Pavo Real

In early Paleoindian times (about 12,000-13,600 years ago) small groups of Clovis and Folsom people camped at Pavo Real for short stays during which they refurbished some of their tools and weapons from the abundant supply of chert they found at the site. Clovis and Folsom are the names of archeological cultures or cultural traditions named by archeologists for their distinctive lanceolate spear and dart tips (projectile points). The finding of a few artifacts below the main Paleoindian layer at Pavo Real suggests that the locality was visited by people at an even earlier time. These may have been early Clovis people, although we are not sure.

Although archeologists think that Folsom culture succeeded Clovis culture, artifacts from the two cultures were found in the same layer at Pavo Real and could not be "stratigraphically separated." In other words, the Folsom materials did not occur in a separate layer overlying that containing the Clovis materials. Perhaps the Clovis and Folsom occupations of the site (and these may well represent more than two visits to the site) were not separated by very much time. Alternatively, sediment accumulation may have been so slow that Folsom peoples camped on the same surface as had Clovis peoples in earlier times, such that the artifacts from both were intermingled.

Throughout the Paleoindian occupations, knappers made use of chert obtained at or very close to the site. Lithic (stone) artifacts are essentially the only preserved cultural evidence upon which we can base our interpretations of the lives and activities of those who camped at Pavo Real. Techniques of stone tool manufacture and aspects of stone tool use can be inferred in some detail; from these inferences, the setting of the site, and comparative data from other early Paleoindian sites, limited interpretations can be made regarding broader aspects of the prehistoric lifeways represented at this site.

Pavo Real was subject to repeated flooding during the era in which the Paleoindian occupations occurred. It is situated on the inside of a bend in Leon Creek on a point bar in a region known for flash flooding. What is preserved of the site is a flat terrace surface bounded on its east by a low bedrock bench and on its west by a steep bank dropping off toward the creek. At the time of the Clovis and Folsom occupations, the site area was minimally 50 meters long and 40 meters wide (about 160 by 130 feet). Materials left by these early occupants were buried by overbank flood deposits across the flat terrace. It is likely that some evidence of the early Paleoindian occupations was washed away at times when high-energy floods eroded the creek bank.

In 1979-1980 when Pavo Real was dug, there was little challenge to the Clovis First theory that Clovis was the first North American culture or to the widely held ideas that Clovis was succeeded over much of its range by Folsom, and that the archeological remains left by both of these were characteristic of nomadic big-game hunters. Clovis hunters were thought to have preyed on mammoths and Folsom hunters on now-extinct species of bison. It was often hypothesized that early Paleoindian hunters may have been responsible for the extinction of these members of the terminal Pleistocene megafauna (large Ice-Age animals).

Excavations at Pavo Real began in the belief that only Archaic archeological deposits were present, a view that was reinforced when a seemingly artifact-free gravel layer was encountered in test units below artifacts and features of early Archaic age. Such gravels were (erroneously) considered by the site's investigators to be the geologic marker of the glacial maximum (peak of last Ice Age) at 17,000 or 18,000 years ago. Near the close of scheduled excavations, a Clovis point was found at the base of the Archaic deposits. A hurried testing ensued in search of additional Paleoindian artifacts and these were, in fact, found. Most of these were found within a flood deposit that was below the gravel then thought to represent the glacial maximum.

The dig was extended, excavation strategy was modified, and several experienced archeologists were added to the field crew, but no specific research goals were identified and pursued. Machines were used to remove the remaining Archaic deposits above the layer (called the Folsom zone or Zone 5) bearing the Paleoindian artifacts. Zone 5 was excavated by hand, and a high proportion of the exposed artifacts was mapped in place. A number of these artifacts included diagnostics (distinctive artifact types) matching Clovis and Folsom assemblages at other, sealed sites.

In much the same way that Clovis and Folsom Paleoindian artifacts were accidentally discovered beneath the Archaic deposits, another component of unknown affiliation was found by chance in deposits underlying the unit containing Clovis and Folsom materials. This lowest cultural component was barely investigated and too little was recovered to suggest what affinities it might have to other early assemblages. For the most part, it seems to date from early Clovis times. (A single, apparently man-made flake was recovered from a still deeper deposit that clearly predates Clovis, but it is hard to say much about a single flake.)

Analysis Results

Two of the preliminary questions addressed by Michael Collins and his research team at the Texas Archeological Research Laboratory were: (1) Were the Clovis and Folsom materials in primary context (where they were originally dropped) or secondary context (moved and redeposited during floods)? and (2) Was there evidence of any stratigraphic separation between the Clovis and Folsom components?

There is no geologic evidence that the cultural materials were transported from elsewhere and deposited together (such as by stream or flood action), nor was there any indication of a stratigraphic break or separation between Clovis and Folsom materials. These geological inferences are entirely consistent with the archeological evidence. The physical condition of the artifacts do not indicate that they were transported nor is there horizontal size-sorting suggestive of stream transport; further, the Clovis and the Folsom artifacts exhibit the same degree of patination (weathering) and surface damage. The most compelling evidence for primary context is the large number of refits, that is, artifact fragments that fit together (conjoin), showing they were once part of the same tool or were made from the same piece of chert. The presence of pieces of large and small knapping debris that conjoin one another and were found close together in Zone 5 is a clear signature of artifacts in primary context.

Within Zone 5 there is no evidence of the vertical separation of Clovis and Folsom artifacts. Geologically, Zone 5 accumulated intermittently over an unknown interval of time beginning before and continuing after the majority of the Clovis and Folsom artifacts were dropped. Probably the best explanation for the component mixing at Pavo Real is that natural deposition was slow enough for all of these cultural materials to accumulate in a very few centimeters of sediment but rapid enough that no stable surface formed.

Various techniques were used to date the Paleoindian deposits and related geological layers at Pavo Real. Tiny fragments of charcoal were found in the site's deeper and earlier deposits, giving the original investigators hope that these could be used to obtain radiocarbon age estimates of the Paleoindian occupations. Unfortunately, the resulting dates fell with early Archaic times, suggesting that the charcoal was introduced into the lower deposits by pit-digging or other disturbances. Years later, attempts to date snail shells (one of the few organic materials that did survive) also were not satisfactory for complicated reasons we won't go into. Finally, a technique called OSL (optically stimulated luminescence) dating was attempted. OSL dating was more successful in that estimates for the absolute ages of Zones 5, 7, and 9 were obtained. Unfortunately, this dating technique does not yield precise age estimates and results n rather wide ranges of possible ages.

It appears likely that the Paleoindian component at Pavo Real was nearly totally excavated. Artifact counts dropped off abruptly along the west, south, and east margins of the excavated area. On the north, chert counts were still high, but the proportion of these that were cultural in origin was dropping off and the percentage of natural or stream-rolled chert fragments was increasing. The exposed area of occupation was roughly 50 m by 12 m. A majority of the artifacts were found in a layer that was no thicker than 20 centimeters (8 inches).

To sum up the key findings regarding the Paleoindian component at Pavo Real: (1) Clovis and Folsom artifacts were found together in a loamy fluvial (water-laid) deposit with no evidence for post-depositional (later) mixing or that they accumulated on a stable land surface; (2) the Clovis and Folsom artifacts were dispersed over an elongate area covering about 550 square meters (almost 6,000 square feet); (3) very little of the Paleoindian component went unexcavated; (4) the Paleoindian artifacts were buried by repeated overbank flooding of relatively low energy; and (5) there is no indication that the stream moved these artifacts prior to their burial.

Because of these factors, the Paleoindian artifacts and two cultural features (both clusters of artifacts) can only be attributed to the Clovis or to the Folsom component on the basis of diagnostic features—lacking these, no specific cultural assignment is possible. Collins made the following assignments: there is 1 Clovis feature, 1 unassigned feature, 145 Clovis artifacts, 57 Folsom artifacts, 99 unassigned artifacts, and 22,933 pieces of unassigned debitage (roughly 16,000 debitage pieces are definitely cultural—the others are probably natural fragments). There are also numerous large stones scattered throughout the component that are too large to have been moved by floodwaters without having also moved lots of artifacts. Therefore, the majority of the large stones are considered manuports, meaning they were carried there by humans and perhaps used as anvil stones or weights to hold down the edges of hide coverings.

Field investigators believed that the two Paleoindian cultural features—concentrations of chipped stone—represented distinct knapping events (occasions when one or more prehistoric flint-workers sat down at the spot to produce or rework tools). The first, Feature P3, consists of small, non-diagnostic flakes from more than one piece of raw material. In the absence of refits, there is no basis for attributing this cluster of flakes to a single knapping event. The other, Feature P4, consists of a core and various debitage pieces shown by refitting to connect with a Clovis blade core and other debitage found outside of the feature.

Pavo Real As A Lithic Workshop

Most of the behavioral evidence preserved and recovered from the early occupational levels at Pavo Real relates to the knapping of stone. This evidence is dominated by tool-making debris produced by breaking up pieces of the local variety of Edwards chert. While it seems likely that the quality of the local chert was somewhat better (less weathered) 13,000 years ago than it is now, it was still plagued with many flaws and a variety of textures. Consequences of the flaws are evident throughout the knapping debris. In the general region, better grades of Edwards chert were readily available at the time to the people who camped at Pavo Real. Was Pavo Real occupied because, or in spite of, the abundant but less-than-ideal chert cropping out at the site?

The bulk of the Paleoindian assemblage from Pavo Real is made up of non-diagnostic lithic debitage (generic flakes, chips, and chunks of chert). Much of the evidence stems from earlier stages of reduction in this material and attests to similarities in the knapping process before the distinctive pieces that characterize Folsom or Clovis take form. There are also numerous simple tools such as unifaces (shaped on only one face) and some bifaces (shaped on both faces) that cannot be assigned to either Clovis or Folsom.

One curious set of Paleoindian artifacts consists of four miniature lanceolate points that cannot be assigned to Folsom or Clovis. Similar small-scale projectile points have been found at other sites in Paleoindian contexts of Clovis, Folsom, and Hell Gap cultures. The four miniature pieces in this collection may include points made by knappers of Clovis or Folsom affiliation, or both.

Clovis Lithic Technology

The majority (72%) of the 202 Paleoindian lithic artifacts at Pavo Real that can be assigned to Clovis or Folsom on the basis of morphology (shape) are of Clovis affiliation. These include fluted projectile points, bifaces, bifacial debitage, blade cores, blades, tools on blades, and blade production debitage. The accompanying photographs provide examples of most of these. The vast majority of the Clovis artifacts are made on the local material.

All of the artifacts inferred to represent Clovis biface technology (as opposed to blade technology) relate to the production, use, maintenance, and discard of fluted projectile points. There are two projectile points that appear to have been discarded at the end of their usefulness. Such pieces are typically found near sources of raw material in association with evidence for point manufacture, exactly the circumstances seen at Pavo Real.

Nine fragmentary Clovis bifaces found at Pavo Real are consistent with the idea that the site was a workshop where weapons were retipped. These fragments appear to be Clovis point preforms (unfinished points) and a fragmentary channel flake (specialized flake that creates the flute). The bifaces are typical of unsuccessful attempts to produce Clovis points.

There is substantial evidence for blade production at Pavo Real. Both of the blade core forms, conical and wedge-shaped, previously recognized in Clovis assemblages are present as well as various kinds of blades, blade fragments, and blade-core preparation flakes. The Clovis assemblage from Pavo Real also includes blade tools such as end scrapers (blades with one end that has been shaped into a circular tool edge). Most of the end scrapers have a single notch on one edge, presumably to facilitate hafting; they also have evidence of having been resharpened.

Dale Hudler examined each of the blades tools from Pavo Real under a microscope and found evidence of use wear (such as edge damage and polish) on many of them. The wear patterns he observed are considered characteristic of contact with both plant and animal materials, with the latter the more common. Such evidence shows that more went on at the site than just tool making. The Clovis people who camped here probably hunted nearby, butchered animals, gathered plant foods, made tools or clothes out of leather and wood, and so on.

Folsom Lithic Technology

Characteristic Folsom lithic artifacts present at Pavo Real include fragmentary Folsom points, aborted Folsom point preforms, channel flakes, thin retouched flakes, ultrathin biface fragments, spurred end scrapers on flakes, large thin unifaces, and multiple gravers. It is possible that some of these could actually be of Clovis origins.

The Folsom artifacts at Pavo Real are exclusively, or very nearly exclusively, made of the local chert. Most of these pieces indicate that the Folsom knappers had problems similar to those encountered by Clovis knappers. There are multiple examples of knapping failures caused by flaws in the raw material. Evidence for early stages in the Folsom reduction technology were not identified in this assemblage. This is almost certainly because they are not distinctive enough to be recognized as such.

Two fragments of Folsom points were presumably removed from their hafts and discarded at the site. Also present is evidence of Folsom point manufacture. This is similar to the Clovis assemblage at the site and is also inferred to be the result of retooling (removing broken or worn out tools from the wooden foreshafts of weapons and replacing these with new ones). Folsom preforms exhibit a number of failures, some resulting from knapping error and some from insurmountable flaws in the raw material. The Folsom channel flakes recovered from Pavo Real seem to have been successfully detached, suggesting that completed Folsom points were probably produced at the site from the local chert.

Bifaces other than preforms include two fragmentary ultrathin bifaces and two thin flakes with bifacial pressure flaking. Ultrathin bifaces, as the name suggests, are large and extraordinarily thin bifaces thought to be cutting tools (knives). Use-wear evidence confirming this function was not found on the Pavo Real tools, perhaps because the evidence for that use was removed by resharpening.

Numerous end scrapers on flakes from Pavo Real exhibit the classic attributes of Folsom end scrapers. Thin retouched flakes and flakes with multiple graver tips are also considered to be of Folsom affiliation. The retouched flakes may be cutting tools while the items called gravers are too delicate to be used in the engraving of any durable material such as wood or bone unless extremely lightweight incisions were produced.

Pavo Real as a Paleoindian Site

At a general level of comparison, the Clovis assemblage compares favorably with Clovis camp sites (as opposed to kill sites and other specialized site types). In Texas, at least five Clovis camp sites are indicated in the ecotone along the Balcones Escarpment (Gault, Wilson-Leonard, Vara Daniel, Spring Lake, and Kincaid—see map in "Site and Its Investigations"). The emerging pattern seems to be that of generalized hunter gatherers who would find an ecotonal setting such as Pavo Real ideal for the variety of resources that could be accessed with relatively little travel. This idea contradicts the prevailing model of Clovis peoples as highly mobile big game hunters.

Pavo Real is consistent with considerable data that now exists on Clovis and on Folsom site distributional patterns. Clovis site distributions and Clovis subsistence data across North America reflect generalized hunting and gathering lifeways, not big-game hunting specialization. To some of us engaged in the search for the origins of human occupancy of North America, it seems highly improbable that this pattern represents the adaptation of the founding populations of the continent (as expected according to the Clovis-First theory). Clovis culture is simply too well adapted to diverse resources in too many kinds of environments to be recent arrivals to those environments.

In its consistency with Clovis site distributional data, Pavo Real reinforces this emerging pattern. The keystone in the interpretive model of rapid Clovis expansion into an empty continent is that specialized big-game hunting is transferable to any habitat where big game are present, but when the evidence is preserved, most Clovis subsistence is based on small animals and probably also included plants. Clovis knappers were intimately familiar with tool stone sources all over the continent, including some that were relatively obscure. Their familiarity with the resources of the continent looks more like the end product of many generations of exploration and learning.

Folsom, on the other hand, seems to be the archeological manifestation of a specialized big-game-hunting way of life. This in itself poses a challenge to archeology since ethnographic analogs are lacking (no comparable human societies survive). Pavo Real as a Folsom site seems consistent with the distributional pattern of other Folsom sites and, in that sense, adds to what is known about Folsom subsistence strategies. As scholars struggle to better understand a truly nomadic big-game-hunting adaptation, this distributional pattern will be pivotal.

Were the Clovis and Folsom components better separated at Pavo Real, undoubtedly more could be said about the structure of their camps and the nature of the activities that transpired there. As it is, Pavo Real provides additional evidence of the regional-scale land-use behaviors of Clovis and of Folsom peoples. Pavo Real is probably also a strong indication that Clovis and Folsom folk camped at numerous localities in the Balcones Escarpment ecotone. Given the long history and current pace of urban and suburban growth in this part of Texas, to say nothing of 13,000 years of erosion, many of these campsites undoubtedly have been lost and such losses will continue.

There are deep alluvial deposits in the valleys of larger streams where they emerge from the dissected edge of the Edwards Plateau and change to a lower gradient along the margin of the Gulf Coastal Plain. These are precisely the settings where early sites have a better chance of survival because they are deeply buried. Sites formed in such settings, where rapid rates of deposition would favor better component isolation, would also have better conditions for preservation of such materials as bone.

Concerted geoarcheological reconnaissance along the alluviated valleys of such rivers as the Brazos, Little, San Gabriel, Colorado, San Marcos, Guadalupe, Medina, Sabinal, Frio, and Nueces as well as lesser streams along the Balcones Escarpment ecotone should identify areas where latest Pleistocene to earliest Holocene age deposits exist. Additional Clovis and Folsom sites will likely be found well-preserved and stratigraphically well-isolated in these settings. Such sites have the potential to expand and refine our understanding of Clovis and Folsom adaptations as expressed in the Balcones Escarpment ecotone and glimpsed at Pavo Real.