
Eocene in the Dover well. Some of this section may be Oligocene. No sili-
ceous fossils were encountered to assist in dating that section. In going
farther northeast from Dover to AMCOR 6011, which is an offshore core, it was
found that the dip of the unit reversed, with diatom zones now dipping to the
northeast as well as seaward. In this core we encounter diatom Zones I
through IV, with younger units or what would principally be the Choptank
Formation in Virginia and Maryland, missing. Again, there is the possibility
of a Lower Mioicene section beneath Zone I, but whether or not this is all
Miocene cannot be ascertained. In correlating across the trough towards the
B-3 C.O.S.T. well, we find that the younger diatom zones are present and in
fact have a complete section of Abbott's (1978) zones from I through VI, with
the V/VI zonal area being quite thick, greater than 150 meters. The other
zones vary somewhat in thickness as related to the nearshore cores. In the
B-3 C.O.S.T. well there is a possible Lower Miocene section beneath Zone I, in
which no microfossils were encountered.

CONCLUSIONS

Utilization of the Abbott (1978) diatom zones in the area show that they can
be used to better define the stratigraphy of the Chesapeake units in the
Maryland and Virginia area. This zonation eliminates many problems that had
existed in the area, such as the correlation of the Richmond Miocene units
with the Fairhaven diatomite. The Fairhaven is equivalent to Zones I and II,
while the Richmond deposits are equivalent to Zone VI. The diatoms should
greatly improve the stratigraphy of the Chesapeake Group. Also the diatoms
confirm correlation of the Calvert Formation with foraminifera zones N6/N7
(lower Fairhaven) and N8/N10 and the Choptank to zones N11/N12.

THE EFFECT OF PREDATION ON MIOCENE MOLLUSC POPULATIONS
OF THE CHESAPEAKE GROUP

Patricia H. Kelley
Department of Geology and Geological Engineering
University of Mississippi
Oxford, Mississippi

ABSTRACT

The high species diversity which characterizes the Chesapeake Group mollusc
fauna resulted, in part, from intense predation upon the most common taxa.
Examination of four of the most common Chesapeake Group mollusc genera indica-
tes that predation represents a significant cause of mortality. Predation
rates within the genera Astarte, Eucrassatella, Anadara, and Turritella range
from 13% to 57%. Naticid gastropods, primarily Lunatia heros but also
Polynices duplicatus, contributed significantly to mortality at nearly all
sizes. Predation was somewhat greater at lower size ranges, but was an
important factor for all but the smallest and largest size categories.

The predator was selective with respect to prey taxa, preferring species of
Astarte, Eucrassatella, and Turritella over Anadara. Drill site preference is
also apparent; 14 of 26 Astarte valves examined for drill site display boreho-
les within the central part of the shell. The predator was also preferential
with respect to sex of Astarte specimens drilled. The smaller, uncrenulated
(male) shells were more susceptible to drilling than larger, crenulated
(female) specimens. No selectivity occurred, however, with respect to right
or left valve. Within Astarte, a significant correlation exists between prey
size and drill hole size; large predators apparently chose large prey.

Intensity of predation varies not only among genera, but also within genera
through time. Within a taxon, predation is comparable for the Calvert and
Choptank Formations. All genera exhibit much lower predation rates within the
St. Marys Formation. The increased gastropod diversity of the St. Marys
Formation may have provided alternative naticid food sources, thereby
decreasing predation upon the taxa examined.

INTRODUCTION

The Miocene Chesapeake Group strata of Maryland are well known for their
remarkable faunal abundance and species diversity. The fauna is largely
molluscan; of 624 species described in the massive Maryland Geological Survey
Miocene volume (1904), 408 are molluscs. Gastropods represent 52% and
bivalves 46% of the Chesapeake Group mollusc fauna, though actual proportions
of these two classes vary among the three Chesapeake Group formations.

Several factors have been suggested to explain such cases of unusually high
species diversity. An abundance of resources, such as space or primary pro-
ductivity, may promote diversity (Brown, 1975). Levin and Paine (1974)
suggested that spatio-temporal heterogeneity may increase diversity.
Environmental stability may foster subdivision of niches, resulting in high
species diversity (Hessler and Sanders, 1967; Sanders, 1969; Slobodkin and
Sanders, 1969).

5
L

0 5mi.
I , I

p

^c

Figure 1. Map of localities collected, west shore of Chesapeake Bay,
Maryland. Locality 1 = Camp Roosevelt, 2 = Willow Beach, 3 = mouth
of Parker Creek, 4 = Governor Run, 5 = Drumcliff, 6 = Camp Conoy, 7
= Little Cove Point, 8 = Langley's Bluff, 9 = Calvert Cliffs State
Park, 10 = mouth of Hellen Creek, 11 = Breedens Point, 12 =
Chancellor's Point, 13 = Windmill Point, 14 = Matoaka Cottages, 15
= Plum Point, 16 = Randle Cliffs Beach, 17 = Kenwood Beach, 18 =
Scientists Cliffs, 19 = Sotterly Point, 20 = Calvert Beach.

k'.

Predation has also been cited frequently as a diversity regulating mechanism.
In general, the distribution of species in nature is irregular; most com-
munities are dominated by one or a few extremely abundant species, with addi-
tional species represented by much smaller populations. If a predator is
introduced into such a community, it will tend to select the more common taxa
as prey. By limiting the abundance of dominant species, predation thus pre-
vents monopolization of resources by common forms, causing diversity to
increase. Case studies indicate such effects occur for herbivores preying on
plant communities in both terrestrial (Harper, 1969) and marine environments
(Paine and Vadas, 1969). Such diversity patterns are well documented for
carnivore-herbivore interactions in the marine environment (Paine, 1966, 1971;
Porter, 1972, 1974).

A preliminary assessment of predation rates within the Chesapeake Group
(Kelley, 1979) indicated gastropod drilllholes characterize one-fourth to one-
half of all specimens in many samples of the most common genera. The present
study determines predation rates for four dominant Chesapeake Group mollusc
genera through time. I also examine predator selectivity with respect to
prey size, taxonomy, drill site, sex, and valve.

MATERIALS AND METHODS

The genera Astarte, Eucrassatella, Anadara, and Turritella comprise four con-
sistently abundant molluscs of the Chesapeake Group. Astarte and
Eucrassatella are heterodont bivalves belonging to the superfamily
Crassatellacea. Anadara is an arcid bivalve, while Turritella is a filter-
feeding mesogastropod.

In order to examine drilling rates through time, I collected three species of
Astarte, representing a single lineage (Blackwelder, written communication).
A. cuneiformis Conrad is abundant within the oldest unit of the Chesapeake
Troup, the Calvert Formation (particularly Shattuck's "zones" 10, 12, and 14).
A. thisphila Glenn characterizes Choptank Formation zones 16 and 17, while the
youngest species, A. perplana Conrad, occurs in the St. Marys Formation. I
also collected three species of Eucrassatella. The genus is represented by E.
melina (Conrad) in Calvert zones 10, 12, and 14, and in the Choptank Formation
by E. turgidula (Conrad) of zones 16 and 17 and E. marylandica (Conrad) of
zone 19.

Anadara subrostrata (Conrad) is found only in zone 10 of the Calvert
Formation, but A. staminea (Say) is common throughout the Choptank Formation.
Specimens of Anadara from the St. Marys Formation have usually been referred
to A. idonea (Conrad). Blackwelder (written communication) has suggested that
specimens from St. Marys zone 22 (Little Cove Point Unit of Blackwelder and
Ward, 1976) should be placed in a new species; however, quantitative analysis
(Kelley, 1979) suggests that retention within A. idonea is appropriate.

Although several species of Turritella occur in the Maryland Miocene deposits,
only T. plebeia Say is particularly common. This species is present in the
Calvert Formation, but is less abundant and more poorly preserved than at
stratigraphically higher levels. Collections employed in this study are from
the Choptank Formation zone 17, and St. Marys zones 22 (Little Cove Point
Unit) and 24.

I obtained samples from twenty localities on the West Shore of Chesapeake Bay
(figure 1). Most collecting sites are from the Calvert Cliffs area along
Chesapeake Bay, but four Choptank localities occur along the Patuxent River or
its tributaries. Two samples from zone 24 were collected at exposures along
the St. Marys River. In addition, specimens from the Carol Jones collection
of the Museum of Comparative Zoology at Harvard University supplemented my
samples of Astarte and Eucrassatella.

In the case of bivalves, I measured shell length of each specimen, using
Mitutoyo dial calipers directly readable to 0.05 mm. I recorded the presence
or absence of a drillhole, and whether the specimen was a right or left valve.
The presence or absence of marginal crenulations is apparently diagnostic of
sex in Astarte (Kelley, 1980); this feature was also recorded. I measured
264 specimens of Eucrassatella. Samples were larger for Astarte and Anadara,
consisting of 540 and 508 valves, respectively.

In addition, a sample of 26 drilled specimens of Astarte cuneiformis from zone
10, Willow Beach (locality 2), was examined in greater detail. I measured the
outer diameter of each drillhole in order to test for a correlation between
prey size and hole size (and thus, indirectly, predator size). I also noted
drill site with respect to a 9-quadrat grid, established by subdividing speci-
mens lengthwise into three sectors (anterior, middle, and posterior) and dor-
soventrally into 3 sectors (dorsal, middle and ventral).

I used Vernier calipers to measure Turritella plebeia shell height, from aper-
ture to apex. Presence or absence of a drillhole was also noted for the 416
specimens examined.

PREDATION RATES

Naticid gastropods were the predators responsible for drilling the four genera
studied, as evidenced by drillhole morphology. Naticid and muricid gastropods
are major drilling predators of Cenozoic molluscs. The two families differ,
however, in drilling habits and in the nature of the holes drilled. Muricids
are epifaunal in habit, and attacks are usually confined to surface-dwelling
prey (Reyment, 1967). On the other hand, naticids are burrowers and thus prey
most heavily upon infaunal molluscs. While muricids drill straight-sided,
cylindrical boreholes, naticids produce truncated parabolic boreholes
(Carriker and Yochelson, 1968). A combination of chemical secretion and radu-
lar rasping is employed in drilling.

Drillholes in the Chesapeake Group material I have examined are naticid in
origin, possessing the parabolic shape. A naticid origin is consistent with
the infaunal life modes of the taxa studied. Two naticid gastropods, Lunatia
heros (Say) and Polynices duplicatus (Say), are common constituents of the
Maryland Miocene fauna. Lunatia heros is an especially prominent gastropod,
ranking second only to Turritella pTlebeia in abundance (Gernand, 1970). It is
particularly numerous in the Choptank Formation.

Predation rates for the gastropod Turritella are calculated as the percentage
of the entire sample represented by drilled individuals. Such calculations
indicate the importance of predation relative to other causes of mortality.
The results obtained may underestimate actual drilling frequencies; drillholes
may weaken a shell, decreasing its preservability (Dudley and Vermeij, 1978).

I assumed the number of valves collected per pelecypod taxon represents twice
the number of individuals present. Each drilled valve, however, represents
one individual killed. Thus, I computed predation rate as 2D/N, where D is
the number of drilled specimens and N is the number of valves. Drillholes may
influence postmortem transport of specimens, affecting calculated predation
rates (Dudley and Vermeij, 1978). A lack of breakage, abrasion, and hydraulic
sorting indicates postmortem transport was not a significant factor for the
material I have examined. Predation rates calculated in this way should
therefore be an unbiased estimate of the contribution of predation to bivalve
mortality.

Average predation intensity within the genus Astarte is 41.8%, ranging from
13.3% for A. perplana to 50.6% for A. thisphila. Predation contributed signi-
ficantly to mortality at nearly alT sizes of A. cuneiformis and A. thisphila
(figure 2). Maximum predation rates occur wiThin the middle size ranges of
the species (18-28 mm for A. cuneiformis and 14-24 mm for A thisphila). Size
selectivity was not extremely important, however, as the sTze distribution of
bored shells is not significantly different from that of the total population.
Comparisons of size-frequency distributions of the total population and the
drilled population were made using chi-square methodologies. For A. cuneifor-
mis, X2=12.13 with 8 degrees of freedom. Both results are nonsignifTcant.

Nevertheless, some protection from predation does seem to be afforded by
reaching larger sizes. Predation rates decrease at sizes above 28 mm for both
species. Intensity of predation is also less at the smallest size ranges.
For example, A. cuneiformis predation at sizes of 18-20 mm is 73.6%, approxi-
mately double that upon sizes less than 18 mm. For A. thisphila, predation is
less important at sizes below 14 mm. Perhaps the nourishment obtained by pre-
dation upon the smallest individuals cannot balance the energy expended in
drilling.

Eucrassatella predation rates are comparable to those for Astarte. E. mary-
landica is characterized by 33.3% drilling, E. turgidula by rates of 417~W,
and E. melina by an intensity of 56.9%. E. melina sample sizes are adequate
for the use of chi-square test. Again, no significant difference occurs bet-
ween the size-frequency distributions of the bored and total populations
(X2=5.74 with 5 degrees of freedom). Nevertheless, for E. melina and E.
turgidula, predation appears to be less important at the smaTlest sizes (below
20 mm) and largest sizes (greater than 80 mm). Predation peaks at 50-70 mm
for E. melina and at 30-50 mm for E. turgidual (figure 3).

Results are somewhat different for the genus Anadara (figure 4). Predation
intensities are much lower than those for Astarte and Eucrassatella. Rates
vary within Anadara from 15.4% (for A. staminea) to 34.4% (for A.
subrostrata). Maximum predation occurs at the smaller sizes; predation peaFs
at 10-20 mm for A. idonea, at 10-30 mm for A. subrostrata, and below 10 mm for
A. staminea.

Predation is also of greatest importance at small sizes of Turritella plebeia.
The average rate of predation for the species is 20.2%, but predation inten-
sity peaks at 42.3% for sizes of 11-13 mm (figure 5). Large size apparently
confers some degree of immunity to predation upon the species. Sizes below 11
mm are also relatively safe from naticid attack.

ASTARTE

FREQUENCY VS. LENGTH (mml

A. perplana

N= 30

A. thisphila

N= 253

12 16 20 24 28 32

A. cuneiformis

N= 257

12

16

20

24 28

36

Size-frequency distributions of Chesapeake Group populations of
Astarte. The distribution of the total population is indicated in
white and that of bored shells in black.

10

Figure 2.

EUCRASSATELLA

FREQUENCY VS. LENGTH Imml

E. marylandica

N = 24

S30

50

70

5-

20-

10-

20-

10

E. melina

N=130

30 50

Figure 3. Size-frequency
Eucrassatella.
cated in white

distributions of Chesapeake Group populations of
The distribution of the total population is indi-
and that of bored shells in black.

10

30 50 70 90

10

70

6;_= I I _i

ANADARA

FREQUENCY VS. LENGTH (mm)

A. idonea ZONE 22
N=56

A. idonea Z 24
N=90

LIL_

30 50

70-

A. subrostrata
N= 128

50

30

10-

30 50

A. staminea
N=234

10

Figure 4. Size-frequency distributions of Chesapeake Group populations of
Anadara. The distribution of the total population is indicated in
white and that of bored shells in black.

20-

20-

10

10

S30

50

S70

30-

10-

10

30 50

TURRITELLA FREQUENCY VS. LENGTH (mm)

80-

60-

40-

20-

14 20 26 32 38

Figure 5. Size-frequency distribution of Chesapeake Group populations of
Turritella plebeia. The distribution of the total population is
indicated in white and that of bored shells in black.

Predation upon these four common Chesapeake Group mollusc genera was substan-
tial, producing 20% to 48% of all generic mortality. Predators showed some
tendency to select individuals from the small to middle size categories of the
taxa. Large size may have provided an escape from drilling, although chi-
square analyses indicate bored samples resemble the total samples in size-
frequency distribution. These results are consistent with patterns of
predation on Miocene and Pliocene populations of the bivalve genus Glycymeris
(Thomas, 1976). Chi-square tests for homogeneity of size distributions of
bored and unbored valves showed no significant differences in 8 of 12 popula-
tions. Where size selectivity occurred, small- to intermediate-sized indivi-
duals were chosen in 3 of 4 cases.

PREDATOR PREFERENCES

The naticid gastropod predators of this study were selective in their choice
of prey. Predator preference is indicated with respect to taxonomy of prey,
site chosen for dirlling, and sex of Astarte specimens. No preference occurs
with respect to valve drilled, however.

Although predation was an important source of mortality for all four genera,
certain taxa were preferred to others. I used chi-square tests of the
equality of two proportions to compare predation rates between taxa. No
significant differences exist between average rates of predation of Astarte
(41.8%) and Eucrassatella (48.4%). Similarly, chi-square analysis indicates
drilling rates for Anadara (21.2%) and Turritella (20.2%) are comparable.
Differences significant at the .001 probability level exist between average
predation rates of Astarte and Anadara, Astarte and Turritella, Eucrassatella
and Anadara, and Eucrassatella and Turritella. Naticids prefer Astarte and
Eucrassatella over Anadara and Turritella by a factor of two. Table 1 sum-
marizes these differences.

In addition to comparing average rates of predation for each genus, I also
examined preferences for individual species at particular stratigraphic levels.
Of the three species collected from zone 10 (Calvert Formation), Eucrassatella
melina is clearly preferred over Astarte cuneiformis and Anadara subrostrata
(differences significant at the .05 level). Within the Choptank zone 17, pre-
dation rates are comparable among Astarte thisphila, Eucrassatella turgidula
and Turritella plebeia. The species Anadara stamina is clearly not pre-
ferred, as its predation rate is less than half those of the other species.
At higher stratigraphic levels, differences are less apparent; chi-square
analyses indicate similar rates of predation among the species present.

Astarte and, especially, Eucrassatella are thus the preferred prey of
Chesapeake Group naticids. Of the four genera considered, Anadara is least
frequently selected, though its average rate of predation is similar to that
of Turritella. These data are consistent with results of neontological stu-
dies. Living naticids are known to attack a wide variety of mollusc species
(Thomas, 1976), and yet definite preferences do occur in their feeding habits.
Reyment (1967, 1971) has demonstrated that Recent naticids of the western
Niger Delta are highly selective predators, choosing such prey taxa as Ostrea
and Cardium. Preference in the present study may be related to the morpholo-
gies of the genera examined. Both Anadara and Turritella are prominently
ornamented with ribs. Dudley and Vermeij (1978) have suggested that ribbing
may provide some immunity to attack, but the nature of this defense is
unclear.

Gastropod predators tend to be selective not only of prey taxa, but also of
the site chosen for drilling (Sohl, 1969; Reyment, 1971; Thomas, 1976).
Position of drillholes is dependent upon the morphology and life habits of the
prey, and also varies with the predator. According to Thomas (1976), muricid
borings on Glycymeris shells are variable in position, occurring either at the
margin or scattered over the center of the valve. Holes drilled by naticids
are confined to smaller areas of the shell in both Recent (Reyment, 1971) and
fossil material. Boreholes in Glycymeris subovata are concentrated dorsally
and slightly anterior to the midlines of the valves. Thomas attributed such a
preference to the life position of Glycymeris, which is oriented with its
posterior margin at the sediment surface. The anterior region would thus be
more accessible to a burrowing predator. Crassatellites undulata Say is
drilled most frequently at the umbones (Sohl, 1969). Small shells (less than
an inch in size) tend to be drilled anteriorly; larger shells are drilled
posteriorly. Change in shape with growth may modify the manner in which the
prey is grasped.

The naticid predator which attacked specimens of Astarte was similarly selec-
tive with respect to drill site. I recorded the drillhole position on a nine-
quadrat grid for 26 specimens of A. cuneiformis (see Table 2). Only 2 of 26
drillholes are ventral in positionT 3 are dorsal and 21 are medial. Of the 21
medial boreholes, 3 are located anteriorly and 4 posteriorly. Thus 14
drillholes occupy the central quadrat. If all quadrats were preferred
equally, an average of 2.9 drillholes should occur in each. A chi-square test
(significant at the .01 level) indicates an obvious preference for drilling
the central portion of the shell. Such selectivity may reflect Astarte soft-
part morphology as well as the manner in which the predator grasps its prey.

Table 2. Number of boreholes per quadrat on Astarte cuneiformis.

DORSOVENTRAL LENGTHWISE SUBDIVISIONS
SUBDIVISIONS Anterior Middle Posterior

Dorsal 0 1 2
Middle 3 14 4
Ventral 1 1 0
*kk*****

Although the predator preferred to drill certain areas of the valve, no pre-
ference occurs with respect to valve drilled. The three bivalve genera show
nearly equal proportions of left and right valves drilled (see Table 3). The
proportion of left valves drilled is 0.528 for Anadara, 0.465 for Astarte, and
0.492 for Eucrassatella. For all three taxa, 95% confidence intervals include
the theoretical proportion 0.5 which would occur if drilling of the two valves
were equally likely. Such a lack of valve preference is characteristic of
infaunal organisms which lie with the plane of commissure oriented approxima-
tely vertically. Either valve is accessible to the burrowing predator.

Table 3. Valves drilled by predators on Chesapeake Group bivalves.

TAXON NUMBER OF NUMBER OF LEFT VALVES 95% CONFIDENCE
DRILLED DRILLED DIVIDED BY INTERVALS FOR
LEFT VALVES RIGHT VALVES TOTAL DRILLED PROPORTION

Anadara 28 25 0.528 0.394 0.662
Astarte 54 62 0.465 0.374 0.556
Eucrassatella 31 32 0.492 0.369 0.615
********

A further preference occurs, with respect to sex of Astarte. Extant Astarte
are protandrous; a sex change from male to female occurs at sizes of about
15-20 mm (Saleuddin, 1964). Sexually dimorphic shell differences appear to
develop in association with the sex change. Ostroumoff (1900), on the basis
of a study of A. sulcata genital products, suggested that individuals with
crenulations in h-e ventral margin are females and smooth specimens are males.
The proportion of crenulated specimens at different stages of development with
Paleocene Astarte populations accords with the percent of females in extant
populations (Kauffman and Buddenhagen, 1969). Discriminant analyses indicate
that the character of crenulations divides Miocene populations into two
distinct groups (presumably sexes). Uncrenulated forms are relatively higher,
but crenulated shells are generally wider (Kelley, 1980). Thus, crenu-
lated specimens are morphologically distinct from smooth individuals; sexual
dimorphism is implied.

Crenulated and noncrenulated valves differ markedly in rate of predation.
Crenulated specimens substantially outnumber smooth shells in the populations
I examined, and yet the number of drilled specimens in the two classes is
nearly equal. Of 184 smooth valves, 51 are drilled (predation rate = 55.4%).
Drilling rate for crenulated specimens is 35.9% (55/306 valves are drilled).
Predation upon smooth "males" is thus significantly greater (X2=5.87, p <.02)
than that upon crenulated "females". The predator preference may not be
directly related to sex, however. Instead, it may be linked to the distinctly
different shape or the generally smaller size of males. Predation peaks for
A. thisphila at 14-16 mm length, and for A. cuneiformis at 18-20 and 14-16 mm,
size classes containing high proportions oT smooth specimens.

RELATION OF PREY SIZE TO PREDATOR SIZE

For the prey species studied, predation is distributed fairly evenly among the
size classes; size-frequency distributions of bored shells are not signifi-
cantly different from those of the total populations. Nevertheless, predator
individuals may prefer certain sizes of prey.

Laboratory studies indicate that maximum drillhole diameter is correlated
positively with predator size (Reyment, 1971). Drillhole size is therefore an
indirect measure of predator size, and can be used to test for a relationship
between predator size and prey size. A logical hypothesis suggests that large
predators choose large prey. Drillhole size should thus be correlated positi-
vely with prey size. While some Recent and fossil samples exhibit this rela-
tionship, nonsignificant correlations also occur, possibly due to the activity
of multiple predator species (Reyment, 1967, 1971).

A sample of twenty-six Astarte cuneiformis specimens from Willow Beach
displays a positive correlation (r=.6616) between shell length and outer
drillhole diameter. The correlation is highly significant (p < .001).
Predator individuals thus selected a specific size range of Astarte, the
larger predators choosing the larger Astarte individuals. Despite this corre-
lation, comparable size-frequency distributions occur for the total population
and its bored component. Such a situation could arise if the size-frequency
distribution of the predator were similar to that of the prey.

PREDATION RATES THROUGH TIME

Predator preference occurs not only among genera at a stratigraphic level, but
also within genera through time (Table 1). Only Eucrassatella shows no sig-
nificant variation in predation rate through time; although rates decrease up
the section, the differences are not significant at the 0.05 level.

Anadara exhibits its greatest predation rates within the Calvert Formation.
A. subrostrata is thinner-shelled than later species, making it easier for the
predator to drill. Predation rates between the two younger, thicker-shelled,
species are similar.

Rates of drilling on Astarte are comparable for the Calvert and Choptank popu-
lations (36.6% and 50.6%). A. perplana, the St. Marys species, exhibits sig-
nificantly lower predation rates (13.3%). A similar situation occurs within
Turritella plebeia. The Choptank predation rate is more than twice that of
St. Marys populations, and the difference is significant at the .001 level.

In general, then, drilling rates for Astarte, Turritella and Anadara are less
in the St. Marys Formation than at stratigraphically lower horizons. In the
case of Anadara, shell morphology (in particular, shell thickness) probably
contributes to this situation. Such is not the case for Astarte, as the St.
Marys form is morphologically similar to the more heavily predated A.
cuneiformis, or for Turritella, as a single species is present throughout ti~e
section. The nature of the prey species thus is not the primary cause of low
St. Marys predation rates.

Low predation rates within the St. Marys Formation must therefore be linked to
conditions external to the genera studied. Conditions of the physical
environment, as indicated by sediment characteristics, were not unusual during
St. Marys time. The dominant St. Marys lithology is a blue, sandy clay to
clayey sand. A similar lithology characterizes zone 14 of the Calvert
Formation, which exhibits high predation rates. Drilling thus does not appear
to be linked with lithology. Thomas (1976) also was unable to discover any
correlation between drilling rates and lithology.

Although predation appears unrelated to physical aspects of the Chesapeake
Group, drilling may vary with the biological environment of the taxa studied.
The St. Marys Formation contains an extremely high diversity of gastropods.
Total species diversities are similar for the three Chesapeake Group for-
mations (Table 4). Yet gastropods represent 64.8% of the common St. Marys
molluscs, but only 46.6% and 45.4% of Calvert and Choptank molluscs, respec-
tively. The proportion of gastropods in the St. Marys Formation is thus
significantly greater than in the Calvert and Choptank units (p <.01).

Table 4. Molluscan diversity in the Chesapeake Group. Data are from Vokes'
tabulation of common Chesapeake Group taxa (1957).

CALVERT FM. CHOPTANK FM. ST. MARYS FM.

NUMBER OF
GASTROPOD 54 44 70
SPECIES

NUMBER OF
BIVALVE 62 53 38
SPECIES

GASTROPODS
PLUS 116 97 108
BIVALVES

Gastropods are the preferred prey of certain naticid species (Reyment, 1971).
The abundance and diversity of gastropods in the St. Marys Formation may have
provided alternative food sources for the naticid predators. The presence of
additional abundant, desirable prey could therefore have lessened the depen-
dence of naticids upon the prey species examined in this study.

CONCLUSIONS

Naticid gastropods were major predators of the dominant Chesapeake Group
infaunal molluscs. Lunatia heros, and possibly Polynices duplicatus, contri-
buted significantly to the mortality of Astarte, Eucrassatella, Anadara and
Turritella. Average predation rates vary from about 20% for Turritella and
Anadara to 48% for Eucrassatella. All but the smallest and largest size
classes were affected substantially by predation.

Eucrassatella and Astarte were the preferred naticid prey. Anadara and
Turritella were less frequently selected; the presence of ribbing may have
provided some immunity to drilling. The predator was also preferential with
respect to drill site, selectively boring the central portion of Astarte
valves. No preference is apparent with respect to valve, as right and left
valves are bored equally. Two morphs of Astarte are unequally drilled,
however. Predation was concentrated upon the uncrenulated, presumably male,
segment of the population.

A significant correlation exists between prey size and drill hole size (and
thus, indirectly, predator size) for Astarte. Predator individuals were
therefore size selective, even though size-frequency distributions of the
bored and total populations are similar.

Predation rates are greatest in the Calvert and Choptank Formations for the
genera examined. Intense predation upon these dominant taxa may have acted as
a diversity-regulating mechanism during Calvert and Choptank time. Rates of
drilling are significantly lower in the St. Marys Formation, possibly due to
the availability of a more diverse gastropod prey fauna.

STRATIGRAPHY AND PETROLOGY OF THE PUNGO RIVER FORMATION,
CENTRAL COASTAL PLAIN OF NORTH CAROLINA

A. Kelly Scarborough, Stanley R. Riggs, and Scott W. Snyder
Department of Geology
East Carolina University
Greenville, North Carolina

ABSTRACT

Up to 30 m of phosphatic sediments of the Middle Miocene Pungo River Formation
were deposited in the northeast-southwest trending Aurora Embayment of North
Carolina. These sediments thin to approximately 15 m over the Cape Lookout
High, a pre-Miocene feature which forms the southern boundary of the Aurora
Embayment. The western and updip limit of the formation parallels the White
Oak Lineament, a regional N-S structure. At this lineament, the formation is
abruptly truncated and thins to a feather-edge; the formation thickens
rapidly to the east and southeast. Deposition of the Pungo River Formation
extended beyond the present updip erosional limit of the formation some
unknown distance to the west of the White Oak Lineament.

The Pungo River Formation consists of the four major sediment packages (units
A, B, C, and D) described by Riggs et al. (this volume), and three lateral
faces (units BB, CC, and DD). Phosphorite sedimentation was concentrated in
units A, B, and C which are laterally correlative throughout most of the study
area. However, the muddy, phosphorite, quartz sands of unit B and possibly
the phosphorite, quartz sands and carbonate sediments of unit C grade downdip
to the southeast into an 11 m thick diatomaceous faces (unit BB). Units A,
B, and C grade into a slightly phosphatic, calcareous, quartz sand faces to
the south, in the area of the Cape Lookout High, which probably represents a
shoaling environment (unit CC). The dolomitic unit D of the northern and
eastern portions of the Aurora Embayment grades laterally into calcareous unit
DD in the central portion of the embayment.

Allochemical phosphate grains of the intraclastic variety dominate all the
units in the formation. However, unit A contains abundant pelletal phosphate
in the fine to very fine sand-size fraction. The highest phosphate con-
centrations were found mid-slope in the west-central portion of the embayment.
Updip to the west, the volume of phosphatic sediments decreases rapidly as the
phosphorite units have been sequentially truncated by subsequent erosion.
Facies changes to the east and south result in decreased phosphate content
within the formation.

The depositional pattern of the regionally persistent and cyclical lithologies
of the Pungo River Formation suggests that units A-C were deposited during a
major transgression. The overlying unit D was deposited during the subsequent
regressive phase. Within the Aurora Area, each of the sediment units is
separated from the overlying unit by a period of increased carbonate and
decreased phosphate deposition. Truncation of the units by erosion took place
prior to the deposition of the Pliocene Yorktown Formation. The extensive
erosion produced an apparent offlap geometrical configuration within the Pungo
River units that actually represent a transgressive or onlap sediment
sequence.

INTRODUCTION

The phosphorites and phosphatic sediments of the Pungo River Formation in the
central Coastal Plain of North Carolina were originally described by Brown
(1958) and correlated with-.the Middle Miocene Calvert Formation of Maryland
and Virginia on the basis of benthic foraminifera. Subsequent work by Gibson
(1967) involving molluscs and benthic foraminifera, and recent work by Gibson
(in press) and by Katrosh and Snyder et al. (this volume), utilizing benthic
and planktic foraminifera, support Brown's correlation. The Pungo River
Formation is also equivalent to portions of the Hawthorn Group of Florida
(Riggs, 1967b, 1979b; Gibson, 1967).

The Pungo River Formation is found only in the subsurface of eastern North
Carolina. Its westward, updip boundary coincides with and parallels a major
north-south structural hingeline recognized by Miller (1971). This hingeline
is called the White Oak Lineament by Snyder et al. (this volume). Miller
(1971) recognized several erosional outliers west of the White Oak Lineament,
which indicates that primary deposition extended beyond the present updip ero-
sional limit of the formation. The formation dips and thickens to the east.
Miller (1971) has mapped the formation northward to Virginia, and eastward and
southward to the continental shelf (figure 1). The northern and eastern
extent of the formation is unknown. To the south, phosphorite sediments of
the Pungo River have been recovered from holes drilled on Bogue Banks,
Carteret County (Steele, 1980), and from vibracones across the continental
shelf in Onslow Bay (Lewis et al., this volume).

The study area is located in the east-central portion of the North Carolina
Coastal Plain, covering the southern half of the Aurora Embayment (figures 1
and 2). The Pungo River Formation, within the study area, unconformably
overlies either the (1) Eocene Castle Hayne Limestone (Miller, 1971; Brown et
al., 1972); (2) Oligocene Trent Formation (Baum et al., 1978) or the River
Bend Formation (Ward et al., 1978); or (3) the lower Miocene Silverdale
Formation or the Haywood Landing Member of the Belgrade Formation (Baum et al,
1978; Ward et al., 1978) depending upon the regional location and one's choice
of stratigraphic terminology. For the purposes of this paper, the underlying
units will be designated only as pre-Pungo River sediments. The Pungo River
Formation is unconformably overlain by the fossiliferous, gravelly, muddy,
phosphatic, quartz sands of the Pliocene Yorktown Formation throughout most of
the Aurora Embayment. Post-Pungo River sediments in portions of the study
area that lie south of the Neuse River have been referred to as the Duplin
Formation (Copeland, 1964). Gibson (in press) interprets the Duplin strata as
biostratigraphically equivalent to the upper Yorktown Formation.

The major objective of this paper is to describe the lithology of the Pungo
River Formation within that portion of the Aurora Embayment extending from the
area of Aurora, N.C. southward and westward into the embayment margins (figure
2). More specifically, the objectives are to (1) correlate the lithologic
units and subunits of the Pungo River as identified in the Aurora Area (Riggs
et al, this volume) throughout the study area; (2) describe the lateral and
vertical variations in the lithologies of the formation within the study area;
(3) describe and classify the phosphate grains of the formation according to
the scheme presented by Riggs (1979a); and (4) evaluate the environmental
controls that led to the accumulation of the Pungo River Formation in the
Aurora Embayment.

LEGEND

COUNTY
BOUNDARY

LOCATION AND
" NUMBER OF CORE
EXAMINED FOR
S THIS STUDY

APPROXIMATE
UPDIP LIMIT OF
MIDDLE MIOCENE
SEDIMENTS
(AFTER MILLER,
1971)

AURORA AREA

SCALE

SKM
0 80 KM

SCAPE
LOOKOUT

0C0C

ONSLOW
BAY

Figure 1. Map of study area showing location of cores; Middle Miocene sediments- mapped by Miller (1971)
extend eastward from their updip limit (heavy line).

PITT

STRUCTURAL CONTROLS

The structural features which have traditionally been recognized as
controlling Tertiary sedimentation in the Coastal Plain of North Carolina are
the Norfolk and Cape Fear Arches, with an associated intervening basinal area
called the Albemarle Embayment (Gibson, 1967; Miller, 1971; Brown et al, 1972;
and Mauger, 1979). The thickest sequence of Tertiary sediments was deposited
in this basin with the sediments thinning or absent over the positive
features.

Miller (1971) states that "the location of economic or potentially economic
phosphate deposits...shows that structural conditions in the basin of deposi-
tion played a prominent, perhaps dominant, role in deposition and con-
centration of the phosphate". The close relationship of structural elements
to phosphorite sedimentation and accumulation has been discussed by Freas and
Riggs, 1964; Riggs, 1967a, 1967b, 1979b, and 1980; Miller, 1971. The Pungo
River sediments were deposited in the Aurora Embayment, a smaller depositional
basin contained within the Albemarle Embayment. The Aurora Embayment is deli-
neated by four structural elements as shown in figure 2. The western and
eastern margins are delineated by north-south trending hingelines defined by
Miller (1971) and Brown et al. (1972). The western hingeline is called the
White Oak Lineament by Snyder et al. (this volume). The northern margin is
delineated by the east-west trending Chowan Arch of Riggs (1976b) and Miller
(1971), which is located just south of the Albemarle Sound (figure 2). The
southern margin is defined by a positive feature located south of the Neuse
River which has been recognized by many workers (Riggs, 1967b; Gibson, 1967;
Miller, 1971; Brown et al., 1972; Snyder et al., this volume). This east-west
trending topographic high, called the Cape Lookout High by Snyder et al. (this
volume), has effected the deposition of the Pungo River Formation and younger
sediments in the southern portion of the study area (holes PON-1, CNN-1, and
CTN-1 in figures 4 and 5). Miller (1971) believes these four structural
features created a restricted "Pungo River basin" or Aurora Embayment into
which phosphate, that was precipitated on the open, seaward side of the
easternmost hingeline, was transported and deposited.

PROCEDURES

Samples representing the Pungo River Formation were gathered from several
sources. The majority of samples were obtained from eight cores drilled by
the International Minerals and Chemical Corporation in 1966. These cores were
sampled at every major change in lithology (Riggs, 1967b). Additional samples
of the Pungo River Formation from the Aurora Area were obtained from
Texasgulf, Inc. and North Carolina Phosphate Corporation (NCPC) from cores
drilled in 1978 and 1979, respectively (figure 1). Sample locations within
each core are indicated by tick marks on the stratigraphic cross-sections
(figrues 4 and 5).

All samples were described using a binocular microscope to determine the
mineralogy and texture, to assign a lithologic name, and to identify the
megascopic phosphate grain types. Petrographic examinations of thin sections
from selected phosphorite intervals were examined to aid in the identification
of phosphate grain types and to supplement mineralogical descriptions of the
lithologies. A textural analysis of each sample was performed according to
the methods outlined by Folk (1974) in order to further characterize the

APPROXIMATE UPDIP LIMIT
OF THE PUNGO RIVER FM.

(AFTER MILLER, 1971)

AURORA AREA

/ CAPE
HATTERAS

N-8 HINGELINE

0

CAPE LOOKOUT HIGH

CAPE LOOKOUT

N-8 HINGELINE

(WHITE OAK LINEAMENT)

N

0 50km
J i

Figure 2. Structural setting of the Aurora Embayment. The embayment is
enclosed by structures 1-4 which are from the following sources:
(1) Riggs, 1967b; Miller, 1971; (2) Miller, 1971; Brown et al.,
1972; (3) Snyder, et al., this volume; and (4) Miller, 1971; Brown
et al., 1972; Snyder et al., this volume.

interrelationships of mineralogy, grain size, and phosphate grain types. The
data produced from the above procedures was then used to define and correlate
the various lithologies throughout the study area (figures 3, 4, and 5), and
to interpret the environmental framework within which the Pungo River
Formation was deposited.

RESULTS

Lithologies and Correlations

Within the Aurora Embayment, the Pungo River Formation consists of seven major
sediment units (figure 3). Four of the sediment packages (units A, B, C, and
D) have been described from the Aurora Area by Riggs et al. (this volume).
These four units have persistent mineralogical and textural characteristics
(Table 1) and thus are laterally correlative throughout most of the study
area. Units BB and CC (fig. 3) are lithologically distinct from units A-D,
are restricted in occurrence, and are lateral faces of units A, B, and C
(figures 4 and 5). Unit DD (figure 3) is a lateral facies of unit D, and it
occurs in the central portion of the study area (figures 4 and 5). It should
be emphasized that the faces change from unit D to unit DD consists primarily
of a change in matrix composition. Unit D is a bioclastic-rich sediment with
a dolomite matrix; unit DD is a calcareous bioclastic-rich sediment (figure
3). The reason for the matrix change, whether primary or secondary, is not
yet understood. However, since many workers attribute dolomitization to post-
depositional limestone replacement effected by magnesium-bearing waters
(Miller, 1971), and since the dolomitic unit D is a lateral faces of unit DD,
it is probable that this faces change represents dolomitization of a primary
calcite or micrite matrix due to groundwater or interstitial water movement
through the bioclastic sediments.

Figure 4 is a north-south geologic cross-section through the study area.
Units A, B, and C of the Pungo River Formation extend from the Aurora Area
(hole NCPC) southward to hole PON-3 with consistent mineralogical and textural
characteristics. The faces change of unit D to unit DD occurs between holes
NCPC and BTN-9. The thickened section of Pungo River sediments at hole PON-3
represents the depositional axis of the Aurora Embayment, but it does not
coincide with the maximum accumulation of phosphate. The phosphate sand con-
tent of units A, B, and C increases northward from hole PON-3, producing a
maximum in the cumulative phosphate content of the formation at hole NCPC. At
hole PON-1, on the southern flank of the embayment, the phosphate content of
units A, B, and C decreases dramatically due to (1) a corresponding increase
in the terrigenous content of units A, B, and C; and (2) an absence of the
quartz phosphorite sands of unit C. The bioclastic sediments of unit DD are
also absent. Hole CNN-1, on the southern margin of the Aurora Embayment, is
dominated by the slightly phosphatic (<10%), shelly, calcareous, quartz sands
of unit CC. This unit, which is the southern faces of units A, B, and C, is
characterized by increased terrigenous and calcareous sedimentation and was
probably deposited in a shoaling environment associated with the Cape Lookout
High. The increased terrigenous sedimentation represented by unit CC
apparently extended northward and downslope to hole PON-1, diluting the
phosphate content of units A and B and producing the mud interbeds in unit B.
The occurrence of unit DD at hole CTN-1 represents the northern extent of
Pungo River sediments from the Onslow Bay depositional system (Lewis et al.,
this volume). The calcareous and terrigenous sediments of unit CC underlie

CENTRAL FACIES: COMPOSITE SECTION, SOUTHERN FACIES: HOLES PON-1, EASTERN FACIES: HOLE PON-4
AURORA EMBAYMENT CNN-1, CTN-1

UNIT THICKNESS LITHOLOGY UNIT THICKNESS LITHOLOGY JNII THICKNESS LITHOLOGY

White, slightly pnosphatic and quartz sandy,
0-7m calcareous, bioclastic shell hash (barnacles,
bryozoans) with <20% calcite mud

ABSENT

-4 1-4 I -J

cream colored, nonindurated to indurated,
fossiliferous and moldic, phosphatic and
quartz sandy calcareous mud or limestone
0-5m interDeds which decrease downward

interbedded, very dark greenish gray, slightly
snelly, quartz pnospiorite sand which becomes
more massive downward

Very dark greenish gray, massive, burrowed
0-6m to mottled, moderately muddy quartz phos-
plorite sand witn mdnor shell material

Light olive green, indurated to semi-indurated,
highly Durrowed and locally silicified,
slightly fossiliferous and moldic, pnosphatic
0-7m and quartz said doloiite mud

B 0-12m moderate olive green, burrowed to mottled,
dolomite muddy, phosphorite quartz sand

Dark olive green, massive and mottled, muddy
0-7m phospoorite quartz sand whici ir locally
gravelly (pihosphorite granules) near nase

A I0-6m

Ligut olive green, indurated to nonindurated,
0-1m highly burrowed and locally silicified,
slightly fossiliferous and moldic, phosphatic
and quartz sandy dolomite mud

Hioderate olive green, burrowed to mottled,
dolomitic, muddy phosphorite quartz sand which
0-5m is locally gravelly (phosphorite and quartz
gravels) near base

Wilte to light gray to ligt olive
green, calcareous silty muds to very
0-17m shelly, calcareous muddy, sometimes
gravelly, slightly phosphatic (410%)
quartz sands

Yellowisn-green, slightly phosphatic
0-7m and quartz sandy, dolosilty bioclastic
shell hash bryozoanss, barnacles,
annelid tubes) to shelly dolomite muds

Light grayiso-green, slightly calc-
areous, slightly posp:latic and
0-1 1m quartz saidy, ditomaceous mud;
diatoin fragments compose up to 70o
of tle sediment

Dark green, gravelly (phosphorite
0-12m granules), nuddy, phoaphorite quartz
sand

NOT RECOVERED

Figure 3. Lithofacies descriptions within the Aurora Embayment

C 0-9m

N 0 T 7
z z z z z
0 0 0 z i-
z m a. 0. o o

15-

20-

25-

30-

35-

40-

45 -

50-

I I 1 1 SEDIMENTS
E- pUNGO RIVER

BIOCLASTIC HASH, CALCITE MATRIX

BIOCLASTIC HASH, DOLOSILT MATRIX

CALCAREOUS MUDS TO SHELLY, CALCAREOUS, QUARTZ SANDS

QUARTZ PHOSPHORITE SAND, CAPPED BY INTERBEDDED
PHOSPHORITE SANDS AND PHOSPHATIC MOLDIC LIMESTONE

MUDDY PHOSPHORITE QUARTZ SAND,
MOLDIC DOLOSILT CAP

DOLOMITIC, MUDDY PHOSPHORITE QUARTZ SAND,
MOLDIC DOLOSILT CAP

PRE-MIOCENE TOPOGRAPHIC
HIGH (POSITION DETERMINED
FROM DATA OF MILLER, 1971)

HORIZONTAL SCALE

0 10 KM

V.E.= 537.5

Figure 4. North-south geologic cross-section through the study area.

DD

cc||I
D NTf

CC

B

A w

_r<

C LL
0O
O.

-0

NW
i
z z
I- I-
CQ CQ

10-

15 -

20 -

25-

30-

35-

40-

45-

50-

55-

60-

65-

70-

75-

80-

"ORK o

DOLOSILT MATRIX

QUARTZ PHOSPHORITE SAND, CAPPED BY INTERBEDDED
S PHOSPHORITE SANDS AND PHOSPHATIC MOLDIC LIMESTONE

SLIGHTLY PHOSPHATIC AND QUARTZ BEARING
BB DIATOMACEOUS MUD

B [I MUDDY PHOSPHORITE QUARTZ SAND

7 DOLOMITIC, MUDDY PHOSPHORITE QUARTZ SAND,
A L MOLDIC DOLOSILT CAP

PUNGO RIVER FM
LITHOFACIES

Northwest-southeast geologic cross-section through the study area.

HORIZONTAL SCALE

0 5 KM
V.E.= 266.5

DD 0-n BIOCLASTIC HASH,
DD r
CALCITE MATRIX

D BIOCLASTIC HASH,

Figure 5.

unit DD and, as in hole CNN-1, reflect sedimentation directly influenced by
the Cape Lookout High.

Figure 5 is a northwest-southeast geologic cross-section through the study
area which shows (1) the sequential truncation of the vertical lithofacies of
the Pungo River Formation by the Late Miocene unconformity (Riggs et al., this
volume); and (2) the occurrence of a diatomaceous faces in hole PON-4 that is
considered equivalent to the upper portion of unit B, and possibly unit C
(figure 3). Truncation of the Pungo River Formation prior to the deposition
of the Pliocene Yorktown Formation was extensive and severe, eliminating the
original western depositional extent of the formation, and producing the
erosional outliers noted by Miller (1971) west of the White Oak Lineament.
Where units A, B, and C are present from hole BTN-12 to hole PON-4 their
mineralogical and textural characteristics are consistent. The calcareous
matrix of unit DD at hole BTN-9 changes downdip to dolomite (unit D) at hole
PON-4. Underlying unit D in hole PON-4 is the slightly sandy (quartz and
phosphate) diatomaceous mud of unit BB. The diatomaceous sediments were pro-
bably deposited as a result of the high productivity associated with upwelling
currents and active phosphorite sedimentation. This depositional regime
existed during the period of maximum transgression when the upwelling currents
would have extended furthest into the Aurora Embayment, and at which point
phosphorite sedimentation would have been greatest (i.e., units B and C).
Gravelly, muddy, phosphorite, quartz sands characteristic of the lower portion
of unit B underlie unit BB.

Regional variations in mean grain size (Mz) of the phosphorite sands of units
A, B, and C are random and minor, varying between extremes of 2.6p and 1.3
(fine to medium sand). No meaningful lateral trends in grain size are
apparent; the variations being either too small to be of regional signifi-
cance, or too erratic. However, vertically within the formation there is a
general fining upward trend in the grain size of the phosphorite sands from
unit A through unit C (Table 1).

CORE HOLES _
Ave. for
UNIT NCPC BTN-11 BTN-9 PON-3 study area

1.9 2.4 2.2 2.3 2.2
m. sand f. sand f. sand f. sand f, sand

2.1 2.2 2.1 1.9 2.0
f. sand f. sand f. sand m. sand f.-m. sand

2.0 2.3 2.0 1.3 2.0
A
f.-m. sand f. sand f.-m. sand m. sand f.-m. sand

Table 1. Grain-size variations of the phosphorite units indicating a general
fining-upward trend. Numbers represent the average phi size of each
unit.

Since the phosphorite sands consist of mixed terrigenous and authigenically-
derived sediments, the grain-size variations are only used to indicate the
relative amount of energy in the depositional environments of the units. The
fining-upward trend from unit A through unit C suggests deposition in
increasingly deeper water further from the shoreline (i.e., deposition during
a relatively rising sea level). This is consistent with the interpretations
of Riggs et al. (this volume).

Phosphate Petrology

The megascopic ( >631) phosphate grains of the Pungo River Formation have been
classified according to the classification of sedimentary phosphorites pro-
posed by Riggs (1979a). He subdivides authigenic phosphate into orthochemical
phosphate mud (microsphorite), analogous to micrite in carbonates (Folk,
1974), and allochemical phosphate grains, analogous to carbonate allochemical
grains. Phosphate allochems consist of intraclasts, pellets, oolites, or
fossil skeletal material (figure 6). Lithochemical and metachemical
phosphorites are products of secondary processes and have not been identified
from the Pungo River Formation.

% ORTHOCHEMS

METACHEMS
RTHOCHEMIC L
PHOSPHORITE
Micros horite
ALL LTH CHEMICAL %OOLITES
YPHOSPHORITE PHOSPHORITE
%ALLOCHEMS %LITHOCHEMS
\ OOLITIC s/o SKELETAL
VARIETIES PHOSPHORITE MATERIAL
SKELETAL
PHOSPHORITE
PELLETAL INTRACLASTIC
-PHOSPHORITE PHOSPHORITE'
% PELLETS % INTRACLASTS

Figure 6 Classification of megascopic sedimentary phosphorites (from Riggs,
1979a).

The majority of predominantly dark brown to black, granule-sized and light to
dark brown, sand-sized phosphate grains of the Pungo River Formation possess
the characteristics of intraclastic phosphate allochems described by Riggs
(1979a). These characteristics include a cryptocrystalline, carbonate fluora-
patite matrix (Rooney and Kerr, 1967), terrigenous and authigenic mineral
inclusions, laminae, mottles, and bored and burrowed sediment surfaces (fig.
7). The intraclasts of the Pungo River Formation, which comprise approxima-
tely 80% of the phosphate grains in the formation are angular to subrounded
and irregular in shape, especially in the coarse sand and gravel fractions.
Rounding and sphericity increase with decreasing grain size, although some
irregularity in shape, such as a flattened side, will persist down to the very
fine sand-size fraction. The presence of inclusions and sedimentary struc-
tures diminishes with decreasing size of the intraclasts due to continued

a, .

Imm

Fi ure 7.

A. Coarse sand to -:ery. coarse sand-sized phosphate intraclasts. Notice
the abundant inclusions of clear quartz and subangular to subrounded
and irregular texture of the intraclast.

B. Thin section of a coarse sand-sized phosphate intraclast. The
disseminated inclusions are dominated by bladed to angular, silt to
fine sand-sized qu:it'tz riin (Q). I ,i very fine sand-sized phos-
phate pellets (P) and a glauconite grain (G) are also present as
inclusions. Note the irregular surface of the intraclast which is
partially due to the brilakinm out of inclusions during transport and
the very fine sand-sized 'hosirtLt- pellets (P) and intraclast (I)
to the coarse sand-sized intraclast.

9S ~"; i~ 4 ~
O'&. ~ ."

Figure 8.

A. Very fine to fine sand-sized phosphate pellets. All grains illustrate
the smooth and polished texture and the ovoid to spherical shape of
pelletal forms.

B. Thin section of very fine to fine sand-sized phosphate pellets show-
ing a regular outline, smooth surface, and a lack of terrigenous
inclusions. The mottled interior of the grains results from the
organic matter (bacterial rods and rod aggregates and microfossil
remains) contained within the carbonate fluorapatite matrix (Riggs,
1979a).

".1::
A ".*";"."' '-;^|-
: :-:: e

- yc~
~

,

fragmentation of the grains during transport (Riggs, 1979a). Inclusions con-
tained in the intraclasts of the Pungo River Formation include the following:
detrital quartz, feldspar, and clay clasts; authigenic phosphate and calcite
allochems, dolomite rhombs, glauconite, and pyrite; and fossil matter such as
teeth, bone fragments, spicules, and microfossil fragments of diatoms,
radiolarians, and foraminifera. Approximately 50% of all inclusions found in
the very fine to coarse sand-sized intraclasts consist of very fine sand to
silt-sized, angular to bladed, quartz grains.. Very fine sand and silt-sized
phosphate allochems comprise approximately 10% of the inclusions, and glauco-
nite grains of this same size comprise approximately 5%. Approximately 30% of
all inclusions consist of microfossil fragments. Feldspar, clay clasts,
calcite, dolomite, and pyrite each comprise approximately 1% of all inclu-
sions. The inclusions occur as disseminated particles throughout the car-
bonate fluorapatite matrix of the intraclasts.

Pelletal allochems comprise the second major group of phosphate grain types
found in the Pungo River Formation (figure 8). The uniformity in size and the
regular geometric shapes of the pellets are the two main characteristics that
distinguish pelletal phosphate grains from highly abraded, fine to very fine
sand-sized intraclasts. The pelletal phosphates are moderate to dark brown in
color, extremely well sorted, and are ovoid, ellipsoidal, and rod-shaped. The
pelletal grains of the Pungo River Formation are dominantly fine to very fine
sand-sized (0.177 mm to 0.063 mm), as are the pelletal grains described by
Riggs (1979a). Pelletal forms consist primarily of a cryptograined phosphate
matrix with varying amounts of inclusions dominated by microfossil fragments
and minor ( <10%) terrigenous inclusions.

All of the units within the study area are dominated by intraclastic phosphate
grains in the fine to medium sand-sized fraction of the sediment. Unit A,
however, is characterized by abundant pelletal allochems, which comprise 30 to
40% and often constitute greater than 50% of the phosphate grains in the very
fine to fine sand-sized range (0.063 to 0.177 mm). In the overlying stra-
tigraphic units B and C, only 10 to 15% of the phosphate grains in this size
range are pellets. The lack of observable concentric layering or micro-
botryoidal textures within the pellets, and the common occurrence of pelletal
grains clustered in what appear to be burrow fillings, strongly suggest a
fecal origin (Riggs, 1979a; Riggs et al., 1980).

Invertebrate and vertebrate skeletal material generally constitutes less than
15% of the phosphate macrograins in any given sample, with a range of 5 to
20%. Indirect evidence of fossils in the form of phosphatic molds and casts
of pelecypods, ostracods, echonoid spines, diatoms, radiolarians, and forami-
nifera are also minor components (< 10%) of most samples. Only an occasional
oolitic grain was recognized in thin-section. "Psuedo-oolites", which possess
a nucleus grain but lack well-defined concentric laminations, comprise from
2-3% of the phosphate grains in many samples.

SUMMARY AND CONCLUSIONS

The Aurora Embayment extends south from the Chowan Arch to the Cape Lookout
High. The westward limit of the embayment is a line approximately coincident
with and parallel to the White Oak Lineament. Up to 30 m of phosphorites and
phosphatic sediments of the Pungo River Formation were deposited in the
eastern poriton of the Aurora Embayment. The formation thins westward to less

than 1 m and pinches out at the White Oak Lineament. The formation thins
southward to approximately 15 m over the Cape Lookout High.

Seven major sediment units (units A, B, C, D, BB, CC, and DD) comprise the
Pungo River Formation within the southern half of the Aurora Embayment.
Throughout most of the study area, the formation consists of units A, B, C,
and DD which are laterally persistent and correlative in terms of mineralogy
and texture. These four units are erosionally truncated eastward from the
White Oak Lineament. Units A, B, and C are the main phosphorite units which
reflect a distinct cyclical pattern of sedimentation as defined by Riggs et
al. (this volume) for the Aurora Area. These vertical trends in sedimentation
within the formation are regional trends and are recognized laterally
throughout most of the study area. These trends include: (1) the decreased
phosphate content and increased carbonate content upward within each of units
A through C; and (2) the increase in phosphate content upward from unit A
through unit C. In addition, there is a slight decrease in grain size upward
from unit A through unit C.

Unit BB is an 11 m thick diatomaceous faces, which contains up to 70% diatom
fragments, that occurs in the eastern portion of the embayment. Facies BB is
considered to be the downbasin equivalent of the mid-slope phosphorite sands
of unit B and possibly unit C. The apparent contemporaneous deposition of the
diatomaceous sediments in the east-central embayment area with the greatest
development of phosphorite sedimentation in the mid-slope area to the west
suggests that both the phosphorite and diatomite sediments were deposited in
response to increasing productivity. Such a depositional regime could reflect
an increasing influence of upwelling currents upsection and westward through
the embayment in response to the Middle Miocene transgression as described by
Riggs et al. (this volume). The transgression culminated with maximum deposi-
tion of phosphate and diatoms in units B, C, and BB when upwelling currents
could have extended furthest into the Aurora Embayment (figure 9).

Unit CC is a shelly, calcareous, medium-grained quartz sand faces that was
deposited in the southern margin of the Aurora Embayment, occurring con-
temporaneously with the deposition of the phosphorite sediments of units A, B,
and C in the central portion of the embayment. Deposition of the dominantly
terrigenous and calcareous sediments on and near the flanks of the Cape
Lookout High effectively diluted and prohibited phosphorite sedimentation
south of hole PON-1 (figures 4, 5, and 9).

The phosphate component of the Pungo River Formation is dominated by
intraclastic allochemical grains. This suggests that there was an orthochemi-
cal microsphorite source which was fragmented, transported, and deposited as
intraclasts within the Aurora Embayment. Unit A, however, contains a signifi-
cant concentration of pelletal phosphate grains which may represent a par-
tially orthochemical environment where the phosphate mud was biogenically
extracted from the water, concentrated, and excreted as fecal pellets. This
interpretation is additional support for the in situ deposition of the
phosphorite of unit A as described by Snyder et al. (thi-svolume).

There is a distinct portion of the study area where optimum phosphate for-
mation and/or deposition took place; this locus occurs about mid-slope on the
west-central embayment margin (figure 9). Updip to the west each faces has b
een sequentially truncated and eliminated by post-Pungo River erosion

BTN-11 *

BTN-9*

PON-3*

Figure 9. Depositional regimes during the period of maximum Middle Miocene sea level
transgression. PHOSPHORITE (unit C), DIATOMITE (unit BB), TERRIGENOUS and
CALCAREOUS SEDIMENT (unit CC)

(figure 5). Downdip to the east, and in the southern embayment margin, there
are major faces changes within each of the units (figures 4 and 5).
Consequently, the primary area of potentially economic phosphate concentration
is found in the west-central mid-slope poriton of the Aurora Embayment where
the maximum thicknesses of units A, B, and C are present.

The regionally persistent nature of the lithologies, the lateral and vertical
facies relationships, and the sedimentation trends exhibited by the Pungo
River Formation within the Aurora Embayment, coupled with the basinal geometry
of each of the lithologies and the subsequent erosional modification of the
sediment units, all support the following interpretation. Units A, B, and C
represent phosphate deposition through a relatively rising sea level which
culminated in the deposition of the rich phosphorite of unit C, the diatomite
of unit BB, and the quartz sands of unit CC. Deposition of the phosphorite
sediments was periodically interrupted by regional increases in carbonate
sedimentation and by non-deposition, as proposed by Riggs et al. (this
volume). Deposition of units D and DD occurred on the following regression
which culminated in a major erosional period and severe truncation of the
Pungo River sediment units across the western margin of the embayment and
across the Cape Lookout High. Thus, the Pungo River units A through C were
originally deposited as an onlap sequence of sediments; the subsequent ero-
sional truncation has produced an apparent offlap geometric configuration
(figure 5).

CYCLIC DEPOSITION OF THE UPPER TERTIARY PHOSPHORITES
OF THE AURORA AREA, NORTH CAROLINA, AND THEIR POSSIBLE
RELATIONSHIP TO GLOBAL SEA LEVEL FLUCTUATIONS

Stanley R. Riggs, Don W. Lewis, A. Kelly Scarborough, and Scott Snyder
Department of Geology
East Carolina University
Greenville, North Carolina

ABSTRACT

The upper Tertiary phosphorites in the Aurora Area occur within the Miocene
Pungo River Formation (units A, B, C, and D) and the Pliocene Yorktown
Formation (lower and upper units). These units are characterized by the
following patterns of sedimentation: (1) Three major erosional unconformities
and four minor unconformable surfaces or hiatuses mark the boundaries between
consecutive units and the under and overlying formations; (2) Indurated car-
bonate sediments, which usually contain either a weathered fossil assemblage
or are completely moldic, cap each unit. The carbonate surfaces locally con-
tain a rock-boring infauna and are often phosphatized; (3) Phosphate sedimen-
tation began in unit A and increased to a maximum through unit C, was
negligable in unit D, was re-initiated in the lower Yorktown, and was non-
existent in the upper Yorktown; (4) Phosphate concentration generally
increases upward within each unit until carbonate sediments become important,
then the phosphate decreases; (5) The dominant carbonate within each unit is
as follows: unit A and B, dolosilt; unit C, calcitic micrite; unit D, dolosilt
with abundant calcite shell material; and both Yorktown units, calcitic
micrite with abundant calcite shells.

This sequence of upper Tertiary sediment units suggests a cyclical pattern
controlled by global eustatic sea level fluctuations. Each depositional unit,
its carbonate cap, and the associated unconformable surfaces correlate with
established third-order sea level cycles. Units A, B, and C appear to repre-
sent the maximum transgressive portion of the second-order Miocene supercycle.
Phosphate sedimentation was coincident with the transgression; the maximum
deposition in the Aurora Area occurred during the highest sea level. Unit D
was deposited only over the eastern portion of the area as a regressive facies
of the supercycle. The Pliocene Yorktown sediments were deposited during the
next supercycle. The lower Yorktown phosphorites coincided with the maximum
transgression while the non-phosphatic upper Yorktown was deposited during the
subsequent regressive phase.

INTRODUCTION

Location

The Aurora Area is roughly a 50 sq. km peninsula which occupies the northern
portion of the U.S. Geological Survey 7.5 minute series Aurora quadrangle map.
The area is bounded on the west, north, and east by Durham Creek, Pamlico
River, and South Creek respectively and extends south to the town of Aurora
(figure 1). The area includes the North Carolina phosphate mining district.
A major open-pit phosphate mine has been in operation since 1964-65; a second
open-pit mine is being prepared to begin production in the near future.

Detailed stratigraphic and sedimentological analyses of many sections in the
active mine and hundreds of core holes drilled by two companies supply the
data base for this paper,

Morphologically, the Aurora Area is situated on the Pleistocene Pamlico
Terrace, east of the Suffolk Scarp on the Outer Coastal Plain Province of
North Carolina. Structurally, the area occurs on the upper mid-slope in the
west-central portion of the Aurora Embayment (figure 1). The Aurora Embayment
is a north-south trending Miocene depositional basin which extends from the
Cape Lookout High on the south (Scarborough et al., Snyder et al., this
volume), northward to the Albemarle Sound area of North Carolina. The western
updip limit of the embayment is the White Oak Lineament (Scarborough et al.,
S.W.P. Snyder et al., this volume), a regional north-south structure which is
coincident with the western boundary limit of the Pungo River Formation of
Miller (1971) and a major hinge zone of Brown et al. (1972).

Previous Work

The lithostratigraphy of the Pungo River Formation in the central portion of
the Aurora Embayment has been adequately and uniformly described by numerous
workers since the deposit was first discovered in 1952 and first described by
Brown in 1958. Brown generated an idealized section for Beaufort County in
which he described 4 phosphatic sand units with 4 intercalatedd calcitic and
dolomitic shell limestones" (Table 1). The intercalated carbonates were
described as highly competent, ranging in thickness from 6 inches to several
feet, and generally becoming thicker towards the base of the phosphorite sec-
tion. Coarse phosphate pebbles were occasionally found on top of the upper-
most limestone layer, which was nearly always present. He interpreted "the
cyclical depositional pattern observed in the phosphorite section" as an occa-
sional breaching of a barrier by normal marine circulation conditions pro-
ducing the intercalated shell limestones in an otherwise "closed or
limited-access basin". He concluded that "the often-recognized association of
phosphorites and underlying limestones is not merely fortuitous; there is a
primary genetic association in many cases".

Kimrey (1964) proposed the name Pungo River Formation for the subsurface
phosphorite units in Beaufort County, North Carolina. A core hole on the
Pungo River near Belhaven was designated to be the type section. He described
the 52 foot thick unit, occurring 224 feet below the surface, as a sequence of
interbedded phosphatic sands, silts, and clays; diatomaceous clays; and
phosphatic limestones. In 1965 Kimrey subdivided the Pungo River into 5 zones
based upon gross lithologies and P205 content with many minor variations in
lithology; 4 of these zones occur in the Aurora quadrangle (Table 1).

Gibson's (1967) detailed section of the initial open-pit at the Lee Creek Mine
considered the upper 43.5 feet of exposed Pungo River and 66 feet of overlying
fossiliferous sands and clayey sands. Seven lithic units which began some
distance above the base of the formation were described in the Pungo River
(Table 1), and 9 units were recognized in the Yorktown Formation. He did not
differentiate the upper shell sequences (units 8 and 9), which have sub-
sequently been identified as the Pleistocene Croatan Formation (Gibson, in
press; Snyder et al., in press).

Miller (1971) studied the lithology and distribution of the Pungo River
Formation throughout the Aurora Embayment. In his description he said that

APPROXIMATE UPDIP LIMIT

OF THE PUNGO RIVER FM.

(AFTER MILLER, 1971)

AURORA AREA

HATTERAS

CAPE LOOKOUT HIGH

ATLANTIC OCEAN

CAPE LOOKOUT

ONSLOW BAY

tN

0 60km
I.,rpI

Figure 1. Location map of the Aurora Area, North Carolina.

,;O

BROWN, 1958 KIMREY, 1965 ROONEY & KERR, 1967 GIBSON, 1967 RIGGS et al., THIS PAPER

BEAUFORT COUNTY AURORA QUADRANGLE LEE CREEK MINE LEE CREEK MINE AURORA AREA

UNITS 7-9 VERY FOSSILIFEROUS
S INTERBEDDED SHELL BEDS, MARL, i SHELL MARLS. UNCONSOLIDATED TO MARL CLAYEY SANCROATAN FM.
IL I I - I-
SAND. & CLAY Z INDURATED SANDY SHELL BEDS, YORKTOWN FM. CALCAREOUS SANDSTONE UNITS 3-6 FOSSILIFEROUS CLAYEY
i UPPER YORKTOWN FM.
O O MASSIVE CLAYS & INTERBEDDED SAND OL SAND
I- I- MARL I.
a E LOWER ML UNIT 2 PHOSPHATIC CLAYEY
O O YORKTOWN FM. O SAND
. REWORKED PHOSPHATE REWORKED PHOSPHATE YORKTOWNFM. LOWER YORKTOWN FM.
PHOSPRTE UNIT 1 CLAYEY SAND &
PHOSPHORITE PHOSPHATE PEBBLES

UNIT 7 YELLOW GREEN SAND &
BRYAZOAN HASH'
UNIT D SHELLY DOLOSILT
DOLOMITIC SHELL LIMESTONE ZONE 1 COOUINA CALVERT FM. COOUINA UNIT 6 YELLOW GREEN SAND &
2 HYDROZOAN HASH
I I.I_ _ I .
UNIT 5 MOLDIC LIMESTONE MOLDIC LIMESTONE
I PHOSPHATIC SAND & DOLOMITIC PHOSPHORITE & COOUINA m UNIT 4 ALTERNATING LIMESTONE Ul INTERBEDDED
> SHELL LIMESTONE & PHOSPHATE > UNIT C
aHOPHOITEAN
PHOSPHATIC SAND ZONE 2 HIGH GRADE PHOSPHORITE PHOSPHORITE UNIT 3 PHOSPHATIC SAND PHOSPHORITE SAND
O O O O 0
a DOLOMITIC SHELL LIMESTONE R G DOLOMITIC LIMESTONE g UNIT 2 DIATOMACEOUS CLAY BURROWED DOLOSILT
ZONE 4 LOWER GRADE CLAYEY PUNGO - z- UNIT B
Z Z U
PHOSPHATIC SAND PHOSPHATIC SAND PHOSPHORITE n UNIT 1 PHOSPHATIC SANDS MUDDY PHOSPHORITE SAND
S-- - ---- - - - - RIVER FM. -- - -- BOTTOM OF PIT a.
DOLOMITIC SHELL LIMESTONE DOLOMITIC LIMESTONE BURROWED DOLOSILT
ZONE 5 PHOSPHATIC CLAY - - - UNIT A
PHOSPHATIC SAND PHOSPHORITE MUDDY PHOSPHORITE SAND

CASTLE HAYNE LIMESTONE CASTLE HAYNE LIMESTONE CASTLE HAYNE LIMESTONE CASTLE HAYNE FM.

Table 1. Lithologic correlation chart of the phosphatic sediment sequence
in the Aurora Area, North Carolina. All boundaries are the
author's interpretations.

"the phosphatic sands are repeated vertically in the section, and are inter-
bedded with diatomaceous clays, calcareous clays, dolomites, and dolomitic
limestones". He described three phosphatic sand units in the Aurora Area; one
at the base of the Pungo River, the second roughly one-third of the way up the
section, and the third unit at or just below the top of the formation. He
found a general. increase in the phosphate content in each phosphorite unit
upward in the section. The lower phosphorites are interbedded with dolomites
while a bryozoan limestone or coquina occurs in the uppermost part of the
Pungo River. Miller believed that the complex interbedding of these rock
types reflects fluctuations in depth within the basin caused by a series of
minor transgressions and regressions.

Most previous workers have adequately described the basic lithologies of the
Pungo River phosphorites and recognized a cyclical pattern in the sedimen-
tation. However, they have not (1) described the inter-relationships of the
various lithologies which define the repetitive sediment packages, (2)
described the associated sedimentary structures, and (3) recognized the signi-
ficance of the cyclical environmental changes within the depositional basin.
These are the objectives of this paper. The interpretation of the cyclical
pattern for the Aurora Area is strictly preliminary at this time; the
interpretation will evolve as detailed core drilling continues throughout the
Aurora Embayment and across the continental shelf in Onslow Bay.

STRATIGRAPHY

Five major stratigraphic formations constitute the upper Tertiary section in
the Aurora Area; these are summarized in Table 2. Each formation represents a
very distinct package of sediments which can be readily subdivided into major
sediment units with well defined characteristics and regional continuity.

Pre-Pungo River Formations

The formation underlying the Pungo River Formation in the Aurora Area is the
Eocene Castle Hayne Limestone. This unit is a very fossiliferous, gray,
quartz sandy, moldic limestone. It lies in sharp unconformable contact with
the Pungo River. The surface of the Castle Hayne is commonly phosphatized, is
extensively bored by hard-rock infauna, and contains remnants of a population
of sessile benthos. Outside of the Aurora Area, in portions of the Aurora
Embayment and Onslow Bay, Oligocene and Lower Miocene sediments directly
underlie the Pungo River (Scarborough et al., Katrosh, et al., Lewis et al.,
and Snyder et al., this volume).

Pungo River Formation

The Middle Miocene Pungo River Formation is ubiquitous throughout the Aurora
Area and ranges from 20 to 25 meters in thickness. The formation is sub-
divided into 4 major sediment units which reflect cyclic patterns of deposi-
ton. These units are recognizable throughout the Aurora Embayment with some
important lateral changes in lithofacies (Scarborough et al., Katrosh et al.,
this volume). The chemical sediment components of the Pungo River (including
the phosphate, dolomite, and calcite minerals) were not deposited uniformly
through time. Rather, their concentration reflects a cyclical pattern which
is the major subject of this paper. Table 3 summarizes a composite section of
the Aurora area.

STRATIGRAPHIC SECTION OF THE AURORA AREA, NORTH CAROLINA

FORMATION

THICKNESS(AVE,)

LITHOLOGY

PLEISTOCENE

PLIOCENE

MIOCENE

EOCENE

POST-CROATAN
SEQUENCE

CROATAN

UPPER YORKTOWN

LOWER YORKTOWN

PUNGO RIVER

CASTLE HAYNE

3-15m

1-25m

2-20m

2-4m

20-25M

Quartz sands; quartz sandy
clays; muds; & peats

Quartz sandy & clayey shell
beds; shelly quartz sands;
& quartz sands

Shelly & clayey quartz silts
& sands

Clayey & shelly phosphorite
quartz sands

Shelly dolomites; clayey &
dolomitic phosphorite & quartz
sands; phosphatic sandy dolo-
mites; & phosphatic & quartz
sandy moldic limestones

Quartz sandy moldic limestones

Table 2. Stratigraphic section of the Aurora Area, N.C. Wavy lines indicate
major unconformities; dashed lines indicate minor unconformities or
hiatuses.

AGE

__
__

hhhhh.Chhh-C~t~chhh~h~LcrC-LrC~I~1 ~L~h~chh)chh~L J-U~

COMPOSITE SECTION OF THE PUNGO RIVER FORMATION IN THE AURORA AREA, NORTH CAROLINA

UNIT THICKNESS(AVE,) LITHOLOGY

LOWER YORKTOWN 2-4m Clayey & shelly phosphorite quartz sand

D O-4M Yellowish-green, slightly phosphatic and
quartz sandy, bioclastic-rich (barnacles,
annelids, & bryazoans) dolosilt

C 5-8m Cream colored, nonindurated to indurated,
I I very fossiliferous & moldic, phosphatic
calcareous mud or limestone interbeds which

LU 3-5m
Interbedded, very dark greenish gray, slightly
i shelly, quartz phosphorite sand which becomes
more massive downward.

'- Very dark greenish gray, massive, highly
2-4m burrowed to mottled, clayey phosphorite
c quartz sand with only minor shell material.

B 8-10m Light olive green, semi-indurated to indurated,
highly burrowed & locally silicified, slightly
fossiliferous & moldic, phosphatic sandy,
C2- dolomite mud.
2-4m
_3 Moderate olive green, highly burrowed to
mottled, dolomite muddy, phosphorite quartz
sand.

5-9m Dark olive green, massive and mottled, clayey,
-3 phosphorite quartz sand which is locally
gravelly (phosphorite granules) near the base.

A 3-5m Light olive green, non-indurated to indurated,
highly burrowed and locally silicified, slightly
fossiliferous & moldic, phosphatic sandy
dolomite mud.

Moderate olive green, burrowed to mottled,
muddy, phosphorite quartz sand.

CASTLE HAYNE Gray, indurated, very fossiliferous & moldic,
quartz sandy limestone.

Table 3. Composite section of the Pungo River Formation in the Aurora Area, N.C.
Wavy lines indicate major unconformities; dashed lines indicate hiatuses.

Yorktown Formation

The Yorktown Formation is a unit with uniform lithologies throughout the
Aurora Area. The bottom of the formation is relatively flat while the upper
surface has considerable topography due to erosional channels and channel
deposits of the Croatan; locally, the channels have completely eroded away the
Yorktown. Snyder et al. (in press) have established the age of the Yorktown
to be Pliocene. The Yorktown is divided into upper and lower lithologic
units, both of which represent regular sedimentation in an open marine con-
tinental shelf environment (Mauger, 1979).

The lower Yorktown is a persistent 2 to 4 m unit that occurs throughout the
Aurora Area. This unit is a dark olive green, poorly sorted, shelly, muddy,
fine to medium phosphorite quartz sand. The phosphate concentration ranges
from 5 to 20%, becoming finer and decreasing in concentration upward in the
unit. The unit commonly contains shelly interbeds of calcareous, articulated
invertebrates. Abundant quartz granules, phosphate granules to pebbles, bone
fragments, and robust shell forms occur in the basal sediments. These are in
sharp contact with the underlying Pungo River, which contains a burrowed and
bored phosphorite pavement (microsphorite) on the surface.

The upper Yorktown contains two sub-units. First, a 12 to 15 m sequence of
light greenish gray, shelly, very muddy, fine to medium calcareous and quartz
sand was deposited. The fossils in some faces of this sub-unit are dominated
by echinoid fragments. This sub-unit grades upward into a 1 to 2 m bed of
greenish gray to cream colored, moldic, calcareous muddy, fine to medium
quartz sand which is often lithified. The fossils are generally highly
weathered and leached and are often dominated by turitellid gastropods and
occasionally by oysters. This latter sub-unit is only preserved locally where
the Yorktown is topographically higher and this faces has not been eroded
away.

Croatan Formation

The Croatan sediments represent a complex sequence of very shelly sands which
were deposited in a Pleistocene coastal system associated with barrier
islands, tidal inlet channels, and extensive shallow, near-shore, marine shelf
environments. The transitional coastal environments truncated and locally
dissected the underlying Yorktown Formation producing a surface with con-
siderable relief. The highly variable thickness of the Croatan sediments is a
direct result of the filling of this irregular topography. The specific
lithofacies, their lateral and vertical relationships, and their geometries
are a direct result of differing energy regimes associated with the sub-
environments of the coastal system. In the Aurora Area, this formation con-
sists of four major sediment units.
1. Blue gray, muddy, shelly sand; shells up to 25% of the sediment;
generally not bedded; 1 to 8 m thick.
2. Light gray, slightly muddy, very sandy shell gravel; shells exceed 25% and
commonly 50% of the sediment; shells large robust forms (corals,
Mercenaria, pectinids, etc.); generally highly bedded or steeply cross-
bedded with channel geometries; 0 to 8 m thick.
3. Light gray, clean to slightly muddy, slightly shelly, fine quartz sand;
shells generally less than 10% of the sediment; shells fragile and deli-
cate forms (Ensis, etc.) in life position, as lag laminae, or as dissemi-

nated fine angular hash; poorly bedded and usually highly burrowed; 0 to 3
m thick.
4. Light gray, slightly muddy, shelly fine to coarse quartz sand; poorly
sorted with general decrease in grain size upward; shell content varies
from a few to 40%; 0 to 15 m thick.

Post-Croatan Sediments

The post-Croatan sediments consist of two distinct lithic sequences. A lower
sequence of interbedded fluvial sands, clays, and organic channel deposits is
overlain by estuarine muddy sands and clays with minor shell beds containing
Ostrea, Rangia, Mulinea, etc. The upper sequence consists of nearshore
marine, burrowed and crossbedded sands with a well developed soil profile and
a zone of weathering superimposed upon it. The post-Croatan sediments are
mainly Pleistocene in age but they are being greatly modified by the Holocene
to modern streams and their associated channel fills and soil profiles.

CYCLIC PATTERNS OF SEDIMENTATION

Depositional Sequences and Unconformities

Vail et al. (1977) defined depositional sequences as stratigraphic units com-
posed of a relatively conformable succession of genetically related strata
which are bounded at the top and the base by unconformities or their correla-
tive conformities. They believe that these depositional sequences (1) repre-
sent predictable successions of rocks deposited during regional or third-order
cycles of relative sea level change, (2) are the basic stratigraphic units,
and (3) are separated by minor interregional unconformities or hiatuses
(i.e., stratigraphic surfaces which represent nonmeasurable periods of geolo-
gic time). A package of depositional sequences resulting from a series of
third-order cycles will usually form a higher order sediment unit or deposi-
tional supersequence which is the product of a second-order cycle (Vail et al,
1977). They describe supersequences as distinct groups of superposed third-
order depositional sequences which are separated by major interregional uncon-
formities (i.e., stratigraphic surfaces which represent measurable periods of
geologic time).

Detailed lithofacies studies of the phosphorites and associated sediments in
the Aurora Area have led to the interpretation of the Pungo River and Yorktown
sediments as depositional supersequences which are bounded by three major
unconformable surfaces (Tables 2 and 3). In the Aurora Area, these surfaces
occur between (1) the Eocene Castle Hayne Limestone and the Lower Miocene
Punto River unit A, (2) the Middle Miocene Pungo River units C or D and the
Pliocene lower Yorktown, and (3) the Pliocene upper Yorktown and the
Pleistocene Croatan Formation. Each of these three surfaces is associated
with major periods of erosion during measurable periods of geologic time and
they separate major sediment packages. The regional erosion associated with
each of these surfaces has produced topography on top of the underlying sedi-
ment units and thereby has, in part, determined the final occurrence and
distribution of the underlying lithofacies.

The basic characteristics of the three major unconformities in the Aurora Area
are summarized below.

1. Eocene Castle Hayne Limestone Lower Miocene Pungo River Formation.
a. This is a fairly regular surface which dips generally east, has about
25 m of slope across the Aurora Area, and contains only minor
topographic undulations.
b. The surface on the indurated moldic limestone hardgrounds was
generally modified by phosphate deposition with minor sulfide and/or
manganese staining. This occurs either as a black surface stain which
decreases downward 25 cm or so, or it accumulated on the surface as a
microsphorite pavement.
c. The hardground surface supported an extensive hard-rock boring infauna
and a population of sessile benthic epifauna.
d. Some minor pebbles occur in the basal section of the Pungo River unit
A.

2. Middle Miocene Pungo River Formation Pliocene lower Yorktown Formation.
a. This is a fairly regular surface which dips generally east with about
15 m of slope across the Aurora Area.
b. The surface contains only minor topographic undulations, however,
local scour holes up to 1 m deep are common.
c. A microsphorite pavement or hardground developed on the sediment by-
pass surface. The microsphorite pavement consists of alternately and
repeatedly deposited phosphorite mud which was subsequently burrowed,
indurated, bored, and torn up to produce extensive phosphorite
intraclast pebbles and cobbles.
d. A gravel lag consisting of phosphate pebble intraclasts, quartz
pebbles, abraded and black-stained vertebrate bones and teeth, and
coarse shell material occurs on the surface.
e. The surface on top of the underlying indurated carbonate units con-
tains an infauna of hard-rock borers.
f. The Pungo River sediments, which were deposited as a marine onlap
sequence to the west, have been severely truncated by an extensive
erosional episode. The resulting major unconformity produced the
apparent stratigraphic offlap pattern described by Scarborough et al.
(this volume).
g. This surface occurs on top of the Pungo River unit C in the western
part of the area and on unit D in the eastern portion. The present
distribution and the variable thickness of unit D is a direct product
of this erosional event.
3. Pliocene upper Yorktown Formation Pleistocene Croatan Formation.
a. Topographic relief exceeds 15 m and is superimposed upon a regional
easterly dip; the surface is an erosional truncation of the upper
Yorktown lithofacies.
b. The erosional lows are often associated with modern drainage systems.
c. Lithofacies of the upper Yorktown moldic, turritellid limestones and
differentially calcite-cemented, fossiliferous sandstones are pre-
served on the Yorktown topographic highs.
d. Local pebble and boulder zones of calcareous cemented sandstones,
moldic limestones, and occasional solution cavities occur in the sedi-
ments immediately above the surface in the topographic lows.

Thus, each of the three major unconformities separates distinctive packages
of marine sediments. The two central packages, the Pungo River and Yorktown
Formations, both contain phosphorites and are composed of a series of smaller
sediment units or depositional sequences. In the Aurora Area the Pungo River

consists of 4 sequences separated by 3 minor unconformities, and the Yorktown
consists of 2 sequences separated by a minor unconformity. These minor uncon-
formities or hiatuses separate units which are conformable, rarely show evi-
dence of significant amounts of erosion, and mark major and abrupt changes in
lithology. The underlying unit is commonly a highly burrowed and fossili-
ferous carbonate which shows the following paragenesis:

1. The cessation of active deposition.
2. Exposure of the unit as a submarine surface during a period of sediment
bypass and nondeposition.
3. Semi-induration to induration of the carbonate.
4. Weathering of the shells which were often completely leached out leaving a
dominantly undeformed, moldic carbonate unit.
5. The local establishment of a hard-rock boring infauna on the surface.

Pungo River Depositional Sequences

Each of the four Pungo River depositional sequences is characterized by simi-
lar vertical sediment patterns. Figure 2 shows the detailed variation in the
three main components (phosphate, carbonate, and terrigenous sediments) ver-
tically through a composite section of two core holes. Figure 3 is a com-
posite of seven core holes which eliminates the subtle variations in
lithofacies within each of the depositional units. An inverse association
exists between the terrigenous and phosphate components, both of which are
inversely related to the carbonate component. Units A, B, and C are terri-
genous phosphorites with a minor fossil component which probably reflects
restricted and somewhat toxic marine conditions (Riggs, 1979b and 1980). Each
of these units grades upward, either gradually or with increasing interbeds,
into an upper carbonate which contains minor terrigenous and phosphate sedi-
ment. The carbonates often contain a large population of a more diverse
fauna, probably representing a more normal open,marine environment of deposi-
tion with decreased amounts of terrigenous imput.

Excluding the carbonate caps within each unit, there is a general decrease in
terrigenous sediment and an increase in phosphorite upward from the base of
unit A to the top of unit C (figure 3). Average phosphate concentrations are 5%
to 20% in unit A, 10% to 40% in unit B, and culminate in unit C with con-
centrations in discreet beds between the carbonate interbeds reaching 60% to
75% of the total sediment. The phosphate concentration decreases to a minimum
within the carbonate sediment of the cap rock within each unit (figures 2 and
3). However, the carbonate portions of each unit, particularly units A and B,
often contain significant phosphate concentrations as clean, well sorted
phosphorite, quartz sands filling burrows. The sands were deposited on the
unconformable surface and subsequently backfilled the extensive, underlying
burrow system developed during the deposition of the carbonate units. The
basic sediment patterns within each of the units A, B, and C and the overall
vertical patterns between units A, B, and C are generally persistent
throughout the Aurora Area. However, minor lateral lithologic variations in
the clay, phosphate, and carbonate content do occur within each unit and the
number and character of the carbonate interbeds in the upper portions of each
unit are variable.

Unit D represents a major change in the depositional regime as phosphate sedi-
mentation declined to a minimum. Generally, there is a thin and minor basal

B o -'. --

0 -.-
S32

o o -

FOSSILIFEROU? MUDDY
QUARTZ SAND

MUDDY PHOSPHORITE
QUARTZ SAND

BIORUDITE HASH IN
PHOSPHATIC DOLOSILT
MOLDIC BIOMICRITE
INTERBEDDED PHOSPHATiC
MOLDIC BIOMICRITE &
QUARTZ PHOSPHORITE SAND

CLAYEY QUARTZ
PHOSPHORITE SAND

PHOSPHATIC DOLOSILT

CLAYE' QUARTZ
PHOSPHORITE SAND

PHOSPHATiC DOLOSILT

PHOSPHATE % TEPR NOUS % CARBON

10 20 30 40 50 60 70 80 90
I I I I I I I I

DOLOSILTY S CLAYE'
PHOSPHORITE QUARTZ SAND .'

SANDY MOLDIC BI MICR'DrTE .

DATA FROM TEXTURAL & MINERALOGICAL ANALYSIS

DATA FROM MEGASCOPIC LOG DESCRIPTION

Figure 2: Relationship of phosphate, carbonate, and terrigenous sediment to
the stratigraphic units and the gamma-ray log of the phosphorite

section (North Carolina Phosphate Corp. core holes H10 and GH8.5).

GAMMA-RAY LOG

~rrr

. -I

._ -

PHOSPHATE
10-0

% TERRENOUS

) 10 20 30 40 50 60 70 80 9011

!* 4 I. . . .

ee== ... .. .. .. .. .. .. .. ... .".. ,. "..

Composite averages of total phosphate, carbonate, and terrigenous
sediments from textural and mineralogical analyses of core holes in
the Aurora Area, North Carolina (core hole data from North Carolina
Phosphate Corp.).

Figure 3.

-- .i; .-..:.: ..;
..-...
-
.-
--- --

sediment that contains up to 5% to 10% phosphate. This grades rapidly upward
into the dominant, very fossiliferous, calcite biorudite in a dolosilt matrix
with no phosphate. Unit D occurs in the down-basin or eastern portion of the
Aurora Area. The distribution pattern along the feather-edge of the unit is
somewhat irregular due to the subsequent erosional truncation of the entire
updip portion of the Pungo River.

The cyclic pattern of deposition and the inter-relationships of each of the
sediment components produces a very distinctive and recognizable signature on
the gamma-ray logs (figure 2). Occasionally, the carbonate portions of units
A, B, and C are not readily picked up on the logs due to the rich phosphorite
sand which has backfilled the burrows in the carbonate sediments. The general
gamma-ray signature for the Pungo River has been known and used for years in
exploration, but it has not been correlated to the depositional sequences.

Yorktown Depositional Sequences

In the Aurora Area, the Yorktown Formation consists of 2 depositional sequen-
ces which are referred to as the lower and upper units (Tables 2 and 3). Both
sequences are distinct lithologic units; the lower unit is a clayey,
phosphorite, quartz sand faces, whereas the upper unit is a very fossili-
ferous, clayey and calcareous, very fine quartz sand without any phosphate.
Snyder et al. (in press) identified the microfaunal suite of both units as
being of definite Pliocene age. This formation is very uniform and occurs
throughout the Aurora Area, except locally where Pleistocene drainages have
eroded into and occasionally through the units, and is almost ubiquitous
throughout the Aurora Embayment (Scarborough et al, Katrosh et al., this
volume).

The lower Yorktown contains significant concentrations of phosphate which
average between 10% and 25% of the sediment. The phosphate is generally most
abundant and coarsest at the base where there is a significant concentration
of gravel. The phosphate rapidly becomes much finer grained upward with a
complete loss of pebbles, and gradually decreases in abundance. The coarse
gravels have been interpreted in the past to be "obviously reworked" from the
underlying Pungo River Formation (Brown, 1958; Gibson, 1967; Miller, 1971).
However, similar types of phosphate have not been found to occur within the
Pungo River except locally in the basal portion immediately above the Castle
Hayne-Pungo River major unconformity. We believe that the phosphate gravel is
related to the development of a microsphorite pavement on top of the
"hardgrounds" formed by the uppermost carbonates in the Pungo River; this
makes it an unconformity-type phosphate as described by Riggs (1980). The
phosphate pavements and the resulting coarse intraclasts were formed on the
Pungo River-Yorktown major unconformable surface during the initial stages of
the Pliocene transgression and thus, represent primary phosphorite sedimen-
tation at that time. In addition, preliminary petrographic work within the
lower Yorktown suggests that all of the phosphate may represent primary depo-
sition during a Pliocene phosphogenic episode.

The top of the lower Yorktown is marked by a major change in lithology and
color. Locally, the surface is marked by a poorly developed and variable zone
of calcite-cemented, partially indurated, and moldic sediments. Mauger (1979)
described a similar zone in the Lee Creek Mine. The major change in com-
position at this poorly developed weathering profile between the lower and
upper units suggests a major change in the depositional regime at this time.

STRATIGRAPHIC RELATIONSHIP TO GLOBAL SEA LEVEL CURVES

Vail et al. (1977) define 3 orders of cycles of relative sea level change.
First-order or "global cycles" are records of world-wide sea level responses
to major, large-scale geotectonic processes which have durations of 100 to 300
million years. The second-and third-order cycles have durations of 10 to 80
million years and 1 to 10 million years, respectively. Vail et al. believe
that some of the second-order cycles may be of sufficient duration and magni-
tude that they may also be a response to geotectonic mechanisms. However,
they believe that glaciation and deglaciation may account for some of the
second-order and many of the third-order cycles, especially in the late
Neogene. They also recognize that there may be other yet unidentified causes
working "in combination with geotectonics and/or glaciation to accentuate or
diminish the changes". For example, Pitman (1978) proposed the idea that
abrupt changes in rates of seafloor spreading could cause sea level fluc-
tuations while the resulting sediment patterns along continental shelves are
dependent upon the rate of sea level change and sediment flux (Pitman, 1979).

Based on their global sea level curves, Vail et al. (1977) have delineated a
series of global highstands and lowstands of the sea. Global highstands are
when sea level is above the shelf edge in most regions of the world and, con-
sequently, are characterized by widespread marine sediments on the shelves and
adjacent coastal plains. Global lowstands are when sea level is below the
shelf edge and are characterized by erosion and nondeposition producing major
interregional unconformities. Vail et al. (1977) identify major highstands at
about 13 million years ago in the Middle Miocene and about 4.5 million years
ago in the Early to Middle Pliocene. Major lowstands occurred at about 6.75
million years ago in the Late Miocene and at 2.9 million years ago in the late
Pliocene to early Pleistocene. Each of these second-order global high and low
stands of the sea is shown on figure 4 and appears to have controlled the
cyclical sedimentation patterns represented in the upper Tertiary section in
the Aurora Area. Thus, the sediments which are Middle Miocene, Pliocene and
Pleistocene in age (Table 2), and which are separated by major unconformities,
appear to fit entirely within 2 second-order cycles, the Te and Tf super-
cycles of Vail and Mitchum (1979).

The major foraminiferal age-dating for the Pungo River Formation has been done
by Gibson (in press). He has found planktic foraminifera indicating zone N.8
(Blow, 1969) widely distributed in the area extending from south of the Neuse
River in North Carolina, northward to the Norfolk area, Virginia. Strata con-
taining planktic assemblages equivalent to zone N.11 (Blow, 1969) are found in
the central and northern parts of the Albemarle Embayment in North Carolina.
These planktic zones correspond to the late Early to early Middle Miocene and
Gibson (in press) assigns an absolute age of 15 to 13 million years ago for
the time of deposition for the Pungo River sediments. As can be seen in
figure 4, this age assignment places the Pungo River into the TM2.2 third-
order cycle of Vail and Mitchum (1979). This assignment coincides with the
maximum stand of sea level of the second-order Te supercycle. However, the
late Early to early Middle Miocene, as plotted on the sea level curves of Vail
and Mitchum (1979), corresponds to an absolute age of 17 to 14.5 million years
for the Pungo River. This puts the Pungo River sediments into both the TM2.1
and TM2.2 third-order cycles of Vail and Mitchum. Brown (1958) concluded that
the foraminifera in the top few feet of the phosphorite section could be dated
Middle Miocene; due to poor foraminiferal preservation and control, the

remainder of the phosphorite section could be considerably older the base
could possibly even be as old as Late Oligocene. Gibson (1967) concurred,
stating that some of the poorly preserved molluscan molds suggest greater
affinities to Oligocene than to Miocene species.

The hypothesis for the depositional history for the Pungo River and Yorktown
cyclical sediment units in the Aurora Area is based primarily upon the
interpretation of the lithostratigraphic units, with the major time control
from the foraminiferal data. In the case of the Pungo River, Gibson's (in
press) foraminiferal dates put general upper age brackets on the formation.
On the basis of recent work, Katrosh et al., Snyder et al. (this volume) have
found that within most of the lower lithofacies there is an extreme paucity of
planktic forms, the foraminiferal preservation is generally poor, the abun-
dance is extremely variable, and the distribution patterns, both regionally
and vertically through the formation, are not clear. Thus, there appears to
be ample justification, particularly in the lower units, to shift the absolute
age assignments of the Pungo River down slightly (a few million years) on the
basis of the lithostratigraphic evidence to allow a "better fit" with the glo-
bal patterns of Miocene sedimentation.

The Yorktown Formation, which unconformably overlies the Pungo River
Formation, has been dated by Snyder et al. (in press) using planktic foramini-
fera. Their age assignment ranged from just below the base of zone N.19 to
the middle of zone N.20 of Blow (1969) and from the early part of zone PL1 to
the middle part of zone PL3 of Berggren (1973). This places the Yorktown into
the Early to early Late Pliocene, or between 4.8 to 3.1 million years ago
(Berggren, 1973). Thus, the Yorktown falls into the Tf supercycle (Vail and
Mitchum, 1979) and represents the third-order cycles TP1, TP2, and TP3 (figure
4).

Thus, by defining the basic lithostratigraphic depositional sequences and
establishing the associated faunal age assignments, periods of major and minor
sea level highstands and lowstands have been identified for the Aurora Area.
Comparison of these cyclical patterns of the upper Tertiary for the Aurora
Area with the established global sea level curves of Vail et al. (1977 and
1979) based upon global patterns of seismic stratigraphic data suggest a
remarkable fit (figure 4).

HYPOTHESIS FOR CYCLIC DEPOSITION IN THE AURORA AREA

The hypothesis is based upon the interpretation of detailed lithostratigraphic
data and superimposed biostratigraphic data for age assignments as follows:
(1) the vertical lithologic changes within each of the depositional sequences;
(2) the changes that occur vertically through the depositional supersequences;
(3) the recognition and location of the major and minor unconformities; (4)
the lateral lithic continuity of both the depositional sequences and super-
sequences through the Aurora Area; (5) the regional stratigraphic continuity
and geometry as developed by Scarborough et al. (this volume); and (6) the
foraminiferal data of Katrosh et al. (this volume). Figure 4 presents a ten-
tative interpretation of the relationship of the upper Tertiary section in the
Aurora Area to the global sea level curve of Vail and Mitchum (1979).

1. A lowstand of the sea during much of the Early Miocene produced a major
unconformity on top of the Castle Hayne sediments. During this period of

CYCLES OF RELATIVE CHANGES
z
OF SEA LEVEL BASED ON O o
COASTAL ONLAP o o O Z i
(AFTER VAIL & MITCHUM, 1979) WJ < 0 0 w
-j w-J :i >-
-a RISING FALLING C U 0 0 .
>- >- 2 LITHOLOGY EPOCH 0O
0.6 0.5 0.4 0.3 o LTHLG EPCH .

TP3
TP2
TP1
TM3.3
TM3.2

TM3.1

PRESENT
SEA t<
LEVEL

~r:::j~::::::~~:~:':~:::"
::~::~::j:::::i::~::::::j:~::Qbi:::;. :~.
I:s.r~ ~.:::
"""""' ~'::~

..---
""."""~:~:~:I:~:~:~:~~,.,:::::::~::i~ ~
^\" :6:::::~:':"::
~~ ~
gwi~~:j::j::j
~:j::~:~::.:f8#:~:~:~X:j~:~:j"""::~:
:~:j ::::::::::~::::~::j
il;:~; :jj~:i:~::j"''li'''~'""'~"'
~:;i
;:::~:~:;:;;~~n:t:
:18~~:S~Si8::;:~:;:: L.:.S:~~Ba::xs::~:::;
'"" x~ ~ LL~
;-~;i:~-
i,,,.,,~
5 *. X~i ~~:
8":"'~:i~t~: ..
iiiiiinslW::::~::::::~::a):::.:j:~:::~::~:~'
~x5:~:.: o:c~;: :'~::::~:~:~:;:::~:s:
:::::::i:,:::
j:~
.;.,,. ;.;.~~.
:~~i::':":o::Ij:qre::::i::::::::~" :~:Y"'~:~
:~:o: ".'.:x:c*.::::~:~:i:j::~:~:i;:~
::::~:::::i::::: :~x
;":' :: '~;;x9:::::":;:#i~~lijl~i#:
'...i.~ :.~:
X
:''Z;''':: ::~::::i::::j:-I:::~i~8ii~iiiiii~
~;:12::~~:~:~:~:i~i
:~i~i~i~i~i~~a.~::":"'":":.';;'
'Xls#a~:':::- x.-.3: #a.~:
;'" i~lH
.,,
:~~~~1':."
::ii:li::~::::::s:::::~:::::::::~:~IIB
:~:Zj~~:::'::181:'::#8iP#51::.;.... :::::::::::j:::I
:o:.;~ :9::s:i~;~:r:::r
'C :~:3ir:i:~::;:::8:::: rc
j
''i ''''
::::::x:~:x:I:::~:::::::::
;ijjj~;uuILLi :---CLLi-

TM1.3

TM1.2

MAJOR

YORK- UPPER
TOWN LOWER

U N CONFORMITY

FOSSILIFEROUS MUDDY QUARTZ SAND

MUDDY PHOSPHORITE QUARTZ SAND- I 1

MAJOR UNCONFORMITY

BIORUDITE HASH IN PHOSPHATIC DOLOSILT

INTERBEDDED PHOSPHATIC MOLDIC BIOMICRITE AND
C QUARTZ PHOSPHORITE SAND
0 CLAYEY QUARTZ PHOSPHORITE SAND
SB PHOSPHATIC DOLOSILT
z CLAYEY QUARTZ PHOSPHORITE SAND
3 PHOSPHATIC DOLOSILT
a. A DOLOSILTY & CLAYEY PHOSPHORITE QUARTZ SAND

MAJOR UNCONFORMITY

-5

- 10

- 15

- 20

_____ ITM1.1 Td _

Figure 4. Preliminary relationship of the upper Tertiary stratigraphic units
in the Aurora Area to the global sea level curves.

TM2.2

TM2.1

TM1.4

TM2.3

exposure, nondeposition, and erosion, these sediments were first indurated
and then leached producing an undeformed moldic limestone. This rock sur-
face formed hardgrounds which were phosphatized and populated by benthic
boring infauna and sessile epifauna during the initial phases of the sub-
sequent transgression.

2. The deposition of the Pungo River supersequence began in the late Early
Miocene and continued through the Middle Miocene. This sediment package
represents the highstand of the sea during the second-order or Te super-
cycle of Vail and Mitchum (1979).
a. Units A, B, and C represent depositional sequences of marine onlap
which were deposited in response to a progressively rising sea level.
b. Units A, B, and C are depositional sequences which represent third-
order cycles deposited in response to relative rises in sea level
associated with Cycles TM1.4, TM2.1, and TM2.2, respectively, of Vail
and Mitchum (1979).
c. Each of the cyclical units A, B, and C is dominantly terrigenous
sediments and phosphorites with a limited and restricted fauna which
culminated in carbonate sedimentation with a richer and less
restricted fauna.
d. The top of each unit is marked by a minor unconformity. These confor-
mable surfaces of nondeposition represent the brief periods of rela-
tive falling sea levels associated with the third-order cycles of Vail
et al. (1977).
e. Unit D follows the same depositional pattern as units A, B, and C.
Unit D is dominated by carbonate with only minor terrigenous and
phosphate sediments at the base. The foraminifera still suggest a
greater water depth with more open marine conditions than units A and
B (Katrosh et al., this volume). Thus, the unit still represents a
highstand of the sea; however, the regional distribution and geometry
and changing depositional regime suggest that unit D was deposited on
the first part (third-order cycle TM2.3) of the regressive phase of
supercycle Te (Vail and Mitchum, 1979).

3. Relative sea level continued to drop producing a major lowstand during the
Late Miocene, which coincides with the third-order cycles TM3.1, TM3.2,
and TM3.3 (Vail and Mitchum, 1979). This period of nondeposition and ero-
sion truncated the updip portions of the Pungo River depositional sequen-
ces, producing the apparent offlap pattern described by Scarborough et al.
(this volume) and the major unconformity between the Pungo River and the
Yorktown Formations.

4. The Yorktown depositional supersequence, which includes both the lower and
upper sequences, was deposited in response to the highstand of the sea
associated with the second-order or Tf super cycle (Vail and Mitchum,
1979).
a. The end of the 4.5 million year erosion period is marked by an exten-
sive development of unconformity type phosphorite (Riggs, 1980). The
leading edge of the third-order cycle of relative sea level rise (TP1
of Vail and Mitchum, 1979) deposited repeated laminae of microsphorite
mud (Riggs, 1979a) which was sequentially burrowed, indurated, bored,
and broken into intraclastic phosphorite gravels and deposited in the
basal sediments of the lower Yorktown sequence along with a coarse
sand and granule, quartz component.

b. The lower Yorktown terrigenous and phosphorite sediments were depo-
sited in response to the relative rise in sea level which is coin-
cident with Vail and Mitchum's third-order TP1 cycle, which culminated
in a sea level maximum about 4 million years ago. The lower Yorktown
contains a significant concentration of intraclastic phosphorite sand
which appears to increase in concentration to the east and south of
the Aurora Area.
c. A minor unconformity, characterized by a moderate weathering zone
(Mauger, 1979), separates the two depositional sequences and marks a
major change in lithology and depositional regimes. This hiatus
appears to be coincident with, or immediately following the sea level
maximum of the second-order global Tf supercycle of Vail and Mitchum
(1979).
d. The upper Yorktown is characterized by a normal marine, very fossili-
ferous, clayey and calcareous, fine quartz sand with essentially no
phosphate. This depositional sequence was deposited during the
slightly lower stand of relative sea level coincident with the
regression of the second-order Tf supercycle and the third-order TP2
cycle of Vail and Mitchum.

5. A major lowstand of the sea between 3 and 4 million years ago terminated
the Yorktown depositional supersequence. This lowstand is coincident with
at least the third-order TP3 cycle and possibly extends into the early
Pleistocene Q supercycle of Vail and Mitchum. The surface sediments of
the upper Yorktown were weathered, indurated, and the fossils dissolved
producing moldic limestones and calcareous sandstones. The distribution
of these sediments was then severely modified by erosion of the superim-
posed drainage system.

6. Subsequent Pleistocene transgressions first deposited the complex faces
of nearshore marine and coastal shell beds of the Croatan Formation on the
irregular upper Yorktown surface.

7. Another Pleistocene transgression, following a major lowstand of the sea,
produced the depositional sequence of fluvial, estuarine, and coastal
terrigenous post-Croatan sediments.

8. The Holocene-Recent weathering and erosional surface, the resulting uncon-
formity, and the associated depositional sequence of fluvial and estuarine
terrigenous sediments are presently being superimposed upon the similar
post-Croatan sediment sequence.

It is self-evident that the upper Tertiary sediments were only deposited and
preserved in the Aurora Area during major highstands of the sea. It is
expected that some younger and older deposits associated with the lowstands do
exist in the seaward and deeper portions of the Aurora Embayment and out onto
the continental shelf. For example, we believe that we will find both younger
and older transitional sequences representing the Pungo River Formation in
Onslow Bay, North Carolina (Lewis et al., this volume).

In the past, the phosphorites of the lower Yorktown have been interpreted as
reworked from the erosion of the underlying Pungo River sediments (Table 1).
However, recent stratigraphic and petrographic work at East Carolina
University suggests that the lower Yorktown phosphorites may represent primary

phosphate sedimentation. Could this portion of the Pliocene (TP1) represent a
phosphogenic period? If this is the case, then the formation and deposition
of phosphorite in both the Pungo River units A, B, and C and in the lower
Yorktown appears to be intimately tied to the transgressive portions of super-
cycles Te and Tf of Vail and Mitchum (1979). Also, phosphate sedimentation in
the Aurora Area appears to increase with the transgression, culminating in
maximum deposition at the time of global sea level maximums.

Riggs (1980) proposed a close and integral relationship between the formation
of major sedimentary phosphorites, their anomalous sediment packages, and
periods of major and increasing tectonism. Pitman (1977) proposed that the
highs in the global sea level curves could be products of major tectonism and
high rates of sea floor spreading. Comparison of the cyclical sediment pack-
ages and their associated phosphate components to the global sea level curves
suggest that there is a very strong correlation of phosphorite formation and
the associated sediment patterns to major oceanic events and the relative
position and movement of sea level.

ACKNOWLEDGEMENTS

This work was partially supported by a grant from the National Science
Foundation (NSF DAR 79-08949). We would like to thank the personnel of North
Carolina Phosphate Corporation, Washington, N.C. and Texasgulf, Inc., Aurora,
N.C. who have allowed us to utilize drill hole samples and supplied access to
the operating open-pit mine for sampling and stratigraphic analysis, respect-
ively. We would specifically like to thank Mr. Ralph Chamness and Mr. Ronald
Crowson of Texasgulf, Mr. Dewey Walker of North Carolina Phosphate
Corporation and Dr. Don Neal, Mr. Doug Ellington, Mr. David Reid, and the many
graduate and undergraduate students and faculty at East Carolina University,
who contributed greatly to our present understanding of this complex sediment
system. Thanks are also due to the many participants of the ongoing programs
associated with the International Geological Correlation Program (I.G.C.P.)
156 on Phosphorites.

SYNTHESIS OF PHOSPHATIC SEDIMENT FAUNAL RELATIONSHIPS
WITHIN THE PUNGO RIVER FORMATION: PALEOENVIRONMENTAL IMPLICATIONS

Scott W. Snyder, Stanley R. Riggs, Mark R. Katrosh,
Don W. Lewis, and A. Kelly Scarborough
Department of Geology
East Carolina University
Greenville, North Carolina

ABSTRACT

The lower part of the Pungo River Formation in the Aurora Embayment (Units A
and B) consists of phosphorite sands and interbedded dolomites that grade
southward into calcareous quartz sands (Unit CC) associated with a pre-Miocene
topographic high. Within this embayment phosphate content decreases southward
and becomes negligible in Unit CC. Units A and B contain sparse foraminif-
eral assemblages dominated by benthic species that indicate nearshore,
shallow water conditions. Planktic specimens are rare to absent in these
units. Similar assemblages persist as the units thicken to the east. The
sporadic occurrence of open shelf species in Units A and B suggests that the
depositional embayment was not restricted; but conditions were clearly not
generally suitable for open shelf species. The frequent dominance of
Buliminella elegantissima, which flourishes in sewage outfall areas in modern
seas, suggests that water chemistry or organic nutrient supply, perhaps
related to phosphate genesis, limited foraminiferal faunal diversity.

Upper Pungo River sediments within the Aurora Embayment (Units C, D, and DD)
consist of phosphorite sands and interbedded phosphatic, quartz bearing,
moldic limestones. Units C and DD also grade southward into the calcareous
quartz sands of Unit CC. These upper units contain richer, more diverse
benthic assemblages that are dominated by middle shelf species. Planktic
specimens are common within these units. Unlike the assemblages of units A
and B, those of Unit C suggest no unusual depositional conditions.
Phosphorites of Unit C are richer in phosphatic sediments than are those of
Units A and B. This enrichment may reflect concentration by physical sedimen-
tary processes. Faunal and sedimentary characteristics suggest that the
phosphate of Unit C was transported, perhaps being derived from adjacent areas
of the embayment or directly from underlying units (A and B), in which the
phosphorites appear to have formed in situ.

INTRODUCTION

The literature on phosphates is voluminous, and theories concerning the origin
of sedimentary phosphate are diverse, often appearing to be somewhat contra-
dictory. Among the theories that have been proposed in the past, several
basic schemes that address the problem of phosphate genesis have received con-
siderable support.

Most of the world's phosphorus resources occur in the form of ancient, bedded
sedimentary phosphorites of marine origin (Manheim and Gulbrandsen, 1979).
The importance of plants and organisms as a direct source for phosphorus has
been emphasized by many authors. Seeley (1866), Brongersma-Sanders (1957),
and Bushinskii (1966) cited the decomposition of marine organisms as a criti-

cal factor in phosphate genesis. Brongersma-Sanders (1957) and Rooney and
Kerr (1967) suggested that episodes of mass mortality might be necessary to
generate sufficient volumes of such decomposing matter. Others have contended
that the upwelling of deep, phosphorus-rich water and its associated high pro-
ductivity rates are the most essential elements for phosphate production
(Kazakov, 1937, 1938; McKelvey, 1963). However, Manheim and Gulbrandsen
(1979) pointed out that high nutrient concentrations are not the only factor
involved in phosphate formation. Such concentrations occur near Antarctica
and in the Gulf of Alaska, but there are no associated phosphorites. Also,
there are phosphorites which have formed in areas where no large scale
upwellings are known to have occurred, such as the Blake Plateau and the
Chatham Rise. Mansfield (1940) and Rooney and Kerr (1967) linked vulcanism
with the occurrence of phosphate, suggesting that volcanic activity might
affect water chemistry or cause mass mortality of marine organisms as a result
of massive ash falls. Some workers have suggested that phosphate forms below
the sediment-water interface, and thus its occurrence may not be directly
related to characteristics of the overlying water mass. The phosphorus-rich
interstitial water of some marine sediments is known to produce sedimentary
phosphorite by the replacement of carbonates (Ames, 1959; Manheim et al.,
1975). Baturin (1971), Cook (1967, 1976), and Bremner and Willis (1975)
viewed replacement as an important mechanism in forming low grade phosphate,
adding that mechanical reworking and concentration would then be necessary to
produce high grade deposits. Miller (1971) and Riggs (1979) demonstrated that
structural and geographic setting strongly influence the accumulation of
significant amounts of phosphate. Riggs (1980) has also postulated a connec-
tion between the origin of economically important phosphate deposits and
regional tectonism. Clearly, phosphorites represent a complex sediment system
in which any of the mechanisms discussed above, or any combination of those
mechanisms, may play a significant role.

Nearly as complex and controversial as the processes responsible for phosphate
genesis are questions concerning the environment of deposition in which
phosphorites accumulate. Shaler (in Penrose, 1888) suggested that phosphates
originate in paludal environments in association with peats. Pevear (1966)
linked phosphorites with phosphorous-rich waters supplied by estuarine
marshes, thus implying accumulations in nearshore, marginal marine environ-
ments. Gibson (1967) and Miller (1971) associated economic phosphate deposits
with shallow shelf, open marine environments ranging in depth from 100 to 200
meters. According to Riggs (1979), coastal environments and shallow water
structural platforms serve as optimum areas for phosphorite formation.
Manheim et al. (1975) documented the formation of contemporary phosphorites in
marine environments ranging to depths of 1000 meters. It appears that
phosphate may originate in a variety of environments as a result of several
different mechanisms or combinations of mechanisms.

Weaver and Beck (1977) pointed out that one very essential concern should be
to distinguish between in situ and transported phsophatic sediments. This
distinction is difficult to recognize, and little has been done to determine
specific criteria that are useful for its recognition. However,
distinguishing between in situ and transported phosphatic sediments has great
potential importance because: (1) it may serve to focus investigations con-
cerning phosphate genesis on in situ deposits, thus eliminating the complexity
of dealing with the complete spectrum of phosphorite accumulations; and (2) it
may aid in understanding the mode of accumulation of phosphatic sediments in a
variety of depositional environments.

OBJECTIVES

It is not our intent in this paper to directly address the question of
phosphate genesis. Rather, we propose to investigate the relationships that
exist between phosphatic sediments and their associated foraminiferal
assemblages within the Pungo River Formation of eastern North Carolina.
Detailed analyses of the physical stratigraphy and petrology of this formation
(Riggs et al., this volume; Scarborough and Riggs, this volume), combined with
analysis of its foraminiferal assemblages (Katrosh and Snyder, this volume),
provide an opportunity to determine the paleoenvironmental conditions in which
individual units of the formation were deposited. This, in turn, may reveal
which units represent in situ deposits and which, if any, appear to be the
result of sediment transport from areas of phosphate genesis into areas that
merely served as passive collection sites for phosphatic sediments.

The term in situ is used here to indicate phosphorites that accumulated within
areas of active phosphate formation. Certainly, minor amounts of transport and
reworking within the geographic confines of each phosphorite unit are not only
possible, but probable. However, those units interpreted as in situ should be
expected to reflect paleoenvironmental conditions that are in some way dif-
ferent from normal marine environments. Regardless of the explanation for
phosphate genesis that one might prefer, phosphorites are characteristically
associated with abnormal sediment packages (Riggs, 1980). Conversely,
transported phosphorites represent accumulations outside of the areas of
active phosphate formation. This does not imply that phosphate formation can-
not continue within adjacent portions of the region. Rather, it indicates
that the phosphorites of such units lay beyond the limits of environments con-
ductive to phosphate genesis. Transported units should, therefore, be
expected to reflect more normal marine conditions. The onset of normal marine
conditions within a sequence of phosphorites may simply indicate that migra-
tion of environments conductive to phosphate formation rather than complete
cessation of phosphate-generating mechanisms.

Perhaps the most direct way to interpret paleoenvironmental conditions is
through analysis of the fauna associated with sediments of the enclosing
units. It is precisely this approach, supplemented by information from the
sediments themselves, that has been employed during the present study.

STUDY AREA

For the purpose of investigating sediment-faunal relationships, the
phosphorite units and foraminiferal assemblages from a portion of the Aurora
Embayment have been selected. The study area includes Beaufort, Pamlico, and
portions of northern Craven counties (figure 1). Economically important
phosphate deposits in the Coastal Plain of North Carolina occur only in this
area, which lies to the north of a pre-Miocene topographic high in southern
Craven County. This high demarcates the southern boundary of the Aurora depo-
sitional embayment, and it represents an area where phosphorite units grade
southward into calcareous quartz sands with little or no phosphate content
(figure 2). Interpretations are based on data from the following cores: NCPC,
BTN-9, BTN-11, PON-1, PON-3, and PON-4 (figure 1). See Katrosh and Snyder
(this volume) for an explanation of the system used to designate the cores.

LEGEND

COUNTY
BOUNDARY

LOCATION AND
NUMBER OF CORE
EXAMINED FOR
THIS STUDY

APPROXIMATE
UPDIP LIMIT OF
MIDDLE MIOCENE
SEDIMENTS
(AFTER MILLER,
1971)

.: AURORA AREA

Figure 1. Map of study area showing location of cores.

P I T TV

.... ,

1o BTN1 I. L0 CAPE

N PON-4S

B AY

CAPE

ONSLOW
BAY -
SCALE

0 50 KM

-J
w 15-
UJ
-J20-
<4
025-
z
430-

:35-
o
0

S45-
-J

w50-

a

z

I.-
Cr o

UIJ

0 -J

C-

p R ,- I -

FINE TO MEDIUM, MUDDY, CALCAREOUS SHELL HASH; MINOR QUARTZ AND
PHOSPHATE

SHELLY, QUARTZ BEARING, PHOSPHATIC DOLOMITE

SHELLY, CALCAREOUS MUDS TO MEDIUM, CALCAREOUS, MUDDY QUARTZ
SANDS; VERY MINOR PHOSPHATE

FINE TO MEDIUM, MUDDY, PHOSPHORITE SANDS INTERBEDDED WITH
PHOSPHATIC, QUARTZ BEARING, MOLDIC LIMESTONES

FINE, DOLOMITIC, MUDDY PHOSPHORITE SANDS CAPPED BY
PHOSPHATIC, QUARTZ BEARING DOLOMITE

FINE TO MEDIUM, DOLOMITIC, MUDDY PHOSPHORITE SANDS
CAPPED BY PHOSPHATIC, QUARTZ BEARING DOLOMITE

PRE-MIOCENE TOPOGRAPHIC
HIGH (POSITION DETERMINED
FROM DATA OF MILLER,
1971)

HORIZONTAL SCALE

0 10 KM

V. E. 537.5

Figure 2. North south distribution of lithologic units within the Aurora Embayment (core NCPC
through core PON-1).

SE 0 E N T S

DO

E]

7

FA

	Title Page
	Front Matter
	Table of Contents
	Introduction
	Main

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
121	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file