Botanical Preservation in Texas Rockshelters: Eagle Nest Canyon (northeastern Chihuahuan Desert) and McCutchen Branch (Lampasas Cut Plain)

By Leslie Bush, Kevin Hanselka, Christina Nielsen, Daniel Rodriguez, and Carol Macaulay-Jameson

**This is the final post detailing different analyses currently being conducted with Eagle Nest Canyon materials.**

Leslie standing next to her poster at TAS.

Leslie standing next to her poster at TAS.


Analysis of botanical samples from rockshelter sites in two different ecological areas of Texas highlight the exceptional value of such sites for paleoethnobotanical research but also the complexities of botanical preservation within and between shelters. In Eagle Nest Canyon, a tributary of the Rio Grande River in the northeastern Chihuahuan Desert, delicate and uncarbonized plant parts such as lechuguilla fibers and bristlegrass chaff are preserved along with the tough, carbonized plant parts that are typically found on open air sites in the area. Preservation does not follow a simple, improving gradient of preservation from the front to the back of the shelters, however. In the more humid climate of central Texas, ancient plant remains at the Barnhill #3 Site (41CV1646) are completely carbonized, or nearly so. Although uncarbonized plant parts are not preserved, the rock shelter provides conditions for the preservation of delicate plant parts such as grass stems and wind-dispersed seeds that are rare to absent on open air sites in the region.

Eagle Nest Canyon

Kelley Cave (41VV164)

Plan map of Kelley Cave showing 2013 excavations and features.

Plan map of Kelley Cave showing 2013 excavations.

Feature 1, identified at the modern surface, consisted of multiple ash lenses with reddened soil at the base. The lack of correlation between bone, debitage, and dung fragments in the Feature 1 level of the cave reflects human digging and rodent turbation (Rodriguez 2015:162). This feature contained less plant material per liter than Feature 6, which was also a burning context.

Feature 4, identified at the surface, dates to roughly 600 cal B.P. It included dense layers of fiber detritus beneath a layer of compacted mud. Sorting and identification of the rich assemblage of plant material in Feature 4 is ongoing, but many leaves and fibers of agave and similar plants and onion bulbs are present. Some carbonized plants are present, but these are mostly wood charcoal and are estimated at ten percent or less of all plant remains.

Feature 6 was encountered at 140 cmbs and dates to roughly 7400 cal. BP.  It is interpreted as a series of overlapping rock-lined pits, with the lower rocks heated in place. No rodent burrows were visible in feature exposures (Rodriguez 2015:126). Insect (termite) and rodent (mouse) feces were recovered in flotation in carbonized form, but the plant remains most likely to represent rodent use (grass and prickly pear seeds) are uncarbonized. It contains more plant material per liter, both carbonized and uncarbonized, than Feature 1.

Summary: The presence of plant remains in particular contexts within Kelley Cave is conditioned both by the types of activities represented (cooking/burning in Features 1 and 6 versus raw plant deposition in Feature 4) and taphonomic processes. Preservation is mostly through carbonization in Features 1 and 6, but a mud drape at the top of Feature 4 led to the preservation of more uncarbonized plant material there. The lower depth of Feature 6 contributed to better plant preservation than in Feature 1 because it afforded better protection from later events such as pit construction and modern looting. The deposition of the mud layer in Feature 4 demonstrates that unique events can produce good preservation of large quantities of uncarbonized plant material at shallow depths in specific locales.

Kelley Macrobot

Eagle Cave (41VV167)

Plan view of Eagle Cave showing the location of the 5 Profile Sections excavated in 2014.

Plan view of Eagle Cave showing the location of the 5 Profile Sections excavated in 2014.

PS4. Deposits in this area reflect mixed discard of refuse from cooking, plant processing and possibly other activities. Three samples, including one from feature context (Feature 2) yielded both carbonized and uncarbonized remains of wood, leaves, bulbs, and seeds. Density of several classes of plant remains (wood charcoal, carbonized leaves and bulbs, and uncarbonized seeds) was higher in the feature context than in other samples.

PS3, located at right angles to PS04, is an ash lens that contained carbonized wood, leaves, and bulbs. The sample comes from beneath a heating element, but it is not clear whether that heating event carbonized PS3 plant material. Uncarbonized plants were limited to three fragments of hackberry seeds, which are particularly durable and occur even in pre-Holocene geological deposits (Wang et al. 1997).

Summary: The presence of botanical material in Eagle Cave deposits is conditioned by the type of deposit and taphonomic processes. In PS4 the concentrated focus of human activity represented by Feature 2 produced a greater density of plant remains than other samples in PS4. In PS3, only carbonized plants were preserved, suggesting that uncarbonized plant material succumbed to taphonomic processes that were not operating in PS4 despite the proximity of the two units, possibly due to higher moisture coming from the shelter wall. Interestingly, PS3 is farther from the cave mouth than PS4, indicating that a simple gradient of increasing preservation from dripline to back wall does not apply.

Eagle Macrobot

Eagle Nest Canyon in Regional Context

Two open-air sites in Val Verde County preserve only carbonized plant remains and few small seeds (see chart below and compare to the materials, especially uncarbonized plant parts, in the two charts above). Eagle Nest Canyon samples have these plant remains in abundance and provide unique opportunities to access this part of prehistoric subsistence (see also A Curious Artifact Comes to Light).

Other LPC Sites


Yucca seeds from Eagle Cave Feature 2.

Yucca seeds from Eagle Cave Feature 2.

Cord-wrapped bundle as initially exposed.

Cord-wrapped bundle as initially exposed.

Barnhill #3 Rock Shelter (41CV1646)

Plan map of Barnhill #3 Rockshelter.

Plan map of Barnhill #3 Rockshelter.

Twenty-six flotation samples from ten features within the shelter and two off-site samples were examined. In this Central Texas shelter, only carbonized archaeological plant remains survived.

Earth ovens yielded wood charcoal, bulb fragments including camas and wild onion, nutshell (hickory, walnut, and acorn), and 123 small carbonized seeds. A corn kernel fragment was present in one earth oven sample.

Charcoal/ash deposits/middens contained wood charcoal, bulb fragments including wild onion, nutshell (hickory, walnut, pecan, and acorn), and 207 small seeds.

The number of samples analyzed from hearths was smaller, but they included wood charcoal, an unidentifiable bulb fragment, and 59 small, wild seeds.

Only a single pit sample was analyzed. It contained hickory and black walnut nutshell, wood charcoal, and 17 small seeds.

Barnhill Macro

Summary: Plant density at Barnhill #3 is conditioned by feature type, with earth ovens and charcoal/ash deposits/middens having higher botanical density than hearths or the pit feature. Although plant density is higher in the southeastern portion of the site, this seems to be due to the unusual density of charcoal in one sample from Feature 11 (F6-2013; 20.92 g/liter) rather than a true reflection of higher charcoal density in the southeastern area as a whole.

Sunflower (left); Seeds from Barnhill #3 Rock Shelter, scale in mm: vetch from Feature 25 (left-center) and maygrass from Feature 7 (right-center); Wild tobacco (right).

Sunflower (left); Seeds from Barnhill #3 Rock Shelter, scale in mm:
vetch from Feature 25 (left-center) and maygrass from Feature 7 (right-center); Wild tobacco (right).

Barnhill #3 Rock Shelter in Regional Context

Uncarbonized plants do not survive at Barnhill #3 as they do in the Chihuahuan Desert shelters, but the protection of the shelter allows for much better carbonized plant preservation than is usual in limestone regions (Braadbaart 2009). Bulb fragments and, especially, small seeds are present in much higher numbers than on other sites in Coryell County.

Barnhill Regional Context


The presence of plant remains in particular archaeological contexts is conditioned at multiple scales of analysis. Plant samples from the three shelters considered here show several variables in play:

  • regional climate (moisture, soil chemistry),
  • local geology (physical protection of site),
  • activity and intensity of ancient use (thermal events, feature v. non-feature context),
  • taphonomic processes (moisture channels through site deposits, insect and rodent burrowing, looting), and
  • unique events (mud drape on Feature 4).


Many thanks to Jack and Wilmuth Skiles and John Barnhill for their good stewardship and site access, and to the Texas Parks and Wildlife Department for use of unpublished data.

**A full PDF is available here: BushEtAl_TAS2015_Macrobotanical_FINAL**

*“Preservation” is used here as a catch-all for the sum total of the processes, natural and cultural, pre- and post- and depositional, that contribute to the presence of plant remains in a particular context.



Braadbaart, F., I. Poole, and A. A. van Brussel

2009    Preservation Potential of Charcoal in Alkaline Environments: An Experimental Approach and Implications for the Archaeological Record. Journal of Archaeological Science 36: 1672–1679.

Rodriguez, Daniel P.

2015    Patterns in the Use of the Rockshelters of Eagle Nest Canyon, Langtry, Texas. Unpublished M. A. thesis, Department of Anthropology, Texas State University, San Marcos, Texas.

Wang, Yang, A. Hope Jahren, and Ronald Amundson

1997    Potential for 14C Dating of Biogenic Carbonate in Hackberry (Celtis) Endocarps. Quaternary Research 47: 337–343.

References for Coryell County Bulb and Seed Densities

 Bush, Leslie L.

2010    Plant Remains from Site 41CV389, Fort Hood, Coryell County, Texas. Report submitted to SWCA Environmental Consultants, Austin, Texas, February 5, 2010.

2011    Plant Remains from Site 41CV286, Coryell County, Texas. Report submitted to Prewitt and Associates, Inc., Austin, Texas, February 5, 2010.

2015    Plant Remains from 2015 Excavations at Barnhill Rockshelter #3 (41CV1646), Coryell County, Texas. Revised report submitted to Department of Anthropology, Baylor University, Waco, Texas, September 28, 2015.

Thoms, Alston V., Douglas K. Boyd, and Karl Kibler

2015    Earth ovens, Geophytes, and Microfossils: Investigating Burned Rock Features and Archeobotanical Remains on Fort Hood, Central Texas. United States Army Fort Hood Archeological Resource Management Series Research Report No. 65. January 2015.

From the Bowels of the Lower Pecos

By Emily McCuistion

Howdy, you may have seen mention of me on this blog and on our facebook page but this is my first official post for the ASWT project. I introduced myself at the beginning of the season, but to refresh I am one of the three interns on the Eagle Nest Canyon 2015 Expedition. My previous archaeology work has been primarily with the National Park Service and Forest Service, neither of which generally have the resources to undertake studies such as this one. This is the first in-depth archaeological excavation I have worked on and I’ve been blown away by the site itself, the other expedition members and the leadership overseeing the project, and our numerous collaborators who contribute their specialized expertise on various and sundry sub-disciplines (chemistry, botany, entomology, geology…the list goes on). It seems that there are infinite directions that project-related studies could go. I love interdisciplinary research and so I’ve been especially taken with the subject on which I am blogging. These delicate resources contain a world of information inside them- I’m glad to introduce the coprolites of Eagle Cave!

Me, poofing sediments off a coprolite.

Me, poofing sediments off a coprolite. I’m wearing a dust mask to protect my airway from the fine particles we stir up.

The study of human coprolites, also known as paleofeces, palaeofaeces, or simply dried-up old poop, was not regarded as important by the archaeological community until the mid-20th century.  Now we realize that we have a lot to learn from coprolites; they can inform us on past diets and ecosystems, the health of an individual, and the manner in which one ate and pooped. Archaeologists can look at the plant and animal macrofossils (visible undigested remains), tiny pollen grains, phytoliths (micro-structures created by plants), viruses, parasites, and even bugs found in coprolites!  Further, coprolites can be directly radiocarbon dated (so we can determine how old they are) and they preserve not only the DNA of the person who made the deposit, but also the DNA of the animals and plants that were consumed. Though perhaps less revealing, shape, color and of course location of the deposit are relevant. The study of odor, though noted by many researchers, has as yet not proved of scientific value.


FN31041 from Eagle Cave. We give each coprolite a unique specimen ID or Field Number (FN). We’ve found that one specimen can be made up of several segments representing one bowel movement.

Victoria excavating a coprolite. After documentation and careful extraction we transport the coprolites to the lab in boxes so that they are not inadvertently crushed.

Victoria excavating a coprolite. After documentation and careful extraction we transport the coprolites to the lab in boxes or by hand so that they are not inadvertently crushed in our packs.

Though I hesitate to mention paleontology- the oft confused (with archaeology) but unrelated study of dinosaurs- the study of dinosaur coprolites pre-dates the study of human. We are indebted to paleontologists for first recognizing the potential in coprolite study. The word coprolite, derived from Greek, was termed in the 1830s by an English paleontologist and translates to “dung stone.” I have yet to hear about human coprolites that have turned to stone, nonetheless the use of the word coprolite to describe human feces in an archaeological context is pervasive.

The Lower Pecos Canyonlands is famous for its coprolite studies, specifically those conducted on specimens excavated from Hinds Cave. Coprolites have been identified at other sites in the region but none have been so well studied as those from Hinds.


Profile of latrine deposit in Hinds Cave. Coprolites are in the upper right corner. Photo from


Hinds Cave is north of Eagle Cave, along the Pecos River. This photo from was taken during the 1970s excavation, as evidenced by the excavation dust swirling from the rockshelter.

Even knowing that coprolites preserve well in the Lower Pecos we didn’t really expect to find them in Eagle Cave – coprolites were not found during the 2014 field season in any of the shelters investigated in Eagle Nest Canyon. Happily, during our excavations this spring we were excited to find an abundance of well preserved coprolites!  Thus far over 80 coprolite specimens have been inventoried during this field season, the vast majority of them excavated from near the front of the rockshelter and approximately a meter or two below the surface.

South Trench Profile of Eagle Cave with point-provenienced coprolites.

South Trench Profile of Eagle Cave with coprolites plotted- click on the image to enlarge.

At Eagle Cave the coprolites are mainly found in areas of discarded fire-cracked rock and plant remains (See: Between a Rock and a Heart Place for info on the fire cracked rock in Eagle). Sometimes they are found within compacted and dry-cracked sediment which we hypothesize may be evidence of urine-soaking, and of which we are taking samples. Sometimes a coprolite is found near a smooth river cobble; speculations about wiping methods ensue…

Matt "Call me John" Larsen was the first to start finding coprolites in the FCR/fiber layers near the front of the shelter.

Matt “Call me John” Larsen, pictured here saying “poop” in American Sign Language, was the first to start finding coprolites in the fiber layers near the front of the shelter.

Analysis of the coprolites is not yet underway but there are some basic observations that can be made about the Eagle Cave specimens. (We are assuming all the specimens we have collected are human; however, it should be noted that other animals have pooped in Eagle Cave over the preceding 9000 years and some of the specimens we have collected may not be human.)

Shape- It has been found that plump, shapely coprolites are often from a fiber/plant-heavy meal while the “loose” ones may be from an individual who had recently consumed a lot of meat. There are, however, many other reasons that one could have had fluid bowel movements: parasites, water containing algae-born toxins, plants with a laxative effect (sotol and lechuguilla – both baked in earth ovens – are known to have this effect on some people), even emotional states- fear has been cited- can cause diarrhea.


The “most-shapely” and largest of our specimens, and a good example of a segmented bowel movement: FN30984.

FN31296 fell a little flat.

FN31296 fell a little flat.

Size is affected by several conditions, including length of time since the last bowel movement and types of food eaten. Meals based on a single type of food can result in smaller stools, though a vegetarian diet can result in a larger stool than an omnivorous diet.

Diminutive FN31109

Diminutive FN31109. Compare it to FN30984 pictured above.

Color can be indicative of diet; the darker specimens often result from meat consumption while the lighter ones can be indicative of a vegetarian and/or carbohydrate-rich meal. When dried, however, colors change. Nonetheless, it is interesting to note color variations.

FN31048 is unusually orange in color.

FN31048 is unusually orange in color.

FN31103 is taupe.

FN31103 appears greyish, though it is important to note that our specimens still have sediment from excavation clinging to them.

FN31063 has a warm brown color, and lots of fibers.

FN31063 has a warm brown color, and lots of fibers.

Macrobotanicals, bones and fur (Macrofossils): Botanists and faunal experts can identify large plant and animal remains within coprolites. Thorough analysis requires breaking down the coprolite and separating it into constituent parts, however, macrofossils are sometimes visible on the surface. Many thanks to our faunal and botanical associates- Christopher Jurgens, Haley Rush, Kevin Hanselka, Leslie Bush, and Phil Dering- for helping to tentatively identify bones and seeds in a couple specimens (meet our zooarcheology and archeobotany heros here: “Come Visit:” A Zoooarchaeologist Is Lured Back to the Lower Pecos and here:The Archaeobotanical Team Forms).

This bone in FN30983 came from a jackrabbit-sized animal. In the words of Charles, "Ouch."

This bone in FN30983 came from a jackrabbit-sized animal. In the words of Charles, “Ouch.”

The diamond-shaped macrofossil in the center of this stool has been identified as a mesquite endocarp.

The diamond-shaped macrofossil in the center of this stool has been identified as a mesquite endocarp. The peak sticking off the backside are fibers from a plant such as lechuguilla or sotol.

FN30984 is chock full of seeds!

FN30984 is full of seeds!

FN30983 contains what looks like prickly pear seeds, the tastiest desert food I've ever eaten (unless you count Jack Johnson's mesquite snickerdoodles).

FN30983 contains what looks like prickly pear seeds, the tastiest desert food I’ve ever eaten (unless you count mesquite snickerdoodles).

Insects: We have yet to see insects in the Eagle Cave coprolites (though a fly pupal case rolled out of a bagged specimen when I removed it for this photo shoot) but we sometimes see their bore-holes. In her 1978 dissertation on Hinds Cave coprolites, Glenna Williams-Dean reported that the experimental stools (a comparative collection of modern specimens left to dry in a rockshelter near Hinds Cave) were up to 95% consumed by insects! When conditions are not suited to an insect, especially a flying insect, it quickly moves on. For this reason they are good indicators of change in climactic and other conditions. (see: Archaeoentomology? for more details)

Holey crap- I believe insects created this swiss cheese affect in FN31021.

Holey crap- I believe insects created this swiss cheese effect in FN31021.

One morning I resumed excavation of a coprolite I hadn't finished exposing the previous day. When dusting off the specimen this beetle crawled out from under it! Perhaps coprophagous (dung-eating) insects are interested in vintage as well as fresh specimens.

One morning I resumed excavation of a coprolite I hadn’t finished exposing the previous day. When dusting off the specimen this beetle crawled out from under it! Perhaps coprophagous (dung-eating) insects are interested in vintage as well as fresh specimens.

As mentioned previously, there are numerous other avenues for coprolite study (e.g., pollen, DNA/genetics, phytoliths, lipids, parasitism, general health, dietary nutrition, seasonality, medicine, food processing techniques, past environment, food trade/exchange, and food storage), but until we begin the analysis of the Eagle Cave specimens I cannot go into great detail.  However, I can provide a brief synopsis (definitely not exhaustive) of some things archaeologists have learned from conducting coprolite studies in the Lower Pecos.

Let us start at the beginning of the digestive system: the mouth. A significant malady affecting ancient Lower Pecos people was dental wear and eventual complete toothlessness by their 40s. Danielson and Reinhard (1998) studied coprolites to examine possible causes for the dental wear and tooth loss in Archaic-age Lower Pecos skeletons. They considered several hypotheses, including that the grit from food-processing stones was wearing down the tooth enamel.  Interestingly, they found no such grit in the coprolites they studied. Rather, calcium oxalate phytoliths (microscopic structures found in certain plants which pass unscathed through the digestive system) are found in abundance in Lower Pecos coprolites. They occur in the agave and cactus plant families (Agavaceae and Cactaceae respectively), and were found to be harder than tooth enamel. Not coincidentally, agave and cactus are a major part of the aboriginal diet of the Lower Pecos people. Other factors probably contributed to dental wear and loss, but a major cause was identified.

The phytoolith pictured up top is prickly pear. The lower is a sotol phytolith (

The phytolith pictured up top is prickly pear. The lower is a sotol phytolith (

So, we have a population of largely toothless people who still must eat. But, must they chew? Specimen FN30984 from Eagle Cave, and some found at other sites, indicate there was not always much chewing – what Dr. Black charmingly calls the “grab it and gulp” method of eating.

That brings us to what they were ingesting. Certainly food availability varies with the seasons and the millennia, but there is coprolite evidence people were eating many of the same foods that are available today (e.g., sotol, lechuguilla, cattail, wild onion, prickly pear, insects, fish, deer, pronghorn, rodents, reptiles, bison [the kitchen ran out of that a few years ago], and too many other foods and medicinal plants to go into at this time – check out the TBH exhibit on the Ethnobotany of the Lower Pecos Canyonlands for more information about Lower Pecos diet.  So far we have only been able to identify a few different seeds in the Eagle Cave coprolites but as we begin detailed analysis of the coprolites we will learn a great deal more.

Finally, when the meal is ready to be passed a decision has to be made on where to release the body’s effluent. I don’t know if there are enough sites in the region with mapped coprolite “deposits” to make assertions about latrine location preferences, but it was found that the latrine at Hinds Cave was at the back of the shelter. The vast majority of coprolites from Eagle Cave have been collected from near the front of the shelter at the top of the burned-rock-strewn talus slope.

Dr. Black takes a close look at the Eagle Cave coprolite collection.

Dr. Black takes a close look at the Eagle Cave coprolite collection.

For more discussion about coprolites from the Lower Pecos and Hinds Cave, check out the TBH exhibit on Hinds Cave to learn more.  So much more on coprolites could (and will) be written, but right now other doodies call!

References Cited:

Danielson, Dennis R. and Karl J. Reinhard                                                                                          1998 Human dental microwear caused by calcium oxalate phytoliths in prehistoric diet of the lower Pecos region, Texas. American Journal of Physical Anthropology 107: 297-304.

Williams-Dean, Glenna                                                                                                                         1978 Ethnobotany and cultural ecology of prehistoric man of southwest Texas. Ph.D. Dissertation, Department of Anthropology, Texas A&M University, College Station.

Cord-Wrapped Fiber Bundle: A Most Curious Artifact Comes to Light

By Kevin Hanselka

Cord-wrapped bundle as initially exposed.

Cord-wrapped bundle as initially exposed.

As discussed in the blog post of March 3, 2014, Leslie Bush and I recently joined Phil Dering as members of the Ancient Southwest Texas Project archaeobotany team. Our role is to study the massive amounts of well-preserved plant remains that are expected to be found in Eagle Cave, Kelley Cave, and Skiles Shelter.  And we have not been disappointed! While excavations and the analyses are ongoing, the bulk matrix samples we have looked at so far are overflowing with fascinating perishable materials reflecting the prehistoric inhabitants’ use of local plants for artifacts, fiber, and food.

In early May I returned to Eagle Nest Canyon to assist in the excavation of a remarkable plant processing feature in Kelley Cave.  When first encountered by Dan Rodriguez and Steve Black last summer, Feature 4 appeared to consist of a thick fiber layer capped by a uniform and seemingly purposeful layer of alluvial mud about 3-5 cm thick.  They hypothesized the mud layer was a prepared surface, presumably created by dumping basket loads of fine Rio Grande alluvium like that found in a flood deposit in adjacent Skiles Shelter.  However, sediment size analysis and comparison led geoarchaeologist Charles Frederick to favor an alternative idea: that the layer is a naturally deposited thin mud drape from the same impressive  flood that left an alluvial layer over one meter thick in the back of Skiles Shelter.  This fits with the fact that Kelley is some 5 meters higher in elevation than Skiles.  A tentative assessment of new radiocarbon dates from both sites suggests this extraordinarily massive flood may have occurred around AD 1350.

Kevin Hanselka pointing to cord-wrapped bundle when initially exposed.

Kevin Hanselka pointing to cord-wrapped bundle when initially exposed.


My interest in Feature 4 is in its layers of plant matter and the information they may contain about the cultural use of plant resources in the shelter. In other words, as Brooke Bonorden so eloquently stated in the March 3rd blog post, I want to assess these deposits for “… well-preserved evidence of full behavioral chains from plant collecting to baking to food consumption to fiber production to weaving and plaiting to finished object to final discard.” Most of the plant remains from Feature 4 probably represent the accumulation of waste products of food prepared in nearby earth ovens.  Mainly  trimmed-off outer leaf bases from baked agave lechuguilla stems or “hearts,” but also discarded agave quids (masticated inner leaf bases)  and other plant remains such as wild onion skins.   The food remains, however, may also be intermingled with manufacturing waste produced as leaves were stripped into fiber for weaving mats or baskets or for twisted twine.

Kevin using dental pick to gingerly free the artifact from the mud and fiber in which it was embedded.

Kevin using dental pick to gingerly free the artifact from the mud and fiber in which it was embedded.

Steve and Dan’s working hypothesis is that the Kelley Shelter occupants cooked plant foods such as lechuguilla, sotol, prickly pear, and wild onion in earth ovens near the mouth of the shelter, and then discarded the waste vegetal debris in dense layers, like that in the upper Feature 4.  A desirable side effect of this patterned behavior would be that plant layers would help suppress dust, making the shelter a more comfortable place to live and work.  We have been kicking around these ideas for several months and Steve had promised he would wait until I came back in early May to further explore Feature 4.

Hanselka takes photos of his handy work.

Hanselka takes photos of his handy work.

And so it was that Steve, fellow visiting archaeologist Drew Sitters, and I began to excavate a 50-x-100 cm unit placed just downslope (west) of the previously exposed mud-capped fiber layer in hopes of encountering buried cooking facilities just inside the shelter drip line.

After removing the disturbed surface layer and remnants of the mud drape/surface and exposing the underlying fiber layer (along with some scattered small burned rocks), we paused to thoroughly document the exposed layer using Structure from Motion (SfM) photogrammetry, as described by Charles Koenig in the January 23rd blog entry. While Steve and Drew took the photos, I busied myself with closely examining the floor of the shelter for interesting perishable artifacts. From my previous experiences in similar sites in Mexico and New Mexico, I know that tiny textile fragments, cordage, and unusual plant parts are often found just lying about the surface in dry rockshelters.

Again, I was not disappointed. In the middle of the previously exposed area of Feature 4 I noticed a small loop of tightly-wound fiber cordage protruding from the dust. Based on the coarseness of the strands, I tentatively interpreted it to be made from twisted lechuguilla leaf fibers. Carefully blowing and brushing away the surrounding dust revealed that the cord was exposed in a drying crack of the mud surface and that in seemed to be encircling a fibrous mass of what appeared to be lechuguilla leaves, and was tied in a simple knot on top. However, the exact nature of the artifact was unclear because it was embedded in the hardened mud drape atop the thick upper fiber layer of Feature 4.

As I later learned, Dan had spotted the same embedded cord last summer when the mud surface was first exposed.  He had left in place to be looked at more closely as the feature was more thoroughly documented. In the intervening months the mud cracks had widened such that I could see the cord was wrapping something.  Given that Dan and Steve had thoroughly documented the upper part of the feature exposure using SfM, Steve authorized me to expose the object in order to thoroughly document it before removal. Although most of Feature 4 will be left unexcavated and carefully protected for the future, Steve decided the collection of the specimen was justified.  Such perishable artifacts are extremely rare in the archaeological record except under the best preservation conditions, and because it will be vulnerable to future destruction by foot traffic in the shelter.  While human visitors can be encouraged to avoid sensitive areas of the site, feral hogs are known to frequent the shelter and wallow out sizable depressions where they sleep.

Cord-wrapped bundle as initially exposed.

Cord-wrapped bundle as initially exposed.

As all experienced archaeologists know, a time-honored rule in fieldwork is that the most interesting finds will most often occur in the last hours of your last day in the field. As it happened, I discovered the artifact about an hour before the end of the work day on a Saturday, and the crew takes Sunday off. What is more, I was scheduled to return home to Austin the next morning, so literally the find was made in the final minutes of my planned field visit before I had to leave Eagle Nest Canyon.  Rather than trying to stay late with the entire crew waiting around for me to finish exposing and removing the artifact (knowing that it would take some time), Steve and I elected to call it a day and return refreshed and alone the first thing the next morning.

Exposing and removing the artifact was a delicate and tedious process, and ultimately took several hours. Although it is well-preserved, the fibrous mass and cordage are very fragile, and they were firmly embedded in the mud drape.  In fact, the cord-wrapped bundle’s mud encasement is probably what protected and preserved it so well in the first place.  I worked to remove it hunched close to the ground, using a dental pick and tweezers (and occasionally the pliers on my multi-tool!) to carefully break up the hardened mud and separate it from the cordage, fibers, and thin fragments of prickly pear leaf epidermis (the “skin”) encircling the bundle. I constantly alternated between picking and tweezing and carefully brushing/blowing away the resulting debris.  Meanwhile Steve photographed the action, took notes, and lined a plastic container with paper towels in which we would cradle the artifact once it was freed.

Kevin lifts bundle out from where it has rested for 650 years.

Kevin lifts bundle out from where it has rested for 650 years.

The revealed artifact proved to be a short “bundle” of fibrous leaf material and a desiccated prickly pear pad bound with a short length of fiber cordage that had been tied into a simple knot on one side. Overall, it was less than 20 cm long and was found to be on top of two additional fragmentary prickly pear pads within the dense vegetal layer in Feature 4. The item must have been sitting on top of the fiber layer when the massive flood event deposited the sheet of mud that then hardened around it, encasing it and preserving it for us to find centuries later (although this is only our preliminary interpretation). Steve took several rounds of SfM photographs as the artifact was exposed and cut free.  The artifact was finally removed by severing the extraneous fibers and leaf fragments that clung to it and fused it to the vegetal layer below. We gingerly lifted it out and carefully put it in the box.

Cord-wrapped bundle awaits further investigation.

Cord-wrapped bundle awaits further investigation.

Back in the ASWT field lab on the Shumla Campus the cord-wrapped bundle awaits thorough documentation and cautious exploration. Next month the archaeobotany team hopes to more thoroughly clean the bundle and examine it in Phil Dering’s laboratory. We will have to strike a delicate balance between exposing it further for identification and interpretation and keeping it intact and preserving it for future study. As the analysis progresses, we will more fully assess what it is and what purpose it served, but for now, I simply classify it as a “bundle of fibrous leaf material, bound by knotted fiber cordage.”

Perishable items such as this more than likely comprised the vast majority of the material culture of the indigenous hunter-gatherer populations who called the Lower Pecos Canyonlands home.  But most perishable artifacts are simply no longer observable in the archaeological record in most settings.  They aren’t preserved in open campsites and, sadly, over the past 80 years most of the rockshelters of the region have been dug into by artifact collectors. The mere fact that this simple artifact has survived the centuries to be discovered and studied is testimony to the spectacular preservation conditions in the Eagle Nest Canyon shelters, and the rare window they provide into the daily lives of the ancient inhabitants of the Lower Pecos Canyonlands.  Once again we must tip our hats to the Skiles family for their stewardship of the canyon.

[Postscript by Steve Black:  We have enjoyed showing the curious cord-wrapped bundle nested in its box to our volunteer and visiting archaeologists.  Several have put forth interesting ideas.  Steve Davis pointed to the fact that the leaf sections protruding from one end of the bundle are charred, and suggested the object might be some sort of torch.  Carolyn Boyd reminded me in many ethnographic accounts from Mexico and the Southwest bundled plant materials were burned to create smoke during rituals. Elton Prewitt pronounced that the prickly pear skin visible within the bundle is part of a prickly pear pouch containing quids and leaf bases. Amy Reid made the excellent suggestion that we get the artifact x-ray CT scanned before dissection. All these ideas are being pondered and I expect the bundle has more than one secret yet to be revealed.

Feature 4 is proving to be much more substantial and complex than Dan and I had realized.  The 2014 Expedition crew is continuing to excavate several small excavation units through what we now think is a large earth oven facility that was used many times.  The upper fiber layer, cord-wrapped bundle, and the mud cap/drape are merely the last episodes of a much longer story.  I suspect the multiple lines of evidence we are developing on Feature 4 will be studied and written about for years to come.]

The Archaeobotanical Team Forms

By Brooke Bonorden and Steve Black

Recently we have formed an Archaeobotanical research team that parallels the Geoarchaeological team discussed in an earlier post, “Geoarchs in Action”.  The Archaeobotanical team leader Dr. Phil Dering has been joined by Dr. Leslie Bush and Dr. Kevin Hanselka as our collaborating experts in archaeobotany, meaning that they specialize in the analysis and identification of macrobotanical remains (plant parts visible to the naked eye) recovered in archaeological excavations.

Phil Dering leads the Archaeobotanical team through Eagle Nest Canyon.

Phil Dering leads the Archaeobotanical team through Eagle Nest Canyon.

Blog followers were introduced to Phil in the recent post, “Winter in the Canyon”.  Phil and Steve have dreamed and plotted about excavating and sampling dry rockshelters in the Lower Pecos for years.  Now that is finally happening, we are generating much more plant work than Phil can handle on his own, so he has recruited Kevin and Leslie to join him.  While Phil has been looking at macrobot samples from Hinds Cave and other Lower Pecos shelters for over three decades, he realizes that fresh eyes will bring new strengths and ideas to our collaborative endeavor.

Leslie Bush, who earned her Ph.D. at Indiana University Bloomington, runs her own macrobotanical consultation service based in Austin and she analyzes samples from many CRM (cultural resource management) projects in Texas.  She is also quite active in the Texas Archeological Society and the Travis County Archeological Society.  Working with Lower Pecos samples will allow Leslie to expand her plant knowledge westward from the Eastern Woodlands and the eastern two-thirds of Texas where she has done most of her previous work.  Kevin, who earned his Ph.D. at Washington University in St Louis, is a bit more familiar with desert plants.  His dissertation was on plant use in Southwestern Tamaulipas and he has also worked in Chihuahua with Dr. Bob Hard of UTSA. Kevin now works as an archaeologist for TxDOT and he jumped at the chance to become involved with ENC plant remains and put his training to use.  How does that saying go, three heads are better than one?

I  (Brooke) was fortunate enough to tag along with the archaeobotanists on their recent visit to Eagle Nest Canyon as they formulated a game plan for dividing and conquering the wealth of plant remains that we are recovering from the ENC rockshelters. From their conversations I learned that archaeobotany is even more intensive than I had thought. Plant species vary considerably across the Lonestar State, so becoming familiar with the types found in different regions takes lots of time and a trained eye.  During their visit, Kevin and Leslie collected samples of flora commonly found in the Lower Pecos Canyonlands to improve the comparative collections in their labs back in Austin.  I hope to help Phil collect more comparative samples this spring, as various plants begin to flower and fruit.  In fact, I hope to be able to help all three of our plant experts as the project unfolds and learn whether archaeobotany might be a viable career route for me.

As mentioned in earlier posts, the mantra for our expedition (and more specifically our excavation methods within Eagle Cave) has become “Low Impact-High Resolution,” meaning that we are attempting to learn as much as possible from the site while simultaneously minimizing our impact upon the remaining intact deposits within the shelter.  To do this we are exposing stratigraphic profiles by digging into disturbed deposits and cutting back until we encounter intact layering.  We are carefully documenting and sampling the intact layers as future posts will detail. Due to the modest size of our sampling areas, much of what we will learn from Eagle Cave will come from intensive analyses of matrix samples from the many strats (stratigraphic layers) we are encountering. Over the course of our expedition, therefore, we will be visited by collaborating experts on everything from geology to plants to bugs to bones in an effort to learn as much as possible from our strats.

Brooke describes the layers she has been exposing in PS 2 (Profile Section 2) at the back of Eagle Cave. The Witte Museum trenched all along this wall in 1936.

Brooke describes the layers she has been exposing in PS 2 (Profile Section 2) at the back of Eagle Cave. The Witte Museum trenched all along this wall in 1936. Leslie and Kevin examine the first intact layers we have encountered in PS2.

The Three-headed Archaeobot visit took place the same day as the Archaeolympics and so while everybody else played with atlatls and fire drills, we (Brooke, Steve, and Tina Nielsen) got to see the Canyon through the eyes of the plant experts.  We visited each of the three rockshelters where ASWT archaeologists have been working, Skiles Shelter, Kelley Cave, and Eagle Cave.   At each we looked at the excavation exposures and Steve explained our working ideas about the stratigraphy and we talked about the plant remains we have recovered. Phil, Leslie and Kevin soon lapsed into rapid-fire plant speak using the specialized terms like “bast,” as Brooke and Tina struggled to take notes and Steve kept nodding, giving the misleading impression that he understood all they were saying.  But as the archaeobotanists talked the talk, we lurkers began to appreciate what they were after and what they would need from us to make it happen.

In addition to being knowledgeable about the plant species across the landscape, archaeobotanists must also be experts in plant anatomy, considering that most of plant remains archaeologists find in the ground are tiny fragments, many of which are burned.  One of the difficult tasks that our experts will face with the ENC samples is identifying the materials in cordage/basketry/sandals, because these plant fibers have been pounded so that the epidermis is removed and only the bast (the insides) remains.  But while it is very difficult and sometimes impossible to identify the plant species represented by processed fiber, we think our sites have the well-preserved evidence of full behavioral chains from plant collecting to baking to food consumption to fiber production to weaving and plaiting to finished object to final discard.  Documenting and understanding these interwoven sequences of behaviors is one of our goals.  The tons upon tons of burned rocks spilling out of the mouths of all of the ENC rockshelters (except Bonfire) show that baking desert succulents like agave lechuguilla and sotol was one of the most common activities in the Canyon.

Kevin (far right) points out scars that suggest the deep mortar hole that goes all the way through this small boulder was quite intentional.  He has seen similar "artifacts" in Tamaulipas.  Incidentally, we think J. Charles Kelley described this very rock in May 1932 when he and E. B. Sayles dug small tests in Eagle Cave.

Kevin (far right) points out scars on the bottom side of this small boulder suggest the deep mortar hole that goes all the way through it was quite intentionally created this way.  Perhaps the open-ended mortar hole served as a “hopper” below which a basket might have been place to catch the pulverized food remains. He has seen similar “artifacts” in Chihuahua. Incidentally, we think J. Charles Kelley described this very rock in May 1932 when he and E. B. Sayles dug small tests in Eagle Cave.

At most archaeological sites archaeobotanical analysis involves identifying charred plant remains—wood charcoal and sometimes burned nuts and seeds—because that is all that is left.  Some charred “macrobotanical” remains can be identified with the unaided eye, but most specimens require careful microscopic examination and comparison with reference samples from known plants.  At Kelley and Eagle Caves, however, many of the plant remains have preserved in a desiccated state due to the dry conditions of the protected rockshelters.  We are finding many kinds of uncharred leafs, flowers, seeds, nuts, fruits, branches, and so on, as well as massive amounts of charred remains.  For our experts, looking at uncharred floral remains under a microscope will be a welcome change from what they see most of the time. Identifying these desiccated plant parts requires a more complete comparative collection that is needed when all you have is charred wood fragments.

Three archaeobotanists examine newly exposed layers at Eagle Cave.  From left to right, Leslie, Phil, and Kevin.

Three archaeobotanists examine newly exposed layers at Eagle Cave. From left to right, Leslie, Phil, and Kevin.

Considering the great lengths that Drs. Dering, Bush, and Hanselka will be going to identify the botanical remains from the ENC sites, you must be curious as to what we hope to learn from them. Plants were used intensively by native peoples not only for food, but for creating paint, medicine, basketry, rope, fermented beverages, sandals, and so much more. The ability to distinguish a large variety of plants across the landscape (and those that were edible from those that weren’t) and utilize them so resourcefully leads me (Brooke) to call prehistoric groups the first true botanists.  Macrobotanical analysis, when used in conjunction with historic ethnographic accounts of Native American groups and experimental archaeology, has proven useful in identifying the seasonality of site occupations (by the presence of plant remains that are only edible/blooming at certain times of the year), reconstructing economic patterns, and identifying plant processing activities.  The extraordinary organic preservation found in the dry rockshelters of the Lower Pecos also gives us a rare window into the many ways plants were used to create “material culture.”

The most shocking part? The archaeobotanists will only need a liter bag of matrix from each layer we sample to make inferences about all of that!  How is that for low-impact, high resolution?