Watson Brake, A Middle Archaic Mound Complex in Northeast Louisiana

40 Pages·2017·6.66 MB·English

TThhee UUnniivveerrssiittyy ooff SSoouutthheerrnn MMiissssiissssiippppii
TThhee AAqquuiillaa DDiiggiittaall CCoommmmuunniittyy
Faculty Publications
10-1-2005
WWaattssoonn BBrraakkee,, AA MMiiddddllee AArrcchhaaiicc MMoouunndd CCoommpplleexx iinn NNoorrtthheeaasstt
LLoouuiissiiaannaa
Joe W. Saunders
University of Louisiana at Monroe, [email protected]
Rolfe D. Mandel
University of Kansas, [email protected]
C. Garth Sampson
Southern Methodist Uniersity, [email protected]
Charles M. Allen
Environmental and Natural Resources Management Division, [email protected]
E. Thurman Allen
Natural Resources Conservation Service, [email protected]
See next page for additional authors
Follow this and additional works at: https://aquila.usm.edu/fac_pubs
Part of the Anthropology Commons
RReeccoommmmeennddeedd CCiittaattiioonn
Saunders, J. W., Mandel, R. D., Sampson, C. G., Allen, C. M., Allen, E., Bush, D. A., Feathers, J. K., Gremillion,
K. J., Hallmark, C., Jackson, H. E., Johnson, J. K., Jones, R., Saucier, R. T., Stringer, G. L., Vidrine, M. F.
(2005). Watson Brake, A Middle Archaic Mound Complex in Northeast Louisiana. American Antiquity,
70(4), 631-668.
Available at: https://aquila.usm.edu/fac_pubs/8240
This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for
inclusion in Faculty Publications by an authorized administrator of The Aquila Digital Community. For more
information, please contact [email protected].
AAuutthhoorrss
Joe W. Saunders, Rolfe D. Mandel, C. Garth Sampson, Charles M. Allen, E. Thurman Allen, Daniel A. Bush,
James K. Feathers, Kristen J. Gremillion, C.T. Hallmark, H. Edwin Jackson, Jay K. Johnson, Reca Jones,
Roger T. Saucier, Gary L. Stringer, and Malcolm F. Vidrine
This article is available at The Aquila Digital Community: https://aquila.usm.edu/fac_pubs/8240
WATSON BRAKE, A MIDDLE ARCHAIC MOUND COMPLEX IN
NORTHEAST LOUISIANA
Joe W. Saunders, Rolfe D. Mandel, C. Garth Sampson, Charles M. Allen, E. Thurman Allen,
Daniel A. Bush, James K. Feathers, Kristen J. Gremillion, C. T. Hallmark, H. Edwin Jackson, Jay
K. Johnson, Reca Jones, Roger T. Saucier, Gary L. Stringer, and Malcolm F. Vidrine
Middle Archaic earthen mound complexes in the lower Mississippi valley are remote antecedents of the famous but much
younger Poverty Point earthworks. Watson Brake is the largest and most complex of these early mound sites. Very extensive
coring and stratigraphic studies, aided by 25 radiocarbon dates and six luminescence dates, show that minor earthworks
were begun here at ca. 3500 B.C. in association with an oval arrangement of burned rock middens at the edge of a stream
terrace. The full extent of the first earthworks is not yet known. Substantial moundraising began ca. 3350 B.C. and contin-
ued in stages until some time after 3000 B. C. when the site was abandoned. All 1 1 mounds and their connecting ridges were
occupied between building bursts. Soils formed on some of these temporary surfaces, while lithics, fire-cracked rock, and
fired clay/loam objects became scattered throughout the mound fills. Faunal and floral remains from a basal midden indi-
cate all-season occupation, supported by broad- spectrum foraging centered on nuts, fish, and deer. All the overlying fills
are so acidic that organics have not survived. The area enclosed by the mounds was kept clean of debris, suggesting its use
as ritual space. The reasons why such elaborate activities first occurred here remain elusive. However, some building bursts
covary with very well-documented increases in El Nino/Southern Oscillation events. During such rapid increases in ENSO
frequencies, rainfall becomes extremely erratic and unpredictable. It may be that early moundraising was a communal
response to new stresses of droughts and flooding that created a suddenly more unpredictable food base.
Los complejos de monticulos de tierra del Arcaico Medio del valle del rio Mississippi son los antecedentes remotos de los
famosos monticulos de Poverty Point, que sefechan mucho mas temprano. Watson Brake es el mas grande y el mas complejo
Joe W. Saunders and Reca Jones ■ Regional Archaeology Program, Museum of Natural History, Department of
Geosciences, University of Louisiana at Monroe, Monroe, LA 71209-6520 ([email protected])
Gary L. Stringer ■ Museum of Natural History, Department of Geosciences, University of Louisiana at Monroe, Monroe,
LA 71209-6520 ([email protected])
Charles M. Allen ■ Environmental and Natural Resources Management Division, 1647 23rd St.l Bid. 2529, Fort Polk, LA
71459 ([email protected])
E. Thurman Allen ■ Natural Resources Conservation Service, 2410 Old Sterlington Road, Suite B, Monroe, LA 71203
([email protected])
Daniel A. Bush ■ P.O. Box 396, Deerfleld, NH 03037 ([email protected])
James K. Feathers ■ Department of Anthropology, University of Washington, Seattle, WA 98195-3100
(jimf@u. washington.edu)
Kristen J. Gremillion ■ Department of Anthropology, Ohio State University, 244 Lord Hall, 214 West 17 Avenue,
Columbus, OH 43210-1364 ([email protected])
C. T. Hallmark ■ Soil and Crop Sciences Department, Texas A&M University, College Station, TX 77843-2474
(hallmark @ tamu.edu)
Edwin H. Jackson ■ Department of Sociology and Anthropology, University of Southern Mississippi, Box 5074,
Hattiesburg, MS 39406-5074 ([email protected])
Jay K. Johnson ■ Department of Sociology and Anthropology, University of Mississippi, University, MS 38677
([email protected])
Rolfe D. Mandel ■ Kansas Geological Survey, University of Kansas, Lawrence, KS 66045-2121 ([email protected])
C. Garth Sampson ■ Department of Anthropology, Southern Methodist University, Dallas, TX 75275-0336
(gsampson @ smu.edu)
Roger T. Saucier ■ Deceased
Malcolm F. Vidrine ■ Division of Sciences, Louisiana State University at Eunice, P.O. Box 1 129, Eunice, LA 70535
(m vidrine @ lsue.edu)
American Antiquity, 70(4), 2005, pp. 631-668
Copyright© 2005 by the Society for American Archaeology
631
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
632 AMERICAN ANTIQUITY [Vol. 70, No. 4, 2005
de estos sitios tempranos de monticulos. Los estudios extensivos estratigrdficos y de bloques de sedimentos taladrados, o sean
corazones, junto con la obtencion de 25 fechas de radicarbono y seisfechas de luminiscencia, muestran que la construction
de monticulos pequenos comenzo aqui hacia 3500 a.C. en asociacion con un arreglo oval de piedras quemadas ubicado al
borde de la terraza del no. La extension espacial de estos primeros monticulos de tierra no ha sido establecida todavia. La
construction sustancial de monticulos comenzo hacia 3350 a.C. y continuo a traves de varias fases hasta despues de 3000
a. C. cuando el sitio fue abandonado. Los once monticulos con sus crestas interconectadas fueron ocupados entre estadios
rdpidos de construction. Las capas de sedimentos se acumularon en algunas de las superficies temporales de estos compo-
nentes de tierra, mientras que el material litico, las piedras fracturadas por fuego y los objetos de arcilla o tierra arcillosa
cocida se dispersaron por todos partes del monticulo. El registro faunistico y arqueobotdnico de los depositos basales demues-
tran que el sitio fue ocupado durante todas las estaciones del aho, idea apoyada por una subsistencia concentrada en la
explotacion de nueces, peces y venado. Las capas estratigrdficas mas superficiales son de una matriz muy dcida, la cual ha
impedido la conservation de restos orgdnicos. El area circundada por los monticulos fue mantenida limpia y libre de despo-
jos, los que sugiere que tenia unafuncion ceremonial. Las razones por las cuales tales actividades se llevaron a cabo aqui no
son claras. Sin embargo, algunas de las fases de construction se correlacionan con algunos de los periodos mejor documen-
tados de aumentos de los eventos de El Nino. Durante aumentos rdpidos en la frecuencia de ENSO, las lluvias ocurren en
forma irregular e imprevisible. Es posible que la construction de monticulos de tierra fuera una respuesta comunal a pre-
siones causadas por una imprevisible escasez de recursos, la cual estuvo ligada a sequias e inundaciones.
discovery of mas sthievree is aLnoathtere i nA Mirssicsshippai i(Ccon neawaary teth al.-
works at Poverty 19 7P7).o Onien mtou n(d1 si6te W(MonCte 5Sa)n o,i dnest rtoyhed ein
lower Mississipp 1i96 8v) maalyl deatye t o (>F500i0g caul Br.Ce. ( H1ay)s 1h995a; s
posed two enduring Sqauundeesrst 1i99o4)n, wsh ilie nall tNhe ootrhetrsh ha vAe bmeenerican
prehistory - when adanted dto bewtweehn 4e05r0 aend 30d50i cdal B .Cs. uch mound-
building begin, and wh Tahuts fatr, rWiatgsong Berarke ehads t hte mhosot esxteen siaveclytivities?
(Ford 1969; Ford an dtes tedW ande dabtedb M id1dle9 Ar5ch6aic; e arGthwiorbks sinon 2001;
Russo 1994a; Webb 1No9rt8h 2Am).e riTca. oHdowaevyer,, p roevniouls ypub lticahtioens first
question can be sati sinfcluadce otnoly ra ciolnyden saedn susmwmareyr ofe rdes.u ltEs arth-
works were construct (Seaudnd eirns e t talh. 1e99 7)v anidc sipneciiatlisyt r eopofrt s Ponoverty
Point at least two mi OllSLe ndatnesi (aBu sbh eanfd oFeratehe rbs 2u00i3l;d Feianthgers began
(ca. 1750 cal B.C.) at t19h97)a, ltith ifcs o(Jrohmnsoni 2d00a0),b flireed ecarothmen opbjelctesx. Six
older radiometrically (S audndaerts eetd al . m1998o), uannd fdau nca (oJamcksopn laendxes are
now known within a 1 S c5ott0 2 00k1).m The groaal dof ithue spr eosenft rePpoortv ise tor itntye- Point
(Figure 1), of which g rLateo awll theesre linJes aofc iknvesstoignati on( 1an6d tWo maCke10) is
the nearest (Gibson av1ai9lab8le 9ot;h erS uanupubnlisdheed resssen teialts inacllud.i n2g 001),
followed by Nolan ( s1am6plMing Ades2ign0, 1me)t h(odTs, .a ndK thied redsueltsr f,ro mper-
sonal communicatio mnor e2 re0ce0nt 5an)al,y seFs. Wree alnso casshessm the arenlev'asnc eBend
Mounds (16OU259) ( Sto aWautsonn Brdakee ofr twso cuerrtent maodle.ls of1 th9e o9ri-4), Wat-
son Brake (16OU17 gin5s )of M(iSddlae Aurcnhaicd meourndbsui ldeingt. al. 1997),
Hedgepeth (16L17) (Saunders and Allen 1994),
and Caney Mounds (16CT5) (Gibson 1991; Saun-
Background
ders et al. 2000). Limited testing shows that their
builders were Middle Archaic hunter- foragers who
The Site and Its Setting
exploited riverine habitats and used locally avail-
able lithic materials li kWaetso nt Bhrakoe1s ise an uovasl aerrdan geamten t nof e1 1a erartbheyn camp-
sites without mounds such as Plum Creek mounds with connecting ridges. These form two
(16OU89) (Saunders 1998; Sheffield 2003 c)u ravnedd rows of earthworks called the north and
Metz Midden (16RI105) (Saunders 2000). Fa rstohutehr mounds (Figure 2). The level enclosure
afield, six other dated Middle Archaic mound b estitweesen them measures about 300 m long by 200
are known in southern Louisiana (Brow nm aancrdoss, and the tallest mound (Gentry Mound, A
Lambert-Brown 1978; Hays 1995; Russ oin aFnigdure 2) is 7.5 m high.
Fogleman 1996; Saunders and Allen 1998; R. Sa uTnhe- earthworks were built on the east rim of an
ders 1994, personal communication 2005), aalnludvial terrace, some of which was probably
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
Saunders et al.] WATSON BRAKE 633
Figure 1. Location of Watson Brake in N. E. Louisiana in relation to other sites mentioned in the text.
formed by the O auquataic cfauhna.i Ats wail l bRe shiowvn,e thre si ted's inuhabri-ing the late Pleis-
tocene. Today, th tanets exrploiitved tehe rful l rcanhge oaf annimnal reesolur cefs, lows about 500 m
farther east of th buet l imsitiedt prees,er vataionn ofd m acraob otmanicali rnemaoinsr tributary known
as Watson Brake refstroictrs oumr knsow ledage osf twhe palanmts copnsuymed . floodplain imme-
diately below and A ntothoer asstet hof tehi s swite's elocsatiton is othfe im mtedhi- e terrace rim. At
the time of site at elyo avcailcablue topolstaonte ini cohernt gr,av elts erhodiings was probably a
clear-running s firodm thee t ercrache scaarp.nnel that flowed about a
meter lower than today, and it was less swampy.
Research History
Complex shifts in major local drainages between
3000 and 2000 TBhe f.irCst re.c ord(eSd viasitu to cthei siete wras in1 19891, 94; Washington
2001) placed nat whuen rsevean mlo unldse anvd a efewe rsidg es awecre ridenoti-ss the mouth of
Watson Brake, t hfiede anrd sekebtchyed b y cReoca Jnonves aendr vistiteid nby g it to a swampy
backwater. These St epehevn Weinlliamtss. S urmface afinyds subggeest edi amplicated in the
abandonment of Potvehrtye Po inst aiget (cea. 2 00a0-t100 0 tB.Ch.) fiors th e tsitie me, but the rela-
tionship has not (Jyonees 2t000 ). bFouer meorne m oufndus anldl oyther riidngesvestigated.
The site is on an ecotone between riverine and were located the following year. John Belmont and
upland mosaics, both extremely rich in edible Reca Jones compiled a map of the site in 1984, and
plants. The upland sector of the site's catchmen Wt atson Brake was first mentioned in print the fol-
supports deer and a full range of small mamma llsowing year (Jones 1985). The age estimate went
and birds, while the riverine sector carries a divers eunchallenged, although Kidder (1991) suggested
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
634 AMERICAN ANTIQUITY [Vol. 70, No. 4, 2005
Figure 2. Surface elevations of the Watson Brake mounds and ridges at half-meter intervals. Inset: group names used in
the text.
that it might be plac epudrch aeseda bry lthiee Arrc haieonlog ictal hCoense rLvanacyt, ean dArchaic,
at ca. 2000 B.C. In 1 s9ub9seq2ue,n tlyp seoldd too thleo Stgatei cof aLolu isaianna, athleysis of
Mound A fill sugges tsouethd m ouand s cMouldi bde tedsteld efo r Athe rfircst htimae.ic age for
the mound (Saunders et al. 1994) and prompted the
Goals
design of field investigations, which ran between
1993 and 1999, follo Rwesearechd p robgryams wmere udesligtneidd toi esstcabilisph lthienary
analyses. Until 1998, afgei oef ltdhe wsiteo, rto kcon fwirma tsh atl tihme mioutnedsd an dto the
north mounds because the landowner of the south ridges were of human origin, to measure the effects
mounds denied access. After the south group wa sof longer and shorter building hiatuses on soil for-
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
Saunders et al.] WATSON BRAKE 635
Figure 3. Locations of all excavated test units and cores at Watson Brake.
mation within 5t0 mh. Oever 1m4,000 ototual stnatiodn t opofgraiphlicl rsead,- and to document
whether earthw inogs rwekre csoll ectwed feor ra siete mrap (aFigiusre e2).d All individually or as
groups. Datable st ratmigraphaic tdesecrirptioinsa anld scor reflatiorns ouse mthe buried soils in
mound and ridg samee v erfticali sclalle.s were recovered with mini-
mal destruction Tthe olayo utt of haugeer p oinets aandr testt phits isw showon rks themselves by
using augers, c oin Friguere s3, t,og ethaern withd th e flielid mlabels. iTetst uenitds test excavations.
When middens with fauna were discovered at the on the north mounds and ridges were aligned within
base of Mound B, it also became possible to exa ma 2-0 m grid, oriented to magnetic north, with each
ine the subsistence base of the site's earliest occu- location predetermined by auger/core/probe results
pants. that revealed buried soils and/or other organics.
Two of the auger/core locations were outside the
grid and their locations recorded with a total sta-
Samples and Methods
tion. Within the north half of the central enclosure,
auger probe layout follows the grid system, with a
Sampling Layout
few placed at 10 m intervals. This was designed to
A site datum was established in the middle of the explore patterns of artifact density inside the enclo-
enclosure and assigned an arbitrary elevation osfure. Auger/core probes into a low rise in the cen-
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
636 AMERICAN ANTIQUITY [Vol. 70, No. 4, 2005
ter of the enclosure w seedrimeen tas wrebre icotllrectaedr firolmy al l phorliazocnse fdor and their
locations shot in later. Dense tree cover in the south mechanical and chemical analyses, and selective
half of the site made the grid system unfeasible, so samples were collected for luminescence dating.
all auger points and test units were arbitrarily placed Intact soil blocks were collected for micromor-
and shot in later by total station. These test units phological analysis.
were also oriented to magnetic north.
Angering and Coring
Excavations
Four devices were employed for sampling the site.
Test units in the north mounds were 1 -x- 1 .5- mA tp sitasmple points marked "Auger" in Figure 3, a
(TU 5 is a 1-x-l-m extension of TU 4) taken hdaonwdn bucket auger was used to collect 17-cm-x-8-
to what was judged to be the Pleistocene t ecrmra cseamples of loose deposit. At points marked
surface. Test units were designed to record s t"rCaCt,i"- a JMC continuous corer was used, in which
graphic details, to recover artifacts, and to o ab tslaiidne hammer pounds a 2-cm-x-9 1 -cm probe then
datable organics in situ. The only earthwor kesx tnroactts it with a foot jack. Points marked "Core"
tested were Mound A, from which continuous d ceonrotee- 5 -cm continuous cores extracted with a foot-
sample segments were extracted, and the otpineryated hydraulic rig, effective to a depth of about
Mound K, which was augered. 5 m. At other points of interest, a 25.4-mm Oak-
Test units in the south mounds were 1.5 m x 1 field push probe was used to evaluate stratigraphy.
m for E, 1 m x 1 m for F through H, and 1 m x .5 In the mounds and ridges, sampling depths var-
m for L. Since no buried organics were encountere died according to whether the interface between the
by preliminary coring, only the artifact-bearing terrace surface and the overlying earthworks could
upper portions (30-40 cm below surface) in each be recognized. In the south mounds where this
unit were sampled. No test units were placed on interface could not be readily observed, coring con-
Mounds I or J because core samples again pro- tinued to below its predicted depth. Coring in
vided good sequences of buried A-horizons with Mound A could not be extended to the terrace sur-
datable organics. The terrace surface under a few face and was abandoned at a depth of 5.6 m. The
earthworks may not have been reached. 42 auger holes inside the enclosure reached a depth
Not shown in Figure 3 are four 1-x-l-m units to of about 45 cm. Those in the low rise (Auger A,
the west of the complex. These offsite excavation sCC-A in Figure 3) went to 138 cm below surface.
were conducted to investigate possible peripheral Deposits from augers and cores in the earth-
occupations, and went to depths of only 10-30 cm .works were described in the field and sampled by
Test units were excavated by cultural/natural soil horizon. These were passed through a 3.2-mm
layers where these occurred. Most visible layers ar emesh or stored for later analysis. The samples taken
buried A-horizons of ancient soils, but concentra- from within the enclosure were processed through
tions of lithics and/or fire-cracked rock were also a .5 -mm screen to recover microdebitage and small
intersected. Deep homogenous fills were more pieces of fire-cracked rock.
common, and these were excavated in arbitrary lev-
Radiocarbon Dating
els. In each unit, excavation was halted within what
was thought to be submound or subridge alluvia lAltogether, 28 radiocarbon samples were collected
deposits. Culturally sterile deposits were no tfrom seven mounds and one ridge. Of these, 19
reached at the bottom of any unit. All deposits were were recovered in situ from test unit excavations
dry-screened through 3.2 mm mesh and all retained and the rest were taken from core or auger sam-
materials, including gravel and concretions, wer eples. Charcoal was assayed in 16 cases where the
collected and processed. Scarce midden deposits buried A-horizon incorporated traces of cultural
were dry-screened into plastic garbage bags and debris. In all but one case (ridge K/A, TU 2), the
taken to the lab for flotation. charcoal occurred as small, scattered particles
All four walls of each test unit were drawn, and rather than large intact pieces. One charcoal sam-
the north and south walls photographed. Features ple in Mound D proved to be a historical intrusion,
such as small pits, depressions, hearths, and post- presumably a burnt tree root, and this date has been
holes were drawn in section and plan form. Bulk omitted.
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
Saunders et al.] WATSON BRAKE 637
One date was run on charred bone from Mound mated from hand specimens. The only core sam-
B, but results were discouraging and no furth eprle textures derived from particle size analysis are
bone samples were submitted. The balance of th ferom the top meter of Core 2 in Mound A (Saun-
dates were run on humates extracted from six buried ders et al. 1994:Table 1).
soils, four organically enriched sediments of buried
Artifact Analysis
A-horizons, and a feature fill, all subjected to rig-
orous pretreatments. Seven of the samples wer Geravel and concretions were separated from arti-
dated by AMS 14C, and the remainder by conven -facts, weighed, size-graded, and stored. Fire-
tional (unextended) gas counting. 313C values wer ceracked rock was then separated from the artifacts,
obtained with all but one date and used to correct weighed, size-graded, and stored. The fired earthen
for isotopic fractionation. All calibrations use th oebjects were weighed, measured in three dimen-
same INTCAL curve (Stuiver et al. 1998). sions, and classified by shape using a typology
detailed in Saunders et al. (1998). Lithics were
Luminescence Dating
weighed, measured, and classified by reduction and
This program concentrated on exploring the fea styi-pological criteria. A reduction sequence analysis
bility of dating mound fill by optically stimulat eddeveloped to deal with biface industries made from
luminescence (OSL). Preliminary work analyz eMdississippi gravels (Johnson 1989; Johnson and
buried soil samples from Mound B (TU 3) a nRadspet 1980) was employed. Finished bifaces were
ridge K/A (TU 2) (Feathers 1997). Subsequent ltyy,ped following local definitions (Ford and Webb
more refined methods were applied to buried s o1i9l56; Webb 1981).
samples from Mounds B, C, and K and ridges K/A
Faunal Analysis
and J/K (Bush and Feathers 2003). In more recent
work, sample columns were collected across bur iFedaunal remains were encountered in four of the
soils and analyzed in 5 cm segments, using bo tnhorth mound earthworks and from two cores in
multigrain and single-grain aliquots of 90-125 u rMnound J. Most of this material was too fragmented
quartz. Segments corresponding to the buried Ato- be identified to taxon, excepting in Mound B,
horizon proved in most cases to be the be swthere preservation was good. The latter samples
bleached. Only well-bleached grains as determin ewdere divided into large (>6.4 mm) and small (6.4
from single-grain analyses were used for datin mg,m-3.2 mm) fragments. Human remains, bone
using a leading edge algorithm. In addition, 35 cm ar-tifacts, antler fragments, gar scales, fish otoliths,
deep topsoil columns were collected from Mound asnd shell were first removed from the large frac-
B and C to study sediment turnover rates in act itvieon. The residue was then identified to taxon and
soils, which is reflected in the frequency of we lel-lement where possible, and MNI counts were
bleached grains. computed from these. Subsamples of the small frac-
tion were drawn from six levels and subjected to
Soil Analysis
the same procedure.
Soils were described in the field using standard Age-at-death was determined for the limited
U.S.D.A. terminology (Soil Survey Staff 1996). deer sample using dentition, epiphyseal fusion
Soil samples were collected from eight excavation onset data (Purdue 1983), and modern reference
units (TU 1-8) for physical and chemical analysis, collections. Season-of-death data were determined
including particle size distribution, pH, cation for fish otolith annuli (Stringer 1998), and weight
exchange capacity, exchangeable Al, organic car- estimates for drum were derived from otolith
bon, and Fe2O3 content. Total exchangeable bases lengths and weights compared to modern reference
(Ca, Mg, Na, K) and exchange acidity were used collections. Mussel shell fragments were identi-
to calculate percent base saturation. Micromor- fied to species using modern reference collections
phological analyses were conducted on soil thin and tallied, as was the single species of aquatic
sections from several mounds and ridges. snail.
Sediment samples from cores were not sub-
Macroplant Analysis
jected to laboratory analysis. Color and pH were
measured in the field and texture classes were esti- All dry-screened midden deposits encountered in
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms
638 AMERICAN ANTIQUITY [Vol. 70, No. 4, 2005
Figure 4. Watson Brake north mounds: profiles of the seven test units showing soil horizons, inclusions, and sediment
sample columns with texture classes. Ap in TU6 refers to Mound A platform.
the north Mounds B H,um aCn R,em aainns d D were floated in a
190-liter flotation device using city water. The
heavy fraction was cNoo blurliealsc wteree denc ouontnere da. I so1lat.e6d h-umman mfrag -nylon
mesh screen, and run moenftsf w erfe rideontmifie d ttoh eleem entta, anndk ag e-wat-daeasth passed
through a fine mesh wbas adegte rmtione dc wohelrle epocsstib let. he light frac-
tion. Once dried, the light fraction was macro-
scopically scanned. Samples of heavy fractions
Mound Construction History
were also scanned, but were found to contain rel-
atively little charcoa lT.h e Pmoourndts ainod nridsge s oidfent iftiehd bey J3on.e2s (-19m85) m dry
screenings were also allc cohntaein cabkundeandt ar tiffacots rand bmurieda otr terurn-ials. Spec-
imens were bagged b cayted spoilsr thoat vvereifny tihee enarthcweor ksa' cnultudra l oirdi- entified
using modern refer geinsn. Tcheey a rec coomlploseedc oft oineo onr msor.e uSnites eofd coat-
thickness was measu rfille, adnd a usoils hiasn degvelo paed ats thce atonp ofn miosnt fgills electron
microscope. (Figure 4). Fill units are numbered sequentially in
Figure 5. Watson Brake north mounds: profiles of test units with numbered earthwork fills over terrace (-tr) deposits.
Textures of the terrace deposits under C/B, B, and B/A (-tr?) resemble fills. Pointers show locations of dated samples.
This content downloaded from 131.95.218.41 on Tue, 29 Aug 2017 19:16:05 UTC
All use subject to http://about.jstor.org/terms