COMPREHENSIVE HISTORIC PRESERVATION PLAN HOUSTON·GALVESTON NAVIGATION CHANNELS, TEXAS PROJECT, GALVESTON, HARRIS, LIBERTY AND CHAMBERS COUNTIES, TEXAS.
By Stephanie L. Perrault and Charles E. Pearson
with contributions by Margaret S. Henson Kay Hudson and Paul Heinrich
Submitted to: U.S. Army Corps of Engineers Galveston District (Contract No. DACW64·91·D·0010, Delivery Order No.1)
Submitted by: Coastal Environments, Inc. 1260 Main Street Baton Rouge, Louisiana MARCH 1993
TABLE OF CONTENTS EXECUTIVE SUMMARy..................................................................... ii LIST OF FIGURES ............................................................................. vii LIST OF TABLES ............................................................................... viii CHAPTER 1: INTRODUCTION .......................................................... Purpose of the Historic Preservation Plan.. . . . . . . . . . . . . . . . . . . . . . . . . . . . The Galveston Bay Navigation System............ .................... Goals............. ..................... ..... ............ ................ Policies................................... ................... ........... Priorities.... .. .. .. .. .. .. .. .. .. .. .. .. .. . .. .. .. .. .. .. .. . .. .. .. . .. .. .. . ... Organization of the Historic Preservation Plan........................ CHAPTER 2:
CHAPTER 3.
OVERVIEW............................................................... Galvest<~n ,I?i.stricts, Corps of Engineers Needs and ResponsibIlities ......................... " . . . . . . . . . . . . . . . . . .. . . .. . . .. . . . Cultural Resources Management Needs......................... Legal Responsibilities.............................................. Summary of Geological History and Natural Setting of Galveston Bay........................................................... Climate ............................................................... Biota ............................................................... Summary of Cultural History of the Galveston Bay Area ......... Previous Archaeological Research. . .. .. .. .. .. . .. .. . .. .. .. .. .. ... Underwater Archaeology .......................................... Native American Culture History ................................ The Prehistoric Period. ... . . .. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. The Paleoindian Period ........................................ The Archaic Period ............................................ Late Prehistoric Period. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Historic Native American Cultures .......................... Euro-american Culture History .................................. Early Exploration and Settlement Prior to 1800............ Early-Nineteenth-Century Occupation and Settlement, 1800-1836 ...................................................... Texas Independence to the Civil War ........................ The Civil War, 1861-1865 .................................... HISTORIC CONTEXTS .............................................. The Historic Context Concept.. .. .. .. ... .. .. .. .. . . . .. .. .. . .. .. .. .. .. .. 1. Late Quaternary Environments, Paleogeography, and the Archaeological Record.................................... Introduction and Perspectives.. .. .. .. . .. .. . .. .. . .. .. .. .. .. .. . Geological Setting. . . . . . . . . . .. . . .. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . Geological Features and Geomorphic Processes.. .... Upland Surfaces and Process ................................ Coast-Parallel Terraces ................................... Prairie Complex. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Prairie Terrace. . .. . . . . . . . . . . . . . . .. . . . . . . . .. . . . . . . . . . . . . . . .. Beaumont Formation - Beaumont Alloformation .......... Surface Modification ......................................
ii i
1-1 1-1 1-2 1-8 1-9 1-10 1-12 2-1 2-1 2-1 2-2 2-7 2-16 2-16 2-18 2-18 2-30 2-31 2-31 2-31 2-33 2-34 2-37 2-38 2-38 2-39 2-41 2-42 3-1 3-1 3-2 3-2 3-6 3-6 3-15 3-16 3-17 3-17 3-20 3-22
Surface Modification ...................................... Effects on Human Adaptation ............................ Effects on the Archaeological Record. . . . . . . . . . . . . . . . . .. Stratigraphy and Chronology of Fluvial Sequences ....... Trinity and San Jacinto River Valleys ................... Galveston Bay ............................................. Fluvial Complexes Beneath Galveston Bay ............ Sea Level Rise: Processes and Chronology ................ Effects on Human Adaptation ............................ Effects on the Archaeological Record ................... 3. Hunter-Gatherer Adaptations to Southeastern Coastal Texas, The Galveston Bay Region, 10,000 - 1700 B.C................................................. Introduction ..................................................... Overview of the Regional Database .......................... Major Problems ................................................ Research Needs and Goals ............................... A Model of Paleoindian and Archaic HunterGatherer Adaptation. . . .. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Adaptation to Changing Coastal and Estuarine Habitats and Resources.. .. . .. .. . .. .. .. . . .. .. .. .. .. .. .. ... Cultural Property Types ....................................... Resource Characteristics and Criteria for Evaluation ...... Stresses on the Resources Base .............................. Treatment Goals, Objectives, and Tools .................... 4. Effects of European Contact on Native Populations in Southeastern Coastal Texas, The Galveston Bay Region, A.D. 1529 - 1850 ................. Introduction ..................................................... Overview of the Regional Database .......................... Major Problems ................................................ Patterns of European/Aboriginal Interaction.. . .. .. . .. .. .... The Results of European Contact and the Archaeological Record .................................... Cultural Property Types ....................................... Resource Characteristics and Criteria for Evaluation ...... Stresses on the Resources Base .............................. Treatment Goals, Objectives, and Tools .................... Special Problems and Suggested Study Units .............. 5. Navigation and Maritime Uses of the Galveston Bay Region During the Historic Period .......................... Introduction ..................................................... Overview of the Regional Databaseand ..................... Major Problems and Research Needs and Goals ........... Navigation in the Galveston Bay Region: 1529 to the Present. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . .. . .. Navigation and Commerce prior to 1800 to 1861 ..... Mexican Texas 1821-1835 and Anglo American Settlement.. .. .. .. .. .. .. .. .. .. .. . .. .. . .. . . .. . .. .. .. .. . .. .... Republic of Texas 1836 - 1846.......................... State Hood 1846 - 1860 .................................. iv
3-22 3-25 3-26 3-27 3-28 3-39 3-43 3-47 3-50 3-53
3-57 3-57 3-57 3-59 3-60 3-61 3-62 3-63 3-63 3-64 3-64
3-65 3-65 3-67 3-70 3-71 3-73 3-74 3-74 3-75 3-75 3-75 3-78 3-78 3-78 3-80 3-81 3-83 3-87 3-93 3-98
Recommended Procedures for the Protection of Cultural Resources in the Galveston Bay Area...................... A Research Design for the Management of Cultural Properties in the Galveston Bay Area.... .... .. .. .. .. .. .. .. .. .. .. .... Research Themes ................................................... Buried and Submerged Cultural Resources ..................... Archaeological Sites. . . . . . . . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . .. . . .. Shipwreck Sites ................................................ Terrestrial Archaeological Sites.. . .. .. .. .. .. .. .. . . .. . .. .. . .. .. ....
6-9 6-11
6-12 6-13 6-13 6-15 6-18
REFERENCES
..................................................................... R-l
APPENDIX 1:
National Historic Preservation Act of 1966 .......... Al-l
APPENDIX 2:
Army Regulation 420-40 ................................. A2-1
APPENDIX 3:
36 CRF Part 800: A3-1
Protection of Historic Properties
APPENDIX 4
Abandoned Shipwreck Act of 1987 and NPS Guidelines ........................................... A4-1
APPENDIX 5:
National Register Bulletin 20 ........................... A5-1
APPENDIX 6:
Programmatic Agreement. ...........•.................... A6-1
APPENDIX 7:
Annotated Listing of Shipwrecks in the Galveston Bay Area .............................. A7-1
vi
the management of the property. Any such contract will contain tenns and conditions necessary to protect the interests of the United States as well as insure adequate preservation of the historic property. The development of a HPP satisfies these responsibilities for a Federal agency such as the COE. Guidelines for the fulfillment of these responsibilities have been developed by the National Park Service and the Advisory Council on Historic Preservation. In recognition of these responsibilities, the Department of the Army has also developed its own guidelines for branches such as the COE. Anny Regulation 420-40 (Historic Preservation) prescribes management responsibilities and standards for the treatment of historic properties. It also presents a format and suggested contents for the development of a Historic Preservation Plan (HPP) in consultation with the ACHP and the appropriate SHPO. The guidelines established in Regulation 420-40 (presented as Appendix B) were followed in the development of this HPP. The Archeological Resources Protection Act of 1979 (P.L. 96-95) was designed to protect cultural resources on public or Indian lands. This law defines the prohibited activities (excavation, removal, damage, or defacement) on public lands and the associated criminal penalties that are enforced by this law. This act requires a pennit for any excavation or removal of archaeological resources from public or Indian lands which is not sponsored by the Federal agency. Any such excavation must be of a scientific nature and all resources removed remain the property of the Federal government. The permit granting authority usually belongs to the land manager responsible for the property.
Summary of Geological History and Natural Setting of Galveston Bay A considerable amount of research has been directed at the geology and natural environment of the Galveston Bay area such that it is reasonably well known. The following section presents a summary of the area's natural setting and its geological history pertinent to gaining an understanding of the major geological and geomorphic features and how they have developed. More detailed discussions on specific aspects of the geology and geomorphology of the study area are provided in following chapters. In particular, these later discussions view
2·7
the geology and geomorphology of the bay area from the concept of "allostratigraphy," an approach which seems to have particular utility in the study of archaeological phenomena_ Galveston Bay, and its associated smaller lakes and bays, is located in the Gulf Coast Province in the upper Coastal Zone of Texas, a region characterized by "several active, natural systems of environments--f1uvial and deltaic systems, marine barrier-strandplain-chenier systems, [and] bay estuary-lagoon systems .. __ " (Fisher et al. 1971 :11)_ Additionally, relict features representative of similar systems active during the Pleistocene are found throughout the region_ Today, a nearly continuous series of marginal, estuarine embayments separated from the Gulf of Mexico by barrier islands and spits characterize the Texas coastline_ The Galveston Bay complex is the largest of these estuarine systems with an area of about 1680 square kilometers (see Figure 1-1, Figure 2-1)_ This estuarine complex consists ofa roughly T-shaped embayment composed of five major elements known as East, Galveston, San Jacinito, Trinity, and West bays_ A large barrier island, Galveston Island, and major spit, Bolivar Peninsula separates the Galveston Bay complex from the Gulf of Mexico (Lankford and Rehkemper 1969:1; White et aL 1985)_ The Galveston Bay bay-estuary-Iagoon system is Holocene in age, created as rising sea levels have flooded older, incised stream valleys in the past 10,000 years or so_ The higher terrain surrounding the bay complex consists of the remains of Pleistocene-age fluvial-deltaic systems, the upper portions of which is termed the Beaumont Formation or Beaumont Terrace (Aronow, 1971; Bernard 1950; Bernard and LeBlanc 1965; Fisk 1944; Saucier 1974,1977)_ Two factors have been the primary controls on producing the current geometry of the Holocene and Pleistocene deposits in the region: 1) inland uplift and seaward subsidence and 2) glacial cycles of the Pleistocene and resultant changes in sea level (Lankford and Rogers 1969:2)_ Fluctuations in sea level have produced the most dramatic impacts on the landscape and these are considered in some detail below. Galveston, San Jacinito, and Trinity bays compose the vertical segment of the "T" of the modern estuarine complex which extends inland about 48 km perpendicular to the coast. This estuarine-bay complex is characterized by a mixing of marine and fluvial processes described in detail by Fisher et aL (1972) and White et aL (1985). The dominance of either system is a function of an array of factors, including river discharge, tidal interchange, water depth, and the location along the axis of the bays between the river mouth and tidal inlet. The Trinity and San Jacinito rivers and other small streams discharge into the heads and along the flanks of these bays. Two shallow, flat-bottomed bays, Galveston and Trinity bays, with an 2-8
1 .~.....
)
.
..
"'~~.
c,.J..c,
,0 ,,"?-
\
La porte
0" ,<:)
~
,,"
c,
"o(j~/,
t
()
'" r'" 1>
--
.J..?"'
--~.
'0
0-<::-
(,,0 ,,"?-
0°
'" '"
~- ~
,Q."
$"
.~
~
i?
:;
.,: 0'<
~
fY
~
o
5
'0 Scale .n miles
Figure 2.1.
Major natural systems in the Galveston Bay area.
20
average depth of 2 to 3 m comprise the bulk of this estuarine-bay complex. These bays were once separated by the east-west trending Red Fish Bar. The bottoms of these bays deepen sharply immediately adjacent to their margins, except for the gently sloping bottoms of the northern heads of San Jacinito and Trinity bays (Lankford and Rehkemper 1969:1-2; Rehkemper 1969a:6-12). Currently, the Trinity River Delta is filling in the head of Trinity Bay. It is a modern, typical bayhead delta dissected by numerous distributary and tidal channels and covered by fresh to brackish marsh. This modern feature has a deltaic plain that covers an area of about 15.5 square kilometers on the eastern side of the Holocene fluvial-deltaic plain of the Trinity River. The construction of this deltaic plain has isolated part of Trinity Bay to form Lake Anahuac. Behind the modern deltaic plain, the progradation of older deltaic plains and their river courses have formed a fluvial-deltaic plain about 10.5 km wide over the last 3,100 years (Rehkemper 1969a:9; Aten 1983). Unlike the Trinity River, the other major stream flowing into the Galveston Bay system, San Jacinito River, lacks a mappable delta at its head because of its small sediment load. Apparently, there has been some accumulation of sediment and resulting shoaling in San Jacinito at the mouth of San Jacinito River. However, extensive dredging, associated dredged material disposal on its banks and islands, and over 4 m of subsidence have obliterated any bayhead delta that may have existed (Lankford and Rehkemper 1969: 1-2; Rehkemper 1969a:9). The cap of the "T" is oriented parallel to the coastline and consists of East and West bays. These bays constitute a coast-parallel lagoon measuring approximately 88 km long and 1.5 to 6.5 km wide. They are quite shallow and flat with an average depth of about 2 m. Shell reefs oriented perpendicular to the axis of the lagoons are scattered throughout both bays. East and West bays are strongly influenced by marine processes and gulf waters through tidal inlets and hurricane overwash. Along both bays, fresh to brackish water marshes extend as far as 6.5 km inland from their shoreward edge. The lagoons and associated marshes are strips of the adjacent Pleistocene coast-parallel terrace which have been flooded by the Holocene transgression and separated from the Gulf of Mexico by the development of Galveston Island and Bolivar Peninsula (Lankford and Rehkemper 1969:1-2; Rehkemper 1969a:6-12). Galveston Island and Bolivar Peninsula separate the Galveston Bay complex from the Gulf of Mexico. This barrier island and spit complex averages about 1.5 to 2.5 km wide and 2-10
has a straight or smoothly arcuate gulf shoreline. Its lagoonal shoreline is highly irregular because of a series of washover deltas. old tidal inlets, storm passes, and distal ends of recurved spits. Well-preserved ridge and swale topography formed by the lateral progradation of this barrier island and spit complex characterize the surfaces of both Galveston Island and Bolivar Peninsula. The ridges and swales parallel the gulf shoreline. Three tidal inlets breach the Bolivar-Galveston barrier complex. These inlets provide limited hydrologic communication between Galveston Bay and the Gulf of Mexico. (Lankford and Rehkemper 1969:2-3; Fisher et al. 1972). The largest of these is Bolivar Roads, a natural pass between Galveston Island and Bolivar Peninsula which connects Galveston Bay with the Gulf of Mexico. The Galveston Bay complex has been extensively modified in a number of ways in historic times. First, over the last century, channel dredging and attendant formation of shallow spoil banks and islands, e.g., the Houston and Texas City ship channels, has significantly modified the natural setting and conditions of the Galveston Bay complex. Second, dredging and the removal of oyster shell over the past century has locally deepened parts of this estuarine-bay complex by as much as 2.5 to 3 m. Third, the removal of groundwater for industrial and urban usage has caused as much as 1.5 m of subsidence at the head of San Jacinito Bay to less than 0.2 m of subsidence within East Bay between 1943 and 1975. This subsidence has significantly increased local rates of shoreline erosion and land loss. Finally, urbanization and dam construction has significantly changed the volume and types of sediment and dissolved solids and gases being delivered by the Trinity and San Jacinito rivers to the Galveston Bay complex. (Rehkemper 1969:11-\3; Fisher et al. 1972; White et al. 1985). The coastal plain surrounding the Galveston Bay complex consists of two major geomorphic terraines. One terraine consists of broad, coast-parallel terraces. These terraces are composed ofrelict alluvial and deltaic plains of Late Quaternary fluvial systems ancestral to the Brazos, Trinity, and San J acini to rivers and a coastal sand ridge system known as the Ingleside sand ridge. The other terraine consists of coast-perpendicular fluvial valleys of these rivers entrenched into these relict systems which have active floodplains flanked by a series of fluvial terraces. As noted earlier, the upper segment of these Late Quaternary alluvial and deltaic plain terrace features are identified as either Beaumont Formation or Beaumont Terrace. There is some controversy over the age of the Beaumont, but recent interpretations argue that deposition of this formation began during the Sangamon Interglacial (prior to 80,000 B.P.) and the uppermost deposits are probably associated with the Mid-Wisconsinan, Farmdalian Interstadial (Saucier 1977). 2- 11
A variety of relict deltaic and fluvial features of the Beaumont can be seen at the surface in the vicinity of the study area. As shown in Figure 2·2, a number of distributary channels which represent ancient fluvial systems can be seen in the vicinity of Galveston Bay. These relict channels represent various meander belts of the Pleistocene-age Brazos and Trinity rivers which may be contemporary to the Deltaic Plain phase (ca. 30,000 to 25,000 B.P.) (Aronow 1971; Aten 1983b, Weinstein 1991b:5). These features were apparently formed during the Farrndalian Interstadial, when sea level was at or near its present level. Subsequent to the Farmdalian high-sea stand, about 25, 000 years ago, sea level began to drop in the wake of Woodfordian glaciation (Fisher et al. 1972:11). Approximately 18,000 years ago, sea levels reached their lowest levels, about 450 ft below the present sea level (Saucier 1981; Saucier and Fleetwood 1970). In response to the fall in sea level, Pleistocene rivers and streams initially extended their floodplain seaward. However, the fall eventually became so dramatic that progradation could not be maintained and streams began to incise their channels into the underlain Pleistocene deposits (Fisher et al. 1972:13). Following the glacial maximum, sea levels began a slow rise, eventually, inundating most of the valleys formed during the low stand. This inundation resulted in the eventual filling of the old river valleys as a progression of deltaic and estuarine systems developed within the valleys and wide-spread deposition occurred. Most of present·day Galveston and Trinity bays constitute the filled Pleistocene age valleys of the Trinity and San Jacinto rivers. A slight reversal in sea level rise occurred between 11,000 and 10,000 years ago (Aten 1983b). This drop was less dramatic than the previous fall, but it was sufficient to cause entrenchment in coastal streams (Weinstein 1991b:5). After about 10,000 years ago, sea level began to rise, with most arguing that it reached its present level about 40-00 to 3500 years ago (Aten 1983b; Lankford 1971; Pearson et al. 1986; Rehkemper 1969). As sea level rose, the valleys of the various rivers in the Galveston Bay area became drowned, and the Galveston Bay estuary developed. Prior to its inundation, the entrenched valley of the Trinity River followed a meandering course through what is now Galveston Bay. The San Jacinto River joined the Trinity between what is presently Smith Point and San Leon, the combined rivers extended southward through the Bolivar Roads area across the Continental Shelf to the Gulf. Lankford and Rogers
2·12
ANCIENT
\ ANCIENT \
BRAZOS
N
j
t 'I
4
"4T~1f
Figure 2-2.
1
,
'0
...I,U
Locations of Pleistocene deltas of Brazos and Trinity rivers (after Barton 1930).
2-13
(1969:41) indicate that the valley was fairly narrow, about 6 miles, and that "Pleistocene outliers isolated by meandering channel erosion stood as hills and ridges above the valley floor." They indicate that the combined Pleistocene age Trinity-San Jacinto had incised its channel to maximum depth of about 130 ft at the entrance to Galveston Bay, and that the Pleistocene Trinity channel reaches a depth of 60 ft at the upper end of Trinity Bay beneath the modern Trinity Delta (Figure 2-3) (Lankford and Rogers 1969:41). Others suggest that the valley entrenchment was deeper. For example, Fisher et a!. (1972) indicate that cores taken in the bay indicate the depth to the now buried and filled relict valley is as much as 260 ft. As sea level rose and the valley of the Trinity-San Jacinto was inundated, large "point bar sand bodies and extensive overbank mud sheets were deposited" within the valleys by the meandering river (Fisher et a!. 1971: 13). Considerable areas of these Holocene meander belt deposits are exposed in the submerged, but as yet unfilled, portions of the Trinity and San Jacinto valleys. By 8,000 years ago, rising sea levels were forming an estuary in what is now Galveston Bay. At that time, infilling of the smaller streams tributary to the bay by locally derived sediments started to occur, a process which continues along some to this day (Lankford 1971:362). Concomitant with the infilling was the development of bars and barriers across the mouths of tributary valleys and within parts of Galveston Bay (Lankford 1971:362363). During the past 4500 years or so, since sea level reached its approximate present level, several changes have occurred in the natural systems and features in the Galveston Bay system. As Fisher et al.( 1971: 14) note these are: (1) Deeper parts of the Trinity and San Jacinto estuaries began to fill with sediment eroded from the walls of drowned valleys; (2) the Trinity and San Jacinto bay-head deltas began their slow filling of the uppermost part of the estuaries; (3) headward erosion by short streams continued within Pleistocene interdistributary areas where significant compaction of mud is occurring; (4) East Bay and West Bay developed as elongate lagoons behind Bolivar Peninsula, which grew southwestward by spit deposition and shoreface deposition from eroded deltaic headlands near High Island, and behind Galveston and Follets Islands, which developed as coalescing, exposed offshore bars that also grew seaward by shoreface deposition; and (5) marshes encroached upon subsiding Pleistocene delta deposits and bay areas that were filled
by
storm-washover
fans 2·I~
and
bay-margin
deposits.
f l>
I
!
!
....uft ....
• 1
it
_11..,.' L ... ,.... HUY"~ ""'_u c.O
Figure 2-3.
,_nor. .., ... ~
Contour map of top of Pleistocene in Galveston area. Pleistocene deposits after Henry, 1956.
2-15
Climate The climate of the Galveston Bay area represents the only truly humid bay system within coastal Texas (Shew et al. 1981:12). Similarly, Galveston Bay receives the most rainfall of any area along the coast, with most precipitation falling during the spring and summer months, eventually reaching a peak during late summer. This peak is then followed by a rapid decrease in October, resulting in a relatively low rate of precipitation throughout the remainder of the fall and winter months (Shew et al. 1981: 13, Fig. 1). Temperatures in the Galveston Bay area are moderated by winds from the Gulf of Mexico, causing mild winters and relatively cool summer nights (Wheeler 1976:2). Mean annual air temperature is 20.5 degrees Celsius at Houston (NOAA 1973, cited in Shew et al. 1981:13), with a maximum in August and minimum in January. Prevailing winds are from the south and southeast, "except in January when frequent high pressure areas bring invasions of polar air and prevailing northerly winds (Wheeler 1976:2). About one-fourth of the days each year are clear, with October having the greatest number of clear days. Cloudy days are frequent from November through May, and partly cloudy days are frequent from June through September (Wheeler 1976:2).
Biota The Galveston Bay system as a whole has an average salinity of 17.3 parts per thousand, the lowest salinity rate of any bay system on the coast of Texas (Martinez 1975 [cited in Shew et al. 1981 :29]). This low salinity is largely a result of the great quantity of freshwater discharged in to the system by the Trinity and San Jacinto rivers in combination with the high volume of local runoff from smaller tributary streams produced by the region's high precipitation rates. This is significant, because several faunal species thrive best in areas where salinity is relatively low and high levels of freshwater inflow occurs. Examples include Rangia clams, an important dietary resource to aboriginal populations, and other species which may have played an important part in the aboriginal diet, such as the blue crab (Callinectes sapidliS) and white shrimp (Panaeus
setiferus), although evidence for prehistoric exploitation of these species is minimal. Several recent summaries of the flora and fauna of the Galveston Bay and Trinity River delta areas have appeared (Dillehay 1975a; Fisher et al. 1972; Gilmore 1974; Mercado-Allinger et al. 1984; Shew et al. 1981; Stokes 1985). In this ovelview it is sufficient simply to provide a brief review of the more important species known from the area. Mercado-Allinger et al. (1984:3-4),
2 ·16
citing data supplied by Dillehay (1975a:166-178), Fisher et al. (1972:70), Gilmore (1974:22), and the Houston Audubon Society and Preservation of Armand Bayou Committee (1974), offer a summary of the species of the region and their study is quoted below: Although the natural vegetation of the project region has been severely altered by twentieth-century agriculture and urbanization, it is possible to define three vegetational assemblages which may have occurred in the area prehistorically: (I) in the Pleistocene uplands, a prairie grassland with species such as little
bluestem (Schizachyrium scoparium), big bluestem (Andropogon gerardi), indiagrass (Sorghastrum avenaceum), and eastern gramagrass (Tripsacum dactyloides); (2) along portions of Clear Creek, the south side of Clear Lake, and especially the north side of Clear Lake, a fluvial woodland with plants such as pecan (Carya illinoensis), willow oak (Quercus phellos), water oak (Quercus nigra), overcup oak (Quercus lyrata), bottomland post oak (Quercus simiUs), water hickory (Carya aquatica), southern red oak (Quercus falcata), American elm (Ulmus americana), Texas sugarberry (Cetis laevigata), palmetto (Sabal minor), American hornbeam (Cmpinus caroliniana), red mulbelTY
(MOrliS
rubra), American beauty berry (Callicarpa americana), flatwood plum (Prunus umbellata), possum-haw (/lex decidua), greenbriar (Smilax spp.), and grapes (Vitis spp.); and (3) along parts of Clear Creek and its tributaries, a brackish to freshwater marsh with plants such as coastal sacahuista (Spartina spartinae), marsh hay cordgrass (Spartina patens), big cordgrass (Spartina cynosuroides), rushes (Juncus spp.), bulrushes (Scirplls spp.), and cattail (Typha latifolia). Likewise, present-day faunal assemblages are only partly representative of those that existed prehistorically. It is clear, however, that the environments described above supported a number of faunal species that were important prehistorically. Animals which occur most commonly in the archaeological record of the region .. .include brackish water clam (Rangia cuneata and R. flexuosa), oyster (Crassostrea virginica), gar (Lepisosteus sp.), catfish (lctalurus sp.), freshwater drum (Aplodinotus grunniens), black drum (Pogonias cromis), sheepshead (Archosarglts probatcephallts), various turtles, alligator (Alligator mississippiensis), various waterfowl, bison (Bison bison), whitetail deer (Odocoileus virginianus), black bear (Ursus americanus), raccoon (Procyon 1010r), opossum (Didelphis marslIpialis), muskrat (Ondatra zibelhica),
2-17
mink (Mustela vison), skunk (Mephitis mephitis), and rabbits (Sylvilaglls
floridanus and S. aquaticus).
Summary of Cultural History of the Galveston Bay Area This section is to place the cultural history of the Galveston Bay area in its appropriate cultural and chronological framework. This section contains only a summary of the region's culture history, not an in depth treatment of the topic and its intent is to provide CaE managers with a handle on the scope and history of cultural resources research within the Galveston Bay area as well as a basic understanding of the prehistoric and historic cultural setting. As noted in Army Regulation 420·40, the overview serves to "determine if the installation [in this case the Galveston Bay Navigation System] has or is likely to have historic properties that may be adversely affected by Army undertakings." This discussion also serves as a platform for the more detailed discussions on specific aspects of the area's culture history contained in the historic contexts presented in the next chapter. A considerable amount of archaeological and historical research has been undertaken in the Galveston Bay region from which information can be drawn. Much of the information presented in this summary has been taken from studies made by Aten (1979; 1983b), Gadus and Howard (1990), Howard et al. (1991), Mercado· Allinger, et al. (1984), Weinstein (1991b),and Weinstein et al. (1988; 1989), to name a few.
Previous Archaeological Research There have been numerous archaeological investigations in the Galveston Bay area of Chambers, Harris, and Galveston counties, many of them funded by the Galveston District CaE. This abundance can be attributed largely to the appearance of federally mandated cultural resource surveys, site assessments, and data-recovery projects within the last 20 years, in conjunction with the boom in construction related to residential development in and around the Galveston Bay area. One bias that should be noted is that the emphasis of most of the previous research has been on prehistoric sites and the prehistoric period, relatively little attention has been paid to non-aboriginal cultural resources. The one exception to this is in the area of historic shipwrecks; a topic of particular importance to this HPP. A number of studies, primarily remote-sensing surveys in and in the vicinity of Galveston Bay, have been undertaken specifically to locate shipwreck remains. These studies are noted in the following discussions. The area under consideration in this review includes all of Chambers and Galveston counties, Harris County east of Houston, Galveston Island and Bolivar Peninsula, 2·18
CHAPTER 3
HISTORIC CONTEXTS
The Historic Context Concept One element in the management of cultural resources, is the "historic context". The term historic context refers to the grouping of resources defined by theme, geographic limit and chronological period. The U.S. Department of the Interior (USDI) has developed specific statements about how historic contexts are to be used in the preservation planning process. Their Standards and Guidelines for Archeology and Historic Preservation state: Decisions about the identification, evaluation, registration, and treatment of historic properties are most reliably made when the relationship of individual properties to other similar properties is understood. Information about historic properties representing aspects of history, architecture, archeology, engineering, and culture must be collected and organized to define these relationships. This organizational framework is called an "historic context." The historic context organizes information based on a cultural theme and its geographical and chronological limits. Contexts describes the significant broad patterns of development in an area that may be represented by historic properties. The development of historic contests is the foundation for decisions about identification, evaluation, registration, and treatment of historic properties [USDI n.d.]. The Guidelines go on to state that a series of preservation goals should be systematically developed for each historic context. These goals are to be prioritized and integrated into the overall preservation planning effort for a given geographic area. The following section presents discussions on five basic historic contexts deemed pertinent to the study and management of cultural resource properties within the Galveston Bay Navigation System. These discussions provide relevant historical information pertinent to understanding the position of the cultural resources within these contexts. The primary goal in developing
these historic contexts is to assist cultural resources managers in developing appropriate priorities and establishing strategies for research and preservation activities related to an aspect of historic navigation within the Galveston Bay Navigation System. The Texas Antiquities Committee began the identification of the resource base of historic shipwrecks in Texas with the initiation of computerized shipwreck reference file in 1972 (Arnold 1982). Although not computerized, TARL has a fairly complete collection of site forms for other cultural resources. These files serve as an important data base for developing historic contexts and in establishing and prioritizing preservation goals. This study presents five historic contexts. These contexts are as follows: Late Quaternary Geology and Environments, Paleogeography, and the Archaeological Record Hunter-Gatherer Adaptations to Southeastern Coastal Texas, The Galveston Bay Region, 10,000 B.C.- A.D. 1700. Effects of European Contact on Native Populations in Southeastern Coastal Texas, The Galveston Bay Region, A.D. 1529 - 1850. Navigation and Maritime Uses of the Galveston Bay Region During the Historic Period.
Historic Contexts 1. Late Quaternary Geology and Environments, Paleogeography, and the Archaeological Record Introduction and Perspectives Understanding the geological history and the scope of environmental change of the Galveston Bay region since the arrival of human populations in the area is prerequisite to understanding the nature of archaeological site distributions and occurrences. This historic context provides a comprehensive discussion on the geological and environmental history of the study area, emphasizing those aspects that relate to the distribution of human populations over time and the occurrence of archaeological sites. The primary goal is to establish a framework which cultural resource managers eventually can use to predict the potential lateral and vertical distribution of prehistoric cultural resources within the area of the Galveston Bay Navigation System and adjacent coastal plain. Second, in addition to predicting the distribution
3-2
of archaeological deposits, assessments will be made concerning the preservation potential of archaeological deposits. Finally, this review of the geomorphology and Quaternary geology of the Galveston Bay area is intended to help guide future geomorphological and geoarchaeological research within the Galveston Bay region. The basis of the following discussions relies on the concept of allostratigraphy in the definition of geological deposits. It is felt that the use of allostratigraphy as a guiding concept in geological interpretation will ultimately prove to be of significant value in the development of models of site archaeological site distributions. In light of this, a rather detailed discussion of allostratigraphy is presented.
Allostratigraphy. The use of allostratigraphy in Gulf Coast geoarchaeological research is new application for this stratigraphic technique. Allostratigraphy is a methodology that can be used to interpret seismic profiles, foundation borings, and other typical offshore geologic data for geoarchaeological studies within the Galveston Bay area and the adjacent continental shelf. Allostratigraphy is a very important and useful, but often either ignored or misunderstood, stratigraphic tool for geoarchaeological research within the Texas Coastal Plain. Allostratigraphy uses the "allofonnation" as the basic unit of analysis. An allofonnation is a mappable body of sedimentary rock or unconsolidated sediments that is defined and identified on the basis of bounding discontinuities. A bounding discontinuity can be either an erosional unconformity or a construction surface (North American Commission on Stratigraphic Nomenclature 1983:865-868). Using this methodology, it is possible to define mappable geomorphic surfaces and sedimentary units that can be used to predict the age, preservation potential, and potential for the occurrence of archaeological deposits within Holocene sediments of Galveston Bay Area and the remainder of the Texas Coastal Plain and Continental Shelf. In order to accomplish the goal of this Historic Context, an extensive review of the Late
Quaternary geology and geomorphology of the Galveston Bay complex was conducted for two reasons. First, although Galveston Bay has been the object of intensive sedimentological and archaeological research, a proper stratigraphic framework is lacking and had to be constructed, because it was lacking. In part, this framework was lacking, because only recently has a methodology, allostratigraphy, been refined by Autin (1989, 1992) Autin et al. (1990, 1991), and Bhattacharya (1992) which has utility for the naming and mapping of Late Quaternary sedimentary deposits within Southeast Texas. Also, the lack of this framework results from the sedimentological orientation of geological research which resulted in the exclusive use of another stratigraphic methodology, sequence stratigraphy, which is powerful way to interpret sedimentary sequences context of relative sea level change, but is inadequate for the purpose of 3-3
defining and mapping these deposits. Finally, contradictory interpretations by Aten (1983), LeBlanc (1991a), Gagliano (1991), Thomas (1991) and others concerning the alluvial deposits commonly called the "Deweyville Terraces or Formation" had to be resolved using a testable hypothesis. Although it was initially thought that certain researchers were wrong and others right, it was discovered that each researcher observed a fundamental part of what became a proposed solution to this problem. With this framework provided by the geological and geomorphological data, the progress of environmental change in the Galveston Bay region since the coming of human populations can be assessed. Further, with this diachronic model, shifting patterns of human adaptation can be more accurately examined as can the resultant archaeological record. The Holocene and Late Pleistocene sediments within many parts of the United States exist as only a thin veneer of sediment or topsoil overlying either unconsolidated sediments or bedrock that predates the human occupation of North America. As a result, these deposits typically are restricted to a thin, relatively uncomplicated, layer of alluvium, colluvium, or residuum. The stratigraphy of such deposits, for the most part, can be described in simple stratigraphic terms without recourse to the complex assemblage of stratigraphic methodology employed normally by geologists. However, the Texas Coastal Plain consists of large coast-parallel terraces comprised of delta, alluvial, and other plains that are the surface expression of thick, unconformity-bounded, sequences of Pleistocene and Holocene deltaic, fluvial, and eolian sediments (Figure 3-1). Within the Galveston Bay region, the Trinity, Brazos, and San Jacinto rivers and the shifting shoreline of the Gulf of Mexico have created thick and intricately stacked sequences of Pleistocene and Holocene shallow marine, estuarine, coastal, deltaic, and fluvial sediments. As a result, multiple systems of independent and formally defined stratigraphic classification systems, e.g., lithostratigraphy, allostratigraphy, and pedostratigraphy, are an essential part of the interdisciplinary research approach used to describe, correlate, and interpret the complex succession of Pleistocene and Holocene sedimentary deposits which have accumulated within the Texas Coastal Plain (Autin et al. 1990,1991; Barton 1930; Bernard et a1.l970; DuBar et al. 1991; Winker 1979). If this stratigraphic analysis is going to be of use to both archaeological research and the
management of known and unknown archaeological properties within the Texas Coastal Plain
3-4
TRINITY BAY
EXPLANATION HOLOCENE SYSTEMS
Fluvial/deltaic sand and mud Barrier island sond ond shell Marsh pI EISTOCENE SYSTEMS
~
Fluvial/deltaic mud "no sand .
N
I o I
o
.
5
•
.
.
10
"
15
20
sand (lnQfeside) HUMAN ACTIVITIES 20mi
15
•
Barrier /slronaptaln
•
25
•
Figure 3-1. Generalized geologic map of the Galveston Bay system Morton 1986:8).
3-5
Spoil and mode land
•
'(Hm
(source:
Paine and
in general and the Galveston Bay area in particular, the stratigraphic nomenclature must be more precisely applied than it has been in the past. For example, criteria, e.g., subsurface stratigraphy and soil geomorphology, in addition to elevation and morphology need to be used to map the distribution of geomorphic surfaces and the sedimentary deposits that underlie them. In addition, a geomorphic surface should not be assumed automatically to have the same age and distribution as the sedimentary strata that underlie it. Furthermore, the different types of stratigraphic units, geomorphic surfaces and formations, should be recognized as separate entities and not hybridized as many previous studies of the geomorphology and geology of the Texas Coastal Plain have consistently done. Finally, the use of models that propose simple one·to·one correlations between the formation of individual paleosols, coast-parallel terraces, delta plains, alluvial plains, and formations with glacial cycles or sea level fluctuations to date or explain the origin of these stratigraphic units should be avoided (Autin et al. 1990, 1991). Five types of stratigraphy, namely morpho stratigraphy (geomorphic surfaces), lithostratigraphy, allostratigraphy, pedostratigraphy, and chronostratigraphy, are important to the definition, correlation, and dating of Pleistocene and Holocene deposits within the Texas Coastal Plain. Three of these types of stratigraphy, geomorphic surfaces, lithostratigraphy, and allostratigraphy, are specifically important to understanding the geomorphology of the Texas Coastal Plain. Two of these types of stratigraphic units, chronostratigraphy and pedostratigraphy, are discussed by Autin et al. (1990, 1991) and the North American Commission on Stratigraphic Nomenclature (1983).
Geological Setting Geological Features and Geomorphic Processes Within the Texas Coastal Plain, a variety of geomorphic surfaces can be recognized. For geoarchaeological and geomorphological research, these geomorphic surfaces can be classified as coast-parallel terraces, fluvial terraces, delta plains, and alluvial plains. Alluvial plains can be further subdivided into meander belts and backswamps. Terraces. A terrace is a relatively flat geomorphic smface that is separated from adjacent geomorphic surfaces by a constructional or erosional scarp. Within the Texas Coastal Plain, two different types of terraces, coast-parallel and fluvial terraces, can be recognized. The coast-parallel terraces are low-relief, gulf-ward sloping geomorphic surfaces separated by low, irregular coast-parallel scarps; they parallel the coast and can be tens of kilometers wide. 3·6
These surfaces represent relict coastal plains consisting of relict, Pleistocene alluvial, delta, and strand plains (Barton 1930; Bernard et al. 1962). A fluvial terrace is a narrow geomorphic surface of purely fluvial origin that parallels an adjacent, modern stream or river. Typically, a fluvial terrace trends perpendicular to the modem coastline and lies between the walls of an entrenched valley and modem meander belt that occupies it. Many investigators studying the Texas Coastal Plain confuse geomorphic surfaces, e.g., fluvial and coast-parallel terraces, with the sediments that form them. As a result, stratigraphic units, e.g., terrace and formation, and names, e.g., Prairie; Deweyville; Montgomery; and Williana, are used interchangeably to refer to both geomorphic surfaces and the underlying sedimentary strata.
Plains. The Prairie and Lissie coast-parallel terraces are complex geomorphic surfaces which consist of an assemblage of smaller constructional geomorphic surfaces formed by the periodic aggradation of fluvial systems and the periodic progradation of either deltaic or strandplain systems. A constructional geomorphic surface, formed by either an active or relict fluvial or deltaic system, is designated as a "plain." Within the Louisiana Coastal Plain, Autin et al. (1988, 1991) demonstrate that the coast-parallel terraces consist of assemblages of relict alluvial, delta, and strandplains. Mapping by Van Siclen (1985, 1991) shows that this is apparently true of the coast-parallel terrace of the Texas Coastal Plain within the project area. Within the older coast-parallel terraces, surficial processes have either greatly modified or obliterated the constructional surface that once formed the surface of the older two coastparallel terraces. An alluvial plain is a geomorphic surface that consists of the active meander belt of a river or stream and its associated flood basins and abandoned meander belts. A meander belt is a surface that consists of an assemblage of constructional landforms created by the meandering of a river while occupying a single course. These constructional landforms include the ridge and swale topography of point bars, natural levee ridges, crevasse splays, abandoned meander loops, and abandoned river courses. A flood basin, also called "backswamp," is an area consisting of swamp, lakes, or combination of both that comprise the low alluvial plain between meander belts (Saucier 1974:10-11). A delta plain is the constructional surface of a delta complex. A delta complex consists of the set of delta lobes fed from a common trunk channel. A delta lobe consists of a set of subdeltas and minor distributaries fed from a major distributary (Coleman and Gagliano 1964; 3-7
Frazier 1967). Recent studies, e.g., Penland et al. (1987:1692), within the Mississippi River Delta region have confused geomorphic surfaces and subsurface sediments by incorrectly extending the definition of a "delta plain" to include both the surface of the delta and sediments that form this surface. By definition, a plain of any type is strictly a geomorphic surface consisting of level or nearly level land. Also, the term "plain" lacks any reference to the deposits that form it. Therefore, in this report, the term "delta plain" is reserved solely for the subaerial, constructional surface of a delta complex. Lithostratigraphy. The basic unit of lithostratigraphy is the formation. A formation
is defined as a mappable body of sedimentary, volcanic, metamorphic, or plutonic rock which can be distinguished and delineated on the basis of its physical character, lithology, and stratigraphic position without reference to its cultural and paleontological content or age. By definition, a formation is recognized by only the physical properties of the lithified or unlithified sediments that compose it (North American Commission on Stratigraphic Nomenclature 1983). Within the coastal plain of Southeast Texas and adjacent Louisiana, various attempts, e.g., McFarlan and LeRoy (1988) and Van Siclen (1985:531,1991:651), have been made to subdivide Pleistocene strata into formations which correlate directly with previously defined and mapped coast-parallel terraces. However, these attempts have repeatedly failed to consistently recognize formations as defined by the rules of stratigraphic nomenclature and demonstrate their correlation with the coast-parallel terraces (DuBar et al. 1991:584; Winker 1979:22-24). Typically, the formations and members defined by such studies, e.g., Bernard and LeBlanc (1965), McFarlan and LeRoy (1988:424-426), and Van Siclen (1985:531, 1991:651), fail to be true lithostratigraphic units (i.e., formations and members) because they are defined as unconformity-bounded, depositional sequences and not by any distinctive and mappable differences in gross lithologic characteristics that would characterize a true formation. Additionally, these three coast-parallel terraces lack any one-to-one correspondence with the 11 high frequency, high amplitude sea level fluctuations that have accompanied the Pleistocene glacial cycles of the last 1.8 million years (Wornardt and Vail 1991:742). Finally, the coastparallel terraces are not the simple depositional surfaces envisioned by Bernard and LeBlanc (1965), Bernard et al. (1962), Doering (1935), Fisk (1939; 1944), and others. Rather, like the High, Intermediate, and Prairie Terraces of adjacent Louisiana, the coast-parallel terraces are apparently very complex geomorphic surfaces, consisting of mUltiple alluvial and deltaic plains formed by separate periods of deltaic progradation and and fluvial aggradation (Aronow 1988:15; Van Siclen 1985:531, 1991:651). The cyclic deposition of fluvial, deltaic, and other 3-8
coastal sediments during the Holocene and Pleistocene Epochs have resulted in sediments consisting of a complex set of interbedded sands, muds, and clays_ The heterogeneous nature of these sediments generally precludes the recognition of stratigraphic units based on gross or unique lithologic characteristics (Dubaret aL 1991; Winker 1979). Therefore, lithostratigraphy is generally an unusable stratigraphic tool within the Plio-Pleistocene sediments of the Texas Coastal Plain. However, on the scale of a site, lithostratigraphic units such as the Layer as defmed by Stein (1990) are useful in archaeological research.
Bounding Discontinuities. Within the Texas Coastal Plain, the four major types of bounding discontinuities that can be used to define and map allostratigraphic units are geomorphic surfaces, fluvial erosion surfaces, flooding surfaces, and ravinement surfaces. As previously discussed, geomorphic surfaces, e.g. coast-parallel terraces, fluvial terraces, meander belts, and delta plains, are the upper bounding discontinuity of depositional sequences of Late Quaternary sediments. Typically, these surfaces are plains formed by the accumulation of alluvial, deltaic, or eolian sediments which exhibit landforms indicative of the processes which formed them, e.g. Bernard and LeBlanc (1965:Figure 4 and 5) and Van Siclen (1985). However, prolonged subaerial exposure of relict constructional plains to weathering and other surficial processes will eventually obliterate any constructional landforms and form a subaerial erosional plain. Where buried intact, a geomorphic surface will be detectable by either laterally persistent paleosols or truncated weathering horizons and abrupt changes in sedimentary facies (Autin et aL 1991:Figure 4, personal communication 1991; Winker 1979:Figure 9). Complexes. Within the Texas Coastal Plain, few of these allostratigraphic units have been either adequately defined or named. In case formal stratigraphic units exist, an informal allostratigraphic unit, the "complex," is used. A complex consists of a single geomorphic surface or temporally related surfaces and associated depositional sequence or sequences. The depositional sequence consists of the deposits of one or more depositional environments and possesses distinct, regionally mappable bounding discontinuities. Typically, the complex is named for the geomorphic surface which forms part of it, although this practice is discouraged by the formal rules of stratigraphic nomenclature for the naming of formal alloformations. After it is named and described as a formal allostratigraphic unit, the use of a complex should be abandoned (Whitney J. Autin, personal communications 1990; Autin et aL 1990; 1991). Coast-Parallel Allostratigraphic Units.
A coast-parallel complex is an
allostratigraphic unit whose surface consists of a coast-parallel terrace. For example, the uppermost coast-parallel complex within eastern Brazoria County consists of a 10 to 25 m thick 3-9
stratiform body of unfossiliferous silty clay with scattered elongated bodies of sand that trend roughly perpendicular to the coast and elongate bodies of shelly sands that parallel the coast. The lower bounding discontinuity of a coast-parallel complex is the paleosol developed within the buried coast-parallel terrace of an underlying terrace or the erosional unconformity at the base of either a buried entrenched valley or meander belt. Gulfward, the fluvial facies grade laterally into coastal and marine deposits (Autin et al. 1991, Winker 1979:44, 50). A fluvial erosion surface is typically an undulating surface cut by either the entrenchment, lateral migration, or both of the thalweg of a river channel. This type of surface is an erosional bounding discontinuity that forms the base and sides of allostratigraphic units associated with fluvial terraces and meander belts. Terrace scarps are subaerially exposed edges of fluvial erosion surfaces (Figure 3-2) (Autin 1989, 1992). A flooding surface is a local disconformity either at the base of or within estuarine deposits created by a permanent rise in sea level. Within the Holocene valley fill beneath Galveston and the other coastal bays of Texas, two types of flooding surfaces, the bayline flooding and intermediate flooding surfaces, have been recognized (Figure 3-3). A bay line flooding surface is a disconforrnity which separates either fluvial or older upland deposits from overlying estuarine deposits across which there is evidence for a permanent rise in sea level. Typically, a bayline flooding surface separates marsh or swamp deposits from overlying bay sediments. An intermediate flooding surface is a disconforrnity which separates younger from older estuarine deposits across which there is evidence of an abrupt increase in water depth without any significant accompanying erosion. The intermediate flooding surfaces are marked by abrupt changes in faunal assemblages, color of sediments, geotechnical properties, and sometimes in lithology and seismic facies (Anderson and Siringan 1992: 10-11; Bhattacharya n.d.; Thomas 1990:99-100; Thomas and Anderson 1989:565). A ravinement surface is a regional marine erosional surface produced by the erosional retreat of the shoreface. The formation of this erosional surface consumes a thickness of underlying sediments of up to several meters thick, the subaerial dune and beach deposits, and the upper shoreface of the barrier island to which the shoreface belongs. Within the Galveston Bay area, where a landward moving shoreline intersects the former subaerial interfluves and other coastal headlands, the ravinement surface cuts a few meters into the sediments of the
3-10
~
'" ?O ...
Abandoned River Channels
~~5 §
Ii
"~ 10 ~ ~ 5
'--
-J.'.,,::
Pb
Pb
'-'
'""' '""'
"" o I o
b
15 \
10
-"1-Erosion~ '\
'\SF" Pb~nDffi"
"J::
~
20
5
Bounadanes
Width (Kilometers) 2 LEGEND Al = Undifferentiated Alluvium , "",."""" = facies boundary Ba = Backswamp Facies - = Bounding disconfonnity c = Channel Fill Facies T77 =Paleosol or soil (depth of Ob = Overbank Facies strips indicates development) Pb = Point Bar Facies 1
Figure 3-2. Hypothetical fluvial alloformations associated with an entrenched valley Heinrich nd).
0 3
(source:
i'====:::t:@w;;:::------ Modem Sea Level------l
~lt!~~~ LEGEND AF = Allofonnation AM = Allomember BFS =Bayline Flooding Surface IFS = Intennediate Flooding Surface RS = Ravinement Surface " " Bayhead Deltas
D
Esturaine Sediments
,,',
"
..
Figure 3-3. Allostratigraphy of estuarine sediments within the entrenched valley of the Trinity and San Jacinto rivers beneath Galveston Bay and on the adjacent Continental Shelf (modified from: Thomas 1990:Figure 2-22).
3-12
coast-parallel terraces. Where landward moving shoreline intersects the valley fill under Galveston Bay, the ravinement surface cuts several meters deeper into the softer estuarine valley fill (Numrnendal and Swift 1987:244-148; Swift 1968; Thomas 1990:211).
Fluvial Allostratigraphic Unit. A fluvial complex or alloformation IS an allostratigraphic unit composed of unconformity-bounded package of fluvial deposits lying within an entrenched alluvial valley. Its upper bounding discontinuity consists of a terrace, which might be buried or partially buried, an erosional unconformity, or a combination of both (see Figure 3-2). The upper bounding discontinuity of an allostratigraphic unit associated with an active river channel is an alluvial plain or meander belt. Both units consist of a basal bounding discontinuity, a body of fluvial sediments that lies between the bounding discontinuities, and an upper bounding discontinuity. Typically, the basal bounding discontinuity is an erosional unconformity formed by scour at the channel bottom and, at the bank, collapse of cutbank of a channel (Autin 1989; 1992). Fluvial sediments deposited by this channel overlie the basal unconformity. Generally, but not always, these sediments consist of a lower part composed of point bar sands and gravels, overlain by finer-grained and vertically accreted natural levee and overbank sediments (Walker 1984). Typically, the upper bounding discontinuity consists of both an exposed or buried fluvial terrace and an erosional unconformity formed during the formation of a younger alloformation (Autin 1989, 1992). The scarp that defines a fluvial terrace is the exposed edge of the basal bounding discontinuity (see Figure 3-2). As a result, a scarp, as reflected by differences in surface morphology, soil development, and thickness of overbank deposits, separates geomorphic surfaces of differing ages. Also, the terrace scarp separates the fluvial complexes consisting of fluvial sediments that can, but not always, differ in type and distribution of facies (Autin 1989; 1992). Two general models have been used to explain the origin of fluvial complexes and alloformations. First, many investigators, e.g., Thomas (1990) and LeBlanc (1991a) have interpreted fluvial terraces and, by implication, their associated fluvial allostratigraphic units in terms of the classic model of Fisk (1944, 1939) for the creation of coast-parallel and other terraces. Fisk (1944) concluded that a terrace is the result of fluvial aggradation followed by a period of fluvial entrenchment. This model implies that the aggradation of a fining-upward sequence in response to a rising base-level, typically a relative rise in sea level, constructs a 3·13
floodplain. The floodplain becomes a terrace when it is abandoned by the fluvial system as a result of entrenchment in response to a dropping base-level, typically a relative drop in sea level. According to this model, each fluvial terrace and its associated allostratigraphic unit is interpreted to represent the fluvial response to a single rise and fall in base level, which is commonly assumed to be sea level (Autin 1992:241). Finally, Autin (1989, 1992) and Blum (1990:80-81) have demonstrated that the formation of fluvial complexes and alloformations is the result of geomorphic processes more complex than simple changes in base levels. Within the Amite River Valley of Louisiana, Autin (1992:240) found that a temporal clustering of cutoffs initiates a period of meander belt instability. This response results from changes in one or more geomorphic influences, e.g., climate, base-level, etc., which cause an imbalance between river hydrology and sediment delivery. Because of the increased rates of channel cutoffs, the channel pattern locally straightens which favors channels avulsion over lateral accretion. Avulsion creates a new channel which truncates the older alluvium and produces the initial lateral boundaries of an alloformation. After a few decades to centuries of instability, a new stable meander belt is established with a channel pattern and slope equilibrium with the new conditions of river hydrology and sediment delivery (Autin 1992:240). Significantly, Autin (1989, 1992) and Blum (1990:80-81) demonstrate that to simply interpret all fluvial alloformations, fluvial complexes, and their terraces solely as the result of rises and falls of sea level is a grossly simplistic explanation that can be wrong as often as it is right.
Estuarine Allostratigraphic Units. A estuarine complex or alloformation is an allostratigraphic unit composed of unconformity-bounded package of fluvial deposits lying within an entrenched alluvial valley (see Figure 3-3). Landward of the shoreface of associated barrier island, its upper bounding discontinuity consists of the bottom of the bay and the delta plain of the bayhead delta filling the flooded valley. Seaward of the of the shoreface of associated barrier island, the upper bounding disconformity is the ravinement surface formed by the trangressing shoreface. The basal bounding disconformity consists of the bayline flooding surface which separates the sediments of this allostratigraphic units from those of the underlying fluvial allostratigraphic units. Subregional bounding disconformities, previously defined as intermediate flooding surfaces, occur within an estuarine deposits that lie between the bayline flooding surface and the ravinement surfaces (see Figure 3-3). An intermediate flooding surface is typically associated with correlatable seismic reflector, nearly flat stmcturally with less than 4 m of relief 3·14
on it, and pinches out against the bayline flooding surface where the structural elevations of both surfaces are equal. The intermediate flooding surface subdivides an estuarine allofonnation into into a series of discontinuity-bounded sedimentary packages, equivalent to allomembers. Each of these packages consist of eustarine-bay deposits with a wedge of bay head delta deposits at its landward edge directly underlying this flooding surface. At the seaward end of the intennediate flooding surface, it commonly truncates barrier island, spit, and tidal inlet deposits where it merges with the ravinement surface.
Upland Surfaces and Processes Fisk (1939,1944) and Fisk and McFarlan (1955) considered each of the coast-parallel terraces, called "coast-wise terraces," to be the result of alluviation followed by a period of extensive fluvial entrenchment. Fisk's model implied that coast-parallel terraces developed by the aggradation of a single depositional sequence in response to rising base level. Lowering of base level was presumed to cause renewed entrenchment, leaving the abandoned coastal plain as a coast-parallel terrace. Thus, the model of Fisk (1939,1944) claims that each coast-parallel terrace and the sedimentary deposits which fonn it are the product of a single depositional cycle and represent an individual sea level cycle (Winker 1979). Based upon Fisk's model many investigators have assumed as fact that each of the mapped coast-parallel terraces which have varied in number between 3 to 4 are fonned by a single depositional cycle. Furthennore, it has been presumed that each of these depositional cycles consists of a lithostratigraphic unit, i. e., fonnation, that should be recognizable on the basis of gross lithologic characteristics. As a result, numerous investigators, e.g., Bernard et al. (1962; 1970), Guevara-Sanchez (1974), Murray (1961), and Solis (1981) have attempted to subdivide the Pleistocene strata that underlie the coast-parallel terraces of Texas and adjacent Louisiana into fonnations that directly correlate with each of of these terraces. However, these attempts have failed to consistently define any such fonnations within the subsurface and demonstrate any correlation with the coast-parallel terraces (DuBar et al. 1991; Winker 1979). Also, the fonnations and members defined in some studies, e.g. Bernard et al. (1962; 1970) and Van Siclen (1985, 1991), fail to be true lithostratigraphic units (i. e. formations and members) as they are defined on the basis of either depositional cycles or bounding discontinuities rather than any distinctive and mappable differences in gross lithologic characteristics of these stratigraphic units.
3-15
In addition to problems with the subsurface stratigraphy, the coast-parallel terraces lack
the proposed one to one correspondence with glacial, eustatic sea level fluctuations envisioned by Fisk (1939, 1944) and Fisk and McFarlan (1955). A review of the Midwest glacial record by Richmond and Fullerton (1986) demonstrates that there has been at least 13 Pleistocene and 3 Pliocene glacial-interglacial cycles instead of the previously accepted 4 glacial cycles. Wornardt and Vail (1991) demonstrate that at least II high frequency, high amplitude sea level fluctuations have accompanied the Pleistocene glacial cycles of the last 1.8 million years. Finally, the coast-parallel terraces are not the simple depositional surfaces envisioned by Fisk (1939; 1944), Doering (1935), and Bernard et al. (1962, 1970), and others. Rather they are complex geomorphic surfaces, each of which has been formed by multiple periods of deltaic and alluvial deposition. Recent research within the Mississippi Alluvial Valley and the coastal plain of Louisiana demonstrates that the individual coast-parallel terraces are allostratigraphic complexes that consist of multiple, geomorphic surfaces. Each of these geomorphic surfaces forms the surface of unconformity-bounded sedimentary sequences, called "alloformations." Because a complex consists of mUltiple components of varying origin, simple casual relationships between the formation of a complex and eustatic sea level events are not valid (Autin et a11990, 1991). Van Siclen (1985,1991) assumes that the aggradation of the alluvial-deltaic plains that he has mapped occurs during rising sea level and subsequent high stands. Current models concerning the development of stratigraphic sequences strongly indicate that plain-wise alluvial aggradation actually occurs during the maximum high stand and between it and the inflection point offalling sea level, the "F inflection point" of Posamentier and Vail (1988: 131). During rising sea level, net alluvial aggradation is limited to the confines of a preexisting entrenched valley as is currently is occurring within the modem Texas Coastal Plain (Posamentier and Vail 1988:143-145).
Coast-Parallel Terraces Starting with Doering (1935), the coastal plain of Southeast Texas has been recognized as consisting of a series of geomorphic surfaces called "coast-parallel terraces." Typically, researchers have mapped these coast-parallel terraces on the basis of topographic expression, degree of seaward slope, degree of preservation of constructional topography, types of soil catenas, and chronologic sequence. Erosional escarpments and depositional on laps form the boundaries between coast-parallel terraces (Winker 1979:15-20). The original coast-parallel 3·16
terraces of Doering (1935) have been redefined and remapped many times as detailed by Winker (1979:15-20). Currently, DuBar et al. (1991:585-587) and Winker (1991) recognize three such terraces, the "Beaumont," "Lissie," and "pre-Lissie" coast-parallel terraces, along the Texas coast. Within Louisiana, their Beaumont Terrace includes the Prairie Terrace of Fisk (1939), Saucier and Snead (1989), and others.
Prairie Complex The lowest of these coast-parallel terraces forms the surface of a coast-parallel complex, designated as the "Prairie Complex" by Autin et al. (1991:556-558) and Saucier and Snead (1989). As previously explained, a coast-parallel complex is an informal allostratigraphic unit that consists of a set of related geomorphic surfaces which form a coast-parallel terrace and the depositional sequences associated with these surfaces. The lowest of the coast-parallel terraces, called the "Prairie Terrace" for this study, by definition forms the surface of the Prairie Complex. The Beaumont Formation (alloformation?) consists of the deltaic, fluvial, and coastal depositional sequences that form this coast -parallel terrace.
Prairie Terrace Within Louisiana and Texas, Autin et al. (1991:556-558) considers the lowest coastparallel terrace that forms the surface of the Prairie Complex to be properly called the "Prairie Terrace." However, with the exception of a small, inland portion of it within Louisiana, Winker (1979:28, 1991) maps this same coast-parallel terrace as the "Beaumont Terrace" within both Louisiana and Texas. These papers agree that the Prairie- Beaumont division is purely arbitrary and only one name should be used for the entire coastwise terrace within its extent across Mississippi, Louisiana, and Texas. However, they disagree on whether this prominent coast-parallel terrace should be called either the "Beaumont Terrace" or the "Prairie Terrace" in its entirety.
In this report, the designation of this coast-parallel terrace in its entirety as the "Prairie Terrace" by Saucier and Snead (1989) is used. Initially, Hayes and Kennedy (1903:27-29) originally defined the "Beaumont" on the basis of its lithology. Later, Deussen (1914, 1924) and Sellards et al. (1932) described the Beaumont as a lithostratigraphic unit. Doering (1935) later informally extended the designation "Beaumont" to the geomorphic surface presumed to be associated with the Beaumont Formation. In contrast, the Prairie Terrace was defined by Fisk (1939) as a geomorphic surface purely on the basis of surface morphology. Fisk (1944) 3-17
later extended the designation "Prairie" to the sediments underlying this terrace without formally describing, defining, and naming it as a valid stratigraphic unit. Because of the original usage of these names and the lack of a formal definition for both the Prairie Formation and Beaumont Terrace, "Prairie" is used to designate the coast-parallel terrace and "Beaumont" is used to designate the sediments that form it. Besides allowing for the consistent designation of this prominent coast-parallel terrace across the entire Gulf Coastal Plain, it also resolves problems with the usage of the term "Beaumont" for both a geomorphic surface and the sediments underlying it noted by Winker (1979:23-24). Finally, it allows for the consistent recognition of the Beaumont Formation which extends eastward into Louisiana as far as Jefferson Davis Parish. The Prairies Terrace is the outermost, lowest, and widest of the coast-parallel terraces which extends from the coastal plain of Mississippi, across Louisiana, Texas, and into Mexico. The continuity of the Prairie Terrace is only interrupted where the floodplains of the Mississippi, Brazos, Colorado, Trinity, Sabine and other rivers cut through it and the South Texas eolian sand sheet buries it. Well-preserved depositional topography that includes relict meander belts, relict delta lobes, and a strandplain and barrier island system characterize the Prairie Terrace. This coast-parallel terrace is crossed by numerous low-relief, silty to sandy meander belts, which often exhibit relict high-sinuosity channel patterns and spread in a radial pattern. The relict high-sinuosity channel systems are often associated with well-defined meander belt ridges (Figure 3-4). Rarely, these channel systems end in recognizable delta plains. However, the majority of this coast-parallel terrace consists of relict flood basins, backswamps, and interdistributary bays in which clayey soils, often vertisols, have developed. Numerous, enigmatic circular to elliptical hillocks commonly called "pimple mounds" cover the Prairie Terrace. A discontinuous series of coastal sand ridges interpreted to be either a relict strandplain or barrier island complex extends across the Prairie Terrace from Mexico along the Texas coast into Southwest Louisiana. These sand ridges have been designated by a variety of names such as "Ingleside Terrace," "Ingleside Barrier trend," Ingleside Strandplain", and "Ingleside Barrier." For this report, it will be simply designated in a nongeneric fashion as the Ingleside sand ridge (Aronow 1971, 1988, 1990; Barton 1930; Bernard et al. 1962, 1970; Dubar et al. 1991; Morton 1988; Price 1933, 1958; St. Clair et al. 1975; Van Siclen 1985, 1991; Winker 1979).
3·18
o
J )
..::-"o
.~::~It ~:~~.~?
.f'...,-- Ul~OOIlU
OIII1EU
-< ~'s",.c:,:"':l;~ , ...... TU'IU.(;oE. f;
~....,.....,...,
, >, .fro,., "!W"., ....,~....
:'-ont
CO~UCft
1"".1.1.....7'$
.
~<,:f:k/.::.: ~I.d~
OF
GULF ....:.
, LEGEND
MEXICO
'.-
IV = Oberlin alluvial plain V = Almeda alluvial plain VI = Eunice alluvial plain VIII = Modern alluvial and deltaic plians Figure 3-4.
Map of meander belt and distributary ridges within Galveston Bay region (modified from Van Siclen nd).
3-19
the western
The Prairie Terrace is a complex geomorphic surface composed of smaller geomorphic surfaces such as alluvial plains, delta plains, and the Ingleside sand ridges (see Figure 3·4). Within the Prairie Terrace, Van Siclen (1985, 1991) has mapped 3 alluvial-deltaic plains, which he calls "coastal terraces," within the Houston-Galveston, Texas region. He interprets these alluvial-deltaic plains to represent the surface expression of members within of the Beaumont Fonnation. However, because the Beaumont Formation consists of unifonn, randomly interstratified sands, silts, muds, and clays, it is highly unlikely that lithostratigraphic units such as members can be recognized within it. These plains, if correctly mapped, more likely are geomorphic surfaces that fonn the upper boundary of allostratigraphic subdivisions similar to those recognized within the Prairie Complex of Louisiana by Autin et al (1991:556558). However, a considerable amount of additional detailed subsurface research will be needed to confinn, define, name, and map the allostratigraphic subdivisions associated with these plains.
Beaumont Formation - Beaumont Allojormation At this time, confusion exists as to how the Beaumont Fonnation, should be defined. The confusion arises, because it was never properly defined with either a designated type section, an accepted thickness, or a specific basal lithologic identifier (Aronow 1988b:3). In addition, within Southeast Texas, the subsurface Pleistocene strata, except for a fairly laterally persistent, thick sand, consist of fairly unifonn, randomly interstratified sands, silts, muds, and clays. On the basis of gross lithology, only two major lithostratigraphic units have been consistently recognized in the subsurface Pliocene-Pleistocene sediments of the literature. The laterally persistent sand has been infonnally named as the "Alta Lorna sand" within Southeast Texas by hydrogeologic studies such as Rose (1943) and Kreitler et al. (1977). These stud.ies have consistently assigned the 150 to 500 m of interstratified sands, silts, muds, and clays that overlie the Alta Lorna sand to the Beaumont Formation (Winker 1979:22-23; DuBar et al. 1991 :585-586). Other researchers, e.g., Bernard et al. (1962; 1970), McFarlan and LeRoy (1988:424426), and Murray (1961:Figure 8.26), have defined the Beaumont Formation and equivalent Prairie Fonnation as the deposition cycle that forms the lowest of the three coast-parallel terraces. As defined by these studies, the Beaumont Formation consists of a sequence of sediments recognized not on the basis of its gross lithologic characteristics, but rather by its bounding discontinuities. As a result, their Beaumont and Prairie Formations fail to be lithostratigraphic units, e.g., a formation, but rather they are allostratigraphic units such as 3 ·20
alloformations or allogroups (North American Commission on Stratigraphic Nomenclature 1983:865-866)_ By such definitions, the Beaumont alloformation forms only the upper portion of the Pleistocene sediments that lie above the Alta Lorna sand. For example, Aronow (1991:9) presumes the Beaumont alloformation to consist of the uppermost fluvial depositional sequence and contemporaneous deltaic and coastal deposits that underlie each of the plains which form the Prairie Terrace. By this definition, the Beaumont alloformation consists only of the upper 15 to 20 m of the Pliocene-Pleistocene sediments that overlie the Alta Lorna sand instead of this entire 150 to 500 m thick sequence. If the Beaumont alloformation is presumed to consist of one to three stacked depositional fluvial sequences, then it would comprise the upper 30 to 60 m of the sediments overlying the Alta Lorna sand (Aronow 1988b:3). Within southern Harris County, the allostratigraphic unit which Bernard and LeBlanc (1965:174) and Bernard et al. (1962:Figure 14) call the Beaumont "Formation" is 40 m thick. Either the Beaumont alloformation or the uppermost part of the Beaumont Formation consists predominantly of clayey, fine-grained sediments. These clayey deposits contain thin, discontinuous beds of sands, silts, and clayey sands and silts and thick, often stacked channel sands. The clayey sediments deposited by the Brazos River often have reddish colors inherited from eroded Permian and Triassic "red beds" within Northeast Texas. Within the fluvial facies, slickenslides, pedogenic carbonate, and plinthite concretions associated with discontinuous, truncated paleosols are common characteristic of these sediments (Aronow 1988b, 1990a, 1990b, 1991a; Winker 1979). The determination of a firm estimate of the age of the Prairie Complex will be a unresolvable controversy until a consensus is developed concerning a specific definition of the Beaumont Formation. For example, as defined by Rose (1943) and Kreitler et al. (1977), the Beaumont Formation probably represents fluvial-deltaic and coastal deposits which have accumulated over the latter part of the Pliocene and all of the Pleistocene Epochs. As defined by Aronow (1991:9), the Beaumont alloformation consists of fluvial-deltaic and coastal deposits which have periodically accumulated during the Sangamonian, Early Wisconsinan, and Middle Wisconsinan Substage (Dubar et al. 1991; Thomas 1990; Winker 1979). Individual depositional sequences and alluvial plains likely accumulated in response to the maximum high stand of sea level during the Sangamonian and the lower than present high stands during the Wisconsinan much like depositional sequences within the Prairie Complex of Louisiana (Autin et al. 1991:558). 3-21
As demonstrated for the Prairie Terrace of Louisiana by Autin et al. (1991:556-558), the Prairie Terrace of the Texas Coastal Plain undoubtedly consists of geomorphic surfaces that formed during both the Wisconsinan and Sangamonian Stages with older geomorphic surfaces and deposits possibly being present. As a result, the question concerning the age of the Beaumont Formation (or alloformation) and associated Prairie Terrace should not be whether they formed during either the Sangamonian or part of the Wisconsinan Stages, but rather when during the Sangamonian and Wisconsinan Stages it formed. However, it has been definitely established that the Prairie Terrace and the sediments which form it were created prior to to the human occupation of the Texas Coastal Plain.
Possible Unnamed Stratigraphic Unit Ongoing work by Frederick (1991, personal communication 1991) infers that eolian sediments covers large portions of the Prairie Terrace of Southeast Texas and Southwest Louisiana. He proposes that this eolian blanket consists of silty sands within the middle coastal plain of Texas which decrease progressively eastward in grain size to almost 100 percent pure silts within southwestern Louisiana. Presumably these eolian deposits have been reworked by various processes to form the innumerable pimple mounds which cover the coastparallel terraces. Similarly, Aronow (1992:2) concludes that the sand and silty sand epipedons greater than 1 m thick and characteristic of grossarenic soils, e.g. the Kennedy and Boy soils series, are nonpedogenic, possibly eolian, in origin. Aronow (1992:2) claims that these soils are associated with subdued, stabilized dune-like topography. Frederick (1991) concludes that these sediments accumulated during Middle and Late Holocene times. Aronow (1992:2) proposes that the sediments formed during periods of aridity between either 6,500 to 4,500 or 1,000 to 800 radiocarbon years B.P. Additional research, including detailed sedimentological and pedological studies are needed to determine the validity, distribution, and age of this unnamed stratigraphic unit.
Surface Modification As the preceding discussion demonstrates, the landscape of the Prairie Terrace consists of relict landforms. With the cessation of coastal processes and the abandonment of individual alluvial and deltaic plains within it, the Prairie Terrace has started to evolve into an erosional coast-parallel terrace. Pedogenic processes, sheet flood erosion and deposition, eolian processes, the development of entrenched drainage systems, and lateral retreat of valley walls have substantially modified the surface of this terrace. The overall effects of these processes is 3-22
to obscure and, eventually, obliterate preexisting constructionallandfonns_ Some of the more controversial and archaeologically significance products of this modification are enigmatic landforms called "pimple mounds" and the previously discussed possible unnamed stratigraphic units (Aronow 1988a, 1990a:3l; Gustavson 1975; Fisheret al. 1972). Pimple Mounds Pimple mounds are innumerable, enigmatic circular to elliptical hillocks that are a common landfonn found not only on the surface of Pleistocene coast-parallel terraces such as the Prairie and Lissie Terraces, but also on the Holocene floodplains of Clear Creek, Buffalo Bayou, Greens Bayou, and other drainages within Harris County. These hillocks are approximately 15 to 60 m in diameter and rise as much as 1.2 m above the intennound terrace surface. Typically, the relief of a pimple mound results from the thickening of the A and E horizons of their sola (Aronow 1988:103, 1990:37-41). The pimple mounds which occur upon the floodplain of modem floodplains have been called by other names such as "floodplain mounds" and "sandy mounds." Because clear differences in size, shape, or structure between the mounds which occur upon floodplains and the coast-parallel terraces have yet to be documented, all of these sandy hillocks found within Southeast Texas are designated "pimple mounds" for this study. Discussions concerning the origin of pimple mounds has generated a immense and diverse literature as documented by Aronow (1990a:37-44) and Washburn (1988). The theories concerning the origin of pimple mounds include: 1. residual hillocks left after either wind or sheetflood erosion possibly with a core of tree-bonded surficial material; 2. accumulations of wind-transported sediment around around clumps of vegetation similar to coppice dunes; 3. eolian accumulations whose sites were started by, or topographically enhanced by, erosional processes; 4. fluvial bedfonns later modified by eolian erosion and deposition; 5. mounds fonned by the "fluffing up" of, or the decreasing the bulk densities of solum materials and centripedal transport of surface materials by burrowing animals; and 6. complex polygenetic landfonns that result from the middle to late Holocene modification of a preexisting eolian drape. At this time, neither a commonly accepted nor solidly documented explanation exists for the origin of pimple mounds (Aronow 1990a, Frederick 1991; Voellinger et al. 1987:91-93). At least two major hypotheses can be discounted for the origin of pimple mounds within Southeast Texas. Numerous paleoclimatic studies all agree that even during the most 3 -23
severe glacial period, e.g., the Wisconsinan glacial maximum between 22,500-14,000 radiocarbon years B.P., paleoclimates were insufficiently cold to form patterned ground (Bryant and Shafer 1977:7-13; Bryant and Holloway 1985). Also, Berg's (1990) proposal that seismic vibrations from earthquakes produced pimple mounds can be discounted, because of the lack of significant seismicity within Southeast Texas (Algermission 1969). More important, the interfering waveforms which produced the simulated "pimple mounds" are artifacts of Berg's (1990) modeling technique that any earthquake within Southeast Texas would fail to produce (Bridget Jensen, personal communication 1992).
Possible Unnamed Stratigraphic Unit The possible unnamed stratigraphic unit is an important modification of the surface of the Prairie Terrace. Because it is might be Holocene in age and, thus, may contain buried Paleo-Indian and Archaic archaeological deposits. In addition, these deposits likely resulted from climatic events significant enough to have impacted subsistence human patterns. Also, it is this unit that forms the bulk of pimple mounds. It is uncertain whether the origin of pimple mounds and this stratigraphic unit are concomitant as is the case of many constructional landforms. Therefore the existence and formation of this unit is a research question that should be addressed separately from questions concerning pimple mound formation (Frederick 1991, personal communication 1992; Voellinger et al. 1987:91-93).
Entrenched Drainage Systems Buffalo Bayou, Clear Creek, and Double Bayou are some of many entrenched drainage systems which have cut into the Prairie Terrace and drain into the Galveston Bay system. Van Siclen (nd, 1985, 1991) proposed that these drainage systems preferentially developed within the interchannel and interdeltaic areas between meander belt and deltaic ridges. Because of their length and height, these ridges act as drainage divides which force any movement of surface water to occur along the relict interchannel and interdeltaic lows lying between them. Presumably, as a result of head ward erosion, well defined stream channels were established. At this time, the history of none of these drainage systems is well known. Van Siclen (nd:Figure 1) has looked at the geology of Buffalo Bayou and illustrates a lOom-thick valley fill of an older course of Buffalo Bayou which he claims to be Early Wisconsinan in age. Forming the terrace and covering this valley fill is a thin red clay. However, this age is based purely upon hypothetical correlations with sea level cycles and lacks any paleontological and 3-24
radiometric control. Similar valley fills have been figured by Van Sic len (nd:Figures II and 12) from White Oak Bayou, a similar entrenched drainage system. The age of these deposits also are unknown, although Late Holocene fluvial deposits are associated with the historic meanders of White Oak Bayou. Because of the intense urban and industrial development which has extensively modified the modern landscape, terraces associated with the older alluvial deposits along Buffalo Bayou could be neither identified nor mapped from modern topographic maps, aerial photography or soil surveys. Possibly, the examination of historic 1foot contour maps might prove more useful in mapping possible Holocene fluvial terraces that might have been associated with Buffalo Bayou. Effects 011 Huma1l Adaptati01l
The upland surfaces surrounding the Galveston Bay system have been exposed sine the first appearance of human populations 11,000 years ago or so. As such, these landforms have always been available for human use and occupation. However, not all areas on these land surfaces have necessarily been desirable to humans. Prior to about 5,000 years ago, during periods of lower sea levels, the Gulf of Mexico and the coastal zone was located much farther to the south. Consequently, none of the varied and rich biological resources associated with coastal biomes would have been available to the regions inhabitants. Presumably, then populations in the area of present-day Galveston Bay were probably much lower than they became after the bay began to be transformed into a coastal estuary. There seems to be no doubt, however, that early populations were living in the region, particularly as evidenced by the early archaeological remains from the McFaddin Beach site, just to the east of Galveston Bay, and possibly by the evidence recovered from the Texas City Dike site. Populations would have been attracted to the resources of the larger river valleys, such as the Trinity and San Jacinto, as well as to some of the smaller streams, if, in fact these were extant. As a result, populations were probably concentrated along desirable landforms within and immediately adjacent to major and minor streams. This pattern of settlement is apparent in the known archaeological record of the region. During the very earliest period of human occupation in the region, the relict, late Pleistocene Trinity and Brazos rivers meander belt systems extant in the area may, also, have been attractive to human populations. These landforms were slightly elevated, and some may have retained wet, or marshy landscapes for a great many years after their abandonment by the Trinity and the Brazos. If so, they would have presented isolated locales of increased 3-25
biological diversity, attractive to a range of wildlife, which in turn would have been attractive to human populations. While this scenario seems reasonable, little attention has been directed toward the archaeology of these features. Most archaeological site found upon the relict alluvial and deltaic plains of the Prairie Terrace should consist of surficial archaeological deposits. However, within Holocene, and terminal Pleistocene (?), alluvial deposits associated with entrenched drainages on these upland surfaces, stratified archaeological deposits can occur. Possible Holocene eolian deposits might have obscured older archaeological deposits. At this time, the origin of pimple mounds and the degree to which they contain stratified archaeological deposits is an unanswered question.
Effects on the Archaeological Record Because the upland landforms all pre· date human occupation, the processes associated with their formation, have had no impact on the archaeological record. However, subsequent alterations to these landforms have occurred, which has impacted the archaeological record. As noted above, with the cessation of coastal processes and the abandonment of individual alluvial and deltaic plains the upland surfaces started to evolve into and erosional coast-parallel terrace. Pedogenic processes, sheet flood erosion and deposition, eolian processes, the development of entrenched drainage systems, and lateral retreat of valley walls began to substantially modify the surface of this terrace. The overall effects of these processes has been to obscure and, eventually, obliterate preexisting constructional landforms.
Erosion of Valley Walls As documented by Paine and Morton (1986), the low to high, steep clay bluffs which form a substantial portion of the shorelines of Galveston, San Jacinto, and Trinity bays have retreated and are currently retreating as a result of shoreline erosion at geoarchaeologically significant rates. For example, their research demonstrates that, except for the Trinity River Delta, that shoreline of Trinity Bay retreated at a rate of 0.8 to 0.9 m per year between 18501852 and 1930. For the same period of time, they found that the shoreline of Galveston and San Jacinto bays retreated at an average rate of 0.7 m per year. Within San Jacinto Bay, the rate of shoreline retreat was as much as 4 m per year between 1850-1852 and 1930, largely as a product of subsidence starting in the 1900s as a result of groundwater pumping and, later, oil production at the Goose Creek Oil Field. At Red Bluff and within the Texas City area, the
3·26
shoreline has retreated at rates as high as 1.4 to 1.5 m per year for the period from 1850·1852 to 1930 (Paine and Morton 1986). As proposed by Aten (1983: 156·157), the enlargement of the Galveston Bay system by the lateral retreat of the valley wall by shoreline erosion would have destroyed much of the archaeological record of the bay margin of this area. The relatively small annual rates of historic shoreline erosion determined by Paine and Morton (1986) would over archaeological periods of time preferentially destroy the older bay margin archaeological deposits. Only when the bay in filled by fluvial·deltaic sediments and the shoreline moved away from the valley wall will it become stabilized and archaeological deposits forming on or adjacent to the bluff edges be preserved (Aten 1983:156-157). Work by Paine (1987a:437-440; 1990:396-398) provides data concerning the timing of valley wall retreat. His geomorphological studies conclude that significant erosional retreat of the Pleistocene bluffs forming the valley walls within the Peggy Lake survey area occurred for a period of time starting around 4,000 radiocarbon years B.P.
As indicated by the
accumulation of colluvial and associated archaeological deposits, these valley walls abruptly ceased to retreat after a brief period of time and have remained stable since then (Paine 1987a:437-440; 1990:396-398). Timing of bluff erosion within the Peggy Lake area supports Aten's (1983:156·157) hypothesis. About 4,000 radiocarbon years B.P., substantial retreat of the valley walls of the San Jacinto River within the Peggy Lake area would have commenced as a result of erosion along the shoreline newly formed by the maximum flooding of the San Jacinto River Valley by eustatic sea level rise. Afterwards, the filling of the newly-created, narrow arm of San Jacinto Bay with fluvial-deltaic sediments would have permanently stopped the retreat of the valley wall by forcing the shoreline responsible for it to move down valley and away from this segment of the valley wall. The age and distribution of archaeological deposits found by Gladus and Howard (1990) within the Peggy Lake survey area are completely consistent with Aten's (1983) hypothesis and known sea level fluctuations within the Galveston Bay complex.
Stratigraphy alld Chrollology of Fluvial Sequellces Just as the development of coast-parallel terraces have been linked to glacio-eustatic base-level cycles, studies of the rivers which drain the Texas Coastal Plain have used the 3·27
concepts and methods of Fisk (1944) to interpret the chronology and genesis of their fluvial terraces. For example, work by Bernard (1950) concerning the Sabine River of Louisiana and Texas and Stickland (1961) concerning the Brazos River of Central Texas have attempted to link the development of fluvial terraces to a fourfold cyclic glacial-interglacial sequence and the
related changes in glacio-eustatic base-level. Although acknowledging that more than 4 glacioeustatic base-level cycles have occurred, other studies, e.g., LeBlanc (1991) and Thomas (1990), still presume that each fluvial terrace and the fluvial complex or alloformation of which the terrace forms the surface represents a fluvial response to an individual glacio-eustatic baselevel cycle. However, in addition to base levels associated with eustatic processes, the geomorphology, stratigraphy, and sedimentary processes are greatly influenced by source-area lithologies, inherent geomorphic controls on sediment delivery and channel form, and baselevel adjustments associated with tectonic and climatic processes. Changes in fluvial discharge and sediment load related to climatic changes alone can produce regionally persistent unconformities within alluvial sequences of coastal plain sequences. Climatic change can produce such regionally persistent unconformities within alluvial sequences over short periods of time, much shorter than those responsible for major glacial-interglacial. As a result, individual fluvial terraces or alloformations and the fluvial terraces that form their surfaces cannot be reliably correlated with sea level events per se (Autin 1992, Blum 1991). The common presumption by many studies concerning the chronology and genesis of fluvial terraces within the Texas Gulf Coast that different fluvial complexes and their fluvial terraces each formed during a separate glacio-eustatic sea level cycle must be proven rather then assumed, because sea level is not the only cause of fluvial terrace development as implied by Fisk (1944).
Trinity and San Jacinto River Valleys The entrenched valleys of the Trinity and San Jacinto Rivers consist of a set of fluvial terraces and an active floodplain. These fluvial terraces represent the surface of fluvial complexes and alloformations, as previously defined, that form the terraces and modem floodplains within both alluvial valleys. The terminology, especially the term "Deweyville," used to designate the fluvial terraces which lie stratigraphically between the Prairie Complex and the deposits which form the modem floodplain varies greatly from study to study (Figure 3-5). Because of this variability, it is impossible to known exactly how a particular study
3-28
Aten (1966b)
modern floodplain
modern alluvialdeltaic plain CD
::> (l
~
o.~
CD
Ill'"
»
0
co
::;:
N
-!
§: CD
- '"
'"'"
roJ
0;-
"co
il "CD
CD
::;:
(lCD
::!
o~
-'"
-!
il CD "()
(l
0.
CD
0
Qd? Terrace B
g
0
::;:
"Tl
CD
0
co
-
'< -<<0 =.: =r
3 el. 0-
co "'- < -co ""'"'.
::>
C
-
CD
Prairie Terrace Beaumont Coastwise Terrace --------
Figure 3-5.
Beaumont Terrace (Formation) Qb
il "-!CD "-
0
co
1::;: Stage 5 Fluvial Terrace "Deweyville Terraces"
::i" ~.
o~·
_:>
° --..:
3",
"'-
e.< °
CD
-!
!!1.
~
Q~§ 3 '< :> ",OCD e. c: (j;
CD
CD
-!
-!
;j
'<
<
I
0
co ::;: co
CD
-'" roJ
~
:E ~
:::
"'" CD
;j
0(D--!
°:>
co
0
~o o:E - CD 3'<
°CD
ffil
c::
'"
III '<
'"c '" 5' <0 0
0
c
"0
g.~
:>
CD -!
CD
il
"
CD
c:!
-!
!!1
-!
iil
CD
"'
CD
B
~ j
Liberty Allogroup
(l
CD
"Tl
0
CD
~
0
CD
::r
CD
-!
-!
il
il "-!CD
CD
"-! CD
"-
CD
"-
-------
::>
."ai' ti):::;:
~3
:>
::!
CD
~
::>
3 el. 0-
a. a. "0 CD
::!
CD
'<
0
-
00
-!
CD
::;: cO-
'"3
!!1.-!
."
oS;
:> -
~
CD 0
CD
III
modern a1luvialdeltaic plain
CD
::;:,
I
allostratigraphy
-------
"co
co c
(1991 )
"'CD"
'"
"
This Report
Aronow
:::
-; co
CD
;;;l
Trinity River Recent Delta and Trinity River floodplain floodplain
~
~-
-!
c:!
LeBlanc (1991a, 1991b
CD
co ::;: co
il
CD co
Terrace A
;j
~ CD
co=l-
- -- - --
~CD
Qd
Thomas (1990)
0
CD
-!
co
()
::>
CD
0
CD-; x co
CD
CD
'<
"
modern a1luvialdeltaic plain
3= "0 CD
0
CJl
Aten (1983)
alluvium Qal
-!
",CJl
'" '"
Aronow et ai_ (1968, 1983)
Gagliano and Thorn (1967)
Beaumont Terrace
Beaumont alluvialdeltaic plain
Beaumont Terrace
;oJ."
@ ~. (")
~.
CD CD
"'
Beaumont
Formation? (Alloformation?
i.-
Correlatiou of geomorphic surfaces mapped within the Trinity River Valley and adjacent coastal plain.
defines tenns such as the "Deweyville Terrace," "Deweyville Fonnation," or even the "modern alluvial plain" unless it includes maps of or elevations for their fluvial terraces. In order to resolve this confusion, these fluvial terraces were remapped for the present study. This remapping used soil surveys for Chambers (Crout 1976), Liberty, Polk (McEwen et aL 1988), and San Jacinto counties (McEwen et aL 1988) and U. S. Geological Survey 7.5 minute maps, many of which postdate 1983. The remapping considered the surface elevation of terraces, associated soils, preservation of surface morphology, and size of preserved fluvial landfonns. In addition, cross·sections were constructed using the UTM grid system, crosssections perpendicular to and along the length of the Trinity River Valley. This reanalysis demonstrated that Aten (1966, 1983:104·116) produced a very perceptively detailed and accurate mapping of the fluvial terraces within the Trinity River Valley (Figure 3·6). First, he clearly recognizes the presences of two groups of temporally related terraces, which LeBlanc (1991b), Thomas (1990), and other studies fail to differentiate. Second, Aten (1966, 1983) clearly appreciates the inapplicability and inappropriateness of the tenn "Deweyville" for designating all of the fluvial terraces, particularly the upper group of fluvial terraces, within the Trinity River Valley. Finally, unlike Aronow (1968, 1982), LeBlanc (1991b), and Thomas (1990), he clearly recognizes the Terrace TI as a distinct fluvial terrace, although partially buried or buried within the lower valley, that is distinct in morphology, age, and origin from the modern floodplain of the Trinity River. On the basis of surface elevation of terraces, associated soils, preservation of surface morphology, and size of preserved fluviallandfonns, Aten (1966,1983:104-116) recognizes two groups of fluvial terraces which are each composed of two terraces (see Figure 3·5, Figure 3·6). According to Aten (1983: 110·111), the upper group of fluvial terraces consists of the Terraces T4 and T3. He presumed both terraces to be strath terraces and noted that they are highly dissected, exhibit normal· size ridge and swale topography, and have been deeply weathered. Aten (1983: 112·114) distinguished a lower group of fluvial terraces, the Terraces T2 and Tl, which are designated as the "Deweyville Terraces." Both terraces exhibit extremely well preserved constructional surfaces characterized by meander radii, ridge and swale topography, channel segments, and meander bights much larger then those exhibited by the nearby Holocene alluvial plain (Aten 1983).
3 ·30
3
Modern alluvialdeltaic plain I To Deweyville Terraces
.....
a
Terrace
ITIIII1
Terrace T2
T,
Beaumont Formation '
Sirath(?) Terrace
Stroth(?) Terrace T4 Deltaic plain phase (uplands)
~ 0,
Figure 3·6.
...
~
..
....
•
• I
Geomorphic surfaces within and adjacent to the Trinity River Valley (modified from Aten 1983:Figure 8.1).
3·31
Allostratigraphy Aten (1983:110) interprets his Terraces T4 and T3, to be strath terraces cut into the Beaumont Formation_ However, seismic sections and foundation borings from Galveston bay clearly indicate that both terraces are associated with fluvial sands which are much as 15 m thick, e_g_ Smyth (1991 :Figures 4-19 and 4-21) and lack gravel. These deposits are related neither to the Beaumont Formation nor the allostratigraphic units to which the Terraces T2 and TI belong_ Therefore, the unnamed alloformations, with Terraces T4 and T3 as their surfaces, lying stratigraphically between the Prairie Complex and the alloformation associated with Terrace T2 are assigned to new and separate allogroup, informally named the "Liberty allogroup" for purposes of this report (see Figures 3-5, 3-6, Figure 3-7). The type location for this allogroup consists of the Terraces T4 and T3 east and northeast of Liberty, Texas.
,
Because these fragmentary terraces are highly dissected and partially buried in places by younger overbank deposits the available stratigraphic and sedimentologic data are insufficient to define, correlate, and map the individual allostratigraphic units within the Liberty allogroup. Because such work is beyond the scope of this study and because of their presumed age, the individual fluvial complexes of the Liberty allogroup are left undifferentiated and unnamed. Although, poorly preserved, the terrace surfaces of the Liberty allogroup apparently exhibit both Trinity-like and Deweyville-like fluvial landforms. Along the Sabine River, Bernard (1950:59-61, 131-134) clearly defines the "Deweyville beds" as the deposits and the single fluvial terrace on which Deweyville, Texas lies. As defined by Bernard (1950:59-61,131-134), the term "Deweyville" designates a single terrace surface and associated beds that can be considered to constitute a single fluvial complex or alloformation with a specific stratigraphic position. By its original definition and the rules of stratigraphic nomenclature, the term "Deweyville" is restrict to the designation of one of the allostratigraphic units associated with the terraces mapped by Aten (1983). As previously explained, alloformations and fluvial complexes, by definition, are correlated on the basis of bounding unconformities rather than the morphology of the fluvial terraces which form their surfaces. Therefore, the term "Deweyville" is restricted to the fluvial complex, designated as the "Deweyville alloformation" in this report, associated with the Terrace T2 of Aten which is considered to be correlative with the Deweyville beds and terrace of Bernard (1950:59-61, 131) within the Sabine River Valley. The fluvial complex of which the Terrace TI of Aten (1983:111) forms its surface, is designated as the Tanner Bayou alloformation from Tanner Bayou where it crosses a fluvial terrace fragment just north of Caspers Ridge within Liberty County, Texas (see Figures 3-5 and 3-6, Figure 3-7). 3-32
50
Prairie Terrace
40
..
'g ~
20
0
10
c
~
~
'" UJ
0 -10
S""';nney Ma"h
Terrace 2
Terrace 4
30
40
Terrace 4 Beaumont aUofonnation Uberty aJlogroup
30
0
?
?
aJlogroup
. ...
i'
!::.. c 10 .g
20
?
Deweyville aJloformation
? Liberty
50
>
ill
-10
-20
-20
25000
20000
10000
15000
5000
Distance (feet) 50
Prairie Terrace
Prairie Terrace
M033 B lit,
40
'".
'"'"
..
o
30
~
20
0
10
c
1 UJ
Te_
Beaumont aJloformation (eba)
t
Big Cane
Modem floodplain over Terrace 1
Modem floodplain over Terrace 1
Creek
tt
Terrace 3
Uberty
allogroup
Tanner Bayou aJloformation
-10
Tanner Bayou alloformation
' ..., ......
..
30 ~
Trinity River
0
50 40 2O~
c
1ii
0
'"
-10
-20
0
10
>
UJ
-20
55000
50000
45000
Distance (feet)
40000
35000
30000
LEGEND
m Marsh or swamp deposits B
.......... ,Inferred geomorphis surface or erosional unconformity
Water
Ota = Trnity River alloformation
Figure 3-7. Cross-section of the Trinity River Valley at Moss Bluff, Texas, showing geomorphic surfaces and allostratigraphic units (modified from Aten 1966).
The presence of oversize constructional fluvial landforms can only be a part of the definition of "Deweyville," because of the rules of stratigraphic nomenclature and, as first postulated by Dr. Armstrong Price in Bernard (1950:58), the conditions which would have produced such landforms very likely occurred during multiple times during the Pleistocene Epoch. For example, ongoing soil-geomorphic research by Dr. Whitney Autin (personal communication 1991) along the Pearl River indicates that the "Upper Deweyville" terrace of Gagliano (1979:2-16) is early Late Wisconsinan in age and the "Lower Deweyville" terrace of Gagliano (1979:2-17) is late Late Wisconsinan to terminal Wisconsinan in age. Both of his "Deweyville" terraces clearly represent separate periods of terrace development that are younger than Deweyville-like fluvial landforms exhibited by the adjacent Prairie Terraces. Similar examples of multiple fluvial terraces with oversize constructional fluvial landforms of distinctly differing ages can be seen along the Nueces River of Texas and Pascagoula River within George County, Mississippi (Bernard 1950:58; Gagliano 1979: 2-13; Williams 1966). The youngest alloformation consists of the modern floodplain of the Trinity River and the alluvium that forms it. For this report, it is informally designated as the Trinity River alloformation, because a time-neutral term, other then "Holocene, " is needed to designate any allostratigraphic unit. The geomorphic development of the alluvial plain that forms the surface of the Trinity River alloformation within the Lower Trinity River Valley has been fully documented by Aten (1983:125-128). The Deweyville, Tanner Bayou, and Trinity River alloformation constitutes an allogroup informally designated as the Fields Bayou allogroup. This allogroup is named for Fields Bayou, a bayou which crosses the surfaces of all three alloformations where they juxtaposed against each other and the Liberty allogroup in northernmost Liberty County. The fluvial deposits of the Fields Bayou allogroup lie between the modern terrace and floodplains and the erosional unconformity at the base of the Deweyville alloformation. This unconformity represents, in part, the maximum, Late Wisconsinan low stand of sea level which occurred during part of Oxygen Isotope Stage 2.
Beneath Galveston Bay, the upper bounding
discontinuity is a bayline flooding surface which separates it from the overlying Galveston Bay alloformation. As previously explained, two very different model have been used to explain the formation of the fluvial complexes, alloformations, and their fluvial terraces which lie stratigraphically between the Prairie Complex and the Holocene deposits of the Trinity River alloformation. First, LeBlanc (1991 a, 1991 b) and Thomas (1990) have both used a strict 3 ·34
Fiskian interpretation, in which fluvial terrace fonnation is solely the result of eustatic sea level changes, to interpret the development of the fluvial complexes within the Trinity River Valley. It is significant that both studies consider the fluvial terrace of the Tanner Bayou allofonnation as a part of the modern floodplain in their studies. Finally, recent work by Autin (1992) and Blum (1991) propose that allofonnations, fluvial complexes, and their fluvial terraces can be formed by changes in fluvial regime related to variations in discharge, sediment load, and different types of base level changes, including eustatic sea level changes. According to the models of LeBlanc (1991 b) and Thomas (1991), the fonnation of the Tanner Bayou and Deweyville allofonnations would predate the last sea level lowstand during the Late Wisconsinan Substage. The Tanner Bayou allofonnation would be at least as old as high sea level stand II, approximately 32,000 to 35,000 radiometric years B.P., and the Deweyville alloformation would be at least as old as high sea level stand III, approximately 40,000 to 50,000 radiometric years B.P. (Figure 3-8). Thomas (1990:88-89), collectively designated as the "Stage 5c Fluvial Terraces," correlates the Deweyville alloformation and Liberty allogroup to an offshore fluvial allofonnations, his 5c fluvial terrace, and concludes that both units are approximately 110,000 years old. Using similar reasoning, Thomas and Anderson (1989:567) proposed that these terraces are at least 50,000 years old. As observed by Aten (1983: 111) for the Tanner Bayou and Deweyville alloformations, either a Late Sangamonian or Middle Wisconsinan age for these allofonnations is contradicted by radiocarbon dates, the excellent preservation of fluviallandfonns, relatively open oxbow lakes, and degree of soil development possessed by Terraces T2 and Tl. However, Rufus J. LeBlanc (personal communications 1991) claims that the radiocarbon dates indicating a Late Wisconsinan age for both allofonnations are all bad dates, because they contradict the age of these fluvial deposits as predicted by his interpretation of the Fiskian model and correlations made by Thomas (1990). On the other hand, the well preserved nature of both Terraces T2 and Tl is completely consistent with these same radiocarbon dates. In contrast, the highly weathered and dissected nature of the Terraces T4 and T3 of the Liberty allogroup, particularly the constructional topography that has been either largely obscured or totally obliterated, is consistent with the Late Sangamonian age for this allogroup (Aten 1983:110) as advocated by Thomas (1990: 88-89). Thus, within the Trinity River Valley, this study hypothesizes that both models are applicable to the fonnation of allostratigraphic units within the Trinity River Valley but at different scales. It is hypothesized that a dramatic drop in sea level, e.g., the Fiskian models of 3-35
THOUSANDS OF YEARS BEFORE PRESENT 160
140
120
100
80
40
60
20
0
VII a b
~
(j) ....
OJ ...... OJ
0 E -20 -' -40 W GJ -60 -' -80
VI
'\
V
" \,, \
«
,, ,, ,, ,
(J)
'.
\
\
,
W-100
1--120
IV
I
I
'
I
, ,, I ( I
J'l
lj,
I
I
I
J I I \
OXYGEN ISOTOPE STAGES 6 I
5e 5d 5c 5b 5a 4
S
Ew
We
\r
3.2 3.4
W
LL
oZ
~ ~
«
3.6 wa: 3.8 ~ f:2 ::J
o 0z 4.2 >0
(
4.0
I
I
,I I I I I I I
WI 4.4 0 l-w 4.6 ro LL 4.8 0
,, , , \
«
,
\
'
,
I ,,
II
\
Z
w-140 a: « -160 0... 0...-180
III
0 f\r~ l
,
~
I
« a:
3.0
\ I I
3
2
1
Wm
WI
H
CHRONOSTRATIGRAPHY H = Holocene Epoch* WI = Late Wisconsinan Substage* Wm = Middle Wisconsinan Substage* We = Early Wisconsinan Substage* Ew = Eowisconsinan Substage* S = Sanganmonian Stage* I = Illinoian Stage* V II - I
= high sea level stands of Moore (1982)
*as defined by Sibrava et al. (1986). Figure 3-8.
Glacio-eustatic sea level record (solid line) and composite oxygen isotope record of deep sea benthonic foraminifera (dashed line) for the past 130,000 years. The latter is an indicator of the ice volume of continental glaciers (adapted from Williams et al. 1981: 161).
3·36
LeBlanc (1991a) and Thomas (1990), caused the formation of unconformity which separates the Liberty and Fields Bayou allogroups. However, within the Trinity River allogroup, periodic instability of the meander belts resulting from factors other then or in addition to changes in rates of rising sea level probably formed individual alloformations and fluvial complexes as demonstrated by Autin (1989, 1992) for the Amite River system and Blum (1990) for the Colorado River system. This model explains the observations of Aten (1983), Thomas (1990), and Thomas and Anderson (1988, 1989). The relative differences in the preservation of surface morphology and degree of weathering noted by Aten (1983: 108-112) between Terraces T3 and T4 of the Liberty allogroup and Terraces T2 and Tl of the Deweyville and Tanner Bayou alloformations is clear evidence of two widely separated periods of fluvial deposition. In addition, this model produces the stratigraphic relationships illustrated by Thomas (l990:Figure 1-ISb) and Thomas and Anderson (1988) within the incised valley system of the Sabine and Trinity Rivers, because the Liberty allogroup correlates with the fluvial complexes which underlie their Sc and, possibly, Sa fluvial terraces and the Fields Bayou allogroup correlates with their "1 Fluvial" deposits. This model is also consistent with published radiometric dates from the Trinity, Neches, and Sabine river valleys. According to this model, the age of the Deweyville and Tanner Bayou alloformations include the 17,000 to 10,000 radiocarbon years B.P. "Deweyville Interval" of Gagliano (1991), a 21,000 radiocarbon years B.P. date from the Deweyville alloformation (Slaughter 1965:9), and a date of "about 13,000" radiocarbon years B.P. from a sand pit in the Deweyville alloformation near Deweyville, Louisiana (Aronow 1986:12). In addition, two dates of Bernard and LeBlanc (196S:149), averaging 17,000 radiocarbon years B.P., were obtained from wood recovered from 40 ft below the terrace of the Deweyville alloformation about three miles west of Deweyville, Louisiana (Gagliano and Thorn 1967:33). Because previous definitions of the term "Deweyville" includes both the Liberty allogroup and the Deweyville alloformation and specific data concerning location ai, material dated, and stratigraphic provenience remain unpublished, it is undetermined whether dates ranging from about 30,000 to 12,000 radiocarbon years B.P. (Aronow 1976:10), from 2S,700 to 13, 2S0 radiocarbon years B.P. (Aronow 1967, 1971), and from greater than 30,000 to 17,000 radiocarbon years B.P. (Bernard and LeBlanc 1965:149) from the "Deweyville" of the Sabine, Neches, and Trinity River Valleys either agree with or contradict this model. Finally, four dates ranging from 13,378±102 (SMU-230S) to 14,208±218 (SMU2236) radiocarbon years B.P. have been obtained from the Audrey alluvium of Ferring 3-37
(1990a:46-47) along Elm Fork of the Trinity River within Denton County, Texas. The Audrey alluvium has Deweyville-like meanders 4 to 5 times as large as those of the modern Elm Creek and consists of sands almost devoid of clay and silt and gravelly basal sands (Ferring 1990a:30).
Origin of Surface Morphology The origin of the oversize fluvial landforms exhibited by the Deweyville and Tanner Bayou alloformations is still controversial. The oversize fluvial landforms and coarser sediments that form the Deweyville and Tanner Bayou alloformations have been interpreted to represent either significantly greater average annual or flood discharge than at present. Estimates from the geometry of meander loops and river courses indicate discharges ranging form 4 to 10 times normal. Various authors have interpreted the origin of such high discharges to have been the result of increased runoff resulting from sources such as increased precipitation, decreased evaporation, and increased frequency of tropical storms. In addition, rates of eustatic sea level rise and amount and type of sediment load may also have significantly influenced the morphology of these landforms (Alford and Holmes 1985; Whitney Autin, personal communication 1992; Baker 1983; Gagliano 1991; Gagliano and Thorn 1967). This origin of the surface morphology remains controversial for three main reasons. First, as previously discussed, the lack of published locational and stratigraphic data for the few radiocarbon dates that have been obtained from these deposits makes the definite determination of the age of these deposits equivocal. As a result, it is difficult to relate the changes in fluvial regime to specific changes in paleoclimate, sea level, and other possible factors. Second, the lack of detailed paleoclimatic information independent of that inferred from changes in channel morphology prohibits the testing of hypotheses relating climatic factors to the morphology of these channels.
Finally, within published reports, the
paleohydrology of the oversize fluvial landforms has been reconstructed solely from the surface morphology of the Deweyville and Tanner Bayou alloformations which might give only the gross, possibly misleading, idea of the hydrology of these fluvial systems. To precisely determine the dynamics of the fluvial systems which created both alloformations, the hydrology of each fluvial system need to be reconstructed from a detailed sedimentological studies of the sediments which form each alloformation. The origin of these oversize fluvial landforms is important because they might record environmental changes significant to human subsistence to which the pollen record is either insensitive or lacking.
3 -38
Galveston Bay Approximately 1 to 19 m of estuarine deposits fill the submerged valleys of the Trinity and San Jacinto River that form the Galveston Bay system (Figure 3-9). These sediments are thickest within the thalweg of the submerged fluvial valleys and pinch out against the valley walls. Except at its updip edge, these estuarine deposits are separated from the underlying fluvial and Pleistocene deposits by a time-transgressive disconformity called a "bayline flooding surfaces (Figure 3-10). A bayline flooding surface is a disconformity which separates either fluvial or older upland deposits from overlying estuarine deposits across which there is evidence for a permanent rise in sea level. On the upper surface of the Galveston Bay alloformation forms the bottom of the Galveston Bay system. However, seaward of Galveston Island, the upper few meters of the estuarine fill is truncate by an regional erosional disconformity called a ravinement surface (Rehkemper 1969a, 1969b; Smyth 1991; Thomas 1990). The estuarine deposits lying between the bayline flooding surface and either bottom of the Galveston Bay system or the ravinement surface form an informal allostratigraphic unit, called the Galveston Bay alloformation. Within the Galveston Bay alloformation, a series of disconformities, called "intermediate flooding surfaces" define an imbricated set of allomembers (Figure 3-10). An intermediate flooding surface is a disconformity which separates younger from older estuarine deposits across which there is evidence of an abrupt increase in water depth without any significant accompanying erosion. The typical allomember consists of four basic sedimentary deposits that consist of a variety of sedimentary facies (Table 1). First, the upstream end of each allomember consists of a wedge of delta sediments which constitute a single buried bayhead delta (McEwen 1963, 1969). Second, the bulk of the valley fill consists of a variety of estuarine silty clays, sandy clays, and clays and shell reefs that have described in detail by by Rehkemper (l969a, 1969b). Third, the middle of each allomember consists of an sequence of tidal-delta and tidal-inlet deposits. Prior to being destroyed by a rapid rise in sea level, a barrier island and spit system was associated with each sequence of tidal deposits. Finally, seaward of the tidal deposits, the allomember consists of shoreface and offshore sediments which grade laterally into a offshore shelf sand body at its toe (Table 1) (Figure 3-10). In case of the youngest allomember, a complete sequence of barrier island deposits, of which Galveston Island and Bolivar Peninsula is part, is present (Rehkemper 1969b; Smyth 1991; Thomas 1990; Thomas and Anderson 1988).
3-39
: ..-TRINITY _RIVER
. :i -, -. ,~
•I
.
I•
SCoUE IN 1llOMfHU
~
.....
~
...
• •o
INDICAtES
con OA'A uno
fOl """ING THtS INTUVAl
+ - ..
I
GALVtSTOill BAY ARt"
f
Figure 3·9.
Isopach map showing thickness of the Galveston Bay alloformation within the Galveston Bay system (source: Smyth 1991:Figure 4.29).
3·40
-
Trinity River Delta
Galveston Island and Bolivar Roads ~.. sl---------I /' .,,;:;:/ ~ ,-,-"",-,_,_,_,_,_ ~"':-:-:-'" .Sabine Bank
sl
.......0OO:.O.O:.:.~~.:\
... :: . .::::::.-;::::::::.~::.~~\ -....::..::....::...: \
,...,tb· 'l5
.......: -- ---1'-
'-':t~~-' .-...-.-...
_"..<"'"
Qgba
_ -J,_ •
-
- 0 _ • • • •: . ' : : : :• •: , . : : :
~-------------------------sl--------------------------------~
~~ineBank ~ -'-'-'-.-.-._._ ...... :: ........... :~. :-::::::::.':::.".':.-::. \
~m.mm ......:::::::::..:::. '........
\
Heald Bank
'-'-'j-'-'-'-'-'-'-'-'-'-.-.~. ~ ._.-'.' . . . .
Unnamed Bank .... _~~._ . ......... r ~ '>.~ ..!", ----...... - ........ ...:.:...:.:-=-.:.. :.: .....
Qgba --
l
~
--
..... "':::.:.:-:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.&'
. _....-:....:.:.:-:.:.:::::::::::::::::::::::~.~ \ - - -- - - - -- ,--
o
••••
- ~-::-'-~~.-.
--.....
........ ~"
~.,
....
"' ... -
::::;,:."
-
4:::;::~ Barrier Island
LEGEND o
--- ............ ....:....:. ....... ~.
~ 8ayhead delta deposits Intermediate flooding surface
~
~
Marine sand shoal
__ ~ Flood tidal delta and ~ channel deposits
~ ~
Ravinement surface
-
Qbg = Galveston Bay alloformation
Bayline flooding surface
Qgba = unnamed allomember
- - s l - - Modem sea level
Figure 3-10. Schematic cross-section of the the Gatveston Alloformation within the Trinity River Valley of Galveston Bay and the adjacent continentat shelf (modified from Thomas 1991:Figure 2-34).
3-41
Table I
3-42
Because a brief period of rapid sea level rise created an individual intermediate flooding surface, each intermediate flooding surface is a time-line that separates a younger from an older sequence of sedimentary deposits (Thomas 1990: 153) and Thomas and Anderson (1988)_ Therefore, each allomember consists of sediments which accumulated during a specific interval of time. As a result, each allomember will reflect a restricted set of environmental conditions and will contain archaeological deposits with a limited range of cultural affiliation. In particular, each bayhead delta should contain archaeological deposits with different ranges of cultural periods.
Fluvial Complexes Beneath Galveston Bay The fluvial deposits of the Fields Bayou and Liberty allogroups underlie the Galveston Bay alloformation within under the Galveston Bay system (Figure 3-11). The submerged and buried alluvial plain of the Trinity River varies greatly in width from 3 to 16 km. This alluvial plain is the buried surface of the Fields Bayou allogroup. The Liberty allogroup consists of three large terrace remnants and other small terrace fragments present along the valley walls of buried Trinity River and San Jacinto River valleys (Figures 3-ll and 3-12). Anderson et al. (1991) have observed on seismic profiles 3 broad terraces separated by 3 to 6 m high scarps which form the buried surface of fluvial sediments belonging to the Liberty Allogroup (Figure 3-12). They interpret each of these terraces and scarps to be a single wave-cut bench eroded into the unconsolidated fluvial sands of the Liberty allogroup during a brief still stand of sea level. According to their interpretations, the lowest terrace lying at -12 m below sea level was cut into the Liberty allogroup after an abrupt rise in sea level starting about 6,500 radiocarbon years B.P. Then about 4,000 radiocarbon years B.P., a higher terrace lying at -6 m below sea level was cut after another abrupt rise in sea level (Anderson et al. 1991:Figure 6; Smyth 1991:Figure 5-5).
Effects on Human Adaptation and the Archaeological Record The age and placement of the fluvial terraces within the Trinity and San Jacinto rivers have several significant implications concerning the occurrence of archaeological deposits within the Trinity River Valley and beneath Galveston Bay. First, if the Deweyville and Tanner Bayou alloformations range in age from about 21,000 to 10,000 B.P. or possibly slightly younger, then the formation of these two alloformations would be contemporaneous with the initial human occupation of the Galveston Bay area. As a result, Paleo-Indian 3-43
A" Northeast East Bay
A'
~----------------------------sl--------------------------------4-0
.;,;:;; I
Undifferentiated Pleistocene Strata
A Southwest West Bay
TGB·C
A'
Undifferentiated Pleistocene Strata
LEGEND
1"':;:;:;:;:::;:;:;: ::::':'::""'1 Tidal Sandsdella
Bounding unconformity
1::::::::1 Estuarine
Facies boundary
.:.:.:.:.:.:.:-: sediments
I:iiliiiiilimilil
~~~~~aS~f~~nIS ~G~C R:h7a::: (1969) Core
I';«':!~~':~':'I Fluvial sands with ~:·:S·~:.:.~:.:·; basal gravels
Qgb = Galveston Bayou alloformation
~ Ruvial
Qfb
=
Fields Bayou al1ogroup
L::jsands
Figure 3-11.
Stratigraphic cross-section at the lower end of Galveston Bay (modified from Smyth 1991:Figure 4-23),
3-44
S"
S'
Smith Point
li o"
UndilferenUated Pleistocene Strata
~~~~--------------------------140 S
S'
Point
"., ~
g
0 10
.c
0. 20
"
0
30
"
"
Undifferentiated Pleistocene Strata
40
LEGEND
r:::::::1 ::::::::
lag sands
- - - - Bounding unconformity
j-:-:':':':':':':I
Estuarine
- - - - Facies boundary
Erosional
':::::::. sediments
I::::::~::::::J Overbank and
:::::::::::::::; marsh sediments
r.·:·}. . · ,.. .:··:1
Fluvial sands with
:,~~::<~.:, /~: basal gravels
1-:........ -: -: -: I sands Fluvial
Figure 3-12.
--si--Bay level
T-59 Foundation Boring
3S
Ogb '" Galveston Bayou aUoformation
Ofb '" Fields Bayou allogroup QIi = Liberty allogroup
Stratigraphic cross-section of Galveston Bay from Smith Point to Dollar Point (modified from Smyth 1991:Figure 4-23).
3-45
archaeological deposits could be expected to occur within the thin overbank deposits which form the upper portion of these complexes. Because the accumulation of archaeological deposits was possibly contemporaneous with terrace construction, both buried and stratified archaeological deposits might be associated with the terraces of both alloformations. In addition, the accumulation of sediment might provide varying degrees of preservation for organic remains associated with these archaeological deposits (Ferring 1990a:58, 1990b). Second, because the fluvial complexes and terraces of the Liberty allogroup predate the human occupation of the Galveston Bay area, any archaeological deposits found upon them will have accumulated after the formation of these terrace surfaces, e.g., the subsequent sites of Ferring (1990a:58, 1990b). Except colluvial and eolian processes, pedogenic processes will be the dominate formation processes effecting these sites. As a result, the cultural material at these sites can be expected to have mixed by pedogenesis and either partially destroyed or modified by weathering (Ferring 1990a:58, 1990b). Finally, if climatic processes were, in part, responsible for the oversize fluvial landforms exhibited by the fluvial terraces of both of these alloformations, they demonstrate that the climate during the Paleo-Indian occupation of the Galveston Bay area was significantly different from present in a manner not readily detected by palynological studies within adjacent parts of Texas. Finally, if the sediments and terraces of the Liberty Allogroup are 110,000 or 80,000 years old, then these deposits predate the human occupation of Southeast Texas. As a result, the occurrence of archaeological deposits should be restricted to their surface and to depths that pedogenic processes would mix surface material down into the terrace deposits. Because of these implications, specific data concerning the age of these allostratigraphic units are needed in order to determine their potential for containing archaeological deposits and paleoclimatic implications. Specific data concerning published dates, e.g., their location, stratigraphic context, standard deviations, and values and standard deviations, e.g., Aronow (1967, 1976:10) and Bernard and LeBlanc (1965:149) needs to be located and published. Also, radiometric dates should be disregarded and discarded for specific reasons, not simply because they contradict a specific model or hypothesis. Finally, additional samples of wood and other organic materials associated with each of the fluvial alloformations and complexes need to be collected, documented, dated, and completely published. Within the Galveston Bay allofornlation, the occurrence of prehistoric archaeological deposits will be restricted to specific depositional facies. Within the bay fill, significant 3-46
prehistoric archaeological deposits will primarily be associated with the subaerial deposits of the natural levee and marsh facies of the bayhead delta as documented by Aten (1983). The various bay facies and the fluvial channel or point bar facies are deposited within subaqueous environments and, as a result, lack archaeological deposits. The sedimentary fill of abandoned channels might possibly contain rare, in situ, archaeological deposits. Within the uppermost bay sediments, historic archaeological deposits, e.g. shipwrecks and the debris washed in by hurricanes, could be encountered. Other then the subaerial deltaic deposits, only the surficial soils and dune sands of the barrier islands and spit depositional sequences contain significant prehistoric archaeological deposits.
Sea Level Rise:
Processes and Chronology
During the Late Wisconsinan, sea level dropped to its maximum low stand and then started to rise again to modern levels (see Figure 3-8). During the Late Wisconsinan glacial maximum, approximately 21,000 to 18,000 radiocarbon years B.P., sea level had apparently dropped over 100 m below present sea level. During this time, the Trinity River entrenched its valley as much as 35 to 40 m below the adjacent coastal plain. Sometime between 16,000 to 18,000 radiocarbon years B.P. sea level started to rise. The episodic nature of this rise through the Wisconsinan Substage and Holocene Epoch is evidenced by the series of bayhead deltas indicating still stands at depths of 14, 20, 29, and 36 m below present sea level (see Figure 3-10) (Thomas 1990; Thomas and Anderson 1989). An unresolved controversy concerning the sea level history of the Texas Coast is whether sea level was as high and, at times, slightly higher or lower than present during the Middle Holocene. For the middle and late Holocene, three distinct sea level histories have been developed for the Gulf of Mexico. First, studies using intertidal peats, e.g., Davies (1980) and Robbin (1984) typically conclude a uniform, asymptotic rise to present day level. For example, Coulombe and Bloom (1983) propose that the average rate of eustatic sea level rise was about 8 mm per year from 10,500 to 6,400 years B.P. and less than I mm per year from 6,400 years B.P. to present within the Gulf of Mexico. Second, studies of barrier island deposits, e.g., Stapor and Tanner (1977), Stapor et al. (1991:836), and Tanner et al. (l989:Figure 8) together with the geoarchaeology of archaeological sites, e.g., Johnson et al. (1986), Paine (l987b), Prewitt and Paine (1988), Russo (1991), and Marquardt (1992), on the other hand, indicate that over the past 3,000 to 6,000 radiocarbon years B.P. sea level has fluctuated both above and below present within a 3·47
range of about 1 to 2 m. These studies have been supported by marine bar and shoal deposits within the New Orleans area lying 0.5 m above sea level which have been dated at about 5,000 radiocarbon years B.P. by Otvos (1978). A reexamination of these deposits by Dr. Gregg Stone (personal communication 1992) and Dr. Frank Stapor (personal communication 1992) have confirmed Otvos' (1978) observations. Finally, studies of the sedimentology and seismic stratigraphy of the deltaic deposits of the Mississippi Delta and the fill of the entrenched valleys of the Sabine-Trinity Rivers show sea level as having risen over the Holocene as periods of short· term rapid rises in sea level alternating with somewhat longer periods of relative sea level still stand. According to most of these studies, e.g. Anderson et al. (1991), Thomas (1990:183-191), and Penland et al. (1988, 1991), modern sea level was reached only about 2,500 to 3,500 radiocarbon years B.P. Also, they conclude that between 4,000 to 6,000 radiocarbon years B.P sea level was at a depth of 5 m below present sea level instead of varying a couple of meters above and below modern sea level. Because the evidence for each group of studies are equally convincing, but contradictory, it is difficult to decide which of these models best describes the Holocene sea level changes along the Louisiana and Texas coasts. However, because they are all internally consistent, but disagree only between methodologies, these sea level histories in some sense could all be accurate as proposed by Stapor et al. (1991:815). The differences in sea level histories might arise because each methodology, apparently, measures a different aspect of sea level change, plus the inability of radiocarbon dating to resolve high frequency fluctuations during a gradually rising Holocene sea level. For example, beach ridges, wave·cut scarps, reworked shells, and archaeological sites record the high stands of individual high frequency fluctuations while peats preferentially record low and rising levels of such high frequency sea level fluctuations (Stapor and Tanner 1977; Stapor et al. 1991 :815). Between 3,000 to 6,000 radiocarbon years B.P., sea level history as inferred from beach ridge and sequence stratigraphic studies is difficult to reconcile. For three reasons, it is definitely a questionable practice as practiced by some studies, e.g. Thomas (1990), to use sea level curves developed by Fairbanks (1989) and modified by Bard et al. (1990) for Barbados to date flooding surfaces and other sea level indicators within the Galveston bay area. First, the 3 to 5 m depth range of the coral, Acropora palma la, used to construct their curve introduces a considerable uncertainty into this curve. Second, Fairbanks' (1989) sea level curve indicates that "present·day" sea level occurs at depth of 3 m below 3·48
modern for unspecified reasons in an area with a local tidal range that fails to exceed 0.7 m. Finally, the sea level history of an island like Barbados will differ considerably from the Texas coastal plain and continental shelf, because of different isostatic responses from the type of crust underlying each area to hydrostatic loading (Bloom 1967). Within Galveston Bay, potential evidence for sea levels varying above and below modern between 3,000 to 6,000 radiocarbon years B.P. has been noted. Henry (1956:29-36) noted possible evidence for such a high stand of sea level in the form of a geomorphic surface, called "Swan Lake Flat," just south of Texas City, Galveston County. Although he concluded that it consisted of sediments deposited by tropical storm surges associated with modern sea level formed it, the similarity of Swam Lake Flat to the flats associated with the Swan Lake Site (41AS16) described by Paine (1987b) demonstrate that additional studies of these deposits in warranted. Aronow (1984) noted the presence of a continuous surficial layer of sandy or loamy sediment with Rangia sp. shells within the area of Sites 41GV14, 41GV15, and 41GV16. He interpreted these sediments to be either eolian deposits, hurricane overwash, or fill. However, they might represent the shoreward edge of Middle Holocene coastal deposits formed during higher than present sea level stands. Unfortunately, because of the intense shoreline erosion along Galveston Bay, conclusive evidence either for or against a higher than modern Holocene high sea stand has largely been obliterated. At this time, Thomas (1990), Anderson et al. (1991), and Smyth (1991) appear to have the best model for the Holocene sea level history of the Texas Gulf Coast and Galveston Bay. However, additional research is definitely required to determine the significance of contradictory evidence offered by Stapor and Tanner (1977), Stapor et al. (1991:815), Paine (1987b), Marquardt (1992), Dr. Gregg Stone (personal communication 1992), and others. Clearly, additional research is required to clearly establish the history of Holocene sea level change and its influence in the preservation of archaeological deposits and contemporaneous settlement patterns within the Galveston Bay area. Whether sea level either varied above and below modern or occurred at a depth of 5 m below present sea level between 3,000 and 6,000 radiocarbon years B.P., has extremely significant implications concerning the preservation of cultural resources and contemporaneous settlement patterns within the Galveston Bay area and the entire Texas Gulf coast. The flooding of the Trinity River and San Jacinto River valleys within the modern Galveston Bay system occurred during the Holocene Epoch. According to Thomas (1990) and Smyth (1991), the first effect of sea level rise consisted of the development of extensive 3·49
marshes covering the alluvial plain adjacent approximately 9,000 to 10,000 radiocarbon years B.P. (Figure 3-13). Approximately 8,000 radiocarbon years B.P., the bayhead delta deposits at 20 m below modern sea level were flooded by a rapid rise in sea level. This abrupt rise in sea level ftrst established widespread estuarine conditions within the Galveston Bay system at a level 14 m below present sea level, at which a bay head delta developed. Sea level continued to rise, possibly, stopping during three brief still stands. During each of these still stands, terraces, interpreted to be erosional terraces, might have been cut into the sediments of the Liberty allogroup and the valley walls of the Trinity and San Jacinto River Valley (see Figure 3-12). The lowest of these terraces which lies 12 m below present sea level was cut approximately 7,000 radiocarbon years B.P. by a brief still stand. The highest of these terraces was cut approximately 4,000 radiocarbon years B.P. as modern estuarine conditions were established. By approximately 2,500 radiocarbon years B.P., Galveston Island and Bolivar Peninsula had become established and the modern Galveston Bay system had become established (Anderson and Siringan 1992:9-10; Anderson et al. 1991: 10-11; Smyth 1991).
Effects on Human Adaptation The currently available archaeological evidence indicates that human popUlations have occupied the Galveston Bay region for the past 11,000 years or so. As discussed above, this was a period characterized by signiftcant environmental changes in the present-day Galveston Bay area, driven, primarily, by rising sea levels and concomitant shifts in plant and animal biomes. Within the Galveston Bay area, as sea levels rose, what had formerly been inland river valley settings, gradually were inundated by gulf waters and transformed into a brackish marsh-estuary system, and, ftnally, much of it was transformed into an open saline bay. Much of the exposed land surface available to populations of 10,000 years ago have been flooded and submerged. These significant landscape and environmental changes significantly influenced human patterns of adaptation to the study area and, additionally, the physical processes associated with these changes have impacted to varying degrees upon the resultant archaeological record. As a result, both the management and study of cultural resources which may be associated with the Galveston Bay Navigation System are dependent upon an understanding of these past environmental changes. Aten (1983) has presented a detailed discussion of the postulated post-glacial environmental settings of the southeast Texas area which is applicable to the Galveston Bay system. His concepts are followed closely in the following discussions. 3-50
....... ··...... ........ .. ........ ·......... ........ ..· ::....... .... ::: .... ........ ·· ...... . ....... ...... .·........ . . . . . ... . .................... ......... . ......... ·.·· ......... ......... . . . . . . ...... ......... . ... .. . . ",".",
'
LEGEND
I::!:!!!~ ..... " Holocene deltaic plain
I:: :: :: ::1 Prairie Terrace
r.·.·.·.1 Undifferentiated .................. swamp and marsh
I::::::::::::::j Undiffereotiated ............... fluvial terraces
E3
~
Holocene floodplain
L:..:...:..:... and incipient delta plains
~
Barrier island
and spit I~I Open gulf or ~
baywaters
Figure 3-13. Speculative reconstruction of Holocene paleogeography within the Galveston Bay area. Constructed from data and figures from Rehkemper (1969a), Thomas (1990), and Smyth (1991).
3-51
It is apparent from the distribution of known prehistoric sites that human populations in the coastal zone even at an early period were involved in the exploitation of a range of coastal resources_ The great abundance of fish and shell fish, and the ease with which they could be exploited, would certainly have been attractive to human populations. The known distribution of archaeological sites in fluvial and estuarine settings of the coastal zone reflect these adaptive patterns_ Within the alluvial plains and estuaries of rivers along the Texas and Louisiana Gulf coast, archaeological deposits commonly are associated with five specific geomorphic contexts_ First, a basic location for the occurrence of archaeological deposits are the natural levees of the modern floodplains of rivers such as the Trinity, Sabine, and Pearl Rivers (Aten 1983; Gagliano 1977:335; Pearson et al. 1986:35-39). Second, archaeological deposits are also commonly found along terrace edges adjacent to either the modern floodplain or large, oxbow lakes on lower terraces along the Sabine and Pearl Rivers (Gagliano 1977:216,334; Pearson et ai, 1986:35-39). Third, Gagliano (1977:336) and Gagliano et al. (1982:39) demonstrate that dense concentrations of archaeological deposits occur along the terrace edges which form the margins of estuaries. Fourth, the deltaic plain of the modern, Late Holocene, Trinity River delta is a prime location for the location of archaeological deposits (Aten 1983). Finally, archaeological deposits typically occur upon the beach ridges and other portions of barrier island-spit complexes such as Galveston Island (Aten 1983; Gagliano 1977:184; Gagliano et al. 1982:33-36). Relying on these perceptions about known site locations, projections can be made concerning archaeological site distributions within the inundated Galveston Bay system, using the allostratigraphic framework discussed earlier. Within the alluvial valley of the Trinity River, four major stratigraphic units, the Liberty allogroup, Deweyville alloformation, Tanner bayou alloformation, and the Trinity River alloformation can be discerned. The Liberty allogroup predates the last low stand of sea level about 21,000 to 17,000 radiocarbon years B.P. Therefore, except where Holocene overbank or colluvial deposits have accumulated upon the fluvial terraces of the Liberty allogroup, archaeological deposits will be restricted to the surface of its terraces. Multicomponent archaeological deposits on its terraces likely consist of mixed assemblages that lack any significant stratigraphic separation. Contrary to recent research, this study tentatively concludes that the Deweyville and Tanner Bayou alloformations are Late Pleistocene to, possibly, early Holocene in age. As a result, archaeological deposits with buried and stratified Paleo-Indian components might occur within the thin blanket of overbank deposits which form both terrace surfaces. Younger archaeological deposits would 3-52
occur as surficial accumulations on their Terrace T2 and Tl. The Trinity River alloformation consists entirely of Holocene fluvial deposits. Its natural levee deposits have a high potential for both buried and surface archaeological deposits. Within the estuarine deposits of the Galveston Bay system itself, prehistoric archaeological deposits will be primarily restricted to the plains of bayhead deltas and the dune sands and surface of beach ridges of barrier island and spit sequences. The delta plains and their associated archaeological deposits have a high preservation potential. The delta plains occur in a specific stratigraphic position within the estuarine deposits that be easily mapped and dated using preexisting seismic and radiometric data (Anderson et al. 1991:Figure 2). During the Holocene transgression, the subaerial portions of relict barrier islands and spits have been completely eroded and redeposited as marine sand shoals. Presumably, any archaeological deposits associated with have been destroyed. In addition, the model prepared by Pearson et al. (1986) for predicting the potential for site preservation within the offshore Sabine River Valley is applicable to the Galveston Bay system. The floodplains of the Trinity and San Jacinto Rivers appear to have been submerged and buried intact. As a result, their is a fairly high probability that the archaeological deposits associated with natural levees and and other fluvial landforms on these floodplains are also intact. The fluvial terraces of the Liberty allogroup and the valley walls and terrace of the Beaumont Formation have been deeply eroded by repeated still stands of sea level. Because any archaeological deposits associated with these deposits would be surface sites, this transgressive erosion would have destroyed any archaeological deposits.
Only within
topographic lows on this surface is there a chance that archaeological deposits have been preserved. As a result, archaeological deposits associated with valley walls and estuarine and terrace margins have very low potential for surviving the initial flooding of the San Jacinto and Trinity River Valleys by rising sea levels.
Effects
011
the Archaeological Record
Rising sea level has had a variable impact upon the pretransgressive archaeological remains existing beneath present-day Galveston Bay. First, as in case of Sabine River Valley of the offshore continental shelf of Texas and Louisiana, archaeological deposits present on the buried floodplains of the Trinity and San Jacinto Rivers have a very high preservation potential (Pearson et al. 1986). Prior to the transgression of the bayline flooding surface, called the "estuarine transgressive zone" by Pearson et al. (1986:224-225), freshwater marsh deposits 3-53
completely bury the floodplain. As sea level rises, the marsh deposits are buried by bay sediments such that the eventual passage of the shoreface over the estuary fails to disturb them if buried sufficiently deep (Thomas and Anderson 1988; Thomas 1990; Smyth 1991). In this aspect, the model developed by Pearson et al. (1986:224·225) for the offshore Sabine River Valley also applies to Galveston Bay. Second, the formation of erosional scarps and terraces during periodic stillstands of sea level greatly effected the potential for the preservation of archaeological deposits on terraces margins. As previously noted, seismic profiles show that during the submergence of Galveston Bay, stillstands might have cut three broad erosional terraces deeply into the unconsolidated fluvial sands of Liberty allogroup during either still stands of or periods of slowly rising sea level (see Figure 3·12). In addition, during the same still stands, the stillstands would have also eroded the stiff Pleistocene clays where the Beaumont Formation formed the valley walls (Anderson et al. 1991: 10). As a result, the bay line flooding surface probably has destroyed any archaeological deposits associated with the former estuarine margins, fluvial terraces, and valley walls. Because the fluvial deposits of the Liberty allogroup predate the human occupation of the Galveston Bay area, its surficial archaeological deposits would have been destroyed along with its terraces. According to Smyth (1991) and Anderson et al. (1991), the center of larger fluvial terrace remnants might remain intact during the transgression of the bayline flooding surface. Where the fluvial terrace is intact, it is possible that archeological deposits might have survived erosion by the bayline flooding surface. Also, within the East and West Bays, the passive flooding of the Prairie Terraces has likely preserved many archaeological deposits. However, any landward movement of the shoreface associated with Galveston Island and Bolivar Peninsula will result in the destruction of any deposits buried beneath the sediments of these bays. Third, like the alluvial plains, marsh deposits apparently bury the bayhead delta plain prior to the transgression of the bayline flooding surface. As a result, the submergence of the bayhead delta beneath Galveston Bay would fail to significantly disturb its surface. Once submerged, the accumulation of bay deposits will protect these deposits (Figure 3-14). The bayhead deltas lie deep enough within the fill of the Trinity and San Jacinto River valleys such that the passage of the shoreface and the formation of a ravinement surface will fail to disturb the bayhead delta and any associated archaeological deposits (Thomas 1990; Thomas and Anderson 1988). However, Aten (1983) documented significant destruction of archaeological 3·54
Trinity River
San JacinlO
River
29.5 0
29.00
o!
10 !
20 !
30 !
40 krn !
L.._ _ _ _ _ _9_5.l.o_o_ _ _ _ _
Gulf of Mexico
...... ... . ....
J. :.:i. :.~1i.l~.~"'~ " ~i:.J.i)
_____ 94.J.'_00_ _ _ _--....128.5
0
LEGEND
o
o o
o
Bayhoad Delta Deposits at Depths 01·14 m Below Saa level Bayhead Delta Deposits at Depths of ·20 m Balow Sea level Bayhead Delta Deposits at Depths of ·29 m Balow Sea level Bayhead Delta Deposits at Depths of ·36 m Below Sea level
I
~::::::: Area of Galveston Bay and continental shelf underlain by .:.:.:.: Gafveston Bay alioformaUon.
Figure 3-14. Distribution and depth of bay head delta deposits within the Galveston Bay alloformation (modified from Thomas 1990:Figure 2·21).
3·55
deposits within the delta plains caused by the enlargement of deltaic lakes and bayous by erosion of their shorelines. Finally, the future preservation potential for archaeological sites upon Galveston Island and Bolivar Peninsula is low_Seismic data, cores, and inspection by scuba divers clearly document that the Sabine Bank, Heald Bank, and the" -29 moo bank are marine shelf sand bodies. The formation of the ravinement surface by the migrating shoreface has completely destroyed the barrier islands and spits as which these sands initially accumulated. The shelf processes have completely reworked and redeposited these sediments as marine shelf sands (Thomas and Anderson 1988, 1989; Thomas 1990:202-204)_ The shoreface and shelf processes would have destroyed any archaeological deposits which might have once been associated with these former barrier island and spit chains. If in the future significant rises in sea occur, then Galveston Island, Bolivar Peninsula, and archaeological deposits upon them, will suffer the same complete destruction as the older barrier islands and spits_ As during preceding still stands, the waves and other processes within Galveston Bay are eroding it shorelines. Currently, the average rate of erosion varies from 1.8 m per year for sand and shell beaches to 1.2 m per year for salt- and brackish-water marshes and 1.0 m pre year for bluffs composed of Pleistocene clay. The specific orientation, bay fetch, and composition of each shoreline determine the rate of retreat for individual stretches of shoreline, while local subsidence, changes in sediment supply, and frequent and intense storms contribute to overall changes in shoreline. Within the Trinity Bay area, land subsidence resulting from the withdrawal of groundwater and, to a much lesser extent oil, has considerably accelerated the pace of shoreline erosion. Changing climate, gradually rising sea level, local and regional subsidence, decreasing sediment supply, recurring storms, and ongoing human activities all promote continued shoreline erosion (Paine and Morton 1986; Morton and Paine 1990). As a result, the continuing destruction of archaeological deposits to shoreline erosion will always be an ongoing problem within the Galveston Bay area.
3-56