Turkeys on the Edge

Tijeras Pueblo Overlook

I’ve written some posts before on the interesting recent research being done on the analysis of DNA and stable isotopes to study the genetics and subsistence of the turkeys of the prehistoric Southwest. A recent short paper adds an interesting dimension to this research, by looking at these issues in a sample of turkey remains from a site on the fringe of the Pueblo world, near its interface with the Plains.

The site in question is Tijeras Pueblo, in the Sandia Mountains just east of Albuquerque, New Mexico. The researchers were interested primarily in looking at the stable isotope chemistry of these turkeys to determine whether they primarily ate maize or wild plants, to try to determine how they were raised. In addition, they looked at the DNA of a subset of them to see if they belonging to a previously identified genetic line of domestic turkeys identified in the prehistoric Southwest, or to a separate line associated with modern wild turkeys. In theory, one might expect that turkeys that ate maize belonged to the domesticated line and ones that ate wild foods belonged to the wild one.

In fact, however, what they found was more complicated and interesting. The turkeys fell into two groups which were quite distinct in their chemistry: one that seemed to have eaten maize and another one that seemed to have eaten wild plants. However, the latter group did not have chemistry quite the same as that of the modern wild turkey specimens they compared it to, and was instead somewhat “intermediate” between the maize-fed ones and the wild ones. This suggested to the authors that these turkeys may have been free-ranged, eating a mix wild plants, some maize, and perhaps also insects, and that some of this free ranging may have been in the cornfields for pest control. Similar husbandry practices are documented in the modern Pueblos but had not previously been identified prehistorically.

Even more interesting, however, was the genetic data. Despite the sharp distinction between subsistence strategies implied by the chemical evidence, almost all of the tested specimens belonged to the domesticated ancient Southwestern lineage, and not the wild one. This suggests that the difference in husbandry practices did not correlate to separate origins of the turkeys, but to something different.

Comparisons to specimens from other areas shed some light on possible reasons for this pattern. The researchers compared these turkeys to some from the Albuquerque area, from the Gallina area, and from Arroyo Hondo Pueblo to the north in the Northern Rio Grande area. Since Tijeras Pueblo is at a relatively high elevation where maize agriculture is somewhat marginal, it might be expected that this environment explains part of the difference in turkey husbandry. And when compared with the nearby but much lower Albuquerque samples and the more distant but comparably high-elevation Gallina ones, there is some evidence for this: the Albuquerque samples grouped with the maize-fed Tijeras ones, and most of the Gallina samples grouped with the free-range Tijeras ones. However, the Arroyo Hondo samples, though also high-elevation, showed a much more maize-based pattern, so there is something more than environmental difference going on here.

The authors suggest that the position of Tijeras Pueblo on the eastern fringe of the Pueblo world, at its interface with the very different cultural world of the Plains, may account for the diversity of the turkey husbandry types shown in their data. Conversely, Arroyo Hondo was further within the Pueblo world, while the Gallina region was culturally distinct in ways that are still poorly understood. The authors recognize, however, that further research will be necessary to flesh out the context of these results. In any case, this is a very interesting paper that adds another little bit to our knowledge of the past.

Read Full Post »

Dating the Macaws

Macaw Feathers and Copper Bell on Display at Chaco Visitor Center Museum

One of the most exciting recent developments in the study of Chaco Canyon is the increasing use of scientific analysis of artifacts and other material remains to test and challenge previous theories based more narrowly on traditional archaeology. This includes the use of radiocarbon dating, which is widely used as a basis for developing chronologies in most other parts of the world but has been underused in the Southwest due to the availability of tree-ring dating for chronology building. Particularly with the development of accelerator mass spectrometry (AMS) radiocarbon dating, which requires vastly less material than earlier methods, however, it is now possible to gain direct dates on a very wide variety of materials, including many artifact types as well as plants and human and animal bones. This allows an independent check on dating of material by association with tree-ring dated architecture and pottery, which has been the traditional approach. The increasing use of AMS on the museum collections excavated decades ago from Pueblo Bonito, in particular, is starting to lead to some unexpected and surprising conclusions. This work is largely being done by archaeologists associated with the University of Virginia led by Steve Plog, in collaboration with colleagues at many other institutions.

One recent paper, about a year old now, reported some surprising results from the dating of the bones of one of the most distinctive species found at Chaco: scarlet macaws. These birds are not native to anywhere near the Southwest, and they must have been brought up from very far south in Mexico. They are disproportionately found at only a few sites in the Southwest, one of which is Pueblo Bonito. Traditionally it has been thought that the importation of macaws was associated with the “florescence” of Chaco, the roughly 100-year period starting around AD 1040 when most of the monumental great houses in the canyon were built and Chacoan influence is seen over a very large part of the northern Southwest. For this study, the researchers dated 14 macaws from Pueblo Bonito: 11 from Room 38, which had the highest concentration of macaw remains at the site, two from Room 78, and one from Room 71. Both of these latter rooms are in fairly close proximity to Room 38 within the site. They also dated four macaws from Mimbres sites in southwestern New Mexico, another area with a relatively high concentration of these birds that lies between Chaco and Mexico and thus could played a role in their procurement, and two from Grand Gulch in Utah, which is on the far fringes of the ancient Pueblo world and yet has produced a few macaw specimens.

The results were surprising, and they challenge the traditional association of macaws with the Chacoan florescence. Six of the Chaco birds dated to between AD 885 and 990 (all dates given here are at 95% probability), well before the florescence and a time when Chaco would have been much less impressive architecturally. This is, however, a time when population in the canyon was increasing rapidly through immigration from various areas that were affected by the big changes at the end of the Pueblo I period, as we have seen in my recent series of posts on Pueblo I. The authors of this paper don’t mention this population movement specifically, but they do suggest that this indicates that the later period of monumental construction and other signs of sociopolitical complexity was the result of a long period of developing complexity, which fits the demographic evidence pretty well.

Room 38, Pueblo Bonito

Six other birds date between AD 970 and 1035, which would put them shortly before or possibly at the very beginning of the florescence and building boom. This suggests that trade relations with the far south continued beyond the initial period when macaws were introduced to the canyon. The final two date between AD 1015 and 1155, which suggests they probably were procured sometime during (or even shortly after) the period of florescence. Overall the dates suggest that macaws were procured throughout most of the period of Chaco’s rise from the period when Chaco was first rising to regional preeminence in the ninth and tenth centuries until its loss of preeminence (I think “collapse” is too strong a term for this still poorly understood phenomenon) in the twelfth.

One thing you may have noticed about those date ranges, however, is that they all overlap. Given the statistical uncertainty of radiocarbon dates, this means that it’s possible that these dates indicate a continuous process of importation of macaws from Mesoamerica. (There is no evidence for breeding of macaws at Chaco, unlike at the later site of Casas Grandes in northern Chihuahua.) The clustering of sets of dates, however, suggests on the contrary that importation was sporadic, with possibly just three individual procurements of multiple birds at a time. And additional complication is that the shape of the radiocarbon calibration curve differs at different times through this sequence, which can lead to certain time periods being over- or under-represented in series of dates. To test these hypotheses, the authors did some simulation of random dates throughout the period in question and compared the resulting distributions with the actual distribution of macaw dates. The results were that the early cluster of dates did conform to what might be expected from the effects of the shape of the curve, the middle cluster had more dates than would be expected and the late cluster fewer. This suggests that while it is possible that procurement of macaws was a continuous process, it does appear that a larger number of birds were imported in the late tenth and early eleventh centuries than earlier or later. Of course, this is a small sample, and these apparent patterns may change with more data.

As for the non-Chaco macaws, one of the Mimbres ones dated to AD 895 to 1020, straddling the first two clusters of dates at Chaco, while the other three all dated from around AD 1015 to 1155, as did the two Grand Gulch specimens. This suggests that macaws were present earlier at Chaco than in areas to either the north or south, which further suggests that at least initial importation of macaws to Chaco didn’t necessarily depend on Mimbres middlemen. Macaws have also been found at Hohokam sites in southern Arizona that appear to be in earlier contexts than the ones at Chaco, but none of these have yet been directly dated.

Room 33, Pueblo Bonito

While it may appear initially surprising, the early dates for macaws at Chaco do actually fit with increasing evidence from other sources suggesting that the rise of Chaco and its social and economic power significantly predated its “florescence” as seen in monumental architecture. This includes a study from a few years ago, from the same group of Virginia researchers, that dated human remains from Room 33 in Pueblo Bonito, including the two burials that were associated with enormous numbers of valuable grave goods, and found those two burials long predated the Chacoan florescence and may in fact have been contemporary with the earliest construction at the great house in the mid-ninth century or even earlier. (That paper really deserves a post of its own, which I keep meaning to write, but this brief summary will have to do for now.)

Taken in conjunction with the evidence for regional population movement in late Pueblo I, this study provides more support for the idea that the influx of populations into the canyon in the late ninth century, some bringing ideas developed in the earlier short-lived villages to the north in Colorado, set the stage for the development of new ideas about social organization and hierarchy which may have led to new ideologies and the importation of both goods and ideas from areas far away. The fact that macaws would have to have come from the south, where the archaeology of areas immediately adjacent to the Chacoan region is much more poorly known than that of comparable areas to the north, points to the importance of developing a better understanding of those areas. We still know very little about the exact routes of trade connections to the south, even as the importance of those connections becomes increasingly apparent.

Watson, A., Plog, S., Culleton, B., Gilman, P., LeBlanc, S., Whiteley, P., Claramunt, S., & Kennett, D. (2015). Early procurement of scarlet macaws and the emergence of social complexity in Chaco Canyon, NM Proceedings of the National Academy of Sciences, 112 (27), 8238-8243 DOI: 10.1073/pnas.1509825112

Read Full Post »

Foreshadow: Animal Bones

McPhee Reservoir, Dolores, Colorado

The first of the shorter, more analytical chapters in Crucible of Pueblos that follow the regional summaries is one by James Potter looking at faunal remains, which in this context basically means animal bones. (I guess this is sort of appropriate for a Halloween post, although animal bones aren’t really as spooky as human ones.) This chapter is basically a series of statistical comparisons of faunal assemblages from different Pueblo I sites, focusing particularly on the large, well-document collections from the Dolores and Animas-La Plata Projects, but also including a few others. Given the focus on these collections, the geographical range of these comparisons is limited to the Central and Eastern Mesa Verde regions. Nevertheless, Potter finds some striking differences between different sites that have interesting implications for understanding their inhabitants’ lives.

The first comparisons are of different villages within the Dolores area. Potter uses two widely used calculations, known as the artiodactyl index and lagomorph index, to compare McPhee Village on the west side of the Dolores River to Grass Mesa Village on the east side. The artiodactyl index is a measure of how common large game animals, such as deer and elk, are within the overall assemblage, and is calculated by taking the number of artiodactyl specimens in the assemblage and dividing it by the number of artiodactyl specimens plus lagomorph (rabbit and hare) specimens. The lagomorph index compares the number of specimens of the two most common lagomorph species, cottontail rabbits and jackrabbits, and is calculated as the number of cottontail specimens divided by the combined number of cottontail and jackrabbit specimens. This is an important measure because cottontails and jackrabbits favor different habitats and have different behavior which can shed light on human land use and hunting practices: jackrabbits prefer open spaces such as those created by clearing land for agriculture, and as a result can often be caught within gardens, while cottontails prefer more sheltered brushy environments. Jackrabbits also run to escape predation while cottontails hide, which makes the former more vulnerable to the kind of communal hunting known to have been practiced by Pueblo peoples in more recent times.

In the case of McPhee and Grass Mesa Villages both indices show little to no difference between the two; indeed they are nearly identical. This suggests that there weren’t major differences between the two communities in land clearing, communal hunting of lagomorphs, or hunting of artiodactyls. This is maybe not surprising, as the two villages are only a few miles apart and in similar ecological settings.

Where they do differ, however, is in another comparison, in this case of the prevalence and diversity of bird remains. McPhee Village has many more bird remains, representing more than twice as many species, than Grass Mesa, despite the overall sample sizes being similar. Furthermore, the avian bones are concentrated within McPhee Village at one particular residential site, known as McPhee Pueblo. This is one of the largest residences in the community and has features that have been interpret as reflecting ritual activity at a level higher than the individual residential group inhabiting the site. This site is considered likely to be a prototype of the “great houses” associated with the later cultural phenomenon centered on Chaco Canyon, where many of the inhabitants of the Dolores area are thought to have gone after the demise of the Pueblo I villages there in the late ninth century AD. The greater number of bird species, and the large number of specimens, at McPhee Pueblo reinforces other indications of the special role this site played in the community. Birds have often been associated with ritual among the Pueblos, with the macaws at Chaco being only one of the most spectacular examples. The fact that there is no similar site at Grass Mesa, and that bird remains are much rarer there overall, suggests significant differences in ritual organization at the two villages despite their proximity, which fits with other evidence suggesting they were settled by people from different cultural backgrounds.

The second major set of comparisons Potter makes addresses change over time, again within the Dolores area. He compares the artiodactyl and lagomorph indices of McPhee Village and the nearby but earlier community of dispersed hamlets known as Sagehen Flats. In this case, the Sagehen Flats sites had much lower artiodactyl indices, which suggests to Potter that this community had more difficultly organizing hunting parties to capture these large animals than the later, larger, and more aggregated community at McPhee. Indeed, it has been suggested that one reason for the formation of the large Pueblo I villages was the opportunity that larger communities provided for more effective hunting of large animals, especially in high-elevation areas close to large populations of artiodactyls.

Sagehen Flats also had a higher lagomorph index value, indicating more cottontails relative to jackrabbits, and suggesting that aggregation at McPhee also included more clearing of land for agriculture, creating the open spaces preferred by jackrabbits. It is also likely that larger communities were more effective at communal hunting, which as noted above would have been easier with jackrabbits. It’s not really surprising that larger communities would have cleared more land for agriculture and conducting larger communal hunts, but this evidence does provide another reason to think that.

Bird remains, on the other hand, were present in very similar proportions at both Sagehen Flats and McPhee, with both much higher than Grass Mesa. This likely results in part from the location of Sagehen Flats near marshes with lots of good habitat for waterfowl, but it’s also noteworthy that the bird remains there, as at McPhee, were heavily concentrated in one habitation site. This site, unlike McPhee Pueblo, doesn’t show other signs of having been exceptionally important compared to others, but it is highly intriguing that there were so many birds there, and it suggests that the pattern of unequal ritual influence seen at McPhee, and later at Chaco, goes back even further, at least in this area.

Durango, Colorado

Next, Potter does a broad comparison of several different site areas, this time treating the Dolores sites as a whole and comparing them to the nearby hamlet of Duckfoot as well as the site clusters of Ridges Basin and Blue Mesa to the east near the modern city of Durango, as well as sites in the Fruitland area to the south near the modern Navajo Reservoir. Starting with the artiodactyl and lagomorph indices, Potter finds high artiodactyl index values at Dolores and Ridges Basin, with much lower ones at Duckfoot and Fruitland. The factors mentioned earlier leading to more effective artiodactyl hunting in larger villages are probably one factor here, with another being elevation, with the higher sites having more artiodactyls than lower ones.

The lagomorph index is highest at Duckfoot and Blue Mesa and lower at Dolores and Ridges Basin, again echoing the pattern seen before where larger villages show evidence for more land clearing and communal hunting compared to smaller, more dispersed sites.

Turning to birds, Potter finds the highest numbers in Ridges Basin, with significantly smaller numbers at Dolores and Duckfoot. (Keep in mind that all of the Dolores sites are lumped together here.) This is likely due in part to the marshy environment of parts of Ridges Basin, but it is also due to much more extensive use of turkeys in Ridges Basin than elsewhere.

Following these rather simple comparisons, Potter does a correspondence analysis of all of the areas comparing categories of animal remains: birds, wild carnivorous mammals, domesticated dogs, lagomorphs, and artiodactyls. This analysis shows that the areas have very distinct associations with particular types of animals. Blue Mesa, Fruitland, and Duckfoot are associated with lagomorphs, Dolores with artiodactyls, and Ridges Basin with both birds and dogs. Potter notes that while Dolores and Ridges Basin have very similar artiodactyl indices, as this analysis suggests, they have very different overall percentages of artiodactyls. The index is thrown off because it uses lagomorph numbers to standardize the artiodactyl numbers, which is problematic in cases like this because the number of lagomorphs also differs a lot between the two areas, with a lot fewer of them at Ridges Basin than at Dolores.

Next, Potter does a detailed analysis of the Ridges Basin community, comparing categories of remains among different site clusters within the basin. He uses a more detailed set of a categories here than in the previous analysis: mammalian carnivores, birds of prey, waterfowl, dogs, turkeys, game birds, artiodactyls, and lagomorphs. The different site clusters show interesting differences in the proportions of these, with the marshy eastern cluster having higher numbers of waterfowl and turkeys. As mentioned above, turkeys are more common throughout Ridges Basin than in other Pueblo I communities, but there are differences in both numbers and context within the basin. The turkeys in the eastern sites are mostly burials, part of a widespread Pueblo practice of burying domestic animals that likely has ritual significance. In some site clusters, however, there is evidence for processing of turkey remains suggested they were used as food. In the north-central cluster there is one pit structure that seems to have been used as a processing area for turkeys and rabbits, and the same site also had turkey eggshells, suggesting strongly that these were domesticated rather than wild turkeys.

Dogs, wild birds, and carnivorous mammals are found mostly as burials throughout Ridges Basin, with some accompanying human burials. This is in contrast to McPhee Pueblo, which as mentioned above had high numbers of wild birds, where remains of ritually important animals like these were found in association with ritual structures. There is no such association anywhere in Ridges Basin, suggesting that while these animals were likely ritually important in both areas, the exact nature of the associated ritual differed.

As for artiodactyls, here as elsewhere they were found in greater numbers at the only aggregated site cluster that can be considered a village: Sacred Ridge. Since this site also has higher numbers of projectile points and processing tools, Potter suggests that the artiodactyls were the result of more effective hunting parties drawn from the larger village population, rather than evidence for special status of the residents of Sacred Ridge or special feasting being conducted there. There are a lot of unusual features to this site, however, so it’s hard to know how to interpret it.

That concludes Potter’s analyses. He ends the chapter with some conclusions that they suggest. First, as seen in multiple analyses, large sites tend to have more artiodactyls than small ones, probably because larger, more aggregated settlements allowed for the building of cooperative hunting parties that were more effective at taking down large game. This was a definite material advantage to community aggregation and the formation of villages, a key characteristic of the Pueblo I period that has led to a lot of questions about why and how it happened. It’s noteworthy, however (although Potter doesn’t note it) that the Pueblo I villages were as a rule short-lived and many seem to have been abandoned under duress, so the greater cohesiveness that allowed for these more effective hunting parties seems to have had definite limits under the circumstances.

Another pattern that emerges is the association of some sites with marshes and the extensive use of waterfowl, and presumably other marsh resources, at these sites. Potter connects this with the general importance of marshes, lakes, and other water places in Pueblo religion and ritual, as well as with the later artificial reservoirs built in the Mesa Verde region. It’s possible that an initial tendency to settle near wetlands because of their practical advantages in terms of resources led over time to a more metaphysical attitude toward watery places, although this remains highly speculative.

There is also a tendency over time for a shift in the contexts in which remains of animals of presumed ritual significance, like wild birds and carnivorous mammals, with early sites such as those in Ridges Basin having them largely associated with burials and the ceremonial “closing” of residential sites, whereas at later sites such as those in the Dolores area they are more associated with communal ritual structures. This suggests a shift in use of these religious symbols from the private to the public sphere, which Potter notes has also been proposed over the same period for the use of red ware pottery, which also likely had ritual significance. This shift may have continued into the rise of the Chacoan system, with its increased focus on monumental architecture presumably associated with public ritual.

Finally, Potter notes the early importance of turkeys in Ridges Basin, which could be due to general environmental differences across the region but may also reflect earlier depletion of large game in this area compared to others. There is a general pattern through Pueblo prehistory of increasing use of turkeys for meat as artiodactyl use declines, presumably in response to overhunting of local populations. On the other hand, one intriguing thing about the greater use of domesticated turkeys at the eastern edge of the Mesa Verde region during Pueblo I is the genetic evidence showing that domestic turkeys in the Southwest are likely more closely related to wild subspecies found to the east than to those found locally. Could the use of turkeys in Ridges Basin reflect early contacts with peoples further east? Potter doesn’t mention this possibility, and I don’t know if there is any other evidence of such contacts, but again, intriguing.

So, yeah, this chapter is a lot more focused than those coming before it, but the results of its analyses are intriguing. As more evidence becomes available from other regions with Pueblo I populations it may be possible to extend these sorts of comparisons further.

Bone Tools at Chaco Museum

Read Full Post »

An Interesting Macaw-Related Fact

Room 38, Pueblo Bonito

In a recent post, I noted the limited distribution of macaw remains within Pueblo Bonito.  While this site has a much higher number of macaws than any other Chacoan site, and more than almost every other site in the prehistoric Southwest, within the site macaw remains were highly concentrated.  All macaws were found in the eastern half of the site, and most were in the eastern part of “Old Bonito” at the northern end of the overall site, particularly in Room 38, which had twelve.  This suggests to me that macaws were closely associated with whatever social group lived in or used that part of Pueblo Bonito.

It’s hard to say what that social group was, but it’s possible that the burials in a complex of four rooms in the northern part of Old Bonito were associated with it.  Associating these burials with the eastern rooms in Old Bonito is perhaps a bit of a stretch, since the burial rooms are actually in the western half of the Old Bonito arc, but they’re just barely on the west side, and there is a separate set of burial rooms at the far western end of Old Bonito that could be plausibly associated with whatever social group lived in or used those rooms.  There is no equivalent set of burial rooms in the eastern part of the arc, although there are a few isolated burials of infants and fetuses.  (Two of the infant burials, in Rooms 306 and 309, were associated with macaws.)  The eastern end of Old Bonito was covered over by later construction and is poorly known, but there is no evidence that it ever held a mortuary complex comparable to the one at the western end.  Given the circumstances, I think it’s plausible that the northern burial complex was associated with the group (or one of the groups) associated with the eastern part of Pueblo Bonito, perhaps in addition to the group associated with the immediately adjacent room suite at the west end of the northern part of Old Bonito, if this was indeed a different group.

Western Burial Rooms in Old Bonito

Analysis of the Pueblo Bonito burials by Nancy Akins for the Chaco Project found that, judging from cranial attributes, the northern and western burial groups were distinct from each other but internally homogeneous.  This suggests that they probably represented kin-based social units, and that the site consisted of at least two of these units, perhaps occupying or using different areas.  Akins couldn’t find a very large sample of burials from other sites in the canyon for comparison, but she was able to compare burials from three small sites in the canyon.  One of these, Bc 59, is across from Bonito on the south side of the canyon near Casa Rinconada, while the other two, 29SJ299 and 29SJ1360, are in Fajada Gap, a few miles east and the location of a substantial community of small houses and two great houses (Una Vida and Kin Nahasbas) in close proximity to Fajada Butte.

While the two burial populations in Pueblo Bonito weren’t particularly similar to each other, they more closely resembled the small site samples.  Specifically, the western burial population was similar to the one from Bc 59, while the northern burial group was more similar to the Fajada Gap group.  Importantly, the two Bonito groups were both more similar to these small-house populations than they were to each other.  This suggests that kinship connections among different sites in the canyon were complicated and didn’t break down on straightforward great house v. small house lines.

Bc 59 from Casa Rinconada

What does all this have to do with macaws?  Well, there is only one small house site at Chaco (as far as I know) that has produced macaw remains, and that site is… 29SJ1360, one of the sites with burials that patterned with the northern burial group at Bonito!  As reported by Peter McKenna in his report on this site, which was excavated by the Chaco Project, a few macaw bones were found in the fill from one of the pit structures.  While there were only a few bones found, they were all unique, suggesting the presence of only one macaw, and from various parts of the body, suggesting that the whole macaw was present.  This fill was only casually screened for artifacts and was later used to backfill the pit structure, so the rest of the macaw is probably still there.  This site also had an unusual architectural feature, a small bin attached to the outside of one of the roomblocks, that according to McKenna looked “remarkably like a parrot bin.”  One important feature that appears to have led to this conclusion was the presence of an adobe “plug” in the north wall, presumably reminiscent of the stone plugs used with “cage stones” at macaw pens at Casas Grandes, where there is substantial evidence for the keeping and breeding of macaws a few hundred years later.

This is all pretty tentative, of course.  Very few sites at Chaco have been excavated, so we have very little sense of the overall distribution of rare finds such as macaw remains.  Still, two separate lines of evidence (biological relationship and association with macaws) seem to point to a strong connection between the northern/eastern part of Pueblo Bonito and at least some sites in the Fajada Gap community, which is not particularly close to Bonito.  Given the rarity of macaws, especially, this seems significant.

Fajada Butte with Green Vegetation

Read Full Post »

Unusual Animals at Pueblo Bonito

Room 38, Pueblo Bonito

Pueblo Bonito is the best-known and most-studied site at Chaco, but there’s still a lot we don’t know about it.  Because it was excavated early in the history of Southwestern archaeology, provenience information for the vast numbers of artifacts found at Bonito is not nearly as precise as would be expected today.  We do generally have information about what was in each excavated room, and often where in the room specific artifacts were, but the careful stratigraphic approaches used today were either totally unknown or in their infancy during the excavation of various parts of Bonito, so interpreting the field notes and site reports can be a challenge.  Partly for this reason, a lot of recent interpretations of Chaco have been based mainly on the more recent and better-documented excavations by the Chaco Project in the 1970s.  This makes Pueblo Alto in particular, the only great house extensively excavated by the Chaco Project, enormously influential in recent interpretations, not always in beneficial ways.  The Pueblo Bonito data has been incorporated into most theories to varying extents, but this often just takes the form of vague gesturing at the elaborate burials and huge quantities of high-value artifacts found there, and sometimes it basically amounts to discounting the importance of Bonito because it is so unlike the other sites.

Still, Bonito is important!  The problematic nature of the documentation notwithstanding, there’s still a ton of data available, and the Chaco Archive has been doing excellent work lately in making it more widely accessible.  Their cool interactive map of the site even allows you to click on a room and see a list of all the features, artifacts, tree-ring dates, and pictures associated with that room.  I’ve been playing around with it a lot lately, and there’s really a ton of interesting stuff in many of the rooms that we don’t hear so much about.

Room 309, Pueblo Bonito

Building on what I was saying earlier about a badger burial at a small site excavated by Earl Morris near Aztec, I decided to look for unusual animal burials or remains that might suggest some patterns in ritual practices or group identities at Bonito.  Many modern Pueblo clans are named after specific animals, and it seems reasonable that some Chacoan social groups (which may or may not be equivalent or ancestral to the modern clans) might have had similar identities that would lead them to leave animal remains in certain contexts that could indicate connections through time between different rooms or sites.  The Chaco Archive database allows you to search for specific types of artifacts, and it even has a special option for non-human burials.  The database doesn’t have all the sites included yet, but it does have all of Bonito, and it’s a powerful tool for finding information about the sites that are included.

Starting with the non-human burials, the ones at Bonito seem to all be of macaws and parrots.  The close connection between Bonito and macaws has long been noted, and Room 38 is particularly known for its large numbers of them, but one thing I hadn’t realized is that, like so much else at Bonito, the distribution of macaws is highly concentrated, not just in a few rooms, but specifically in a few rooms on the east side of the site.  The macaw burials, in addition to the two in Room 38, are in Rooms 71, 78, and 306, all of which are in the eastern part of Old Bonito.  Not all of these are actually formal burials, but they are all complete skeletons.  Extending the search to individual bones adds Rooms 249, 251, 309, and 312, as well as Kiva J and the east mound in front of the site.  Again, these are all on the east side of Bonito, although not just in Old Bonito this time.  Rooms 309 and 312 aren’t technically in Old Bonito, but they are among the rooms added right in front of it, and are very close to Rooms 306, 71, and 78, which also had macaws.  Rooms 249 and 251 are in the block of rooms added onto the southeast part of the site over an earlier extension that apparently built over part of the eastern end of Old Bonito (this part of the site is very complicated and its construction sequence is poorly understood).  Kiva J is one of the six blocked-in kivas between this block and the plaza.  And, of course, the east mound is the easternmost of the two mounds.

Kiva J, Pueblo Bonito

What does all this mean?  Many have suggested that the number of macaws at Bonito indicates the possible presence of a macaw clan like the one known today at Zuni.  If this is indeed the explanation for all the macaws, and it seems plausible given the restricted distribution of them to just a few sites at this time and the contexts in which they are found, it seems that this clan probably lived in or had claims on the eastern part of Pueblo Bonito, and that this association held not just in the earliest stages of the site but even after it was expanded.  Perhaps members of this clan were the initial residents of the eastern suites in Old Bonito, then when those rooms were converted to other uses as the site was expanded they moved into the new southeast wing.

One question that might be raised at this point is whether this distribution is actually specific to macaws.  Maybe all exotic birds and animals are concentrated in this part of the site, which would suggest that there might be something special about the eastern half of the site but not necessarily anything tied to a specific clan.  Some research into the layout of the rooms has shown that the southeast corner is unusual in not being divided into obvious room suites, whereas the southwest corner seems to be.  Maybe instead of the macaw clan living in the eastern half, everyone lived in the western half and they used the eastern half for macaw-related (and other) ritual.

Room 330, Pueblo Bonito

One way to test this would be to look at other unusual animals.  Finding animals of ritual importance beyond obvious exotics like macaws is tricky, because many animals were certainly used for mundane purposes and their remains are therefore all over.  Dogs and turkeys were kept domestically, so their remains probably wouldn’t indicate anything special about social groupings, and game animals such as rabbits and deer might have interesting implications for access to different kinds of meat but, again, not necessarily specific symbolic implications.  That basically leaves animals that don’t serve a clear subsistence or other utilitarian purpose but are nevertheless found in sufficient numbers to suggest something more than mere chance is behind their presence.  The best example I’ve found: bears.

You basically never hear about bears in discussions of Chaco.  They are not present in the area now and probably weren’t in antiquity either, and their remains are certainly rare at Chaco but not entirely absent.  At Pueblo Bonito, bear remains are mostly concentrated on the west side of the site, in stark contrast to the macaw remains on the east side.  There are some artifacts made of bear bone, including two apparent gaming pieces, one each from Rooms 267 and 290 (both on the east side), but there are also unworked bear bones, especially jaws and feet, particularly concentrated in Rooms 92, 102, and 109, which are part of the same suite of rooms in the west wing of Old Bonito.  Room 92 also had a bear hide and mass of hair that is probably also from a bear.  Another room in this part of the site, Room 330, had a grizzly bear jaw.  Another bear jaw was in Room 66 and a claw was in Room 10; both these rooms are on the east side of the site.  So not as clear-cut as the macaw evidence, but still a strong suggestion that people with some sort of connection to bears lived in the western part of the site.  The “bear-paw” motif is well-known in rock art, and George Pepper, who excavated these rooms, reported that Room 97 (the room under Room 92) had similar “bear paws” drawn on the smoke-blackened plaster.  Finally, Kiva Q, the great kiva in the west plaza, contained a (dedicatory?) cache of objects that included bear paws.  This is all very suggestive, though of course not totally dispositive.

Kiva Q and West Plaza, Pueblo Bonito

There may be other examples of these sorts of patterns that could give us a better sense of who exactly was living at Pueblo Bonito and what other people at which other sites they had particularly close ties to.  Despite the fact that this information has been available for a long time, it’s only now that it’s starting to become widely available in a useful form.  New analytical techniques are revolutionizing our understanding of Chaco in all sorts of ways, but one of the most important contributions technology can make is just to make existing information available so it can be assembled, analyzed, and compared to information from elsewhere.

Bear Paw at Three Rivers Petroglyph Site

Read Full Post »

Casas Grandes Macaw Breeding

Wupatki Pueblo

One of the many similarities between Chaco and Casas Grandes in Chihuahua, in addition to the similar types of effigy vessels, is the presence of significant numbers of scarlet macaw skeletons at both sites.  As with most of these parallels, the evidence at Casas Grandes is more impressive in scale, with hundreds of macaws found as compared to dozens at Chaco.  It has long been argued that Casas Grandes may have been breeding the macaws, rather than importing them all directly from areas to the south, and that it may have even been the source of most of the macaws found elsewhere in the Southwest.  When it was thought that Casas Grandes was contemporaneous with Chaco, it seemed like an obvious source of the macaws found there and in the Mimbres area in between, but it is now known that the rise of Casas Grandes came well after the decline of Chaco, so wherever the macaws at Chaco came from, it wasn’t Casas Grandes.

So, leaving the question of Chaco and other early sites with macaws aside, what was going on with the macaws at Casas Grandes.  Were they indeed being bred and traded to other Southwestern sites, such as Wupatki in northern Arizona?  Or were they being imported directly from somewhere in Mesoamerica, either for trade or for local use?  A recent paper sets out to look at this question using stable isotope ratios, a type of analysis we have been seeing a lot of lately in Southwestern archaeology.  Specifically, the paper looks at the ratio of carbon-13 to carbon-12, widely used to determine the diet of an animal and specifically whether it was dependent on maize, and the ratio of oxygen-18 to oxygen-16, which reflects the chemical makeup of the water the animal drank and can be used as a rough proxy for the environment in which it was raised.  We’ve seen carbon isotope analysis recently used to show that Anasazi turkeys subsisted mainly on maize.  Oxygen isotope analysis is used less in the Southwest, but it is particularly useful here because the ratio varies with elevation, humidity, and proximity to the ocean, all of which are major differences between the high, dry Casas Grandes area and the humid coastal lowlands to the south where the scarlet macaw has its natural range.  The ratio thus gives a good sense of whether the birds were brought up from the south or raised right at Casas Grandes.  Strontium isotope ratios can pinpoint places of origin with more precision, but to be useful they have to be accompanied by a major sampling effort to determine the ratios in the various possible source areas.  The conclusion of this paper suggests that strontium analysis may be in the works, but as a first pass oxygen isotope analysis is a good way to go.

The carbon isotope analysis found that the macaws subsisted almost entirely on maize, although the very youngest nestlings seem to have been fed a more varied diet.  Clearly, then, these were not wild macaws captured in their natural habitat and brought north.  In addition, the oxygen isotope analysis found that all but one of the macaws had spent its life in the Casas Grandes area.  The exceptional macaw had an oxygen isotope ratio suggesting a more humid, low-lying environment, so it was likely traded up from somewhere in Mesoamerica.  This suggests that the macaws were indeed being bred at Casas Grandes, although trade relationships with Mesoamerican communities did involve the acquisition at least occasionally of additional birds.

This is all very interesting for understanding Casas Grandes, but what I find most intriguing about the paper is the possibility of using the same techniques on macaw bones from other parts of the Southwest, especially Chaco and other sites pre-dating Casas Grandes.  This paper seems to show pretty conclusively that macaw breeding was going on at Casas Grandes, but it’s still a very open question whether that was an innovation or the continuation of an earlier Southwestern tradition.  I think the Mimbres area is the best place to look for earlier macaw breeding, although obviously Chaco is a possibility as well.  The exact techniques used in this paper would really only be useful for determining if macaws were bred in captivity and if they came from Mesoamerica or somewhere in the Southwest, but the addition of strontium analysis would allow more precise identification of breeding areas within the Southwest, if indeed there were any earlier than Casas Grandes.

Somerville, A., Nelson, B., & Knudson, K. (2010). Isotopic investigation of pre-Hispanic macaw breeding in Northwest Mexico Journal of Anthropological Archaeology, 29 (1), 125-135 DOI: 10.1016/j.jaa.2009.09.003

Read Full Post »

So What Did the Turkeys Eat?

Escalante Pueblo

As if on cue, given that I’ve been talking about turkey husbandry and stable isotope testing of human remains, a paper in the latest issue of the Journal of Archaeological Science combines the two, using similar stable isotope techniques on turkey remains from sites in southwestern Colorado to determine what the turkeys were eating.  The idea here is to test two possible hypotheses that have been proposed before:

1. Turkeys were allowed to roam free in agricultural fields, eating whatever they found, especially insect pests that threatened the crops
2. Turkeys were instead kept in pens or otherwise confined in domestic contexts and fed table scraps or other leftover/surplus portions of human food.

The assumptions behind the tests were that if Theory 1 is correct, the turkeys would have isotope ratios significantly different from those of humans, reflecting a diet based on local plants and insects or other small animals, whereas if Theory 2 is correct, the turkeys would have isotope ratios similar to those of humans, since they would be eating the same types of food, mostly maize.  The main sample of turkey bones tested came from Shields Pueblo near Cortez, Colorado, which was occupied primarily during the Pueblo II and Pueblo III periods (ca. 1020 to 1280 AD) and has been extensively excavated by Crow Canyon Archaeological Center.  This site is particularly well suited to this kind of study because it has one of the largest and best documented samples of faunal remains known in the area.  Bone samples from jackrabbits and cottontails found at Shields were also tested to provide a control sample of animals that would certainly have been mainly eating local plants rather than maize.  The ratios were compared to those from earlier tests on human remains from the region, as well as a set of turkey bones from a wide variety of other sites in southwestern Colorado dating over a long period of time, including the outlying Chacoan great house known as Escalante Pueblo.

Blocked-In Kiva at Escalante Pueblo

The results were quite straightforward, and they clearly supported Theory 2.  The isotope ratios were very similar to those from human samples, showing extensive reliance on maize and little to no meat consumption, and quite different from the rabbit samples, which showed more reliance on native plants.  This was the case not only for the Shields Pueblo samples but also for those from other sites, and there was no clear evidence of change over time.  All of the turkey remains gave results in the same narrow range, regardless of site or time period.  The obvious conclusion is that turkeys were being kept in pens or other confined spaces in domestic contexts, and were fed table scraps or surplus cornmeal.  The authors also suggest another possible conclusion:

There is another behavioral explanation (besides feeding of surplus maize/table scraps) for the similarity in stable carbon and nitrogen isotopes between turkeys and humans. It is possible that turkeys were fed human fecal waste. The practice of feeding waste to livestock is not unheard of in prehistory.

. . .

While there are no records of turkey being fed human waste, it is common for turkeys to eat their own waste as well as the waste of other fowl. Given that maize does not completely digest in the human gut, it is certainly possible that some food was obtained in this way.

The only evidence they cite for this interpretation, however, comes from Korean pig raising, and as they note there is no evidence at all that anything similar was ever done with turkeys, in the Southwest or elsewhere.  Given that lack of evidence, I find this idea implausible.  Table scraps from human meals and/or surplus or specially prepared cornmeal seem like much more reasonable ways turkeys would have been fed.

Rawlings, T., & Driver, J. (2010). Paleodiet of domestic turkey, Shields Pueblo (5MT3807), Colorado: isotopic analysis and its implications for care of a household domesticate Journal of Archaeological Science, 37 (10), 2433-2441 DOI: 10.1016/j.jas.2010.05.004

Read Full Post »

More Turkey Stuff

Rooms in Northwest Part of Salmon Ruin

In looking into recent research on Southwestern turkeys, I found an interesting paper from 2007 by E. Bradley Beacham and Stephen R. Durand about turkey eggshell.  Specifically, they came up with a new technique for analyzing archaeological eggshell to determine whether or not the egg had hatched.  The idea behind it, confirmed by an experiment they did with modern wild turkey eggs, is that bird embryos take the material to develop their skeletons from the interior portion of their eggshells, so the longer an embryo has been developing in the egg the more reduced the inside of the shell will be.  This is clearly apparent in microscopic examination of modern eggshells.  The usefulness of this technique is that it can theoretically determine if turkeys were being bred, as opposed to simply being kept captive, if large numbers of eggs had hatched.  Beacham and Durand connect this to a longstanding dispute in Southwestern archaeology over how to interpret the evidence of archaeological turkeys, with one side arguing that turkeys were introduced from Mexico as already domesticated animals, with the wild turkeys currently known in the Southwest possibly descending from escaped domesticated turkeys that went feral, and the other side arguing that turkeys were domesticated in the Southwest from local wild turkeys, possibly quite late after many years of having been captured in small numbers for their feathers but not bred.  We now know from genetic evidence that both of these theories are wrong, although the former is apparently closer to being true, and it now seems that turkeys were introduced as domesticated animals, probably from the east rather than the south, and that this happened quite early, at least by Basketmaker II and possibly during the Late Archaic.

The archaeological eggshells Beacham and Durand analyzed using this technique were from two rooms at Salmon Ruin, a major Chacoan outlier on the San Juan River near present-day Bloomfield, New Mexico.  These rooms were excavated in the 1970s, and they were selected for this analysis because they contained significant quantities of well-preserved turkey eggshell.  Eggshell is very fragile and needs very good preservation conditions to survive, so the excellent preservation conditions at Chacoan great houses like Salmon are important in doing this kind of research.  The sampled shell fragments came from strata clearly identified with the three originally designated periods of occupation at Salmon: Primary or Chacoan (ca. 1088 to 1125 AD), Intermediate (ca. 1125 to 1190), and Secondary or Mesa Verdean (ca. 1190 to 1280).  These turned out to have quite different results.  The Primary shells overwhelmingly showed no evidence of having hatched, and little evidence of having held embryos long enough for material to be taken from them at all.  The Intermediate shells, on the other hand, mostly did show evidence of having held embryos for significant periods, and about half of them seem to have hatched.  The Secondary shells, most of which came from a different room from the other two samples, were more like the Primary ones in that they showed little evidence of embryo development and almost none of hatching.

Salmon Ruins Sign

This is odd, and hard to interpret in terms of domestication or intensity of use.  The sheer amount of Secondary eggshell throughout the site, compared to the much smaller amount from earlier periods, suggests increased use of turkeys during the thirteenth century, which is consistent with evidence from throughout the Southwest showing increased turkey use, especially for meat, during this period.  The shell evidence, however, while it does show that breeding seems to have taken place during the Intermediate period, doesn’t really fit with this increased overall use.  It’s possible that this is simply due to a sampling issue, and that further study of a larger sample of Secondary eggshell would show more evidence for breeding.  It is also possible, as Beacham and Durand note, that this increased use of turkeys involved the consumption of eggs as well as meat, and that the Secondary sample analyzed is associated with this use of eggs rather than with breeding.  Something similar may also be going on with the Primary sample, although Beacham and Durand, who clearly seem to favor a late onset of breeding and domestication, prefer to see it more as evidence that turkeys were not yet fully domesticated at that time.  As noted above, however, the genetic evidence argues strongly against this interpretation.

Given the genetic evidence, it is not clear how to interpret the eggshell evidence in this paper, but it offers an interesting new way to look at turkey eggshell in well-preserved contexts, and give the frequency of these contexts in the Southwest further use of the technique may further refine our understanding of the issues surrounding turkey use in the prehistoric Southwest.

BEACHAM, E., & DURAND, S. (2007). Eggshell and the archaeological record: new insights into turkey husbandry in the American Southwest Journal of Archaeological Science, 34 (10), 1610-1621 DOI: 10.1016/j.jas.2006.11.015

Read Full Post »

The Turkey Connection

Wild Turkey

In a comment to the previous post, Alan Reed Bishop brings up an issue closely related to the recent evidence for early maize cultivation in Chaco Canyon: the introduction of domesticated turkeys to the Southwest.  A recent study of archaeological turkey remains found that the majority of the turkeys found in Southwestern archaeological sites are genetically distinct from both the local subspecies of wild turkey and the subspecies found in Mexico that was domesticated there and is ancestral to the modern domestic turkey.  Instead, the Southwestern domestic turkeys were closest genetically to two subspecies of wild turkey found to the east and southeast, in the southern Plains and the eastern US.  This strongly implies that turkeys were domesticated somewhere to the east and then introduced to the Southwest as domestic animals, presumably through long-distance trade contacts.

The earliest remains used in that study were coprolites from Turkey Pen Cave in Grand Gulch, Utah, and date to the Basketmaker II period.  Some of these coprolites were also directly dated by AMS; the earliest had a 95% confidence interval of AD 20 to 200.  Like the bones from other sites analyzed in the same study, the Turkey Pen Cave coprolites indicated that most of the turkeys kept there belonged to the domesticated lineage, apparently non-local, that showed strong similarities to wild subspecies further east.  In addition, an earlier study of Basketmaker II subsistence in the Cedar Mesa/Grand Gulch area, using a variety of lines of evidence including coprolites, found that corn agriculture was already as central to the Basketmaker II subsistence system as it would be in later Pueblo times.  The presence of domesticated turkeys as well as corn agriculture as well-established aspects of Basketmaker II society seems to imply that both were introduced earlier, perhaps in the Archaic period, and the accumulating evidence for Archaic maize throughout the Southwest supports this supposition.  Less study has been done of turkeys, however, and while the DNA study refers to alleged Archaic turkey remains from the Southwest, the references are to obscure sources that I have not been able to track down.

Regardless of when domesticated turkeys were first introduced to the Southwest, they presumably were not introduced along with maize.  Turkeys seem to have been introduced from the east, while maize definitely came from the south.  It is possible that both came from northeastern Mexico, where one of the wild turkey subspecies similar to the Southwestern type is found, but there is basically no evidence for direct contact between that area and the Southwest, and the earliest evidence for maize there apparently dates to approximately the same period as the earliest Southwestern maize, suggesting that agriculture was not introduced to this area early enough to make it the vector for transmission to the Southwest.  It is much more likely that maize was introduced through western and/or northern Mexico, areas with extensive evidence for contact with the Southwest throughout prehistory.  So it seems quite clear that the introduction of turkeys and corn were separate events, but it seems equally clear that both were in fact introduced from elsewhere, probably during the Late Archaic, and it is striking that they seem to be present together from quite early on, at least on the Colorado Plateau.  (Turkeys are conspicuously absent from early agricultural sites in southern Arizona, which is another piece of evidence suggesting that they were not introduced from the south.)  I’m not really sure what the upshot of all this is, but it’s certainly interesting stuff.

Matson, R., & Chisholm, B. (1991). Basketmaker II Subsistence: Carbon Isotopes and Other Dietary Indicators from Cedar Mesa, Utah American Antiquity, 56 (3) DOI: 10.2307/280894

Speller, C., Kemp, B., Wyatt, S., Monroe, C., Lipe, W., Arndt, U., & Yang, D. (2010). Ancient mitochondrial DNA analysis reveals complexity of indigenous North American turkey domestication Proceedings of the National Academy of Sciences, 107 (7), 2807-2812 DOI: 10.1073/pnas.0909724107

Read Full Post »

Woof!

My Mom's Dog Mimi at Chaco

In comments to the previous post ben asked about the use of dogs as draft animals.  I replied that they were so used in conjunction with the travois, especially on the Plains, but that the dogs in the Southwest and in Mesoamerica were smaller than Plains dogs and not able to pull any substantial loads.  This reminded me that I’ve never done a post on dogs, and that I probably should.

The only domesticated animals the Chacoans and other prehistoric Southwestern peoples had were dogs and turkeys.  The first detailed study of Southwestern dogs, and the only one I am aware of, was done by Harold Colton of the Museum of Northern Arizona and published in 1970.  This paper is one of the earliest examples of the use of statistics in Southwestern archaeology.  Colton used statistical techniques to compare various measurements of dog remains from various parts of the Southwest and various time periods.  It isn’t clear if any of his 110 specimens came from Chaco, but it appears that none did.  This is unsurprising given that he mostly used specimens in the MNA collections, which contain little Chacoan material.  Nevertheless, his sample includes most parts of the Southwest and most periods, so it is reasonable to presume that any generalizations about Southwestern dogs resulting from it can be applied to Chaco as well.

The specific statistical techniques Colton used are rather different from those commonly used in archaeological publications today, and as a result I have a hard time evaluating them.  Nevertheless, some of the patterns he found seem clear.  Briefly, the early specimens, dating before AD 800, were all small, and some lacked the first premolar on the lower jaw.  After 800, however, a second type of dog, larger and never missing the first premolar, appeared, first in the Rio Grande Valley and later further west.  It is not clear from the presented data when the large dog would have reached the Chaco area, but I suspect it would not have been until after the main florescence of Chaco between AD 1030 and AD 1130.  This distribution in time and space suggests that the larger dog was introduced from the Plains, where large dogs are known from early on.  These large dogs apparently interbred with the small dogs, as the average size of the small dogs increased over time beginning at the time of the introduction of the large dogs.

Colton’s sample size was small and his conclusions tentative, but some interesting patterns emerge nevertheless.  I think the most interesting is the rather early introduction of small dogs to the Rio Grande Valley, which suggests substantial contact with the Plains as early as AD 800.  Considerable Plains-Pueblo contact is known for the later period, starting around AD 1325, when massive immigration into the Rio Grande area from other parts of the Pueblo world substantially shifted the geographic distribution of Pueblo peoples, but before that the Rio Grande Valley was sparsely populated and relatively little is known about it.  The introduction of the large dogs, however, suggests that Plains contacts were of longstanding importance in the region, and that the intensified contacts after 1325 may have been the result of the increased population rather than a change in the basic structure of regional relationships.  Other evidence for the early importance of contact between the eastern Pueblos and the Plains comes from the discovery that domesticated turkeys in the Southwest were genetically closer to subspecies of wild turkey found further east than to the local subspecies.

In light of this evidence, I’ll modify my response to ben to say that the dogs available to the Chacoans probably would not have been suitable for pulling loads, but that it is possible that the large Plains dog was already available in Chacoan times.  There’s no evidence that I know of, however, that dogs were used as beasts of burden in the Southwest at any point in prehistory.  I’ll look for more recent studies on the issue of Southwestern dogs to see if I can get more information.  I don’t know of any such studies, but there may well be some.

Colton, H. (1970). The Aboriginal Southwestern Indian Dog American Antiquity, 35 (2) DOI: 10.2307/278144

Read Full Post »