Pavo Real

Pavo Real was first recognized and recorded in 1970 by two high school students, Bill Fawcett and Paul McGuff, long after the original two-lane highway 1604 had been built across Leon Creek. They had found the site after TxDOT cleared a new, much wider right-of-way in anticipation of expanding the road into an eight-lane thoroughfare. The two students were searching the area for archeological sites because they could see that subdivisions, roads, and shopping centers would soon obliterate most traces of the many prehistoric campsites present in northern Bexar County. They hoped the information they recorded would be of use, as indeed would prove to be the case. (Fawcett and McGuff both went on to study archeology in college and become professional archeologists.)

The original site record was brief and noted the presence of "fire pits" near the creek that had created "a rise above the rest of the site." On the surface it appeared similar to many other poorly known sites in the area—just another seemingly unremarkable place where prehistoric people had left behind evidence of camping and tool making. A few years later, Dr. Thomas R. Hester, then professor of anthropology at the new, nearby campus of the University of Texas at San Antonio, visited the site and recognized that the site might have deeply buried deposits dating back to at least the early part of the Archaic period. Hester alerted TxDOT archeologists to the site's potential and urged the agency to investigate the site prior to the construction of the highway.

In the late 1970s, federal and state laws and regulations governing cultural resources were in place, but the field of contract or cultural resource management (CRM) archeology was still young and the "standard operating procedures" were not yet standardized. Because some of the funding for the highway expansion came from the federal government, the impact of the construction on potentially important "historic properties" (such as archeological sites and old buildings) would have to be considered under the National Historic Preservation Act of 1966. Because site 41BX52 lay unalterably in the path of the planned highway, TxDOT archeologists began to evaluate the site's significance by carrying out test excavations.

Before discussing the investigations further, let's take a look at what archeologists call the "site setting" of Pavo Real.

Site Setting

The general location of the Pavo Real site figures prominently in the region's historical landscape as well in as San Antonio's modern transportation network. As part of it's unique natural setting, the site was located on or near important but little known historic and, very probably, prehistoric trails. The valley of Leon Creek, which heads (begins) only 22 kilometers upstream from the site, forms a natural corridor linking the Hill Country of the southern Edwards Plateau, the Blackland Prairie to the east and the Gulf Coastal Plain to the south. The site is located within the Balcones Fault zone just below the Balcones Escarpment, once known as the margin of Apachería and Lomería Grande. The edge of the Edwards Plateau is also known as the Balcones Canyonlands because of the many deeply entrenched streams and rivers that drain the plateau.

Historically, an old trail ran through the natural pass formed by the Leon Creek Valley and linked Bexar (modern San Antonio) and a Spanish Colonial presidio popularly known as San Sabá (Presidio San Luis de las Amarillas) in present-day Menard County. Early settlers called this pass, la puerta (de las casas) viejas (roughly translated as, "gateway to the old houses" or "old pass"). The nearby pass and the immediate vicinity of the Pavo Real site became a documented Comanche trail leading northward from the springs of San Pedro Creek in San Antonio. In the early history of the settlement, the historical trail also may have been called the camino de tehuacanas (an apparent reference to the Tawakonis, a Wichita group commonly associated with north-central Texas throughout the 18th century). By the mid-19th century, this ancient trail became a major route of German immigration from San Antonio into the Hill Country.

Today, the vicinity of the Pavo Real site is the intersection of two major regional thoroughfares. Interstate Highway 10 passes through the Leon Creek Valley following the same basic route as the historical trail and today, a few hundred years east of Pavo Real, carries an average of about 181,000 vehicles per day. Loop 1604 West runs along the Balcones Escarpment and today it carries some 162,000 vehicles daily over Leon Creek and the former location of the ancient campsite.

In addition to Pavo Real's location on a natural corridor, three other factors help explain why prehistoric peoples stopped to camp at Pavo Real: water, chert, and ecotone. Leon Creek itself would have had water it in year-round, even during droughts, at least in its deeper holes. Until the early 20th century, Leon Springs (10 kilometers upstream) and three other minor springs flowed freely (since the early 1900s, groundwater pumping has severely reduced or halted the flow of all springs in northern Bexar County). Today very little water flows in Leon Creek except at times of significant rainfall when it is subject to flash flooding. The finding of water snails in the site's Paleoindian deposits hint that there was a permanent pool of water at or near the site early in its history. The relatively well-watered valley would have provided habitat for many of the plants and animals upon which prehistoric peoples depended.

The site area also has an abundance of chert, the hard silica-rich rock often called "flint" that prehistoric peoples used to fashion projectile points and many tools. Chert seams are visible in the limestone bluff opposite the site, cobbles and fractured chert pieces are common in the gravel bed of Leon Creek, and, most importantly, chert was available on the site along a natural limestone bench that was exposed in Paleoindian times. The chert present along the bench ranged from pebbles to small boulders up to 40 centimeters (about 16 inches) across. As chert goes, the local material is not of a particularly high quality. It is highly weathered (turning it white and eventually soft and chalky), often fractured, and sometimes contains internal masses of a harder chert that makes flint knapping difficult. Most of it is white to gray in color. Despite its uneven quality, the Pavo Real chert was used intensively by the Paleoindian groups who camped at the site and less intensively by later Archaic peoples.

The third factor is that the site is located along a broad ecotone or zone of mixed plant and animal communities sharing characteristics of the Edwards Plateau (Balconian biotic province) to the north, the Blackland Prairie (Texan biotic province) and the Gulf Coastal Plain (Tamaulipan biotic province) to the south. North of the site, the plateau is rocky, hilly, and has mostly thin soils. The vegetation is drought-resistant oak-juniper savanna on uplands and slopes with hardwood forests in the better-watered canyons and valleys. Along perennial streams are found cypress, pecan, willow, and other trees dependent upon abundant water. Among the streams and rivers that form the Balcones Canyonlands, Leon Creek is rather small and intermittent. Not far east of the site is a narrow band of the Blackland Prairie, which was originally dominated by tall grasses. South of the site, is the expansive Gulf Coast Plain, which is covered mainly in drought-resistant thorny scrub brush and grasses except along streams.

In summary, the Pavo Real site was located on a natural transportation corridor in a location that provided prehistoric peoples with convenient access to the basic necessities of hunting and gathering life: water, tool-making materials (chert and wood), plants and animals. But don't get the wrong idea—Pavo Real was not a particularly idyllic or favorable spot, it was just in a general location that had lots of things going for it. There are (or were) dozens of other prehistoric sites located along this stretch of Leon Creek, some of them were probably more-favored spots, to judge by the relative quantities and densities of refuse. Many major prehistoric occupation sites occur along the Balcones Escarpment, especially at the major springs and along the larger streams and rivers. At times of drought, the Balcones Canyonlands may have served as critical resource areas because of the permanent water and wealth and diversity of other natural resources.

Investigations

When TxDOT archeologists began to investigate Pavo Real, they thought it was a typical Archaic campsite, as it indeed proved to be (in part). Archeologist Jerry Henderson was placed in charge of the fieldwork assisted by Glenn T. Goode and several other TxDOT archeologists. Initial backhoe testing in late May 1979 revealed the presence of several shallowly buried middens and deeper deposits suspected to date to the Early Archaic (about 7,000-9,000 years ago). As the TxDOT work progressed, the Archaic deposits at 41BX52 were found to include at least one well-developed burned rock midden (Feature 4) as well as several smaller middens, all located along the terrace edge nearest the creek. A hearth field of undetermined extent occurred in the vicinity of the annular (ring-shaped) midden, the only area of the site where the Archaic deposits were sampled reasonably well.

Following the backhoe trenching in May, 1979, the TxDOT archeologists began hand excavations at Pavo Real to examine the middens and dig below them in hopes of finding an Early Archaic component. [Most burned rock middens date to the Middle Archaic period (about 5,000-7,000 years ago) or later. The lifeways of earlier Archaic peoples was very poorly known in the late 1970s and was the subject of considerable interest.] For most of the summer of 1979, the archeologists, aided by local TxDOT workers opened up a series of small excavation units, mostly 2-x-2 meter squares, in different parts of the site. They concentrated most of their effort in and around the site's largest burned rock midden (designated Feature 4).

By mid-August time was running out; fieldwork was scheduled to be completed before the end of August. Thus far the site hadn't turned up much that was particularly interesting or different from previously excavated sites. It was a typical burned rock midden site dating mainly to the Middle and Late Archaic periods with some evidence of Early Archaic occupation. But preservation conditions were poor—almost nothing was found except burned rocks, dart points and other chipped stone tools, and lots of tool-making debris. In a few cases some charcoal was found in some of the hearths that might yield radiocarbon dates.

Friday August 17th would be the last day for the local workers, and the Austin-based archeologists expected to wrap things up the following week. It was apparent that many of the excavation units could not be completed within the remaining time. Then, unexpectedly, a Clovis point was found on August 15th within what had previously been considered an Early Archaic deposit (only 45 centimeters, or 18 inches, beneath the surface). This find was potentially important because Clovis sites were rare—was this an isolated find?

The archeologists were given a several-week extension to determine whether intact Paleoindian deposits were present. By late September they had found a Folsom point and other early artifacts and determined that an isolated Paleoindian component was present, sandwiched between two gravel lenses. A Gradall (a large precision-digging machine often used in highway construction) was then brought in to remove most of the remaining Archaic deposits north and east of Feature 4. Despite continued pressure to hurry up and finish the dig so that highway construction could proceed, the dig was extended by several months to investigate the Paleoindian materials.

The second half of the dig, from October, 1979, to January, 1980, concentrated exclusively on the site's Paleoindian component, which rested in a sandy layer between two gravel lenses. Henderson was still in charge of the work, but additional experienced professional archeologists were hired on to help. At the time, the archeologists thought the early materials dated mainly to the Folsom period and so they called the sandy layer the Folsom zone. To investigate the zone, they laid out a series of 2-x-2-meter squares that systematically covered most of the area (in contrast to the spotty approach taken with the Archaic dig). Instead of digging using mainly shovels, as was done with the Archaic dig, they dug mainly with trowels and carefully plotted the locations of many of the artifacts they found.

In addition to the archeological dig there was also geological work done to try and understand how the Paleoindian component formed. This mainly consisted of using heavy machines to dig deep, but narrow, trenches through various parts of the site followed by detailed geological recording. Unfortunately, the geological work was not coordinated with the archeological work, resulting in several missed research opportunities. For example, the deposits identified geologically as those most likely to show stratigraphic separation between Clovis and Folsom occupations were not targeted by the archeological excavations. Another missed opportunity was the chance to effectively sample a deeply buried deposit beneath the main Paleoindian deposit. Flint flakes (tool-making debris) were found in a deep geological test, but this was not followed up by excavation. These materials could date to earliest Clovis times or possibly even earlier.

One very far-sighted thing that did come from the geological work was the taking of four soil monoliths. Archeologist Grant Hall from the Center for Archaeological Research at nearby UT-San Antonio lent his expertise in this effort. Soil monoliths are columns of sediment that were carved out of the walls of some of the trenches, solidified with a plastic resin, wrapped in burlap, and then encased in wooden boxes built around the columns. The accompanying photographs show how this was done. Such samples provide "witness columns" of intact sediments for future studies. And, in fact, archeologists and soil scientists 20 years later were able to take samples from the archived Pavo Real soil monoliths and apply sophisticated new analytical techniques to study how the site formed and to date the site's deposits.

After the Pavo Real dig ended in early 1980, the materials were processed in TxDOT's archeological laboratory and Jerry Henderson began initial analysis. Unfortunately, other agency priorities intervened, personnel changed, and the project remained uncompleted for two decades.

In 2000, archeologists from the Texas Archeological Research Laboratory began to analyze the Pavo Real materials under contract to TxDOT. Noted Paleoindian expert and geoarcheologist Michael B. Collins directed the studies of the Paleoindian materials. Pavo Real's Archaic materials were studied by Steve Black, who had directed the excavation of another Archaic burned rock midden site in northern Bexar County, the Panther Springs Creek site, in 1979-1980 while the work at Pavo Real was underway. Dale Hudler managed the project and undertook a variety of special studies of projectile points and other materials. Expert consultants were called in to do other specialized studies, particularly those of the soil monoliths. The work was completed in 2003, culminating in a thorough scientific report and this web exhibit.