University of Texas at Austin wordmarkCollege of Liberal Arts wordmark
Texas Beyond History
TBH Home
Coastal Main
Prehistoric Texas Main

McFaddin Beach: This Site Is All Washed Up!

Photograph courtesy of Matthew White.

McFaddin Beach is a 32-kilometer long stretch of sandy beach in Jefferson County on the upper Texas Gulf coast, extending from High Island on the west to Sea Rim State Park on the east. Sabine Pass and the Louisiana state line lie another 19 km farther to the east. Although this public beachfront parcel has been assigned a trinomial archeological site designation (41 JF 50), it is not really an archeological site, but rather a place where artifacts and animal bones have been washing ashore for many years. Archeologists use the terms redeposited or secondary context to refer to accumulations like this.

The Texas coastline is a dynamic environment where sand can either be added or subtracted by currents and storm surges, and McFaddin Beach is a place where net erosion of sand has been taking place, flanked on either side by areas of net accretion. Despite this, it is clear that the artifacts are washing ashore from the Gulf, not eroding out from deposits behind the beach. A few of the artifacts have marine organisms (barnacles or bryozoa) attached. Backhoe testing (in 1983 and 2004) on both sides of Highway 87 has shown that the Pleistocene Beaumont Formation lies less than 2 meters below the ground surface, overlain by sterile Holocene sand and clay representing swamp and marsh deposits. Radiocarbon and OSL (optically stimulated luminescence) dates on the Beaumont/Prairie Formation range from roughly 28,000 to about 135,000 calendar years B.P. or more.

Although artifacts from essentially the entire known span of Texas prehistory are represented, the most striking aspect of this place are the extraordinary numbers of Paleoindian and Early Archaic  projectile points that have been discovered here. McFaddin Beach alone has produced more Clovis points than any single county in the state of Texas — over a hundred of them. Its nearest competitor, Gaines County in west Texas, has less than a third as many known Clovis points. Even the Gault site in central Texas has produced less than half as many Clovis points, despite years of intensive excavation. As a result, McFaddin Beach shows up as a conspicuous hot spot on both state and national distribution maps of Paleoindian points.

This peculiar concentration of material inevitably leads to several questions:

  • does the high density of projectile points correspond to an actual concentration of people (for example, a tight cluster of major base camps), a Paleoindian population density unparalleled elsewhere in the state?
  • or are coastal processes simply concentrating artifacts from a much larger area of the seafloor, giving a misleading impression of true artifact density?
  • or are the large numbers of artifacts simply a result of consistently intense scrutiny by beachcombers over the past several decades?
  • what was the attraction for early human populations in this particular part of the coastal plain?
  • and are the fossil bones related in some way to the Paleoindian occupations, or are they much older?
  • To address these questions, we must understand something about how the site has come to light, the geologic history of the Gulf coast and its salt domes, and the postglacial history of falling and rising sea level.

    History of Investigation

    Some of the earliest collections from the beach were made by local residents Odis and Beverly Beckham, especially after Hurricane Carla in 1961. Dr. Russell J. Long, a biology professor at Lamar University, contacted local collectors, studied their collections, and published the first summary of the site in 1977, a short report issued by the Spindletop Museum. Two years earlier, Jeffrey Russell, a student of Long’s, completed an inventory and study of the fossil vertebrates in the beach collections as a Master’s thesis at Lamar University; it remains the major source of information on the fossils. In the early 1980s, the establishment of the Strategic Petroleum Reserve facility at the Big Hill salt dome (inland from McFaddin Beach) led to construction of a brine disposal pipeline across the marsh, the beach, and out into the Gulf. Survey and backhoe trenching (1982-83) by Coastal Environments, Inc., in advance of this pipeline provided the first formal investigation of the geology immediately inland from the beach, and showed that the artifacts were not coming from this area. In 1983, avocational archeologist Paul Tanner of Port Arthur began keeping detailed locational records of artifacts found on the beach, and over time became the chief field researcher for the site. At about the same time, the Minerals Management Service in the US Department of the Interior became concerned about the possible impact of petroleum exploration and recovery on submerged archeological sites on the continental shelf, and commissioned some studies of the seafloor geology. These did not include the McFaddin Beach area, but they greatly add to our knowledge of the undersea geology. A 1986 study by Charles Pearson and others from Coastal Environments reconstructs the submergence history of the drowned Sabine River valley south of McFaddin Beach.

    In 1991, a conference on the site was organized in Port Arthur by Dee Ann Story, Paul Tanner, and Ellen Sue Turner, and 27 local collections were brought for examination, photography, and preliminary documentation. These included both artifacts and vertebrate fossils. In 1992, the Bureau of Economic Geology (UT Austin) sank a series of geologic cores along the coast, including three on or just offshore from McFaddin Beach. In 1994-96, Melanie Stright, an archeologist with the Minerals Management Service (US Dept. of the Interior), began a detailed study of five of the best-documented collections, the ones with recorded dates and locations for the finds. Her study included 880 artifacts, or about a third of the total known from the site at the time. She examined sources of toolstone, degree of wear, resharpening, the spatial position of the artifact along the beach, ages of the artifacts based on typology, and functional classes of artifacts. Completed as a Ph.D. dissertation at American University, her study was published in 1999 by the Minerals Management Service and remains the chief source of information on the site. In 2004, David Driver of Moore Archeological Consulting carried out a survey of a proposed wetlands area at McFaddin National Wildlife Refuge and excavated five backhoe trenches (1.0-1.7 m deep) on the inland side of the highway, finding no cultural material. Since the 1991 conference, Paul Tanner has continued monitoring the site and has recorded, among other things, several more Clovis points from the beach.

    Geologic and Sea Level History


    In This Exhibit:

    Map of McFaddin Beach.
    Map of McFaddin Beach. Source: Stright, Lear, and Bennett (1999:Fig. 1).
    photo of a chipped-stone projectile point
    A chipped-stone projectile point lies amid broken shell and other beach detritus washed onshore. Photo by David Crain.
    photo of Backhoe Trench 1,
    Backhoe Trench 1, dug in August, 1983, was one of several backhoe trenches dug through the beach and marsh sediments before pipeline construction. No archeological deposits were found.
    photo of McFaddin Beach
    McFaddin Beach. Photo by Thomas R. Hester, 1991.
    illustration of Geologic cross-section of Clam Lake immediately inland from McFaddin Beach
    Geologic cross-section of Clam Lake immediately inland from McFaddin Beach, reconstructed by coring through the marsh. Adapted from Morton et al. (1999:Fig. 3).
    photo of Sue Turner and Russell Long at the McFaddin Beach Conference
    Archeologist Ellen Sue Turner and Biology Professor Russell Long at the McFaddin Beach Conference. TARL Archives.
    illustration of the approaching Gulf shoreline in relation to McFaddin Beach, at 15,000 B.P.
    The approaching Gulf shoreline in relation to McFaddin Beach, at 15,000 B.P., before earliest Clovis occupations. (Images adapted from illustrations by K. Milliken from Anderson 2007:Fig. 2.16). Click to enlarge.
    illustration of the Late Paleoindian shoreline at 10,000 BP
    The Late Paleoindian shoreline at 10,000 B.P. as the flooding began to fill the Sabine River valley distant from McFaddin Beach. See shoreline at later time periods: 8000 B.P. and 5500 B.P.

    It seems clear that artifacts and fossils are arriving on the beach from a submerged, offshore source area, perhaps at no great distance or depth in the Gulf. It is also clear that both the present beach area and the offshore source were actually high and dry parts of the inland coastal zone until relatively recently in geologic history. This was an interfluvial area between the Trinity River valley to the southwest and the Sabine River valley to the northeast. During the Pleistocene, or last ice age, so much water was sequestered in major continental ice sheets (like the Laurentide ice sheet in North America) and in alpine ice sheets rimming the earth’s mountain ranges, that global sea level was drastically lowered.

    Some of the latest and most interesting research on sea level history comes from  Barbados, on the western edge of the North Atlantic, and it is based on isotopically dated samples of four species of drowned corals, one of which is known to grow at a restricted depth range. Uranium-thorium dating is used to fix the ages of coral reefs that were killed by postglacial sea level rise. These new studies suggest that maximum glaciation (and consequently, minimum sea levels) occurred about 26,000 calendar years ago, with sea level about 125 meters lower than at present. This is about 5000 years earlier and five meters deeper than the consensus values that most geologists have customarily used (21,000 calendar years, or 18,000 radiocarbon years for the Last Glacial Maximum, and a 120 m lowering). At 26,000 B.P., the Gulf shoreline in the upper Texas coast region would have been perhaps 200 km to the southeast of its present position. It should also be noted that this is just the latest lowering of Gulf sea level during the Pleistocene — there were earlier highstands and lowstands.

    By 14,000 years ago, sea level was rising rapidly because of ice sheet melting. Assessing the rate of rise in the Gulf is complicated by the fact that eustatic sea level indicators in the Gulf of Mexico tend to plot higher than contemporaneous indicators elsewhere in the world (for discussion, see the reference by Simms and others in “Sources”). The oldest radiocarbon dated shell or peat samples cored from the northern Gulf are about 19,700 calendar years old. If we rely on the Barbados coral data, by about 13,300-13,500 calendar years B.P. (when the archeological record at McFaddin Beach begins with the deposition of Clovis points at the source area) sea level still stood about 65-68 m below the present level, the shoreline was still 175 km away, and the site area still well inland.

    At about 10,000 calendar years B.P., when various types of Late Paleoindian projectile points were being deposited at the site, sea level had risen to 20 meters below the present level. At about 7700 cal B.P., a barrier shoreline developed about 55 km southeast of the McFaddin area. The rate of sea level rise (and flooding of the continental shelf) was especially rapid in the early Holocene from about 9000 to 7000 cal B.P. Without knowing where the actual source area for the McFaddin artifacts is located, it is impossible to say exactly when the area was flooded and rendered inaccessible for habitation, but there are some artifacts from the beach that Stright classifies as “transitional Archaic” (about 2350-1250 cal B.P.) and Late Prehistoric (about 1250-400 cal B.P.). By the time these were discarded, the source area may well have been flooded, which raises the possibility that these most recent artifacts might have been discarded on the beach, rather than washing ashore like the earlier material.

    During the period of lowered sea level, the combined channels of the Sabine River and Calcasieu River incised the exposed continental shelf, running southwestward roughly parallel to the present coast and about 30 km out from it, joining the Trinity River channel, then turning southward. The Deweyville terrace system that flanks these rivers continues onto the continental shelf, running under the waters of the Gulf. As sea level rose, the river valley was flooded, with the contact between fresh and sea water turned into an estuary. Marine coring in the 1980s by archeologists penetrated a Rangia shell deposit and a bone concentration with burned and unburned bone (water snake, amphibian, fish), fish scales, seeds, and nutshell. These may be bayside archeological sites occupied as the Gulf waters flooded what had been an estuary. A radiocarbon assay of 8055 ± 90 radiocarbon years B.P. on some of the shell equates to about 8500 cal B.P. (marine database, using standard marine reservoir age). Estuarine deposits dating from 7400 to 7400 cal B.P. (buried under Sabine Bank) and 8400 to 7700 cal B.P. (buried under Heald Bank) have been cored offshore. At 7700 cal B.P., the shoreline was still some 50 km away; by 5300 B.P., it had approached to about 40 km away; and by 2800 cal B.P., was about 5 km away from McFaddin Beach. By 1500 cal B.P. (or about 450 A.D.), the shoreline was approximately in its present position. Two lines of vibracores punched across the valley axis contain pine, oak, juniper, cypress, grass, sedge, and other  pollen types suggesting vegetation similar to that found on the coastal plain today.

    What happened after the middle Holocene, or about 6000 cal B.P., is in dispute.  One group of geologists insists there has been a fluctuating series of late Holocene highstands — periods during which sea level actually stood higher than today and covered part of what is now dry (well, dry to marshy, perhaps) land, evidenced by a series of inland beach ridges radiocarbon dated to that period. Another group of geologists insists that the beach ridges are really storm surge deposits, and document Late Holocene climatic intervals with increased storminess. They maintain that sea level has risen more or less smoothly to its present level and has never onlapped the land during the Holocene.

    When glaciation initially lowered sea level, the exposed continental shelf would have been contaminated with salt, but rainfall and runoff would have flushed the salt back into the Gulf long before human populations arrived. This broad, exposed plain (at least 175 km wide) might have been covered with grassland interfingered with gallery forests developed on the Deweyville terrace systems lining the river valleys. The poorly dated vertebrate fossil faunas from the region include various grazing animals like mammoth, bison, horse, and pronghorn. Lowered sea level also led to greatly increased groundwater flow from the continental interior to the shelf by removing hydrostatic pressure. According to one estimate, in glacial periods every 50 km of coastline received an extra 40 cubic kilometers of fresh water annually due to this effect (see the reference by Faure and others). Clusters of occupation sites may have formed around springs on the exposed shelf.

    The McFaddin Beach Salt Dome, located about 1.6 km offshore and buried under about 400 meters of Pleistocene and Holocene sediments, may be relevant to the geologic history of the site. Although the salt itself is Jurassic in age, many of the salt diapirs in the Gulf have been active during the Holocene. If the McFaddin Beach dome has experienced active upward movement during the Holocene, it may have resulted in some uplift of the seafloor sediments, possibly exposing archeological deposits to erosion by storm surges. Joints and faults may also be associated with diapir growth as well, and salt domes are often associated with saline springs. If there were saline springs associated in some way with this deeply buried salt during the period of lowered sea level, the springs might have been attractive as a salt lick for a variety of vertebrate species.

    Vertebrate Fossils

    The vertebrate fauna found on the beach is similar to other poorly-dated Rancholabrean faunas of southeast Texas and adjacent Louisiana (the Rancholabrean is the most recent of the Pleistocene land mammal ages, beginning at least 300,000 years ago), such as the Damon Mound, Texas City Dike, Avery Island, and Sims Bayou faunas. A fragment of proboscidean tusk from McFaddin Beach was assayed at 11,100±750 radiocarbon years before present (the 13C value is -24.3 ‰) by the Gulf Oil Corporation, but the assay was done over 30 years ago, and it is not clear if the material dated was the collagen fraction, or the collagen plus apatite. This assay equates to about 12,000-13,850 calendar years ago. Many of the fossils are mineralized or stained by organic or mineral deposits.

    Although this assay is squarely in the middle of the Clovis time span, it would be a mistake to assume all the McFaddin fauna is of Clovis age. It would not be at all surprising if a much better dating program were to reveal that the beach fossils vary widely in age, with many of them much older than Clovis material, and a few much more recent. The list includes several extinct species: sabertooth cat, cave bear, ground sloth, pampathere (giant armadillo), mammoth, mastodon, peccary, llama, tapir, and extinct horse. Some of these (Holmesina, Smilodon, Tremarctos, Neochoerus, Eremotherium, Mylohyus) are suspected to have become extinct well before the Clovis time span. Two mammoth teeth with possible human modification, some abraded or chipped long bone fragments from unknown species, and a bone projectile point have been found on the beach.

    The McFaddin Beach fauna is clearly not an unbiased vertebrate sample, because the only small mammals present are a cotton rat and a prairie dog. Fossil bones from small animals like rodents have probably been overlooked by beachcombers. But it is interesting to note that aquatic animals make up a significant part of the list: gar, catfish, sunfish, various turtles, alligator, otter, beaver, raccoon, capybara, and tapir all live in or near freshwater habitats. According to Paul Tanner, fossil turtle shell, horse bones, and deer antler are most common. Man of the terrestrial species (prairie dog, cotton rat, cottontail, bison, horse, peccary, pampathere) were either grazers or rooters, living in open grassy or brushy habitats rather than woodland. A few others  come from wooded (mastodon, ground sloth) or woodland gap (tapir) habitats, and these may be from gallery woodlands that lined streams crossing the exposed continental shelf.

    There are perhaps two different models to explain the origin of the beach fauna. Damon Mound, a Brazoria County salt dome, furnishes an example. At Damon Mound the sediments overlying the salt diapir are thought to be point bar and levee deposits of Beaumont age. The fauna is very similar to the McFaddin fauna, and the bones probably belong to animals that lived and died in the ancestral Brazos or San Bernard River valleys, either in the immediate vicinity or having been redeposited downstream by flooding. Upward diapir growth during the Pleistocene and Holocene has pushed the overlying alluvial sediments upward into a prominent mound, exposing them to erosion and quarrying activities. If this model applies also to the McFaddin dome (which is buried under the seafloor, unlike Damon Mound), it could indicate that most of the McFaddin bones predate and are unrelated to human occupation.

    The second model is more speculative because so little is known about the McFaddin dome. If joints or faults allowed saline springs to develop some 400 meters above the McFaddin dome while the continental shelf was exposed, they might have been attractive to a variety of herbivores. Many herbivores resort to salt licks (both wet and dry) and resort to geophagy (earth eating) to obtain supplemental elements (such as sodium, magnesium, iodine, and carbonates) during some seasons. This often happens during the transition from low-quality winter forage to rapidly greening spring vegetation, and it is especially important to females who have to meet the demands of lactation, growth, and weight regain during the spring season. Forage, especially early spring forage, may contain toxins (these are plant defense strategies to ward off grazers) or high levels of acidity. Clay consumption may help to buffer rumen acidity and neutralize secondary plant compounds such as tannins and alkaloids. In Kenya, licks frequented by African elephants contain measurably higher levels of sodium and iodine in the clay that is eaten.

    Some researchers have also suggested that licks serve a social function, as a “meetup place” for gregarious herbivores. In British Columbia, as Ayotte and others have noted, “wet licks are associated with groundwater springs, often becoming treeless areas of deep mud after years of use by moose and elk.” The resemblance to some of the mammoth kill sites at cienegas in the western US or the Quaternary bone deposits at Saltville, Virginia,  is striking, and it could well be that salt licks on the coastal plain attracted not only mammoths, bison, and horses, but also Paleoindian hunters, who probably would have been thoroughly familiar with the congregation times and gender composition of herds. A Master’s thesis study by Laura Abraczinskas of the spatial association between salines and proboscidean finds in Michigan was inconclusive, in part because of incomplete documentation of some of the sites. All of this is simply speculative, however, until we know more about the geology of McFaddin Dome.

    Artifacts from the Beach

    A more detailed view of the approaching shoreline from the early Holocene to the present.
    A more detailed view of the approaching shoreline from the early Holocene to the present. Positions are labeled with calendar (not radiocarbon) age. Note that the 2800 and 1500 cal B.P. shorelines are inland of their present positions because the Sabine delta had not yet built out. Sabine and Heald banks formed during the Holocene over the filled and submerged Trinity/Sabine/Calcasieu valleys. Adapted from Rodriguez et al. (2004:Fig. 15).
    It seems clear that artifacts and fossils are arriving on the beach from a submerged, offshore source area, perhaps at no great distance or depth in the Gulf. It is also clear that both the present beach area and the offshore source were actually high and dry parts of the inland coastal zone until relatively recently in geologic history.
    Radiocarbon assays of submerged peat and shell beds in the northern Gulf
    Radiocarbon assays of submerged peat and shell beds in the northern Gulf extend far enough back in time to allow reconstruction of sea level for Late Paleoindian to modern occupation periods. For Clovis and Folsom occupations, we must rely on reconstructions like those based on uranium-thorium dated corals from Barbados (not shown here). The vertical axis shows depth below modern sea level; the horizontal axis shows calendar (not radiocarbon) years before 1950 AD. The red curve is a best-fit line for rising sea level based on radiocarbon assays (adapted from Milliken, Anderson and Rodriguez (2008:Fig. 4). The left-hand intercept of the curve is below the zero point because dated materials include basal peats and swash-zone shells that actually reside slightly below sea level. Graphic by Ken Brown.
    photo of Archeologist Mike Collins examining a collection brought to the McFaddin Beach conference with Dennis Stanford and Russell Long
    Archeologist Michael Collins examines a collection brought to the McFaddin Beach conference with Dennis Stanford and Russell Long. Photo by Ellen Sue Turner.
    photo of the leg bone of extinct camel found on McFaddin Beach
    Fossil tooth washed up on McFaddin Beach. TARL Archives.
    Leg bone of extinct bison found on McFaddin Beach. 2004 copyrighted photo by T.W. Proctor, Curator, Proctor Museum of Science.
    pdf chart of Taxa identified from McFaddin Beach and other southeastern sites
    Taxa identified from McFaddin Beach and other southeastern sites. Chart by Ken Brown.
    This partition of time-diagnostic artifacts is based on Melanie Strightís inventory of five collections from McFaddin Beach, but uses a somewhat different scheme for period assignment. Artifacts that could not be assigned to a specific period are omitted. Here, as in Strightís own classification, Paleoindian artifacts are most abundant.
    photo of bone awl,
    Bone awl, Murray Brown collection. Photo by Thomas R. Hester, 1991.
    Obverse and reverse views of Clovis point from Morris Weeks collection
    Obverse and reverse views of Clovis point from Morris Weeks collection. Photo by Thomas R. Hester, 1991.
    Obverse and reverse views of large Clovis perform
    Obverse and reverse views of large Clovis perform; Patrick Frank/Sandy Massina collection. Photo by Ken Brown, 1991.
    Obverse and reverse views of Clovis point
    Obverse and reverse views of Clovis point (CN 5, Edwards chert) from Joe Coen collection, discovered March, 1991. Photo by Ken Brown, 1991.
    Artifacts from the beach northeast of High Island, from the Jean Lane collection. Photo by Ken Brown, 1991.

    Although Melanie Stright studied only the five best-documented collections, compared to the 27 that were examined at the 1991 conference, her research remains the chief source of information on the artifact assemblage. When looking at time-diagnostic artifacts (mostly projectile points, but also including such things as Clovis blades and beveled knives), the most striking aspect of the assemblage is the large proportion of Paleoindian artifacts — probably about 45% of the time-diagnostic items. Late Archaic artifacts (defined here as after about 2800 cal B.P. and before the Late Prehistoric) probably account for almost 32%. Early and Middle Archaic artifacts are much less well represented, although it must be admitted that very little is known about this span of time in East and Southeast Texas. Stright’s study sample included 53 San Patrice, 36 Scottsbluff, 27 Dalton, 21 Clovis, 13 Plainview, 13, Pelican, 3 Hell Gap, 2 Folsom, 2 Angostura, and several other Paleoindian types. Are Stright’s five collections representative of the entire beach assemblage? We can probably assume they are, although we should remember that there are now five times as many Clovis points documented as in her study sample. Presumably the proportional representation of different time periods would remain about the same if we could inventory every artifact found at the beach.

    For most sites in Texas with long occupational histories, it is usually the case that post-Paleoindian material vastly outnumbers the Paleoindian collections. The usual explanations are that early populations were highly mobile and thinly scattered, not burdened with much material culture, and that early components have had a much longer time to suffer the vicissitudes of site destruction. A collection where Paleoindian material dominates is highly unusual. It is also clear that in Stright’s study sample, the earliest Paleoindian types (Clovis, Folsom, Midland) are less numerous than somewhat later Paleoindian types like Scottsbluff, Angostura, and Early Side-Notched, and much of the Paleoindian material consists of eastern woodland artifact types such as Dalton, San Patrice, and Pelican. According to archeologist Dennis Stanford, many of the Clovis points resemble eastern rather than western manufacturing styles.

    Even more so than the rest of the Texas coast, McFaddin Beach lies in a region where knappable stone of any size is very scarce. The nearest bedrock sources are about 160 km away. Aside from these, petrified wood, fossil palm wood, and Citronelle gravels in Louisiana are some of the nearest sources. The nearest rivers, the Sabine and Neches, do not cross the Edwards Plateau and carry no Edwards chert  in terrace gravels. In 1995, geoarcheologist and lithic sourcing expert Larry Banks examined the five collections analyzed by Stright, to document the original source areas of the rocks used to make the artifacts (764 artifacts were examined). Using his extensive personal reference collection of rock types, Banks discovered some of the most interesting data about the site yet to emerge. Banks found that Edwards chert and quartzitic chert accounted for almost 72% of the total; about 65% of  this lacked visual evidence of gravel origins. The nearest Edwards bedrock outcrops are about 274 km away, while Edwards gravels occurs in terraces flanking the major rivers from the Brazos southward. Artifacts made from gravels could be identified by the presence of cobble cortex or circular weathering bands, but Banks suggests many artifacts could be made from gravel cherts without any remaining visual clues to their origin. The next most abundant sources were petrified wood or fossil palm wood, and Tecovas jasper from the Texas panhandle. Curiously, although Tecovas jasper is well-represented, Alibates silicified dolomite seems to be absent. Ouachita Mountain sources represent 16 items, Fisher Quartzite 9, the Arbuckle Mountains 6, and the Ozark Mountains 5.

    The biggest surprise, however, is the wide variety and immense geographic area from which the rock types are drawn. Banks identified 59 different named sources, plus unidentified gravels, Ogalalla quartzite (widespread over the entire southern Plains), petrified wood,  and palm wood. There are Plainview points of Minnelusa/Casper/Madera chert (Wyoming), Weeping Water Creek chert (Nebraska), and possibly from Pinetop (Ouachita Mountains) or St. Joe (Ozark Mountains) sources; a Dalton point of Weeping Water Creek chert; a Clovis point and a drill made on a small Scottsbluff point, of Knife River chert; and a scraper possibly made of Chuska (New Mexico) chert. The most distant sources are almost entirely represented by Paleoindian artifacts. Found by Jean Lane, but not included in Banks’s study is an unidentified fragmentary projectile point made of obsidian. Thomas R. Hester submitted the point to the Lawrence Berkeley Laboratory for analysis using Precise X-Ray Fluorescence, and the source was identified as Zacualtipan, Hidalgo, over 1000 km to the south

    Also unexpected is the fact that many of the projectile point types with eastern affiliations are made of Edwards chert, either from far-distant outcrops or from moderately distant gravel sources (terrace gravels of the Guadalupe, Colorado, or even Colorado and Brazos rivers, perhaps). Many, perhaps most, of the San Patrice and Pelican points, and other types such as Big Sandy, Dalton, Motley, Epps, Gary, and Delhi-like are made of Edwards chert. It is clear that much of the lithic industry, not only in Paleoindian but also in later times, was supported by raw materials drawn from far away and from a diverse geographic supply area. Is this typical of sites in the region, or is it unique to McFaddin Beach? We should remember that comprehensive sourcing studies like this one by Larry Banks are quite rare. In most cases where lithic sources are pursued, studies focus on a single exotic artifact, or at least a small sample. Studies of very large samples like the five McFaddin Beach collections are rare, especially in stone-poor regions of east Texas and the eastern US. If we had more such studies, we might find that toolstone movement across the landscape was much more extensive than expected in post-Paleoindian times.

    Spatial Analysis of Artifacts

    Obverse and reverse views of Folsom point
    Obverse and reverse views of Folsom point (Keokuk variety of Boone Chert, Oklahoma) from Pitcher collection. Photo by Ken Brown, 1991.
    photo of lanceolate and other Paleoindian dart points
    Examples of lanceolate and other Paleoindian dart points from the Tanner Collection at TARL.
    photo of an unidentified fragmentary projectile point made of obsidian
    Unidentified fragmentary projectile point made of obsidian from Zacualtipan, Hidalgo, found by Jean Lane and sourced by Thomas R. Hester. The very heavy weathering rind hides the glassy interior and probably makes it look extensively beach-rolled.
    photo of a variety of projectile points found by Jesse Fremont
    A variety of projectile points found by Jesse Fremont. Photo by Ken Brown, 1991.
    photo of two large trimmed blades
    Two large trimmed blades collected by Victor Garcia. Photo by Ken Brown.
    One of avocational archeologist Paul Tannerís maps showing plotted locations of artifacts he has collected from McFaddin Beach over some 25 years.

    As a major focus of her dissertation, Melanie Stright studied the spatial distribution of artifacts on the beach to look for clues to the possible location and nature of sites offshore. Using plottings supplied by the five collectors who contributed data, she compiled a composite map of artifact locations. By measuring GPS coordinates on landmarks used by the collectors, she was able to compute UTM (Universal Transverse Mercator) coordinates for each artifact. Importing the data into a GIS (Geographic Information Systems) program allowed her to look for clustering of artifacts by age, date of collection, artifact weight, and other variables. Clustering of artifacts could indicate 1) environmental processes that redistribute artifacts systematically, 2) remnant patterning that hints at locations of sites offshore, or 3) systematic collector bias.

    When all the artifacts in the beach were plotted together, a major cluster of artifacts along the eastern half of the beach and a less prominent cluster along the western half appeared, with an area (“the gap” at the center where only seven artifacts were found. Is this area less productive because the beach is a bit steeper here, or because this area is less accessible to collectors?

    Artifacts are believed to come ashore as a result of seafloor erosion associated with major storm surges. Stright found that 11 major tropical storms occurred between 1970 and 1989, plus another in 1995, the years that her study collections had been acquired. During the spring and summer months, from April to September, a strong longshore current sweeps southwestward along the coast, a current that would tend to pick up any artifacts brought into shallow water by storm surges, and redistribute them toward the Bolivar Peninsula. During the winter months, from October to March, a weaker current sweeps in the opposite direction and would probably have much less effect. In support of this pattern, Stright noted that artifacts from the western cluster tended to be more heavily worn. Anyone who has been swimming or surfcasting at places like Galveston or Padre Island along the Texas coast knows how strong this longshore current can be. Swimmers and fishing weights usually come ashore some distance from where they started out.

    Human disturbance might be another possible source of artifact movement. Paul Tanner noted an unusually large number of artifacts along the western portion of the beach after the winter of 1983, after construction of the Big Hill brine disposal pipeline. While examining artifacts at the 1991 conference, archeologist Mike Collins noted that most of the older artifacts (particularly Clovis points) showed little evidence of wear, while many of the younger, Archaic artifacts showed abrasion from rolling in the surf and chemical pitting from exposure to seawater. He suspected that this might be evidence of scouring of stratified sites, with the younger artifacts having been exhumed several years ago, and with and older and more deeply buried Paleoindian artifacts just now beginning to be exposed by scouring. Paul Tanner has also made the same observation.

    In addition to studying visual plots of artifacts on maps of the beach, Stright also did a nearest-neighbor analysis of artifact locations. Nearest neighbor is a statistical test used in geography, wildlife biology, and botany to look for clustering of discrete objects. The distance between each pair of objects is computed and compared to a sample consisting of the same number of randomly located objects. The distribution can be clustered, random, or regular. Stright found that the data set as a whole was clustered, and likewise the Paleoindian artifacts were clustered — perhaps just representing the two major clusters at each end of the beach — but all other subgroups (such as Late Archaic or Early Archaic) had “regular” distributions. These results are perhaps more a result of the nearly one-dimensional sampling area and the smaller numbers of the post-Paleoindian artifacts than anything else. The beach is, after all, 32 kilometers long but only 25 meters wide.

    Lessons Learned

    McFaddin Beach is a textbook case for the importance of provenience and context in archeology. If we could actually see the areas of seafloor where the artifacts and fossils are coming from, we would know if any of the vertebrate fossils are associated with Clovis artifacts. Are any of these the remains of Clovis-aged megafaunal kills, or are they much older fossils scoured from some other location and brought by chance to the same stretch of beachfront? Are the Paleoindian artifacts coming from one or a few large campsites with prolonged occupations (something like the Gault site in central Texas) or from a much larger number of smaller, more widely scattered sites? Are the sites stratified, or are the Paleoindian and Archaic artifacts arriving from entirely different places on the submerged landscape? Why is there so little chipping debris in the collections? Is it because flakes, blades and flake fragments are too hard to see on the beach, or less interesting to collectors, or is it simply because toolstone was scarce and most reduction was done closer to the rock source? Why is there so little evidence of Clovis blades? Is the McFaddin Beach salt dome in any way relevant to the archeology or the artifact distribution? We would probably have answers to all these questions and more if we could actually see the submerged sites themselves, but that will have to wait for better underwater exploration technology. For now, all the really important questions remain unanswered because of the lack of provenience and context. We should also point out that because many of the beach collectors allowed their collections to be studied by archeologists, and some of them even kept map records of their finds, we nevertheless know a lot more than we would otherwise.

    Credits and Sources

    The McFaddin Beach exhibit was written by archeologist Kenneth (Ken) M. Brown. Susan Dial and Heather Smith developed the exhibit.

    Brown’s interest in archeology can be traced back to his teen years participating in Texas Archeological Society field schools, and he has been at it ever since. Awarded a  Ph.D. from the University of Texas at Austin, he has worked in the field for nearly 40 years. Ken's dissertation on the Berger Bluff site provides unique insight into the Late Pleistocene and Early Holocene climatic history of the coastal plains of Texas. A Research Fellow at the Texas Archeological Research Laboratory, Brown has pursued interests in paleoenvironmental analysis and the study of wooden artifacts. For this exhibit, he drew on the sources listed below as well as information gathered during visits to the McFaddin Beach site and examination of various artifact collections.

    Print Sources

    Gulf Coast Geology

    Anderson, John B.
    2007   The Formation and Future of the Upper Texas Coast: A Geologist Answers Questions About Sand, Storms, and Living by the Sea. College Station, Texas A&M University Press.

    Faure, Hugues, Robert C. Walter and Douglas R. Grant
    2002   The Coastal Oasis: Ice Age Springs on Emerged Continental Shelves. Global and Planetary Change 33(1):47-56.

    Simms, Alexander R., Niranjan Aryal, Yusuke Yokoyama, Hiroyuki Matsuzaki, and Regina Dewitt
    2009   Insights on a Proposed Mid-Holocene Highstand along the Northwestern Gulf of Mexico from the Evolution of Small Coastal Ponds. Journal of Sedimentary Research 79: 757-772.

    Salt Licks, Salt Domes, and Megafauna

    Abracczinskas, Laura M.
    1992   The Distribution of Pleistocene Proboscidean Sites in Michigan: An Update of Records and a Co-Occurrence Analysis of Their Relation to Surface Saline Water. Unpublished MS thesis (Zoology), Michigan State University.

    Autin, Whitney J., Richard P. McCulloh and A. Todd Davison
    1986   Quaternary Geology of Avery Island, Louisiana. Transactions, Gulf Coast Association of Geological Societies 36:379-390.

    Ayotte, Jeremy B., Katherine L. Parker, and Michael P. Gillingham
    2008   Use of Natural Licks by Four Species of Ungulates in Northern British Columbia. Journal of Mammalogy 89(4):1041-1050.

    Flores, Mary D.
    2022   Identification of Pleistocene Fauna from McFaddin Beach. Unpublished MS thesis (Biological Sciences), Sam Houston State University.

    Gagliano, Sherwood M.
    1967  Occupation Sequence at Avery Island. Baton Rouge, Louisiana State University Press, Coastal Studies Series No. 22.

    Hamlin, H. Scott
    2006   Salt Domes in the Gulf Coast Aquifer. Chapter 12 in Robert E. Mace, Sarah C. Davidson, Edward S. Angle and William F. Mullican, III (eds.), Aquifers of the Gulf Coast of Texas. Texas Water Development Board, Report 365.

    Jenkins, John T. Jr.
    1978   Geology and Paleontology of the Upper Clastic Interval at Damon Mound. Pages 45-61 in Evelyn M. Etter (ed.), Damon Mound Field Trip Guidebook. Houston Geological Scoeity.

    Jones, Robert L. and Harold C. Hanson
    1985  Mineral Licks, Geophagy, and Biogeochemistry of North American Ungulates. Ames, Iowa State University Press.

    Njiiri Mwangi, Peter, Antoni Milewski and Geoffrey M. Wahungu
    2004   Chemical Composition of Mineral Licks Used by Elephants in Aberdares National Park, Kenya. Pachyderm 37:59-67.

    Saunders, Jeffrey J.
    1977   Late Pleistocene Vertebrates of the Western Ozark Highland, Missouri. Illinois State Museum, Reports of Investigations 33.

    Geologic and Sea Level History

    Balsillie, James H. and Joseph F. Donoghue
    2004   High Resolution Sea-Level History for the Gulf of Mexico Since the Last Glacial Maximum. Tallahassee, Florida Geological Survey, Report of Investigations 103.

    Cohen, Denis, Mark Person, Peng Wang, Carl W. Gable, Deborah Hutchinson, Andee Marksamer, Brandon Dugan, Henk Kooi, Koos Groen, Daniel Lizarralde, Robert L. Evans, Frederick D. Day-Lewis, and John W. Lane, Jr.
    2010   Origin and Extent of Fresh Paleowaters on the Atlantic Continental Shelf, USA. Ground Water 48 (1): 143-158.

    Gagliano, Sherwood M., Charles E. Pearson, Richard A. Weinstein, Diane E. Wiseman, and Christopher M. McClendon
    1982   Sedimentary Studies of Prehistoric Archaeological Sites. Criteria for the Identification of Submerged Archaeological Sites of the Northern Gulf of Mexico Continental Shelf. Preservation Planning Series. Baton Rouge, Coastal Environments, Inc.

    Milliken, K. T., John B. Anderson, and Antonio B. Rodriguez
    2008   A New Composite Holocene Sea-Level Curve for the Northern Gulf of Mexico. Pages 1-11 in John B. Anderson and Antonio B. Rodriguez (eds.), Response of Upper Gulf Coast Estuaries to Holocene Climate Change and Sea-Level Rise. Special Paper 443, The Geological Society of America.

    Morton, Robert A., Jack L. Kindinger, James G. Flocks, and Laura B. Stewart
    1999   Climatic-Eustatic Control of Holocene Nearshore Parasequence Development, Southeastern Texas Coast. Gulf Coast Association of Geological Societies Transactions 49:384-395.

    Peltier, W. R. and R. G. Fairbanks
    2006   Global Glacial Ice Volume and Last Glacial Maximum Duration From an Extended Barbados Sea Level Record. Quaternary Science Reviews 25(23-24):3322-3337.

    Simms, Alexander, Kurt Lambeck, Anthony Purcell, John B. Anderson and Antonio B. Rodriguez
    2007  Sea-Level History of the Gulf of Mexico Since the Last Glacial Maximum With Implications for the Melting History of the Laurentide Ice Sheet. Quaternary Science Reviews 26(7-8):920-940.

    Warny, Sophie, David M. Jarzen, Amanda Evans, Patrick Hesp and Philip Bart
    2012   Environmental Significance of Abundant and Diverse Hornwort Spores in a Potential Submerged Paleoindian Site in the Gulf of Mexico. Palynology 36(2):234-253.


    McFaddin Beach

    Arnold, J. Barto III and Thomas J. Oertling
    1995   Upper Texas Coast Underwater Archaeological Reconnaissance Project: Galveston, Chambers, and Jefferson Counties. The International Journal of Nautical Archaeology 24(3):199-204.

    Bever, Michael R. and David J. Meltzer
    2007   Exploring Variation in Paleoindian Life Ways: The Third Revised Edition of the Texas Clovis Fluted Point Survey. Bulletin of the Texas Archeological Society 78:65-99.

    Driver, David
    2004   Re: 30-Acre Wetland Creation Project, McFaddin National Wildlife Refuge Archeological Reconnaissance and Assessment, Jefferson County, Texas (MAC PN 04-85). Letter Report, Moore Archeological Consulting, Inc.

    Hester, T. R., F. Asaro, F. Stross and B. Giauque
    1992   On the Beach: Trace Element Analysis of an
    Obsidian Artifact from Site 41JF50  La Tierra 19(2):2-5.

    Hester, Thomas R., Michael B. Collins, Dee Ann Story, Ellen S. Turner, Paul Tanner, Kenneth M. Brown, Larry D. Banks, Dennis Stanford, and Russell J. Long
     1992   Paleoindian Archaeology at McFaddin Beach, Texas. Current Research in the Pleistocene 9: 20-22.

    Long, Russell J.
    1977   McFaddin Beach. The Patillo Higgins Series of Natural History and Anthropology No. 1, Spindletop Museum, Lamar University.

    1986   Two Clovis Points From McFaddin Beach, Texas. Ohio Archaeologist 26(1):9.

    Patterson, Leland W.
    2000   Comments on a Study of McFaddin Beach Artifacts. La Tierra 27(4):15-19.

    Pearson, Charles E., David B. Kelley, Richard A. Weinstein, and Sherwood M. Gagliano
    1986  Archaeological Investigations on the Outer Continental Shelf: A Study Within the Sabine River Valley, Offshore Louisiana and Texas. Baton Rouge, Coastal Environments, Inc.

    Pearson, Charles E., George J. Castille and Richard A. Weinstein
    1982  Cultural Resources Survey of the Big Hill Storage Complex Pipeline, Nederland to the Gulf of Mexico Jefferson County, Texas. Baton Rouge, Coastal Environments, Inc.

    Pearson, Charles E., and Richard A. Weinstein
    1983   Cultural Resources Addendum Big Hill Storage Complex Pipeline, Nederland to the Gulf of Mexico Jefferson County, Texas. Baton Rouge, Coastal Environments, Inc.

    Russell, Jeffrey D.
    1975   Identification of Pleistocene Fossils From McFaddin Beach, Texas. Unpublished MS thesis, Lamar University.

    Shelley, Steven D.
    1981   Cultural Resources Survey of the Proposed McMoran Exporation Company S/D Facility McFadden Marsh National Wildlife Refuge, Jefferson County, Texas. New World Research, Inc., Report of Investigations 45.

    Stright, Melanie J., Eileen M. Lear, and James F. Bennett
    1999   Spatial Data Analysis of Artifacts Redeposited by Coastal Erosion: A Case Study of McFaddin Beach, Texas. US Department of the Interior, Minerals Management Service (2 volumes).

    Tanner, Paul and Ellen S. Turner
    1993   Two Bone Artifacts From McFaddin Beach. La Tierra 20(3):17-18.

    Turner, Ellen S. and Paul Tanner
    1994   The McFaddin Beach Site on the Upper Texas Coast. Bulletin of the Texas Archeological Society 65:319-336.


    Matthew White Photography

    Ground stone plummet, Victor Garcia collection; found between Sea Rim State Park and the cattle pens. Photo by Ken Brown, 1991.
    photo of Early Archaic points
    San Patrice and Pelican points from the Tanner Collection at TARL.
    map of sources of rock used to manufacture artifacts during the Paleoindian period
    Sources of rock used to manufacture artifacts during the Paleoindian period, including such distant sources as Knife River Flint (30, North Dakota), Weeping Water Creek (59, Nebraska), Spanish Diggings (44, Wyoming), Minnelusa (35, Wyoming), and Chuska (5, New Mexico). Source: Stright, Lear, and Bennett (1999:Fig. 10).
    map showing plotted locations of Clovis artifacts
    This map showing plotted locations of Clovis artifacts illustrates the kind of locational data Melanie Stright was able to assemble for artifacts from different time periods, using the plotting data given her by the five collectors (Tanner, Louvier, Coen, Brown, and Fremont) who plotted their finds. Source: Stright, Lear, and Bennett (1999:Fig. 36).
    map showing plotted locations of Clovis artifacts
    Archeologist Dee Ann Story discusses theories with exhibit author Ken Brown. Photo by Ellen Sue Turner. TARL Archives.
    Cover of two-volume study of McFaddin Beach
    Avocational archeologist Paul Tanner, shown with his collections from McFaddin Beach donated to the Texas Archeological Research Laboratory. His 25-year study of the site, including the detailed documentation of dozens of other collections, has been recognized by the Society for American Archaeology with the 2009 Don Crabtree Award. .
    Cover of two-volume study of McFaddin Beach
    Cover of two-volume study of McFaddin Beach by Melanie Stright, Eileen M. Lear, and James F. Bennett .
    image of forms created by Paul Tanner
    One of hundreds of forms created by Paul Tanner to document McFaddin Beach artifacts held among area collectors. Each form contains a sketch of the artifact and its measurements. The work of Tanner and other avocational archeologists has brought together important data from widely dispersed collections. TARL Records.
    photo of the gathering at McFaddin Beach Conference, 1991
    Gathering at McFaddin Beach Conference, 1991. From left, Richard Weinstein, Michael B. Collins, Dee Ann Story, Paul Goldberg, Thomas R. Hester, and Dennis Stanford. Enlarge to see additional participants. Photo by Ellen Sue Turner.