|
 |
| |
| |
 |
|
| |
| |
| |
| |
| |
|
|
|
|
| |
Dating Wupatki Pueblo: Tree Ring Evidence |
|
|
Introduction: Overall Dates of Human Presence in the Wupatki Area Wupatki Pueblo is blessed with a large number of tree-ring dated beams that give a good general picture of when the pueblo was built, occupied, and abandoned. However, please keep in mind that available tree-ring dates represent only a small fraction of the full time span of human activities in and around Wupatki Pueblo. Tree-ring dates come from roof beams that were cut and used by people during the period when people built pueblo-style dwellings (ca. AD 900 to 1275 in this area). Humans were using and occupying the Wupatki area long before the prehistoric pueblo period (ca. AD 550 to 1275). It has been documented that Wupatki National Monument was used by Paleoindian peoples some 11,000 years ago (Downum 1993; Stanford 1999), and there is scattered evidence of Archaic peoples (dating to ca. 10,000 to 4,000 years ago) in the Monument and its surrounding area (Anderson and McDonald 1990; Bradley and Neff 1994). This is not too surprising, because nearby springs probably made the immediate environment of Wupatki Pueblo attractive during a number of time periods by people having several different lifeways. Unfortunately, we have not yet found much direct evidence of a very early human presence in the immediate vicinity of Wupatki, probably because the intense pueblo occupation of the 12th and 13th centuries A.D. has obscured or destroyed any artifacts or features that were originally present. After the advent of agriculture ca. 3,500 years ago, human populations on the Colorado Plateau grew in size and expanded their territories. One of the places occupied during the early days of corn farming was the Wupatki area. An early agricultural period occupation near Wupatki Pueblo itself has been documented by Basketmaker II or Basketmaker III-age storage cists (ca. 1500 B.C.to A.D. 700) in the lowest excavated levels of the Amphitheater (Colton 1946:60). Continuing pre-eruption pueblo occupation (i.e., occupation prior to the eruption of the Sunset Crater volcano in about A.D. 1064) has been documented by the recent NPS survey of Wupatki National Monument (Anderson 1990; Downum and Sullivan 1990). No doubt the Wupatki-area springs provided an attraction for early farmers, just as they would have for earlier Paleoindian and Archaic peoples. Immediately post-eruptive occupations (i.e., post A.D. 1064), dating just prior to the earliest tree-ring dates from Wupatki Pueblo, have also been documented in various locations of the monument. One of these sites, the Heiser Springs Pithouses, is located several hundred meters southeast of Wupatki Pueblo. Tree-ring cutting dates in the A.D. 1090s establish this site as the earliest tree-ring dated occupation in the Monument, dating about 15 years before the earliest tree-ring cutting dates from Wupatki. By far the most intensive period of occupation is represented by the tree-ring dates from Wupatki Pueblo, which span the first decade of the 1100s through the latter part of the 1200s. These dates generally hold true for occupation of the greater Wupatki area, which witnessed a population boom in the early 1100s, and was abandoned by no later than the last quarter of the 1200s. Composition of the Wupatki Tree-Ring Sample The full range of Wupatki’s tree-ring dates allows us to reconstruct a general profile of the pueblo’s occupation span and to identify some specific periods of intensive construction activity. The overall range of tree-ring dates from Wupatki is presented in Tables below, in the form of stem-and-leaf plots showing how the dates are distributed by decade. Another web page presents detailed information on the attributes of the tree-ring specimens themselves. A total of 170 tree-ring dates now is available from Wupatki, consisting of 155 that previously have been reported (Robinson et al. 1975:92-95) and 15 new date determinations obtained during a 1997 tree-ring sampling project conducted by NPS archaeologist Thomas Windes (Dean 1997). The 170 tree-ring dates include 83 “cutting” dates (i.e., dates where the outside ring of the tree-ring specimen marks the last ring that was grown in the life of the tree).
Where Did Wupatki’s Roof Beams Originate? Many of the tree-ring dated beams found at Wupatki are exotic species that do not grow near the pueblo. Obviously, the builders of Wupatki, as did other ancient pueblo people in the region (Dean 1969:80-81), had some definite preferences regarding the type of trees they believed were proper to use for roofing their pueblo. Evidently they felt that high-elevation conifers were the most appropriate choice. They rejected use of one of the most commonly available local tree species, cottonwood, which grew in some abundance around local springs and along the banks of the Little Colorado River, only about 10 km (6.2 miles) to the east. (Cottonwood beams are present, but extremely rare at Wupatki.) Tree specimens from Wupatki come from a wide variety of tree species and represent a wide range of sizes and forms, including large primary beams, smaller poles, and split planks. The tree species represent wood types that should be considered “exotic” to Wupatki pueblo, in the sense that none of the dated beams came from species growing at or near the site today. Only six of the 170 dated specimens were Pinyon Pine, the species that today grows closest (probably within 3 to 5 km [1.9 to 3.2 miles]) of Wupatki. Nearly half (n=79, 46.4%) of the dated specimens were identified by the Laboratory of Tree-Ring Research at the University of Arizona as “Spruce/Fir” (not further differentiated by species, due to the difficulty in doing so, even with high-power magnification; see Betancourt et al. 1986). The closest places where these species can be found today are the northern slopes of O’Leary Peak and Darton Dome, about 20 km (12.4 miles) southwest of Wupatki Pueblo, but they are most abundant in the San Francisco Peaks, more than 30 km (18.7 miles) to the southwest of Wupatki. The next most common species was Ponderosa Pine (n=66, 38.8%), followed by Douglas Fir (n=19, 11.2%). Ponderosa Pine is available today about 10 km (6.2 miles) south of Wupatki, at the lower elevation limits of the Ponderosa Pine forest. Douglas Fir can be found only along the margins of the Bonito Lava flow, the northwestern slope of Sunset Crater, and the northern slopes of O’Leary Peak and Darton Dome, about 20 km (12.4 miles) southwest of Wupatki. All of these tree species also could also have been obtained as driftwood in the flood plain of the Little Colorado River, an issue that we will consider next.
Possible Effects of Exotic Tree Species on Interpreting Wupatki’s Dates It is important to consider the origins of exotic tree species issue in some detail, because it is possible that trees coming from a long distance away might have been cut (or might have died naturally) long before they were used to roof the pueblo. Two questions arise that are relevant to interpreting the dates obtained from Wupatki’s exotic roof beams: 1. Did the builders of the pueblo obtain their roof beams from driftwood in the Little Colorado River? (a problematic situation if true, because driftwood is made up of trees that died on various occasions and were transported down the river at various times during individual flood events), and 2. Would there have been a “lag effect” or hiatus (Dean 1978:227) caused by the time involved in cutting, stockpiling, and perhaps trading for trees obtained several miles from the pueblo? Regarding the first question, it is possible but it seems unlikely that most of the spruce, fir, ponderosa pine, and other conifer beams at Wupatki were obtained as driftwood in the Little Colorado. Most of the dated beams show cut marks from stone axes, the presence of bark, and other attributes that are not consistent with use of driftwood specimens, which often are stripped of bark and scarred and polished by water action. The scarcity of cottonwood in the pueblo also indicates that wood resources in general were not coming from the river flood plain. If Wupatki’s residents were habitually getting their roof beams from the river floodplain, it seems reasonable to expect that they would have gathered cottonwood beams along with the conifers, since both would be readily available there. In addition, some of Wupatki’s wooden roof beams showed unusual growth patterns in their rings, interpreted as indicating the effects of Sunset Crater ash fall (Smiley 1958). If these patterns indeed represent the effects of Sunset Crater ash, this again speaks against the idea that the beams were obtained as driftwood, because there is no plausible mechanism for alluvial transport into the Little Colorado River of pine or fir trees growing in the ash-fall area. (The Little Colorado River flows from southeast to northwest, and is entirely out of the main area of Sunset Crater ash fall – it is therefore impossible to imagine that high-elevation species growing in the ash fall area would have been uprooted by flooding and brought by tributary streams into the Little Colorado River.) The second question, involving a possible “lag effect” in the time between cutting of trees and their use in the pueblo, is more difficult to address. Ancient pueblo people are known to have cut and stockpiled beams in anticipation of future building events, sometimes several years in advance (Dean 1969:77). The fact that Wupatki’s roof beams apparently were cut in locations many miles removed from the Pueblo introduces the possibility that they were cut in remote locations, stockpiled there and allowed to season, and then transported back to the pueblo after they had dried (and thus had become considerably lighter and easier for people to carry – remember, the ancient Wupatkians had no beasts of burden). In addition, there is the interesting issue of whether or not the Wupatki roof beams were obtained by trade. During the time of Wupatki’s construction, in the 12th and 13th centuries A.D., the landscape over which Wupatki’s builders would have had to travel to obtain their roof beams was densely settled by people of several different cultural or ethnic groups. Ceramics and other material traits suggest that people living in these areas were affiliated with at least two cultural traditions (Sinagua and Cohonina) that may have been distinct from the majority of the general Wupatki area population (who seem to have been affiliated mostly with the Kayenta ancestral pueblo tradition). The Sinagua and Cohonina were organized into numerous small communities that may or may not have been entirely friendly with each other, let alone people from a more distant community like Wupatki. Thus, it can be concluded that the Wupatkians either had safe passage through these territories to obtain roof beams, or they obtained them through exchange. We know very little about the specific mechanisms by which prehistoric peoples might have obtained their wood resources, especially as concerns long-distance treks into “foreign” territory to obtain roof beams (e.g., Betancourt et al. 1986). Hence, it cannot be ruled out that some of Wupatki’s roof beams were obtained by trade with neighboring populations. If so, it seems possible that there would have been some lag in the time (perhaps a few years) between the cutting of trees and their final use as roof beams. At present there is no good way to estimate whether or not such potential time lags significantly skew our understanding of when Wupatki pueblo was built. The best we can do is examine the distribution of dates within individual rooms at Wupatki and see if we can detect any glaring disparities in the clusters of dates within a single room or adjacent rooms. Unfortunately, at Wupatki most rooms do not show strong clusters of dates that would clearly indicate coordinated building activity. Instead, when we look at tree-ring dates room by room, we see a range of dates that are often difficult to interpret with respect to possible stockpiling, construction, beam-reuse, room and roof remodeling, and roof repair. Many of the roofs at Wupatki appear to have been remodeled and repaired over a considerable span of time (in some cases, seven or eight decades). The relatively short possible lag time involved in stockpiling (usually no more than a year or two) or exchange (unknown, but probably also no more than a few years) is probably minimal by comparison. Interpreting Wupatki’s Tree-Ring Dates The overall set of tree-ring dates from Wupatki indicates that the earliest identifiable tree-cutting activity took place in A.D. 1106. This date, however, is represented by only a single tree-ring specimen. Because the date is an isolated occurrence, and no other tree-ring dates are documented for the next 25 years, it is possible that the 1106 date may have come from use of a dead tree or re-use of a beam from another location. The specimen that produced this date was a small pole of Ponderosa Pine, approximately 7.2 cm (2.8 inches) in diameter, so it easily could have been acquired by breaking off a small, dead pine tree. (Laboratory of Tree-Ring Research files did not indicate whether or not there was evidence that the beam had an axe-cut end, or was broken.) The specimen did not exhibit bark, as some other Wupatki samples do, so there was no positive evidence that this specimen came from a freshly-cut roof beam that was immediately brought into the pueblo. Thus, although this beam shows some possible construction activity at Wupatki in the first decade of the 1100s, a 25-year gap between the 1106 date and the next latest tree-ring cutting date (at A.D. 1131) suggests that any early period of construction was quite limited in scale. The next cutting dates come at A.D. 1131 (represented by two specimens), and they mark the first significant tree-ring dated construction events at Wupatki. As shown in Tables below, the 1130s are the decade during which the building activity at Wupatki begins in earnest. There is a strong cluster of dates at A.D. 1137, and another cluster at A.D. 1138. The number of tree-ring cutting dates decrease slightly during the A.D. 1140s, but clusters at A.D. 1145 and 1146 suggest important construction activities in the mid-decade. The total number of tree-ring dates, as well as cutting dates, decreases during the 1150s, but clusters of cutting dates at A.D. 1153 and 1155 suggest some low intensity construction activities during the mid-1150s. The A.D. 1160s have the most tree-ring dates and the most cutting dates of any decade at Wupatki, with a strong cluster of dates at A.D. 1160, and smaller clusters at 1161, 1167, and 1168. Tree-ring dates decrease in number in the 1170s and 1180s, with clusters evident at A.D. 1174 and 1183. Tree-ring dates then increase slightly from the 1180s to the 1190s, with the 1190s showing relatively weak clusters at A.D. 1190, 1192, and 1197. By the 1200s, construction activity seems nearly to have ceased at Wupatki pueblo. Only two cutting dates are documented for the first decade of the 1200s, and the next decade shows only four dates, three of them cutting dates. No tree-ring dates of any kind are documented for the A.D. 1220s, 1230s, and 1240s. New dates, obtained in 1997, indicate that people were cutting trees (presumably for the purposes of constructing and remodeling rooms) at least as late as the A.D. 1250s and 1260s. All of these late dates are non-cutting dates, and all come from unknown locations at the site, so it is known only that the cutting of trees continued at least as late as A.D. 1260. These dates, while not previously documented, are not too surprising, given that Wupatki Pueblo has produced relatively abundant ceramic artifacts (e.g., Kayenta Black-on-white, Kayenta Polychrome, Kiet Siel Gray, and St. Johns Polychrome) known to date to the last half of the A.D. 1200s (Breternitz 1966; Downum 1988). Some Jeddito Plain and Jeddito Black-on-yellow sherds (types believed to date after about A.D. 1300) also show a continuing use or low-level occupation during the 14th century. (It is unknown, however, whether any of the late non-cutting dates came from trees that were felled during this time. As non-cutting dates, it is known only that they were cut after the last dated ring on the specimen.) Distribution of Burned Roof Beams It is interesting to note the temporal and spatial distribution of burned roof beams at Wupatki, because these give important clues about the nature of Wupatki’s occupation and abandonment. Some ruins in the Southwest are known to have been destroyed or partially destroyed by fire, and a common pattern at many late prehistoric pueblos is that they were catastrophically destroyed by fire upon abandonment. In general it is difficult to say if this was done intentionally by the pueblo’s inhabitants, or if the fire was set by enemies. In some cases, for example the Point of Pines Pueblo (Haury , there is clear evidence that at Wupatki only seven out of 170 dated wooden beams were burned (specimens F-1613, 3030, and 3048, and WPT-229, 322, 331, and 345). Four of these beams produced cutting dates (two at A.D. 1137, one at A.D. 1145, and one at A.D. 1146), one gave a “near” cutting date (probably within only one or a few years of A.D. 1136), and the remaining two provided non-cutting dates (at A.D. 1090 and 1130). Several things are immediately noteworthy about this distribution. First, both the cutting and non-cutting dates are among the earliest in the entire sequence at Wupatki. Also, when provenience information was available, the dates from burned specimens occurred in stratigraphically early positions, i.e., below later rooms and surfaces. The areas producing burned specimens were also widely separated, occurring on both the west (at the base of Room 36) and east (floor of Room 45B and beneath Rooms 80 and 81) sides of the pueblo. These facts suggest that multiple, widely separated structures were burned at Wupatki (the fire thus was not a single conflagration that spread across adjacent rooms), and that the burning(s) took place relatively early in the construction sequence. It is not known whether the burned specimens are evidence of a one-time, intentional burning event where several rooms were set afire at once, or a series of unrelated, accidental events that were separated in time. Given the absence of subsequent burning events at Wupatki, however, multiple accidental fires seem unlikely. Pueblo rooms do not burn easily. Finally, the temporal and spatial distribution of burned wood specimens indicates clearly that the final abandonment of Wupatki was not accompanied by burning. Wupatki thus was not destroyed by fire, but was allowed to decay naturally and fill with windblown sediment, wall rock, collapsed roof beams, and other materials. Prior to excavation in the 1930s, several rooms (e.g., Rooms 35 and 41) still had intact roofs that remained in their original positions, a testament not only to how well the pueblo was built, but also to an orderly and gradual abandonment of the pueblo (as opposed to destruction by fire, a fate that is relatively common for many of the later, PIV-age [AD 1300 to 1540] pueblos in the northern Southwest). Summary Wooden roof beams at Wupatki Pueblo have produced 170 tree-ring dates. Eighty-three of the dates are “cutting” dates, or dates that record the true death date of the tree (presumed in most instances to have been caused by a human being cutting down the tree). The wood from Wupatki came predominantly from high-elevation conifer trees. These beams likely came from upland and mountain settings in the vicinity of O’Leary Peak, Sunset Crater, and the San Francisco Peaks. Some specimens might have been obtained as driftwood from the Little Colorado River, but Wupatki’s residents did not commonly go to the river floodplain to get their wood. More likely, they went on long-distance woodcutting expeditions or they obtained their roof beams in trade with neighboring cultural groups. Minor construction at the pueblo may have begun around A.D. 1106. The “boom” period of Wupatki’s construction commences in the A.D. 1130s and continues into the 1190s. Significant building stops by A.D. 1215. Three very late dates, probably representing trees cut down after A.D. 1260, mark the end of prehistoric building activity. This last set of dates comes from unknown locations within the pueblo, and cannot be easily interpreted except to note that the dates are accompanied by late prehistoric ceramics (dating at least as late as the tree-ring dates, and possibly later). The late cutting dates therefore may reflect a small-scale continuing occupation or a limited re-occupation of Wupatki that continued as late as the 14th century. Wupatki was not burned when the pueblo was abandoned. It was instead allowed to decay naturally, a rather unusual fate when compared to many other late prehistoric pueblos in the area (most of which were burned, either intentionally as part of the abandonment process or by enemies). |
|
|
|
||
