Site Description

Clovis Occupation in Central Texas

by Michael B. Collins and Thomas R. Hester

Gault is a large site with a conspicuous, dark midden over an area approaching 0.8 by 0.2 km. It occupies a verdant cove along the upper reaches of Buttermilk Creek, which is entrenched more than 15 m into the limestone bedrock of the Lampasas Cut Plain. Ten kilometers downstream to the east, Buttermilk Creek emerges from the limestone upland and joins Salado Creek on the western margin of the Black Prairie. The Lampasas Cut Plain is a subregion of the Edwards Plateau and the Black Prairie is a subregion of the Gulf Coastal Plain. A major ecotone marks the boundary between the coastal plain and the limestone plateau, two regions of contrasting bedrock, topography, soils, flora, fauna, and even climate to a minor extent. The dissected eastern fringe of the plateau is notable for its numerous springs and extensive outcrops of Edwards chert, both of which occur at the Gault Site. Hunter-gatherers living in this ecotone could optimize their access to a wider array of plants and animals than could be found in either single region. At a much smaller scale, the mesic valley floor at the site contrasts in resources with the surrounding xeric uplands. Several small, perennial springs well up along the lower slopes of the valley walls and at places along the creek in the valley floor. Willow, elm, burr oak, walnut, and bois d'arc trees along the valley floor stand in marked contrast to scrub oak, mesquite, and cacti on the immediately adjacent uplands.

Cultural materials comprising the site occur north and south of the easterly-flowing creek. They are exposed on the surface, shallowly buried in thin soils on the upland surface, at the surface (as a midden), and more deeply buried in fluvial deposits on the valley floor. There is also cultural material on and in colluvial slope deposits along the valley margins.

Burned rock and large quantities of chert comprise a major portion of the midden debris at the Gault Site; very little in the way of bone, shell, or charcoal is present. As so often happens, the high visibility and richness of the Gault midden has attracted artifact collectors and commercial looters over most of the past 80 or 90 years (Collins 1998; Hester and Collins 1998). Early in that era of relic-collecting, word of the site reached Professor J. E. Pearce, founder of the Department of Anthropology at the University of Texas, and he conducted limited excavations at the site in 1929-30 (Figure 1). Pearce's work and the digging by most of the collectors and looters were confined to the massive, dark-colored midden deposits. TARL houses the materials recovered by Pearce, a collection that has been studied repeatedly over the years (for example, Elton Prewitt's definition of the Andice Dart Point type was based in large part upon specimens from the Gault Site [Prewitt 1983]).

Back to Top

Other than a brief effort by archeological staff of the Texas Historical Commission to document the active looting at the site in 1988 (Figure 2), no other professional work was undertaken until 1991. That effort was in response to reported finds of Clovis-age engraved stones at the site in 1990 by David Olmstead, a collector (Collins et al. 1991).

Olmstead had dug well below the dark midden into sub-midden valley alluvium and recovered four engraved stones and two Clovis points, evidently in direct association (Figure 3). After casting and photographing these six objects, Pete Bostrom of the Lithics Casting Laboratory in Troy, Illinois, contacted Hester and Collins at the University of Texas at Austin to suggest that we contact Olmstead and learn more about the finds. As a result, in 1991 the Texas Archeological Research Laboratory (TARL) conducted a brief, controlled excavation close to the spot of Olmstead's finds near the southwestern edge of the site. TARL corroborated the essential facts of Olmstead's account by finding engraved stones with diagnostic Clovis artifacts (Collins et al. 1992; Hester et al. 1992).

In the area of Olmstead's find and the 1991 TARL excavation, the stratigraphic section consists of a colluvial toe-slope overlying valley fill deposits. The latter have been strongly affected by the upwelling of spring water. In fact, an elevated water table at the time of excavation prevented the project from reaching the base of culture-bearing deposits in the valley floor. From the base upward, the exposed section (Figure 4) consists of bedrock and four late-Quaternary units (Zones 1 through 4) overlying limestone bedrock (Collins et al. 1992; Hester et al. 1992). Zone 1 on the slope is silty clay with carbonate granules, lacking in cultural materials. Zone 2 expresses weathering at its upper surface and has downslope, fluvial (2a) and upslope, colluvial (2b) facies. Zone 2a is a gray silty clay with carbonate granules and limestone clasts; only the upper decimeter of Zone 2a was investigated due to the ground water. One complete Clovis point, Clovis blade segments, engraved stones (Figure 5), a fragmentary chipped stone adze, an unidentified lanceolate point fragment, flakes of quartz and chalcedony, and bifacial performs were recovered. Evidently, it is from lower in this zone that collectors have recovered engraved stones and Clovis diagnostic artifacts at times of lowered water table. Zone 2b is a massive clay loam with deeply weathered limestone clasts that yielded only a few cultural flakes. The intermediate deposit, where Zone 2b is intermixed with overlying Zone 3, contains various late Paleoindian and early Archaic artifacts. Zone 3 is a clay loam with carbonate granules with considerable cultural material of late Paleoindian and early Archaic affiliations. The surface unit, Zone 4, consists of a dark anthrosolic clay loam with artifacts of late Archaic Age.

Uncontrolled digging by collectors and commercial looters continued from 1991 until new owners acquired a portion of the site in 1998. Although they immediately closed the site to collectors, the landowners began to dig for relics for their own collection. When they encountered mammoth bones associated with stone tools in July 1998, they contacted Collins and Hester and invited TARL to visit the site and investigate this discovery. Limited excavations were carried out in the late summer and fall of 1998 by staff, students, and volunteers under auspices of TARL and the Vertebrate Paleontology Laboratory of the University of Texas at Austin. The landowner's finds were made near the southern margin of the valley in proximity to a small spring. They had exposed several mammoth bones and an extraordinary number of stone artifacts exhibiting Clovis attributes. Upon investigation, additional stone artifacts (a Clovis point [Figure 6]), blades, blade cores, and flakes were found in association with mammoth bones and a mandible (Figure 7) and an unidentified bone that, because of its size, is also suspected to be mammoth. In addition, the project gained a general understanding of the stratigraphic sequence in this part of the site (Collins 1998, 1999; Hester and Collins 1998; Lundelius 1998). The section here consists of better-defined stratigraphic units and has been more fully investigated than the one examined in 1991. Correlations can probably be inferred between the two, but nothing remains of the section in the vicinity of our 1991 excavations because looters destroyed it shortly after the work ended.

Back to Top

In proximity to the 1998 mammoth bone finds, excavations have penetrated a suite of diverse geologic deposits, paleosols, and anthrosols. Most of the geologic strata contain artifacts and some also contain fossil bone. Deposits vary from boulder gravels to fine clays that formed in several different depositional environments. Buried soils that formed on once-stable land surfaces are observable in several exposures across the site, the best-defined being an early Holocene paleosol. This soil formed with little in the way of human influence. The single most prominent aspect of the site is a massive anthrosol (midden) extending from the present surface to depths of almost a meter. This midden was formed over the last 6,000 years or more and reflects extensive and, at times, intensive human occupations of the locality.

From the base upward, the geological sequence consists of bedrock, two deposits of gravel, clay, a paleosol, colluvium, and an anthrosol (Figure 8). The section is less than 2 m thick and represents approximately the last 12,000-radiocarbon years of time. Each of the zones and any cultural materials contained in this section are described individually below.

Bedrock is Cretaceous limestone. The contact between the underlying Comanche Peak Formation and overlying Edwards Formation is exposed along the valley walls at the site, and the lowest zone exposed in our excavations is saprolitic Comanche Peak limestone (Zone 7). The upper surface of the Comanche Peak is unconformably overlain by Quaternary valley fill (Zones 6C through 1).

Typically in the area of our excavation a dense deposit (Zone 6C; actually a series of lenses) of well-rounded pebbles, cobbles, and small boulders of chert and limestone lies in direct contact with bedrock. A few unidentifiable pieces of water-rolled bone fragments, bits of mammoth tooth enamel, and well-preserved horse teeth occur in this zone; no cultural artifacts are present. Limonite and manganese coat the surfaces of most bones and clasts in this zone.

In places this zone appears to be a gravel of water-transported clasts and in others it may consist of, or include, angular colluvial clasts that became rounded in place by dissolution in groundwater. Teeth and poorly preserved bones of mammoth, horse, and bison are found in and on these cobble lenses. Chipped stone artifacts are present among the rounded limestone clasts. Among theses are numerous pieces with clear Clovis affiliations, including points (Figure 9), blade cores, blades, tools made on blades, and an adze (Figure 10). These chert artifacts do not exhibit damage from stream transport. Also present as lenses, both in and above the rounded limestone clasts, are deposits of ultra clays (Zone 6B). These exhibit a blocky structure that may indicate soil development, but they could also be lenses of ponded clay that, because of their nearly pure nature, developed a structure that mimics that seen in clay-like subsoils. These, too, contain abundant chipped stone artifacts including Clovis diagnostics. Some of the fossil animal bones and teeth are found at the base of these clay lenses. In places a lens of limestone clasts has been identified above the clay; this is depicted in Figure 1 as zone 6A.

Back to Top

Next in the sequence is a clay rich deposit, typically around 60- to 70-cm thick, that may be largely fluvial in origin with some intermixing of ponded sediments and colluvial material (Zone 5). It contains solution-rounded limestone clasts that decrease in frequency with distance from the bedrock valley wall. There are multiple lenses of chipped stone artifacts in this zone, evidently representing multiple intervals of occupation, and possibly even discrete living surfaces. Most of the lithics are lying flat and have pedogenic carbonate pendants on their lower surfaces. Clovis points (Figure 9), blades (Figure 11), preforms (Figure 10), channel flakes, distinctive bifacial thinning flakes (including overshot flakes), and tools made on blades have been found throughout this zone. Faunal material is virtually absent.

A strongly expressed paleosol formed at the top of the clay-like zone. The organically-enriched A horizon of this soil is some 20 cm thick and it is almost devoid of artifacts (Zone 4A); beneath this is a carbonate-enriched subsoil (Zone 4B). Only one artifact from Zone 4A is potentially time-diagnostic; it resembles Wilson points (ca. 10,000- 9,000 B.P.) as defined at the nearby Wilson-Leonard site (Collins 1998). This soil is believed to be correlated with the Royalty Soil; the sediments, the clasts, and lithic artifacts are all stained a yellowish color. This is tentatively believed to be the result of iron carried in ground water. It is also thought that the surface of the water table fluctuated from a low level near bedrock to a high level near the top of the Royalty Soil. This fluctuation meant that the lowest portion of the Quaternary section remained saturated most of the time, the intermediate portion was alternately wet and dry, and that the upper portion was saturated only infrequently. Bone is found only in the lowest portion of the section where it is believed to have remained saturated most of the time.

Delicately engraved limestone clasts (Figure 3) are among the artifacts found at the Gault site. A few of these are of Clovis age whereas others clearly date to the early Archaic. This suggests that this art form was in vogue for at least 3,000 years.

Another clay rich deposit (Zone 3), believed to be largely colluvial in origin, overlies the Royalty Soil. This zone lacks the iron staining and contains angular, as distinct from the previously-described rounded, limestone clasts. These angular clasts decrease in frequency and size with distance from the bedrock valley wall. In this zone are found a large number of artifacts including many diagnostic of the earliest part of the Archaic tradition in Central Texas. These include Clear Fork tools, a Waco sinker, and many examples of such projectile point types as Angostura, Hoxie, Gower, Uvalde, Martindale, and other bifucate-stemmed forms (Figure 9). A generalized age of ca. 8,800 to 8,500 years is indicated by this array of materials. Because it is consistently bracketed by temporally-diagnostic artifacts, the above-described paleosol is correlated with the Royalty Soil. At Gault, this soil immediately overlies Clovis artifacts, contains what may be a Wilson point, and is overlain by a zone containing artifacts datable to ca. 8,800 B.P. Numerous radiocarbon dates for the Royalty soil at Fort Hood range from ca. 11,325 to 8,260 B.P. (Nordt 1992); however, most fall in the interval 10,000 to 9,500 B.P.

A distinct layer of broken snail shells (Zone 2) is the next recognized zone in the section. No specific interpretation is offered for this zone at present.

Back to Top

The uppermost zone in this section at Gault is an anthrosol (Zone 1), typically about 30 cm in thickness. This dark-colored, clay rich zone contains burned rocks, rare pieces of burned animal bone, and moderate numbers of chipped shone artifacts. Comparatively few diagnostic pieces have been recovered. One potsherd as well as projectile points of late Archaic affiliation including Ensor, Fairland, and Fairland-like forms (Figure 9) have been noted.

It would appear that stratigraphic Zones 7 through 4 represent valley alluviation, possibly correlative with the Georgetown Alluvium at Fort Hood (Nordt 1992, 1995) followed by stabilization of the valley floor and soil formation (Royalty Paleosol, Zone 4A). Subsequently, alluvial fans and colluvial slopes (Zone 2 and 3) extended onto the top of the Royalty Soil and eventually stabilized with formation of another soil (Zone 1). Humans frequently returned to this locality for centuries at the end of the Pleistocene and for millennia during the Holocene.

The Clovis record is notable for its evidence of multiple occupations, large occupation area, vast amount of cultural debris, diverse tool kit, and debris from all stages of knapping bifaces and blades. Of particular interest are blades used for cutting meat and scything grass (Inman and Hudler 1998) as well as the adze, a wood-working tool. The Clovis materials from Gault will significantly revise the long-popular interpretation of Clovis as nomadic mammoth hunters.

Back to Top