Quaternary Geology Of The Fort Polk Area, Sw Louisiana (2002)

BRI Bulletin

51

Pliocene Surface Stratigraphy in the Fort Polk Region: Implications for Louisiana Surface Geology Paul V. Heinrich 1 and Richard P. McCulloh 1

Abstract Recent completion of a project of nearly threeyear's duration to investigate and map the surface geology of ten 7.5-minute quadrangles encompassing the Fort Polk military reservation leads to the conclusion that, in contrast to the prevalent convention of the last several decades, strata assignable with fair confidence to the Pliocene are mappable in Louisiana. The investigation found that the subdivision of the Fleming section into six alternating sand- and clay-rich members of formation rank, following previous authors, continues to be the most effective classification of surface and shallow-subsurface Miocene strata. In accordance with recent work done at the Louisiana Geological Survey, however, the highest elevation and stratigraphically lowest Citronelleequivalent units classify as members of the Willis Formation of Texas, and are assigned to the Pliocene. The next younger units incising the Willis classify as members of the Lissie Formation of Texas and are assigned to the middle Pleistocene. The results of the investigation support the premise that previous failure to recognize Pliocene surface units in Louisiana, in distinct contrast to surface geologic mapping in adjacent states, was an artifact of the classification of Quaternary surface units prevailing here. Recognition of Pliocene strata at the surface could have manifold implications, such as for amplifying the interpretation of paleoclimates via correlation of Neogene surface-stratigraphic units with published coastal-onlap curves.

1

Louisiana Geological Survey, Louisiana State University, Baton Rouge, Louisiana 70803

Introduction The research presented in this paper is a summary of portions of a more comprehensive investigation of the surface geology of the Fort Polk region in northern Vernon Parish and southern Natchitoches and Sabine parishes by McCulloh and Heinrich (1999), in particular, the chapter Pliocene Series (Heinrich and McCulloh, 1999), and funded by the Joint Readiness Training Command at Fort Polk. This research is built on earlier work by Welch (1942), Rogers and Calandro (1965), Andersen (1960, 1993), and most recently by Hinds (1998a, b, c, and 1999). The study area, consisting of ten 7.5-minute quadrangles, is shown in Figure 1.

This research had a threefold objective. First, it was to provide the U.S. Army with basic geologic data essential to the successful conduct of its environmental programs. Second, it was to enlarge upon previous research by Hinds (1997, 1998a, b, c, and 1999) concerning the surficial geology of the Fort Polk region. Finally, this work was to refine and continue previous geologic mapping by the Louisiana Geological Survey within this area (Louisiana Geological Survey, 1993; Snead et al., 1998). The study area was covered using a standard four-wheel drive vehicle to inspect areas accessible along roads. Selected intervening areas where exposures appeared likely were accessed on foot. However, a widespread cover of surficial deposits and thick vegetation conceals the stratigraphic units being mapped and makes exposures scarce. Outcrops occurred as artificial exposures along roads and trails, and natural exposures along the beds, cutbanks, and valley walls of streams and rivers. Typically, exposures were small and inconspicuous and found in the ditches along roads. The better and sometimes sizable exposures consisted of roadcuts and natural bluffs. During fieldwork, notes were recorded on copies of the topographic base maps and in field books for each of the ten 7.5-minute quadrangles in the study area.

52

BRI Bulletin

The field observations were used to delineate stratigraphic boundaries on the basis of gross lithology and internal features. Boundaries were modified on the basis of topography from 7.5-minute quadrangles and distribution of soils as mapped in Natural Resources Conservation Service (NRCS) parish soil surveys for Natchitoches, Sabine, and Vernon parishes (Guillory, 1997; Martinet al., 1990; and Muse and Cooley, in press). Finally, 1:2,000 black and white Edgar Tobin Aerial Surveys and 1:20,000 black and white Agricultural Stabilization and Conservation Service aerial photographs were consulted for preparation of the mapping.

Only a small number of researchers have studied the Pliocene and Pleistocene strata within the study area. Welch (1942) conducted the first and most extensive study of the project area at a time after it had been clearcut of forest but lacked topographic maps. Using the terminology of Fisk (1938), Welch (1942) classified the deposits of the Upland Allogroup as the Williana Formation and the deposits of the Intermediate Allogroup (Lissie Alloformation) as the Bentley Formation. Welch (1942) considered both units to be of Pleistocene age based upon these correlations. On the basis of substantial field work, Andersen (1960) and Bianchi

Vernon

Figure ]-Location map and index to 7.5-minute quadrangles covered by the study area (from Fig. I, McCulloh and Heinrich, 1999:2; reprinted by permission).

BRI Bulletin (1982) further refined Welch's (1942) original Pleistocene framework of the area. Bianchi (1982) made a significant contribution in the recognition of stepped topography associated with Welch's (1942:62) "Pleistocene" deposits. Later, Rogers (1982 and 1993, personal communication) proposed for many years that surface Pliocene strata potentially occurred in these deposits based on their lithological characteristics and elevation of the older and higher Pleistocene remnants. Investigating Roger's (1982 and 1993, personal communication) ideas, Hinds (1998a, b, c, and 1999) mapped these deposits in detail, now considered to be the Upland Allogroup within the areas of Slagle and Fort Polk, Louisiana. Besides providing detailed descriptions of these deposits, he identified outcrops possibly of Pliocene age within the Upland Allogroup. Snead et al. (1998) used the mapping of Hinds (1998b, c) and preliminary results from this investigation in their preparation of the Quaternary Geologic Map of Louisiana (Snead et al., 1998), as part of a cooperative agreement with the U.S. Geological Survey under its State Geologic Mapping Program (STATEMAP).

Regional Setting The Fort Polk area lies in the West Gulf Coastal section of the Coastal Plain Province of Thornbury (1965). Its gulfward part consists of coast-parallel Quaternary terraces. The study area lies in the area of transition between fluvial terraces to the north and coast-parallel terraces to the south. The northern part of the Coastal Plain Province consists of deeply dissected Tertiary uplands characterized by a series of coast-parallel cuestas and lowlands inland of the terraces. The cuestas are commonly called "wolds." Less commonly, the lowlands are called "vales." The cuestas are characterized by steep inland-facing escarpments and gentle gulfward slopes. The southernmost cuesta, the Kisatchie Wold, dominates the Fort Polk region almost completely. Within the Fort Polk region, the Kisatchie Wold consists of two distinct cuestas. The Kisatchie Wold encompasses all of the Main Post and most of the Birds Creek, Fort Polk, Fullerton Lake, Lacamp, Slagle, and Simpson South quadrangles. A well-defined escarpment 30 to 50ft (9 to 15 m) high forms the southern edge of this cuesta. This escarpment separates the highly dissected cuesta

53

from the flatter, less-dissected coast-parallel terrace surface to the south. The southern cuesta rises in elevation from about 270 to 290 ft (82 to 88 m) along its southern edge to an elevation of 400 to 440ft (122 to 134m) along its crest. Northward of this crest, the Kisatchie Wold drops rapidly into the lowland formed by the valleys of Bayou Castor, Cypress Creek, and Calcasieu River. Topographic profiles along the crests of ridges between the major drainages show that the southern cuesta rises as a series of topographic steps instead of as a gradual slope. A series of northwest-southeast trending drainages have cut valleys that deeply dissect these steps. The study area lies downdip of the Sabine uplift, the north Louisiana salt basin, and AngelinaCaldwell flexure to the north; the Toledo Bend flexure and Gulf Coast salt and growth-fault basin lie to the south (Figs. 2 and 3 ). The Angelina-Caldwell flexure is marked at the surface by a zone of faults mapped in Louisiana by Andersen (1960, 1993). The stream net over much of the study area shows rectangular drainage that is suggestive of potential structural control.

Figure 2-Regional geologic and tectonic framework of study area (from Fig. 8, McCulloh and Heinrich, 1999:19; reprinted by permission).

In the study areas, Tertiary strata range in age from Eocene to some high and discontinuous remnants of oxidized, sandy strata herein assigned a Pliocene age. The Tertiary section consists of varying proportions of sand, silt, and mud, with sand

(JJ

+:-

PROJECTION OF STUDY AREA

H ARKANSAS

I

LOUISIANA

LOUISIANA

I

TEXAS

GULF OF MEXICO

CCJ

;;o CCJ

c

([) ~. :::::J

0

50 MILES 0

50 KILOMETERS

VERTICAL EXAGGERATION = 20.8

Figure 3-Dip section through the northern Gulf Coast along 94° west longitude. The study area of this report lies 60 km (37 miles) east of this line of section, and spans units ranging at the surface from the uppennost or farthest downdip portion of the outcrop belt of the Claiborne Group to the lower or updip portion of the outcrop belt of strata of Plio-Pleistocene age. (Geopressure on this diagram refers to "hard" geopressure, equivalent to 13 ppg drilling mud weight or the approximate base of the transition zone). (Redrawn and adapted from the American Association of Petroleum Geologists, 1990; reprinted by permission.)

BRI Bulletin

constituting the greater volume, and reflects deposition ranging from fluvial to shallow marine, with overall increasingly terrestrial character upsection. Quaternary units comprise strata underlying Pleistocene terraces and Holocene alluvium; various surficial deposits of Quaternary age form thin veneers on other map units, and with a single exception, were not mapped for this investigation. Pliocene and Pleistocene units associated with terrace surfaces are divisible into three main groups, here classified as allogroups (units characterized by bounding unconformities): Upland, Intermediate, and Prairie (Table 1). Directly underlying the Plio-Pleistocene strata is the Fleming Formation. As defined by Fisk (1940) for Rapides Parish and mapped by Welch (1942) for Vernon Parish, the members of the Fleming consist of alternating coarser-grained, fluvial-dominated lithofacies and finer-grained, more marine-influenced lithofacies. Fisk (1940) interpreted the coarsergrained 1nembers as fluvial and the finer-grained members as estuarine. Hinds made the refinement of interpreting Fleming members as representing upper and lower deltaic plain deposition, and the latter, corresponding to the finer-grained members, as potentially the updip expression of subsurface marine shale tongues and associated flooding surfaces corresponding to periods of maximum transgression during the Miocene. The vertebrate fossils discovered recently at Fort Polk occur in one of these overall finer-grained subunits, the Castor Creek, but within a notably coarsergrained subinterval consisting of conglomerate and sandstone. Although the Fleming is traditionally classified as a formation, in Louisiana and in the study area in particular, the unit is more appropriately described as of group rank and consisting of formation-rank subunits.

Upland Allogroup One of the more widespread and recognizable stratigraphic units within the Coastal Plain Province is the Upland Allogroup. It is part of a regionally extensive, coast-parallel belt of coarse-grained sediments that have been mapped along the Coastal Plain Province from Virginia to Texas. Typically, the sediments of the Upland Allogroup occur as erosional remnants capping isolated hilltops and comprising the ridge crests of interfluves (Autin et al., 1991).

55

The deposits that comprise the Upland Allogroup have gone through a variety of name changes. Initially they were described by Hilgard (1860) as the Orange Sand Formation and later as the Lafayette Formation of McGee (1891). Later, the Lafayette Formation was abandoned and subdivided into the Citronelle Formation of Matson (1916), the Willis Formation of Doering (1935), the Bentley and Williana Formations of (Fisk 1938, 1940) and others. Since the definition of these formations, there have been seemingly innumerable proposals as to their ages and correlations. Mapping by Snead and McCulloh (1984) reconstituted these formations into an informal morphostratigraphic unit called the High Terraces. Autin et al. (1991), redefining the High Terraces as an informal unconformity-bounded unit, renamed it as the Upland Complex. As a preliminary step towards defining the Upland Complex as a formal allostratigraphic unit, Snead et al. (1998) renamed it as the Upland Allogroup. As defined by Autin et al. (1991), the Upland Allogroup consists of alluvial deposits of early Pleistocene and Pliocene age. The Pliocene-age sediments consist of the Citronelle and Willis formations, which crop out, respectively, east of the Mississippi River Valley and west of the Calcasieu River. Between the Mississippi and Calcasieu rivers, the Upland Allogroup consists of the early Pleistocene-age alluvial deposits designated by Fisk (1938, 1940) as the Williana and Bentley formations. The Upland Allogroup consists of sediments that differ from the underlying Miocene and overlying Pleistocene strata. First, the Upland Allogroup consists of sediments that are typically coarser than the sediments of the Fleming Formation underlying it and those of the Intermediate Allogroup that onlap onto it. The coarser nature of the Upland Allogroup is commonly, but not always, manifest in the presence of either abundant chert gravel or quartz granules (Matson, 1916; Doering, 1935, 1956; Forney, 1950; and Autin et al., 1991). Second, sediments within the Upland Allogroup are characteristically very highly weathered unlike younger and older strata. These sediments are characterized by deep reddish, brownish, and purplish colors of ferric iron oxides not seen in younger or older strata (Doering, 1956; Aronow, 1982; and Autin et al., 1991).

BRI Bulletin

56

Stratigraphic ColuiiJn of Fort Polk Region Rock - Stratigraphic Units

Time - Stratigraphic Units

.,

E Cl)

Cl)

.,

......

.i:

>,

~

Stage

Cl)

~

Group

Holocene

Formation

Member

alluvium + Biq Brushy

undifferentiated

c..

c

:;J

Q)

ca;

Q) +.J

+.J

c:: s..

"' 01 £.2

u

Calabrian

·w V)

('\S

~~~ Q)

Upper

i;lc..

0::

Ci

(undifferentiated alluvium)

<(

0

::::J

e

:~

a;

('\S

·-

:;J

E ~

01 0

] e

Lissie

Q)=

Lower

_E<(

a;

Piacenzian

cQ) u

~~~ "0

.Q

Zan clean

a..

c

c..

:;J

0 ~

"' o_g>

I....

Q

g

Messinian

~~~

Tortonian Blounts Creek

undifferentiated

Castor Creek

undifferentiated

Williamson Creek

undifferentiated

Dough Hills

undifferentiated

Carnahan Bayou

undifferentiated

Aquitanian

Lena

undifferentiated

Chattian

Catahoula

undifferentiated

Q)

c::

g 0 Q)

z

a;

ca;

~

Serravallian

u ""0 .Q :0 ~ ~

I....

a;

~

c:n c

.E Langhian

a;

u::

Burdigalian

0

~

Willis

:::J= <(

_J

('\S

·-e ~

Gravel Hill Allomember TowPr Roilrl Allomember Duaout Road Allomember Kisatchie Allomember Fort Polk Allomember

I....

Q

a;

ca;

g

u

0

.2"1

0

c:n

I....

I....

~

..0

0

Q)

_J

c::

::::J

a;

Rupelian

V)

~

u

undifferentiated

>

~ 0

Q)

~ 0..

c::

I....

a;

g9

0V)

Priabonian

~

u

('\S

undifferentiated

ca; u 0

a;

u.J

a; ""0

~

~

c

I....

Bartonian

0 ..0 C\S

Cockfield

undifferentiated

u

Table ]-Stratigraphic classification of units mapped for this investigation (from Table 1, McCulloh and Heinrich, 1999:23; reprinted by permission.

BRI Bulletin Third, the Upland Allogroup characteristically lacks its original constructional surface. Its surface consists of rolling ridges and ravines produced by deep erosion. This erosion has removed the original depositional surface of the units that it comprises. The only indication that a unit's depositional surface is left might be concordant summits along the crest of major drainage divides. The Upland Allogroup erodes into resistant hills, which creates well-defined cuestas and sharp ridges. In contrast to the Upland Allogroup, the Intermediate Allogroup exhibits remnants of its original depositional surfaces in the form of flat-crested ridges. Finally, the soils associated with the Upland Complex consist primarily of ultisols with some alfisols. The soils developed within the Upland Allogroup are typically far more mature than those developed in either the Fleming Group or the Intermediate Allogroup (Forney, 1950; Autin et al., 1991; and Dubar et al., 1991). Autin et al. (1991) and Autin (1993) attribute the degree of weathering to a regionally pervasive geosol developed within it as the result of intense long-duration weathering of sandy to loamy sediments. According to Autin (1993 ), this geosol has been recognized from east Texas to west Florida and northward within the Mississippi Alluvial Valley. Willis Formation Traditionally, the deposits of Upland Allogroup within south-central Louisiana have been correlated with the Williana Formation of Fisk (1938, 1940). For example, both Welch (1942) and Hinds (1998a, 1999) correlated these sediments with graveliferous fluvial deposits capping hills across the Calcasieu River that Fisk designated as the Williana and Bentley formations. Similarly, Bernard (1950) classified strata equivalent to the Upland Allogroup in southeast Texas as belonging to the Williana Formation of Fisk (1938, 1940).

In contrast, Doering (1956) observed that along strike, Fisk's (1938, 1940) Williana Formation in Ra pides Parish lies below the level of the deposits of the Upland Allogroup in the Fort Polk region. In Figures 3 and 4, Doering (1956) showed that the surface of the Upland Allogroup in south-central Louisiana lies as much as 300 ft (90 meters) above the surface of the Williana Formation on the opposite side of the Calcasieu River. These figures also show that the that gulfward slope on the "surface" of the

57

Upland Allogroup in south-central Louisiana is drastically steeper than the gulfward slope on the surface of the Williana Formation of Fisk (1938, 1940). Winker (1991) also illustrated the same differences in the stratigraphic position of the Upland Allogroup in south-central Louisiana and the Williana Formation across the Calcasieu River. He classified the Upland Allogroup in the Fort Polk region as his preLissie surface(s) and the Williana Formation of Fisk (1940) as being part of his Lissie surface. Of these two interpretations, a carefully constructed cross section, Figures 4 and 5, confirms those made by Doering (1956) and Winker (1991) concerning the stratigraphic separation between the type Williana and Bentley formations east of the Calcasieu River and the Upland Allogroup in south-central Louisiana. The sediments of the Upland Allogroup within south-central Louisiana lie as much 200 ft (60 m) above the top of the Williana Formation across the Calcasieu River in Rapides Parish. The base of the youngest and lowest sediments of the Upland Allogroup in the study area lies barely level with the top of the Williana Formation. The base of the Williana Formation lies about 50 ft (15 m) below the base of deposits in the study area. The great difference in stratigraphic position and lithology between the sediments in the study area and the Williana Formation in Rapides Parish demonstrates that they are different units of formation rank within the Upland Allogroup. In terms of lithology and stratigraphic position, the strata of the Upland Allogroup are identical to the Willis Formation of Doering (1935) (Forney, 1950; Bernard and LeBlanc, 1965; and Aronow, 1982). For this reason, they are correlated with and designated as the Willis Formation in this investigation. In addition, the deposits of the Williana and Bentley formations are deeply entrenched into both the underlying Miocene below the level of the Upland Allogroup strata west of the Calcasieu River. This demonstrates that the Williana and Bentley formations are different stratigraphic units from the Willis Formation and younger than it. Lithology Within the study region, two different facies, a sandy facies and gravelly facies were discerned. Bianchi (1982:74) also noted these facies which he labeled, respectively, "fine sandy facies" of his "high coastwise group," and the "graveliferous facies" of

t.n

co Lacamp Section East

West Intersection with Birds Creek Section

500

1200 1500 1800 2100 2400

Feel

500

Puwf

j 400

Q)400~~

tf .!: c: 300 0

~

,._ "' ~~~ ~ g~

.c:

&l .§ (/)

100

93 07' 30"

I

::J

0

(/)

[ij 200

Williana Formation of Fisk (1940)

(/)"'

Mfcc

c::;:o ~,._

~

c.

E~

~;.:

Mfcc

\

300

200

Ui

92 52' 30"

Mfcc

93 00' 00"

100

Legend for Figures 4, 8, 9, and 10 Stratigraphic Units Quaternary System Holocene Series

lEI lim

Hua

Unnamed Alluvium

Hbb

Big Brushy Formation

O:l

;:;o

O:l

Pleistocene Series

Pp

i!

Prairie Allogroup undifferentiated Intermediate Allogroup

Pilu

~ ~

Pill

Tertiary System

Pliocene Series

[GJ

Puw

[jill

Puwg

[£']

Puwt

D D D DMfb D Mfcc

Puwd Puwk Puwf

Castor Creek Member

Figure 4-Cross section along crest of Kisatchie Wold and across Calcasieu River Valley. Horizontal scale= 1:125.000. Vertical exaggeration= 52X (from Fig. 18, Heinrich and McCulloh, 1999:49; reprinted by permission).

c

(D ~.

::::J

BRI Bulletin

59

93'15'00" 3r15

Slagle N

+

Elmer

31

Melder I

Melder

31'00'00"1---------~-----+-----'----l---.=::::t-.L.--------~--------.131

93'15'00"

93'00'00"

92' 45'00"

'00'00" 92'37'30"

LEGEND Main Post area of Fort Polk ~

Cross-section line

•

Small-size town Medium-size town

Figure 5-/ndex map showing locations of cross sections relative to the main post at Fort Polk, Louisiana (from Fig. 19, Heinrich and McCulloh, 1999:52; reprinted by permission).

his "low coastwise group." The gravelly facies is restricted to the youngest and lowermost units of the Willis Formation within the study area.

Sandy Facies The sandy facies consists of mostly very fine to medium sand and muddy sand. Purplish, pebble to cobble-size clasts of clay and silty clay, quartz granules and minor beds and lenses of silt and clay occur throughout this facies. In many outcrops, the sandy facies is largely massive, having been homogenized by plant roots and weathering. Where preserved, these sediments exhibit sedimentary structures that include planar cross-bedding, trough cross-bedding, ripple lamination, climbing ripple lamination, and scour-and-fill structures. Of these sedimentary structures, medium-scale trough cross-bedding is the most common. The fine-grained sediments occur as discontinuous beds and lenses. No upward fining or coarsening trends were observed in the sandy facies. The sandy facies exhibits bright and variable colors. Although predominantly various shades of red, the sandy facies can vary from reddish brown and weak red, to strong brown and yellow and pale red-purple

(Hinds 1998a, 1999). Purplish, fine-grained, rounded rip-up clasts occur in almost every outcrop of the sandy facies in varying abundance. Frequently, these purplish clasts of clay or silty clay are abundant enough to form thin beds of rip-up clast conglomerate. They occur concentrated along cross-beds and at the base of scours. The purplish rip-up clasts range in diameter from 2 to 5 em (0.8 to 2 inches). Rip-up clasts as large as 8 em (3 inches) in diameter were observed, whereas Hinds (1997) reported rip-up clasts as large as 25 em (10 inches). Whitish clay or mud clasts of sand size commonly occur on and delineate crossbeds. In places, the sandy facies is gravelly. Typically the gravel consists of subrounded to subangular quartz clasts 1 to 2 em (0. 4 to 0. 8 inches) in diameter. The granule gravel occurs either scattered throughout the sand or discrete gravel stringers on cross-beds. The only fossils observed in the sandy facies are rare pieces of petrified wood observed at two locations in the Fort Polk area. In both cases, the petri-

BRI Bulletin

60

fied wood occurs at the base of the Willis Formation near its contact with the Fleming Group. In contrast, Bernard (1950) observed that petrified wood was common in the sandier parts of the Willis Formation. The petrified wood consisted of large logs and stumps, which Bernard judged to be native to the Willis Formation.

Gravelly Facies The gravelly facies consists of a heterogeneous mixture of muddy, medium-to-coarse sands, gravelly sands, sandy gravel, and gravel. Like the sandy facies, exposures of the gravelly facies frequently are massive, having been homogenized by plant roots or weathering. Where preserved, well-developed medium to thick sets of trough cross-bedding characterize the sandy gravels and gravels. Sand beds within this facies exhibit tabular and trough cross-bedding of varying scales, and ripple lamination. Purplish clay to silty clay rip-up clasts identical to those found in the sandy facies are abundant within the gravelly facies. No evidence for any overall upward fining or coarsening trends within the gravelly facies was observed during fieldwork. The gravel consists of granule- to pebble-size chert gravel. As noted by Lenzer (1982), the gravel consists mainly of "dense" brown and brownish gray chert with minor amounts of other types of cherts, sedimentary quartzite, and metamorphic quartzite. Quartz, which is abundant as granules in the sandy facies, is noticeably absent among the rock types (Lenzer, 1982; Hinds, 1998a, 1999). The largest pebbles observed ranged from 4 to 5 em (1.5 to 2 inches) in diameter. The chert gravel contains the only fossils so far found in the gravelly facies of the Willis Formation. These fossils consist of the molds of rare brachiopods, crinoid stems, and corals, and other Paleozoic marine invertebrates. These fossils are not native to the Willis Formation, but rather were derived from Paleozoic carbonate rocks from which the chert was eroded and redeposited. The study of these fossils might provide some evidence concerning the source of the chert gravels found in the gravelly facies. The colors exhibited by the gravelly facies typically consist of various shades of red to yellowish brown. In a single outcrop, colors can include dark reddish brown, yellowish red, and brownish yellow surrounding numerous purplish rip-up clasts. In some

outcrops of both the gravelly and sandy facies, they consist of bleached-looking, grayish sediments not unlike those of the Catahoula and Fleming in coloration, with red mottling in places. These bleached sediments are typically homogenous mixtures of sand, mud, and in places, gravel. Abundant root molds and haloes are commonly associated with these sediments.

Contacts Within the Fort Polk area, the Willis Formation unconformably overlies the Fleming Group. In outcrops, the contact between the Willis Formation and Fleming Group is extremely irregular and in many places channeled (Fig. 6). Within small areas, the basal contact has been mapped as having 10's of feet of relief upon it. Regionally, the contact is a lowangle angular unconformity that northward progressively truncates the underlying Fleming Group. At the northern edge of the main Willis outcrop belt, the Blounts Creek Formation is completely missing and the Willis Formation rests directly on the Castor Creek Formation. The Fleming Group directly underlying the Willis Formation often shows evidence of weathering. Typically, the Fleming Group lying within a meter of the contact exhibits the development of purplish alteration and rectangular blocky peds. Commonly, the overlying sediments of the Willis Formation contain purplish rip-up clasts derived from the underlying weathered sediments of the Fleming Group. Sediments of the Willis Formation incorporating purplish rip-up clasts overlying weathered, fine-grained Fleming Group sediments, from which the clasts were derived, were observed at numerous localities. Welch (1942, Fig. 11) depicts these relations at this contact exposed in a railroad cut near Pickering (Fig. 7). Not all of the observed contacts between the Willis Formation and Fleming Group exhibit this alteration below the contact. In a few exposures, bleached sediments of the Willis Formation overlie relatively unweathered fine sands or other sediments of the Willis Formation. Within the study area, the upper contact of the Willis Formation consists of the modern land surface and another major unconformity between it and the overlying Intermediate Allogroup. The concordant summits of knolls along the crests of interfluves define a series of deeply eroded steps that characterize the upper boundary of the Willis Formation

BRI Bulletin

61

.f

4

3

·:~

.., 1..

2

Q)

+J

Q)

~

EXPLANATION Yellow loam Brownish yellow 1OYR 6/6 Sand, overall very fine to fine, some sparse medium to very coarse grains. Yellowish red 5YR 5/6-8, with abundant white 1OYR 8/1 to very pale brown 1OYR 8/2 burrow mottles. Red 7 .5R 4/6mud rip-up clasts near base.

Clayey very fine to fine sand Light-gray 2.5Y 7/2 with red 1OR 4/8 mottles. In places beneath contact with overlying sand at north end of cut, mud shows alteration to red 7.5R 4/6 in interior cores of peds, with light gray 2.5Y 7/2 remaining along ped surfaces. Apparent dip 4°. Cross-bedded sand Overall very fine to medium, with some grains ranging from coarse sand to granules. Reddish yellow 5YR 6/6, with very pale brown 1OYR 8/4 burrows and root molds. Red 7.5 4/6 mud burrow casts up to 0.5 m below upper contact.

Figure 6-Combined outcrop sketch and measured section (no horizontal scale) along roadcut, SEINW Sec. 30, T. 5 N., R. 9 W, Peas on quadrangle. Mottled sediment of the Upland Allogroup is mantled by yellow loam and cuts out two units of the Carnahan Bayou Fonnation of the Fleming Group (from Fig. 13, Heinrich and McCulloh, 1999:34; reprinted by pennission).

BRI Bulletin

62

(Bianchi, 1982). Where the Lissie Formation of the Intermediate Allogroup onlaps onto the Willis Formation, it completely truncates the Willis Formation. Geological mapping indicates that the removal of the Willis Formation is complete and the Lissie Alloformation rests directly on the sediments of the Fleming Group. Bernard (1950) also observed a similarly deeply eroded unconformity between the Lissie Alloformation and Willis Formation in Jasper and Newton counties, Texas.

Overlying Strata Either overlying the Willis Formation or occupying stream valleys incised through it, within the Fort Polk area, are younger fluvial sediments of the Intermediate and Prairie Allogroups. In this area, the Intermediate Allogroup consists entirely of the Lissie Alloformation and the Prairie Allogroup consists of undifferentiated fluvial deposits that form terraces of relatively small extent. Unnamed allu-

NORTH

SOUTH

FEET

300

SELLERS STORE

WILLIS FORM _-UN ATION CONFORMITY-

---

250 BLOUNTSCRE

EK MEMBER OF FLEMING FORMATION

200 ONE MILE

EXPLANATION 1. 2. 3. 4. 5.

6.

7. 8. 9.

Light-gray silty clay. Sandstone lentil and ironstone at base of weathered material. Weathered, silty clay and residual terrace sand, 2 feet. Sandy siltstone lintil, thickness 0-6 feet, well exposed in railroad cut. Small remnant of Willis Formation; 0-10 feet red, massive sand exposed above highway, cross-bedded. Contact with underlying Blounts Creek member irregular; chert gravel and clay balls at contact. Sandy siltstone, 0-6 feet thick, underlain by unconsolidated silty clay and overlain by about 6-10 feet of orange, red and purple mottled, weathered, argillaceous silt; distinct contact with Willis Formation and suggestion of pre-Willis soil horizon. Massive sand, thicker down hill, is result of soil creep and occurs extensively on outcrops of both terrace formations and on sandy Miocene strata. Massive red sand, gravel at base. Gray, weathered silty clay.

Figure 7-Unconformable relationship of Willis Formation and underlying Blounts Creek Formation of the Fleming Group along U.S. Highway 171 in Kansas City Southern Railway cut at 2.4 km (1.5 miles) south of Pickering (Sec. 3, T. IS., R. 9 W, New Llano quadrangle). (Redrawn and adapted from Fig. 11, Welch., 1942:29).

BRI Bulletin

vium of Late Pleistocene to Holocene in age underlies the flood plains. However, it is not considered in this discussion, as the details of these deposits are discussed in McCulloh and Heinrich (1999). Intermediate Allogroup The Intermediate Allogroup of western Louisiana consists of the fluvial deposits of the Calcasieu and Sabine rivers, their tributaries, and various coastal plain streams. This allogroup consists of strata that Bernard (1950) classified as the Bentley and Williana Terraces in southeast Texas; mapped as the Lissie and Oberlin formations by Doering (1956), and as the Oakdale and DeRidder surfaces by Snead et al. (1995). The Intermediate Allogroup of southwest Louisiana consists of the Elizabeth, Oakdale, and Lissie Alloformations.

The Intermediate Allogroup lies stratigraphically and topographically between the Prairie and Upland allogroups. The surface of the Intermediate Allogroup lies topographically higher than the terrace surfaces of the Prairie Allogroup and topographically lower than the dissected uplands underlain by strata of both the Upland Allogroup and Fleming Group. Although heavily dissected, terrace surfaces of the Intermediate Allogroup are recognizable as flat-topped ridge crests. However, erosion has long since removed any constructional fluvial or coastal landforms. Slightly south of the Fort Polk region in west-central Louisiana, the Intermediate Allogroup consists of 100 to 300ft (30 to 90 m) of red, brown, and buff interbedded sand, silt, and clay. In general, the sediments of the Intermediate Allogroup are much sandier and have more mature soils, mostly alfisols, than sediments that characterize the Prairie Allogroup (Autin et al., 1991; Dubar et al., 1991; and Snead et al., 1998). Lissie Alloformation Within the study area, the only unit of the Intermediate Allogroup present is the Lissie Alloformation. It consists of the deposits of the Calcasieu and Sabine rivers, their tributaries, and various coastal streams. In this region, the Lissie Alloformation is associated with two coast-parallel terraces. A higher and older terrace forms a narrow belt along the edge of the Tertiary Uplands. The younger terrace is the northern edge of the extensive, gulfward-dipping ramp-like surface that extends southward from the study region into Calcasieu Par-

63

ish. This terrace surface dips to the south at 4 ft/mi (0.8 rnlkm) except where it is offset by fault-line scarps. The Lissie Alloformation extends eastward along the entire Texas coastal plain, where it was named into Mexico as illustrated by Bernard and LeBlanc (1965) and Winker (1991). Only within the Cooter's Bogs area in the southeast corner of the study area were small outcrops of the Lissie Formation observed. Welch (1942) mapped both of these terraces as the Bentley Terrace as he did most of the Lissie Formation within the Fort Polk region. In this area, small exposures of the Lissie Formation consist of grayish clayey silt and very fine sand with yellowish brown mottles. In a hand specimen, these sediments superficially resemble sediment of the Fleming. They contrast sharply with the sediments of the Willis Formation, which are pervasively stained, even cemented, with reddish, brownish, and purplish ferric iron oxides. In places, yellow loam mantles these sediments. Prairie Allogroup (Undifferentiated) The Prairie Allogroup consists of fluvial, deltaic, estuarine, and beach deposits that lie topographically below the Intermediate Allogroup and topographically above the Deweyville Allogroup. Currently, this allogroup is considered to be late-tomiddle Pleistocene in age. The terrace surfaces associated with it exhibit little dissection and commonly show relict constructional topography (Autin et al., 1991; Dubar et al., 1991).

Within the Fort Polk region, the Prairie Allogroup consists of fluvial sediments that form discontinuous terraces found within the valleys of local drainages. Hinds (1998a, 1999) noted that the deposits of the Prairie Allogroup are recognized as relatively flat areas lying slightly above the level of the Holocene flood plain. It can be inferred from the types of soils, as described by Guillory (1997) and Martin et al. (1990) associated with the terraces of the Prairie Allogroup that it consists largely of silts, silty sands, and sands within the Fort Polk region. Discussion There are two major interpretations for the stratigraphy of the Willis Formation within the Fort Polk region. First, Welch (1942) and Doering (1956) infer that the Willis Formation consists of a single gulfward-tilted stratigraphic unit with a highly dissected terrace that both dip below the deposits of

0"'1 ~

FORT POLK SECTION

Puwd

Northeast

Southwest

300

500

Pilu

Puwf

w 400 If

200

I

.!:

§

300

300

North Fort Polk

~

"' I

[iJ 200

~~~~

100

~~ "'

Mfcc

~~&~

&~

I ~&~ I"' I ~~~

200

~

~

100

~=="---""====i;...--'==~5 Milos

Figure 8-Cross section along crest ofintelfluve on which Fort Polk and North Fort Polk lie. Horizontal scale= 1:125,000. Vertical exaggeration= 52X. (Refer to legend on P. 58.) (From Fig. 23, Heinrich and McCulloh, 1999:61; reprinted by permission).

CJ:1

;;a CJ:1

c

([) ~.

:J

Q)

>

~

I

500

Puwf / Puwk Puwk profile /off-section ~\------------.

400

co Q) (J)

Q)

South

BIRDS CREEK SECTION

500

400

300

i~ 200

300

Intersection with Lacamp Section

Mfb

Pilu

~~, U)~~ ~

,.-

~ ~

,.-

j t-- mot--

200

Mfb

Mfb

100

100

5 Miles ====----====----~== 0 300 600 900 1200 1500 1800 2100 2400

Feet

=:=::~~--===---==--6 Kilometers

Figure 9-Cross section along crest of interfluve separating Birds Creek and Little Sixmile Creek-Sixmile Creek drainages. Horizontal scale Exaggeration= 52X (Refer to legend on p. 58.) (From Heinrich and McCulloh, 1999:63; reprinted by permission.)

= 1:250,000. Vertical

North

cu >

~ Cll Q)

en

Puwd

400

I

Fullerton Branch Section

400

I

300

300

§g 0

-g

200

200

Mfb Mfb

05 Q)

LL

500

~

~ Cll

400

100

South North

Fullerton Section

Q)

en

§g

100

500

.-

400

II

Puwg? 300

300

0 ..0

Cll

05200~

~

c: 0

100

c:

""1.8 ""

200

,t~~ ~~ ..c:

Figure 10-Cross section along crest of inte~fiuve that forms drainage divide benveen the watersheds of Big Brushy Creek and Ten Mile Creek, and cross section along crest o.fside inte1jluve. Horizontal scale= 1:250,000. Vertical exaggeration= 52X. (Refer to legend on p. 58.) (From Fig. 25, Heinrich and McCulloh, 1999:65; reprinted by permission.)

. ---

0

2

3

4

100

OJ

;:;:o OJ

II-

~

0'1

VI

66

BRI Bulletin

the Lissie Alloformation. They both showed the Lissie Alloformation conformably onlapping the deposits of the Willis Formation. In contrast, Bianchi (1982) illustrates the Willis Formation as forming a single pediment cut into the Fleming Group with a relatively flat terrace surface. Instead of dipping beneath the deposits of the Lissie Formation, the Willis Forillation and its associated terrace surface are truncated by the Lissie Alloformation. In southeast Texas, Bernard (1950) combines aspects of both models in that he interpreted the Willis Formation as a single gulfward-dipping stratigraphic unit truncated by the Lissie Alloformation. The 7.5-minute geologic quadrangles and cross sections prepared for this study show a modified interpretation of the ideas of Bianchi (1982). The topographic profiles of the cross sections confirm the existence of the stepped topography discussed by Bianchi (1982). At least five stepped surfaces are shown by the topographic profiles. Integration of the data from the geologic mapping into the cross sections revealed that the base of the Willis Formation generally parallels the concordant summits of each of the steps (Figs. 5 and 8 through 10). In addition, the cross-sections show that abrupt changes in the elevation of the basal contact of the Willis ForIllation coincide with the scarps separating each step. Based upon these observations, the steps are interpreted to represent a series of five relatively flat coast-parallel terraces underlain by fluvial deposits of the Willis formation. Each terrace is lower than the next terrace to the north. Correspondingly, its associated fluvial deposits are more deeply downcut into the Flen1ing Group relative to the terrace to the north. These terraces and associated fluvial deposits are inferred to be separated by erosional contacts (Figs. 8 through 10); as such, they are interpreted as allostratigraphic units. On this basis, five informal allostratigraphic units are recognized and mapped: the Fort Polk allomember, Kisatchie allomember, Dugout Road allomember, Tower Road allomember, and Gravel Hill allomember (Figs. 8 through 10). The Fort Polk and Kisatchie allomembers consist of the sandy facies. The Dugout Road, Tower Road, and Gravel Hill allomembers consist of the gravelly facies. The deep dissection of these deposits by local drainages has reduced the surface of each allomember to a series of concordant summits. In case of the highest

and oldest allomember, the Fort Polk allomember, all evidence of its surface has been obliterated. Its fluvial deposits have been reduced to erosional remnants capping the east-west crest of the southern cuesta and isolated summits along the interfluves extending southward from it (Figs. 8 through 10). North of the main post of Folk Polk, the Willis Formation occurs as widely scattered, discontinuous outliers of limited extent. These outliers form only the highest summits, commonly with maximum elevations exceeding 400ft (122m).

Origin The deposits of the Willis Formation are generally agreed to be fluvial in origin (Welch, 1942; Doering, 1935, 1965; Bernard, 1950; and Autin et al., 1991 ). However, insufficient data exist to discern the specific type of fluvial system that created these deposits. Regardless of whatever type of fluvial system deposited the Willis Formation, the contrast between the coarse-grained Willis Formation and the underlying and overlying units shows that it represents a radically different and unusual sedimentological regime. The widespread distribution of correlative coarse-grained and deeply weathered strata, the Citronelle Formation, indicates that environmental factors influencing the deposition the Willis Formation affected all of the northern coastal plain of the Gulf of Mexico. What the specific factors, e.g., eustatic sea level change, epeirogenic uplift, or paleoclimatic changes, that are involved remain unresolved. The reddish, brownish, or purplish ferric iron oxides that pervasively stain, even cement, the sediments of both the sandy and gravelly facies show that the entire Willis Formation has been highly weathered to very deep depths. The deeply weathered nature of the Willis Formation reflects a number of influences. These include: (1) the high initial permeability of the coarse parent materials; (2) the thickness of the drained soil column above the water table; (3) the comparatively great age of the strata within the Willis Formation; and (4) contemporaneous weathering of these strata during their accumulation as a result of warm Pliocene paleoclimates (Aronow, 1982). The general lack of fossils within both the sandy and gravelly facies of the Willis Formation implies that conditions during the deposition of the Willis were generally not conducive to the preservation of

BRI Bulletin fossils. Because of the coarse-grained nature of the Willis Formation, oxidation removed organic matter and leaching destroyed bones and shells during and after deposition. Also, the accumulation of the Willis Formation during a period of relatively hot and humid climate would have greatly accentuated these processes and largely precluded the formation of fossils. Rare petrified wood is preserved because silicification occurs relatively soon after burial.

Age Because age-diagnostic fossils have not been found in the Willis Formation, it cannot be directly dated at this time. Instead, what is known about the age of the stratigraphic units that unconformably onlap it and which it unconformably overlies must be used to infer its approximate age. Also, it is presumed that the major unconformities that separate the Willis Formation from the underlying Fleming Group and overlying Intermediate Allogroup represent major eustatic and/or tectonic events significant enough to have been recognized and dated. The Lissie Alloformation of the Intermediate Allogroup, which unconformably onlaps the Willis Formation, is likely of Middle to Early Pleistocene in age. In Texas, the Lissie Formation contains Pleistocene vertebrate remains that include Bison latifrons (Harlan), Mammuthus columbi (Falconer), Mammuthus imperator (Leidy), Equus excelsus Leidy, Equus francisi Hay, Equus complicatus Leidy, and Equus semiplicatus Cope (Forney 1950). Kukla and Opdyke (1972:569) reported sediments with reverse magnetic polarity from the "Lissie," "Bentley," and "2nd terrace," which Dubar et al. (1991) suggested are equivalent to the Lissie Formation. Finally, Winker (1979, 1991) correlates the surface of the Lissie Formation to the top of the R6 Reflector, a regional seismic reflector in the offshore. This regional reflector lies immediately below the Trimosina A (denticulata) bethic foraminifera datum that dates from 0.6 to 0.65 million years ago (Dubar et al., 1991; Winker, 1979, 1991). This is consistent with this reflector and the top of the Lissie Alloformation (Formation) and Intermediate Allogroup being associated with a major regression within the Gulf of Mexico that occurred about 0.80 million years ago. The break between the Prairie and Intermediate Allogroups would not only be associ-

67

a ted with a major regression that started at this time, but also a significant change in the frequency of sealevel fluctuations (Fig. 11 ). As discussed previously, the Williana and Bentley formations of Fisk (1938, 1940), which are found east of the Calcasieu River and west of the Mississippi River, lie downcut below the level of the Willis Formation and the correlative Citronelle Formation. They are erosional remnants of fluvial deposits that apparently filled earlier alluvial valleys of the Mississippi and Red rivers entrenched deeply through earlier deposits of the Willis and Citronelle formations. Woodward and Gueno (1941) interpreted the presence of very large boulders of chert, quartzite, sandstone, and petrified wood within the Williana as being the result of downstream transport by icerafting. Pebbles of igneous rock and Baraboo Quartzite in gravels of the Williana Formation indicate the presence of glacial outwash within it (Fisk, 1939). Although, as hypothesized by Autin and others (1991), these gravels might belong to pockets of glacial outwash included in the Williana Formation along the sides of the Mississippi River Valley, the above pebbles imply that the gravels are in part of glacial-outwash origin. Being older than the Lissie Formation, the Williana Formation would have formed during Early Pleistocene or Late Pliocene glaciations. This would imply that the Willis Formation is no younger than Late Pliocene. Vertebrate faunas described from both the Goliad Formation in Texas and the Fleming Group in Louisiana constrain the lower limit of the age of the Willis Formation. Schiebout (1994, 1997) estimates the age of the vertebrate fauna from the Castor Creek Formation of the Fleming Group to be between 12.5 and 14.5 million years old. Given the thickness of the Blounts Creek Formation lying between it and the vertebrate-bearing beds in the underlying Castor Creek Formation, the Willis Formation is significantly younger than this fauna. In Texas the Willis Formation overlies the Goliad Formation, which contains the Lapara Creek Fauna, which Prothero and Manning (1987) estimated to be 10 to 11 million years old. Interpreting these faunas together suggests that the Willis Formation is significantly younger than 10 million years old. A maximum age for the Citronelle and Miccosukee Formations, east of the Mississippi Alluvial Valley, is provided by other vertebrate faunas.

BRI Bulletin

68 Geologic Time (Million years)

Eustatic Curves

Epoch 200

100

Relative Change of Coastal Onlap 0 m

shoreward

Regional Stratigraphy Southwest Louisiana

seaward

(])

c

(])

(.)

0

+-'

en

1

(])

a.. 2 (])

c

(]) (.)

0 3

(L

4 5

6 7

Q) +-'

ctS

8 9 10

11 12 13

_J

(])

c

(]) (.)

0 ~ Q)

-c -c ~

14 Om

Figure ]]-Suggested correlation of stratigraphic units in southwest Louisiana with sea-level and coastal onlap cycles of Paleo-Data, Inc. (1993). Data from Paleo-Data, Inc. ( 1993). (From Fig. 26, Heinrich and McCulloh, 1999:69 reprinted by permission.)

BRI Bulletin

Huddlestun (1988) argued that the Ashville local fauna of north Florida, which consists of late Barstovian vertebrate fossils, came from strata underlying the Miccosukee Formation. Thus, the Miccosukee Formation postdates a middle Miocene (late Barstovian) age of about 11 to 16 million years (Woodburne and Swisher, 1995). In southwest Alabama, the Mauvilla local fauna occurs in paralicalluvial deposits of the "Miocene coarse clastics" of Raymond (1985:64), which underlie the Citronelle Formation. The Mauvilla local fauna consists of late Miocene (early Hemphillian) vertebrates that date these sediments to about seven to nine million years ago (Hulbert and Whitmore, 1997; Woodburne and Swisher, 1995). These vertebrate faunas indicate that the Citronelle and Miccosukee Formations are younger than middle-to-late Miocene age. Within southeast Texas, Morton et al. (1988) correlate the base of the Willis Formation with the Bigenerina "A" (Bigenerina floridana) benthic foraminifer datum. Allowing for the arbitrary drop in datum, Morton et al. (1988) correlated the Willis Formation with the middle and upper portions of the Bigenerina shale. The correlation of the Willis Formation with the transgressive Bigenerina shale and equivalent benthic foraminifer datums implies that it correlates with major sea-level highstands that occurred during the Early Pliocene (Paleo-Data, Inc., 1993) (Fig. 11). Kukla and Opdyke (1972) reported the presence of normally magnetized sediments within the Willis Formation. They interpreted the normal magnetism of its sediments to represent the Gauss polarity epoch, Chron C2An, 2.6 to 3.5 million years ago. However, normal magnetism is also consistent with Chron C3n, 4.1 to 5.2 million years ago, which includes many of the Early Pliocene sea-level highstands. It is not determinable which of these periods of normal magnetism these normally magnetized sediments belong to and whether sediments with reverse magnetism are also present within the Willis Formation. Finally, scanty palynological data indicate a Pliocene age for the Citronelle Formation. Otvos (1997) recovered trace amounts of Sciadopytis (Japanese umbrella pine) pollen from peats exposed in pits dug into the Citronelle Formation at Vancleave, Alabama, and Mossy Head, Florida. The presence of Sciadopytis pollen precludes a Pleistocene age for the

69

Citronelle Formation because the Japanese umbrella pine became extinct in North America prior to the end of the Pliocene (Otvos, 1997 and 1998). For example, Sciadopytis pollen occurs in the Plioceneage Yorktown Formation of North Carolina and Virginia and the lower Beavertown Formation of Delaware and New Jersey and is absent in the overlying Late Pliocene and Pleistocene strata (Groot, 1991; Cronin et al., 1993 ). The period of geologic time covering the Pliocene and late Miocene epochs, three to six million years ago, is judged to be the most likely time during which the bulk of the Willis Formation in the Fort Polk region accumulated. The unconformity at the base of the Willis Formation correlates with a major drop in sea level about 10.5 million years ago (Fig. 11 ). The drastic drop in sea level shifted deposition tracts southward (Wornardt and Vail, 1991; Paleo-Data, Inc., 1993 ). This regression would have ended the accumulation of the Blounts Creek Formation of the Fleming Group within the Fort Polk Area. Between 5.8 and 10.5 million years ago, sea level was either significantly below or close to present sea level. As a result, even during highstands, the edge of coastal plain deposition would have remained south of the Fort Polk region (Fig. 11). During this time, erosion of the Fleming Group would have occurred in response to sea-level drop and ongoing uplift. Between 5.8 and 3.0 million years ago, during a series of highstands, sea level rose as much as 87 to 120 ft (25 to 35m) above present sea level. The inner edge of coastal plain deposition would have transgressed during the highstands back into the Fort Polk region and resulted in the deposition of fluvial sediments (Wornardt and Vail, 1991; Paleo-Data, Inc., 1993). If so, the Willis Formation could represent this period of high sea levels, while individual allomembers represent individual highstands. As discussed by Otvos (1997, 1998), the Citronelle formation contains evidence of sea-level highstands higher than modern. This evidence consists of 0.5 to 7.0-m-thick units composed of muddy sands and muddy, pebbly, fine-to-coarse sands deposited in inshore and nearshore environments. The marine origin of these sediments is shown by the presence of either Ophiomorpha burrows produced by ghost shrimp; the tubes of polychaete worms; the internal molds of veneroid and~ other shallow marine bivalves; or some combination of them. Otvos

BRI Bulletin

70

(1997) observed these marine beds as occurring near both the base and top of the Citronelle Formation. Huddlestun (1988) and Otvos (1998) have correlated the Citronelle and Miccosukee Formations with Upper Pliocene, not Lower Pliocene, sea-level highstands. They argue that the Miccosukee Formation and the correlative Citronelle Formation in adjacent Florida is of late Pliocene (Piacenzian) age. These arguments are based upon the Citronelle Fortnation overlying the fossiliferous Upper Pliocene Jackson Bluff Formation in the Florida panhandle and on Huddlestun's (1988) correlation of the Miccosukee Formation with the microfossil-dated Cypresshead Formation of eastern Georgia (Otvos, 1988). Because of uncertainties in correlations and facies changes between Florida and south-central Louisiana, a Late Pleistocene age for the Citronelle Formation in Georgia and Florida fails to preclude an Early Pliocene age for the Willis Formation in southcentral Louisiana. However, it does imply that the younger allomembers of the Willis Formation likely are of Late Pliocene age. If Late Pliocene age deposits exist within the Willis formation, the younger, gra veliferous Tower Road and Gravel Hill allomembers, which closely resemble the Citronelle Formation in lithology, would be the most likely candidates for such deposits. In this case, the major unconformity between the Willis Formation and the onlapping Lissie Formation would represent a major lowstand of sea level that occurred about 1.8 million years ago (Fig. 11.) Summary

A detailed study of the Upland Allogroup shows that the strata cropping out west of the Calcasieu River consist of fluvial deposits that are correlated to and designated as the Willis Formation in Texas. The Willis Formation is a separate and older stratigraphic unit from the Williana and Bentley Formations as defined by Fisk (1938, 1940) in central Louisiana. The Willis Formation is considered a stratigraphic equivalent of the Citronelle Formation in Mississippi and of Upland Allogroup strata in the Florida Parishes of Louisiana. Within the Fort Polk region, the Willis Formation consists of two well-defined sedimentary facies and five allostratigraphic units. On the basis of lithology, the Willis Formation can be divided into

sandy and gravelly facies. Mapping of these strata and stepped surface topography along its interfluves indicates that the Willis Formation consists of five members separated by major, scarp-forming unconformities. These allomembers form five coastparallel terraces. The contact between the Willis Formation and the Fleming Group is an angular unconformity and it is unconformably onlapped by the Intermediate Complex. The Willis Formation is considered to be Early and possibly Late Pliocene in age. It is hypothesized that it represents fluvial deposition that occurred during sea-level highstands 3.0 to 5.8 million years ago and possibly as late as 1.9 million years ago. Acknowledgments

The research discussed in this paper was funded by the U.S. Army Corps of Engineers (USACE), Joint Readiness Training Command and Fort Polk divisions, on a contract administered and managed by Stephen Austin, Fort Worth District, under Contract No. DACA63-95-D-0051, Delivery Order No. 0008. The results of this research were submitted to Prewitt and Associates Consulting Archaeologists, Inc., Austin, Texas, as McCulloh and Heinrich (1999). The USACE gave us permission to publish figures contained that report. We also acknowledge the American Association of Petroleum Geologists who granted permission to publish Figure 4 of this report. Many people gave to us assistance that was integral in the conduct of this research, and without which it could not have been completed. The most frequent and essential help with operational matters was given by Jim Grafton, Bob Hays, and Gina Lay of the Environmental Learning Center at Fort Polk, and by Barry Oswald and the other Range Control staff who provided general scheduling of our access to land on the military reservation. Dr. Charles H. Stagg, chief of Environmental and Natural Resources Management Division of the Directorate of Public Works at Fort Polk, was uniformly supportive of our efforts. Mr. McCann provided crucial help in our scheduling of access to military land during training rotations. The project technical managers contributed to this paper with their patient understanding and technical oversight of this research. Dr. A. Frank Servello, formerly of the U.S. Army Corps of Engineers (USACE), Fort Worth office, served as project tech-

BRI Bulletin nical manager for the greater part of the period during which field work was conducted. In addition to his normal duties, he contributed an avid interest in this research. Stephen Austin of the USACE, Fort Worth office, following Servello's departure and our completion of most of the field work, served as project technical manager for the remainder of the period of performance. Faculty, staff, and students at Louisiana State University (LSU) also assisted us in our research. David Hinds, a graduated master's student from the Department of Geology and Geophysics, provided fruitful consultation on the results of his own geological investigations. Dr. Judith Schiebout, associate curator, Museum of Natural Science, contributed helpful discussion and support on many occasions. John Anderson, who directs the Cartographic Information Center administered by the LSU Department of Geography and Anthropology, provided invaluable access to and assistance with assorted maps and images in that archive. Several U.S. Geological Survey (USGS) personnel provided information and assistance in the form of consultation, file data, and copies of USGS reports. Mark Gremillion of the Fort Polk office, and John Lovelace, Larry Prakken, Rob Fendick, Roland Tollett, Darlene Smothers, and Wendy Lovelace of the Water Resources Division, Baton Rouge district office, assisted us with access to subsurface data in the USGS files. M. Earl Stewart, wildlife biologist with the U.S. Department of Agriculture, provided access to aerial photos and maps in the files of the National Forest Service, Vernon Ranger District, Kisatchie National Forest. The following employees of timber companies were active in the study area and helped provide critical access to large tracts of nonmilitary land: Robert H. Crosby of Crosby Land and Resources, Mandeville, Louisiana; Darwin Foster of Temple Inland Inc., Diboll, Texas; Mike Hudson of the Temple Inland Inc. office in De Ridder, Louisiana, who gave a positive account of our work to Darwin Foster and helped us obtain permission for access; Rick Leeper, forester with the Temple Inland Inc., De Ridder office, who provided access to gated properties; and Bob Nolan of the Boise Cascade office in Provencal, Louisiana, who gave us permission and

71

keys to access gated properties. For producing the illustrations in-house at the Louisiana Geological Survey (LGS), we are indebted to the staff of the LGS Cartographic Section under the direction of John Snead, cartographic manager. Lisa Pond, research associate, produced most of the illustrations for this paper. Production and assistance with the production of a number of the illustrations was provided by Robert Paulsell, research associate; Edward Koch, research associate; Edwin B. "Bud" Millet, cartographic supervisor; and David W. Griffin, research associate. Margo Olinde, LGS editor, provided helpful consultation regarding the editing and formatting of this paper. Finally, acknowledgment is made to two geologists who have had a positive shaping influence on this research through their previous investigations and their helpful conversations about them over a period of 15 years: H. V. Andersen (Professor Emeritus, LSU Department of Geology and Geophysics), who mapped the geology of Sabine and Natchitoches parishes; and James E. Rogers (USGS, retired, now a consulting ground-water hydrologist), who has investigated the surface and shallow subsurface geology in Vernon Parish and many other parts of the state.

Bibliography American Association of Petroleum Geologists, 1990, Gulf coast regional cross section. Tulsa, Oklahoma: American Association of Petroleum Geologists. Three oversized sheets. Andersen, H.V., 1960, Geology of Sabine Parish: Louisiana Department of Conservation, Geological Bulletin No. 34, 164 p. Andersen, H.V., 1993, Geology of Natchitoches Parish: Louisiana Department of Conservation, Geological Bulletin No. 44, 22 7 p. Aronow, S. 1982, Surface geology, in C. L.Neitsch, ed., Soil Survey of Jasper and Newton Counties, Texas: U.S. Department of Agriculture, Soil Conservation Service, Washington, D.C. p. 102107. Autin, WJ., 1993, Quaternary geology of the lower Red River Valley, in W.J. Autin and C.E. Pearson, eds., Quaternary Geology and Geoarchaeology of the Lower Red River Valley: A Field Trip: South Central Cell Friends of the Pleistocene, Baton Rouge, Louisiana. p. 5 24.

72

BRI Bulletin

Autin, W.J., Burns, S.F., Miller, B.J., Saucier, R.T., and Snead, J.I., 1991, Quaternary geology of the Lower Mississippi Valley, Chapter 18, in R.B. Morrison, ed., Quaternary Nonglacial Geology: Conterminous U.S.: The Geology of North An1erica, v. K 2, Geological Society of America, Boulder, Colorado. p. 547-582. Bernard, H.A. 1950. Quaternary Geology of Southeast Texas: unpublished Ph.D. dissertation, Department of Geology, Louisiana State University, Baton Rouge. 165 p. Bernard, H.A., and LeBlanc, R.J., 1965, Resume of the Quaternary geology of the northwestern Gulf of Mexico province, in H.E. Wright, Jr., and D.G. Frey, eds., The Quaternary of the United States: Princeton University Press, Princeton, New Jersey. p. 137-224. Bianchi, T.H., 1982, Geomorphic setting in the Fort Polk region: an abstract. Section 3 in A.F. Servello, ed., U.S.L. Fort Polk Archaeological Survey and Cultural Resources Management Program: unpublished report prepared for U.S. Anny Corps of Engineers, Fort Worth district, Texas, under contract no. DACA63-76-C-0253, p. 71-76. Cronin, T.M., Willard, D.A., Dowsett, H.J., Ishrnan, S.E., Holtz, T.R., Jr., and Liddicoat, J.C., 1993, The Yorktown Formation of Virginia: implications for late Pliocene climate and sea level history: Geological Society of America, Abstracts with Programs, v. 25, no. 6, p. 334. Dubar, J. R., Ewing, T.E., Lundelius, E.L., Jr., Otvos, E.G., and Winker, C.D., 1991, Quaternary geology of the Gulf of Mexico Coastal Plain. Chapter 19 in R.B. Morrison, ed., Quaternary Nonglacial Geology: Conterminous U.S.: The Geology of North America, v. K 2. Geological Society of America, Boulder, Colorado. p. 583-610. Doering, J., 1935, Post-Fleming surface fonnations of southeast Texas and south Louisiana: Bulletin of the American Association of Petroleum Geologists, v. 19, no. 5, p. 651-688. Doering, J., 1956, Review of Quaternary surface formations of Gulf Coast region: Bulletin of the American Association of Petroleum Geologists, v. 40, no. 8, p. 1816 1862.

Fisk, H.N., 1938, Geology of Grant and La Salle parishes: Louisiana Department of Conservation, Geological Bulletin No. 10, 246 p. Fisk, H.N., 1939, Igneous and metamorphic rock from Pleistocene gravels of central Louisiana: Journal of Sedimentary Petrology, v. 9, no. 1, p. 20-27. Fisk, H.N., 1940, Geology of Avoyelles and Rapides Parishes: Louisiana Department of Conservation, Geological Bulletin No. 18, 240

p. Forney, L.B., 1950, The Willis Formation of the Texas Gulf Coast: Unpublished master's thesis, School of Arts and Sciences, University of Houston, Houston, Texas. 39 p. Groot, J.J., 1991, Palynological evidence for Late Miocene, Pliocene, Early Pleistocene climate changes in the middle U.S. Atlantic coastal plain: Quaternary Science Reviews. v. 10, no. 2 3, p. 147-162. Guillory, C.M., 1997, Soil Survey of Sabine Parish: U.S. Department of Agriculture, Natural Resources Conservation Service, Washington, D. C. 194 p. Heinrich, P.V. and R.P. McCulloh, 1999, Pliocene Series, p. 43-69 in R.P McCulloh and P.V. Heinrich, Geology of the Fort Polk region, Sabine, Natchitoches, and Vernon Parishes, Louisiana. Final report prepared for the U.S. Army Corps of Engineers, Forth Worth district, under contract no. DACA63-95-D-0051, delivery order no. 0008. Hinds, D.J., 1997, A report to accompany the geologic map, Fort Polk area, western Louisiana, scale 1:24,000: Unpublished report, Department of Geology and Geophysics, Louisiana State University, Baton Rouge. 49 p. plus plates (includes one 1 :24,000-scale geologic map). Hinds, D.J., 1998a, Neogene Stratigraphy and Depositional Environments of the Fort Polk and Slagle Area, Western Louisiana: Unpublished master's thesis, Department of Geology and Geophysics, Louisiana State University, Baton Rouge, Louisiana. 100 p. plus plates (includes two geologic maps at an approximate scale of 1:53,333).

BRI Bulletin

Hinds, D.]., 1998b, Geologic map of the Fort Polk area, Vernon Parish, Louisiana. Geologic quadrangle map no. GQ-1: Produced in cooperation with U.S. Geological Survey, EDMAP program, under assistance award No. 1434-HQ-96-AG01535. Scale 1:24,000. Louisiana Geological Survey, Baton Rouge, Louisiana. Hinds, D.]., 1998c, Geologic map of the Slagle area, Vernon Parish, Louisiana. Geologic quadrangle map no. GQ-2: Produced in cooperation with U.S. Geological Survey, EDMAP program, under assistance award No. 1434-HQ-97-AG01753. Scale 1:24,000. Louisiana Geological Survey, Baton Rouge, Louisiana. Hinds, D.]., 1999, Neogene stratigraphy and depositional environments of the Fort Polk and Slagle areas of western Louisiana: Report of Investigations No. 99-01. Louisiana Geological Survey, Baton Rouge, Louisiana. 60 p. Hilgard, E.W., 1860, Report on the Geology and Agriculture of the State of Mississippi: (publisher unknown) Jackson, Mississippi. 391 p. Hulbert, R.C., Jr., and Whitmore, F. C., Jr., 1997, Early Hemphillian mammals from the Mauvilla local fauna, Alabama: Journal of Vertebrate Paleontology. v. 17, no. 3, p. 54 Huddlestun, P.F., 1988, A revision of the lithostratigraphic units of the coastal plain of Georgia: the Miocene through Holocene: Bulletin No. 104. Georgia Geological Survey, Atlanta, Georgia. 162 p. Kukla, G.]., and Opdyke, N.D., 1972, American glacial stages in paleomagnetic time scale: Geological Society of America Abstracts with Programs. v. 4, p. 569-570. Lenzer, ]., 1982, Geomorphology of the Fort Polk Military Reservation. Appendix 4, in P.M. Thomas, S. Shellyey, L.J. Campbell, M.T. Swanson, C.S. Weed, and J.P. Lenzer, Cultural Resources Investigations at the Fort Polk Military Reservation, Vernon, Sabine, and Natchitoches Parishes, Louisiana: New World Research Technical Report no. 69. Unpublished report prepared by New World Research, Fort Walton Beach, Florida, for National Park Service under contract no. C-54052(80). Atlanta, Georgia: National Park Service. p. 41 to 449.

73

Louisiana Geological Survey, compiler, 1993, Alexandria 1:250,000 geologic quadrangle: Unpublished map plus explanation and notes. Prepared in cooperation with U.S. Geological Survey, COGEOMAP program, under cooperative agreement no. 14-08-0001-A0878. Scale 1:250,000. Martin, P.G., Butler, C.L., Scott, E., Lyles, J.E., Mariano, M., Ragus, ]., Mason, P., and Schoelerman, L., 1990, Soil Survey of Natchitoches Parish: U.S. Department of Agriculture, Soil Conservation Service, Washington, D.C., 192 p. Matson, G.C., 1916, Pliocene Citronelle Formation of the Gulf Coastal Plain: U.S. Geological Survey Professional Paper 98. U.S. Government Printing Office, Washington, D.C., p. 167-192. McCulloh, R.P., and Heinrich, P.V., 1999, Geology of the Fort Polk region, Sabine, Natchitoches, and Vernon Parishes, Louisiana: Final report prepared for U.S. Army Corps of Engineers, Fort Worth District under contract no. DACA63-95-D-0051, delivery order no. 0008. 119 p. plus plates and appendices (includes ten 1 :24, 000-scale geologic rna ps). McGee, W.J., 1891, The Lafayette Formation: U.S. Geological Survey Annual Report 12, Part 1, U.S. Government Printing Office, Washington, D.C. p. 347-521. Morton, R.A., Jirik, L.A., and Galloway, W.E., 1988, Middle-upper Miocene depositional sequences of the Texas Coastal Plain and continental shelf: geologic framework and hydrocarbon plays: Report of investigations no. 174. University of Texas Bureau of Economic Geology, Austin, Texas. 40 p. Muse, B., and Cooley, M., Soil Survey of Vernon Parish, 1999: U.S. Department of Agriculture, Natural Resources Conservation Service, Washington, D.C. (in press). Otvos, E.G., 1997, Northeastern Gulf coastal plain revisited: Neogene and Quaternary units and events-old and new concepts. Guidebook, Louisiana to northwest Florida field trip, October 18-19, 1997. New Orleans: New Orleans Geological Society, New Orleans, Louisiana. 143 p.

74

BRI Bulletin

Otvos, E.G., 1998, Citronelle Formation, northeast Gulf coastal plain-stratigraphic and age issues: American Association of Petroleum Geologists Bulletin. v. 82, no. 9, p. 1789. Paleo-Data, Inc., 1993, Gulf of Mexico chronostratigraphic correlation chart. New Orleans: Schlumberger and Geco-Prakla. 1 sheet. Raymond, D.E., 1985, Depositional sequences in the Pensacola Clay (Miocene) of Alabama: Bulletin no. 114, Alabama Geological Survey, Tuscaloosa, 87 p. Rogers, J.E., and Calandro, A.J., 1965, Water resources of Vernon Parish, Louisiana: Water Resources Bulletin No. 6., Louisiana Geological Survey, Baton Rouge, Louisiana. 104 p. plus plates (includes one 1:126,720-scale geologic map). Schiebout, J.A., 1994, Fossil vertebrates from the Castor Creek Member, Fleming Formation, western Louisiana: Transactions of the Gulf Coast Association of Geological Societies. v. 44, pp. 675-680. Schiebout, J.A., 1997, Paleofaunal survey, collecting, processing, and documentation at two locations on Fort Polk, Louisiana: Unpublished report prepared for U.S. Army Corps of Engineers, Fort Worth district, under contract no. DACW63-90-D-0008, delivery order no. 13. U.S. Army Corps of Engineers, Fort Worth, Texas, 92 p. Snead, J.I., and McCulloh, R.P. (compilers), 1984, Geologic 1nap of Louisiana: Louisiana Geological Survey, Baton Rouge, Louisiana, scale 1:500,000. Snead, J.I., Heinrich, P.V., and McCulloh, R.P. (compilers), 1995, De Ridder [Louisiana portion] 30 X 60 minute geologic quadrangle (preliminary): Unpublished map plus 16-p. expanded explanation and notes. Prepared in cooperation with U.S. Geological Survey, STATEMAP program, under cooperative agreement no. 1434-HQ-94-A-01233. Louisiana Geological Survey, Baton Rouge, Louisiana, scale 1:100,000.

Snead, J.I., Heinrich, P.V., McCulloh, R.P., and Autin, W.J. (compilers), 1998, Quaternary geologic map of Louisiana: Unpublished map plus 12 p. expanded explanation and notes. Prepared in cooperation with U.S. Geological Survey, STATEMAP program, under cooperative agreement no. 1434-HQ-97-AG-01812. Louisiana Geological Survey, Baton Rouge, Louisiana, scale 1:500,000. Thornbury, W.D., 1965, Regional Geomorphology of the United States: John Wiley and Sons, Inc., New York. 609 p. Welch, R.N., 1942, Geology of Vernon Parish: Louisiana Department of Conservation, Geological Bulletin No. 22, 90 p. Winker, C.D., 1979, Late Pleistocene FluvialDeltaic Deposition on the Texas Coastal Plain and Shelf: unpublished master's thesis, Department of Geology, University of Texas, Austin, 187 p. Winket; C.D., (compiler) 1991, Quaternary geology, northwestern Gulf Coast, Plate 8 in R.B. Morrison, ed., Quaternary nonglacial geology: Conterminous U.S. The Geology of North America, v. K-2. The Geological Society of America, Boulder, Colorado, scale 1:2,000,000. Wornardt, W.W., Jr., and Vail, P.R., 1991, Revision of the Plio-Pleistocene cycles and their application to sequence stratigraphy and shelf and slope sediments in the Gulf of Mexico: Transactions of the Gulf Coast Association of Geological Societies. v. 41, p. 719-744. Woodburne, M.O., and Swisher, C.C., III, 1995, Land mammal high-resolution geochronology, intercontinental overland dispersals, sea level, climate, and vicariance, in W.A. Berggren, D.V. Kent, and M.P. Aubry, eds., Geochronology, Time Scales and Global Stratigraphic Correlation: Unified Temporal Framework for an Historical Geology: Special Publication No. 54. Society of Economic Paleontologists and Mineralogists (Society for Sedimentary Geology), Tulsa, Oklahoma. p. 335-364. Woodward, T.P., and Gueno, A.J., 1941, The sand and gravel deposits of Louisiana: Louisiana Department of Conservation, Geological Bulletin No. 19, 429 p.

Basin Research Institute

Bulletin A publication of the Louisiana Geological Survey Special Millennium Issue 1999-2000 Louisiana Geological Survey Chacko J. John, Director and State Geologist Board of Advisers Frank W. Harrison, Jr., Chair Senator Max T. Malone Terry Ryder Karen Gautreaux James M. Coleman James E. Rogers William E. Marsalis The BRI Bulletin Staff Editorial Science Director/Chacko J. John Bulletin Editor/Margo Olinde Production Manager/John 1. Snead Graphic DesignslEdward Koch & Lisa Pond Contributing Writers Ron Zimmerman John Echols Rick McCulloh Paul V. Heinrich Technical SupportlReed Bourgeois

Distribution Cherri Webre Telephone: (225) 388-8328 Fax: (225) 388-3662 Web Site, http://www.lgs.lsu.edu The BRI Bulletin is distributed to academic, industrial, and professional persons and groups associated with geological research and applications. Back issues may be ordered by contacting the BRI business office.

Volume 9 • June 2000 Contents Predicting A North Louisiana Jurassic Pinnacle Reef Trend From Palinspastic Reconstructions Ron Zimmerman ................. .... ... ....... .. ............ 1 Coalbed Methane: Louisiana's Unexplored Energy Resource John B. Echols ............ ................................... .. 18 A Summary Of The Total Oil Generation Potential Of Louisiana's Hydrocarbon Systems Ron Zimmerman ......... ............. ... .. ..... .. .... ....... 28 De Quincy Fault-Line Scarp, Beauregard and Calcasieu Parishes, Louisiana Paul V. Heinrich .... ........................... ..... ..... ..... 38 Pliocene Surface Stratigraphy in the Fort Polk Region: Implications for Louisiana Surface Geology Paul V. Heinrich and Richard P. McCulloh ...... 51

The Louisiana Geological Survey grants permission for fair, nor-forprofir use of SRI Bullerin contenrs. The SRI Bulletin is published yearly by LGS, whose offices are locared ar Louisiana Srare University, Room 208, Howe-Russell Geoscience Complex, Saton Rouge, LA 70803-4101.

Geological data and inrerprerations shown in this bullerin have been carefully compiled and interprered by rhe Louisiana Geological Survey of Louisiana Srare University. The interpretations represent our best judgement, however, we cannor and do nor guarantee the accuracy or correctness of any geological or other information.