You are on page 1of 63

.• U.S.

. National Technicallnform.lion Service

...•. AD;'A033 ~ 42





. MAy 1970

o , -:.'-
. ,-"_: '-, '-. .-. - - -


__ ", ~;t..-- SPRIfIGFIflO, VA. Z2161. ". __

. 'Kt$\.,~,.x I~mt.l< ""X":<


. prepared by U. S. Army Engineer W.terw.ys Experiment St.tion, Vicksburg, Mississippi

'"-. "
~ < ' •

This document has been approved for public felease and sale; Its distribution Is unlimited
, -, - ':""'.


by 0

R. T. Soucier, A. R. Flee~wood

. ., ....

~- -: - .' -- '

W· o

. ',no 00··

May 1970

Prepared by U. S. Army Engbeer W~lerways Experiment Station, Vicksburg, _Mississippi


This dot'Jment has been approved for public releare and sale; lis dis(rlbul/on Is unUmliej
I 0.3
, W? '-\ ,.,..-..
No- &,.'lO-I~-

?'his paper was prepared by Dr .R. T. Saucier and ~!r. A. R. Fleetwood

of the G<!ol.ogy Branch, Soils Division, U. S. Army Eng~neer Waterways Exper-
iment Station (ImS), for publication in the Bulletin of the Geological
Society of America. Observations instrmnental. in the formulation of the
,concepts and concluid.onsiiresented'iii this' paper were made during large-
scale engineering-g~ologic mapping conducted by the WES Geology
Branch for the U. S'. Army Engineer DistrIct, Vicksburg. ?'he paper was
reviewed and approved by the Office, Chief of Engineers •
Director of WES ,durin" preparation of this paper was COL Levi A. Brown.
Technical Director was Mr'. F. R. Brown.

. ",

•• 70027
.Abstract: Efforts to differentiate Quaternary alluvial deposits along the

Ouachita River between Camden, Arkansas) and Monroe, Louisiana, accordinf5: to

environments of deposition have indicatcd the presence of five fluvial ter-

races, three of which are considered to comprise the De\1eyvillc terrace

sequence •. This sequence, lying stratignphically between the Prairi" ter-

raCe and the Holocene floodplain, can be distinguished partly by large

. meander scars that ere characteristic of the terrace ;rhere it is present on

several Coastal Plain streams. Previous itIVestigations have established

the age of the Doweyville terraces as being "ctween 13,000 and 30,000 years
before present; hOl·,ever, opinions vary regarding mode of origin. Horphologic

characteristics of and stratigraphic relationships bet;reen the terraces

along the Ouachita River indicate that a close causa:;' relationship existed

between the terraces and major episodes of aggradation and degradation in

the Hississippi alluvial valley, but changes in hydrhulic regimen brought

.. about by climatic change ,rere also significant.

The chronological sequence indicated for the On?chita Valley includes

(a) dep,'sltion of the Prairie terrace forr.tation during the Sangamon Inter-

glacial stage, (b) primary valley degradation and deep stream entrenchment
during 'laxing Early Vlisconsin glacIation, (c) alluvir.~ion during waning

Early I'iisconsin glaciation with maximuTn valley aggradation occurring during

the Fanr~alian Substage, (d) secondary valley Jceradation and moderate en-

trenchment during waxing Late ,/isconsin glaciation,and "(e) alluviation and i .-


valley aggradation dudng wanine Late Viisconsin glaciation and continuing

into the Holocene.

Rapid deposition of glacial oubrurh in the !.t'ssissippi Valley, includ-

ing the development of an alluvial cone by the Arkansas River, resulted in

the damming of the 10\;er end of the Ouachita Valley during the Farmdalian



, .'
,r, '

Substage. A 500- to 700-sq-mile perennial lake, called Lake Monroe, was

,t created by the danuning: shoreline features and lacustrine plain remnants
}-- of the lake comprise what is recognized' as the highes't, and intermediate

Deweyville terraces • The l(»rest De>leyville terrace >las created subsequently

>. during a period of >laxing glaciation and valley entrenchment and under the

"influence of a pluvial climate. Elnpirical equations suggest that stream

discharges were 6' or 7 times greater than present and that precipitation
, -

was more' than t>lice ,that of present during this period.

""" ; ", <

"- .. ' .-

1-- -
. I
I .. /

Introduotion • • • • • • • • • .' • • • • • • • 1

Background • • • • • • • • • • • • • 1 :

i Scope . 1

~ l-!ethodology 2
. 'i~

Acknowledgments ' . • 3' "

Geographic and Geologic Setting. 1

General stratigraphy and Nomenclature. • 6
Prairie and Montgomery Terraces 6 ~t
D""eyvillc Terrace. 7
i ., "-~"
Terrace JllOrpho1ogy and Lithology
· • • • • 8
!: . . '.~,)
Introduct ion. ' • • • 8 ,,'

:Prafrie Terrace • • 8
:i'j "

'Deweyville Terraces 1 and 2 11

• •
Deweyville Terrace 3. • • • • • •• 16 '~';ff
Holocene Floodplain.

Geologic History
• • • • 18

• • • •
Bttckground. • • , • 21

Interpretation. • • • 23 .' ~

Sengamon Interglacial stage. . 23


E\J.r1y Hisconsin glaciation
.' • • • 24 -1-
Farmdalian Substage, • ., 26
Late Vlisconsin glaciation. • 29
I-Ioodfordian Substage • 30
The Recent , ,. • 32
Discussion • • • • • • • • 33
Extinct Lakes • • • • 33
/lge and ():>.'igin of De;teyvllle Terraces • • • • • • • • 35
Deweyville Paleoclimatology and PaleohJ~rology.
• • • • • 37
References Cited • . ,-. • • ·. ", • • 46

.:, .'



. Figures:

1.' . 'Location of study area and relationship to the Lower 14ississippi

... ...
. - -.-

'.' .' _. ....

• • • • • • ;'. ~ . . --_: • . : ; '. - .', ~ . t·
. -. , .. 5 -. '-"'<

2. ,Reconstructed do,mvalley (longitudinal) terrace and floodplain

... ... -.- ..... . . . .


, ' ,
~ . .........'.-... .
. -', .
Morphology' of a typical J.ak'e beach or barrier and adjacent surfaces
located about 7 miles southl1est of Monroe. La •• • • • • • • • • • 13
4; &tent and configuration
, of Lake Monroe and associated features • • 14

t Lol1er Hississippi Valley chronological concepts (a) as proposed by
~. -
, ,

"1- Fisk (1944). ",nd (b) as applied by Saucier (i968) to sealevel

:fluctuation curve proposed by Curray (1965) • • • • • • • • • • •

Diagranunatic illustrations of stages in the development of the
.,. OuachitaValley •• : •• • • • • • • • • • • • • .- . . . .- . . .
EXtent· and conf:tgaration of the Far:n<!nlia;. Substage, Arkansas River'
,cone (I·lacon Ridge) as related to the formation of Lake Monroe • •

-,1. Quaternary geology. Ouachita River area, Arkansas and Louisiana • • 50
A comparison of De ;;eyville and Holocene meander belt features • • • 51


1. Holocene anJ De>leyville (Qtd stage) strea;n geometry measurements. 40

, 2. Holocene and Del1eyville (Qtd
stage) st,'eam discharges calculated
from empirical equat ions. • • • • . . . . . . . . . . . . . . . ..

- , '-
~ --'.-

. Background

The general stratigraphy of Quaternary depositional terraces of. both

fluvial anumarine origin in the central Gulf Coast area has been recog-

.nized for about 50 years. Nomenclature introduced by Fisk (1939) is most

widely accepted as is his interpretatlon of the origin of the tc'rraces in

relation to "axing and "aning glaciation and major sea level variations

(Fisk, 194Q). Alternate interpretations and nomenclature have been of-

fered (Doering, 1956); ho;rever, the most serious objections raised have

been in relation to the ages of terraces and their correlation with spe-

cific glacial or interglacial stages (Tro;,bridc;e, 195!1; Leighton and

Willman, 19119; Durham and others, 1967), particularly the older terraces •
. .
.Bin,e the;rork by Fisk, one of the fe" significant modifications to the

late Quaternary terrace stratigraphy has been the recognition by Bernard

(1950) and the realization by Gagliano and Thorn (1967) of the geographic
extent and chronologie significance of the De"eyville terrace in the Gulf

-and Atlantic coastal plain areas. The mode of origin and relationship of

this terrace to older terraces and to the Holocene floodplain of the

OuRchita River are of major concern in this paper.


Delineation of the Quaternary formations along the Ouachita River was

accomplished by FleetVlood during a three-year project conducted by the

U. S. Army lligineer Haterways E:<periment Station, Vicksburg, mss., for

the U. S. AI7IlY llieineer District, VIcksburg. The primary purpose of the

project ;,as to differentiate the QuatQrnary alluvial sequences according to

env~ronments of deposition md to assess tl.d engineering properties of the

sediments, both surficial and buri~d, in each environment. De~ailed litho-

logic and stratigraphie

. .
data developed in the project ~re available .~n report

form (Floet>lood, 1969) and are not presented in this pa·~e,..

The basic i:--tont of the ButhOl'3 is to point out a CBu"~l relationship

bet>leen the Quaternary terrace sequence of the Ouachita Vo',lley Hl,d major :

episodes of aggradation and degradation in the KLs>issippi alluvial valley.

Saucier (1968) has recently presente'l tentative evidence of a previously

unrecognized v/isconsinan stage cycle of vallcy cutting alYl filling in. the

1o>1or J.lississippi Valley area and has 0ffered a revlsed chrollclcgy: the

Ouachita Valley tcn'ace sequence can be sati.Rfactcrily explained only in

terms of the new chronolor,y and offers consi,derable additional substantiat-

ing evidence.

The secondary intent of this paper is to call. '\ttention to certain

rather unusual geomorphic features that, l.n one case, indicate a period

of >lell devclloped pluvial climate and, in another case, in:ii.cate ah ear- -- -.

lier !,0riod of' ,ride3prend '-alley flooding and Ihke formation related to

1o>ler Nississippl Yalley sedimentation patterns o:1d rates.


The extent of terrace mappine (Plate 1) is coincident

>lith the extent of available large-scale topoV-aphic mapping b:r the U. S.

Army Corps of Engineers and the U. S. Geological Survey. Initial terrace

identification on1 delineation 'ISS perforJOc.Q using contours >lhile cross-

valley ane] do'tm-vn,11ey profilen, Sh0'tlir.g both actunl and recon:Jtructed-

sur.face elevations, ;Iere used for modifications oy,d refinements. Final

delineations were achieved through an examination of the physiographic ex-

. pressions of. the terraces on aerial photOgl'aphs and photo mosaics.

Subsurface data in the form of logs of wells and borings were .used to

. establish terrace Htholo!\y anil permitted estimates of such parameters "s

depths of river scour and valley entrenchment. Field reconnaissances of the

study area proved to be generally unre>mrding because of the low· relief,

heavy vegetation cover, lack of access to many large areas, and infrequent

outcrops 'or road cuts.

In essence., this has been an intensive investigation of regional

physiography. Geomorphological investigation techniques proved to be the

most effective, and the geomorphic expression of landforms proved to be

the most significant factor in discerning terrace stratigraphy and infer-

ring modes of origin.


Special thanks are extended to Dr. Charles R. Kolb for his review of ..;
the mam,script and to both him and Mr. H. B. steinriede, Jr., for their

valuable critical discus.,ions of the problem that helped crystallize our

thinking. Nrs. Naomi J. Rhodman and ~!rs. Hary E. Skipworth assisted >lith
typing, and ~!r. J. A. Sherlock provided editorial review. Supervision by

Hr. Nels J. Nyman of the ext'msive drafting assignment is appreciated. Ap-

prova1 by the Office, Chief of Engineers, of the manuscript for publication

is acknO>lledged.


The Ouachita Piver flo'ds generally southeashlard over an airline dis-

'tance of about 300 miles from its hea:"raters in the Ouachita Mountains of

west-central Arkansas to where it becomes tributary to the Red rover in

east-central louisiana (Fig. 1). It meanders freely and has created 0.2-

to. 5-mile-Hide floodplain in the Coastal Plain beh/con Ho.lvcrn, ArK. '. and

sterlingten,' la. South ef sterlingten, the river enters th~J.!ississippi

alluvial valley and fle"s as an undcrfit stream in an abandened Helecene

meander belt ef the Arkallsas River (ng. 1 nnd Plate 1). .'

The Quaternary valley er'the Ouachita River is quite ~,ssymetrical

betHeen Camden, 'Ark., and ~,!onree, La. (Plato 1). Threughout mest ef this

distance, the present river is at the base or within 5 to. 10 miles of the

western valley wall "hich rises abruptly to. heights ef 100 to. 150 ft abeve

the floodplain. Upland fermat ions Hest ef the rivo" are efEocene age

(ClaiberneGreup) and are composed of thinly-bedded and mestly uninduratcd

clays, silts, and fine sands of shalle" marine or deltaic erigin. Ellst

ef the present river, ene must cre'·s a 20- to. 25-mile-wi(l,e, step-like

sequence ef low, flat terraces befere enceuntering pre- Quaternary ferma-

tions at similar elevations.

The Eocene formatiens strike reughly nerthenst-seuth"est, dip gently

to. the. ~outheast, and have been enly moderately affected structurally. At

least ene major fault zone cross",s the Ouachita Valley near Smackover, Ark.;

ho,tevcr, an almost complete absence of evidence of f"ulting in Q>l!lternary

deposits indicates general quiescence since Tertiary times. DifferentIal

eresien of the Eocene formatiens has resulted in a maturely dis&ected up-

land landscape ,~th faint suggestiens of cuestas.

The Ouachita River floodplajn north of Sterlington 1s subject to

annual flooding and consequently is uninhabited and heavily timbered '.-lith

deciduous hard'iJOods. Numerous shoals- and bars characterized the river at

t---= - - - - - - - - - - - - - - - - - - --r'--,§':,,,f, ,.



L 0 J~ ..~

GilL F
o! , I •

Figure 1. location of st\ldy are:l and relationship to the

lower l-lissis3i.ppi Valley.
low water in its natural state, but a series of small locks and dams now

provides a navigable ~hannel to Camden. The terraces are sparsely populated;

widespread pine forests support an extensive lumbering industry. S.outh of

sterlington, the higher and ''lider natural levees of the old ·Arkansas Hiver

courses are extensively cultivated and industrialized and well-populated,

. with Monroe being the largest urban con,plex.


Prairie and Hontgomery ,Terraces

The Prairie terrace formation (Fisk, 1939) is the most are ally exten"

dve Quaternary terrace in the Ouachita River area (Plate 1) and also the

most easily identified terrace in regard to surface morphology and relative

topographic position. Although the Prairie terrace Has originally assigned

by 'Fisk to the Peorian (Bradyan) Interglacial stage,several recent invPd-

tigations in the Gulf and Atlantic Coastal Plain areas have indicated evidence

- for the assignment oithe Prairie or it" equivalents (Beaumont of Texas and

Pamlico of Florida and Georgia) to the Sangamon Interglacial stage (Gagliano

and Thom, 1967; Hoyt and others, 1968; Schnable, 1966). Evidence for such

an ';Issignment is present in the LO\'lel' W.ssiss;ppi Valley (Saucier, 1968) and

also in the Ouachita Valley.

Maturely dissected remnants of a terrace situated topographically

higher than the Prairie occur in the extreme northeast part of the mapped
area. Little at-Gent ion has been devoted to this terrace; ho,rever, super-

ficially it appears to moet the crit~ria set forth by Fisk (1939) for the

Montgomery terrace.

. Deweyville Terrace, '

''''' The name De,reyville was first applied by Bernard (1950) to !\ low dep-

: ositional terrace along the lower Sabine River in Texas and Louisiana.
• .- • • M

,R\,cognition of the Deweyville (1S a distinct terr!\ce formation was based on

'1ts"intermediate position bctween the Prairie terrac~ .and the Holocene

floodplain and the presence of distinctive meander scaro several times'

, -larger in width and radii than those of the present river. Bernard con~

side,'ed a pluvial c;timatc expianation for the large meander features but

concluded that they were lUore likely attributable to a rapid rise it; sea
y' .

level in Late "is cons in times.

'. In ,this initial work on the Deweyville, Bernard recognized similar

,·t~rracesala'ng the Brazos, TrInity, and Pearl Rivers of the central Gulf
,: COast 'ar~a' and w~s aware of a possibly related terrace along the Ouachita

'River. "Somewhat' lat'er; Bernard and L,Blanc (1965) expressed an opinior"


8S have Gagliano and Thom (1967) and Saucier (1968),that the Deweyville

't,erl:'ace l'robably'correlates with braided stream surfaces in the Mississippi

~lluvial valley. Published (Pearson and others, 1966, p. 458) and unpub-

lished radiocarbon dates from along several OJlf Coast rivers suggest for-

mation of tho De."'yvi:.. ... e cerrace between about 13,000 and 30,000 years

before present.

As many as three levels or sub-terraces have been recognized (Gagliano

and Thom, 1967; Gagliano, 'personal cOliullunication) along the Sabine and

Trinity Rivers where large-scale topographic mapping is available and has

been examiried in detail. Along the Ouachita Rivor, three terrace levels

OCCur between the Prairie tenace and the Holocene floodplain. Although

7 "
only the lowest level, designated as the Qtd level (Plate 1), exhibits large
meander scars, all three levels are considered to comprise the De"eyville
. :,'.-

~terrace. Evideneeconsidered more conclusive than has heretofore been avail-

able will be presented later in this paper to demonstrate a close relation-

ship of all three terrace levels to the Mississippi alluvial valley braided

surfaces and to further substantiate the suggested time span.



Each Quaternary terrace formation, including the Holocenc alluvium,

is a depositional sequence consisting of a fine-grained to~stratum (clays

and silts) and a coarse_grained substratum (sands and grf.vels). The

topstratum-substratum thickness ratio of about 1: 2 or 1:3 remains rather

. uniform throughout the area except on the older terraces where portions of

the topstratum have been removed by erosion. Topstrat1JJll deposits primarily

represent overbank deposition in natural levee, backswamp (flood-basin),~ and

abandoned channel or course envirorunents, ~lhereas substratum deposits -were

laid dO>n! in point bar, channel lag, and possibly braided-channel bar en-

virorunents (Allen, 1965; ;"leehrood, 1969). The only significant exception

to this occurs in relation to thc dO'n!stream portions of the Qtd l and QtO'2

terraces where much of the topstratum represents' deposii;ion in" a lncustrine

environment. /. -.I
~ .
. ,",;;
Prairie Terrace

Interfluve areas of the Prairie terrace formation are,!ufficiently

~rell preserved to permit reconstruction of former elevations and slopes,

tlltho'.lgh differential erosion has reduced the once neurly flat relict

floodplain toa gently rolling to moderately hilly landscape. As indi-

. - _.

eated'in Figu;e' 2, th:' formation ex'nibits a reconstructed dOWllvalleY,slope

·I• .
·.that· waxc;; southw,,,'" !'rom about 0.33, f't/mile above Fel.senthal, Ark., to

over 0.75 f't/milebelow Honroe. The convex slope. profile is quite similar

r·~~: " to that of .the ·present river at natural low water; the two remain seParated
~:" "
bY about 110 f't over a distance of about 90 miles.
. .

t . Subsurface data indicate that the Prairie' terrace formation was de-

posiLt':a oa utl eL'o"lollal surface on Claiborne Group deposits that was a1-

most certainly formed as a result of ent.renchment during a preceding glacial

'stage, probably the Iilinoian. Local relief on the erosior! surface appears

"',to ·"e.les8 than 40 f't •. In view of a lack of any indications of soil pro-

file de.velcpment or subaerial ;reathering characteristics, it is assumed

that erosion took place 'oeneath an alluvial COVer as a result of lateral

and vertical stream corrasion •

. '. - . . . - . .

'. :Poss!ble' di.ffer'mtlation of the Prairie terrace formation according

. to parent stream is stlsgested by a distinct variation in the thickness

of the terrace formation. All of the formation in Arkansas and that por-

tion located ;;est of the Ouachita River in Louisiana (Plate 1) has a thick-

n~ss varying !'rom about 60 to 75 f't and was undoubtedly deposited by an


ancestral Ouachita River. The portion of the terrace formation situated

east of the river in Louisia~a, which includes the Bastrop Hills, has an

average thicYJless of about 140 ft, suggesting deposition by the larger an-

cestral l,lississippi River.

Even YThel'e essentially undissected, the Prairie terrace forma.tion has

/::"?:~y:~~,l',~~~~" ,·,~,!i,:~I_~~r:!~~.~l'':':"'""''~'''''''-':·~''''''"''''w


zoo ICAJ.I.OtN

--, ,
====f rtrJCNTHAt.
'I I
1 _ _-£PR.#IRIl: (Of,} TERR.#Ct:

J"" '- •••~

. -
-. - -- c:.. ,
__ ~_;.- '''J TERRACE
..... "'"-- Q

. ~

... tOO
--..:.__ ___
-...L_ ---...
_~ .c.f"'~ TERRACE I
-~---t-~---- 1 - -_ _
--r--oI' I--

o :. 7~

: . . .-..-.. ---J
It T----l--_
NA1'VRAL (.ow WA1'~oR PIWF'I(.E ----.. __
I I OF OUAChITA RlYeA II" -....-...... I
~r 1 '-"",

Ol ! I !
o zo ",0 ~ a<;l ,00 I~
AP'I'f!O;I(I""''I'C t.ll!.tAC.t Al.O~ loIN! C'"' StCTION ((;.l!1IIC",,o,!.IZto IIttVtI\ COUft5.C)

Figure 2. Reconstructed downva.lley (longitudinal) terrace and

floodplain profiles.


.","'~'" ....... ' ~~~"'-"~'~'",~:''':.o.;.",;,...~,~~.~... (.:~~~.:..rZ3~~~;\-:;';';-:::iJ:ii~~:;,~.:~,'.:~~~~.J<",·,~~l~.:.:~i,~~~;.>i:.Ui~:'~';::;;;:,~",;';~;,~-.i::~">ij.:';'--';i'';r'';;:~''';~;':ili:~~V.~:W$~

a completely amorphous surface expression. Relict channel scars are to-
, '

tally lacking; hence ,no indication of former stream patterns or dimensions

can be obtained.

Deweyville Terraces 1 ani 2

The highest DeHeyvilleterrace formation, the Qtd terrace, is wide-

spread north of the Ouachita River in Arkansas and occurs at scattered ,
", localities on bota sides of the river in Louisiana. ' It is only slightly

dissected a.q compared llith the Prairie terrace fOi'lllation. Where large con-

tinguous areas exist, such as along Horo Creek and in the Flatwods terrace

'area (Plate 1), ,interfluves exhibit relief of less than J ft over areas of
several square miles. The intel'lllediate De'18yville terrace formation, desig-

nat~d as theQtd terrace, occurs along the 10>ler Saline Rival' in Arkansas
'and aiong the Ouachita River essentially orily
. south of the Saline River.
, ,.- -". ..
: ,-.

'Surfacerelief on this terrace is comparable to that on the Qtd terrace:

" Upstream from Felsenthal, the Qtd terrace has a rather uniform dO>ll1-
. :.'~ - - . l
valleynlope of about 0.50 ft/mile (Fig. 2). Ho,rever, near Felsenthal th~

slope abruptlY,declines to less than 0.1 ft/mile, and south of Alabama

Landing, the surface slope is zero. The Qtd terrace exhibits a similar
decline in do,mvalley slope and it also finally attains a zero slope south

of Monroe.
This decline in slope was an enig~a to the authors for an appreciable

period during the study, parti'cularly ,rnen considered in comparison ,lith

the slopes o~ the other terraces and ths Holocene floodplain. The only

possibly tenable explanation for this ;tas that the Ouachita River during

this time was shallo,ring its gra!lient and agerading its floodplain in 1'0-

sponse to rapid alluviation ill the Miss;'ssippi alluvial valley.

Considerable evidence has subsequently been discerned that supports this

hypothesis, by far the most significant evidence baing in tha form of a

series of 10\, ridges such as the one illustrated in Fi[,'ure 3. These:

ridges, delinentedin red in Plate 1, occur on both sides of the Ouachita

River south of the Arkansas-Louisiana state line. Regardless of length

or . location, they all have a relatively constant width of 1000 to 2OO0ft,

they have conspicuously accordant crest elevations varying only from about

103 to 105 ft msl,~hey stand at least 10 ft and usually ro ft. above the

adjacent terrace levels, they al,~ys occur on the Qtd

terrace at its con-
tact with the Qtd terrace, and, with but one exception, drainage is de-
veloped inllnediately adjacent and parallel to the ridges on the higher ter-

. race side. Profiles nonnal to the ridges all sho>! relative elevation dif-

i'erences and slopes simi}-ar to those included in Figure 3.

rt· has been concluded that these features arc beach ridges that formed
.. around a large possibly seasonal but more likely perennial lake for >!hich

the authors intrClluce the name Lake 140n1'oe. The pro'oable extent of the lake

:. is indicated in Figure II.

The exceptional flatness, both local and regional, of the main extent

of the Qtd terrace belo>! the state line and its position relative to the
beach ridges suggest that it "as a backbeach or backbarrier flat that >Ins

subject to inundation only during extremely high lake staees. The Qtd
terrace is interpreted as being the relict lake bottom or lacustrine plain.

Baline River is considered to mark the approximate northernmost extent of

the lake; the sloping Qtd terrace north of this point c.long the Ouachita .1
is the relict floodplain of the anoestral river before it discha,ged into

the lake. M~ander scars alYl occasion'al abandonf..:d channels ~.:irllilar :in size


____ '. Qfd l T~RRACE


~ ~ ______~ I Qld,2TEMACE 8'

cll::~_------------~~-- ________________~l ,

Figure 3. !.!Orphology of a typical lake beach cr barrier and

adjacent surfaces located about 7 miles soutlnTest of I·:onroo, La.


.:'...-., ','J




1£II(H£$ M"l." ....JUI.;{R

1I'&'c0(8tI-C1I OR II"-Co(·

,1\000",(\I.. [ (U[tfT or

STATUTE "'<t.U o LA"'! "'ON~[ AT NOII",,lL

.~X3.t >II

Figure 4. Extent and confiBuration of

Lake lolonroe nnd associated features.


to those of the present river are discernible on this terrace in the }.lora

Bay area, >lhile the backbeach or backbnrrier flat portions of the twrace

have a complctely amorphous surface expression.

At normal stage, extinct Lake Nonroe is estim"tcd to have had an area

of about 500 sq miles (Fig. 4) and 11 mean maximum depth of 20 to 25 rt. The

area of the lake may have been as much as 700 sq miles at extremely high

stages; hO;Tever, the dcpth Hould have been only 5 to 10 rtgre,atcr. These

are~ determinations ,are based on the assumption that the lake did not ex~end

appreciably beyond a line draHn connecting the southernmost extent of the

Qtd terrace on the east and >lest sides of the Ouachita River (Plate 1).
The lithology of thc Q"d and Qtd terraces va1'1_es little from the
l 2
characit"ristic fine-grained topstratum/coarse-grained substratum sequence

previously mentioned,.- The fact that the sequence is present to the southern-

most extent of the terraces, including the area occupied by Lake Nonroe, in-

dicates a l'luvial orlgin for most of the terrace sediments in that area. , ,
Probable lacustrine and backbesch-flat sediments apparently overlie' the

fluvial sequence as a thin veneer and have been observed at only tHO loca-

tions: 'in the FlatHoods Terrace area, >There the Qtd terrace topstratum is
unusually thick and contains large quantities of silt-size material, and in

,an area about 6 miles south;rest of J.lonroc, >There the .wne terrace exhibits

extensive deposits of fine sands and silts in the topstratum. Samples from

these bl0 areas were analyzed, but no sedimentologie or paleontologic evi-

dence was discerned that confirmed deposition in a lacustrine or lacustl'ine- ' ;

related environment.
The Qtd terrace is er:)sional rather than depositional in at lcaot
one area Bouth:'lCst of ll.onroe. All terraeemorpholot>ic criteria are


- -------
· satisfied; ho;rever, outcropping sedirnents a:~e the ironotone-bearing, finely-

b.edded clays and silts of the Cockfield formation of the Claiborne Group

.. (FleehfOod, 1969). It is asswned that this occurrence is attributab.le to

·"subaqueous planation by nearshore currents in. Lake Monroe.

Jurlging from a fOil scattered r"wI cuts, aggregate pits, and borings,

.. .
. .
beach ridges appear to bc composed prlInarilyof irregularly stratified

· sands and gravels undoubtedly reworked from nearby fluvial terrace d"posits.

The only observed e,xceptIon to this is the 8-mile-1ong, north-south trend-

· ing ridge called Praid.e de Butte located north of Bastrop, 1M. (Plate 1).

· A exposure in a canal bank cutting throuGh a portion of thIs

ridge sho',ls only massive to faintly-stratified, well-sorted f:ne sands.

· The 11.tho10gy, morphology, and 10~ntion of Prairie de Butte sUiSse at it was

a barrier spit that pm'Ually enclosed an arm of Lake ).lonroe extendins up

· the floodplain Of Chetni n- ,,-Haut Creek. Bayou Bartholomew in this area 1s

afaj,r1y recent occurrence not related to the lake.

Deweyville Terrace 3

The lowest of the Deweyville terraces is intermittently present on

both sides of the river from at. least 12 mil es north of Cwr.den downstream .

to 12 miles south of the Arkansas-Louisiana state line (Plate 1). The

'do,mvalley slope profile (Fig. 3) is slightly concave, the slope dec~eas­

ing from about O.60ft/mile above Catrflen to about O. ?(:. ft/mile just north ~..
of the do;mstream-most point. South of thh peire, thc Qtd terraee for- I,
mation undoubtedly is present at 1eMt intennittcntly in the subsurfacc,
being buried by the Holocene natural levees alone abandoned Arkansas River
Courses such 8.3 the Bayou BartholoDl'3~1 course.

Disqectl.on of the Qtd terracc in comparable in maenitude to that of

the Qtd terrace in Arkansas; however, there is one majordif'f'erence.
Drainage patte,-"s on the 'itd terraLc no;; ref'l"ctlittle or nothing of
the control or inf'luence once exerted by primary depositional featu;:es of'

. the former floodplain, "hereas on thb Qtd tertace, the control of the
relict features is strikingly evident •

.. Rel~ct features present on the 'itd terrace include numerous aban-

doned" channels·, fn>:~:-;~ents 0:[' abe.ndonod courses, arut extensive areas of

. point bar ridge and. s,,":le topography (meander scrolls). These features

are the most outstanding and diagnostic characteristics of the terrace

and are present throughout the mapped extent of the terrace.

Too aba!Jdoned channels and related point bar topography on the Qtd
terrace betJ.lCen Calion, Ark., and I~cro Bay are visible on the photomosaic

·-in ·Plate 2. These featUl'es illustrate ;;ell the significant size differ-

ences bet·"een the De,lCyville-stage stream and the present stream that are

so typical not only along the Ouachita but along most Gulf Coast rivers •

.Measurements in this area indicate that the 'itd terrace chamlels are ap-
proximatt·ly 3 times "iaer than the rc-esent· stream channel and that the

maande! . re.dii and meander ·,Javelengths are probably about 2 times greater

than those of the present stream. These average size differences arc not

so great as they are along certain other Gulf Coast streams; hm1ever,

. they nill ;;ould fall l1ithin estimated range> (Gagliano and Thorn, 1967).

On the basis of sparse and poor quality data, it appears that the

'itd terrace fomatiop substratum is proportionately thicker than is. the.

case in the other De;;eyville formations and that the complete fluvial

sequence might be somel"hat thicker than in the other formations, These

difi'erences are not sUl'prisin€ cOllsid?ring the larger stream channels aud -

the inferred greater discharges and thnllfeg depths associated ,lith the

Qtd terrace.
The lithology of the QW te~racc formation topstratum and substratum'
is not Rno'"" todiffel' from that of the other terrace formation." 'in any

respect. No reliable data arc available to indicate the natur~ of the dc-

-posits filJ.illg the large aban'loned chmlllcls and courses; it can only be

,, that thcy are filled primarily w:i1;h fine-grained sedimOllts "(clays .,

and silts) as in" the Holocene channels and courses of the Ouachita anrc ,,
, ;~


Hississippi Valleys.


The only portion of the HolOCene floodplain of the Ouachita River that

can be considered typical of that stream is the part extending dOlmstrcam

to about the mouth of the Saline River. Below this point, the stream has

been profoundly influenced directly or indIrectly by the Arkansas River.

This influence is manifested in the slope, ,lidth, morphology, and llthology

of the Holocene alluvial sequence.

The dmmvalley slope of the floodplain abovc the Saline River (sec-

tion mile 56, Fig. 2) is uniform at a rat. of about 0.55 ft/mile. Flood-

plain limits are easily defined by a low scarp or other phyJiograp~ic ex-
pression, and thc floodplain consis+.s of the present meander belt and short
segments of u3ually not more than one abandoned meander belt. All meander

belt segments exhibit ll\.unerous abandoned chanr.els or cutoffs. - Floodplain

relief, i.e. the difference betHeen natural levee crests and abandoned

channels or point bar sHales, is generally not more than 15 ft.

South of Sterlington, the Ou<'.chita River is an underfit stream

flowing in an abandoned ArJ<ansas River meander belt. The profiles along the

Holocene floodplain (average level) and natural levee crests in Figure 2

illustrate well the dHfel'ences in eleva',i"n and slope that are prcsent

<belo" eterl1ngton. Before the developmcnt of the Bayou Bartholomew COUrse

of the Arkansas River, the Ouachita River floodplain is assUmcd to have

<continued to slope dOlmstream below sterlington at about the samc rate as ,

it now does above the mouth of the Saline River. HO\;ever, the major

<change illustrated in Figure 2 result cd whon the Arkansas River aggraded

to a level sufficiently high !;o permit it to divcrl through a narrow gap

in the Prairie terrace fonnation at the north end. of the Bastrop Hills

(Plate 1) into the Chemin-a-llaut Creek floodplain thence into the Ouachita

floodplain. The Arkansas River initially achieved a major steepening of

<gradient through this di<rersion but rapii!:t: adjusted by quickly building

up the< fioodpiain to hei£hts of 20 to 30-f't above its <former-elevation.

« • The relatively large natural levees along the Arkansas River course,

as compared with the (.'l,achi<'<[. River levees, are but ono expression of the

substantially higher sediment load (parlicularly snt-size material) of

the fonner streBm. Because of the Arkansas River's appreciably greater

. .
all~viatl.on capacity, an extensive alluvial cone was built by the dver
upstream into the Ouachit a Valley north of Sterl; ngton i'roni the Bayou

Barlholome" course. The effect or the development of< the cone- and tne

corresponding shullo;7Ing «f the gradient of the Ouachita River (Fig. 2)

has been one of alluvial dr01ming and is manifested by the large,

frequently-inundate,", st<ampy 10\Tland or fleod basin called Grand. Marais

(Plate 1) I<{hich is located immediately upstream from the cone bct'<fcen

l'elsenthal and Alabama Landing, La. Within the flood basin, the Ouachita

, J.iver natural levees are poorly devcloped,thc floodplain relief is mini-

mal" and the river has abandoned meander belts more frequently than any-

." "here else. In sharp contrast to this, the Ouachita River exhibits

characteristics of an entrenched stream such as a less sinuous channel,

few abandoned channels or cutoffs, and a nano\{ floodplain with no aban-

doned meander belts "here it flows through the cone south of Alabama

Landing (section miles 70 to 85, Fig. 2).

Investigations conducted near the turn of the century first called

attention to the anomalous behavio"' I)f the Ouachita River bctHeen Felsen-

thnl and sterlington (Veatch, 1906). Faulting was postulated as the cause

of the Grand Mara1s; hO\;cver, neither physiographic nor subsurface evidence

has ever been' discerned to substantiate this. Moreover, thc downvalley


terrace profiles constructed for this study (Fig. 2), indicating no verti-

, cal displacements ,are considered as conclusive evidence that none of the

Quaternary formations in this area.have been significantly affected by

". faulting.
Local variations in the thickness of the Holocene topstratum are

appreciable; ho"ever, it is possible to detect a dOl-mstream increase in

, average thickness from about 20 ft above Calion to about 50 ft below

MOnroe., As HOuld be expected, a major change in the character of the top-

stratum occurs in the vicinity (,f S,terlington. Upstream from this point,

the Holocene topstratum is corr.posed essentinlly of scdiments deposited in

natural levee, abandoned channel, and abandoned course environments.

Sediments deposited in a bnekswamp or floodbasin environment are volumet-

rically insignificant and have not been delineated (F'leehlOod, 1969).


DoWnstream from sterlil".gton, backswamp deposits comprise 75 percent or more

of the topstratum: natural levee deposits and abandoned channel and course

deposits are coni'ined to the narrow and widely separated meander bel,!;s.

The Holocene sUbstratum deposits generally cannot be precisel;- de-

- lineated either upstream or dOWllstrearu from sterlint>ton. In many cases the

: Holocene sequence is apparently thinner than the Qtd terrace for-nation sc-
3 ,
-- quence; consequently,- Holocene sands and gravels are underlain as well as

flanked by Qtd ter:ace -formation sand~;"and gravels. Only in those cases

whE,,-'e the Holocene sequence overlies Cl'l.ibo:cne Group deposits off the flank

of the Qtd floodplail1 can the substratum sequence be defined. A substra-

tum thickness of ro to 40 ftl_s typical of these 10cation3.


llackground - .:~

The most ,rldely accepted and almost traditional correlation of His-

consinan stage events in the Lm-rer ~!ississippi Valley &rea with waxing and

waning - glaciation and corresponding eustat ic sea level variations was first

proposed by Fisk (194 11) and has been modified little since that time. As

indicated in Figure 5a, entrenchment of the l·!ississippi alluvial valley

'- . .

J~nd tributaries) is correlated with I-Iaxing Late lUsconsin glaciation and a

falling sea level. Filling of the entrenched valley, first by braided

streal1lS choked I-lith coarse-grained glacial outwash then by meandering streams

carrying a smaller and finer load, is correlated with waning glaciation and

a rising and finally a static sea level.

In 1968, Saucier presented evidence which indicated that _some of the

fans or cones in the Mississippi Vall~y built by braided streams are


, .-<.,. .:-;,' __ ,. '_ ..• _r'_ -" .

60,000 40,000 20,000 ~.OOO 0
100 .
.J "C ~V
200 '~" ... ..:y .;,~
.J O>-Itd
'.;i",t ~::!
td~oo:: ..-'"-" /,~ ~tn% ... ~<If
~ ~>'0n.Z
. OJZlt..O "<- <} oZ~w
p- Q", z'"
~S z~;:°i=

v, 500
r t .... a:U
0'" '"
... ;,



9w 6g.000 40,000 20,000 5,000 0

._J'"'" 100 ~;t

w ~:s ?,
\\ t:
...'"'" ~

f->- \



x-' wO
.." 500

--- oZI-~~0Z
0: 0<
<C-. U~ :i
.. 0:
~::lI.<~w f-'"
z- >~u.~
>-:3 - :>0°0::- °
JZ.z;:) ",w
« -'i=OV"l ;,:>
E~ 0:'" >-~~8~~ Ow "'<-
>-Z~:> "Z
'" 90:
..2:. ~:!~QJ:ii) Zr
0: ;'1-
ii: ... J5~;?~8
81- "'-
.J>' '"
",0 w .. ",0
0"- .. 0 ~Ut.LO)V)o. ",0 ~a~t:i 0:'

Figure 5. Lm",r l,!ississippi Valley chronological concerts (,,) as

proposed by Fisk (194:,), and (b) as applied by Saucier (1968) to
sea level fluctuation ourye proposed by Curray (1965).

appreciably older than had been thought and must be relo.teu to 0. preceding

period of· ""ning glaciat ion. A conuept of late Quaternary chronology de-

veloped by Currey (1956) based on evidence from continental shelf areas

appeared to offer the only satisfactory explanation for the 10'10 phases of

braided-stream deposition ·and hlo of velley entrenchment obsel'Veu

·by Saucier .(1968) to have occurred follOuing deposition of the Prairie ter-
race formation. The correlation of alluvial valley events with Curray's

sea level curve and.chrono10gy is indicated in Figure 5b.

Varying interpretations could be offered. for many indiVidual asP3cts

of the Ouachita Valley terr"ce sequence if considered separr.tely. Hollever,

the sequence or aggregate situation indicates basic D.{lreement only

with the chronology indicated in Fi(wre 5b' The follo,ling paragraphS con-

tain the auti'')r"' interpretations of the origin of the Ouachita Valley

. terraces based on this chronoloGY.

C Interpretation

Sangamon Interglacial Stap;e. EverY"lhere in tho Gulf Coast eret., the

Prairie terrace formation (or equivalent) displays evidence of ha'ing been

formed during a relatively long interglacial stage with a stable climate

and little or 1:0 variation in sea level. The coast"ise plain portion of

the formation has conspicuous and "ell preserved beach, lagoon, and bar-

rier island f(,:ltures. The riverine plain or relict floodplain pOl'tions of

tnT formatiolis on tJ.l Gulf Coast streams, includine; the Ouachita River, 1'1
indicate attaimnent of a rather advanced state of maturity, with 10'1 flood-·

p:Lain r"lief and a thick topstrll',um formed by rather s 10'1 accumu1at ion of

sedlments throu~h overb"nk depooitiolh In the Ouachita Valley, paleo-

channel evidonce is lack~n~, and it can only be a3sumed that the stream at
c; -

, ,

,thist'1me (i.e. the Sangamon InterglaGiel stage) was s1milar in size and

,character to the present rive,.. ' Figure 6a diagrammatically illustrRtes '

, "probable <ronditions in the Ouachita River area dUring this stage.

, Early Wisconsin glaciatIon. With the onset 01: the Early ~lisconsin

'glaciation,theancesj;ral Ouachita River began to degrade its valley and

become entrenched. This was probably in part a direct respOllse to a

, '

10'lering 01: the river's local base level (i.e. the Mississippi River) be-

cause of adjustment to a falling sea level and in part a -result of a

. climatically-induced change in river regime,'

" ,
The degrailational stage probably persisted until or shortl,y beyond

the max1mum extent 01: the Early Hisconsin glaciation or the Altonian Sub-

&tage (Fig. 51», a :period 01: possibly 15,000 years or more. During this

, , time,- - much but

.. probably
, not all
. , 01: the Prairie terrace fonnati6n
. 'las
rnovedby erosio~. It is logical to assume that at' least isolated areas
. of substratum remained. In those areas where the Prairie fonnation was

_probably completely removed and erosion affecte.dthe underlying Claibol'lle.

Group - deposits,
the erosional contact
, sho>ls no evidence of sub"erial
weathering. This suggests that the basio mechanism of erosion or entrench-'

ment may have been scouring or corrasion by shi-rting and seasonally-'

. enlargi:ig stream channels.

",Figure 6b illustrates the probable conditions existing ;[:,' the, Ouachita

Valley immediately follo'ling the Altonian Substage during the stages

of the retreat or "taning of the' Early \'Iisconsin Glaciation. Althougli the

Ouachita River carried no glacial out'tash, early deposition consisted al-

most entirely of sands and gravels, implying braided-stream deposition.

Sed1ments deposited later are noticeably 1:iner grainc" '1d undoubtedly

.--------,---~-"-,---"'--~--. - !
;' ::. :" '::-,:,.' . ',:. . ~"

;'7':,..:.';,.,...~'-'"~ . "
~';;"-.' ~
~"C'""" ...,. ~ . ~



,1,. -:;--.,:- :: =.--.:.
{~,'~\'t;;.~~'D _

~ ~~. ~--'



,~'. .~"-~ SC:t-~..~~::.~--:-
.~.,,;. -~~:..-.

r ........"" --....
"-"' ...'~...--::>Ot.~ _ ,_ _ '::.? ~1 .._~~~:9~.:­
Q ....
I ~0.~~E""~~~
.',.. " . .- .;.~ ~ .. .

--l . t PRrs£NT (THE I1t'LOCE1I.'O
o Figure 6. Diagral!l!l'.atic illustrations of stages in the developr.:ent of the Ouachita Valley.


>, '.
reflect a change from a braided to a mean~ering recime. Assuming the time

.. scale sho,m in Fi(lUre 5b to be correct, the period of alluviationfollowing

the Altonian Substage lasted for about 10,000 years. Probable conditions

.inthe Ouachita Valley near the end of this period of alluviation and prior

to the Farmdalian Substage are illustrated in Figure 6c.

Farmdalian· Substage. It is necessary to consider the post-Altonian

Substage events in the /fdssissippi alluvial valley to interpret conditions

in the Ouachita River area durincthe Farmdalio.n Substage (Fig. 5b).

Throughout the period of ,,,,,,inc Early l-1isconsin glaciation, thc ancestral

Mississipp.i and Ohio Rivers carried copiou" quantities of glacial outwash

and rapidly aggraded the valley through deposition of coarse-grained al-

luvium. The axis of the Mississippi River valley train' is indicated by

drainage patterns as being west· of and parallel to Crowleys Ridge in

. the. Hestern I.o;(lands (Fig. 1) thence across the pr0sent Mississippi River

course into and parallel to the eastern side of the Yazoo Basin. The Ohio

River valley train apparently developed east ofCrm<leys Ridge and merged
. . .
. ·.. with the Mississippi River valley train in the upper part of the Yazoo

Basin. During this time, the Arkansas River apparently had a significantly

greater discharge because of greater preCipitation in its drainage basin

ar.d/or melhrater from Cordilleran glaciers. Below Little Rock, Ark., the

river ,~s a braided stream carrying large quantities of sands and gravels

and it developed 0. long, narrow alluvial cone extending along the "estern

side of the Hississippi Valley for a distance of over 200 miles. The

upper part of the cone has been largely removed by erosion or buried by

Holoceno meands r belts; however, much of the 10,;01' part remains and forms

a 101'( ridge called Kacon Ridge (Fie;. i.).

. '. The c><tellt' and topographic expression of Macon Ridge at the latitude

at which the Ouachita River enters the Mississippi Valley are shown in

. Figure 7 •. Four levels, designated A through n, are recognized; the eastern-

~'. '. . - - .. -

'most level (lev~l A), '. the highest and oldest and the westernmost level
- '~.., -~,>.

(ie~e1 n) is the lo>lest and youngest. Preserved braided drainage patterns

:/"',.\;(;1' . 'indicilteth~t' the levels are terraces resulting from progressive westward
:silift:lng: of the anc~stral Arkan~asRiver while it "as B10"ly degrading its
····fl.oodJ.ll~ii'...Radiocarbon dates report~d by Saucier (1968) from near the sur-· "

"f'aCE,'of!t£lcon Ridge 8 miles southeast of Winnsboro, La. (Fig. 7) indicate

the deposits of'level C to be about 30,000 years old. Inferred rates of

"sedimentation and stream migration and. relative stratigraphic positions sug-

gestCthatthe deposits in level A are not more than af'ew thousand years

·:olderthanthis. on the 'b'lsis of topographic position and the radiocarbon

dates, level A of Macon'Ridge is considered to represent the greatest extent

of agg~adation in the Mississippi Valley during the Farmdalian Substage.'

Alluviation in the Ouachita Valley (Fig. 6c) can be assumed to have

'beeninequilib~ium >lith alJ:uviation (Mississippi River yalley train devel~p­

'ment and Arkansas River cone development) in theUississippi alluvial v?l1ey

during much of tho period of' "aning Early Wisconsin glaciation. Ho"ever,

. with the attainment of full development of Hacon Ridge during the Fanndalian

Substage, disequilibrium apparently resulted. It is felt that the rate of

deposition of the last few tens of feet of sediment in Macon Ridge culminat-

ing in level A "as so rapid that aggradation of the Ouachita floodplain

could not keep pace. The result of thi, ·"as the creation of what >res ef-

fective1y a dam across the 10>rer end of the Ouachita Valley in the vIcinity

of Homoe and the consequent formation of first a swampy basin similar to


o OCv.1:,"Vl1..u:n'lAIICt:!.

o """. . ,,'t:t.T't""VoCI;

~ U!'<O,r". Tt:II;TU..... ot-oS.TS

-n-... ~gg~~~~I»

1...rvo. 0 ltVEt. C

t)I~TANa:. Io.m.t:!.
;: Fisure 7. Extent and .configuro:tion of the Fa...'""mda.lian Substage'
Arkansas River cone (Macon Ridge) as related to the formation
i of Lake Monroe.
,.,"'-'"- .

"!t: ' """
!: - .--'
I. the Grarrl )-fal'ais and finally lake Monroe. Profile A-A I, Figure 7, illus-

. 'tr'ates that .the highest point. on level A of l-facon Ridge (i.e. the cloaest

approxhflfltion to the original surface) are dthin 2 0)." 3 ft of the rccon-

structcd moan w1l.>dmum level ef lake Honroe, lor;' eleval;bn or about 97 it

,mal.. There is no reason to doubt that level A of l-facon Ridge once extended

as far ,mat M the limit indicated in Figure 1.

.lake ).!onroo was probably perennial for much of its Hfo; ho;:evcr, large

seasonal vflriatioM. in lev"l mmt have occm:red as a result of the Ouachita

!liver dischargo bc1ng augmented I-lith backwater 1'rom the Arkansas or l.Jisais-

sippi Rivers dud ng epring floods. Discharg~ from the lake quite likely

occurred by ,my of one of the active braided channels of the Arkanso.. Iliver

only during relatively 10>1 staees on that river. Figure 6d illustratc" tho
Msumcdcon(litiona in the Ouachita Valley durin(l the Farrrdalian Subnta(lc.

lete l'lisconoin glaciution. The lo;rcr levels of l-facon Ridee (lovels

Jl.:D) cculil have bcen formedcither during the Farmdallan Subotl!6e or 8lightly,

later during the enact of late VliscoMin glaciation. !t is ponsible that

with a decreasc in the discharge of the Arko.l1AllS River such as could have

occurred because of termination of melt;rnter.ar4/or a change to a drier

(interglacial) climate durill(l tho Farrr"lalian Substage, the river becrunc

more compotentto transport its load an'\ began to slmlly deerade ita flood-'

plain. On the other hand, with waxing late Hioconnin glacIation, the

Arkanall3 River as well as thc Mississippi River would have '.cgun steepening
their gl'llIlientiJ "'rl degrading their flocdplains in responsc to falling aea I
level. Xf the Farl«laHan Sub,d;aee 'IUS as brief a3 suggested by Currny

(1965) (Fl,g. 5b), it is llkcly that conditions of cli1r.ate and glaciation

that >loulil permIt a significant decrease in diGchnree did not develop;

-. ,-

,; , hence ,tho latter intorprctatl,on i8 favored. In eithor Calle, the results

would have been tho awne. The alluvial-ridgo barder responsIble for

creating and maintoinlne; Lak(; Monroe for posdbly an J one M Bever"~

: thousand yearn ,inS destroyed 01' brca~hcd suffIciontly to allo>l the lako
--~ .
r', .
' to drain •.

With continued ,/llxlng e;laejed;ion and a r03ultant major oU1tetic

'lo>lering of sea loval bet>recn 2O,vOO and 25,000 years ago (Fig. 5b), th8

Arkansas lliv~r eventually boc"!J1o eignificantly entrenched. l'he axis of

the ontrenohr,d· vallcy io kn",m to be bet>leon tho >lestern edge of 14.. 00n

Ridge' (Fl£>. 7' al¥l the Tortiary uplands or oldertel'l'ace remnaut" (c.g.,

the Bastrop Hills). In the Ouachita Valley, the ancestral Ouachit" Rivor

unque'stionably ;:I\8l\ffectcd by the falling bnoelevel and also >/llS forced

todeerade its floodplain. llo>lcvel', during thi3 interval, tho ·.'iver also

was forced to nceornmodnte 1t"clf to a sienificantly greater discharge.

The resulting wlnptntion to t.ho change in hydraulic rogime in mf.l1J.fcotcd

by the Qtd
3terrncc. As the floodrlnin reprenented by this terrace dc-

"velopod, large £\rOWl of the lacuBtrine plain that resulted from tho drain-

ing of lAke Nonrco ,mr(l dcatroyed by lateral nJr;~aHon of the river. The

rel(>,tiomhip that dcv<llopc(l botwcn the Qtd terrace unO. the lncustrinc
plain (Qtd terrace) is diallraJranatically illustratod in Fle;ur" 60. The
downvalley terrr co profiles included in Figure 2 Ldicnte that net dcgrn-
dntion by the river during wa;d.nll !ate ;/;.:;oonsil. glaci"tinn took place

along tho entire strotch of river included in this study. c,j

.l:!9Zli'ordian SubstfJr;.,. The projected do',rnvallf,y dope of the QI;d
ter- Ii
3 1
race sugcoats that the f,rknnSB" River nt that tIme had entrenched itself

to about the level or the prc3cnt floodplnin (i.p. an elcv.atIon of "baut

65 :!'t msl) some 10 to 15 mHes south of ).1onroe whel'b the Ouachita River

probably became tritJtary to tho Arkansas Riv'}r. The max:iJnum depth of

entrenehrnent 01' degradation attained later at 01' shortly a:!'tor the n:axi-

.mum extent of the Late Hisconsin glaciation (the Hoodfordian Substage)

is not kno,m ,·lith certainty; ho,·tever, i f the thickness of the Holocene

topstratum (backsHamp and natural levee deposits) in this aroa is sub-

. tracted from the floodplain elovation, a probable minimun elevation of

20 ft msl is indlcated for the dcepest entrenchment.

Hith the Arkansas lliver near Monroe entrenched to an elevation of

20 ft msl or 10"er, the Ouachita River ,·taS forced to accommodate and

apparently did so by further degrading its floodplain. The Indicated

minimum degradatIon of the Ouachita River floodplain compatible ,·lith

changes in the Arkansas River is a level about 15 ft belm·t the prosent

floodplain (excluding the effects of tho 'Bayou lJartho10",c" coursn).

It has not been possible to determine tho precise character of the

Ouachita River a';; the time of maximunentrenchment. The floodplain >lldth

..: necessarHy was equal i;o or less than tho present floodplain "idth,

suggesting that the river during ':;he l'Toodfordian Substage might have

beon some',that smaller in size (i.o. a lesser (lischaree) than l.t ;tUS "hen

the Qtd terrace ',ta8 formed. Sca1l9ps. in the Qtd

terrace alo)1g the
3 3
contact Hith the present floodplain but . -t attributable to observable

Holocene floodplain channels suggest that the river ,taS l;~sically a

. meandering stream at or shortly after the timG of ma.xilnun entrenchment.

, The Recent. ~ Hhen the Late I>isconsin glaciation began waning and sea

'level began rising, the proceas of entrenchlJ\ent was halted and floodplain
aggradation becalJ\e dOlJ\inant. Evidence indicates that the Mississippi and

,Ohio River valley trains dcveloped east of Cro,lleys Ridge and trended

thro~gh the Yazoo Basin into Louisiana east of Macon Ridge (Fig. 1). The

.. Arkansas River began buHding a cone frOlJ\the point where the river flows

.;·'oUt'ofthe OzarkMoUntains province ~e~r Little Rock. Possibly because of .

the development of this cone and the rapid filling of the upper part of
. ,

the entrenchment, the Arkansas laver in a rather early stage of >raning

, glaciationdiverted frOlJ\ the entrenched vallcy west of ~!B"on Ridge to a

'llfore easterly course that trendcd lJ\ore directly toward the ~lississippi

River. Thus deprived of its prilJ\ary source of alluvium, the 10lfer part

'of the entrenclunent ,including th_ area adjacellt to the Ouachita valley,

'agg~adedslowlY through the accumulatJ.on of mostly rill~~grained sediments

iiltroduced by lJ\inor streams and derived from backwater flooding by the

" Mississippi Ri.ver. Presumably the Ouachita River Has capable of

", •. _--'----
lIt has been rather conunon practice in the central Gulf Coast area

to define the Recent as being tile period of time follOlan!> the last major

10\'1 stan;! of sea level (Russell, 19110) rather than as being the time fol~ ,,
loWing the cessation of continental glaciatIon. According to the former 1,
def'inition, most Recent deltaic and alluvial 'lalley deposits Hould be '~

separaple from Pleistocene deposits by an erosional unconformity, Hhereas,

according to the latter definition, no persistent time-stratigr"phic di- -

vision Hould be present.. Russell's definition is considered convenient

for use in this p:tper at this point. '

aggradi.lll'; its :rlood\11aill at a corresponding rate also through the accu-

mulatioll of l?rimarily fine~grained materials.

,The Arkansas River ceased building a cone and became a tightly. meandering

'"stream carryir.g a much smaller sediment load perhaps by as early as 10,000

" years ago. During the following several thousand years, meander belts >!ere

,formed extending for more than 200 miies southeast and south of Little Rock
into, Lquisiana along beth sideBof Macon Ridge. lJayouMacon and lJoeuf River

'now each occupy abandoned Arkansas River meander belts in the latitude of

Monroe (Fig. 7). The meander belt occupied by lJayau J.!acon is the older of
the t>lO; hO'-/ever, archaeological evidence (1I'cbb, 1968) indicates both >!ere

>!ell developed and possibly even abandoned by the Arkansas River by about

'3000 to 3500 years ago. It >tas about this time or slightly later that the

Arkansas River developed the Bayou Bartholomew course by diverting through

the gap in, 1;he Prairi~ terrace formation at the north end of t)le Bastrop

,Hills and into the Ouachita River floodplain near Sterlington. Abandonment

, of ,the Bayou Bartholomew course by the Arkansas River probably took place

"about 1000 to 1500 years ago.

"Figure 6r diagrammatically illu"tO'ates conditions in the Ouachita

valley as they exist today and, in effect, as they have been during the

- past 1000 years or more. During this time, the Ouachita River floodplain

has probably aggraded only a few feet through the accumulation of backs>!am?

and natural levee deposits. Course changes and formations of cutoffs have

probably been minor and infrequent.


Extinct Mkcs
Numerous sHampy" depressions and occ(uional1y small lakes are present.

in the Hississippi alluvial valley and the floa:lplains of its major trib-

utaries as a result of more rapid alluviation by one streomservineto

block or· dam the mouth of a smaller st:;.~orun, a .process sometimes refe~red

to as alluvial ~ro',ming. Typically, 1;he drowning is caused by the growth

of natural levees along a newly developing meander belt of a major stream

located a short distance from (Le. several miles or less) and trending

roughly parallel to a terrace or other upland. Small streams dischareing

'from the' upland are ponded in their lower reaches becauso of an inability

to aggrade their 'floodplains to the level of the natural J,evees.

Lake ~!onroe is unusual because of its large size and the fact that it

, is the first extinct lake to, be rec(lgniz8d as hovine been crcated in the

Hissisdppi alluvial valley area as a result of the development of alluvial

cones or valley traIn:. (braided stream deposition). Subsequent to its

discovery, all examinat:ton of topographic quadraneles and aerial photographs

revealed the probable former existence of similar but smaller extinct lakes,

on the Hatchie, Obion, and Forked neer Rivers in western ~ennes.c~ and pos~

',aibly also all the Big Black River in 'Nest-central MissIssippi. Each of these

streams has _in its valley a low terrace that has a very 10" 01' nil do,mvalley

slope in proximity to the l.fississippi alluvial valley. The projected slopes

of these terraces indicate that the str,eems "ere graded to the approximate

highest elevations of the Farmdalian Substage valley trains of the Ohio and

Hississippi Rivers. SuffIcient work has not been done to ~etermine if shore-
line features are present along the sides of the valleys of these smaller

streams. The situation on the lIatchle River is particularly similar to

that on the Ouachita River in that there is not only a terrace ",h1ch ap-

pears to be a correlative of the Q,td -'Qtd levels but there are aloo fluvial
l 2

.' - . .
~'=i_~""''''''''''',",,",;'''''''''"'',",,''<>c"C,.''~:~c,,<,,,,-,,<".~\?\;,'''-L-~ ..... :,.~"_.";'.~'cloo~~ ..~""~A:=d.J";";;":"'"",",-.W....:....2w-T'."".'~""''';'';;'~~':''''';'';;''''~oV.".r...>:~"'--.-O-"o-...=;-~.e..• ~,.. ;:.t:'-~~~'~:V_~o,.-!-.·.;.__,j
features analogous to those of the Qtd level in the same position relative
3 ,
to the olde}' level.
othor streams in tho Mississippi alluvial valley area Vlhose ph~sical

settings and locations relative to tho alluvial valley Vlould have been con-

ducive to, the formation of lakes during Farmdalian Substage vallcy}rain

development uould include 'i;he Coldwater, Tallahatchic, Yalob,:,sha, and

Yokona Rivers of north\{estern ·Mississippi. Each "r these rivers has at,

least one terrace; h.o"ever, no attempt has been made to analyze terrace

g·lopes or morphology •.
An extinct lake quite analoeo11s to Lake Monroe is kno,m to have ex-

istedalong the 101-Ier o;'io, Tennessee, and Clark Riwrs in Kentucky and

Illinois for 40 ",iles above If,etropolis, Ill. (Finch and others, 1946 ).

Primary evidence for this lake, named Lake Paducah, is in the form of rel- .

. iet bay-mouth bars and beaches highly similar to those around Lake Monroe.

The age of Lake Paducah is indicated by radiocarbon dates as beine about

21,000 years (Olive, 1966) and rapid alluviation in the Ohio valley isa

.possible explanation for its formation.

Thus far, there is no evidence that lakes formed on eastern tribu-

tar1es to the ~lississippi Valley during the early Recent when valley train

developed because of \{aning Late Wisconsin glaniation. It Is possible t.hat

the, did develop and that their shorelines eJ.ther coincide I·lith older simi-

lar features or are no\{ buried beneath Holocene floodplain deposits.

Age and Orie1n of De'16yville Terraces

Radiocarbon dates on matm ia15 from the DCHeyvil1e terraces along

several Gulf Coast streams 'very from ~3,000 to 30,000 years before present.


Along the Atlantic Coast, radiocarbon'dat'es indicate formation of equivalents

of this terrac~ b'etwecll 17,000 and 36,000 years b~fore pr"sent (Gagliano and

Thom, 19(7). In sp1,te of the considerable time span indicated by the· dates,
~,. .
',there has be~n "strong tendency to ';ant to correlate'the Deweyville ,lith

events that occurred during a period of on1~i a few thousand years at the end

cirthe ti;"~,span ~uchasrising sea ievel (Bernard and LeBlanc, 1965)' or at

,the begil1lling of the time span, i.e. the Farmdalian Substage (Gagliano and

Thom, ' 19b7; Saucier, ·1968).

The authors feel that this tendency has been motivated in large meas-

exposure to popular Fiskian concepts in which fluvial terrace for-


o m'ation in the Lower H:l.ssissippi Valley area is assumed to have occurred only

interglacial stages or advanced stages of waning glaetiation (Fisk, 0

" Terrace stratigraphy along lobe OUachita River strollgly indicates

that terrace formation took place during late interglacial times and con-

tinued through much of a peri,od of waxing glaciation. The Qtd terrace,

o more typical of the ,De,reyvilleterraces as they occur on other Gulf and.

'Atlantic Coast streams, is undoubtedly a glacial rather than all interglacial

si?age terrace.

bvidellce indicates that the Qtu and Qtd

terraces call be correlated
l 2
with the culminat ion of alluviat ion in the Hiss iss ip1'i alluvial valley
during the Farmdaliall Substage some 23,000 to 30,000 years ago. The Qtd
terrace formed at some later time but before the maximum extent of Late

Wiscon& ~r. glaciat:'.on (,roodfordian Substage) about 18,000 years ago. Topo-

graphic relationships and terrace slopes indicate that the Qtd terrace
was formed while the river was adjusting to a falling base level broueht
about by cntrencIDnent of streams in the Mississippi alluvial valley.

The De"eyville terraces along the Ouachita River differ from those

of othor streams along the Gulf and Atlantic Coasts in regard to certain

~:~> causal fa~tors, but not in regard to their effects. Alluviationor ag- -

~ gradation follOl-/Cd by ·entrencIDnent or degradation in the I>lississippi al-

luvial valley "as essential to the formation of terraces in the OuaLhita

~ Valley, >lhereas along the other streams, a coincident high sea level stand-
.".- ..
follo>led by falling sea level apparently had similar effects. The fact

that Lake lobnroe fOl7ned in the Ouachita Valley and the Qtd and Qtd
l 2
. races are basically lacustrine in or:lgin in the 10>ler part of the valley is

not considered significant in this regard. Its formation is considered as

. simply a local manifestation of an other,lise regional aggradational trend.

~ It >las ·largely for thb reason that the authors designated the Qtd and
Qtd 2 terraces as· De"eyville terrac"s rather than intrcduce ne" names for


ve"eyville Paleoclimatology and Paleohydrology

Although evidenee from the Ouachita Valley could be interpreted as

indicating that degradation of the valley during De>leyville times Has a

direct response to entrenchment in the Hississippi alluvial ·valley, doubt

has been expressed recently that there >las sufficient time during the "1s--

consinan stage for streams to have adjusted their eradients to falling base

levels (either entrenchments or sea level) for allY appreciable distance

above their mouths. Parallelism bet'tlecn gradients of terrace surfaces for

long distances do,lllvalley Hith sudden steepening of gradients near the

coast, such as illuRtrated by Thom (1967) for streams in South Carolina, is

one form of evidencc that has been cited by proponents of this idea.

'If the correlation of the Qtd terrace with waxing Late Wisconsb
glaciation ia correct, this would mean that the terrace was formed during
, '

,-,-.- , aperiod of 'Pluvi~l climate. This is certainly indicated by the large

, relict meanders on the terrace. Schumm (1965) indicates that although

opinions vary considerably, the consensus is that an increase in annual

precipitation of 10 in. and a decreas~ in mean annual temperature of 100 F '

could have occurred in the southeastern United states duril1(l the last

"glaciation. Using these values, Schumm (1965, p. 791) has concluded from

an assessment of the effect of a cooler and Hettel' climate on runoff, sedi-

ment yield,and sediment concentration, that stream incision or qe(lradation

Hould have resulted. He calls attention to an example of stream incision

,'dux'ing late interglacial to mid glacial time folloHed by stream deposition

,during full glacial to early interglacial time along the Red River in Texas

observed by Frye an<i Leonard (1963).

Thus, it'appears reasonable that the DeHeyville terrace of·the Ouachita, .

.as well as of all other Gulf and Atlnntic Coastal Plain streams, may have

originated as a result of two different processes (i.e., eustatio sea level

variations and climatic change) operating simu1tancously to produce similar


Elnpirical relationships beb/ecn certain stream geometry parameters and

stream discharge developed by Carlston (1965) and Schumm (1969) can be ap-

plied to the situation in the Ouachita River area to derive a quantitative

estimate of the paleodischarge of the Qtd -sta g e stream. It "as decided

that the relative validity of th" varioU3 equation>: could be tested by ap-

plying thcm first to measurable parame"ters of the present river and then

comparing the results 11ith known discharge values.

A~ indicated by Schumm (J.969), geometry parameters such as channel

width, depth, gradi.ent, and "avelenc;th appear to be determincd by both

discharge and type of sediment load moved throueh the channel. DatA arn

not a~cilable in the case of the Ouachita River area to indicate thc

'nature of the sediment load in either the present river channel or the

'Qtdo-stage channels., However, because of the close s :imilarity bct"ecn


'the'lithology of the Qtd terrace and, that of the Holocene alluvial ae-
quence, it has been assumed for comparative purposes that the .edDaent

, load cf the river was the same dm'inll the Qtd stage as it has been during
the Holocene. It is further assumed, therefore, that observable vur;,a-

tions in geometry paraloeters are entirely a result of dlfferent discharges.

'As illustrated in Figure 2, there is no significant differcnc" be-

tween the Q~d3 terrace gradient and the Holocene floodplain gradient.

Although it is not possible to reconstl'\,ct the sinu(,sity of the Qtd -stagc

river, it has been assumed that the value of this parameter was essen-

tially the same as that of the present river. Thud, meander wavelength,

"channel width, and channel depth emerge as the only variable parameters

for "hich empirical equations are available that can be measurnd for both

the Qtd -stage and the Holocene channels.

Measurements of the geometry parameters of the present rhel' ch"nne~

and Holocene aban<loned channels (Table 1) 'Iere made beb;een Crurrlen and the

mouth of the Saline River. This areal r0striction was necessary to avoid

the complicating factor of added discharge from a major tributary and

also to avoid an area in which the morphology of the present r1 vel' is

kno,m to have been strongly influenced by the Arkansas, Rlvel'. To permit


.---- .
Table 1. .Holocene and Deweyville (Qtd stR1"e) stream Geometry l1easuremcnts

Location Meander Channel Channel Observed

wavelength, width, depth, mean annual
Lm' ."', in d, in discharge,
feet feet 'Q:, in CfB
in feet

Present o.~ehita
River 9,200 .500 35 14,000

Holocene abandoned
channels 6,800 500 35
Qtd3 terrace aban-
done1 channels 19,100 1,350 7b;!:


.. . ".


. -~ !',
a valid. correlation, all· Qtd terrace st~eam channel measurements '/erc re-
stricted to the same area.
. The meander ,mvelength (L ) value f:»' the present river wac obtained
by ealculat'ing tho average meander radius end mult,_plying by four. This

. procedure ,ias determined through trial and error to be the only practical

method, and it was employed only along those stret"h~s of river th·t ap-

.. peared· to be re!.tively free from e "ternal eomI-ii,cating controls such.s

·,:implngement of .channel against valley wall. The channel width ('I) valuo ,-, '

··.is an average ofa number ·01' measurements made from aerial photogrAphs

and represents the bankfull '/idth. The channel depth (d) value is also

an average derived from hydrographic surveys and represents the interval

from the bankfull elevation to the average maximum channel elevation •

. A cursory examination Of mopo and aerial photographs l'cvenl.ed a dis-

tinctive difference in size between the radii of active river channel

-meanders and those of Holocene abandoned channels. In vIew of the fact

that all Qtd terrace channel parameter measurements necessarily could

be made only from abandoned channels, values for Lm' w, and d were obtained

for the Holocene abandoned channel.s to permit a more va1id correlation.

As indicated in Table 1, the abandoned channels have 1\ significantly

smaller Lm value than active· channel meanders but are identical in regard

to values for If and d.

Values fdr Lm and w for the Qtd
terrace abandoned channels were ob-
tained from mops and aerial photographs. To obtain a sufficient number

of illdividual measurements, it >las necessary to usc all discernible chan-

nels. Therefore, it has been necessary to make the assumption that all
. channels ,regor(llel<s of Bh~ vllriatioll, repres\,nt a static discharge con-

dition and do not represent a transition froln or to a lesser discharge.

, ,

Tha~ dvalue for tho Qtd terrace abandoned channels is considered to be

the le&oo reliable value obtained. Since no precise depths are kno,m, the

valuc cou1.d be derivcd only by infercnce from such observations as the

thickness of the terrace alluvial sequence.
~ The mean annual discharge (Q) v~lue indicated in Table 1 for the
, ,..resent' river was obtained by subtrncting the discharges of thc SaUne

River, Bayou Bartholome'l, and Bayou d' Arbonnc from the kno,m discharge

('27-year record) of the Ouachita River at If,onroe. These str""ms are the

only tri,butarics of consequence in the area, therefore the value for Q

should bc corroct to Hithl.n les3 than 1000 cfB.

Discharges calculatcrl from equations 12 and 22 of Carlston (1965,

p. 875 and 879) andequatl.ons lJ and 10 of 8ch= (1969, p. 257-258) are

"included in Table 2. In both equations of 8chUlllln, 0 value, !~, is in-

cluded "hieh represents the percent of silt and clay (sedin.ents finer

" .than 0.07[' mm) 1,11 the sediments fonning the perimeter 'of the channel..

No quantitative data aro available on the nature of the sediment load in

the Ouachita Rlver arM. To overcome this problem, the wldth/depth ratio

(w/d) "as substituted for M. ""According. to Schumm (personal communication),

the channel shope or ,,/d closely reflects the type of sediment load.

The laree variation in val" 18 for Q sho,m ir. Table 2 is not uncx-
pected considering the nature of U',e equations and the possible major inoe-

curacies in ecomotry purlllnDtcr mc'~ urem~nts used in them. Hhile neither

the equatiol13 or Carlston nor those of Schmnm could produce values for the
'-'.-. F . Table 2. Holoceno. and DC'ICyvillc (Qtc13 staB") str'Oam Dinclw.rllcn Cu.lcula-
'. . tec1 f,'om Empirical F.quationB

Location Schumm (1969)

Q,in efs, calcu1atec1 Q, in cfa, calculated

from: from:

( 1)
·m ....
..,(2). L (3)

Present Ouachita
River 16,300 10,700 3If,200 12,500
Holocene abandoned
channels 8,500 ",O/fOO lh,ooo 12,500
Qtd3 terrace aban-
doned channa ls 80,000 .' 92,900 ,Glf/IOO 229,200

-O.31~ -'
(l)r,=106.1 q<'.46 (3) L=
m 1390 \ 7if
" ...
'. Ir)
(If) ··37 0?9
" "4 t.J_

lei) .•


- -'


:-l&41§&~ti&'1~(#$nff*¥J-44';'t;-.;, ~(jw'?-'iN&i'''';Te'·*V~(L#''''{i'*-'''~i;MUi4¥.Jb*~I.i;frWW_~*~;jw.x? ft{K·.'f-i-g--'"ftt'{4f,"~,:,;r.~~oM>t?.·t>-W ";{;(.:;;; -~tl~~
, t ... ·

present river closcJ.y dvplicatingthe kno,m discharge, those of Carloton

.hm, conniderably :,008 variation. In the nuthb:'s' opinion, the valuos for

Ii for the Qtd terracoabandOnG-1 channel. derived from Carlston's equa-

, tions, are probably morc nearly corr.oct. Theillability to derive an accu-

rate value fr)r d IIhich is un illteeral port of bothof Schumm's equat10M

'. "., , (c-<li; not Carlston' g) is one factor to be consi'lerod as is the fDcl; that
,. , '". ~.;,>:..:.
Sclnimm'a equations, unlike 'those of Carlston, result from an analy.ia of
'rivera that are, for the most port, located in Bubhumid and semiarId ......
regions (Schumm, 1969, p.259). The validity of Schumm's equatior,. is

further negated by the conspicuouuly anomalous value 1'0" Q of 56!1, 700 cfa

lIodetonnined, fr~m r,!'l for the Qtd 3 terrace channels. This value is nlTrKwt

identical with thc valUe for '1i for tho obviously much larger J.!ississippi

River at Vicksburg, Mi86. "

.. .,.
. According to Dury (1965) "ho modo an
... - , .
to est:ltnoto
.: ... '. . . . . .
'paleodischarge by uRing ~,!", relotion"hip that ""iots bet;leen meander char-

:acteristic'J of medern rtv,'co and their discharge, a value for

. - .'.
Ii of 5 to

'lO'timessrcater than the presfmt valuo 'la3 found to be cotronon for a lllrge

number of ~l;reams repreaenting a ;/ide geographic coverage. A simnar

rati6 of pnleed ischarl3e to presont din charge was observe'l by SchU1~'Il (1968)
. in. a study of an Australian river system some"ohat analogous to the Ouachita

River. Both values for ~ calculated by usIng Carlston's equations fall

within this range, whereas. thoae calCUlated using Schu1I1m's equotionJ greatly

"xcood the range.

AssIJrrdng a value for ~ of 80,000 to 90,000 cfs for the Qtd -oto go
stream to be correct, tl.e paleoclimatic lmplicatir,no ar", significan f,. Under

· .. ,;
present climatic conditions "ith a mean annual precipitation of about 52 in.

and a mean annual temperature of ae-,ut 630 F, the mean annual runoff from

tho Ouachita drainage basin above the mouth of the Saline River is approxi-

mately 18 in. Hith a value for Q of 80,000 to 90,000 cfs, the mean annual
runoff "ould be approximately 100 to 115 in. Using curves presented by

Schurrun (1965·, p. 784), it call be seen that regardless of "hat the mean an- . .,

nualtemperature may have been, a runoff of this magnitude probably required

~. a mean annual.. precipitation of at least 100 in. This value is considerably

more than has been postulated bySchurrun and others for this area during a

pluvial climate. Although one could argue that even Carlston's equations

yield anomalously high paleodischarge values and thereby exaggerate inferred

runoffs and precipitation, it must be remembered that the De;reyville terraces

.; on other Gulf and Atlantic Coastal Plain streams have reli"t channels that

are even larger in comparison to the present channels thah they are along

the Ouachita River.


Allen, J. R. L., 1965, A reviel; of the origin and characteristics of Re-

cent al)'uvial sediments: Sedimentology, v. 5, no. 2, p. 91-191.

Bernard, H. A., 1950, Quaternary geology of southeast Texas: Ph.D. dis-

sert., Louisiana StateUniv., Baton Roue;e, la., 165 p.

Be';nard, H. A., and LeBlanc, R. j., 1965, Reswne of .the Quaternary geology.

-Of the 1I0rth"ester11 Gulf of Nexico, 1E. Wright, H. E., Jr., an-i Frey,_

D. G., ~tors, The Quaternary of the Uni'voo States: Pdnc~to'l Univ.

Press, p. 137-185.•

. Carlston, C. VI., 1965, The relation of free meander ;(eometry to stream

discharge and its geomorphic implications: Am. Jour. Sci., v. 263,.

p. 864-885.
Curray, J. R.-; 1965, late Quaternary history, continental shelves of the

. -United States, in)Tright, H. E. ,Jr. ,. and Prey, D. G.; Editors, The

Quaternary of the United States: Princeton Univ. Press, p. 723 c 735.

Doering, J. A. ,.1956, Review of Quaternary surface formations of Gulf' .

Coast region: Am. Assoc. Petrolewn Geologists Bull., v. 40, no. 8,

p. 1816-1862.
Durha'1l C. 0., Jr., Noore, C. H., Jr., and Parsons, B., 1';;67, An agnostic
view of the terraces: Natchez to Uel{ Orleans: Geol. Soc. America

Field Trip GuidebOOk, lIm( OrleaM, 1967, p. EI-E22.

Dury, .G. H., 1965, Theoretical imPlications of unierfit streams: U. S.

Geol. Survey Prof. Paper 452-C, lf3 p.

Finch, H. 1., Olive, H. H., and Holfe, E. H., 1964, 'Ancient lake in western


.,;.,. <.

--~-- --
Kentucky and southern Illinois, in Geological Survey Research, 1"964;

U. S. Geol. Survey Prof. Paper 501-C, p. C130-C133.

Fisk, II. N., 1939, Depositional terrace slopes in Lou~siana: Jour. Geo-

morphology, v. 2, no. 2, p. 181-200.

__'_'191111, Geological investigation of 'the alluvial valley of the Ib;lCr

Nississippl hive)': Vicksburg, l.liss., U" S. Army Corps Ell[;ineers,

Miss. River Comm., 78p.

Fisk, II. 1h, and NcFarlan, E. j Jr., 1955, Late ~uaterJlary deltaic deposits

of the Nississippi River; Geol. Soc. America Special Paper 62,

p. 279-302.
Fleetuood, A. R., 1969, Geological investi.,ation of the Qu""hita River

,area, lo;1er Uississippi Valley: Vicksburg, /.liss., U. S. Army Engl'.

, • .. j

Watenlays Experimetit sta., Tech. Rept. s~69-2, 24 p.

'Frye; J. C., and' Lcionard, A. B., 1963,' Pieistocenegeology of Red River

basin in Texas: Univ. Texas llur. Econ. Geology Rept. Invest. No.

49, 48 p.
Gagliano, S. H., and Thom, B. G., 1967, De;reyvllle terrace, Gulf and At-

lantic coasts: louisiana State Univ., Coastal Studies Bull. 1,


p. 23-41.

Hoyt, J. H., Weimer, R. J., and Henry, V. J., Jr., 1968, Age of Late-

Pleistocene shoreline deposits, coastal Georgia,p. 381-393 in HoI'-

rison, R. B., and Hright, H. E., Jr., Editors, f.!eans of ~orrclation

of Quaternary success:i.ons: Proc. Intcrnut. Assoc. for Qu~ternat'Y

ri~o~areh, VII Internnt. Cong., v. 8, 631 p.

Leighton, H. N., and Hillman, H. B., 1949, Late Ceelczoie geology of Nis-

sinslppi valley: Itinerary of 'Fiel", Conf., J\U1C 12-"5, 1949.

Olive, W. ~1., 1966, lake Paducah, of late Pleistoccne age, in "estern

Kentucky and southern Illinois, in Geoloeical Survey Research, 1966 :

u. S: Geol. Survey Prof. Paper 550-D, p. D87-D88.

Pearson, F. J.; Jr., Davie, B. M.. and l'amers, M. A., 1966, Unive!'sity of

Texas !'adiocar','on dates IV: Rad:,.oellrbon, v. 8, p. lt53-466.

Russell, R. J., 1940, Quaternnry history of Louisiana: aBol. Soc.

America Bull., v. 51, p. 1199-1234.

SaUcier, R. T., 1~68, A new chronology for braided stream surface forma-

tion' in te.e Lo"er l-iississippi Valley: Southeastern Gcology , v. 9,

no. 2, p. 65-76 •
Schnablc, J. E., 1966,The evolution and development of part of the north-

west Florida coast: Florlda State Univ., ~'he Sedimentologlcal Re-

search lab., Cont., No. 12, 231 p.

Schumm, S. A., 1965, Quaternary paleohydrology, !!l Hright, 11. E., Jr.,

and ~'rey, D.G., Editors, ,The Quaternary of the United states:

Princeton Univ. Press, p. 783-794.

-. __' _ 1968, River adjustment to altered hydrologic rcgimen-Murrumbideee

River and paleochannels, Australia: U. S. Gcol. Survey Prof. Paper

598, 65 p.
_ _ 1969, River metamorphosis: Jour. Hydr. Div., Am. Soc. Civil Engl'.

Proc., v. 95, no. HYl, P. 255-273.

?hom, B. G., 1967, Coastal and fluvial landforms; Horryand. Marion Coun-

ties, South Carolina: Louisiana State Univ., Coastal Studies lnst.

Tech. Rept. 44, 75 p.

Tro"bridee, A. C., 1954, Hississippi River and Gulf' Coast terraces and

, , '

" t';SW&;'''"'''il:~''#~'&i;-''*':'~-:=~~""'''''''~''~~ F' ' ' ' ' ' ' ' ' ' ' ' '"41' '«'='Wi' ' ' i' ' "';' "' ' ' '-.w=,~"~_' '~>' ' ' ' ';' ' 'o, ~ ,="c,"'>
aedDnents as related to Pleistocene history-a problem: Geol. Soc.

Amedea Bull., v. 65, no. 8, p. 793-812.

Veatch, A. C,., 1906, Geology and underground water resources of northern

Louisiana and southern Arkansas: U. S. Geol. Survey Prof. Paper 46,
422 p.

Webb, C. n., 1968, The extent and content of Poverty Point culture: Amer-
icallAntiquity, v. 33, no. 3, p. 29'(-321.

,~, ...

". -

C ••.· •..


~ -

~: f

.. "

, - " _. _.-

~ ---
-~- ·t~:-"-. ~ e'§j: Y&i"·"if·M,:"":,*,.-;fiii1.~~~-"",,,~~~£;.ti~'~filtd;h;---;:&~~~f; ,;,-,"",.. ,~~;:,:c""'r,i~$Jk';"a..i.~~..ii#..'-'i~..t.¢..:,;....;ri,.,;;~":;';r---':,.b.~"",:';;:'.=-,"-,,~,,,-'.;d


. ,. 1


» • ~

,. . -.---,' 1

. ····i

>. .•._.<~"lc~AW;ilii~ili'l')~;~'>Co''''4.".,"",~,.L,,,,,,,",,~....~L~'';'''i''&-~w.',.\-J
· .l>!{



Figure 1. Quaternary geology, Ouachita Rh'er art;l. Arkansas and Louidana.



Plgme 1. Quaternary grology; Ou~chita River ar(';J, Arkam21 arid Louidana.

I Grological Society of America- Bulletini v. 81, no. 3

_ _ ____ .. _ ... _ ... _~__ - \r*&%~2-lAi8&i-*"'hWAAW1&;&j.Wfjj;'4i#+-'{'-""~k"'':;i?'*riff.'Ws·~el;§ti;i:;.XW1klA-,"'.r·'.:;~t'ibV~~'¥idg{jd,":¢i;~~1l_-
;. r:
L: -"
i::' ',-.<!':;
~;::,...-' , ....
r<;c '~'., . ".:
r,., ',_. ::-......-. . /".
l ' · /;_.."""'"--;, ',_.

V"~ "",:::::: ., i
(~~;~(':': \ :>_ J.', '•• ,

't. _. .--r_,,/-'
:' - ,,:;:~' -', ',', I
-~"- -~--- -"
'. -. :'-:- .'), -~- .~~~~> ~
-'" ;
;'.: ~::~~.-':'~:;/?<~/.-.:'

- .'-
t-> .. '. - -,-.,.
, , j
I~~':,-:,>: _ \:~'~',>_: ,: ',_~~;~~~

\,~;;~~<: :- ~ l.

L~ . . . . . , -. . -. . . . .__'_-__. ~~ ,-
Figure 1. Photomosaic. Figure 2. Interpretation. -t,