The Geological Society of America Field Guide 42

Geology and geomorphology of the Carolina , Chesterfi eld County,

Christopher S. Swezey U.S. Geological Survey, 12201 Sunrise Valley Drive, MS 926A, Reston, Virginia 20192, USA

Bradley A. Fitzwater G. Richard Whittecar Old Dominion University, Department of Ocean, Earth, and Atmospheric Sciences, 4600 Elkhorn Avenue, Norfolk, Virginia 23529, USA

ABSTRACT

This two-day fi eld trip focuses on the geology and geomorphology of the Caro- lina Sandhills in Chesterfi eld County, South Carolina. This area is located in the updip portion of the U.S. Atlantic Coastal province, supports an ecosystem of longleaf pine (Pinus palustris) and wiregrass (Aristida stricta), and contains three major geologic map units: (1) An ~60–120-m-thick unit of weakly consolidated sand, sandstone, mud, and gravel is mapped as the Upper Cretaceous Middendorf Formation and is interpreted as a fl uvial deposit. This unit is capped by an uncon- formity, and displays reticulate mottling, plinthite, and other paleosol features at the unconformity. The Middendorf Formation is the largest aquifer in South Caro- lina. (2) A 0.3–10-m-thick unit of unconsolidated sand is mapped as the Quater- nary Pinehurst Formation and is interpreted as deposits of eolian sand sheets and derived via remobilization of sand from the underlying Cretaceous strata. This unit displays argillic horizons and abundant evidence of bioturbation by veg- etation. (3) A <3-m-thick unit of sand, pebbly sand, sandy mud, and mud is mapped as Quaternary terrace deposits adjacent to modern drainages. In addition to the geologic units listed above, a prominent geomorphologic feature in the study area is a north-trending escarpment (incised by headwater streams) that forms a mark- edly asymmetric drainage divide. This drainage divide, as well as the Quaternary terraces deposits, are interpreted as evidence of landscape disequilibrium (possibly geomorphic responses to Quaternary climate changes).

Swezey, C.S., Fitzwater, B.A., and Whittecar, G.R., 2016, Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina, in Doar, W.R., III, ed., , Structures, and Landforms in Central South Carolina—Field Guides for the 2016 GSA Southeastern Section Meeting, Columbia, South Carolina: Geological Society of America Field Guide 42, p. 9–36, doi:10.1130/2016.0042(02). © 2016 The Geological Society of America. All rights reserved. For permission to copy, contact [email protected].

9 10 Swezey et al.

INTRODUCTION age. In a geologic map of the Blaney quadrangle in Richland and Kershaw Counties, Ridgeway et al. (1966) described the sandhills This fi eld trip provides an overview of the geology and geo- as uniform sand of probably Tertiary age. In Lexington and Aiken morphology of the Carolina Sandhills in Chesterfi eld County, Counties, Kite (1987) provided brief descriptions of the sandhills, South Carolina. Most of the data presented in this paper are based and stated that their age was unknown. In a summary paper cover- on recent detailed geologic maps of the Middendorf and Patrick ing both South Carolina and , Nystrom et al. (1991) quadrangles of Chesterfi eld County (to be published by the South provided succinct descriptions of the sand thickness, topographic Carolina Geological Survey). These two quadrangles encompass expressions, and elevations, and estimated the sand to be of late most of the Carolina Sandhills National Wildlife Refuge and the Miocene age. In Chesterfi eld County, Leigh (1998) described six Sand Hills State , and contain the following three major profi les among the sandhills, and estimated the surfi cial sand to geologic map units: be of Tertiary age. (1) An ~60–120 m- (197–394-ft-) thick unit of weakly consoli- Along with uncertain age estimates, the depositional envi- dated sand, sandstone, mud, and gravel that is mapped as ronment of the Carolina Sandhills has been the subject of much the Upper Cretaceous Middendorf Formation. The unit has speculation. In South Carolina, speculation on the depositional several distinct lithologic facies, and it extends to the east environment of the surfi cial sand has ranged from eolian to sub- in the subsurface, where it forms one of the larger aquifers aqueous, and some researchers have postulated specifi c subaqueous in the southeastern . This unit is interpreted as environments ranging from fl uvial to marine (Cooke, 1936; Ridge- a fl uvial deposit that accumulated during a base-level low- way et al., 1966; Kite, 1987; Nystrom and Kite, 1988; Nystrom et stand and subsequent early transgression. al., 1991; Leigh, 1998). Many of the statements about depositional (2) A 0.3–10-m- (1–33-ft-) thick unit of unconsolidated sand environments, however, have been simply conjectures without that is mapped as the Quaternary Pinehurst Formation. much descriptive information to support the interpretations. This unit forms the “sandhills” of the region, and it extends Despite these uncertainties, several authors have noted a across most of the Middendorf and Patrick quadrangles. close association of the Carolina Sandhills with outcrops of This Quaternary sand is interpreted as eolian sand sheets underlying Cretaceous sand and sandstone. Cooke (1936), for and dunes derived from the underlying Cretaceous Mid- example, noted that the area of the Congaree Sand Hills corre- dendorf Formation during the last glaciation when climate sponds to the area in which the Cretaceous Tuscaloosa Forma- conditions were colder, windier, and more arid. tion (what is now called the Middendorf Formation) is exposed. (3) A <3-m- (10-ft-) thick unit of sand, pebbly sand, sandy Ridgeway et al. (1966) noted that the sandhills in Richland and mud, and/or mud that is mapped as Quaternary terrace Kershaw Counties overlie sand of the Cretaceous Middendorf deposits adjacent to the modern drainages. Formation, and they stated that the two units have similar grain The three map units listed above have different geomorpho- sizes, as well as similar abundance and composition of heavy logical expressions and have been subjected to different pedogenic minerals. They speculated that the overlying sand (“sandhills”) processes, many of which are discussed during this fi eld trip. was derived from reworking of the underlying Cretaceous sand. The Carolina Sandhills is a 15–60-km- (9–37-mi-) wide physiographic region of rolling sandy topography that extends ROAD LOG AND STOP DESCRIPTIONS from the western border of Georgia to central North Carolina along the updip (north and west) margin of the Coastal Plain This fi eld trip is focused on the Carolina Sandhills region in province in the southeastern United States (Fig. 1). The rec- Chesterfi eld County, South Carolina (Fig. 2). In this county, the ognition of the Carolina Sandhills as a separate physiographic sandhills occur on a relatively high plateau of Cretaceous strata, region dates from at least the work of McGee (1890, 1891), who bounded to the west by Paleozoic schist of the province described sand hills on the inner coastal plain of South Caro- and bounded to the east by the east-facing Orangeburg Scarp (Swift lina and North Carolina. Holmes (1893) delineated the “sand-hill and Heron, 1969). This scarp was named for a site at the City of country of the Carolinas” as a distinct geological region of the Orangeburg in Orangeburg County (South Carolina), where the inner coastal plain between the Savannah River of South Caro- toe of the scarp lies at ~64 m (210 ft) above sea level (Colquhoun, lina and the Neuse River of North Carolina, and Darton (1896) 1962). In Figure 2, the Orangeburg Scarp is a north-northeast– noted that the sandhill sediments extend across both South Caro- trending linear feature visible in western Lee County and western lina and Georgia. Cooke (1936) later used the term “Congaree Darlington County, where the topography changes abruptly from Sand Hills” to denote the extensive sand accumulations across ~60–80 m (197–262 ft) elevation (green, yellow) to 100–120 m the inner coastal plain of South Carolina. (328–394 ft) elevation (orange, red). The Orangeburg Scarp is Relatively few descriptions have been published on the Caro- interpreted as the shoreline formed by wave erosion at a time of lina Sandhills in South Carolina, and age estimates have been rather high sea level during the middle Pliocene (Dowsett and Cronin, tentative. In fi eld trip guides for Richland and Kershaw Counties, 1990). East of this shoreline lie extensive low-relief Coastal Plain Johnson (1961) and Otwell et al. (1966) provided brief descriptions terraces (composed of coastal and marine sediments) that are domi- of the sandhills, which they referred to as being of “post-Eocene” nated by oval topographic depressions called Carolina Bays, and Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 11 related geomorphological features formed by eolian processes (e.g., DAY 1 (2 April 2016) Ivester et al., 2002, 2003; Moore et al., 2014). West of the Orange- Drive from Columbia to the Carolina Sandhills National burg Scarp, no broad low-relief surfaces exist where Carolina Bays Wildlife Refuge. could form, and eolian sand sheets and dunes formed on the rolling, dissected uplands of the “sandhills” region. Upon arrival at the Carolina Sandhills National Wildlife Ref- There are ten stops scheduled for this fi eld trip (Fig. 3). The uge, the day begins with an overview of the refuge, followed by fi rst six stops are scheduled for Day 1, and are located in the a stop at an outcrop of the Cretaceous Middendorf Formation, western portion of the study area. The four stops scheduled for and a stop among the Quaternary eolian dunes (“sandhills”) that Day 2 are located in the eastern portion of the study area. are now stabilized by vegetation. After lunch at Lake Bee and a

84°W 80°W 76°W

Eocene 38°N = Area of Carolina Sandhills Virginia Crater (andAPPALACHIAN outcrop of Cretaceous MiddendorfPLATEAU Formation) = Province boundary = Counties in VALLEYFigure 2 AND North Carolina = StructuralRIDGE feature = Middendorf quadrangle PIEDMONT COASTAL and Patrick quadrangle PLAIN

34°N South Carolina Cape Fear Arch

34°N Georgia

COASTAL PLAIN Alabama ATLANTIC OCEAN Southeast Georgia Embayment

Southwest Georgia 30°N Embayment Pe ninsular Arch

0 50 100 miles Florida 30°N 0 50 100 kilometers

Ocala Arch GULF OF MEXICO

84°W 80°W

Figure 1. Location of the Carolina Sandhills in the southeastern United States. Outline of the Carolina Sandhills is from Griffi th et al. (2001, 2002). Locations of the Cape Fear Arch, Peninsular Arch, and Ocala Arch are from LeGrand (1961). Outline of Eocene crater is from Powars et al. (2015). The Middendorf quadrangle is immediately to the west of the Patrick quadrangle. 12 Swezey et al.

80°W

NC Piedmont Coastal Plain CHESTERFIELD SC Cheraw core MAP LOCATION Patrick core

Pee Dee River McBee

Hartsville

KERSHAW

DARLINGTON

Orangeburg Scarp

LEE ELEVATION 220 m 200 m 80°W North

100 m 0 5 10 miles 05 10 km 34°N 34°N 10 m

Figure 2. Shaded relief map of Kershaw, Chesterfi eld, Lee, and Darlington Counties, South Carolina (Albers projection, 1983 North American Datum). Two-meter elevation data are derived from LiDAR point cloud data (available online at www.dnr.sc.gov/GIS/lidar.html, accessed 19 July 2015). The Orangeburg Scarp is a north-northeast–trending linear feature visible in western Lee County and western Darlington County, where the topography changes abruptly from ~60– 80 m (197–262 ft) elevation (green, yellow) to 100–120 m (328–394 ft) elevation (orange, red). Black rectangles show the locations of the Middendorf quadrangle on the west and the Patrick quadrangle on the east, and both quadrangles are shown in greater detail in Figure 3. Blue circles show locations of cores discussed at Stop 6. NC—North Carolina; SC—South Carolina.

look at nearby outcrops of the Cretaceous Middendorf Forma- pine trees in the southeastern United States generally prefer tion, there is a stop at an outcrop displaying extreme pedogenic sandy substrates, and the names of many of the towns in the alteration (“plinthite”) at the top of the Middendorf Formation, Carolina Sandhills region have some form of the word “pine” and another stop near the Middendorf type section at a pit where (e.g., Pinehurst, Southern Pines). Within the Carolina Sandhills the U.S. Geological Survey (USGS) obtained a 85.3 m (280 ft) of Chesterfi eld County, the predominant substrates are sand of deep core in the Middendorf Formation. the Quaternary Pinehurst Formation and the underlying Cre- taceous Middendorf Formation. Accordingly, the vegetation STOP 1: Headquarters of Carolina Sandhills National of the refuge is dominated by trees of longleaf pine. There is Wildlife Refuge in Middendorf Quadrangle also a scattered understory of scrub oak (turkey oak, dwarf post (N 34° 30′ 13.48″, W 80° 13′ 29.92″) oak), and a groundcover of wiregrass (Christensen, 2000; Ear- ley, 2004; Sorrie, 2011). On outcrops of clay (within the Cre- Many of the plant species of the Carolina Sandhills show taceous Middendorf Formation), however, blackjack oak may distinct associations with the underlying geology. For example, be more common than other trees (Askins, 2010). In addition, Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 13 eld

34°37'30''N 34°30'N

80°00'W

80°00'W North 0 10 1

2 miles Juniper Creek Juniper 2 kilometers Figure 24 lidar.html, accessed 19 July 2015). Location lidar.html, 9 Patrick 0 0 221 m m 17 200 m 150 m 100 m m 50 Figure 11 ELEVATION 8 6 y eld trip stops. NWR—National Wildlife Refuge. Wildlife eld trip stops. NWR—National 7

80°07'30''W 80°07'30''W eld County, South Carolina (Albers projection, 1983 North American Datum) showing fi American Datum) showing South Carolina (Albers projection, 1983 North eld County,

Big Black Creek Middendorf Middendorf Quadrangle Quadrangle Patrick 2 1 3 4 FIELD TRIP STOPS: 1 = Carolina Sandhills NWR Headquarters Road Wire on Old 2 = Middendorf Formation Drive Wildlife on 3 = Pinehurst Formation Bee at Lake 4 = Middendorf Formation Road 4 Wire on 5 = Middendorf Formation core pit near Middendorf type section 6 = Patrick Scotch Road off 7 = Drainage divide at Sugarloaf 8 = Middendorf Formation on Isaac Road 9 = Pinehurst Formation propert Wilkes on and Pinehurst Formation Terraces 10 = Figure 10

80°15'W

80°15'W

34°30'N

34°37'30''N STOP 5: STOP 3.7 km (2.3 miles) southwest of image is shown in Figure 2. White circles and adjacent numbers (1–10) denote fi in Figure 2. of image is shown Figure 3. Shaded relief map of the Middendorf and Patrick quadrangles, Chesterfi Figure 3. Shaded relief map of the Middendorf and Patrick online at www.dnr.sc.gov/GIS/ point cloud data (available from LiDAR data are derived elevation Two-meter trip stop localities. 14 Swezey et al.

pond cypress and sweetgum are present on terraces adjacent to Wildlife Service (Askins, 2010). The original purposes of the ref- some of the creeks. uge were to provide habitat for migratory birds, to provide wildlife- Stop 1 provides an opportunity to discuss the ecological sig- dependent recreation opportunities, and to demonstrate land man- nifi cance and history of the Carolina Sandhills National Wildlife agement practices that would enhance the conservation of natural Refuge, a region known for an ecosystem of longleaf pine (Pinus resources. During the 1950s and 1960s, loblolly pine and slash pine palustris) and wiregrass (Aristida stricta). Longleaf pine is able (which are not native) were planted to replace much of the longleaf to grow in a wide variety of settings, and the “sandhills” is just pine. Today, however, the loblolly pine and slash pine are being har- one type of longleaf pine community (Fig. 4). The longleaf pine vested, these areas are being replanted with longleaf pine, and the and wiregrass ecosystem is maintained by frequent low-intensity refuge is managed primarily to restore and maintain the longleaf fi res that facilitate the reproduction of the trees, wiregrass, and pine and wiregrass ecosystem (Askins, 2010). associated plants (Earley, 2004; Askins, 2010). The longleaf pine is particularly well adapted to fi re because of features such as STOP 2: Outcrop of Middendorf Formation in Pit on Old thick bark, large seed size, inconsistent seeding, fall seed sprout- Wire Road within the Carolina Sandhills National Wildlife ing, and slow growth during the early years of the tree. Many Refuge in Middendorf Quadrangle (N 34° 31′ 36.89″, details of the role of fi re in this ecosystem are not well under- W 80° 12′ 57.71″) stood, but it is known that fi re reduces the amount of ground lit- ter, reduces competition from other tree species, returns nutrients Stop 2 is a borrow pit located off Old Wire Road (Fig. 5) that to the soil, and inhibits certain pathogens that affl ict longleaf pine displays good exposures of the Middendorf Formation. This road is and other associated species (Earley, 2004; Askins, 2010). the remnants of an old stage-coach route along which was strung an As described in Earley (2004) and Finch et al. (2012), the early telegraph wire line. During the American Civil War, General longleaf pine and savannas of the southeastern United William T. Sherman’s Union army marched along Old Wire Road States once formed one of the more extensive woodland eco- from Columbia to Cheraw and eventually into North Carolina. systems in North America. When the Spanish arrived during The pit at Stop 2 is one of the better outcrops in the wild- the early 1500s, longleaf pine was the dominant tree that cov- life refuge, and it shows the predominant Middendorf lithology ered ~60% of the Atlantic and Gulf Coastal , covering of grayish-red (5R 4/2) to dark yellowish-orange (10YR 6/6) a wide swath of every coastal state from the James River in medium to coarse sand or sandstone.1 At most locations, the sand Virginia to the shores of Lake Okeechobee in Florida and west or sandstone is moderately to poorly sorted. The degree of lithifi - into southeastern Texas. The longleaf pine was the dominant cation is quite variable over short distances, ranging from loose tree over ~242,800 km2 (93,745 mi2) of the southeast, and the sand, to sand or sandstone held together by kaolin matrix, to iron- longleaf pine occurred together with other pines and hard- cemented sandstone. woods on an additional 121,400 km2 (46,873 mi2). Today, how- Fossils are not common in the Middendorf Formation. Berry ever, only 1.4% of the Atlantic and Gulf Coastal Plains support (1910, 1914) identifi ed fossil leaves in a clay bed at the Midden- longleaf pine. In 1996, it was estimated that only 11,938 km2 dorf type section (near Stop 6), and estimated the age of the leaves (4609 mi2) remained of longleaf pine forest, and much of this to be early Late Cretaceous (possibly Cenomanian–Turonian). remaining area consisted of small fragments of land (Earley, Pieces of petrifi ed wood have been identifi ed at several outcrops 2004). In other words, the longleaf pine ecosystem has experi- in the Middendorf and Patrick quadrangles, and some of these enced a decline of ~95%. pieces have been identifi ed as conifer wood (Fitzwater et al., The longleaf pine is exceptionally strong and resistant to 2014a). A core in the Middendorf Formation at Cheraw State rot, and the tree was harvested for timber (especially ship build- Park, which is ~32 km (20 mi) northeast of Stop 2, has yielded ing). The longleaf pine was also a major source of tar, pitch, and Late Cretaceous palynomorphs including abundant fern spores turpentine. Prior to 1860, however, the cutting of longleaf pine (Cicatricosisporites, Gleicheniidites, Foveotriletes), as well as timber and getting it to market was a slow and diffi cult process. gymnosperm pollen (Taxodiaceae and Pinuspollenites), angio- By the time of the American Civil War (1861–1865), the longleaf sperm pollen (Momipites), and freshwater algae (Schizosporis). pine forest was essentially extirpated from Virginia and north- These palynomorphs were identifi ed by Christopher Bernhardt eastern North Carolina, but much of the longleaf pine forest to (USGS), and are reported in Swezey et al. (2015). the south remained relatively intact. After 1865, however, the On the east wall of the pit at Stop 2 (Fig. 6), the following lumber industry focused on the southeastern United States and two stratigraphic units are apparent: (1) a lower unit of medium with the use of railroads, essentially destroyed the remaining to coarse sand with east-dipping cross-bedding; and (2) an upper longleaf pine forest within 50 years (Earley, 2004). unit of medium to coarse sand with approximately horizontal During the 1930s, the U.S. Federal Government purchased planar bedding. The sand-sized grains consist predominantly of agriculturally poor and/or damaged land in the Sandhills region of Chesterfi eld County (South Carolina), and in 1939, established this land as the Carolina Sandhills National Wildlife Refuge to be man- 1Unless otherwise stated, the color nomenclature used in this guide is from the aged by what is now the U.S. Department of the Interior Fish and Geological Society of America rock color chart by Goddard et al. (1963). Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 15

Figure 4. Longleaf pine (Pinus palus- tris) in sandy soil of the Carolina Sand- hills National Wildlife Refuge. Circled twenty-eight-cm- (11-in-) long hammer at base of small tree in foreground pro- vides a sense of scale.

quartz with some mica and opaque minerals, and in places, there 6) reveal that such clay beds become increasingly rare lower in is a matrix of kaolin clay among the sand grains. In addition, the the formation. lower unit with cross-bedding contains some clasts of kaolin clay. On the west wall of the pit (Fig. 8), the strata consist of On the south wall of the pit (Fig. 7), the same two strati- medium to coarse sand composed predominantly of quartz, but graphic units are apparent, and beds (lenses) of clay are also vis- primary sedimentary structures are not apparent and separate ible. An ~3-m- (9.8-ft-) long by 0.3-m- (1.0-ft-) thick clay bed depositional units cannot be distinguished. This outcrop in which (lens) is located at the contact between the lower unit with cross- sedimentary structures are not apparent is typical of most out- bedding and the upper unit with horizontal planar bedding. Clay crops of the Middendorf Formation. beds (lenses) of this size are relatively common within the upper The strata exposed at Stop 2 are interpreted as fl uvial sediments, portion of the Middendorf Formation, but core data (seen at Stop primarily channel deposits. The scarcity of distinct fi ning-upward

Figure 5. Stop 2 borrow pit showing the Upper Cretaceous Middendorf Forma- tion. View is toward the south. Vehicle provides a sense of scale. 16 Swezey et al.

upper unit: horizontal planar bedding

Figure 6. East wall of pit at Stop 2 showing cross-bedding in the Upper lower unit: cross-bedding Cretaceous Middendorf Formation. Twenty-eight-cm- (11-in-) long hammer provides a sense of scale.

sequences suggests that point bars (which are typical of mean- plain deposit, or some form of post-depositional accumulation of dering rivers) may have been relatively scarce. Another possibil- clay within the strata. The sand- and gravel-size clasts of clay are ity is that the upper portions of fi ning-upward sequences have interpreted as rip-up clasts eroded from larger clay beds and/or been eroded by subsequent channel deposits. Nevertheless, the clay lenses. Most of the clay matrix among the quartz sand grains relatively coarse grain size and the east-dipping cross-bedding is thought to be post-depositional. The absence of primary depo- are consistent with a proximal fl uvial system fl owing from the sitional features in the west wall of the pit is attributed to post- Piedmont toward the coast. The clay lens might represent the depositional pedogenic alteration. More evidence of pedogenic fi nal stage of accumulation in an abandoned channel, a fl ood- alteration of the upper Middendorf Formation is seen at Stop 5.

upper unit Figure 7. South wall of pit at Stop 2 clay lens showing clay bed (lens) at the contact be- tween the lower unit with cross- bedding and the upper unit with horizontal planar bedding, Upper Cretaceous Middendorf lower unit Formation. Twenty-eight-cm- (11 in-) long hammer provides a sense of scale. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 17

Figure 8. West wall of pit at Stop 2 showing Upper Cretaceous Middendorf Formation without evidence of obvi- ous primary sedimentary structures. Twenty-eight-cm- (11-in-) long hammer provides a sense of scale.

STOP 3: Overlook of “Sandhills” (Pinehurst Formation) optically stimulated luminescence (OSL). These observations on Wildlife Drive within the Carolina Sandhills National and data provide an informed framework for interpreting the ori- Wildlife Refuge in Middendorf Quadrangle gin and signifi cance of the “sandhills.” (N 34° 31′ 33.30″, W 80° 13′ 39.89″) Note the sign that displays information about the red- cockaded woodpecker (Fig. 9). More than 30 plant and animal Stop 3 is a parking area on the east side of Wildlife Drive species associated with the longleaf pine ecosystem are listed as within the refuge. This site provides a good location to observe threatened or endangered (Askins, 2010). In ways that are still the “sandhills” that cap the rolling landscape in both the Midden- poorly understood, many of these species require interactions dorf and Patrick quadrangles, and to present data derived from with the longleaf pine and with frequent low-intensity fi res. The geologic mapping, LiDAR, ground-penetrating radar (GPR), and most famous endangered species of the refuge is probably the

Figure 9. View of “sandhills” landscape (Quaternary Pinehurst Formation) at Stop 3. Vehicle provides a sense of scale.

red-cockaded woodpecker sign 18 Swezey et al. red-cockaded woodpecker (Picoides borealis), which prefers to the sediment is of medium sand (upper) to coarse sand (lower) excavate nesting and roosting cavities in living trees of longleaf size, using the sediment size terminology of Wentworth (1922) pine that are 80–120 yr old (Askins, 2010). The loss of longleaf and Folk (1954, 1980). Sieving analyses using mesh sizes at pine habitat has corresponded with a rapid decline in the red- 0.5 phi increments (following Folk and Ward, 1957; Folk, 1966) cockaded woodpecker population, and in 1970 the red-cockaded revealed grain sizes ranging from fi ne sand (lower) to coarse woodpecker was listed as an endangered species. Within the Car- sand (lower), with the most frequently occurring grain size olina Sandhills National Wildlife Refuge, a white band has been being medium sand (upper) to coarse sand (lower), which is 1.5– painted around trees that have been identifi ed as having nesting 0.74 phi (0.35–0.59 mm). Sediment sorting values (sφ) calculated or roosting cavities for these woodpeckers (Askins, 2010). from cumulative percent curves using the sorting formula of Folk Stop 3 provides an excellent view of the “sandhills” that and Ward (1957) ranged from 0.76 to 1.65 (moderately sorted to overlie the Cretaceous Middendorf Formation. The “sandhills” poorly sorted). The sand-sized grains consist predominantly of are a surfi cial unit of unconsolidated sand that is present across quartz with some mica and opaque minerals. Most of the quartz all of the Middendorf quadrangle and the Patrick quadrangle, grains of medium sand size and coarser were subrounded to sub- except along the modern creek fl oodplains and the lowest terrace angular, ranging from high sphericity to low sphericity (using the above some of the modern fl oodplains (seen at Stop 10). At many sphericity and roundness terminology of Powers, 1953). Textural locations, this unconsolidated sand is <1.5-m- (4.9-ft-) thick and maturity of individual samples ranged from immature to subma- does not display any characteristic morphological expression. ture, using the sediment textural maturity terminology of Folk Images derived from LiDAR data, however, reveal that in areas (1951, 1954). of higher elevation (such as Stop 3), the sand forms subdued hills Outcrops of the unconsolidated sand do not display primary of up to 6 m (19.7 ft) relief with steeper sides on the east and sedimentary structures, although some structures are visible in southeast (Fig. 10). In these areas, the total thickness of the sand GPR data that were obtained by Kerby M. Dobbs (Fitzwater et al., can be more than 6 m (19.7 ft). 2014b). Several GPR traverses across the sandhills, for example, The unconsolidated surfi cial sediment of the “sandhills” is have revealed 2–5-m-thick sets of southeast-dipping cross-bedding composed of grayish-orange (10YR 7/4) sand and muddy sand. at depths below 2 m (6.6 ft) (Fig. 11). In most examples, the dip Visual inspection of samples in the fi eld indicated that most of direction of the cross-bedding is consistent with the dip of the

North

ELEVATION 221 m 200 m 150 m 2 100 m 3 50 m 17 m

0 1 mile

01 km Figure 10. Detail of shaded relief map shown in Figure 3. This detailed image shows the subdued hills of the “sandhills,” with steeper sides on the east and southeast sides of the hills. White circles and adjacent numbers (2, 3) denote fi eld trip stops. The location of Figure 10 is shown in Figure 3. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 19

Thin white lines show the trace of southeast-dipping cross-bedding in the lower Pinehurst Formation Northwest Southeast 0 10 20 30 40 50 60 m 0

4

Pinehurst Formation Pinehurst 8 Formation

12 m Middendorf Formation Formation Middendorf

Figure 11. Ground-penetrating radar (GPR) traverse over one of the sandhills in the Patrick quadrangle, Chesterfi eld County, South Carolina. Location of traverse is shown in Figure 3.

steepest side of the sandhill. The GPR data also show that there is the range of 73–19 ka, coincident with the last glaciation in the considerable relief on the unconformity that caps the Cretaceous Northern Hemisphere. Additional details and the signifi cance of Middendorf Formation. For example, up to 5 m (16.4 ft) of relief these OSL ages are presented and discussed at Stop 9. is present on the unconformity in this portion of the refuge. In Chesterfi eld County, the unconsolidated surfi cial sand The unit of unconsolidated surfi cial sand within the Carolina that is mapped as the Quaternary Pinehurst Formation is inter- Sandhills is mapped as the Quaternary Pinehurst Formation. Con- preted as eolian sand sheets and dunes derived from the immedi- ley (1962) gave the name Pinehurst Formation to a 3–46-m- (10– ately underlying Cretaceous sand and sandstone, and stabilized 150-ft-) thick unit of non-fossiliferous sand and gravel on top of the by vegetation under prevailing climate conditions. This interpre- higher coastal plain hills in Moore County, North Carolina. Bartlett tation is supported by the OSL ages, the sandhill locations, the (1967) later examined the Pinehurst type section and delineated a sandhill morphology as visible in LiDAR imagery, and the GPR lower unit of sand and clay that he mapped as the Cretaceous Mid- data. Furthermore, the OSL ages from the Carolina Sandhills are dendorf Formation, which is capped by an unconformity, above coincident with other OSL ages from known eolian deposits and which lies loose sand. Bartlett (1967) restricted the term Pinehurst features elsewhere in the southeastern United States (e.g., Marke- Formation to the loose sand that overlies this unconformity. A sub- wich and Markewich, 1994; Ivester et al., 2001; Swezey et al., sequent summary by Nystrom et al. (1991) stated that widespread 2013; Moore et al., 2014). deposits of sand extend discontinuously across the upper Coastal These new data eliminate several alternative interpretations Plain from northern Aiken County (South Carolina) northeastward that have been suggested for the surfi cial deposits in the Caro- to the North Carolina state line, and that these deposits are the lina Sandhills. For example, the OSL ages rule out the possibility southwestern continuation of the Pinehurst Formation as redefi ned of the sand being beach deposits, because sea level was likely by Bartlett (1967). This unit of sand overlies an unconformity with to have been well below the elevation of the Sandhills during signifi cant local relief. the time of the last glaciation. The OSL ages rule against fl uvial OSL dating, which provides an age of the last time that a deposits, because the sand blankets most of the landscape and particular quartz grain was exposed to sunlight, has provided is not spatially associated with obvious Quaternary fl uvial chan- a numerical chronology for the sand. Using this technique, 11 nels. The OSL ages also rule out the possibility of the sand being samples of the Pinehurst Formation in the Middendorf and Pat- associated with the Chesapeake Bay impact crater, which formed rick quadrangles have yielded mean OSL ages ranging from during the Eocene. ca. 73–19 thousand years (ka) ago and 6 samples have yielded The location of the sand of the Pinehurst Formation puts some mean OSL ages from ca. 12–8 ka (Swezey et al., 2016). Four- constraints on possible sediment sources and rules out some pos- teen of these ages were determined in 2012–2014 by Shannon sible depositional environments. As stated above, the sand is not A. Mahan (USGS, Denver, Colorado), and three of these ages located adjacent to obvious Quaternary fl uvial channels that might were determined in 2015 by George A. Brook (University of have provided a sediment source. However, the sand is located Georgia, Athens, Georgia). Most of these 17 OSL ages are within exclusively in areas where sand or sandstone of the Cretaceous 20 Swezey et al.

Middendorf Formation is near the surface (as noted by Cooke, sional spots of pale red (5R 6/2). This sandstone forms a distinct 1936; Ridgeway et al., 1966). Furthermore, the relatively coarse facies of the Middendorf Formation. The sandstone consists of grain size and poor sorting of the sand suggest that the sand of moderately sorted medium sand composed primarily of quartz the Pinehurst Formation has not traveled very far from its source. (Fig. 12). Fossils have not been found in this facies of the Mid- This spatial association of the unconsolidated Quaternary sand dendorf Formation. The red spots are caused by an iron cement, with outcrops of the Cretaceous sand and sandstone (and the and some of these spots are centered on vugs with diameters of relatively coarse grain size and poor sorting of the Quaternary ~0.5 cm (0.2 in). Primary sedimentary structures are not obvious, sand) strongly suggests that the Quaternary sand is derived from but there is a hint at this outcrop, and other nearby outcrops, that the underlying Cretaceous strata. As noted by Ridgeway et al. the unit consists of relatively horizontal to slightly undulatory (1966), the similarity of grain sizes and heavy mineral suites also beds that are 0.2–0.6 m (0.7–2.0 ft) thick. supports the interpretation that the Quaternary unconsolidated This unit forms distinct ledges that crop out at ~107 m sand was derived from the underlying Cretaceous sand. (350 ft) elevation throughout the western portion of the Midden- In relatively fl at, topographically high areas, the unconsoli- dorf quadrangle. At a location 0.3–0.6 km (0.2–0.4 mi) down- dated sand of the Pinehurst Formation forms subdued hills of up stream from Lake Bee along the north side of Hemp Branch, this to 6 m (19.7 ft) relief, with steeper sides on the east and southeast unit forms a very prominent series of ledges, and water seeps (Fig. 10). The location of the steeper sides of the hills is consis- from beneath these ledges at times when much of the rest of the tent with the eastward dip directions of cross-bedding in GPR wildlife refuge is relatively dry. These ledges provide some of data, thus suggesting that the hill morphology is indeed relict the very few locations where one might attempt to measure strike depositional topography of bedforms rather than pure erosional and dip, and they show that the strata in this area are very near topography unrelated to depositional processes. to horizontal. Regional context suggests that there is probably a LUNCH: Lake Bee picnic area (N 34° 34′ 41.41″, very slight dip to the east. W 80° 14′ 11.69″). The yellowish-gray (5Y 8/1) to grayish-pink (5R 8/2) sand- Three species of carnivorous plants are present within the Car- stone exposed at this stop is interpreted as fl uvial sediments. In olina Sandhills National Wildlife Refuge, and the largest of these the absence of primary sedimentary structures, however, it is dif- plants may be seen along the edge of Lake Bee. The perennial yel- fi cult to situate this facies more precisely within the fl uvial sys- low pitcher-plant (Sarracenia fl ava) or “trumpet” plant forms leaf tem context. Nevertheless, the facies is widespread in the western tubes up to 1.0 m (3.3 ft) tall, and grows in moist that are poor portion of the Middendorf quadrangle, and it occurs at a very in nitrogen (Sorrie, 2011). The pitcher-plants compensate for this predictable elevation in the upper portion of the Middendorf For- nitrogen defi ciency by trapping insects in leaf tubes and digesting mation. This facies is also found at elevations of 64–85 m (210– the insects using chemicals secreted by the leaves. 280 ft) in the Patrick quadrangle. Various adaptations allow these carnivorous plant species to compete in nutrient-poor soils such as are common in the STOP 5: Outcrop of Middendorf Formation (with Plinthite) sandhills and in wetlands. Groundwater seeps and toe-slope on Wire Road 4 within the Carolina Sandhills National springs used to be much more abundant across the Coastal Plain Wildlife Refuge 3.7 km (2.3 mi) West of the Middendorf when the longleaf pine and wiregrass savanna were the domi- Quadrangle (N 34° 29′ 40.72″, W 80° 16′ 51.93″) nant upland vegetation. The more dense foliage in present native forests intercepts and transpires much greater amounts of rain- Stop 5 is located on the south side of Wire Road 4, ~3.7 km fall, soil moisture, and shallow groundwater than savannas with (2.3 mi) west of the Middendorf quadrangle and 2.4 km (1.5 mi) widely space trees and open understory (e.g., Riekerk 1983; Win- east of Highway 151, within the southwestern portion of the ref- ston, 1995; Steven and Toner, 2004). In one Coastal Plain site in uge. This stop shows an excellent exposure of the uppermost por- southern Virginia where oak-pine vegetation was removed and tion of the Cretaceous Middendorf Formation where it is deeply longleaf pine savannas are being restored, the recharge to shallow weathered (Fig. 13). This stop provides an opportunity to discuss groundwater systems increased by 30% during years with normal the characteristics of plinthite-bearing soil profi les in the Coastal precipitation (McLeod et al., 2012). The resultant higher water Plain and their geomorphic and stratigraphic signifi cance. For tables and increased number of groundwater-seepage wetlands, this stop, the color nomenclature used is from the Munsell Soil along with the frequent fi res that kill competing vegetation, allow Color Chart (Munsell Color, 2000). the carnivorous plants to thrive. Several distinct beds are visible in this outcrop of the Mid- dendorf Formation. Most of these beds were originally sandy, but STOP 4: Middendorf Outcrops at Lake Bee Picnic Area have been greatly modifi ed by pedogenic processes. Primary sed- within the Carolina Sandhills National Wildlife Refuge in imentary structures are not obvious because of the overprinting Middendorf Quadrangle (N 34° 34′ 41.41″, W 80° 14′ 11.69″) of pedogenic clay and remobilized iron in the soil. The network of irregular lines and blobs marked by tan, yellow, orange, and Stop 4 at the Lake Bee picnic area shows outcrops of yellow- red is called reticulate mottling. At this site, the reticulate mot- ish-gray (5Y 8/1) to grayish-pink (5R 8/2) sandstone with occa- tling dominates a meter-thick zone across the entire exposure. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 21

Figure 12. Yellowish-gray to grayish- pink sandstone of the Upper Cretaceous Middendorf Formation at Stop 4 near Lake Bee. Twenty-eight-cm- (11-in-) long hammer provides a sense of scale.

Within this zone are strong brown (7.5YR 5/6) masses of oxi- such as occurs in the zone of a fl uctuating water table. Because dized iron, and light brownish-gray (10YR 6/2) iron-depleted the water table is no longer close to the surface at this well- masses. Beneath this zone for at least another meter, the reticu- drained site, the iron masses and depletions are interpreted as late mottling becomes less dominant but is still common. Above relict redoximorphic features (Morton, 1995). the zone of reticulate mottling, the Middendorf strata are a more Plinthite masses are iron-rich mixtures of kaolinitic clay uniform red (10R 5/8) sandy loam with a few depleted mottles, with quartz, and these masses commonly have well-developed many of which resemble root paths or animal burrows. reticulate mottling (Soil Survey Staff, 2006). What is now termed According to Soil Survey Staff (2006), reticulate mottling “plinthite” was called “laterite” in some older soil literature forms by the dissolution and then precipitation of iron when the (Schaetzl and Thompson, 2015). In plinthite, when the soil is soil experiences alternating reducing and oxidizing conditions, moist, the concentrations of iron particles can be soft enough

Figure 13. Plinthite at the top of the Upper Cretaceous Middendorf Forma- tion at Stop 5. Twenty-eight-cm (11-in-) long hammer provides a sense of scale. 22 Swezey et al. to be sliced with a shovel. If plinthite dries and hardens, then have removed kilometers of landscape prior to the Miocene, and the iron particles cannot be re-softened, and the resulting larger some regions of the Piedmont may have been eroded by marine hardened masses are called ironstone (Soil Survey Staff, 2006) or transgressions during the Miocene (e.g., Weems and Edwards, petroplinthite (Eze et al., 2014). If eroded and transported from 2007). Even so, climates in the Carolinas during some parts of an outcrop, then plinthite and/or ironstone particles and masses the Miocene were notably warmer than present (e.g., Poag and may survive transport by slope processes, currents, and/or waves, Sevon, 1989), and areas even as far north as Delaware may have and may be found as reworked clasts in younger sediment accu- been subjected to tropical to subtropical conditions for millions mulations (e.g., Whittecar et al., 1993, 1994). of years (e.g., Ward, 1998). Thus, deep and intense weathering At the Stop 5 outcrop, the plinthite and reticulate mottling conditions during the Miocene may have developed long enough are interpreted as pedogenic features that formed on the uncon- for existing relict landscapes to form some of the plinthite and formity at the top of the Middendorf Formation. At many sites reticulate mottling seen in outcrops at this stop. throughout the Middendorf and Patrick quadrangles, the upper- In summary, plinthite in the inner coastal plain is a clear most Middendorf Formation and the unconformity that caps it indication of ancient soils that formed at sites with high fl uctuat- show evidence of a substantial pedogenic overprint, but typically ing water tables. Such plinthite may be the result of deep tropi- only the lower portions of deeply weathered soil profi les with cal weathering that occurred during pre-Pliocene epochs. The occasional redoximorphic features have been preserved. Other plinthite could also have formed during a warm but cooler-than- indicators of intense and/or long-lasting chemical weathering tropical climate, if given enough time and a very stable landscape. occur throughout the formation (e.g., clay pseudomorphs or “ghosts” of feldspar sand, clasts and beds of mud that has altered STOP 6: Patrick Core Pit 305 m (1000 ft) East of Ruby– to kaolin). As mentioned previously, GPR data show evidence Hartsville Road within the Sand Hills State Forest in of substantial relief on this unconformity (at least 5 m or 16.4 ft Middendorf Quadrangle (N 34° 33′ 10.98″, W 80° 07′ 40.76″) locally), suggesting that much erosion has taken place in addition to the pedogenic modifi cations. Stop 6 is located 305 m (1000 ft) east of Ruby–Hartsville One interpretation of these features is that they developed Road within the Sand Hills State Forest. As described by Askins during periods of intense and long-lasting weathering under (2010), this property was once part of the Carolina Sandhills humid tropical conditions. Plinthites are common in many oxi- National Wildlife Refuge. In 1991, however, a portion of the Car- sols that form today in humid tropical and subtropical environ- olina Sandhills refuge was transferred from the Federal govern- ments within poorly and somewhat poorly drained soils (Eze et ment to the South Carolina Forestry Commission, thus forming al., 2014; Schaetzl and Thompson 2015). High temperatures, a the Sand Hills State Forest. landscape with abundant moisture, and a fl uctuating water table Stop 6 is a borrow pit that provides an excellent exposure seem to be necessary conditions for their formation (e.g., Osher of the unconformity between the Upper Cretaceous Midden- and Buol, 1998), and the availability of iron aids the process dorf Formation and the overlying Quaternary Pinehurst Forma- considerably (e.g., Smith, 2007). Thus the presence of plinthite tion (Fig. 14). This pit is also the location where the USGS may indicate that the landscape at this site used to be a lowland drilled an 85.3-m- (280-ft-) deep core in March 2014. This pit in a tropical or subtropical climate. The common occurrence of is located 0.3 km (0.2 mi) northeast of the type section of the plinthitic soils on dissected uplands (such as here in the Caro- Middendorf Formation. lina Sandhills) would indicate a geomorphic history of stream This stop is an excellent place to discuss stratigraphic nomen- incision and topographic inversion that changed lowlands into an clature. Darton (1896) described a several-hundred-ft-thick unit upland plateau (Eze et al., 2014). of Cretaceous sand, sandstone, and clay extending from Colum- In the southeastern United States, more than 50 soil series bia to Cheraw, and he designated this unit as the Potomac Group contain plinthitic features (Shaw et al., 2011). These plinthitic after similar strata described and named by McGee (1886) from soils are present in high-elevation Coastal Plain settings, and they outcrops along the Potomac River near Washington, D.C. Sloan are reported from Pliocene- and Miocene-age surfaces (e.g., Cady (1904, 1907), however, indicated that a precise equivalent of the and Daniels, 1968). In southeastern Virginia, Miocene deposits Potomac strata is not present in South Carolina, and he subdi- along the Fall Zone contain in situ plinthite concentrations, but in vided the approximately equivalent South Carolina strata into Pliocene surfaces, the ironstone particles formed from plinthite (1) a lower unit of white and yellow sand with clay beds and kaolin are rounded clasts that have been reworked within shore-zone (collectively named the Hamburg Beds after the now-abandoned sediments (Whittecar et al., 1994). community of Hamburg in Aiken County) and (2) an upper unit Although it is possible that some of the features of chemical of white, yellow, orange, pink, and purple sand with clay beds weathering in the Middendorf Formation date from periods as old and kaolin (collectively named the Middendorf Beds after the as the Cretaceous, it is generally very diffi cult in the southeastern community of Middendorf in Chesterfi eld County). In subse- United States to demonstrate that an existing landscape surface quent studies, Berry (1910, 1914) used the term “Middendorf and its related weathering phenomena are older than the Miocene arkose member of the Black Creek Formation” in a study of Cre- (e.g., Whittecar et al., 2016). Stream incision and slope processes taceous plant fossils near the community of Middendorf. Cooke Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 23

26.5 ± 1.8 ka

Quaternary Pinehurst Formation Figure 14. Stop 6 outcrop showing the 46.0 ± 2.7 ka Upper Cretaceous Middendorf Forma- unconformity tion, capped by unconformity, which is overlain by Quaternary Pinehurst For- mation. Shovel and hoe provide a sense Cretaceous Middendorf Formation of scale. Also shown are weighted mean OSL ages determined by Shannon A. Mahan (USGS).

(1926) used the term “Middendorf Formation,” and he stated that Wright, 1976). Since then, Woollen and Colquhoun (1977) used this formation should include both the “Hamburg Beds” and the the term Middendorf Formation in Chesterfi eld and Darlington “Middendorf Beds” of Sloan (1904, 1907). Cooke (1936) desig- Counties, Owens (1989) used the term Middendorf Formation on nated a specifi c type locality of the Middendorf Formation as an his 1:250,000-scale geologic map of northeastern South Carolina outcrop on the railroad line at the crossing of the Hartsville–Ruby and southeastern North Carolina, and Prowell et al. (2003) used Road, 2 mi northeast of Middendorf, in Chesterfi eld County, the term Middendorf Formation in the vicinity of the Middendorf South Carolina (the type locality is 0.3 km [0.2 mi] southwest of type locality. Recent geologic maps by Fitzwater, Swezey, and Stop 6). Cooke (1936) also stated that the Middendorf Formation Whittecar of the 1:24,000-scale Middendorf quadrangle and the in South Carolina extends across Georgia and is identical to the Patrick quadrangle have also used the stratigraphic term of Mid- Tuscaloosa Formation, which is an older stratigraphic term from dendorf Formation. Alabama (Smith and Johnson, 1887) that should take precedence. In conjunction with geologic mapping of the Middendorf Swift and Heron (1969), however, indicated that the term “Tus- and Patrick quadrangles, the USGS drilled a core in March 2014 caloosa Formation” has been used in the Carolinas to describe at Stop 6 (Fig. 15). This core (Patrick core, CTF-320) started at two lithologically distinct bodies of strata that are separated by a ground elevation of 112.8 m (370 ft) above sea level (ASL) an unconformity. The lower strata consist of numerous laterally and reached 85.3 m (280 ft) total depth. In this core, the strata persistent units of gravelly sand grading up to sand and muddy are predominantly sandstone to pebbly sandstone (sand, where sand exposed along the Cape Fear River and other areas near poorly lithifi ed), except at 38.1–26.8 m (125–87.9 ft) depth Fayetteville, North Carolina. They suggested that the term “Cape where clay is the predominant lithology. This thick interval of Fear Formation” be applied to these strata, following nomencla- clay can be seen in outcrop at a kaolinite mine off Bullard Ford ture proposed by Stephenson (1907, 1909, 1912). In contrast, Road (south of Little Beaverdam Branch) in the southeast por- the upper strata consist of heterogeneous units of pebbly sand, tion of the Middendorf quadrangle. Fining-upward sequences clean sand, muddy sand, and clay beds that are widespread in the and cross-bedding are visible in some sandstone units of the Carolinas. Swift and Heron suggested that the name Middendorf Patrick core. Other lithologies include 0.2–5.0-m- (0.7–16.4-ft-) Formation be applied to these strata. Indeed, many publications thick beds of clay to silty clay, <1.5-m- (4.9-ft-) thick units continued to use the term Middendorf Formation (e.g., Heron, of alternating laminations of sandstone (sand, where poorly 1958a, 1958b, 1959; Ridgeway et al., 1966; Swift and Heron, lithifi ed) and clay, and <1.0-m- (3.3-ft-) thick units of pebbly 1967, 1969; Howell and Zupan, 1974), and eventually the U.S. sandstone to gravel. An assemblage of coarse to very fi ne sand Geological Survey reinstated the stratigraphic term “Middendorf with disrupted bedding, mottled texture, and downward taper- Formation” in South Carolina and North Carolina (Cohee and ing vertical structures (interpreted as a paleosol) is present in 24 Swezey et al.

Figure 15. Diagram of the Patrick and Cheraw cores (Cretaceous Middendorf Formation), Chesterfi eld County, South Carolina. In this fi gure, both cores are hung relative to sea level. The distance between the two cores is ~25 km (15.5 mi). See Figure 2 for core locations. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 25

the Patrick core at 85.3–81.7 m (280.0–268.0 ft) depth or 27.5– pollen (Taxodiaceae and Pinuspollenites), angiosperm pollen 31.1 m (90.2–102.0 ft) ASL. (Momipites), and freshwater algae (Schizosporis). In modern A similar core was drilled by the USGS in July 1995 at environments, Taxodiaceae does not tolerate high salinity, and is Cheraw State Park, which is located ~25 km (15.5 mi) northeast typically found on intermittently fl ooded, poorly drained fl ood- of the Patrick core. This core (Cheraw core, CTF-081) started at plains (i.e., additional evidence suggesting a fl uvial rather than a ground elevation of 58.8 m (193 ft) ASL, reached the contact of estuarine or marine deposit). These palynomorphs were iden- the Middendorf Formation over Paleozoic schist (meta-siltstone) tifi ed by Christopher Bernhardt (USGS), and are reported in at 66.4 m (218 ft) depth, and reached a total depth of 74.4 m Swezey et al. (2015). (244 ft) in the schist. As with the Patrick core, the strata in the Drive from McBee to Hartsville (going over Orangeburg Scarp). Cheraw core are predominantly sandstone to pebbly sandstone. Spend the night in Hartsville. Fining-upward sequences and cross-bedding are visible in some sandstone units, and other lithologies include <1.5-m- (4.9-ft-) DAY 2 (3 April 2016) thick units of clay to silty clay, and <1.5-m- (4.9-ft-) thick units of Drive from Hartsville to McBee (going over Orangeburg Scarp) alternating sand and clay laminations. An assemblage of coarse and then to the Sand Hills State Forest near the town of Patrick. to very fi ne sand with disrupted bedding, mottled texture, and downward tapering vertical structures (interpreted as a paleosol) The second day of the fi eld trip begins with driving northwest is present in the Cheraw core at 38.1–36.3 m (125.0–119.1 ft) from Hartsville to McBee along Highway 151. Near the boundary depth or 20.7–22.5 m (67.9–73.8 ft) ASL. Pebbles of smoky between Darlington County and Chesterfi eld County, the road goes quartz and milky quartz up to 5 cm (2 in) diameter are present up the Orangeburg Scarp (Fig. 19), an east-facing scarp where the at the base of the Middendorf Formation in the Cheraw core (Fig. 16). The two cores do not reveal signifi cant changes in lithology such that separate stratigraphic terms might be warranted, and thus the term “Middendorf Formation” is applied to all of the cored intervals above the Paleozoic schist. However, grain size is generally coarser in the lower part of the section. For example, the predominant lithologies in the lower Middendorf Forma- tion (below the paleosols) are coarse sand and gravel, whereas the predominant lithologies in the upper Middendorf Formation (above the paleosols) are medium sand and clay. The Middendorf Formation is interpreted to be fl uvial sedi- ments derived from highlands to the west including the uplifted margins of Mesozoic rift basins and/or the Appalachian Moun- tains. A diagram of facies associated with a macrotidal estu- ary helps to portray this interpretation (Fig. 17). The sandy and gravelly units of the Middendorf Formation are interpreted as primarily channel deposits, whereas the thick clay units are inter- preted as fl oodplain deposits and abandoned channel deposits. In a sequence stratigraphic context, the Middendorf Formation is interpreted as fl uvial sediments that accumulated during base- level lowstand and subsequent early transgression. The initial base-level fall resulted in relatively coarse fl uvial sediments prograding into the basin across the unconformity on Paleo- zoic schist. As the system changed from lowstand conditions to early transgression, fl uvial sediment grain size became relatively fi ner and a greater proportion of fl oodplain muds was preserved (Fig. 18). The fl uvial interpretation of the Middendorf Formation is consistent with the leaf fossils identifi ed by Berry (1910, 1914) at the Middendorf type section, and with the pieces of petrifi ed Figure 16. Lowermost portion of the Cheraw core (CTF-081), show- wood found in nearby outcrops. Furthermore, silty clay at 65.5 m ing unconformity between Paleozoic schist (meta-siltstone) and over- lying Cretaceous Middendorf Formation. The numbers on the wooden (215.2 ft) depth in the Cheraw core has yielded Late Cretaceous blocks denote depth from surface in feet. The unconformity is placed palynomorphs including abundant fern spores (Cicatricosi- at 66.4 m (218 ft) depth, and multi-colored quartz pebbles are present sporites, Gleicheniidites, Foveotriletes), as well as gymnosperm on this unconformity. 26 Swezey et al.

Estuarine Inlet Fluvial Funnel Channel

Figure 17. Facies associated with a Inlet Sand Estuarine Mud Estuarine Estuarine Fluvial macrotidal estuary, such as the Gironde Tidal Bar Channel and Channel and estuary in southwestern France (Allen, and Tidal Flat Tidal Flat 1991). The Cretaceous Middendorf Sand & Mud Sand & Mud Sand, Gravel, & Mud Formation is analogous to the upstream (“fl uvial”) portion of the diagram.

Estuarine Tidal Bar Estuarine Channel Fluvial Channel and Tidal Flat and Tidal Flat and Floodplain

Tidal Flat: rippled Tidal Flat: rippled Floodplain: levee sand, flaser bedding, sand, flaser & crevasse splay fining up to mud bedding, & mud fine sand and mud Tidal Bar: upward- Channel: upward- Channel: thickening sets of thinning sets of conglomerate medium sand with medium sand with fining up to cross-bedding cross-bedding cross-bedded and ripples, and ripples, coarse to abundant clay clay laminations medium sand laminations

topography changes abruptly from ~60–80 m (197–262 ft) eleva- STOP 7: Drainage Divide on Road between Scotch Road and tion to 100–120 m (328–394 ft) elevation. This geomorphic fea- Sugarloaf Mountain within the Sand Hills State Forest in ture was named for a site at the City of Orangeburg in Orangeburg Middendorf Quadrangle (N 34° 35′ 08.10″, W 80° 08′ 10.44″) County of South Carolina (Colquhoun, 1962), and is interpreted as a shoreline terrace formed by wave erosion at a time of high sea Stop 7 is an overlook that shows the drainage divide on the level during the middle Pliocene (Dowsett and Cronin, 1990). road to Sugarloaf Mountain within the Sand Hills State Forest After driving northwest up the Orangeburg Scarp and onto in the Middendorf quadrangle (Fig. 20). This drainage divide the plateau of Cretaceous strata, there is a fi eld trip stop at a is a north-trending escarpment incised by headwater streams. major north-trending drainage divide, followed by a stop at The divide separates the predominantly south-fl owing drain- Sugarloaf Mountain to look at various facies of the Cretaceous ages associated with Big Black Creek from the predominantly Middendorf Formation, and then a stop along Isaac Road to east-fl owing drainages associated with Juniper Creek and similar look at an exceptional exposure of the Quaternary Pinehurst tributaries of the (Fig. 2). The recognition of this Formation. After lunch back at Sugarloaf Mountain, the next escarpment is an excellent example of how LiDAR data are trans- fi eld trip stop is on private property to look at terraces along forming our ability to resolve topographic features in the Coastal Juniper Creek and to examine contacts between these terraces Plain. On the ground, this drainage divide appears to be a rela- and sand of the Pinehurst Formation. tively subtle escarpment of low relief (Fig. 20), whereas in the Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 27

IV - Highstand Systems Tract (HST) HST

III - Late Transgressive Systems Tract (TST) TST

II - Early Transgressive Systems Tract (TST) LST I - Lowstand Systems Tract (LST)

As one goes from the Lowstand Systems Tract to the Early immature mature hydromorphic channel floodplain Transgressive Systems Tract, a greater proportion of flood- soil soil soil deposit deposit plain mud is preserved (cf., upper portion of Patrick core).

Model of fluvial sequence stratigraphy as outlined by Wright and Marriott (1993).

IV - Highstand Systems Tract (HST): increased density of channel sand bodies, but finer grain size than during LST; high rates of floodplain reworking; possible lacustrine deposits; better developed soils, but reduced preservation potential; reduced rate of creation of accumulation space. III - Late Transgressive Systems Tract (TST): greater proportion of floodplain mud; fewer and more isolated channel sand bodies; weakly developed soils; increased rate of creation of accumulation space.

II - Early Transgressive Systems Tract (TST): multi-story channel sand bodies; floodplain mud prone to reworking by channels or by stripping; rising base level may favor hydromorphic soils; low rate of creation of accumulation space. I - Lowstand Systems Tract (LST): Predominantly coarser-grained channel deposits; mature soils on terrace surfaces.

The Cretaceous Middendorf Formation is a Lowstand Systems Tract (I) and Early Transgressive Systems Tract (II). Sequence Stratigraphy Context/Chronology of the Cretaceous Middendorf Formation: (1) Fall in Sea Level = Fluvial systems prograde into basin over unconformity on Paleozoic schist; (2) Lower portion of fluvial strata = Relatively coarser grained point bar deposits interpreted as a Lowstand Systems Tract; (3) Upper portion of fluvial strata = Relatively finer grained (and contains more clay beds - floodplain deposits) interpreted as an Early Transgressive Systems Tract (TST).

Figure 18. Model of fl uvial sequence stratigraphy; modifi ed from Wright and Marriott (1993).

LiDAR imagery, this escarpment appears as a prominent feature through these fl uvial systems yet exists, and so it is unclear characterized by broad arcuate embayments carved by stream whether the divide asymmetry is infl uenced more by Quaternary head waters on the eastern side of the divide (Fig. 3). changes in climate and/or sea level, or whether longer-term fac- The origin of the marked asymmetry across this divide may tors are more important. stem from several factors, including Quaternary climate changes, Another reason for the drainage divide asymmetry could be erosion-resistant strata, and long-term migration of the drainage resistant beds within the Middendorf Formation. Morphologi- divide. The watersheds on both sides of this divide are tributar- cally, the embayments along the divide resemble those described ies to the Pee Dee/Yadkin drainage, a large fl uvial network that by Dunne (1990) that form by groundwater sapping and fi rst- drains part of the Blue Ridge and Piedmont provinces in the Car- order stream erosion below resistant cliff-forming strata. It is rea- olinas and then crosses the Coastal Plain province. During times sonable to ponder if the scarp at the drainage divide is supported of sea-level lowstand, episodes of headward erosion may have in part by iron-cemented sandstone (such as that seen at Sugar- occurred along tributaries such as Big Black Creek and Juniper loaf Mountain; Stop 8) or by weakly cemented sandstone (such Creek that join the Pee Dee River in the Coastal Plain province. as that seen at Stop 2). All evidence to date from mapping, how- As discussed in greater detail at Stop 10, the sequences of ter- ever, suggests that moderate or strong cementation only occurs at races along Juniper Creek indicate that a modest amount of such the scale of individual outcrops, with one exception being trace- incision occurred episodically during the Pleistocene. Whittecar able for a kilometer (Stop 4). No examples of such cementation et al. (2013) suggested that such incision events along streams over long distances are known from the slopes along the drainage could infl uence landscape stability along the drainage divide. divide. Furthermore, strong cementation of the Middendorf For- However, no regional-scale mapping and correlation of terraces mation is not present at depth in the Patrick core or Cheraw core. 28 Swezey et al.

the divide should migrate over geologic time. χ maps suggest that the Big Black Creek–Juniper Creek divide in the study area may be in a state of disequilibrium, albeit a relatively modest one when compared to divides between larger drainage basins in the Coastal Plain and Piedmont provinces of the Carolinas. If more detailed analyses of χ across this region confi rm the disequilibrium, then this confi rmation would support the observations that overall topography across most of the Patrick quadrangle is younger than the higher elevation topography within the Middendorf quad- rangle. The unusually thick, deeply weathered soils on the broad interfl uvial regions within the Middendorf quadrangle (see Stop 5) would also support this disequilibrium hypothesis.

STOP 8: Sugarloaf Mountain within the Sand Hills State ′ ″ Figure 19. The Orangeburg Scarp on Highway 151 near the border Forest in Patrick Quadrangle (N 34° 35 15.24 , between Darlington County and Chesterfi eld County, South Carolina. W 80° 07′ 04.73″) View is to the northwest. Stop 8 is located at Sugarloaf Mountain, a prominent out- crop of the Cretaceous Middendorf Formation in the Patrick quadrangle (Figs. 21 and 22). This outcrop has been known for The most plausible explanation for the drainage divide asym- a long time, and a measured section of the strata here was pub- metry is long-term migration of the divide caused by different lished by Sloan (1904). slopes of the entire watershed. Streams fl owing west through the Near the base of Sugarloaf Mountain, beds or lenses of clay Big Black Creek system travel ~86 km (53 mi) to join the Pee form a distinct facies of the Middendorf Formation. Such clay Dee River at 14 m (46 ft) elevation (0.16 m/km or 0.87 ft/mi), beds are more common in the upper Middendorf Formation, and whereas Juniper Creek tributaries travel ~28 km (17 mi) to join similar clay beds or lenses were seen at the pit at Stop 2 and in the Pee Dee River at 26 m (85 ft) elevation (0.92 m/km or 5 ft/ the Patrick core (CTF-320) at Stop 6. These clay lenses and beds mi). The steeper watershed gradient of the Juniper Creek tributar- consist predominantly of kaolinite. Although some kaolinite in ies allows them to be more erosive and to drive the divide west- the Carolina Sandhills is likely to be of diagenetic origin (e.g., ward over time. Willett et al. (2014) developed a parameter, χ, kaolin matrix around quartz sand grains), most beds and lenses of that allows a determination of the degree of disequilibrium that kaolinite are interpreted as fl oodplain deposits or the fi nal stage exists between streams across divides, and the direction in which of sediment accumulation in an abandoned channel. It is possible

Figure 20. East-facing arcuate embay- ment along north-trending drainage divide at Stop 7. View is to the south. The white line traces the edge of the escarpment, and a change from rela- tively higher ground to topographically lower ground on the escarpment face. This escarpment is very diffi cult to por- tray in a photograph, and it may be help- ful to refer to the location of Stop 7 and the east-facing escarpment in Figure 3. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 29

Figure 21. Block of iron-cemented sandstone that has come down from Sugarloaf Mountain at Stop 8. Sugar- loaf Mountain is located on the left side of the photograph. Vehicle provides a sense of scale.

that these beds and lenses were composed initially of a different this site shows a buried Bt horizon of a paleosol in soil profi les clay mineral, and have subsequently been altered to kaolinite. of the Candor association (Morton, 1995) within the Pinehurst At Stop 8, the most prominent facies of the Middendorf Formation. This stop provides an opportunity to discuss these Formation is that of iron-cemented cross-bedded sandstone to pedogenic features, the signifi cance of the OSL ages obtained, pebbly sandstone. Blocks of this facies form the top of Sug- and the potential use of soil maps in the development of geo- arloaf Mountain, and other blocks are scattered around the logic maps. For this stop, the color nomenclature used is from base of the mountain where they appear to have rolled down the Munsell Soil Color Chart (Munsell Color, 2000). from above (Fig. 21). This facies is found at several locations The Quaternary Pinehurst Formation is surfi cial sand with throughout the Middendorf and Patrick quadrangles, but has not weak to moderate weathering (i.e., soil lamellae to brown Bt been identifi ed in the subsurface. Furthermore, the iron cemen- horizons) that drapes over most of the uplands and terraces in tation is restricted to relatively coarse-grained strata. The cross- the Middendorf and Patrick quadrangles. The thickness of this bedded sandstone to pebbly sandstone is interpreted as fl uvial sand unit ranges from tens of centimeters to several meters. How- sediments (primarily channel deposits, as depicted in Fig. 17), ever, on the recent geologic map of the Patrick quadrangle (to be and the iron is attributed to some form of diagenetic shallow published by the South Carolina Geological Survey), the Pine- groundwater circulation. After the sandstone is cemented by hurst Formation is portrayed only where the sand thickness is iron, the sandstone is more resistant to erosion and forms topo- greater than 2 m (6.6. ft). Although it could be a considerable graphic highs across the landscape. challenge to defi ne an approximate map boundary at 2 m (6.6 ft) sand thickness, the defi nitions of two of the soil series mapped in STOP 9: Isaac Road Outcrop of Pinehurst Formation in this county by Morton (1995)—the Alpin sand and the Condor Patrick Quadrangle (N 34° 37′ 17.67″, W 80° 03′ 34.35″) sand—meet this geologic map criteria (Table 1). Other soils in the area also have notable sand mantles (Table 1), but the sand Stop 9 is located on the east side of Isaac Road, and one thickness of these other soils is generally <2 m (6.6 ft). For exam- wonders if the name of this road has a specifi c association with ple, surfi cial sand in the Troupe soil is up to 2 m (6.6. ft) thick, the longleaf pine. When the sap of the longleaf pine was distilled, and surfi cial sand in the Ailey sand and Vaucluse loamy sand is the resulting product was called rosin. There were 12 different 0.5–1 m (1.6–3.3 ft) thick (Morton, 1995). grades of rosin, and “Isaac” was the name of a particular grade Buried surfaces and soil horizons within their characteristic (Earley, 2004). profi les are present in most of the soil series of the Carolina Sand- Stop 9 is an exceptional outcrop of the Quaternary Pine- hills. The Troupe, Ailey, and Vaucluse soil series each contain hurst Formation (Fig. 23). As with most outcrops of this unit, subsoil horizons with reddish colors (e.g., 2.5YR) that are usu- primary sedimentary structures are not visible but there is evi- ally found in deeply weathered deposits of the Cretaceous Mid- dence of bioturbation by vegetation, as well as pedogenic fea- dendorf Formation. When exposed in outcrop, these deep hori- tures including soil lamellae and argillic horizons. Specifi cally, zons often contain relict redoximorphic features and structures 30 Swezey et al.

Figure 22. Measured section by Sloan (1904, p. 107–108) at Sugarloaf Mountain. Sloan (1904) also mentioned that 6.1 m (20 ft) of “drab clay” was exposed in a pit at the base of Sugarloaf Mountain. The top of Sugarloaf Mountain is 156.4 m (512 ft) above sea level, and thus this section is stratigraphically above all of the units in the Patrick core. that could indicate the presence of a truncated paleosol. The The brown colors in these horizons (e.g., 10YR) suggest that they Alpin sand has no buried soils within the uppermost 2 m (6.6 m), developed during the Quaternary. but the Candor sand contains a brown Bt horizon on the top of a Other soil types occur in only a very few places (<10%) buried sand bed. Overlying this argillic horizon is a full A-E-Bt-C on the uplands (Morton, 1995). The Noboco loamy sand (oxy- profi le that has formed in the surfi cial sand sheet or sand. aquic paleudult), for example, with its strongly oxidized nodules Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 31

and reduced masses in its upper Bt horizon, appears to be a soil formed where the Middendorf sediments now have no overlying Quaternary sediments. Terrace deposits (discussed later at Stop 10) have their own suite of soil series that formed on sand and mud deposited by streams. The following three weighted mean OSL ages were obtained from the Pinehurst Formation at this site: 67.1 ± 5.5 ka, 56.1 ± 4.9 ka, and 61.3 ± 2.8 ka. These three ages were determined in L ages. 2014 by Shannon A. Mahan (USGS). As mentioned at Stop 3, most of the OSL ages from the Pinehurst Formation range from ca. 73–19 ka, coincident with glacial conditions in the North- ern Hemisphere. It is important to remember, however, that the individual OSL ages indicate only the last time that a particular sample was exposed to sunlight, and not the total time of sedi- ment mobilization. No OSL ages from the Carolina Sandhills are younger than 8 ka, and thus it appears that since this date, the eolian sand sheets and dunes have remained stabilized by vegeta- tion and subjected to pedogenic processes. In a broader context, the interpretation of the Carolina Sand- hills as eolian sand sheets and dunes is consistent with other evi- dence for eolian activity in the southeastern United States during the last glaciation and early deglaciation. Specifi cally, parabolic eolian dunes in river valleys of the Coastal Plain from Georgia through Delaware were active ca. 40–19 ka (Markewich and Markewich, 1994; Ivester et al., 2001; Swezey et al., 2013). Addi- tional evidence of eolian activity during this time comes from Carolina Bays, which are oval depressions on the Coastal Plain interpreted as having formed by eolian defl ation ca. 44–10 ka (Ivester et al., 2002, 2003; Moore et al., 2014). LUNCH: Picnic area at Sugarloaf Mountain (N 34° 35′ 15.24″, W 80° 07′ 04.73″).

STOP 10: Tommy Wilkes Farm in Patrick Quadrangle (N 34° 34′ 46.49″, W 80° 01′ 32.35″) This is private property. Please obtain consent from the owner before proceeding onto the property.

Stop 10 is located on one of the larger terrace complexes in the study area and provides a good location to discuss the relations between Quaternary terrace deposits, the Pinehurst Formation, and pedogenic features. For this stop, the color nomenclature used is from the Munsell Soil Color Chart (Mun- sell Color, 2000). Throughout the Patrick quadrangle, the following three groups of terraces exist along streams and creeks: (1) low-level terraces that are 0–4.6 m (0–15 ft) above creek level (ACL), for most creeks within the Middendorf and Patrick quadrangles;

Figure 23. Stop 9 along the east side of Isaac Road, showing outcrop of the Quaternary Pinehurst Formation and weighted mean OS outcrop of the Quaternary Pinehurst Formation Figure 23. Stop 9 along the east side of Isaac Road, showing (2) mid-level terraces that are 4.6–9.1 m (15–30 ft) ACL; and (3) high-level terraces that are >9.1 m (30 ft) ACL. The fact that there are three groups of terraces suggests that there have been at least three episodes of geomorphologic disturbance that have led to terrace formation. The terrace deposits are <3 m (9.8 ft) thick, and are composed of moderately sorted to poorly sorted pebbles, pebbly sand, sand, 32 Swezey et al.

TABLE 1. SOIL SERIES CHARACTERISTICS AND INTERPRETATION OF IMPORTANT UPLAND SANDY SOILS SERIES IN THE MIDDENDORF AND PATRICK QUADRANGLES, CHESTERFIELD COUNTY, SOUTH CAROLINA Soil Series: Alpin sand* Candor sand* Troupe sand Ailey sand Vaucluse sand

Sand (Q) >2 m thick

Sand (Q) 1–2 m thick

Sand (Q) 0–1 m thick ?

Buried brown (Q) paleosol

Buried, eroded red (K) weathered sediment Typic quartzi- Arenic paleudult Grossarenic Arenic Typic Soil classification psamment kandiudult kanhapludult kanhapludult Data from Morton (1995). *Denotes the soil series used to delineate the boundaries of the Quaternary Pinehurst Formation.

silt, and/or clay. At many sites, these terrace sediments are arranged the base of the C horizon at a depth of 1.9 m (6.2 ft) for OSL in a general fi ning-upward sequence. The pebbles range up to dating, and yielded a weighted mean age of 12.1 ± 2.0 ka. An 1.9 cm (0.7 in.) diameter, and are composed of smoky quartz, milky abrupt boundary separated the C horizon from the underlying quartz, and clear quartz. Some clasts of light-gray clay up to 2 cm 0.8-m- (2.6-ft-) thick unit of brownish-yellow (10YR 6/8) loamy

(0.8 in.) diameter are also present. The sand, which ranges in size sand. This loamy sand, which was interpreted as a Bb horizon, from fi ne to coarse, is composed of quartz with minor amounts of was sampled at a depth of 2.0 m (6.6 ft) for OSL dating, and mica. Terrace sediments that are closer to the modern fl oodplains yielded a weighted mean age of 29.2 ± 4.9 ka. A wavy boundary are typically brown and yellow, whereas higher terrace sediments separated the Bb horizon from an underlying sandy loam that contain reddish-yellow and brown colors. contained mottling. This sandy loam was exposed in the bottom At Stop 10 (Wilkes farm site), a backhoe was used to dig two 0.3 m (1.9 ft) of the trench. trenches into sandy hillslope deposits identifi ed through GPR Geologic mapping throughout the Patrick quadrangle has and borehole data (Fig. 24). The trenches were ~18.3 m (60 ft) shown that sand of the Quaternary Pinehurst Formation is pres- apart, and located at ~34° 34′ 44.2″ N, 80° 01′ 43.2 ″ W at ~61 m ent on the mid-level terraces and the upper-level terraces, but not (200 ft) elevation. OSL ages at this site were determined in 2015 on the low-level terraces (i.e., terraces that are <4.6 m [15.1 ft] by George A. Brook (University of Georgia). ACL) or on the modern fl oodplains. This observation places The fi rst trench (“Trench 1” in Fig. 24) was 2.5 m (8.2 ft) some constraints on the ages of the terraces. The terrace deposit deep and contained only sand of the Quaternary Pinehurst For- in the second trench, for example, was a “high-level” terrace at mation. The soil profi le in this trench consisted of a thin O hori- an elevation of 15.2 m (50 ft) ACL. This terrace was intersected zon, a 0.2-m- (0.7-ft-) thick brown (10YR 5/3) sandy Ap horizon, at a depth of 2.6 m (8.5 ft) in an auger hole located 3.5 m (11 ft) a 0.3-m- (1.0-ft-) thick dark yellowish-orange (10YR 6/6) sandy west-northwest of the second trench. As stated previously, these E horizon, and a 1.0-m- (3.3-ft-) thick brown (7.5YR 5/8) loamy high-level terraces throughout the Patrick quadrangle are over- sand B horizon. The B horizon was separated from the under- lain by sand of the Pinehurst Formation, and thus the terraces lying C horizon by a gradual wavy boundary with visible root are older than the Pinehurst Formation. In this particular trench, structures. The underlying 1-m- (3.3-ft-) thick sandy C horizon the Bb horizon (loamy sand) in the Pinehurst Formation yielded had a color of brownish yellow (10YR 6/6). One OSL sample a weighted mean OSL age of 29.2 ± 4.9 ka, which is similar to was taken at a depth of 2.5 m (8.2 ft) within the C horizon, and a weighted mean OSL age of 33.8 ± 3.4 ka obtained from the yielded a weighted mean age of 33.8 ± 3.4 ka. C horizon (sand) in the Pinehurst Formation of the fi rst trench. The second trench (“Trench 2” in Fig. 24) was 3 m (9.8 ft) It is thought that the top of the B horizon in the second trench deep and contained sand of the Quaternary Pinehurst Forma- (Fig. 24) is the paleosol depicted in Figure 25. tion (Fig. 25). The soil profi le in this trench consisted of a thin O horizon as well as a 0.4-m- (1.3-ft-) thick brown (10YR 5/3) End of fi eld trip. Begin drive back to Columbia. sandy Ap horizon. Directly below the Ap horizon from 0.4 to 1.9 m (1.3–6.2 ft) depth, there was a white (10YR 8/1) and yel- ACKNOWLEDGMENTS low (10YR 8/6) sandy C horizon. Within the C horizon and start- ing at a depth of 0.8 m (2.6 ft) below the surface, there were The authors extend sincere thanks to Bill Clendenin, Scott How- multiple brownish-yellow (10YR 6/8) soil lamellae that were up ard, and Will Doar of the South Carolina Geological Survey to 1 cm (0.4 in) thick in places. This material was sampled at for much cooperation, support, and encouragement. We also Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 33

Figure 24. GPR transect and interpretation of stratigraphy at Stop 10. Location of transect is shown in Figure 3. Paleosol within Pinehurst Formation is the top of the B horizon in the second trench (Fig. 25), and denotes at least one episode of landscape stabilization during accumulation of the Pinehurst Formation.

sand (A horizon)

sand (C horizon)

12.1 ± 2.1 ka

29.2 ± 4.9 ka loamy sand (buried B horizon) sandy loam with mottling

Figure 25. “Trench 2” at Stop 10, showing pedogenic horizons within the Quaternary Pinehurst Formation and weighted mean OSL ages. Blue and red tape on hoe mark increments of 0.1 m (0.3 ft). 34 Swezey et al. thank Allyne Askins and Nancy Jordan of the Carolina Sand- Earley, L.S., 2004, Looking for Longleaf: The Fall and Rise of an American Forest: hills National Wildlife Refuge and Brian Davis of the Sand Hills Chapel Hill, North Carolina, University of North Carolina Press, 336 p. Eze, P.N., Udeigwe, T.K., and Meadows, M.E., 2014, Plinthite and its associ- State Forest for property access, logistical help, and additional ated evolutionary forms in soils and landscapes: A review: Pedosphere, encouragement. We are grateful for additional property access v. 24, p. 153–166, doi:10.1016/S1002-0160(14)60002-3. granted to us by Tommy Wilkes, Kemp McLeod, and Fraser Finch, B., Young, B.M., Johnson, R., and Hall, J.C., 2012, Longleaf, Far as the Eye Can See: Chapel Hill, North Carolina, University of North Carolina Almond. Brad Fitzwater was funded by a USGS EdMap grant, Press, 176 p. and he thanks Rick Goshen, Steve Stone, Bryte Shoup, Katy Fitzwater, B.A., Swezey, C.S., Whittecar, R.G., Garrity, C.P., and Mahan, S.A., Ames, and Ben Hiza for assistance in the fi eld. The Patrick core 2014a, Geologic mapping of the Middendorf and Patrick quadrangles, Chesterfi eld County, South Carolina: Geological Society of America was obtained by USGS drillers Gene Cobbs, Jeff Gray, and Tony Abstracts with Programs, v. 46, no. 3, p. 82. Burkart, and Joe Gellici ran geophysical logs in the core hole. Fitzwater, B.A., Whittecar, R.G., Dobbs, K.M., Swezey, C.S., Garrity, C.P., OSL dating was conducted by Shannon Mahan of the USGS and and Mahan, S.A., 2014b, Ground-penetrating radar (GPR) and optically stimulated luminescence (OSL) data from Quaternary sands in the Caro- by George Brook of the University of Georgia. Kerby Dobbs of lina Sandhills, Chesterfi eld County, South Carolina: Geological Society Maser Consulting obtained the GPR data. Frank Dulong of the of America Abstracts with Programs, v. 46, no. 6, p. 306. USGS provided assistance in the identifi cation of clay minerals. Folk, R.L., 1951, Stages of textural maturity in sedimentary rocks: Journal of Sedimentary Petrology, v. 21, p. 127–130, doi:10.2110/jsr.21.127. This manuscript benefi tted from reviews by USGS geologists Folk, R.L., 1954, The distinction between grain size and mineral composition Mark Carter and Lucy Edwards. Any use of trade, fi rm, or prod- in sedimentary rock nomenclature: The Journal of Geology, v. 62, p. 344– uct names is for descriptive purposes only and does not imply 359, doi:10.1086/626171. Folk, R.L., 1966, A review of grain size parameters: Sedimentology, v. 6, endorsement by the U.S. Government. p. 73–93, doi:10.1111/j.1365-3091.1966.tb01572.x. Folk, R.L., 1980, Petrology of Sedimentary Rocks: Austin, Texas, Hemphill REFERENCES CITED Publishing Company, 170 p. Folk, R.L., and Ward, W.C., 1957, Brazos River point bar: A study in the signifi - cance of grain size parameters: Journal of Sedimentary Petrology, v. 27, Allen, G.P., 1991, Sedimentary processes and facies in the Gironde estuary: A p. 3–26, doi:10.1306/74D70646-2B21-11D7-8648000102C1865D. recent model for macrotidal estuarine systems, in Smith, D.G., Reinson, Goddard, E.N., Trask, P.D., De Ford, R.K., Rove, O.N., Singewald, J.T., Jr., and G.E., Zaitlin, B.A., and Rahmani, R.A., eds., Clastic Tidal Sedimentol- Overbeck, R.M., 1963, Rock-Color Chart: New York, Geological Society ogy: Canadian Society of Petroleum Geologists Memoir 16, p. 29–40. of America. Askins, A.H., 2010, Carolina Sandhills National Wildlife Refuge Comprehen- Griffi th, G.E., Omernik, J.M., Comstock, J.A., Lawrence, S., Martin, G., God- sive Conservation Plan: U.S. Department of the Interior Fish and Wildlife dard, A., Hulcher, V.J., and Foster, T., 2001, Ecoregions of Alabama and Service–Southeast Region, 234 p. Georgia: U.S. Geological Survey, 1:1,700,000 scale map, 1 sheet. Bartlett, C.S., Jr., 1967, Geology of the Southern Pines quadrangle, North Caro- Griffi th, G.E., Omernik, J.M., Comstock, J.A., Schafale, M.P., McNab, W.H., lina [M.S. thesis]: Chapel Hill, North Carolina, University of North Caro- Lenat, D.R., MacPherson, T.F., Glover, J.B., and Shelburne, V.B., 2002, lina at Chapel Hill, 101 p. Ecoregions of North Carolina and South Carolina: U.S. Geological Sur- Berry, E.W., 1910, A Cretaceous Lycopodium: American Journal of Science, vey, 1:1,500,000 scale map, 1 sheet. ser. 4, v. 30, p. 275–276, doi:10.2475/ajs.s4-30.178.275. Heron, S.D., Jr., 1958a, Clay minerals of the outcropping basal Cretaceous Berry, E.W., 1914, The Upper Cretaceous and Eocene Floras of South Carolina beds between the Cape Fear River, North Carolina, and Lynches River, and Georgia: U.S. Geological Survey Professional Paper 84, 200 p. South Carolina: Clays and Clay Minerals, v. 7, p. 148–161, doi:10.1346/ Cady, J.G., and Daniels, R.B., 1968, Genesis of some very old soils—The CCMN.1958.0070108. Paleudults: 9th International Congress of Soil Science: Transactions, v. 4: Heron, S.D., Jr., 1958b, The stratigraphy of the outcropping basal Cretaceous Sydney, International Society of Soil Science, p. 103–112. formations between the Neuse River, North Carolina, and Lynches River, Christensen, N.L., 2000, Vegetation of the southeastern Coastal Plain, in Bar- South Carolina [Ph.D. diss.]: Chapel Hill, North Carolina, University of bour, M.G., and Billings, W.D., eds., North American Terrestrial Vegeta- North Carolina at Chapel Hill, 155 p. tion (2nd ed.): Cambridge, UK, Cambridge University Press, p. 397–448. Heron, S.D., Jr., 1959, A small basement cored anticlinal warp in the basal Cre- Cohee, G.V., and Wright, W.B., 1976, Changes in Stratigraphic Nomenclature taceous sediments near Cheraw, South Carolina [South Carolina Division by the U.S. Geological Survey, 1975: U.S. Geological Survey Bulle- of Geology]: Geologic Notes, v. 3, no. 4, p. 1–4. tin 1422-A, 84 p. Holmes, J.A., 1893, Geology of the sand-hill country of the Carolinas: Geologi- Colquhoun, D.J., 1962, On surfi cial sediments in central South Carolina—A cal Society of America Bulletin, v. 5, p. 33–35. progress report [South Carolina Division of Geology]: Geologic Notes, Howell, D.E., and Zupan, A.-J.W., 1974, Evidence for post-Cretaceous tec- v. 6, no. 6, p. 63–80. tonic activity in the Westfi eld Creek area north of Cheraw, South Caro- Conley, J.F., 1962, Geology and Mineral Resources of Moore County, North lina [South Carolina Division of Geology]: Geologic Notes, v. 18, no. 4, Carolina: North Carolina Division of Mineral Resources Bulletin 76, 40 p. p. 98–105. Cooke, C.W., 1926, Correlation of the basal Cretaceous beds of the southeast- Ivester, A.H., Leigh, D.S., and Godfrey-Smith, D.I., 2001, Chronology of ern United States, in Shorter Contributions to General Geology: U.S. inland eolian dunes on the coastal plain of Georgia: Quaternary Research, Geological Survey Professional Paper 140, p. 137–139. v. 55, p. 293–302, doi:10.1006/qres.2001.2230. Cooke, C.W., 1936, Geology of the Coastal Plain of South Carolina: U.S. Geo- Ivester, A.H., Godfrey-Smith, D.I., Brooks, M.J., and Taylor, B.E., 2002, Caro- logical Survey Bulletin 867, 196 p. lina Bays and inland dunes of the southern Atlantic Coastal Plain yield Darton, N.H., 1896, Artesian Well Prospects in the Atlantic Coastal Plain new evidence for regional paleoclimate: Geological Society of America Region: U.S. Geological Survey Bulletin 138, 232 p. Abstracts with Programs, v. 34, no. 6, p. 273–274. Dowsett, H.J., and Cronin, T.M., 1990, High eustatic sea level during the mid- Ivester, A.H., Godfrey-Smith, D.I., Brooks, M.J., and Taylor, B.E., 2003, Con- dle Pliocene: Evidence from the southeastern U.S. Atlantic Coastal Plain: centric sand rims document the evolution of a Carolina Bay in the middle Geology, v. 18, p. 435–438, doi:10.1130/0091-7613(1990)018<0435 coastal plain of South Carolina: Geological Society of America Abstracts :HESLDT>2.3.CO;2. with Programs, v. 35, no. 6, p. 169. Dunne, T., 1990, Hydrology, mechanics, and geomorphic implications of ero- Johnson, H.S., Jr., 1961, Line stratigraphy northeast of Columbia, S.C. [South sion by subsurface fl ow, in Higgins, C.G., and Coates, D.R., eds., Ground- Carolina Division of Geology]: Geologic Notes, v. 5, no. 5, p. 81–87. water Geomorphology; The Role of Subsurface Water in Earth-Surface Kite, L.E., 1987, Cretaceous and Tertiary stratigraphy of the Gilbert 15-minute Processes and Landforms: Geological Society of America Special Paper quadrangle, Lexington and Aiken Counties, South Carolina: South Caro- 252, p. 1–28, doi:10.1130/SPE252-p1. lina Geology, v. 31, no. 1, p. 17–27. Geology and geomorphology of the Carolina Sandhills, Chesterfi eld County, South Carolina 35

LeGrand, H.E., 1961, Summary of geology of Atlantic Coastal Plain: American America–Soil Science Society of America (ASA-CSSA-SSSA) Inter- Association of Petroleum Geologists Bulletin, v. 45, p. 1557–1571. national Annual Meeting (Oct. 16–19, 2011; San Antonio, Texas); Leigh, D.S., 1998, Evaluating artifact burial by eolian versus bioturbation pro- online at https://scisoc.confex.com/scisoc/2011am/webprogram/ cesses, South Carolina Sandhills, USA: Geoarchaeology, v. 13, p. 309– Paper67175.html. 330, doi:10.1002/(SICI)1520-6548(199802)13:3<309::AID-GEA4 Sloan, E., 1904, A Preliminary Report on the Clays of South Carolina: South >3.0.CO;2-8. Carolina Geological Survey, Series 4, Bulletin 1, 171 p. Markewich, H.W., and Markewich, W., 1994, An Overview of Pleistocene and Sloan, E., 1907, Geology and mineral resources, in Watson, E.J., commissioner, Holocene Inland Dunes in Georgia and the Carolinas—Morphology, Dis- Handbook of South Carolina; Resources, Institutions and Industries of the tribution, Age, and Paleoclimate: U.S. Geological Survey Bulletin 2069, State: South Carolina Department of Agriculture, Commerce, and Immi- 32 p. gration, p. 77–145. McGee, W.J., 1886, Geologic formations: Report of the Health Offi cer of the Smith, E.A., and Johnson, L.C., 1887, Tertiary and Cretaceous Strata of the District of Columbia for the Year Ending June 30, 1885: Washington, Tuscaloosa, Tombigbee, and Alabama Rivers: U.S. Geological Survey D.C., Government Printing Offi ce, p. 19–20, 23–25. Bulletin 43, 189 p. McGee, W.J., 1890, The southern extension of the Appomattox Formation: The Smith, R.C., 2007, Hydromorphology and plinthite characterization of some American Journal of Science, 3rd series, v. 40, p. 15–41. Alabama Coastal Plain soils [Ph.D. diss.]: Auburn, Alabama, Auburn Uni- McGee, W.J., 1891, The Lafayette Formation: 12th Annual Report of the United versity, 159 p. States Geological Survey to the Secretary of the Interior 1890–’91: Wash- Soil Survey Staff, 2006, Keys to Soil Taxonomy (10th ed.): U.S. Depart- ington, D.C., U.S. Geological Survey, p. 347–521. ment of Agriculture, Natural Resources Conservation Service, 332 p.; McLeod, J., Whittecar, G.R., and Dobbs, K.M., 2012, Hydrogeologic condi- online at www.nrcs.usda.gov/Internet/FSE_DOCUMENTS/nrcs142p2 tions that affect restored pitcher plant bogs, southeastern Virginia: Geo- _052172.pdf. logical Society of America Abstracts with Programs, v. 44, no. 4, p. 59. Sorrie, B.A., 2011, A Field Guide to Wildfl owers of the Sandhills Region: Moore, C.R., Brooks, M.J., Mallinson, D.J., Parham, P.R., Ivester, A.H., and North Carolina, South Carolina, Georgia: Chapel Hill, University of Feathers, J.K., 2014, Rapid scour, sand rim construction, and basin migra- North Carolina Press, 378 p. tion of a Carolina Bay in southeastern North Carolina: Geological Society Stephenson, L.W., 1907, Some facts relating to the Mesozoic deposits of the of America Abstracts with Programs, v. 46, no. 3, p. 96–97. Coastal Plain of North Carolina: The Johns Hopkins University Circular Morton, R., 1995, Soil Survey of Chesterfi eld County, South Carolina: Wash- no. 7, p. 93–99. ington, D.C., U.S. Soil Conservation Service, 206 p. Stephenson, L.W., 1909, Cretaceous geology of the Carolinas and Georgia: Sci- Munsell Color, 2000, Munsell Soil Color Charts: New Windsor, New York, 10 p. ence, v. 30, no. 760, p. 124–125. Nystrom, P.G., Jr., and Kite, L.E., 1988, The Pinehurst Formation in South Stephenson, L.W., 1912, The Cretaceous formations, in Clark, W.B., Miller, Carolina: Geological Society of America Abstracts with Programs, v. 20, B.L., Stephenson, L.W., Johnson, B.L., and Parker, H.N., eds., The no. 4, p. 307–308. Coastal Plain of North Carolina: Raleigh, North Carolina, E.M. Uzzell Nystrom, P.G., Jr., Willoughby, R.H., and Price, L.K., 1991, Cretaceous and & Co., North Carolina Geological and Economic Survey Report 3, Tertiary stratigraphy of the Upper Coastal Plain, South Carolina, in Hor- p. 73–171. ton, J.W., Jr., and Zullo, V.A., eds., The Geology of the Carolinas: Knox- Steven, D., and Toner, M., 2004, Vegetation of upper Coastal Plain depres- ville, University of Tennessee Press, p. 221–240. sion wetlands: Environmental templates and wetland dynamics within Osher, L.J., and Buol, S.W., 1998, Relationship of soil properties to parent a landscape framework: Wetlands, v. 24, p. 23–42, doi:10.1672/0277 material and landscape position in eastern Madre de Dios, Peru: Geo- -5212(2004)024[0023:VOUCPD]2.0.CO;2. derma, v. 83, p. 143–166, doi:10.1016/S0016-7061(97)00133-X. Swezey, C.S., Schultz, A.P., Alemán González, W., Bernhardt, C.E., Doar, Otwell, W.L., Johnson, H.S., Jr., and Heron, S.D., Jr., 1966, The D. T. Duncan W.R., III, Garrity, C.P., Mahan, S.A., and McGeehin, J.P., 2013, Qua- kaolin property, Richland and Kershaw Counties, South Carolina [South ternary eolian dunes in the valley of the Savannah River, Jasper County, Carolina Division of Geology]: Geologic Notes, v. 10, no. 1, p. 9–15. South Carolina: Quaternary Research, v. 80, p. 250–264, doi:10.1016/j Owens, J.P., 1989, Geologic Map of the Cape Fear Region, Florence 1° × 2° .yqres.2013.06.007. Quadrangle and Northern Half of the Georgetown 1° × 2° Quadrangle, Swezey, C.S., Fitzwater, B.A., Bernhardt, C.E., and Whittecar, G.R., 2015, North Carolina and South Carolina: U.S. Geological Survey Miscella- Cretaceous stratigraphy in the updip portion of the U.S. Atlantic Coastal neous Investigations Series Map I-1948-A, 2 sheets. Plain, Chesterfi eld County, South Carolina: Geological Society of Amer- Poag, C.W., and Sevon, W.D., 1989, A record of Appalachian denudation ica Abstracts with Programs, v. 47, no. 7, p. 795. in postrift Mesozoic and Cenozoic sedimentary deposits of the U.S. Swezey, C.S., Fitzwater, B.A., Whittecar, G.R., Garrity, C.P., Mahon, S.A., and middle Atlantic continental margin: Geomorphology, v. 2, p. 119–157, Dobbs, K.M., 2016, The Carolina Sandhills of the southeastern United doi:10.1016/0169-555X(89)90009-3. States are eolian sand sheets and dunes that were active during the last Powars, D.S., Catchings, R.D., Horton, J.W., Jr., Schindler, J.S., and Pavich, glaciation: Geological Society of America Abstracts with Programs, v. 48, M.J., 2015, Stafford fault system: 120 million year movement history of no. 3, session 23. northern Virginia, in Horton, J.W., Jr., Chapman, M.C., and Green, R.A., Swift, D.J.P., and Heron, S.D., Jr., 1967, Tidal deposits in the Cretaceous eds., The 2011 Mineral, Virginia, Earthquake, and Its Signifi cance for of the Carolina Coastal Plain: Sedimentary Geology, v. 1, p. 259–282, Seismic Hazards in Eastern North America: Geological Society of Amer- doi:10.1016/0037-0738(67)90064-4. ica Special Paper 509, p. 407–431, doi:10.1130/2015.2509(23). Swift, D.J.P., and Heron, S.D., Jr., 1969, Stratigraphy of the Carolina Creta- Powers, M.C., 1953, A new roundness scale for sedimentary particles: Jour- ceous: Southeastern Geology, v. 10, p. 201–245. nal of Sedimentary Petrology, v. 23, p. 117–119, doi:10.1306/D4269567 Ward, L.W., 1998, Mollusks from the lower Miocene Pollack Farm site, Kent -2B26-11D7-8648000102C1865D. County, Delaware: A preliminary analysis, in Benson, R.N., ed., Geology Prowell, D.C., Christopher, R.A., Waters, K.E., and Nix, S.K., 2003, The chrno- and Paleontology of the Lower Miocene Pollack Farm Fossil Site, Dela- [sic] and lithostratigraphic signifi cance of the type section of the Mid- ware: Delaware Geological Survey Special Publication 21, p. 59–131. dendorf Formation, Chesterfi eld County, South Carolina: Southeastern Weems, R.E., and Edwards, L.E., 2007, Post-middle Miocene origin of modern Geology, v. 42, p. 47–66. landforms in the eastern Piedmont of Virginia: Stratigraphy, v. 4, p. 35–48. Ridgeway, D.C., McCauley, J.F., and Colquhoun, D.J., 1966, Geology of the Wentworth, C.K., 1922, A scale of grade and class terms for clastic sediments: Blaney Quadrangle, South Carolina: South Carolina Geological Survey The Journal of Geology, v. 30, p. 377–392, doi:10.1086/622910. Map Series MS-11, scale 1:24,000, 1 sheet. Whittecar, G.R., Emry, S.R., Daniels, W.L., and Hodges, R.L., 1993, Bed- Riekerk, H., 1983, Pine Tree Evapotranspiration: University of Florida Water rock weathering, soil development, and geomorphic evolution of the Old Resources Research Center Publication no. 62, 37 p. Hickory district, Dinwiddie County, Virginia: Virginia Journal of Schaetzl, R.J., and Thompson, M.L., 2015, Soils: Geomorphology and Genesis Science, v. 44, p. 157. (2nd ed.): New York, Cambridge University Press, 778 p. Whittecar, G.R., Pontier, N.K., Corrigan, P.B., Herman, S.W., Henderson, P.N., Shaw, J.N., Kelley, J.A., Owen, J., Rodekohr, D., Smith, R., and Stone, H., Gravette, J.V., and Luchetti, P.A., 1994, Soils and sediments in high-level 2011, Plinthite properties and implications in southeastern U.S. coastal Tertiary alluvium, Gholsonville, Virginia: Virginia Journal of Science, plain soils: American Society of Agronomy–Crop Science Society of v. 45, p. 88. 36 Swezey et al.

Whittecar, G.R., Fitzwater, B.A., Swezey, C.S., and Garrity, C.P., 2013, Geo- Woollen, I.D., and Colquhoun, D.J., 1977, The Black Creek and Middendorf morphic responses of the Carolina Sandhills (Chesterfi eld County, South Formations in Darlington and Chesterfi eld Counties, South Carolina, their Carolina) to Quaternary climate changes: Geological Society of America type areas [South Carolina Geological Survey]: Geologic Notes, v. 21, Abstracts with Programs, v. 45, no. 7, p. 192. no. 4, p. 164–197. Whittecar, G.R., Newell, W.L., and Eaton, L.S., 2016, Landscape evolution in Wright, V.P., and Marriott, S.B., 1993, The sequence stratigraphy of fl uvial dep- Virginia, in Bailey, C., and Eaton, L.C., eds., The Geology of Virginia: ositional systems: The role of fl oodplain sediment storage: Sedimentary Martinsville, Virginia, Virginia Museum of Natural History (in press). Geology, v. 86, p. 203–210, doi:10.1016/0037-0738(93)90022-W. Willett, S.D., McCoy, S.W., Perron, J.T., Goren, L., and Chen, C.Y., 2014, Dynamic reorganization of river basins: Science, v. 343, 1248765, doi:10.1126/science.1248765. Winston, R., 1995, Effects of vegetation differences on hydrology in a longleaf pine savanna: Geological Society of America Abstracts with Programs, v. 27, no. 7, p. 307. MANUSCRIPT ACCEPTED BY THE SOCIETY 3 MARCH 2016

Printed in the USA