Water Canyon

Project name: Water Canyon
Site type: Prehistoric hunting site
Period: Late Paleoindian
Project director: Robert Dello-Russo


Paleoindian Research at the Water Canyon Site (LA 134764)

The occurrence of Paleoindian remains in New Mexico is a relatively rare thing. In the ARMS database we have just over 1,200 Paleoindian sites and isolated projectile points, of which the majority may be points only. These 1,200 sites and points constitute less than 1 percent of all documented sites in the state. Of that very small portion, there are probably fewer than 20 Paleoindian sites that have been professionally excavated to any degree. More interesting still is the fact that good bone preservation at open, excavated Paleoindian sites is restricted to those sites along the eastern flank of the state, on the edge of the southern High Plains. These include the type site for the Clovis Culture, Black Water Draw; the type site for the Folsom Culture, the Folsom site; Milnesand; Ted Williams; Elida; San Jon; and others. No sites in the remainder of the state, except Sandia Cave and perhaps Pendejo Cave (both sheltered sites), have good bone preservation, although some (such as Ake, Boca Negra, and Mockingbird Gap) have produced a few small bone fragments or pieces of tooth enamel. The discovery of the Water Canyon site in Socorro County may be, then, not only the first opportunity in west-central New Mexico to investigate an intact, stratified Paleoindian site with well-preserved faunal remains, but also one directly associated with a robust record of paleoclimatic and paleoenvironmental change.

I first documented the Water Canyon site in 2001 as principal investigator for Escondida Research Group during an inventory in advance of a proposed construction project (Dello-Russo 2001). This undertaking, by the Energetic Materials Research and Testing Center (EMRTC), a division of New Mexico Tech in Socorro, was never realized, but the archaeological survey allowed for the discovery of ten sites with only Middle Archaic, Early Archaic, and/or Late Paleoindian components. The fact that the Water Canyon site is on the 9,000 acre EMRTC facility is beneficial, since the locked perimeter fence ensures the preservation of the site. However, it also requires that we have escorts and badges when working at the site.


Site Description

The Water Canyon site is in a portion of a structural basin between the Magdalena and Socorro Mountains. From a geomorphological perspective, the site rests at a point where a large bajada fan flanking the east side of the Magdalena Mountains is truncated by small incised channels and other erosional effects of the Water Canyon drainage. One incised channel or arroyo (now associated with Locus 1) was found in 2001, just northwest of the original artifact scatter (now Locus 4). At 1 to 2 m below the surface, this arroyo exhibits a buried, very dark and extensive sediment lens. Dark deposit in arroyoAt the time of discovery, this was estimated to date to the Early Holocene geological epoch. Similar deposits in other areas of the American West have been referred to as “Black Mats” (Haynes 2008). This deposit represents a fossil wet meadow.


Initial Fieldwork (2008–2010)

In the spring of 2008, I led Vance Holliday and Bill Reitze (University of Arizona), Patrice Walker (Escondida Research Group), and Bruce Huckell (University of New Mexico) to the site to collect hand-auger sediment samples for radiocarbon dating. We also discovered a bone protruding from the buried wet meadow in the north cutbank of the arroyo. Subsequent 14C dates for three bulk sediment samples and one charcoal sample ranged from cal BP 8700 ± 160 (Early Archaic/Early Holocene) to cal BP 13,025 ± 175 (Clovis/Late Pleistocene), while Late Paleoindian–era dates bracketed the bone. These dates both supported and exceeded our previous age estimates for the wet meadow deposit and, together with the bone, gave us our first provocative glimpse of the site’s potential.

In the spring of 2009, the fall of 2009, and the spring of 2010, we continued work at Water Canyon to refine our understanding of the site. Table 1 lists personnel who worked at the site during one or more of these three sessions. Our research is decidedly interdisciplinary and aims at evaluating the nature and extent of the wet meadow deposit, determining the character of both cultural and faunal assemblages, assessing site chronology, understanding the site’s soil formation history, discovering whether the site was occupied at multiple times during the Paleoindian period, and utilizing the proxy data archive in the wet meadow deposit to assist in the reconstruction of the regional paleoclimate and paleoenvironment. This latter research goal is important because (1) few significant such reconstructions currently exist in this part of the Southwest for the Late Pleistocene–Early Holocene transition, and (2) such reconstructions may facilitate not only our view of Paleoindian adaptations but also our understanding of environmental responses to present-day climatic shifts.

We have, to date, developed a contour map of the entire site and mapped and field-analyzed all surface artifacts flagged in 2009 utilizing a total station. We now see the site comprising four surface artifact concentrations: Loci 1, 2, 3, and 4, encompassing about 8.9 hectares (22 acres). Diagnostic surface artifacts include a Clovis point base, a Late Paleoindian point base (Scottsbluff? Belen?), and a Late Paleoindian point midsection.Clovis point base Late Paleoindian point baseLate Paleoindian point midsectionWe have completed 21 hand-excavated 1 by 1 m test units (12 in Locus 1 and 9 in Locus 3), 6 mechanically excavated backhoe trenches (Locus 1, Locus 2, Locus 3), and 49 mechanical soil cores utilizing a Giddings soil coring rig. These excavations have enabled us to collect assemblages of subsurface lithics and faunal materials, a suite of 26 pollen samples from hand excavations and soil cores, a suite of 19 bulk sediment samples for flotation processing and macrobotanical analysis, 19 radiocarbon samples (including bulk sediment, charcoal, and bison bone), two samples of red ocher that may have cultural associations, two OSL (optically stimulated luminescence) pilot samples; and five stratigraphic profile illustrations from trenches and five from hand-excavated units. The sediments in all mechanical cores have been described by Vance Holliday.Giddings soil coring rig


Lithic Artifact Assemblage

In all, 100 lithic artifacts were recovered from the excavations during the spring 2010 field session. The majority of lithic artifacts are debitage, including sharpening and biface thinning flakes. Of interest was the recovery from Locus 1 of a rhyolite early-stage biface midsection exhibiting one large outrepassé flake scar, and the in-situ recovery of a cobble chopper from Locus 3. The biface is spatially associated with bison bones from Locus 1. All the lithic artifacts recovered in 2010 were from Locus 1, Locus 3, and the west wall of Trench 4. These augment the 26 artifacts recovered in 2008 and 2009, and the 257 surface artifacts field-analyzed during 2009. Also associated with bison bones is an obsidian biface thinning flake sourced to the Mule Creek quarry, in southwest New Mexico (Shackley 2010) and silicified rhyolite flakes sourced to the Black Canyon quarry, just southeast of the Water Canyon site (Dello-Russo 2003). No diagnostic lithic artifacts were recovered from subsurface contexts from 2008-2010. Laboratory analysis of artifacts recovered thus far is on-going at OAS.


Faunal Assemblage

Eighty-three bones, bone fragments, and tooth enamel fragments recovered from the 2009 excavations and 54 bones, bone fragments, and enamel fragments recovered from the 2010 excavations were analyzed by Nancy Akins (OAS) and Gary Morgan (New Mexico Museum of Natural History). All of the hand-excavated bones recovered in 2010 came from Locus 1 on the south side of arroyo, while bones recovered in 2009 came primarily from the north side of the arroyo and Test Unit 1-5 on the south side. All these were recovered at elevations 1 to 2 m below the surface. Taxa in the faunal assemblages include Bison sp., large mammal (nonbison), Neotoma (wood rat), Salienta (toad), and Hawaii miniscula (a land snail typically found in floodplains and wet meadows; identified by Peter Kondrashov, Kirksville College, Missouri). Bison skeletal elements include distal femur, calcaneum, rib, teeth, tooth enamel (n = 15), sesamoids, cranium, tibia, carpal/tarsal, humerus, mandible, cervical vertebra, long bones, and flat bones.Bison distal femur The unfused epiphysis on the distal femur indicates that the animal was no more than five years old. Some of the bone fragments were calcined, suggesting they had been culturally burned, and this is consistent with the nearby presence of a possible thermal feature (see below). Three sets of bone fragments were identified in Cores 10-01, 10-02, and 10-03 by Russell Greaves (Harvard University) at depths of about 3.70 m below the surface. While no definitive bone architecture could be identified for these fragments, Greaves suggested in the field that, given the bone size and the cancellous tissue morphology, the bones were from at least a sheep- or goat-size animal, but most likely were from a deer or larger animal.


Radiocarbon and OSL Samples

In all, 19 14C samples were recovered by hand excavations (primarily Test Units 1-11 and 1-12) and mechanical cores (Table 2). Two dated charcoal samples were examined by Pamela McBride (OAS) and were determined to be Quercus (oak). The relatively high density of charcoal samples from Test Unit 1-11 suggests that they might represent the remains of a thermal feature. Chronometric analysis of these samples will test that possibility. Along with these charcoal samples, 18 additional bulk sediment samples are in the process of being dated.

Two sediment samples for OSL dating were taken from the west wall of backhoe Trench 3 during the fall 2009 field session. The final analyses of these samples have recently been completed, providing calendar dates of 9370 ± 630 cal BP and 9430 ± 670 cal BP, which are consistent with radiocarbon dates on the oak charcoal from nearby hand-excavated Test Unit 1-11 (9280 ± 160 cal BP) and on the bison bone (9155 ± 135 cal BP) (Table 2). The vast majority of samples dated thus far have a Late Paleoindian (Early Holocene) age. One sample (FS 23; 12,300 ± 350 cal BP) has a Folsom-era age, and another (FS 32; 13,025 ± 175 cal BP) has a Clovis-era age.


Pollen Sampling and Analysis

Susan Smith (Northern Arizona University) and Stephen Hall (Red Rock Geological Enterprises) retrieved two pollen samples from Locus 1 (Test Unit 1-12), two pollen control samples from the modern surface, and one pollen sample from Locus 3 (Test Unit 3-07). The latter sample was recovered from beneath the in-situ cobble chopper. They also retrieved 7 pollen samples from Core 10-26 and 14 pollen samples from Core 10-6, in parallel with the previously mentioned datable sediment samples collected from the same cores.

Analyses of all pollen samples are being undertaken by Susan Smith at the Bilby Research Center. Results of our initial pilot study indicate that pollen is indeed preserved in many of the samples and reveal that the Late Paleoindian–age deposits provide evidence of a mosaic plant community at the time, including a sagebrush steppe environment, a wet meadow, and a possible gallery forest of birch and maple. Subsequent pollen analysis supports the previous findings and indicates the additional presence of marsh elder and possibly cherry. The most surprising development is the discovery of spores (thought to be from filamentous algae) at the base of some of the mechanical cores. These spores, which are showing up at about 3.7 m below the surface—the same deep level as the unidentified bone—suggest the presence of an algal layer similar to the Black Mat seen at the Murray Springs Clovis site in Arizona (Haynes and Huckell 2007).


Seismic Testing and the Groundwater Flow Model

In the fall of 2010, Mark Person, geohydrologist from the New Mexico Institute of Mining and Technology (NM Tech); three students from the institute; and I laid out a line of seismic geophones to the south of Locus 1 in a west–northwest to east–southeast orientation. The seismic test, using a “Betsy gun,” was initiated to search the subsurface at the Water Canyon site for the presence of a water table. While this data was to be used in conjunction with a groundwater flow model being developed at NM Tech to augment our understanding of the prehistoric water regime over the Pleistocene–Holocene transition, the results have thus far proved inconclusive.


Summary and Discussion for 2008-2010

The field seasons at Water Canyon have been quite a success, given the number of excavated test units, backhoe trenches, and mechanical cores taken, the number of scientific samples recovered, and our discoveries to date. Our current understanding of the site has improved tremendously in terms of the site stratigraphy, the nature and orientation of the buried wet meadow deposit, and site chronology. Analytical results are beginning to provide us with more details about the young bison that is represented by the numerous skeletal elements excavated thus far. Because the deposit where the majority of the bones are located is bisected by the arroyo, the different average elevations for bones on the north side of the arroyo compared to those on the south side suggest (1) the wet meadow deposit slopes rather steeply from southwest down to the northeast; and/or (2) the bison bones on the south side of the arroyo come from a different individual than those on the north.Locus 1 in 2011

Sediment and soil descriptions from the mechanical cores and backhoe trenches, as well as stratigraphic profile illustrations taken from the hand-excavated test units and the backhoe trenches, are helping us refine our model of site formation processes. For example, a large depression in the south-to-north profile of Trench 5 suggests the presence of an ancient arroyo cut. This same possibility is echoed further to the east, where the three mechanical cores exhibit bone fragments at levels around 3.7 m below the surface. Since these bone samples are substantially deeper than the remainder of bones excavated in Locus 1, they may represent a different individual(s), and they might sit at the base of a deep and ancient arroyo channel. Expanding on that idea, it is perhaps possible that the bones comprise part of a bone bed. Future mechanical excavations will be required to evaluate that possibility. Additional OSL dating of overburden sediments will expand our understanding of the tempo of sediment erosion and deposition.

Paleoenvironmental proxy data recovered at Water Canyon include pollen, macrobotanical samples, and faunal materials (amphibians, snails). We can also gain insights into the ancient environment by analyzing the ratio of stable carbon and/or nitrogen isotopes found in radiocarbon dated materials, bison bone, bison tooth enamel, and occasionally in the shells of snails. The relationships between time and delta 13C values for radiocarbon-dated materials at Water Canyon is illustrated here. The trend in this figure indicates that the abundance of plants utilizing the C3 photosynthetic pathway (more efficient in wet and cool environments) decreases from the Late Pleistocene into the Early Holocene, while the relative frequency of C4 and CAM pathway plants (more efficient in brighter, warmer, and drier environments) increases.

In addition, we are working toward a more robust model of climatic change over the Pleistocene–Holocene transition and the responses of past vegetational communities to that change by the use of transformational computations between the results of on-site pollen analyses and the analysis of samples taken systematically across an elevational gradient (Rio Grande to top of Magdalena Mountains). The following is a simple example. While the Water Canyon site today (5,300 ft elevation and 8 in of annual precipitation) supports a juniper-savannah community, pollen from the Late Paleoindian component indicates that the site at that time supported a much more mesic environment, an analog of which exists today at ca. 8,600 ft elevation (with about 27 inches of annual precipitation). The drying trend reflected here parallels the same trend seen in the delta 13C values.



Continued interdisciplinary research at the Water Canyon site is clearly warranted by the results to date, and we are eagerly anticipating new and significant discoveries from on-going and future collaborative efforts.



Dello-Russo, R. D.
2003    Geochemical Comparisons of Silicified Rhyolites from Two Prehistoric Quarries and 11 Prehistoric Projectile Points, Socorro County, New Mexico. Geoarchaeology 19:237–264.

2001    A Cultural Resources Inventory of 472 Acres in Socorro County, New Mexico: The Archaeology of the EMRTC / GLINT Project Area. Report ERG-2001-3. Submitted to EMRTC/NMIMT by Escondida Research Group, Socorro.

Haynes, C. V., Jr.
2008    Younger Dryas “Black Mats” and the Rancholabrean Termination in North America. Proceedings of the National Academy of Sciences 105(18):6520–6525.

Haynes, C. V., Jr., and B. B. Huckell (editors)
2007    Murray Springs: A Clovis Site with Multiple Activity Areas in the San Pedro Valley, Arizona. University of Arizona Press, Tucson.

Shackley, M. S.
2010    Source Provenance of Obsidian Small Debitage from the Water Canyon Paleoindian Site (LA 134764), Socorro County, New Mexico. Geoarchaeological XRF Laboratory, University of California, Berkeley.



Funding for the radiocarbon date on the bison bone collagen was provided by a grant from the New Mexico Archaeological Council (NMAC). I appreciate their flexibility, patience, and generosity with this project.

Other research at the Water Canyon site has been generously underwritten by the following institutions and donors: Museum of New Mexico/Office of Archaeological Studies; University of Arizona Departments of Anthropology and Geology; Escondida Research Group, LLC; Argonaut Archaeological Research Fund; Curtiss T. and Mary G. Brennan Foundation; Energetic Materials Research and Testing Center; various donations from the Friends of Archaeology, Museum of New Mexico Foundation; University of California Berkeley XRF Laboratory; New Mexico Institute of Mining and Technology; New Mexico Museum of Natural History; Red Rock Geological Enterprises; VaprNet Anesthesiology Network; Lowes of Santa Fe; and a private donation in the name of George McJunkin.