Fire on the Mountain: Earth Oven Features in Nevada

Sheep Range roasting pit with banana yucca (foreground) and Joshua trees (background) growing around it.

Editor’s Note: This blog piece was written in 2016, but is only now being posted owing to …  the distractions of ASWT research.  Spencer, a stalwart member of the 2016 ENC crew, has returned to southern Nevada where he is again working on the Desert National Wildlife Refuge.

By Spencer Lodge

Hello everyone, this blog is a little different than previous posts.  Instead of focusing on work we accomplished in Eagle Nest Canyon, I will highlight what I have learned about the earth oven facilities I recorded in southern Nevada as part of my 2016 M.A. thesis at Texas State University, “Fire on the Mountain: Roasting Pits in the Sheep Range on Desert National Wildlife Refuge.” (Google the title if you would like to read my thesis.) My study area represents an interesting contrast with the Lower Pecos Canyonlands, a contrast I appreciate all the more after spending six months (Jan-June 2016) helping to excavate and document the copious evidence of earth oven cookery at Eagle Cave and Sayles Adobe.

Before moving to Texas for grad school, I worked on the Desert National Wildlife Refuge located some 20 miles north of Las Vegas. While working there I recorded nearly 200 roasting pits (i.e., ring middens) throughout the Sheep Range, the primary mountain range on the Refuge. When researching graduate programs, I was attracted to Texas State University due to the focus on earth oven technology by Dr. Black. Even though 1,000 miles separate southern Nevada and southwest Texas, I find the similarities in earth oven technology between both areas to be quite interesting.

My Study Area

Satellite image of the Sheep Range within the Desert National Wildlife Refuge (red) where my research was conducted.

My study focused on the Sheep Range, located roughly 20 miles north of Las Vegas in southern Nevada, in the fuzzy boundary between the Great Basin to the north and Mojave Desert to the south. The Great Basin, known for its internal draining system which resulted in pluvial large lakes during the Late Pleistocene and Early Holocene, is evidenced by the Desert Dry Lake located directly northwest of the Sheep Range. Flora associated with the Mojave Desert is located at various ecoregions spread across the Sheep Range, including Utah agave, Joshua Trees, Banana Yucca, and Mojave Yucca.

Utah agave (left) and banana yucca (right), two of the plants that were likely baked in Sheep Range roasting pits.

The extremes are very apparent in southern Nevada, from the bone dry playa beds up to pinion covered mountains. This region is dry, receiving 2-7 inches of rain per year in the lowlands, and up to 16-25 inches in in upper elevations (Mayer et al. 2012:30). Reliable water sources in the immediate area consisted of springs scattered throughout the Sheep Range. Due to the aridity and extreme nature of the landscape, prehistoric peoples in the region were highly mobile with relatively low population density. In fact, the Southern Paiute (or Nuwuvi) who inhabited southern Nevada at the time of Euroamerican contact had the lowest population density of any group in the Great Basin.

Ethnographic accounts for the use of roasting pits by Southern Paiute suggest this method of cooking was used for numerous plant foods, including agave, various species of yucca, and green pinyon pine cones.

View of a roasting pit (foreground) surrounded by Joshua trees and black brush with the Desert Dry Lake in the background.

Recording and Studying Roasting Pits

While conducting a survey on the Desert National Wildlife Refuge where the Sheep Range is found, I documented my first roasting pit. These earth oven facilities are comprised of rock, carbon-stained sediment, and charcoal, all by-products of hot-rock cooking. A typical roasting pit has a circular shape with a sunken central depression where the oven was built. To my surprise, a significant amount of the thermally-altered rocks were white, allowing us to spot the features several hundred meters away even with a pinyon tree growing from the center.

Even more surprising, the density of white rock allowed me to find more than 250 roasting pits using aerial imagery. The process I used to find these sites was rather straight-forward. Using Google Earth, I scanned the canyons and alluvial fans extending from the Sheep Range, staying at an average eye elevation of 7,000 ft. Potential roasting pits were marked to be verified in the field.

A volunteer writes the photo log as I (background) search for associated artifacts.

In the field, roasting pits were measured, photographed, and the surrounding landscape was surveyed for associated material culture or additional roasting pits initially unidentified. Three primary types of measurements were taken: the diameter of the midden, the diameter of the sunken depression, and the size of the cooking pit when identifiable. Height measurement were taken as well, but deemed unreliable as I was unable to know exactly where the roasting pit midden ended and the topography underneath began.

Schematic model of roasting pit showing the terminology used by Lodge (2016).

Observations focused on the type of vegetative zone and topographical setting in which the feature was constructed, as well as how one roasting pit differed from those around it. (For example: Was it smaller? Did it appear more eroded or infilled? Was it located closer to resources?) Once a roasting pit was recorded, we either hiked or flew to the next one on my list. I was fortunate to have the aid of a helicopter to access the more isolated roasting pits, which was terrific given the rugged and generally undeveloped nature of the Sheep Range. In total, 193 roasting pits were recorded and another 30 potential cooking features were identified. After the end of this field project, I recorded an additional 10 roasting pits.

Example of a roasting pit that is smaller, more ephemeral than most, and partially filled in with sediment. Visible in the background is the helicopter that allowed me to access remote features.

Analyzing Roasting Pits

I was unable to excavate or perform any destructive forms of analysis to the feature or cultural materials found in association. Instead I decided to analyze roasting pit distribution and size measurements throughout the Sheep Range using a combination of statistics and ArcGIS. Statistical tests were used to see if roasting pit size varied significantly according on the vegetative zone it was built in, and to test the usefulness of my identification method in areas with poor visibility (tree cover). ArcGIS was used to see if roasting pits were more often built in clusters or equally distributed, and if clustered, whether “hot spots” of use could be identified. I also conducted thermal testing on rock samples and used X-Ray Defraction (XRD) to determine the material type as well as the reason why thermally-altered rocks turn and remain white.

For both statistical and GIS analysis, I looked at midden size as an indication of use. That is to say, since each cooking event results in additional waste in the form of spent rock, charcoal, carbon stained sediment, etc., I infer that roasting pits with larger middens were used more often than those with smaller middens. I measured the length and width of each midden (Exterior), as well as the interior depressions.

Example of a well-defined roasting pit with the central cooking depression covered in vegetation.

 

Both methods of analysis taught me several things. First, roasting pits built in lower elevation vegetative communities (such as the Creosote Brush Community) were smaller on average than those built in higher elevation communities (such as Mixed Shrub or Pinion/Juniper Communities). Since both fuel and food resources diminish along with elevation, higher elevation sites allowed people to revisit locations more frequently, increasing midden size more quickly in the process.

Second, GIS analysis indicates roasting pits were more often built in groups, as opposed to constructed evenly throughout the Sheep Range. Hot-Spot GIS analysis also highlighted certain portions of the Sheep Range where larger roasting pits were concentrated, suggesting these locations were more often frequented for baking foods.

Scatter plot showing the relationship between roasting pit size and elevation.

Cooking Features in Nevada and Texas

Like the burned rock middens of the Lower Pecos, the roasting pits of southern Nevada are synonymous with earth oven facilities. The term earth oven facility refers both to the cooking pit where foods are baked and to the associated debris created from cooking (spent rocks, charcoal, etc.). Roasting pits are comprised of a central depression where foods were baked and a surrounding ring midden.

One critical difference between earth oven features in southern Nevada and southwest Texas is the amount of excavation they have received. Here in Nevada, roasting pit excavation has been minimal compared to the extensive work done in Texas. Recovered botanical remains from several roasting pits suggest agave, yucca, and meat were all cooked within. Just over 60 radiocarbon dates have been obtained from southern Nevada roasting pits, mostly dating to the past 2,000 years and as far back as 3,800 B.P. In comparison, there are now hundreds of radiocarbon dates from earth oven facilities in southwest and central Texas! Ring middens are also present in Texas, however they tend to have larger accumulations of fire-cracked rock and occur in both open air and rockshelter settings. While earth ovens have been found in rockshelters in Nevada, roasting pits are found only in open settings.

Color Change

Perhaps the most striking difference between the earth oven facilities of the Lower Pecos Canyonlands and those of southern Nevada is how visible they are on the landscape. Roasting pits in the Sheep Range are overall much easier to spot due in part to their color. When heated to temperatures exceeding 875° C, dolomite and limestone, the preferred material type for pit roasting in the Sheep Range, changes color from gray to white. Roasting pit middens are not entirely comprised of white rocks, but rather a mixture of bleached and natural colored rocks. However, even a relatively minimal amount of white rocks intermixed within a midden allows roasting pits to stand out against the otherwise drab landscape.

An up close example showing the mixture of white and natural colored rock found with a roasting pit.

Limestone was also commonly used in earth ovens in southwest Texas, but in contrast to southern Nevada, burned white rocks are not commonly found. For example, while excavating in Eagle Cave in 2016 I encountered less than ten white limestone rocks among the many hundreds of fire-cracked rocks I handled. Instead, the burned limestone rocks in the Lower Pecos are typically dark gray in color after use, sometimes exhibiting a pinkish hue.

Nevada dolomite before (left) and after (right) being heated to 900 C.

Size Variation

Another reason why the Sheep Range roasting are apparent on the landscape is due to their size. On average, roasting pits were nearly a meter in height at their peak with a maximum recorded height of over two and a half meters! When you consider that roasting pits were commonly built in areas lacking tall vegetation, they simply had nowhere to hide.

A very impressive two and a half meter tall roasting pit in the Sheep Range.

Conclusion

Since wrapping up the 2016 field season at Eagle Nest Canyon, I have returned to southern Nevada to work once again at the Desert Refuge. Using what I’ve learned researching earth ovens in Eagle Nest Canyon, I intend to continue investigating roasting pits throughout the Range (I’ve already recorded another ten features, bringing my total to 203!).  I also hope to explore the experimental use of earth ovens and perhaps one day I’ll have the opportunity to excavate one of the roasting pits of southern Nevada.   Until then I’ll keep looking up at the Sheep Range and thinking about times not so long ago when the Chemehuevi, a Southern Paiute group, frequented the area.

“One could tell from great distances when people gathered mescal [and] see fires on all the mountains.”

 

Lodge, Spencer N.
2016    Fire on the Mountain: Roasting Pits in the Sheep Range on Desert National Wildlife Refuge. M.A. thesis, Anthropology, Texas State University.